Multifunctional 45S5 Bioglass® -Based Composite Scaffolds for Bone Tissue Engineering

Multifunctional 45S5 Bioglass® -based composite scaffolds for bone tissue engineering

Phillipa Jane Newby

Department of Materials, Imperial College London (UK)

A thesis presented for the degree of Doctor of Philosophy (PhD)

Imperial College London

November 2013

Declaration

This thesis is a representation of my own original research work. Whenever contributions of others have been made, every effort has been made to indicate and acknowledge their input to the work, including reference to the relevant literature, and acknowledgement of collaborative research and discussions.

The copyright of this thesis rests with the author and is made available under a Creative Commons Attribution Non-Commercial No Derivatives licence. Researchers are free to copy, distribute or transmit the thesis on the condition that they attribute it, that they do not use it for commercial purposes and that they do not alter, transform or build upon it. For any reuse or redistribution, researchers must make clear to others the licence terms of this work.

Phillipa J Newby

Abstract

In the evolving field of bone tissue engineering, bioactive glass-ceramic scaffolds continue to draw significant attention as they can provide a 3D template that promotes new bone formation. These scaffolds have proved that they can loosely mimic cancellous bone structure as well as exhibiting appropriate mechanical properties and bioactivity to support the growth of new bone tissue.

This work concentrates on improving the functionality of 45S5 Bioglass® -derived scaffolds produced by the foam replication technique, by introducing silver and copper through molten salt ion exchange to provide an antibacterial function and through the use of gelatin coatings to improve compressional strength, whilst not impeding the original scaffold’s bioactivity.

The results confirmed that the introduced modifications have enhanced the scaffolds functionality as well as improving existing properties of the scaffolds such as mechanical competence. Overall the compressive strength of the scaffolds was improved with the addition of copper and silver ions and by the presence of gelatin coatings. After immersion in SBF, the presence of the additives did not impede the formation of hydroxyapatite on the scaffolds and hence the bioactivity. The characterization of the doped and coated samples showed that they have similar behavior to those of the base scaffold or have improved them marginally in terms of wettability and surface topography.

During cell culture studies, using HPLSC’s, rMSC’s and hMSC’s, the silver doped scaffolds caused a slight cytotoxic effect, whilst the copper and silver/copper samples caused a significant cytotoxic effect. Bacterial studies showed that the metal ion-doped materials produced an antibacterial effect against the E.coli, S.aureus and K.pneumoniae. The present study has contributed to the enhancement of the standard 45S5 Bioglass® -derived scaffolds and has provided a base for future in vitro and in vivo work involving these materials in bone regeneration strategies.

Acknowledgements

I would like to thank my project supervisors Professor Aldo R Boccaccini and Professor Eduardo Saiz Gutierrez for all of their advice and guidance on this project; without their continued support and help this project would not be possible. I also gratefully acknowledge the financial support of the Engineering and Physical Sciences Research Council (EPSRC).

Thanks also go to Dr X Chatzistavrou for many useful and enlightening conversations during the course of this project along with the rest of the research group (Deborah, Darma, Mikey, Bo, Fatamah, Tayyab, Vivianna, Lutz and Melek) as they provided many occasions to discus and thrash out issues we were having. I would like to extend thanks to department of materials at Imperial in general for support and advice during my time there, in particular Norma and Professor David McPhial.

Thanks go to Dr I Thompson, King’s College London for providing the 45S5 Bioglass powder that forms the back bone of this project. Thanks to Dr Reem El-Gendy and Dr X Yang for agreeing to carry out cell culture test on the basic 45S5 Bioglass® scaffolds and the silver doped scaffolds and to Leoni and Dr U Kneaser’s research group for collaborating with me on the second set of cell culture work, to Anette and Dr H Hȕber’s research group for the collaborating with me on the E.coli investigation and finally to Dr T Hammer for his collaboration on the rest of the bacterial work. All of the collaborators where helpful in explaining how the experiments should work and the techniques used to obtain the results.

Lastly thanks go to my family, Mum, Dad and Danielle and my partner and his family, Robert, Sarah and Andrew, and my friends (Nick, Tasha, Zeeba, Mark, Abi, Jen and Richard) as without their continuing support, love, laughter and their ability to keep me sane this project wouldn’t be possible especially during many bumps in the road.

Table of Contents

Chapter One ...... 30 1. Introduction ...... 30 Chapter Two...... 33 2. Literature review ...... 33 2.1 Introduction ...... 33 2.2 Tissue engineering: the general principles ...... 33 2.3 The need for tissue engineering ...... 34 2.4 Bone tissue engineering ...... 35 2.5 Challenges faced in bone tissue engineering ...... 41 2.6 Properties for an ideal bone tissue engineering material ...... 42 2.6.1 Biological properties required ...... 42 2.6.2 Structural properties required for an ideal bone tissue engineering material ...... 44 2.7 The progression of materials in bone tissue engineering ...... 44 2.7.1 Calcium phosphate and its derivatives ...... 45 2.7.1.1 Hydroxyapatite ...... 45 2.7.1.2 Amorphous calcium phosphates ...... 46 2.7.1.3 Tricalcium phosphates ...... 46 2.7.1.4 Biphasic calcium phosphates ...... 46 2.7.2 Glass-ceramics in bone tissue engineering ...... 47 2.7.2.1 Bioactive glasses ...... 47 2.7.3 Other materials used in bone tissue engineering ...... 49 2.8 Bioactive glass processing ...... 50 2.8.1 Producing bioactive glasses by melt quenching ...... 50 2.8.2 Producing bioactive glasses by sol-gel processing ...... 51 2.9 The need for scaffolds in bone tissue engineering and additional properties ...... 51 2.9.1 Additional properties for a scaffold - Porosity ...... 52 2.9.2 Additional properties for a scaffold – Wettability ...... 53 2.9.3 Additional properties for a scaffold – Surface topography ...... 54 2.10 Current scaffolds in bone tissue engineering ...... 55 2.10.1 Glass-ceramic scaffolds for bone tissue engineering ...... 55 2.10.2 45S5 Bioglass®- derived glass-ceramic scaffolds for bone tissue engineering ...... 56 2.10.3 Other bone tissue engineering scaffolds ...... 56 2.11 Scaffold production ...... 57

2.11.1 Foam Replication technique to produce scaffolds for bone tissue engineering ...... 58 2.11.2 Sol-gel processing to produce scaffolds for bone tissue engineering ...... 58 2.11.3 Freeze drying to produce scaffolds for bone tissue engineering ...... 59 2.11.4 Solid free-form fabricated scaffolds for bone tissue engineering ...... 60 2.12 Multifunctional scaffolds for bone tissue engineering ...... 61 2.13 Antibacterial scaffolds ...... 62 2.13.1 Antibacterial properties of 45S5 Bioglass® -derived glass-ceramic scaffolds ...... 63 2.13.2 Antibacterial properties of silver ...... 64 2.13.3 Copper ions as an antibacterial agent ...... 64 2.14 Angiogenic scaffolds ...... 65 2.14.1 Angiogenic properties of 45S5 Bioglass® ...... 66 2.14.2 Copper ions as an angiogenic agent ...... 66 2.15 Molten salt ion exchange (MSIE) ...... 67 2.15.1 Incorporating Silver ions into 45S5 Bioglass® -derived glass-ceramic scaffolds using molten salt ion exchange ...... 69 2.15.2 Incorporating Copper ions into 45S5 Bioglass® -derived glass-ceramic scaffolds using molten salt ion exchange ...... 70 2.15.3 Dual ion incorporation into 45S5 Bioglass® -derived glass-ceramic scaffolds using molten salt ion exchange ...... 70 2.16 Polymer coated scaffolds for bone tissue engineering ...... 71 2.16.1 Poly (D, L-lactic acid) coated scaffolds ...... 73 2.16.2 Gelatin as a potential coating agent ...... 74 2.17 In vitro cell culture studies to test the behaviour of bone tissue engineering scaffolds ...... 76 2.18 Bacterial testing on bone tissue engineering scaffolds ...... 77 2.19 Summary of the literature review ...... 78 Chapter Three...... 80 3. Aims and objectives of the project ...... 80 3.1 Aims ...... 80 3.2 Objectives ...... 80 Chapter Four ...... 83 4. Materials, Methodology and Characterisation techniques ...... 83 4.1 Materials ...... 83 4.1.1 45S5 Bioglass® ...... 83 4.1.2 Foam replication technique materials ...... 83 4.1.3 Molten salt ion exchange techniques ...... 84 4.1.4 Biopolymers for scaffold coating ...... 84

4.1.5 Other materials used...... 84 4.1.6 Materials for the human periodontal ligamental stromal cell study ...... 85 4.1.7 Materials for the rat bone-marrow stromal cell and human bone-marrow stromal cell studies ...... 85 4.1.8 Bacterial study materials ...... 86 4.2 Methodology ...... 86 4.2.1 Foam replication technique ...... 87 4.2.2 Molten salt ion exchange processing ...... 88 I Silver molten salt ion exchange process ...... 88 II Copper molten salt ion exchange process ...... 91 III Silver and copper combined molten salt ion exchange process ...... 92 4.2.3 Biopolymer coating ...... 92 I Poly (D, L-lactic acid) coating ...... 92 II Gelatin coating ...... 93 4.2.4 Combination scaffolds ...... 95 4.3 Structural Characterisation techniques ...... 96 4.3.1 Porosity studies ...... 96 4.3.2 Simulated body fluid studies ...... 97 4.3.3 Scanning electron microscopy ...... 98 4.3.4 Energy dispersive X-ray spectroscopy ...... 98 4.3.5 X-ray diffraction analysis ...... 98 4.3.6 Fourier transform infra-red spectroscopy...... 99 4.3.7 White light interferometery ...... 99 4.3.8 Wettability ...... 101 4.3.9 Inductively coupled plasma mass spectrometry ...... 102 4.3.10 Mechanical properties ...... 102 4.3.11 EDX depth studies ...... 103 4.4 Biological Characterisation techniques ...... 104 4.4.1 Cell preparation and extraction for human periodontal ligamental stromal cells ...... 104 4.4.2 In vitro cell seeding using human periodontal ligamental stromal cells ...... 105 4.4.3 Biological characterisation using human periodontal ligamental stromal cells ...... 106 4.4.4 Cell preparation and extraction for rat bone-marrow stromal cells and human bone-marrow stromal cells ...... 106 4.4.5 2D cell cultures using rat bone-marrow stromal cells and human bone-marrow stromal cells ...... 107 4.4.6 3D cell cultures using rat bone-marrow stromal cells ...... 110

4.4.7 Biological characterisation using rat bone-marrow stromal cells and human bone-marrow stromal cells ...... 111 4.4.8 Bacterial preparation, counting and growth curves ...... 113 4.4.9 Minimum inhibitory concentration testing using metal salts against Escherichia Coli ..... 116 4.4.10 Testing the bacteriostatic or bactericidal capabilities of metal salts ...... 118 4.4.11 Growth inhibition using supernatants against Escherichia coli ...... 118 4.4.12 Agar diffusion to show growth inhibition of Staphylococcus aureus and Escherichia coli ...... 119 I Growth inhibition demonstrated through agar diffusion on Escherichia coli ...... 119 II Growth inhibition demonstrated through agar diffusion on Staphylococcus aureus ...... 120 4.4.13 Growth inhibition on 3D scaffolds using Klebsiella pneumoniae ...... 120 4.5 Statistical Analysis ...... 121 Chapter Five ...... 122 5. Structural results and discussion ...... 122 5.1 General remarks ...... 122 5.2 45S5 Bioglass®- derived glass-ceramic scaffolds ...... 122 5.3 Silver molten salt ion exchanged scaffolds ...... 134 5.3.1 Preliminary processing of silver molten salt ion exchange scaffolds ...... 134 5.3.2 Structural characterisation of silver molten salt ion exchange scaffolds ...... 137 5.4 Copper molten salt ion exchanged scaffolds ...... 151 5.5 Combination molten salt ion exchanged scaffolds...... 162 5.6 Polymer coated 45S5 Bioglass®- derived glass-ceramic scaffolds ...... 174 5.6.1 Preliminary processing of gelatin coated 45S5 Bioglass®- derived glass-ceramic scaffolds ...... 174 5.6.2 Structural characterisation of polymer coated 45S5 Bioglass®- derived glass-ceramic scaffolds ...... 179 5.7 Polymer coated molten salt ion exchanged 45S5 Bioglass®- derived glass-ceramic scaffolds 191 5.7.1 Structural characterisation of polymer coated copper molten salt ion exchanged scaffolds ...... 191 5.7.2 Structural characterisation of polymer coated combination molten salt ion exchange scaffolds ...... 200 5.8 Summary of results from the structural characterisation ...... 210 Chapter Six...... 214 6. Biological Results and Discussion ...... 214 6.1 General remarks ...... 214 6.2 In vitro cell biology using human periodontal ligamental stromal cells ...... 214 6.2.1 Cell viability assessment using live/dead markers ...... 215

6.3 In vitro cell biology using rat bone-marrow stromal cells and human bone-marrow stromal cells ...... 218 6.3.1 Preliminary toxicity and biocompatibility study of low concentration molten salt ion exchange scaffolds using rat bone-marrow stromal cells ...... 219 6.3.2 The effect of pH change ...... 222 6.3.3 Biocompatibility and toxicity study using low concentration molten salt ion exchange scaffolds ...... 225 6.3.4 Biocompatibility and toxicity study using extremely low concentration molten salt ion exchange scaffolds ...... 227 6.4 Antibacterial assessment of molten salt ion exchange scaffolds ...... 230 6.4.1 Growth curves of Escherichia coli ...... 231 6.4.2 Copper nitrate vs copper chloride salts in a growth inhibition study using Escherichia coli ...... 234 6.4.3 Minimum inhibitory concentration testing using metal salts against Escherichia coli ...... 235 6.4.4 Metal salts growth inhibition study using Escherichia coli ...... 238 6.4.5 Growth inhibition capabilities of molten salt ion exchange scaffolds using Escherichia coli ...... 244 6.4.6 Agar diffusion studies using Escherichia coli ...... 247 6.4.7 Agar diffusion studies using Staphylococcus aureus ...... 250 6.4.8 Growth inhibition studies using Klebsiella pneumoniae ...... 252 6.5 Summary of results from the cell biology and bacterial studies ...... 253 6.5.1 Summary of cell biology work ...... 253 6.5.2 Summary of bacterial studies ...... 254 Chapter Seven ...... 257 7. Concluding remarks ...... 257 7.1 General observations on 45S5 Bioglass®- derived glass-ceramic scaffolds ...... 257 7.2 Molten salt ion exchange scaffolds ...... 259 7.3 Polymer coated scaffolds ...... 261 7.4 Composite scaffolds ...... 262 7.5 Summary ...... 263 Chapter Eight ...... 265 8. Future work ...... 265 8.1 Expanding the use of the foam replication technique ...... 265 8.2 The use of molten salt ion exchange to introduce multifunctional ions into scaffolds ...... 266 8.2.1 Other candidates for molten salt ion exchange processing ...... 267 8.3 The use of polymers in combination with bioactive glass scaffolds ...... 268 8.4 Gelatin as a drug delivery vehicle ...... 269

8.4.1 Introducing antibiotics and growth factors into gelatin coatings ...... 270 8.4.2 Therapeutic metal nano-particles – Copper oxide and Iron oxide ...... 271 8.5 Combining molten salt ion exchange scaffolds with polymer coatings ...... 272 8.6 Issues brought forward by this work ...... 273 8.6.1 pH issue with the base scaffolds ...... 273 8.6.2 Sterilisation ...... 273 8.7 Further structural characterisation of the scaffolds ...... 274 8.7.1 Molecular modelling using the Reverse Monte Carlo method ...... 274 8.8 Further in vitro characterisation ...... 275 8.8.1 Cell culture characterisation ...... 276 8.8.2 Bacterial characterisation ...... 277 8.9 In vivo characterisation ...... 277 8.9.1 In vivo studies using 45S5 Bioglass® –derived scaffolds using human dental pulp stromal cells ...... 278 8.9.2 In vivo study using 45S5 Bioglass® –derived scaffolds in an arteriovenous loop model... 279 References ...... 282 Conference papers and publications ...... 324 Papers ...... 324 Conferences...... 325 Book chapters...... 325 Appendices ...... 326 Appendix A – Background theory ...... 326 A.1 Multifunctional scaffolds ...... 326 Appendix B – Experimental methods ...... 332 B.1 Pellet production ...... 332 B.2 Preparation of samples for EDX depth studies ...... 333 B.3 Extraction of tooth pulp to obtain human periodontal ligamental stromal cells ...... 334 B.4 Extraction of rat bone-marrow stromal cells ...... 335 B.5 Extraction of human bone-marrow stromal cells ...... 336 B.6 Standard I Agar plate production ...... 337 B.7 Bacteria counting ...... 338 Appendix C – Results ...... 341 C.1 Silver molten salt ion exchange scaffolds ...... 341 C.2 Bacterial assessment of molten salt ion exchange scaffolds ...... 345 C.3 Polymer coated 45S5 Bioglass® -derived glass-ceramic scaffolds...... 351

C.4 Polymer coated molten salt ion exchanged 45S5 Bioglass® -derived glass-ceramic scaffolds ...... 353 Appendix D – Future Work ...... 355 D.1 Reverse Monte Carlo modelling ...... 355 D.2 The Arteriovenous loop...... 360 D.3 Therapeutic metal nano-particles – Copper oxide and Iron oxide ...... 361 Appendix E – Statistical Analysis ...... 363 Appendix F – Copyright permissions ...... 365

List of Figures

Figure 1:- Overview of the development of multifunctional BBG scaffolds within the frame-work of this project ...... 31 Figure 2:-Transplant waiting lists and the number of operations carried out over a ten year period for the UK (24) ...... 35 Figure 3:- Schematic diagram showing the functions of bone in the human body (65) ...... 36 Figure 4:-Schematic diagram of the cross section of an adult human bone showing the structure in detail (78) (75) ...... 37 Figure 5:- Bone remodelling cycle under normal conditions (65) (81) ...... 39 Figure 6:- Types of bone implants currently available as treatment options for significant bone trauma (84) ...... 39 Figure 7:- Ternary diagram of SiO2 – Na2O - CaO which shows where the optimal compositions for bioactive behaviour lie (3) (185) (186) (187) (122) ...... 49 Figure 8:- Diagram showing the different functions that are directly related to the pore structure a scaffold. (209) ...... 53 Figure 9:-Illustration of the contact angle used to quantify the wettability of a sample along with the relationship of the wettability and the hydrophilic or hydrophobic nature of a sample (212) (213) (215) ...... 54 Figure 10:-Evaluation of compressive strength against porosity for scaffolds fabricated using A) Freeze casting, B) Foam replication, C) Sol-gel and D) solid free-form fabrication with data for glass- ceramics only, scaffolds coated in a polymer and MSIE samples. Data is shown in appendix A ...... 59 Figure 11:- Comparison of elastic modulus against compressive strength for a variety of biomaterials including composites, polymers, bone and bio-ceramics (121) ...... 62 Figure 12:- Schematic representation of the chemical strengthening of glass through molten salt ion exchange showing the exchange process between potassium in the melt and sodium in the glass ...... 68 Figure 13:- Fracture toughening mechnasiums observed for polymer coated ceramics in the literature showing A) crack bridging, B) crack deflection, C) crack tip shielding and D) pull-out. (345) ...... 72 Figure 14:- Structure of PDLLA (350) ...... 73 Figure 15:- Structure of gelatin (357) (358) ...... 75 Figure 16:- Overview of tasks and experiments carried out in the framework of the present project .. 82 Figure 17:- Sintering profile for the fabrication of BBG scaffolds by the foam replication technique described in Chen at al. (4) ...... 88 Figure 18:- Flow diagram of the MSIE process applied to incorporate silver into Bioglass® - based scaffolds ...... 89 Figure 19:- Flow diagram of the optimised MSIE process for the incorporation of therapeutic ions into Bioglass® -based scaffolds ...... 91 Figure 20:- Flow diagram of the coating process used to coat the BBG scaffolds with PDLLA ...... 93 Figure 21:- Flow diagram representing the process used to coat BBG scaffold samples with gelatin in this work ...... 94 Figure 22:- Definitions of RMS, Ra and PV when applied to the roughness of a sample (401) ...... 100 Figure 23:-Schematic diagram illustrating the fixing, polishing and preparation of samples ready for the EDX depth analysis ...... 103 Figure 24:- Preparation of the 2D culture, showing the placement of the well insert and the sample scaffold ...... 108 Figure 25:- Flow chart showing the initial protocol used for the 2D culture study ...... 109 Figure 26:- Flow chart showing the optimised protocol used for the 2D cell culture study ...... 110

Figure 27:- Flow chart for the 3D cell culture study ...... 111 Figure 28:- Colony forming units (CFU's) of E.coli showing the process of diluting the E.coli suspension, plating it and then counting the CFU’s after incubation...... 115 Figure 29:- SEM images of BBG scaffolds produced using the foam replication technique at magnifications of A) x50, B) x300, C) x1400 and D) x10000 ...... 123 Figure 30:-White light interferometry results for a sintered BBG pellet showing A) a 2D white light microscopy intensity plot, B) a 2D solid plot coloured version of the white light microscopy intensity plot shown in A), C) a 3D oblique plot showing the surface roughness of the sample in terms of colour, D) a sample cross section across the pellet showing the surface profile and E) a 2D coloured surface map of a BBG pellet showing the peaks and valleys in terms of different colours ...... 125 Figure 31:- FTIR analysis of sintered BBG scaffold including the associated characteristic bond peaks (413) (414) (415) ...... 126 Figure 32:- XRD spectrum of sintered BBG scaffold (at 1100oC for two hours) confirming the crystalline nature of the material (4) (5) (389) (416)...... 127 Figure 33:- EDX analysis of a BBG scaffold to determine the elemental composition of the sample127 Figure 34:- SEM images of BBG scaffolds after immersion in SBF for A) one day (magnification x600), B) three days (magnification of x1500), C) seven days (magnification of x1500), D) 14 days (magnification of x1500), E) 21 days (magnification of x1500) and F) 28 days (magnification of x170) (arrows indicating HA) ...... 129 Figure 35:- EDX analysis of a BBG scaffolds which have been immersed in SBF for A) 1 day, B) 3 days, C) 7 days, D) 14 days, E) 21 days and F) 28 days showing high levels of Ca and P which confirms the formation of HA ...... 131 Figure 36:-FTIR analysis of BBG scaffolds immersed in SBF for 1, 3, 7, 14, 21 and 28 days ...... 132 Figure 37:- XRD analysis of BBG scaffolds which have been immersed in SBF for 1, 3, 7, 14, 21 and 28 days ...... 133 Figure 38:- pH changes in supernatants due to the presence of BBG scaffolds with A) pH changes in SBF and B) pH changes in cell culture media (DMEM/F-12). * Indicates a statistically significant change between the SBF and CCM (p≤0.05) using ANOVA analysis...... 134 Figure 39:- EDX spectra for silver doped BBG scaffolds with concentrations of silver nitrate in the salt bath; A) low concentration of silver nitrate, B) medium concentration of silver nitrate and C) high concentration of silver nitrate as shown in section 4.2.2.I ...... 135 Figure 40:- XRD analysis of silver doped scaffolds obtained from different silver concentrations in the molten salt bath showing the peaks indicative of sintered/crystallised 45S5 Bioglass®

(Na2Ca2Si3O9), silver and silver phosphate ...... 136 Figure 41:- BBG scaffolds immersed in a low silver concentration molten salt bath for A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes showing the presence of non-desired salt residue from the salt bath (arrows indicating the salt residue) ...... 138 Figure 42:- Porosity of sintered BBG scaffolds before being doped with silver through MSIE, after MSIE but before being subjected to the washing protocol and after MSIE and after being subjected to the washing protocol * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 138 Figure 43:- BBG scaffolds immersed in a low silver concentration molten salt bath for A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes ...... 139 Figure 44:- XRD characterisation of BBG scaffold after introducing silver through MSIE. Spectra for the three concentrations with an immersion period of 15 minutes are shown ...... 140 Figure 45:- FTIR spectra showing medium concentration of silver MSIE scaffolds with immersion periods of 15, 30, 45 and 60 minutes ...... 141

Figure 46:- EDX analysis of silver doped BBG scaffolds showing low concentrations samples with an immersion periods of A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes, medium concentration samples immersed in the salt bath for E) 15 minutes, F) 30 minutes, G) 45 minutes and H) 60 minutes and high concentration samples immersed in the salt bath for I) 15 minutes, J) 30 minutes, K) 45 minutes and L) 60 minutes ...... 142 Figure 47:- Depth EDX analysis of silver doped BBG pellets covering all concentrations of silver nitrate in the molten salt bath and all immersion periods investigated. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 142 Figure 48:- BBG scaffolds immersed in a medium concentration molten salt bath for A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes and then immersed in SBF for 14 days. Images have a magnification of x1500 (arrows indicating HA) ...... 145 Figure 49:- XRD characterisation of BBG after introducing silver through MSIE. Spectra for the three concentrations investigated with an immersion period of 15 minutes after immersion in SBF for 14 days are shown ...... 146 Figure 50:- FTIR spectra showing low concentration silver MSIE scaffolds with immersion periods of 15, 30, 45 and 60 minutes after immersion in SBF for 14 days ...... 146 Figure 51:-EDX analysis showing the elemental composition of the silver doped samples over the three concentrations with an immersion period of 60 minutes with A) low concentration of silver in the salt bath, B) medium concentration in the salt bath and C) high concentration of silver in the salt bath ...... 147 Figure 52:- Depth EDX analysis of silver doped BBG pellets covering all concentrations of silver nitrate in the salt bath and all immersion periods in the salt bath after 14 days in SBF. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 148 Figure 53:- Wettability data for silver doped BBG scaffolds showing results for pre and post immersion in SBF for 14 days (Data shown are mean values and standard deviation of at least 3 measurements). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 150 Figure 54:-ICP-MS data for silver MSIE scaffolds for all concentrations of silver in the molten salt bath and over all immersion periods. * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis...... 150 Figure 55:- BBG scaffolds immersed a variety of different concentrations of copper nitrate molten salt bath for 15 minutes with A) low concentration salt bath (magnification of x150), B) low concentration salt bath (magnification of x1000), C) medium concentration salt bath (magnification of x150) and D) high concentration salt bath (magnification of x150) ...... 152 Figure 56:- Offset XRD spectra of the three concentrations of copper doped BBG scaffolds ...... 152 Figure 57:- FTIR spectra of copper doped BBG scaffolds of all concentrations after immersion in the molten salt bath for 15 minutes ...... 153 Figure 58:- EDX analysis of copper doped BBG scaffolds for all three concentrations of copper in the salt melts with A) low concentration, B) medium concentration and C) high concentration ...... 154 Figure 59:- Depth EDX analysis of copper doped BBG pellets. * Indicates a significant change between the change in concentration (p≤0.05) using ANOVA analysis...... 154 Figure 60:- BBG scaffolds immersed in a variety of different concentrations of copper nitrate molten salt bath for 15 minutes with A) low concentration salt bath, B) medium concentration salt bath and C) high concentration salt bath after immersion in SBF for 14days. All images have a magnification of x300 ...... 156 Figure 61:- Offset XRD spectra of the three concentrations of copper doped BBG scaffolds after immersion in SBF for 14 days ...... 157

Figure 62:- FTIR spectra for all concentrations of copper doped scaffolds after 15 minutes in the molten salt bath and after immersion in SBF for 14 days ...... 158 Figure 63:- EDX analysis of copper doped BBG scaffolds for all three concentrations of copper in the salt melts with A) low concentration, B) medium concentration and C) high concentration after immersion in SBF for 14 days ...... 159 Figure 64:- Depth EDX analysis of copper doped BBG pellets after immersion in SBF for 14days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 160 Figure 65:- Wettability data for copper doped BBG scaffolds showing results for samples before and after immersion in SBF for 14 days (Data shown are mean values and standard deviation of at least 3 measurements). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 161 Figure 66:- ICP-MS analysis of copper ion release from copper doped BBG scaffolds immersed in PBS for the three different concentrations of copper nitrate salt used in the MSIE study. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 162 Figure 67:- SEM images of BBG scaffolds immersed in the combination silver nitrate, copper nitrate and sodium nitrate salt bath for A) 15 minutes (magnification of x400), B) 30 minutes (magnification of x400), C) 45 minutes (magnification of x400) and D) 60 minutes (magnification of x150) ...... 163 Figure 68:- XRD patterns of BBG after MSIE introducing both silver and copper for immersion periods of 15, 30, 45 and 60 minutes ...... 164 Figure 69:- FTIR spectra for silver/copper doped BBG scaffolds obtained by MSIE with different salt bath immersion periods ...... 165 Figure 70:- EDX analysis of the silver/copper doped BBG scaffolds with immersion periods in the salt bath of A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes ...... 165 Figure 71:- Depth EDX analysis of silver/copper doped BBG pellets after immersion in salt bath for 15, 30, 45 and 60 minutes. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 166 Figure 72:- BBG scaffolds immersed in the combination silver nitrate, copper nitrate and sodium nitrate salt bath for A) 15 minutes (magnification of x1000), B) 30 minutes (magnification of x1000), C) 45 minutes (magnification of x1000) and D) 60 minutes (magnification of x1500) after immersion in SBF for 14days ...... 168 Figure 73:- XRD analysis of BBG scaffolds after MSIE processing introducing silver and copper for immersion periods of 15, 30, 45 and 60 minutes after immersion in SBF for 14 days ...... 169 Figure 74:- FTIR spectra for silver/copper doped BBG scaffolds with different salt bath immersion periods after immersion in SBF for 14 days ...... 170 Figure 75:- EDX analysis of the silver/copper doped BBG scaffolds with immersion periods in the salt bath of A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes after immersion in SBF for 14 days ...... 170 Figure 76:- Depth EDX analysis for silver/copper doped BBG pellets after immersion in the salt bath for 15, 30, 45 and 60 minutes followed by immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 171 Figure 77:- Wettability data for the combination ion doped BBG scaffolds showing spectra for samples before and after immersion in SBF for 14 days (Data shown are mean values and standard deviation of at least 3 measurements). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 173 Figure 78:- ICP-MS data for the silver/copper doped samples obtained by different immersion periods in the salt bath, A) showing the release of silver from the samples and B) showing the release of copper from the samples. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 174

Figure 79:- BBG scaffolds coated with gelatin with A) a single coating of 5wt% type A gelatin (magnification of x50), B) a single coating of 6wt% type A gelatin (magnification of x850), C) a single coating of 5wt% type B gelatin (magnification of x550) and D) a single coating of 6wt% type B gelatin (magnification of x2500) with a coating immersion time of one minute (arrows indicate the gelatin coating)...... 176 Figure 80:- BBG scaffolds coated with gelatin type A with A) two coats of 5wt% type A gelatin (magnification of x350), B) three coats of 5wt% type A gelatin (magnification of x350), C) two coats of 6wt% type A gelatin (magnification of x450) and D) three coats of 6wt% type A gelatin (magnification of x450) with a coating immersion time of one minute for each coat (arrows indicate the gelatin coating) ...... 177 Figure 81:- BBG scaffolds coated with gelatin type B with A) two coats of 5wt% type B gelatin (magnification of x700), B) three coats of 5wt% type B gelatin (magnification of x400), C) two coats of 6wt% type B gelatin (magnification of x900) and D) three coats of 6wt% type B gelatin (magnification of x350) with a coating immersion time of one minute (arrows indicate the gelatin coating) ...... 178 Figure 82:- BBG scaffolds coated with PDLLA with A) a 5wt% coating (magnification of x50), B) a 5wt% coating (magnification of x430), C) a 5wt% coating (magnification of x250) and D) a 5wt% coating (magnification of x1000) (arrows indicate the PDLLA coating) ...... 180 Figure 83:- XRD spectra for gelatin and PDLLA coated BBG scaffolds ...... 180 Figure 84:- FTIR spectra for PDLLA coated BBG scaffolds before and after immersion in SBF for 14 days with A) showing both spectra indicating the peaks corresponding to PDLLA and B) depicting the distinctive peaks characteristic of BBG scaffolds ...... 181 Figure 85:- FTIR spectra for gelatin coated BBG scaffolds ...... 182 Figure 86:- EDX analysis of polymer coated BBG scaffolds including A) Type A 5wt% gelatin coated scaffold, B) Type A 6wt% gelatin coated scaffold, C) Type B 5wt% gelatin coated scaffold, D) Type B 6wt% gelatin coated scaffold and E) PDLLA coated scaffold...... 182 Figure 87:- BBG scaffolds coated with gelatin type A with A) a 5wt% coating (magnification of x750), B) a 5wt% coating (magnification of x900), C) a 6wt% coating (magnification of x1900) and D) a 6wt% coating (magnification of x900) after immersion in SBF for 14 days (Black arrows indicating the gelatin and white arrows indicating the HA) ...... 185 Figure 88:- BBG scaffolds coated with gelatin type B with A) a 5wt% coating (magnification of x550), B) a 5wt% coating (magnification of x100), C) a 6wt% coating (magnification of x600) and D) a 6wt% coating (magnification of x450) after immersion in SBF for 14 days (Black arrows indicating the gelatin and white arrows indicating the HA) ...... 185 Figure 89:- BBG scaffolds coated with PDLLA with A) a 5wt% coating (magnification of x50) and B) a 5wt% coating (magnification of x1700) after immersion in SBF for 14 days (Black arrows indicating the PDLLA and white arrows indicating the HA) ...... 186 Figure 90:- XRD spectra for gelatin coated BBG scaffolds and PDLLA coated 45S5 Bioglass® - derived scaffolds after immersion in SBF for 14 days...... 187 Figure 91:- FTIR spectra for gelatin coated BBG samples after immersion in SBF for 14days showing A) the full spectra and B) the area of the spectra showing the peaks indicative of the BBG ...... 188 Figure 92:- EDX analysis of polymer coated BBG scaffolds including A) Type A 5wt% gelatin coated scaffold, B) Type A 6wt% gelatin coated scaffold, C) Type B 5wt% gelatin coated scaffold, D) Type B 6wt% gelatin coated scaffold and E) PDLLA coated scaffold after immersion in SBF for 14 days ...... 189 Figure 93:- Wettability data for the gelatin coated BBG samples showing spectra from before and after immersion in SBF for 14 days (Data shown are mean values and standard deviation of at least 3

16 measurements). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 190 Figure 94:- 45S5 Bioglass® –derived glass-ceramic scaffolds doped with copper and then coated with gelatin with A) a 5wt% coating of type A gelatin (magnification of x550), B) a 6wt% coating of type A gelatin (magnification of x250), C) a 5wt% coating of type B gelatin (magnification of x700) and D) a 6wt% coating of type B gelatin (magnification of x900) (arrows indicate the gelatin coating) . 192 Figure 95:- XRD analysis of gelatin coated copper-doped 45S5 Bioglass® -derived scaffolds ...... 193 Figure 96:- FTIR analysis of gelatin coated copper doped BBG scaffolds showing A) the full spectrum and B) the part of the spectrum exhibiting the peaks indicative of the copper-doped scaffold ...... 194 Figure 97:- EDX analysis of polymer coated copper-doped BBG scaffolds including A) Type A 5wt% gelatin coated scaffold, B) Type A 6wt% gelatin coated scaffold, C) Type B 5wt% gelatin coated scaffold and D) Type B 6wt% gelatin coated scaffold ...... 194 Figure 98:- BBG scaffolds doped with copper and then coated with gelatin with A) 5wt% coating of type A gelatin (magnification of x300), B) 6wt% coating of type A gelatin (magnification of x500), C) 5wt% coating of type B gelatin (magnification of x170) and D) 6wt% coating of type B gelatin (magnification of x250) after immersion in SBF for 14days (Black arrows indicating the gelatin and white arrows indicating the HA) ...... 196 Figure 99:- XRD analysis of gelatin coated and copper-doped BBG scaffolds after immersion in SBF for 14 days...... 197 Figure 100:- FTIR analysis of gelatin coated copper doped BBG scaffolds after immersion in SBF for 14 days showing A) the full spectrum and B) the part of the spectrum that presents the peaks indicative of the doped scaffold ...... 198 Figure 101:- EDX analysis of polymer coated copper-doped BBG scaffolds including A) Type A 5wt% gelatin coated scaffold, B) Type A 6wt% gelatin coated scaffold, C) Type B 5wt% gelatin coated scaffold and D) Type B 6wt% gelatin coated scaffold after immersion in SBF for 14 days ... 198 Figure 102:- Wettability data for the gelatin coated copper doped BBG scaffolds before and after immersion in SBF for 14 days (Data shown are mean values and standard deviation of at least 3 measurements). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 200 Figure 103:- BBG scaffolds doped with copper and silver and then coated with gelatin with A) a 5wt% coating of type A gelatin (magnification of x750), B) a 6wt% coating of type A gelatin (magnification of x900), C) a 5wt% coating of type B gelatin (magnification of x700) and D) a 6wt% coating of type B gelatin (magnification of x250) (arrows indicating the gelatin coating) ...... 201 Figure 104:- XRD analysis of gelatin coated silver/copper-doped BBG scaffolds ...... 202 Figure 105:- FTIR analysis of gelatin coated silver/copper doped BBG scaffolds showing A) the full spectrum and B) the part of the spectrum which shows the typical peaks indicative of the doped scaffold ...... 203 Figure 106:- EDX analysis of polymer coated silver/copper-doped BBG scaffolds including A) Type A 5wt% gelatin coated scaffold, B) Type A 6wt% gelatin coated scaffold, C) Type B 5wt% gelatin coated scaffold and D) Type B 6wt% gelatin coated scaffold ...... 203 Figure 107:- BBG scaffolds doped with copper and silver and then coated with gelatin with A) a 5wt% coating of type A gelatin (magnification of x150), B) a 6wt% coating of type A gelatin (magnification of x2000), C) a 5wt% coating of type B gelatin (magnification of x160) and D) a 6wt% coating of type B gelatin (magnification of x600) after immersion in SBF for 14days ...... 205 Figure 108:- XRD analysis of gelatin coated silver/copper-doped BBG scaffolds after immersion in SBF for 14 days ...... 206

Figure 109:- FTIR analysis of gelatin coated silver/copper doped BBG scaffolds after 14 days immersed in SBF with A) showing the full spectrum and B) showing the part of the spectrum which shows the peaks indicative of the doped scaffold ...... 207 Figure 110:- EDX analysis of polymer coated silver/copper-doped BBG scaffolds including A) Type A 5wt% gelatin coated scaffold, B) Type A 6wt% gelatin coated scaffold, C) Type B 5wt% gelatin coated scaffold and D) Type B 6wt% gelatin coated scaffold after immersion in SBF for 14 days ... 208 Figure 111:- Wettability data for the gelatin coated silver/copper-doped BBG scaffolds showing spectra from before and after immersion in SBF for 14 days (Data shown are mean values and standard deviation of at least 3 measurements). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 210 Figure 112:- HPLSC’s attachment and viability on the surface of 3D BBG scaffolds with A) and b) two different BBG scaffolds. Viable cells are stained green and dead cells are stained red after a total of 10 days of cell culture ...... 215 Figure 113:- HPLSC's attachment and viability on the surface of low concentration silver MSIE 3D BBG scaffolds with a ranged of immersion periods in molten salt: A) immersion period of 15 minutes, B) immersion period of 30 minutes, C) immersion period of 45 minutes and D) immersion period of 60 minute. Viable cells are stained green and dead cells are stained red after a total of 10 days of cell culture ...... 216 Figure 114:- HPLSC's attachment and viability on the surface of medium concentration silver MSIE 3D BBG scaffolds with a ranged of immersion periods in molten salt: A) immersion period of 15 minutes, B) immersion period of 30 minutes, C) immersion period of 45 minutes and D) immersion period of 60 minutes. Viable cells are stained green and dead cells are stained red after a total of 10 days of cell culture ...... 217 Figure 115:- HPLSC's attachment and viability on the surface of high concentration silver MSIE 3D BBG scaffolds with a ranged of immersion periods in molten salt: A) immersion period of 15 minutes, B) immersion period of 30 minutes, C) immersion period of 45 minutes and D) immersion period of 60 minutes. Viable cells are stained green and dead cells are stained red after a total of 10 days of cell culture ...... 217 Figure 116:- 2D seeded rMSC’s on: A) control well with no sample, B) BBG scaffold, C) low concentration silver-doped BBG scaffold, D) low concentration copper-doped BBG scaffold and E) low concentration combination ion BBG scaffold after one day of contact with the cells ...... 220 Figure 117:- 2D seeded rMSC’s on: A) control well with no sample, B) BBG scaffold, C) low concentration silver-doped BBG scaffold, D) low concentration copper-doped BBG scaffold and E) low concentration combination ion BBG scaffold after three day of contact with the cells ...... 220 Figure 118:-Image illustrating an increase in the pH which leads to a change in the colouration of the CCM from A) the typical orange colouration to B) the bright pink colouration which indicates the activation of the phenol red indicator in the media ...... 221 Figure 119:- rMSC’s seeded onto pre-conditioned scaffolds: A) control well with insert one day after seeding, B) DMEM/F-12 pre-conditioned scaffold one day after seeding, C) 0.9% NaCl pre- conditioned scaffold one day after seeding, D) control well with insert two days after seeding, E) DMEM/F-12 pre-conditioned scaffold two days after seeding and F) 0.9% NaCl pre-conditioned scaffold two days after seeding ...... 224 Figure 120:- 2D rMSC cultures with inserts containing A) control well with no insert and samples, B) wells containing BBG scaffolds, C) wells containing low concentration silver-doped BBG scaffolds, D) wells containing low concentration copper-doped BBG scaffolds and E) wells containing low concentration combination ion-doped BBG scaffolds and then subjected to an AB assay. For copper- doped BBG scaffolds there was no detected metabolic activity after 14 days and for silver/copper-

18 doped BBG scaffolds no metabolic activity could be detected after 7 days. * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis...... 225 Figure 121:- 3D rMSC cultures containing A) control well with no samples, B) wells containing BBG scaffolds, C) wells containing low concentration silver-doped BBG scaffolds, D) wells containing low concentration copper-doped BBG scaffolds and E) wells containing low concentration combination ion-doped BBG scaffolds and then subjected to an AB assay. * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis...... 226 Figure 122:- 2D culture seeded with hMSC’s onto: A) BBG scaffold after one day, B) extremely low concentration copper-doped sample after one day, C) extremely low concentration combination ion- doped sample after one day, D) BBG scaffold after five days, E) extremely low concentration copper- doped sample after five days, F) extremely low concentration combination ion-doped sample after five days, G) BBG scaffold after ten days, H) extremely low concentration copper-doped sample after ten days and I) extremely low concentration combination ion-doped sample after ten days ...... 228 Figure 123:- 2D hMSC cultures with inserts containing A) BBG scaffolds, B) extremely low concentration copper-doped samples and C) extremely low concentration combination ion-doped samples and then subjected to an AB assay after one, five and ten days of being seeded with hMSC’s. For copper-doped and silver/copper-doped BBG scaffolds no metabolic activity could be detected after ten days. * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis...... 229 Figure 124:- Absorbance measure over a full visible light spectrum for E.coli grown in a variety of different media, with the yellow band showing the ideal wavelength range at which the absorbance should be measured ...... 232 Figure 125:- Growth of E.coli in both MHB and DMEM/F-12 with A) showing the full incubation period and B) showing the growth curves for the first five hours; the absorbance was measured at 650nm.(The lines connecting the dots are only to help the eye). * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis...... 233 Figure 126:- E.coli bacteria per ml measured using the CFU plating method taking samples at three o time points, plating onto standard I agar plates and incubated for 24 hours at 37 C and 5%CO2. (The lines connecting the dots are only to help the eye) * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis...... 233 Figure 127:- E.coli suspended in DMEM/F-12 with a range of different concentration solutions of o copper nitrate or copper chloride after being incubated for 24 hours at 37 C at 5% CO2: A) copper chloride series with an Δ absorption of 0.001, B) copper chloride series with an Δ absorption of 0.01 C) copper nitrate series with an Δ absorption of 0.001, and D) copper nitrate series with an Δ absorption of 0.01 with concentrations of i) control with no E.coli, ii) control with E.coli, iii) 0.05mM of copper salt, iv) 0.1mM of copper salt, v) 0.5mM of copper salt and vi) 1mM of copper salt ...... 235 Figure 128:- Standard I agar plates showing; A) E.coli suspension with silver nitrate at a concentration 0.05mM and additional sodium thiosulfate, B) E.coli suspension with silver nitrate at a concentration 0.05mM, C) E.coli suspension with copper nitrate at a concentration 0.05mM and additional sodium thiosulfate and D) E.coli suspension with copper nitrate at a concentration 0.05mM after being incubated in suspension for 24 hours, plated and further incubated for 24 hours at 5% CO2 and at 37oc using an initial E.coli change in absorption of 0.01 ...... 237 Figure 129:- The change in absorption of the bacterial suspension due the presence of a range of concentrations of silver nitrate over time which gives an indication of the growth rate of the suspended bacteria. * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis...... 238 Figure 130:- The change in absorption of the bacterial suspension due to the presence of a range of concentrations of copper nitrate over time which gives an indication of the growth rate of the

19 suspended bacteria. * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis...... 239

Figure 131:- Graph showing CO2 levels required to change the pH for given levels of NaHCO3 found in DMEM/F-12 to give a suitable pH value for cells and bacteria to grow ...... 240 Figure 132:- DMEM/F-12 and E.coli suspensions with A) silver nitrate with concentrations of iii) 0.01mM iv) 0.05mM v) 0.1mM and vi) 0.5mM, B) copper nitrate with concentrations of iii) 5mM iv) 2mM v) 1mM vi) 0.5mM and vii) 0.1mM, C) combination of silver and copper nitrate in a ratio of 1:1 with concentrations of iii)1 mM iv)0.5mM v)0.1mM vi) 0.05mM and vii) 0.01mM, with i) and ii) in each picture showing a plain media control and an E.coli suspension control, respectively ...... 241 Figure 133:-Suspensions of E.coli in DMEM/F-12 with supernatants added from scaffolds immersed in DMEM/F-12 for A) 1 day B) 3 days C) 7 days and D) 28 days after 24 hours of incubation at 37oc and 9.5% CO2 ...... 245 Figure 134:- Results of the agar diffusion study to show the growth inhibition of silver nitrate solutions with A) showing a plain media disk, 0.1mM, 0.05mM and 0.02mM concentrations of silver nitrate on filter paper disks and B) showing a plain media disk, 1mM, 0.5mM and 0.2mM concentrations of silver nitrate on filter paper disks ...... 248 Figure 135:- Result of the agar diffusion study to show the growth inhibition of dilutions of copper nitrate inoculated onto filter paper disks. Concentrations of 2mM, 1mM and 0.5mM were used along with a control disks inoculated with media ...... 248 Figure 136:- Results of the agar diffusion study to show the growth inhibition of: A) BBG scaffold, B) Low concentration silver-doped BBG scaffold, C) Low concentration copper-doped BBG scaffold and D) Low concentration combination ion-doped BBG scaffold in the presence of E.coli bacteria 249 Figure 137:- Results of agar diffusion to show the growth inhibition properties of: A) BBG scaffold, B) Low concentration silver-doped BBG scaffold, C) Low concentration copper-doped BBG scaffold and D) Low concentration combination ion-doped BBG scaffold in the presence of S.aureus bacteria after an incubation period of 24 hours ...... 251 Figure 138:- Author’s assessment of the contribution to the field of BTE based on the results of the present work ...... 264 Figure 139:- Metal nano-particle loaded gelatin coated BBG scaffolds: A) gelatin type A coated scaffold, B) gelatin type A coated scaffold loaded with copper oxide nano-particles, C) gelatin type A coated scaffold loaded with iron oxide nano-particles, D) gelatin type B coated scaffold, E) gelatin type B coated scaffold loaded with copper oxide nano-particles, and F) gelatin type B coated scaffold loaded with iron oxide nano-particles...... 272 Figure 140:- RMC model of silicon oxide showing the atomic structure as extrapolated from EXAFS data and other constraints (616) ...... 275 Figure 141:- The AV loop was directed through a Teflon isolation chamber which was filled with the BBG scaffold pieces and fibrin gel with a fibrinogen concentration of 10mg/ml and a thrombin concentration of 2 I.U./ml. (318) With A) showing the vascular pedicle in place in the cage before being covered in more BBG pieces and fibrin gel and B) showing the Teflon cage before the lid is placed...... 279 Figure 142:- Macroscopic appearance of the inside of the Teflon cage, (shown in Figure 141) after three weeks upon explanation. (318) ...... 280 Figure 143:- The material removed from the Teflon cage after being subjected to hematoxylin and eosin (H&E) staining to show newly formed tissue in the sample. The arrows are indicate the vascular pedicle that passed through the BBG scaffold and fibrin gel sample with a magnification of x25. (318) ...... 280 Figure 144:- Micro CT analysis giving a 3D representation of the newly formed vessels found within the material removed from the Teflon cage showing the final network structure of the vessels

20 including the original vascular pedicle (318) (Micro CT images obtained by Dr T Fey from the Institute of Glass and Ceramics, University of Erlangen-Nuremberg, Germany) ...... 281 Figure 145:- Skematic of a counting plate showing the areas that cells and bacteria are counted in . 339 Figure 146:- Counting chamber used in bacteria and cell counting ...... 339 Figure 147:- Skematic of which cells/bacteria can be included in the counting process ...... 340 Figure 148:- 45S5 Bioglass® –derived glass-ceramic scaffolds immersed in a medium concentration molten salt bath for A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes Images have a magnification of x150...... 341 Figure 149:- 45S5 Bioglass® –derived glass-ceramic scaffolds immersed in a high concentration molten salt bath for A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes Images have a magnification of x150...... 341 Figure 150:- Low concentration of silver MSIE scaffolds with immersion periods of 15, 30, 45 and 60 minutes ...... 342 Figure 151:- High concentration of silver MSIE scaffolds with immersion periods of 15, 30, 45 and 60 minutes ...... 342 Figure 152:- 45S5 Bioglass® –derived glass-ceramic scaffolds immersed in a low concentration molten salt bath for A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes and then immersed in SBF for 14 days. Images have a magnification of x1500 ...... 343 Figure 153:- 45S5 Bioglass® –derived glass-ceramic scaffolds immersed in a high concentration molten salt bath for A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes and then immersed in SBF for 14 days. Images have a magnification of x1500...... 343 Figure 154:- High concentration of silver MSIE scaffolds with immersion periods of 15, 30, 45 and 60 minutes after immersion in SBF for 14 days ...... 344 Figure 155:- Medium concentration of silver MSIE scaffolds with immersion periods of 15, 30, 45 and 60 minutes after immersion in SBF for 14 days ...... 344 Figure 156:- Control samples taken after four and 24 hours of exposure E.coli and being plated onto o standard I agar plates and incubated for 24 hours at 37 c at 5% CO2: A) initial bacterial absorbance of 0.01 and 0.001 after four hours of exposure, B) initial bacterial absorbance of 0.001 after 24 hours of exposure and C) initial bacterial absorbance of 0.01 after 24 hours of exposure using DMEM/F-12 as the suspension media ...... 345 Figure 157:- Control samples taken after four and 24 hours of exposure E.coli and being plated onto o standard I agar plates and incubated for 24 hours at 37 c at 5% CO2: A) initial bacterial absorbance of 0.01 and 0.001 after four hours of exposure, B) initial bacterial absorbance of 0.001 after 24 hours of exposure and C) initial bacterial absorbance of 0.01 after 24 hours of exposure using MHB as the suspension media ...... 345 Figure 158:- Samples taken after four hours of exposure to varying concentrations of copper chloride o salt and being plated onto standard I agar plates and incubated for 24 hours at 37 c at 5% CO2: A) concentrations of copper chloride of 1mM and 0.5mM with an initial bacterial absorbance of 0.01, B) concentrations of copper chloride of 0.1mM and 0.05mM with an initial bacterial absorbance of 0.01, C) concentration of copper chloride of 1mM with an initial bacterial absorbance of 0.001, D) concentration of copper chloride of 0.5mM with an initial bacterial absorbance of 0.001, E) concentration of copper chloride of 0.1mM with an initial bacterial absorbance of 0.001 and F) concentration of copper chloride of 0.05mM with an initial bacterial absorbance of 0.001 using DMEM/F-12 as the suspension media ...... 345 Figure 159:- Samples taken after four hours of exposure to varying concentrations of copper chloride o salt and being plated onto standard I agar plates and incubated for 24 hours at 37 c at 5% CO2: A) concentrations of copper chloride of 1mM with an initial bacterial absorbance of 0.01, B) concentrations of copper chloride of 0.5mM with an initial bacterial absorbance of 0.01,C)

21 concentrations of copper chloride of 0.1mM with an initial bacterial absorbance of 0.01, D) concentrations of copper chloride of 0.05mM with an initial bacterial absorbance of 0.01E) concentration of copper chloride of 1mM with an initial bacterial absorbance of 0.001, F) concentration of copper chloride of 0.5mM with an initial bacterial absorbance of 0.001, G) concentration of copper chloride of 0.1mM with an initial bacterial absorbance of 0.001 and H) concentration of copper chloride of 0.05mM with an initial bacterial absorbance of 0.001 using MHB as the suspension media ...... 346 Figure 160:- Samples taken after four hours of exposure to varying concentrations of copper nitrate salt, an E.coli suspension in DMEM/F-12 and being plated onto standard I agar plates and incubated o for 24 hours at 37 c at 5% CO2: A) concentration of copper nitrate of 1mM with an initial bacterial absorbance of 0.001, B) concentration of copper nitrate of 0.5mM with an initial bacterial absorbance of 0.001, C) concentration of copper nitrate of 0.1mM with an initial bacterial absorbance of 0.001, D) concentration of copper nitrate of 0.05mM with an initial bacterial absorbance of 0.001, E) concentration of copper nitrate of 1mM with an initial bacterial absorbance of 0.01, F) concentration of copper nitrate of 0.5mM with an initial bacterial absorbance of 0.01, G) concentration of copper nitrate of 0.1mM with an initial bacterial absorbance of 0.01, and H) concentration of copper nitrate of 0.05mM with an initial bacterial absorbance of 0.01 using DMEM/F-12 as the suspension media ... 346 Figure 161:- Samples taken after four hours of exposure to varying concentrations of copper nitrate salt, an E.coli suspension in MHB and being plated onto standard I agar plates and incubated for 24 o hours at 37 c at 5% CO2: A) concentration of copper nitrate of 1mM with an initial bacterial absorbance of 0.001, B) concentration of copper nitrate of 0.5mM with an initial bacterial absorbance of 0.001, C) concentration of copper nitrate of 0.1mM with an initial bacterial absorbance of 0.001, D) concentration of copper nitrate of 0.05mM with an initial bacterial absorbance of 0.001, E) concentration of copper nitrate of 1mM with an initial bacterial absorbance of 0.01, F) concentration of copper nitrate of 0.5mM with an initial bacterial absorbance of 0.01, G) concentration of copper nitrate of 0.1mM with an initial bacterial absorbance of 0.01, and H) concentration of copper nitrate of 0.05mM with an initial bacterial absorbance of 0.01 ...... 347 Figure 162:- Samples taken after 24 hours of exposure to varying concentrations of copper chloride salt, an E.coli suspension in DMEM/F-12 and being plated onto standard I agar plates and incubated o for 24 hours at 37 c at 5% CO2: A) concentration of copper chloride of 1mM with an initial bacterial absorbance of 0.01, B) concentration of copper chloride of 0.5mM with an initial bacterial absorbance of 0.01, C) concentration of copper chloride of 0.1mM with an initial bacterial absorbance of 0.01, D) concentration of copper chloride of 0.05mM with an initial bacterial absorbance of 0.01, E) concentration of copper chloride of 1mM with an initial bacterial absorbance of 0.001, F) concentration of copper chloride of 0.5mM with an initial bacterial absorbance of 0.001, G) concentration of copper chloride of 0.1mM with an initial bacterial absorbance of 0.001, and H) concentration of copper chloride of 0.05mM with an initial bacterial absorbance of 0.001 ...... 347 Figure 163:- Samples taken after 24 hours of exposure to varying concentrations of copper chloride salt, an E.coli suspension in MBH and being plated onto standard I agar plates and incubated for 24 o hours at 37 c at 5% CO2: A) concentration of copper chloride of 1mM with an initial bacterial absorbance of 0.01, B) concentration of copper chloride of 0.5mM with an initial bacterial absorbance of 0.01, C) concentration of copper chloride of 0.1mM with an initial bacterial absorbance of 0.01, D) concentration of copper chloride of 0.05mM with an initial bacterial absorbance of 0.01, E) concentration of copper chloride of 1mM with an initial bacterial absorbance of 0.001, F) concentration of copper chloride of 0.5mM with an initial bacterial absorbance of 0.001, G) concentration of copper chloride of 0.1mM with an initial bacterial absorbance of 0.001, and H) concentration of copper chloride of 0.05mM with an initial bacterial absorbance of 0.001 ...... 348

Figure 164:- Samples taken after 24 hours of exposure to varying concentrations of copper nitrate salt, an E.coli suspension in DMEM/F-12 and being plated onto standard I agar plates and incubated for 24 o hours at 37 c at 5% CO2: A) concentration of copper nitrate of 1mM with an initial bacterial absorbance of 0.01, B) concentration of copper nitrate of 0.5mM with an initial bacterial absorbance of 0.01, C) concentration of copper nitrate of 0.1mM with an initial bacterial absorbance of 0.01, D) concentration of copper nitrate of 0.05mM with an initial bacterial absorbance of 0.01, E) concentration of copper nitrate of 1mM with an initial bacterial absorbance of 0.001, F) concentration of copper nitrate of 0.5mM with an initial bacterial absorbance of 0.001, G) concentration of copper nitrate of 0.1mM with an initial bacterial absorbance of 0.001, and H) concentration of copper nitrate of 0.05mM with an initial bacterial absorbance of 0.001 ...... 348 Figure 165:- Samples taken after 24 hours of exposure to varying concentrations of copper nitrate salt, an E.coli suspension in MHB and being plated onto standard I agar plates and incubated for 24 hours o at 37 c at 5% CO2: A) concentration of copper nitrate of 1mM with an initial bacterial absorbance of 0.01, B) concentration of copper nitrate of 0.5mM with an initial bacterial absorbance of 0.01, C) concentration of copper nitrate of 0.1mM with an initial bacterial absorbance of 0.01, D) concentration of copper nitrate of 0.05mM with an initial bacterial absorbance of 0.01, E) concentration of copper nitrate of 1mM with an initial bacterial absorbance of 0.001, F) concentration of copper nitrate of 0.5mM with an initial bacterial absorbance of 0.001, G) concentration of copper nitrate of 0.1mM with an initial bacterial absorbance of 0.001, and H) concentration of copper nitrate of 0.05mM with an initial bacterial absorbance of 0.001 ...... 349 Figure 166:- Supernatents plated after an immersion period of 28 days in DMEM/F-12, added to an E.coli suspension for 24 hours and then plated onto standard I agar plates with sodium thiosulfate; A) plates showing the supernatents taken from sample 1 with samples including a plain 45S5 Bioglass® - glass-ceramic scaffold, the lowest concentration MSIE samples of each type and B) plates showing the supernatents taken from sample 2 with samples including a plain 45S5 Bioglass® -glass-ceramic scaffold, the lowest concentration MSIE samples of each type ...... 349 Figure 167:- Supernatents plated after an immersion period of 7 days in DMEM/F-12, added to an E.coli suspension for 24 hours and then plated onto standard I agar plates with sodium thiosulfate; A) plates showing the supernatents taken from sample 1 with samples including a plain 45S5 Bioglass® - glass-ceramic scaffold, the lowest concentration MSIE samples of each type and B) plates showing the supernatents taken from sample 2 with samples including a plain 45S5 Bioglass® -glass-ceramic scaffold, the lowest concentration MSIE samples of each type ...... 350 Figure 168:- Supernatents plated after an immersion period of 3 days in DMEM/F-12, added to an E.coli suspension for 24 hours and then plated onto standard I agar plates with sodium thiosulfate; A) plates showing the supernatents taken from sample 1 with samples including a plain 45S5 Bioglass® - glass-ceramic scaffold, the lowest concentration MSIE samples of each type and B) plates showing the supernatents taken from sample 2 with samples including a plain 45S5 Bioglass® -glass-ceramic scaffold, the lowest concentration MSIE samples of each type ...... 350 Figure 169:- Supernatents plated after an immersion period of 1 day in DMEM/F-12, added to an E.coli suspension for 24 hours and then plated onto standard I agar plates with sodium thiosulfate; A) plates showing the supernatents taken from sample 1 with samples including a plain 45S5 Bioglass® - glass-ceramic scaffold, the lowest concentration MSIE samples of each type and B) plates showing the supernatents taken from sample 2 with samples including a plain 45S5 Bioglass® -glass-ceramic scaffold, the lowest concentration MSIE samples of each type ...... 351 Figure 170:- FTIR spectra for PDLLA coated BBG scaffolds before and after immersion in SBF for 14 days from Figure 84 showing just the section of the spectra relating to the PDLLA peaks ...... 351 Figure 171:- FTIR spectra for gelatin coated BBG scaffolds from Figure 85 just showing the peaks relating to the gelatin coating ...... 352

Figure 172:- FTIR spectra for gelatin coated BBG scaffolds after immersion in SBF for 14 days from Figure 91 just showing the peaks relating to the gelatin coating ...... 352 Figure 173:- FTIR spectra for gelatin coated copper-doped BBG scaffolds from Figure 96 just showing the peaks relating to the gelatin coating ...... 353 Figure 174:- FTIR spectra for gelatin coated copper-doped BBG scaffolds after immersion in SBF for 14 days from Figure 100 just showing the peaks relating to the gelatin coating ...... 353 Figure 175:- FTIR spectra for gelatin coated silver/copper-doped BBG scaffolds from Figure 105 just showing the peaks relating to the gelatin coating ...... 354 Figure 176:- FTIR spectra for gelatin coated silver/copper-doped BBG scaffolds after immersion in SBF for 14 days from Figure 109 just showing the peaks relating to the gelatin coating ...... 354 Figure 177:- FTIR analysis of BBG scaffolds coated with a) 6wt% type A gelatin with copper nano- particles, b) 6wt% type A gelatin doped with iron oxide nano-particles, c) 6wt% type B gelatin doped with copper nano-particles and d) 6wt% type B gelatin doped with iron oxide nano-particles ...... 361 Figure 178:- BBG scaffolds coated with gelatin with A) a type A 5wt% coating with iron oxide nano- particles (magnification of x500), B) a type B 5wt% coating with iron oxide nano-particles (magnification of x250), C) a type A 5wt% coating with copper oxide nanoparticles (magnification of x 300) and D) a type B 5wt% coating with copper oxide nanoparticles (magnification of x160) ...... 362 Figure 179:- Permission granted for the use of the images and data shown in Figure 137 and Table 49 from Dr Timo Hammer ...... 366 Figure 180:- Permission granted for the use of the images shown in Figure 116, Figure 117, Figure 119, Figure 120, Figure 121, Figure 122 and Figure 123 from Dr Leonie Strobel ...... 366 Figure 181:- Permission granted for the use of the images shown in Figure 112, Figure 113, Figure 114 and Figure 115 from Dr Reem El-Gendy ...... 367 Figure 182:- Permission granted for the use of the images shown in Figure 141, Figure 142, Figure 143 and Figure 144 from Dr Andreas Arkudas ...... 367 Figure 183:- Permission granted for the use of the image shown in Figure 12 from the publisher John Wiley and Sons ...... 368 Figure 184:- Permission granted for the use of the image shown in Figure 11 from the publisher Elsevier ...... 368 Figure 185:- Permission granted for the use of the image shown in Figure 4 from the publisher Springer ...... 369 Figure 186:- Permission granted for the use of the image shown in Figure 7 from the publisher Springer ...... 369

List of Tables

Table 1:- The different types of alloplastic graft (50) ...... 40 Table 2:- Concentrations, ratios and quantities of sodium nitrate and silver nitrate used in the preliminary study of the silver MSIE process ...... 89 Table 3:- Concentrations, ratios and quantities of sodium nitrate and silver nitrate used in the main study of the silver MSIE ...... 90 Table 4:- Concentrations, ratios and quantities of sodium nitrate and copper nitrate used in the study of the copper MSIE ...... 92 Table 5:- Concentrations and quantities of gelatin used in the preliminary studies ...... 95 Table 6:- Chemicals used for the production of simulated body fluid (SBF) ...... 97 Table 7:- Washing protocol used for the cell biology work using rMSC’s and hMSC’s ...... 108 Table 8:- Working dilution table of the dilutions required to get the required number of CFU’s on an agar plate ...... 115 Table 9:- Metal salts concentrations used in the minimum inhibition study ...... 117 Table 10:- Average compressive strength of BBG scaffolds after immersion in SBF. * Indicates a statistically significant change due to the immersion period (p≤0.05) using ANOVA analysis...... 130 Table 11:- White light interferometry results for the surface roughness of BBG pellets after immersion in SBF for a number of days. * Indicates a statistically significant change due to the change in immersion period (p≤0.05) using ANOVA analysis...... 130 Table 12:- Compressive strength of silver MSIE BBG scaffolds. (113) * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 143 Table 13:- White light interferometry results for the surface roughness of silver doped 45S5 Bioglass® –derived glass-ceramic pellets * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 144 Table 14-: Compressive strength of silver MSIE BBG scaffolds after immersion in SBF for 14 days. (113) * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis. 148 Table 15:- White light interferometry results for the surface roughness of silver doped BBG pellets after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 149 Table 16:- Compressive strength of copper MSIE BBG scaffolds (samples prepared by MSIE for 15 minutes). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 155 Table 17:- White light interferometry results for the surface roughness of copper doped BBG pellets. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 155 Table 18:- Compressive strength of copper MSIE BBG scaffolds after immersion in SBF for 14 days (samples prepared by MSIE for 15 minutes). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 160 Table 19:- White light interferometry results for the surface roughness of copper doped BBG pellets after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 161 Table 20:- Compressive strength of silver/copper-doped MSIE BBG scaffolds obtained by immersion in lower concentration salt bath for the given time periods. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 167 Table 21:- White light interferometry results for the surface roughness of combination ion doped BBG pellets. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 167

Table 22:- Compressive strength of combination MSIE BBG scaffolds after immersion in SBF for 14 days. Samples were produced by MSIE, immersing them in the lowest concentration salt bath for the given periods of time. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 172 Table 23:- White light interferometry results for the surface roughness of combination ion BBG pellets after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 172 Table 24:- Porosity data for gelatin type A coated BBG scaffolds for a range of polymer immersion periods. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 175 Table 25:- Porosity data for gelatin type B coated BBG scaffolds for a range of polymer immersion periods. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 175 Table 26:- Porosity data for multiple coats of gelatin on BBG scaffolds. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 177 Table 27:- Compressive strength of polymer coated BBG scaffolds. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 183 Table 28:- White light interferometry results for the surface roughness of polymer coated BBG pellets. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 183 Table 29:- Compressive strength of polymer coated BBG scaffolds after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis. 189 Table 30:- White light interferometry results for the surface roughness of polymer coated 45S5 Bioglass® –derived glass-ceramic pellets after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 190 Table 31:- Compressive strength of copper doped gelatin coated BBG scaffolds. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 195 Table 32:- White light interferometry results for the surface roughness of gelatin coated copper doped BBG pellets. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 195 Table 33:- Compressive strength of gelatin coated copper doped BBG scaffolds after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 199 Table 34:- White light interferometry results for the surface roughness of gelatin coated copper doped BBG pellets after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 199 Table 35: Compressive strength of silver/copper-doped polymer coated BBG scaffolds. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 204 Table 36:- White light interferometry results for the surface roughness of gelatin coated, silver/copper-doped BBG pellets . * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 204 Table 37:- Compressive strength of silver/copper-doped polymer coated BBG scaffolds after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 208 Table 38:- White light interferometry results for the surface roughness of gelatin coated, silver/copper-doped BBG pellets after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis...... 209 Table 39:- pH changes throughout the washing procedure ...... 222

Table 40:- Observations made after introducing a longer pre-conditioning phase based on section 4.4.5 ...... 223 Table 41:- CFU counts from copper salt suspensions taken after four and 24 hours using both media types ...... 236 Table 42:- Colony forming units counted after being exposed to varying concentrations of silver nitrate in the E.coli suspension after 24 hours of incubation (TMTC meaning too many to count) ... 243 Table 43:- Colony forming units counted after being exposed to varying concentrations of copper nitrate in the E.coli suspension after 24 hours of incubation (TMTC meaning too many to count) ... 243 Table 44:- Colony forming units counted after being exposed to varying concentrations of combination of copper nitrate and silver nitrate, in a 1:1 ratio, in the E.coli suspension after 24 hours of incubation (TMTC meaning too many to count) ...... 243 Table 45:- Colony forming units counted after being exposed to supernatants from immersed plain o scaffolds after 24 hours incubating at 37 C at 9.5% CO2 ...... 245 Table 46:- Colony forming units counted after being exposed to supernatants from immersed low o concentration silver MSIE scaffolds after 24 hours incubating at 37 C at 9.5% CO2 ...... 246 Table 47:- Colony forming units counted after being exposed to supernatants from immersed low o concentration copper MSIE scaffolds after 24 hours incubating at 37 C at 9.5% CO2 ...... 246 Table 48:- Colony forming units counted after being exposed to supernatants from immersed low o concentration silver/copper MSIE scaffolds after 24 hours incubating at 37 C at 9.5% CO2 ...... 246 Table 49:- Results of bacterial experiments carried out on 3D scaffolds loaded with K.pneumoniae and incubated for 18 hours at 36oC ...... 252 Table 50:-Relevant metallic ion candidates for MSIE scaffolds ...... 268 Table 51:- Collated data for a variety of materials fabricated by one of the four fabrication methods discussed in chapter two, showing the change in porosity against compressional strength with FC meaning freeze casting, FR meaning foam replication, SFF meaning solid free-form replication and SG meaning sol-gel processing ...... 326 Table 52:- Collated data for a variety of materials fabricated by one of the four fabrication methods discussed in chapter two, with the addition of a polymer coating, showing the change in porosity against compressional strength with FC meaning freeze casting, FR meaning foam replication, SFF meaning solid free-form replication and SG meaning sol-gel processing ...... 330 Table 53:- Collated data for a variety of materials fabricated by one of the four fabrication methods discussed in chapter two, subjected to MSIE coating, showing the change in porosity against compressional strength with FR meaning foam replication ...... 331 Table 54:- Permission status of copyright of figures, tables and data used in this work ...... 365

List of Equations

Equation 1:-Porosity of an uncoated scaffold ...... 52 Equation 2:- Mechanism of silver-sodium molten salt ion exchange in a glass ...... 69 Equation 3:- Molten salt ion exchange process to introduce silver into a pre-fabricated BBG scaffold ...... 69 Equation 4:- Mechanism of copper-sodium molten salt ion exchange in a glass ...... 70 Equation 5:- Molten salt ion exchange process to introduce copper into a pre-fabricated BBG scaffold ...... 70 Equation 6:- Mechanism of copper/silver-sodium molten salt ion exchange in a glass ...... 71 Equation 7:- Molten salt ion exchange process to introduce copper and silver into a pre-fabricated BBG scaffold ...... 71 Equation 8:- Porosity of polymer coated scaffolds (210) ...... 96 Equation 9:- RMS of a sample being characterised using white light interferometry (401)...... 100 Equation 10:- Ra of a sample being characterised using white light interferometry (401)...... 100 Equation 11:- AB reduction equation ...... 112 Equation 12:- Correction factor for the AB reduction equation ...... 112 Equation 13:- Calculation of the amount of pre-culture stock solution which is required to be transferred to the stock solution ...... 114 Equation 14:- Number of bacteria per ml to work out dilutions for plating the bacteria ...... 114 Equation 15:- Number of bacteria per mL from counting CFU’s ...... 116 Equation 16 :- MC modelling – the probability of the original configuration ...... 356 Equation 17 :- MC modelling – the probability of the new configuration after one atom move ...... 357 Equation 18 :- MC modelling – Constraint of whether an atom stays in the new position or moved back to its original position ...... 357 Equation 19 :- RMC modelling –Experimental Structure factor relationship ...... 357 Equation 20 :- RMC modelling – Probability of the difference in experimental structure factors in terms of the standard deviation of the normal distribution ...... 357 Equation 21 :- RMC modelling – Probability of the real structure factor...... 358 Equation 22 :- RMC modelling –Standard deviation in terms of the scattering vector ...... 358 Equation 23:- RMC modelling - Constraint of whether an atom stays in the new position or moved back to its original position ...... 358 Equation 24 :- RMC modelling – Radial distribution function for the initial configuration of atoms 358 Equation 25 :- RMC modelling – Transformation of the radial distribution into the total structure factor ...... 359 Equation 26 :- RMC modelling – Difference between the experimental structure factor and the starting structure factor ...... 359 Equation 27 :- RMC modelling - Constraint of whether an atom stays in the new position or moved back to its original position ...... 359 Equation 28 :- RMC modelling - Constraint of whether an atom stays in the new position or moved back to its original position plus the new coordination factor ...... 360 Equation 29:- Mean of a data set ...... 363 Equation 30:- Standard error on the mean ...... 363 Equation 31:- Group mean of multiple data sets ...... 363 Equation 32:- ANOVA analysis of multiple data sets ...... 363 Equation 33:- ANOVA analysis of multiple data sets ...... 364 Equation 34:- ANOVA analysis of multiple data sets ...... 364 Equation 35:- ANOVA analysis of multiple data sets ...... 364

Equation 36:- Student T Test of two data sets ...... 364

Chapter One

1. Introduction

The demand for transplantable tissue currently far out strips the supply of donor organs considering the increasing volume of tissue/organ transplants that are required due to an ageing population. The scientific community is devoting efforts to develop a range of strategies to produce alternative solutions to combat the increase in demand for transplantable tissue. One area that has rapidly expanded is the use of artificial materials to produce scaffolds for tissue regeneration. (1) In the field of bone regeneration the use of bioactive glass-ceramic scaffolds is gaining increasing interest for novel replacement/regeneration therapy, which is the research area in which this work has been carried out.

Bioactive glass -derived glass-ceramic scaffolds have attracted in the last few years increased scientific interest. (2) They are highly suitable 3-D porous structures for bone tissue engineering (BTE) which provide mechanical support to the surrounding tissue whilst promoting new bone growth; in essence they fulfil the basic criteria set out for BTE. These base scaffolds not only exhibit a high degree of bioactivity and osteoconductivity but also suitable biodegradability. A family of scaffolds is based

® on the material Bioglass (45% SiO2, 24.5% Na2O, 24.4% CaO and 6% P2O5 in wt %). (3) These Bioglass® -derived scaffolds (4) (5) developed for the first time in 2006 still need to be extensively improved for realistic consideration for clinical studies.

Thus this work branches out from previous research carried out with 45S5 Bioglass® -derived glass-ceramic (BBG) scaffolds as the starting biomaterial. Produced using the foam replication technique, (4) (5) these scaffolds are attractive as starting structures for developing advanced multifunctional scaffolds with optimised properties for use in BTE. These BBG scaffolds have a pore structure that tried to mimic the trabecular bone porosity and they have an ability to induce the formation of hydroxyapatite (HA) on their surfaces in physiological conditions which by its nature encourages bone attachment and promotes bone regeneration. However these scaffolds exhibit certain drawbacks, for example in terms of their mechanical competence as they do not have the required compressive strength required for the safe handling and for load bearing applications. Moreover these scaffolds have not yet been fully tailored to further enhance their bone regeneration capabilities and the scaffolds structural integrity including their resistance to crack propagation (work of fracture) has not been optimised. Apart from the structural similarity to bone and the ability to form HA on their surfaces, BBG scaffolds have only recently been considered to incorporate other functions which are relevant for tissue engineering applications such as antibacterial effect or drug delivery ability. The ultimate aim of

30 this work is thus to develop multifunctional BBG based composite scaffolds suitable for BTE and exhibiting a range of properties (and combination of properties) focusing on antibacterial effect coupled with bioactivity and improved structural and mechanical integrity. A range of approaches were explored in this project in order to produce multifunctional scaffolds that achieve the projects aims, which are summarised in Figure 1.

Figure 1:- Overview of the development of multifunctional BBG scaffolds within the frame-work of this project

As shown in Figure 1 this work followed two paths to evolving the BBG scaffolds using molten salt ion exchange (MSIE) to introduce therapeutic metal ions, initially investigated by Di Nunzio et al (6) on other silicate systems and expanding on the use of polymer coatings, which has been previously researched, (7) by investigating the use of gelatin.

Processing techniques such as foam replica technique to fabricate the BBG scaffolds, MSIE to introduce metal ions into the scaffolds structure and slurry dipping to obtain the novel polymer coatings have been utilized in this project. Simulated body fluid (SBF) studies were carried out to determine the bioactivity of the scaffolds. Along with mechanical testing of the scaffolds, characterisation techniques have ranged from scanning electron microscopy (SEM) to Fourier transform infra-red spectroscopy (FTIR), porosity studies, surface analysis and characterisation of the degradation behaviour. Combined with biological and bacterial testing, a comprehensive study about the modified scaffolds in comparison to the BBG scaffolds has been carried out.

Metal ions such as silver, zinc and copper have been shown to have some unique biological properties being, for example antibacterial and in some cases angiogenic. (8) (9) (10) (11) Considering the emerging field of using functional therapeutic ions in BTE (12) and the potential biological advantages provided by these ions and their controlled release, the present project is of relevance in the broader field of ion releasing scaffolds. In this project metal ions were incorporated into the surface structure of the scaffold, through MSIE, and the effect of ion incorporation on the structure and bioactive behaviour of the scaffolds was investigated.

This work also explored the use of polymer coatings to improve the mechanical properties of the base scaffold and to provide scaffolds with drug delivery ability, e.g. with the polymer coating acting as a possible drug carrier. An established synthetic polymer such as poly-DL-lactic acid (PDLLA) and the novel use of gelatin, comparing both gelatin type A and type B, were considered to explore gelatin’s suitability to coat BBG scaffolds, assessing how the polymer coatings can improve the scaffolds properties.

The thesis is organised in the following manner. An extensive literature review is presented in chapter two, covering the current approaches and techniques which are relevant to this research field, in order to justify the materials and techniques used in this work. Chapter three presents the aims and objectives of this work and the strategies and experimental approaches employed to achieve these. Chapter four describes the materials and characterisation techniques used within this work. Chapter five presents the structural characterisation results and discussion of the different types of scaffold developed, whilst chapter six describes the results and discussion from the antibacterial studies and biological characterisation of the different types of scaffolds. Chapter seven contains the concluding remarks presenting a summary of the key results obtained in this work and finally chapter eight makes suggestions for future work in general as well as describing some of the developing work that has been started as immediate follow-up research activities originating from the presented work.

Chapter Two

2. Literature review

2.1 Introduction

This chapter focuses on the key concepts which illustrate the science backbone of this work, including an overview of the current research surrounding tissue engineering concentrating in particular on BTE, the need for bioactive materials and current strategies for their use, the biology of bone and how this relates to BTE and a summary of the materials involved in the development of BTE scaffolds. This chapter will provide information behind the rationale for the choice of the materials and processes used throughout this work for the development of multifunctional scaffolds for BTE.

2.2 Tissue engineering: the general principles

Tissue engineering (TE) is an interdisciplinary research field which incorporates interaction between scientists of different backgrounds including biologists, engineers, material specialists, medical doctors and computer modelling experts since the mid 1980's. (13) (14) (15) This area of research can be described as a field that strives to produces techniques and tissue constructs by combining biomaterials, cells and signalling molecules that can be used in the repair or replacement of human tissue and organs. (16) (17) (18) TE utilizes thus principles from a variety of fields, such as cell biology, (19) material science, (20) biochemistry (21) and medicine (22) and it provides a different approach than more traditional therapeutic methods. (23) In orthopaedics, TE is important considering the increasing need for better therapeutic approaches, technologies and biomaterials to overcome the limitations of current orthopaedic implants. (24) (25)

The TE field is split into several areas depending on the type of tissue that is being investigated; this work concentrates on bone tissue engineering (BTE). Other areas include cardiac tissue, (26) cartilage, (27) lung tissue, (28) liver tissue, (29) intestinal tissue, (30) nerve tissue (31) and dentistry. (32) In principle, the two strategies deployed fall under two general categories:-

 The use of artificial matrices  The use of natural (decellularized) matrices (33)

The artificial matrices are manufactured from an appropriate biomaterial and then seeded with the appropriate cells (34) and this is the category that the work present here will fall under. Natural matrices are collagen-rich structures of animal origin (35) which are produced by chemical and mechanical manipulation of the tissue to remove the cellular components (decellularized scaffolds). (36) (37) (38)

Tissues consists of three components: the cells, the extracellular matrix (ECM) and signalling systems and since the premise of TE is to mimic a specific tissue, any engineered tissue must have a similar triad of components. (15) (39) In this work, the ECM is the component that will be the main focus as this is the tissue component being mimicked by the scaffold.

TE has enjoyed success in creating complex constructs that fulfil the criteria of tissue regeneration and repair. In recent years the research community has focused on tissues that have a predominately biomechanical function because of the challenges that these tissue types present and the extreme demand for them. (40) (41) (42)(24) This indicates that any developed material needs to not only assist the repair of the tissue but also needs to support the surrounding tissue by taking on the role of the host tissue. This requirement means that scaffolds have to be able to mimic the properties and structure of the host tissue in order to encourage and support the new tissue. (42) This set of principles is highly relevant for BTE as the biomechanical properties of bone are critical to their function in vivo. (42) If the scaffold is designed to follow these principles, other problems that can occur after implantation can be overcome, such as the possibility of implant rejection and of an inflammatory response.

2.3 The need for tissue engineering

At the present the demand for transplantable tissue far out strips the supply of donor organs due to the sheer volume of tissue/organ transplants that are currently needed. (43) (44) In the UK there were 5532 patients on the organ transplant waiting list in March 2000, with 2311 transplants being completed. This rose to 7997 and 2645, on the waiting list and transplants being carried out respectively, a decade later, as shown in Figure 2. (45) (40) (41) Other types of surgeries that are in great demand are reconstruction/repair procedures, such as bone grafts and muscle grafts, for the repair of tissue due to significant trauma, infection or cancer. (46)

Figure 2:-Transplant waiting lists and the number of operations carried out over a ten year period for the UK (24)

Suitable engineered tissues therefore are the logical way forward to reduce this difference between those patients waiting for transplants and those actually them. (47) (48) It is expected that the number of people dying whilst waiting for an organ transplant can be reduced by using engineered tissues as the replacement therapy. (49) (50)

2.4 Bone tissue engineering

The skeletal system is an integral part of the human body, acting not only as a scaffold but also as the storage reservoir for elements, including 99% of the body's calcium. (51) Bone is the main focus of this project. There are a number of reasons to focus the attention to bone regeneration approaches:-

 Bone defects are extremely common (52) and can be highly debilitating for the patient (53) (54)  A shortage of viable material to use for bone grafts (55) (56) due to the increasing volume of bone defects, (57) ranging from shattered bones (58) to facial reconstruction, (55) that require surgical repair (59)  Simpler bone fractures can heal under conventional therapies and do not require major surgery, (46) but the more problematic bone defects, caused by significant trauma, (60) infection (61) or cancer (46) are difficult to repair surgically (62) and affect the quality of life of the patient due to loss of function of the bone. (63)

To better understand the task in hand, the structure, composition, behaviour and the role of bone in the human body need to be explored. Indeed the whole concept of scaffolds for TE hinges on how

35 well the material mimics the structure and properties of the tissue it will be replacing, supporting and regenerating. (64)

A skeletal system has a range of functions that are categorised into mechanical, synthesis and metabolic. (65) Figure 3 shows the functions of bone in the human body and their grouping. The bones in the human skeletal system can be split into several regions with different structures; long bones which are generally concerned with movement, (66) (67) the tibia and femur for example, (62) and flat bones which have more of a protection role, (68) (69) for example the bones of the skull and the rib cage. (65)

Figure 3:- Schematic diagram showing the functions of bone in the human body (65)

Bone is a complex composite matrix, which comprises of approximately 95% collagen I

combined with HA (Ca10 (PO4)6(OH) 2) crystals. (70) Bone is formed of two layers, a dense outer layer called cortical bone and an inner spongy layer called trabecular bone. (71) (72) Approximately 80% of a human’s skeletal mass is made up by cortical bone and it is characterised by its smooth surface and closed cell network. (71) (72) Cortical bone is a dense structure with a porosity of about 10% (65) which is important as it acts as a protective layer to the trabecular bone. (73) Trabecular bone is made up of an open porous network allowing blood vessels, nutrients and cells to pass through the bone. (71) (72) Trabecular bone accounts for the remaining 20% of the total bone mass in the human skeletal system but has a surface area ten times that of cortical bone. (71) (72) This type of bone has porosity in the range of 50-90%. (65) (70) (74) (75) The compressive strength of cortical and trabecular bone ranges from 130 to 180MPa (76) and 0.8-11MPa (77), respectively. (33) The mechanical properties of the struts contained within both types of bone are very similar with a compressive strength of around 136MPa. (4) Figure 4 shows the structure of an adult human bone including the two layers discussed above.

Figure 4:-Schematic diagram of the cross section of an adult human bone showing the structure in detail (78) (75)

Within bone there are four cell types present: osteoblasts, osteocytes, osteoclasts and bone lining cells each with their own unique purpose. (70) (74) Bone lining cells are believed to be inactive osteoblasts which cover every available surface of the bone structure, acting as a protective barrier against certain ions and are ready to become active when required. (70) Osteoblast cells are cuboid when mature, are responsible for synthesizing the bone matrix (the osteoid), contributing to bone mineralisation and they deposit the bone matrix at a rate of 1µm per day in an adult. (70) (75) Osteoclast cells are highly specialized and multinucleated cells whose sole purpose is the resorption of bone. (79) Osteoblast cells regulate the osteoclast cells activity, which can absorb as much as 200, 000 µm3 of bone per day, which is roughly the amount of bone that is formed by seven to ten generations of osteoblasts. (75) Osteocytes are derived from osteoblast cells and are buried in the bone matrix during bone formation. Osteocytes are the primary cells found in bone and these cells initiate the recruitment of osteoblasts and osteoclasts to a site which needs remodelling or repairing, acting as damage sensors. (70) (75)

In order to understand how bone regenerates after a trauma, it is essential to understand how a bone remodels itself under normal conditions. This is a continuous process of bone resorption by the osteoclasts coupled with bone formation by the osteoblasts. This cycle of resorption and repair takes approximately three to six months to complete in an adult skeleton (75) and occurring at different rates depending on the type of bone (cortical or trabecular) and the location of the bone. (70) Around 1-3% of an adult’s skeletal system will be replaced each year up until the age of between 30-35 years old and this will ensure a constant density in the body’s bone. (65) The ratio between bone formation and bone

37 resorption will change in favour of the bone resorption after the age of approximately 35 years meaning the bone density will start to decrease. (65) This constant updating of the skeletal bone has evolved to help the skeleton to adapt when the mechanical conditions change whilst still maintaining the ability to participate in mineral regulation. (70) Normal bone remodelling cycle has five stages (65):-

 Quiescence  Activation  Resorption  Formation  Mineralisation

The remodelling cycle under normal conditions is depicted in Figure 5.

Quiescence is the pre-remodelling state of the bone, with a layer of inactive bone lining cells on the upper most surface of the bone with a thin collagenous membrane underneath the cells which separates them from the mineralised bone. (65) (80) The activation phase starts with the removal of the membrane so the osteoclasts can start the process of bone resorption. Mononuclear cells travel to the cleared site, attach to the bone surface and then fuse together to become multinuclear osteoclasts. The mature osteoclasts begin the resorption of the mineralised bone matrix resulting in a cavity, this is the resorption stage. (81) (82) This process stops when the cavity reaches a depth of between 60-100µm. (81) In the formation phase, osteoblasts produce and deposit the components of ECM with the collagen fibres organising themselves into the lamellar structure of bone. (81) The initial rate of deposition is high at 2-3µm per day and is due to the high density of osteoblasts present at the start of this stage. (81) Once this un-mineralised osteoid reaches a width greater than 20µm the mineralisation phase can start. The mineralisation process starts at a rate of 1-2µm per day in order not to interrupt the formation process. The mineralisation continues, after the formation phase has ceased, until the boundary between the existing bone and new bone is removed. The remaining cells on the bone are then no longer required and they deposit themselves on to the surface of the bone, deactivate and become bone lining cells and returning the cycle to the quiescence phase. (65)

This process is the body’s mechanism for bone remodelling, but is only part of the process of repair undertaken with a bone suffers a significant trauma. The four stages of the repair process are (82) : -

 Inflammation  Soft callus formation  Hard callus formation  Remodelling

Figure 5:- Bone remodelling cycle under normal conditions (65) (81)

All of these stages need to be taken into consideration when designing a material/implant that is going to replace, support and help to rejuvenate or regenerate the bone. When there is no other option but to replace a large portion of bone several treatment options are currently available and these can be thought of as either synthetic or natural bone grafts. Figure 6 illustrates the different types of bone graft/implant that are currently being used in bone replacement therapy.

Natural bone grafts have been around since the mid-17th century (83) from work carried out by Anthonie van Leeuwenhock and Job van Meekeren (83) through the implantation of a piece of dog skulls into a soldiers skull and are more widely used than the synthetic bone grafts (84) which have only been around since the discovery of bioactive properties of certain ceramics (3) in particular bioactive glasses. (85)

Figure 6:- Types of bone implants currently available as treatment options for significant bone trauma (84)

Autograft implants are considered the gold standard of bone implants (86) for a number of reasons, the main one being that this type of implant is taken from the patient who requires the bone graft. (86) This type of graft avoids the problem of the body rejecting the implant but as with all transplants there is still the risk of infection. (84) Autograft implants are usually gathered from the iliac crest of the patient. (87) Although this type of implant is the implant of choice it does have its problems including that there is only a very limited amount of bone that can be harvested to use as a graft and that the harvesting procedure is not only painful it comes with its own set of complications. (84)

Allograft implants are harvested from a deceased human donor. (88) This type of implant has the advantage over Autograft implants that it doesn’t require a secondary surgical site as the implant is coming from another source. (89) However since this implant comes from another human it does suffer from the problem that the host body can reject the implant, (90) although the closest match in terms of blood type and antibodies is implanted. (91) This type of implant also suffers from the risk of infection due to the implantation of a foreign object into the patient; (92) this also induces an inflammatory response from the tissue surrounding the implant. (84)

Xenograft implants are grafts that come from non-human sources and were first used back in the 16th century. (93) These types of grafts are not yet widely used in modern medicine as they normally induce an immune response from the host’s body and the graft doesn’t induce new bone growth. (84) However extensive research is being carried out in improving the decullularisation process to make this type of scaffold/implant a more realistic alturnative.

Alloplastic implants can be inorganic, synthetic, organic or a combination of these and can be implanted as a treatment for bone trauma instead of the other natural grafts; (94) Table 1 shows the different types of alloplastic grafts.

Table 1:- The different types of alloplastic graft (50)

Group Type of Graft Examples References 1 Biological - organic Decellularized bone matrix (95) (96) Calcium phosphates (97) (98) 2 Synthetic - inorganic HA (99) (100) Bioglass® (101) (102) Collagen (103) (104) 3 Synthetic - organic Poly-L-lactide (105) (106) Poly(D,L-lactide-co-glycolide) (107) (108) 4 Composites Composites of groups 1-3 (109) (110)

This project is concentrating on alloplastic grafts from group four, Table 1, classed as composite materials.

2.5 Challenges faced in bone tissue engineering

Some traumas to bone, such as open fractures, result in infection and contamination. (111) It is extremely difficult to control the resulting infection and treat the underlying bone defect with a graft at the same time. (112) This results in the patient’s treatment occurring in two parts, firstly controlling the infection and then once that has cleared up the bone graft surgery can go ahead. This is of course is expensive and results in an extended stay in the hospital for the patient. (112) This also needs to be taken into consideration during the design phase of the implant, such as introducing an antibacterial element into the scaffolds atomic matrix to reduce the likelihood of post-operative infections. (113)

Another problem with implants is the body’s immune response after implantation which will lead to the body rejecting the implant. (114) Until recently it has been accepted by the medical community that the possibility of an implantation procedure occurring without the need for immunosuppressant drugs was impossible unless the bone graft was an autograft implant. (115)

Unfortunately the problem of rejection can’t always be solved as the mechanism of the body rejecting the implant is the same as the body’s reaction to foreign organisms such as viruses and bacteria. (116)This is because the body’s immune system can’t distinguish between dangerous bacteria and a perfectly safe bone graft or implant. (117) (118) The need for implants to not be rejected is important because rejected implants require a longer stay in hospital increasing the cost to the hospital and increasing the discomfort for the patient. (119) The main solution to the immune response would be to produce an implant that is sufficiently inert not to produce an immune response but will fulfil the criteria for a TE implant that was set out in section 2.2.

In recent years the developments of biomaterials, seeded with cells, have shown not to induce such a severe immunological response from the patient. (115) The implants that are produced in this way have not been rejected, with over 50,000 cases of this type of implant occurring and no rejections. (115) These implants are normally alloplastic and in order not to induce an immune response the implants should elicit a positive reaction at the interface with the surrounding tissue. (115) The BBG scaffolds that this work will be focusing on fall under this type of implant, e.g. Bioglass® is a bioactive material that induces new bone growth (3) without producing a severe immune response in the body. (120) These specific characteristics of bioactive glasses, more specifically Bioglass®, will be discussed section 2.7.2.

2.6 Properties for an ideal bone tissue engineering material

In order for a material to be considered as a viable candidate for TE it needs to fulfil certain criteria. These are general criteria which not only apply to scaffolds for BTE but to other types of implants as well. These are split into two general areas (7) (121):-

 Biological

 Structural

If a material exhibits suitable biological and structural properties and also tries to mimic the requirements of the tissue it is trying to replace then it represents a candidate for tissue engineering.

2.6.1 Biological properties required

Before describing how the biomaterials are classified in terms of their interaction with the surrounding tissue, it must be highlighted that materials must be biocompatible. Biocompatible materials do not release elements in a concentration that is toxic and must not trigger an immune response from the host tissue which could lead to reactions such as inflammation and a septic rejection. (122) Whilst allowing the material to perform the desired function and generating the appropriate beneficial cellular response. (123)

How the material interacts with the surrounding tissue has a direct influence on how the material is used as an implant (or scaffold) or part of an implantable device and this interaction is classified into three types:-

 Bioinert  Bioactive  Bioresorbable/Biodegradable

Bioinert materials do not interact with or cause a response in the surrounding tissue. These materials do not release any toxic elements but they also do not have a positive influence on the host tissue making the materials biocompatible. (122) (123) Examples of bioinert materials used in orthopaedic and dental applications are alumina (Al2O3), (124) zirconia (ZrO2), (124) Ti-29Nb-13Ta- 4.6Zr titanium alloy (125) and Ti-6Al-4V titanium alloy. (125) Such bioinert materials are not usually applied to fabricate BTE scaffolds due to their lack of bioactivity, as discussed below.

Bioactivity in general is the materials ability to interact with the surrounding tissue to enhance the tissues response and the tissue/material bonding process (126) and specifically in the case of BTE materials the capacity to induce the formation of carbonated HA on its surface when exposed to simulated body fluid (SBF) or implanted in vivo is the is a well-documented process that is considered evidence of the bioactivity of a material used for BTE. (127) (128) (129) The process of HA formation on glass-ceramics can be described in a number of stages:-

I). The exchange of cations, for example sodium, from the surface of the glass-ceramic with hydrogen ions from the physiological solution. (130) (128) (125)

II). The loss of soluble silica in the form of Si(OH)4 from the glass-ceramic to the surrounding solution caused by the breaking of Si-O-Si bonds at the interface. (130) (125) III). Formation of silanol (Si-OH) groups on the surface of the glass-ceramic. (130) (125) (129)

IV). Polycondensation of the silanol groups to form a SiO2 – rich layer on the surface of the glass- ceramic. (130) (125) (129)

2+ 3- V). Ca and PO4 groups migrating to the surface of the glass-ceramic through the SiO2 – rich layer

to form a CaO-P2O5-rich film on top of the SiO2-rich layer. (130) (128) (125) (129)

- 2- - VI). The amorphous Ca-O-P2O5 film crystallizes through the incorporation of OH , CO3 , or F anions. (130) (128) (125) (129)

The final interaction class concerns whether a material is bioresorbable or biodegradable. Biodegradable materials breakdown due to structural and chemical degradation, dispersed in vivo and partially metabolised by the body. (131) (132) (133) Bioresorbable materials also breakdown due to structural and chemical degradation, are dispersed in vivo but are entirely removed from the body through metabolisation or through natural pathways. (131) (132) (133) Examples of these types of materials include calcium phosphates, (130) hydroxyapatite (130) (131) and polylactic-polyglycolic acid (PLGA) copolymers. (131) (132) (133) It should be noted that materials that are both bioactive and biodegradable/bioresorbable and the materials of choice for fabricating BTE scaffolds.

Ideally, once implanted the material must degrade at a suitable rate, a rate that matches the growth rate of the new tissue surrounding the implanted scaffold. (134)This means the rate of degradation should be controllable if the material is to be considered for a range of tissue engineering applications. (134) The degradation productions should be non-toxic and not cause an inflammatory response in the host tissue to keep the material biocompatible. (135)

An ideal implanted material also needs to facilitate cell attachment, proliferation and differentiation. (136) (137) The material must not produce any toxic by-products or cause an inflammatory response once implanted into the host body. (136) (138) The chemical make-up of a material is therefore extremely important in order to determine if the material will behave in an appropriate manner when implanted. These statements only hold true for a single material and not a

43 whole implantable device as these will be made up of a number of components. (87) (139)Whilst these materials on their own might be biocompatible in a biological system once assembled into an implant device they whole ensemble might prove not to be biocompatible. (87) (139) The material ideally should provide a biocompatible surface that can act as a framework on which new bone growth can occur (Osteoconductive). (140) (141) (142) The material ideally should be capable of inducing the differentiation of stem cells into osteogenic bone forming cells, osteoblasts and osteocytes, (Osteoinductive). (140) (141) (142) The material should ideally be able to induce new bone growth by eliciting intracellular and extracellular responses at the interface between the material and the host tissue by inducing the integration between the host bone tissue and implant in the form of a continuous layer at the interface (Osteoproductive). (143) (144) (145)

2.6.2 Structural properties required for an ideal bone tissue engineering material

The mechanical properties of the material are important as the material has to provide support to the surrounding tissue during its regeneration. (146) (147) The material needs to have mechanical properties (elastic modulus, compression strength and fracture toughness) that closely mimic those of the tissue surrounding the implant. (148) (149) Bone repair involves interrelated chemical, biological and mechanical processes occurring simultaneously, with many of the chemical and biological processes being dependant on certain mechanical conditions. (150) (151) (152) Producing a material with mechanical properties that mimic those of bone with reduce the loss of the surrounding bone tissue due to stress-shielding. (153) (154)

Other properties such as porosity, wettability and topography are more related to the type of implant produced, in this works case a scaffold, and will be discussed in the sections 2.3.7, 2.3.8 and 2.3.9 respectively.

2.7 The progression of materials in bone tissue engineering

Since bone consists largely of crystalline hydroxyapatite (HA) and other related calcium phosphates (CaP), the logical step is to develop bone substitute materials and tissue engineering scaffolds from these materials. Both HA and CaP exhibit good biocompatibility due to the fact that they are chemically similar to bone. However, this similarity is also a hindrance as HA is stable within the body and in tissue engineering the scaffold should be biodegradable. (155) This fact leads researchers to investigate other materials that could be used in BTE such as silicate glasses and glass-ceramics, for

44 example Bioglass®, and other biomaterials based on a combination of bioceramics and biodegradable polymers, forming composites.

2.7.1 Calcium phosphate and its derivatives

Calcium phosphate ceramics (CPC’s) have been studied extensively for more than 40 years as bone substitute materials (156) due to their many desirable properties which match those of an ideal BTE material, as described in section 2.6, such as biocompatibility and bioactivity. (157) CPC’s can be split into four groups:-

 Hydroxyapatite (HA)  Tricalcium phosphates (TCP’s)  Amorphous calcium phosphates (ACP’s)  Biphasic calcium phosphates (BCP’s)

Modifications to the processing methods of CPC’s have given rise to a range of CPC’s with distinct differences in chemical and physical properties leading to a family of suitable materials for use in BTE. (157)

2.7.1.1 Hydroxyapatite

HA is a stable form of calcium phosphate in the form of Ca10(PO4)6(OH)2, as described in section 2.6.1, and is naturally found in both bones and teeth as a major component accounting for between 60% to 65% of the mineral content. (158) HA exhibits some of the desirable properties for scaffolds described in section 2.6, being thus a suitable candidate for BTE. (159) (160) As described in section 2.6 a key indicator in establishing a materials suitability for BTE is through its bioactivity which is determined by the formation of HA surface layer making HA scaffolds bioactive and exceptionally biocompatible. (159) (160) Although it should be noted that HA requires close proximity to the target area and bone cells in order to induce osteoconduction. (156) Two significant areas where HA has a disadvantage when it comes to its use as a BTE scaffold is its comparatively low mechanical strength and its long degradation period. (159) (160) (161) Due to its properties HA is also considered to coat implants, (161) (162) as a drug delivery vehicle (162) and combined with polymers or composite scaffolds. In fact HA coatings on implants for hip and knee joints have become the gold standard in joint replacement therapy (161) due to the imparted bioactivity achieving a faster healing process and a quicker recovery. (162) It should also be noted that on-going work to improve and change the

45 properties of HA through ionic substitutions, including substituting the carbonate with a phosphate to improve the solubility and bioactivity, (163) substituting the fluoride for hydroxide to increase the stability (163) and incorporating silicon ions to enhance osteoconduction. (163)

2.7.1.2 Amorphous calcium phosphates

The vast majority of ACP’s are the first solid phase to appear when calcium and orthophosphate containing aqueous solutions are mixed. (164) (165) ACP’s have a wide range of compositions and therefore a range of applications including implant coating and injectable cements. (164) (165) They have many of the desirable properties listed in section 2.6 including a bioactive nature. In addition, ACP’s does not produce an immune or inflammatory response in the host tissue (166) and can be tailored to have the required degradation behaviour. However they are difficult to produce in any other form than a powder and is not mechanically strong. (164) (165) Most of the most recent work with ACP’s has concentrated on making composites to improve the mechanical properties whilst keeping the ACP’s desirable properties. (164) (165)

2.7.1.3 Tricalcium phosphates

TCP’s come in two forms, alpha-TCP which is formed at high temperatures and beta-TCP. The high temperatures required to produce alpha-TCP has made it rather unattractive in the research community and this has led to the community to focus on the beta-TCP as a BTE material.

TCP materials tend to be more soluble than the other CPC due to their composition and along with the other desirable properties such as bioactivity and biocompatibility makes this material suitable for BTE. In vitro and in vivo studies have shown that TCP’s exhibit suitable cell seeding capabilities (167) being biodegradable and comparable to the other CPC’s. (168)

2.7.1.4 Biphasic calcium phosphates

BCP’s are composed of a layer of a stable CPC, normally HA to provide stability, and a more soluble phase such as a TCP. (157) This type of CPC has the advantage that it can be tailored to contain layers that will complement each other whilst enhancing the desirable properties of this class of materials such as bioactivity. (157) Introducing a more soluble CPC to HA gives the material a

46 controllable and tailorable resorption/solubility (157) Specific properties, including those discussed in section 2.6, depend on the individual phases used. (163) These layered materials have shown better osteogenic cell expansion and differentiation compared to TCP’s and HA on their own. (169) (170)

2.7.2 Glass-ceramics in bone tissue engineering

Glass-ceramics are formed by controlling the crystallisation of suitable silicate glass compositions by heat treatment. (171) The resulting glass-ceramic has better mechanical properties than its parent glass and will exhibit other unique properties, meaning that glass-ceramics can find other applications than amorphous glasses. (171) There is currently a wide range of glass-ceramics used in the biomedical materials field, especially in bone regeneration, as the properties of glass-ceramics can be tailored for this purpose. (171) Among them, four key materials have been extensively researched; Apatite-wollastonite (A-W), Ceravital®, Bioverit® and 45S5 Bioglass® –derived glass-ceramics. These materials fulfil the requirements for improved bone substitute and tissue engineering scaffold materials. In this work the main material of interest is a glass-ceramic derived from thermal treatment (sintering) of 45S5 Bioglass®.

2.7.2.1 Bioactive glasses

A new class of materials, based on a silicate system, with high bioactivity and a remarkable ability to bond to both hard and soft connective tissue was discovered by Hench in 1969. (76) These bioactive glasses have very specific compositions containing at least 55wt% SiO2 in the system of SiO2-

Na2O-CaO-P2O5. They are categorised as class A biomaterials as they are osteoconductive, osteogenic and osteoproductive compared with class B biomaterials (such as HA) which are just osteoconductive. (76) During reactions on the bioactive glasses surface, soluble ions of Si, Ca, P and Na are being released and they induce extracellular and intracellular responses (172) and form HA in physiological solutions. (127) (128) (129) The first developed bioactive glass composition, which is known as 45S5

® Bioglass , contains (in weight %) 45% SiO2, 24.5% Na2O, 24.4% CaO and 6% P2O5 (3) and this is the composition that will be used as the base material in this work.

Bioactive glasses have found niche application as bone fillers/cements (173) and small non- load bearing implants (76) over the last forty years. It has taken this long to establish the materials credentials and carry out the necessary clinical trials required for its use in humans. (76) The advantages of these materials have made them highly suitable for many biomedical applications, PerioGlas®, used

47 in the treatment of periodontal disease, and NovaBone® in bone fillers. (76) These materials have also been used in middle ear bone replacement implants and more recently they are being considered a promising scaffold material for BTE. (174)

The main attribute of bioactive glass, including melt-derived and sol-gel glasses, is their bioactive behaviour and their ability to induce the formation of HA, as described previously. (76) (175) This ability to bond to bone once the material has been implanted is crucial as it will avoid loosening of the scaffold once implanted. (76) (175)

Bioactive glasses have extra properties, compared to HA, which makes them more suitable for BTE. These include properties such as providing support for enzyme activity, (176) vascularisation and angiogenesis, (177) and osteoblast adhesion (178) (179) as well as cell differentiation. (180) (181) An observation is that as the bioactive glass structure biodegrades, the dissolution products influence the osteoblast cycle, the genes that control osteogenesis and the production of growth factors. (172) (182)Bone mineralization and gene activation are influenced by the released silicon and calcium ions from the glass (or glass-ceramic) in suitable concentrations. (172)

Researchers have been working on the development of the ideal composition of bioactive glass in order to achieve the best trade-off between properties, since the bioactivity of the material is composition-dependant. (183) (184) With bioactive glasses being composition-dependant, the scientific community has studied a vast range of compositions by keeping the 6wt% P2O5 constant and changing the relative concentrations of the other components. (3) This approach is summarised by the ternary diagram shown in Figure 7. (3) (185) (186) (187) (122)

By keeping the composition of the bioactive glass within the bone bonding area, in Figure 7, the best trade-off between mechanical properties, bioactivity and the rate of resorption/degradation can be achieved. (3) This composition diagram in Figure 7 also gives the option of tailoring the materials for particular criteria for applications in a specific area of BTE.

Figure 7:- Ternary diagram of SiO2 – Na2O - CaO which shows where the optimal compositions for bioactive behaviour lie (3) (185) (186) (187) (122)

2.7.3 Other materials used in bone tissue engineering

In addition to polymer/bioceramic composites mentioned above and discussed in the literature, (121) other materials that have come to the front of BTE scaffold research include zirconia, alumina and titanium oxide. Zirconia has been used in solid dental implants for some time and work is being carried out to develop zirconia porous scaffolds for BTE. Zirconia demonstrates good biocompatibility and there is evidence that the host bone responds positively to the material, however zirconia lacks the desirable resorption and bioactive properties that are required for BTE (188) and therefore it should be coated with a bioactive material.

Titanium-based materials have been used in dental implants and exhibited some of the desirable properties described in section 2.6 with superior mechanical properties compared to the CPC’s and bioactive glasses described earlier. However titanium-based materials lack the crucial bioactivity which is key for BTE applications to encourage bone regeneration and the formation of HA. (189) Titanium- based scaffolds have been also investigated, however as is the case with zirconia, to increase the bioactivity bioactive glass coatings maybe required. (190)

Another possible candidate to be considered is mineralised collagen, which is also found in our skeletal system. It has been found to be more potent than TCP in terms of cell seeding efficacy, osteogenic marker generation, encouraging 3D cell alignment and cell infiltration. (191) Current efforts

49 concentrate on developing strategies to improve the mechanical properties (mainly compressive strength) of mineralised collagen, for example fabricating bioactive glass/collagen composites. (192) (193)

2.8 Bioactive glass processing

There are two established techniques for producing bioactive glass powders that can then be turned into functional scaffolds; Melt quenching (5) and sol-gel processing. (2) Both are discussed in this section.

2.8.1 Producing bioactive glasses by melt quenching

Melt quenching is the process of producing high quality bioactive glasses by essentially melting together the component materials (in the required quantities to achieve the desired composition), quenching the melt and grinding the melt into powder. (2) (194) (195) For 45S5 Bioglass® high purity analytical grade powders of SiO2, Na2CO3, CaCO3 and P2O5 are melted together in a platinum crucible at 1400oC for four hours then for a further five hours at 950oC to decarbonise the melt. (5) (194) (195) The melt is then either quenched in deionised water (for powder production) (5) (194) (195) or in graphite moulds for the production of rod or monoliths. (2) The melt quenched in deionised water, for powder production, is then milled with ethanol. (5) (194) (195) This quenching process, in either deionised water or graphite moulds, is quick enough to maintain the desired amorphous structure at this stage of scaffold production. (195) The final step is to anneal the powder at 480oC for eight hours to remove possible internal stress in the melted powder caused by the milling process. (195) Throughout this process the chemical composition is checked for impurities using inductively coupled plasma mass spectrometry. (194) (195)

This technique can be used to create bioactive glasses in a wide variety of compositions

including SiO2-P2O5-CaO-Na2O-CaF2, (196) CaO-B2O3-P2O5 (197) and CaO-MgO-SrO-SiO2-P2O5-

CaF2. (198)

The 45S5 Bioglass® powder used in this work, provided by Dr I Thompson (Kings College, London, UK), was produced using the melt quenching technique.

2.8.2 Producing bioactive glasses by sol-gel processing

The second processing technique is a wet chemical low temperature approach called the sol- gel method. This process allows the control of the materials homogeneity and final chemical composition. (175) (2) Generally a solution containing compositional precursors undergo polymer-type reactions at room temperature to form a gel, which is then dried, heated to form a glass before being milled into a powder. (2) Typical compositions produced using this technique have fewer components

than those produced by melt quenching and include 58S (58wt% SiO2, 33wt% CaO and 9wt% P2O5),

(199) 68S (68wt% SiO2, 23wt% CaO and 9wt% P2O5), (200) 77S (77wt% SiO2, 14wt% CaO and 9wt%

P2O5) (199) and 91S (58wt% SiO2 and 9wt% P2O5). (200)

This processing technique has only been recently applied to 45S5 Bioglass® compositions and other similar compositions containing Na2O, as the room temperature approach has little need for Na2O ability to reduce the melting point of the composition. (2) However due to the greater need for a material to be biodegradable, compositions including Na2O are now being processed using the sol-gel technique

as the addition of Na2O increases the solubility of the composition. (2)

The 45S5 Bioglass® is synthesised by a process of hydrolysis and polycondensation of an aqueous solution of tetraethyl orthosilicate (TEOS), triethyl phosphate (P(OEt)3), calcium nitrate tetrahydrate and sodium nitrate with HNO3 to hydrolyse the TEOL and P(OEt)3 to produce the sol. (201) The sol is cast into an air tight mould and a process of agglomeration and bonding of the nano- particles continues until a gel is formed. (201) This gel is aged which allows the bonds within the gel to strengthen whilst it shrinks. The gel is dried and stabilized before being milled into a powder that can then be used in scaffold production. (201)

Sol-gel processing does provide a great versatility in glass processing by simply changing the pH of the sol to produce powders, monoliths or nano-particles. (2)

2.9 The need for scaffolds in bone tissue engineering and additional properties

The focus of this project is the development and characterisation of scaffolds for BTE and so far the discussion has centred on the materials that are currently used and not on the scaffolds themselves. Any scaffold suitable for BTE needs to have the properties already discussed in section 2.6 but also requires other properties including 3D interconnected porosity, surface roughness and a wetting ability.

Scaffolds provide structural support to cells and new tissue by acting as a temporary extracellular matrix compared to bulk materials. (202) Essentially, scaffolds allow for new bone to grown and to infiltrate them and to gradually replace the scaffold as it degrades whilst still providing mechanical/structural stability. (203) Scaffolds should accurately mimic the bones extracellular matrix allowing for almost seamless replacement during regeneration (204) (205) and they can be tailored in size and shape to fit into different bone defects. (206) (207) Surface properties have been empirically linked to cell attachment and proliferation for decades meaning that, for example surface chemistry and surface roughness are important for the biological compatibility of the scaffold. (208)

2.9.1 Additional properties for a scaffold - Porosity

Porosity is a key parameter for a scaffold as it determines the applicability of the scaffold in a given tissue engineering area. Figure 8 demonstrates the reasons why porosity is a parameter that is important in tissue engineering showing the different functions related to a porous structure. (209)

Pore parameters such as orientation, interconnectivity and morphology have to be tailored to suit the needs of the tissue in question. Ideally a scaffold for BTE requires a pore size ranging from between 200 to 500µm and pores need to be interconnected. (207) This size range is required for cells and nutrients and other required elements to reach the target site and for waste material to be removed. (209) In addition, the scaffolds should be manufactured to maximise the porosity that should be achieved without compromising the mechanical properties of the scaffold. (210) (211) This pore size range (200-500µm) is required so that the cells and molecules involved in bone growth can infiltrate the scaffold. (210) (207) The porosity range of a human trabecular bone is between 50 and 90%, (65) (70) (74) (75) so any scaffold has to have porosity in this range in order to provide the seamless transition between implant and the surrounding tissue. (210) The porosity of a scaffold can be calculated using Equation 1.

Equation 1:-Porosity of an uncoated scaffold

Where ρscaffold and ρmaterial are the densities of the scaffold and the material that the scaffold is made of respectively, mscaffold is the mass of the scaffold and vscaffold is the total volume of the scaffold.

Figure 8:- Diagram showing the different functions that are directly related to the pore structure a scaffold. (209)

As stated in section 2.6 the mechanical competence of a biomaterial needs to be comparable to that of the tissue it is replacing and this is clearly applicable to a scaffold. Thus the porous nature of the scaffold has to have mechanical properties similar to those of bone. In the context of BTE the scaffolds need to have compressive strengths in the range 0.8-11MPa in order to successfully mimic trabecular bone (77)(33) whilst maintaining a suitable porosity. (210)

2.9.2 Additional properties for a scaffold – Wettability

Wetting is the ability of a liquid to maintain contact with a solid sample surface. The degree of wetting is determined by a force balance between adhesive and cohesive forces. (212) (213) This property is important in the field of biomaterials because it is significant to determine protein absorption and cell adhesion (214) when the materials come in contact with the biological environment.

An ideal contact angle for cell attachment should fall between 45 and 90 degrees which will imply that the material being tested has a hydrophilic surface as this is suitable for cell adhesion. (208) (214) Lower contact angles indicate a more hydrophilic nature compared to large contact angles which indicate that the surface has a more hydrophobic nature. (215) Figure 9 illustrates the relationship between the interfacial tensions acting on the sample surface and the liquid droplet as well as the relationship between the contact angle and the wetting capabilities of the sample’s surface. (212) (213)

Figure 9:-Illustration of the contact angle used to quantify the wettability of a sample along with the relationship of the wettability and the hydrophilic or hydrophobic nature of a sample (212) (213) (215)

Wetting plays an important role in cell adhesion, spreading and growth especially in the initial stages where the implant/scaffold is conditioned by proteins and the cells start to attach themselves to the scaffolds surface. (216) (217) (218) With a direct correlation between the contact angle and cell attachment it should also be noted that the cell attachment is also dependant on the stiffness of a substrate as cells required support to attach to a surface, for example rigid materials such as a glass- ceramic compared to less rigid materials such as polymers. (208) This justifies the use of glass-ceramic as the base material for the scaffolds produced in this work as they provide a more suitable support system for the surrounding tissue. (208)

2.9.3 Additional properties for a scaffold – Surface topography

The surface topography also has an effect on the suitability of a scaffold for use in BTE and it is directly related to the contact angle. (208) The change in surface roughness directly affects the contact angle and therefore the cell attachment of the surrounding tissue. (208) Osteoblasts are sensitive to the surface roughness and in some cases an increase in surface roughness has decreased healing times in vivo. (219) In the literature the general consensus is that the scaffolds surface needs to have a certain roughness in order to enhance cell adhesion and proliferation. (220) White light interferometry (221) can be used to investigate the surface topography of a material for BTE with two distinct values sets for

54 enhanced cell adhesion reported in the literature. (222) (223) (224) (225) The first set of RMS (root mean square) values corresponds to the macro-topography in the range of 300nm to 800nm (222) (223) (224) (225) and the second set of RMS values correspond to the micro-topography of the surface in the range of 1.1nm to 88nm (226) (227) (228) and both are important to enhance cell attachment and proliferation. In this work the focus is on the macro-topography of the surface of the scaffold as it is this roughness that supports the initial cell attachment. (222) (223) (224) (225)

2.10 Current scaffolds in bone tissue engineering

This section charts the progress that has been made in fabricating scaffolds for BTE using materials described in section 2.7 with the properties described in sections 2.6 and 2.9.

2.10.1 Glass-ceramic scaffolds for bone tissue engineering

As described in section 2.7, the logical step from the naturally occurring HA and calcium phosphates is to use synthetic glass-ceramic bioactive glasses which are then fabricated into three- dimensional scaffold ready for implantation, however due to their disadvantages including their long degradation time (described in section 2.7) they aren’t being used in this work. Several key glass- ceramics have come to the for front of the current development of glass-ceramic scaffolds for BTE including apatite-wollastonite (A-W), 13-93 bioactive glass and the original bioactive glass 45S5 Bioglass® (BBG).

A-W derived glass-ceramic scaffolds have been manufactured using a variety of fabrication techniques including adding a plastic porosifer, (229) foam impregnation, (230) (231) wood powder burnt out during sintering, (232) three dimension printing (233) and laser sintering. (234) (235) All these scaffolds have produced evidence that A-W derived glass-ceramic scaffolds show the potential to be used in BTE showing suitable porosities, precipitation of HA after immersion in SBF, in vitro assessments have shown cell attachment and growth without cytotoxic effects and mechanical properties similar to those observed for trabecular bone. However these properties have not been found in one scaffold type making it difficult to identify the most suitable scaffold production method for A- W glasses. Further research is required to improve this potentially important scaffold type before it can be considered for implantation in particular improving its resorption/degradation properties. (236)

13-93 bioactive glass-ceramic scaffolds are a more recent addition to the BTE scaffold family, with a range of fabricated techniques being utilized to produce porous three dimensional constructs including laser sintering, (237) freeze extrusion fabrication (238) and foam replication. (239) Results for this scaffold material have showed promising results with growth and differentiation of osteoblasts, (240) suitable porosity, (239) suitable mechanical properties, (239) bioactivity (241) and variable degradation rates. (241) However more work is required on these scaffolds before they can be implanted as their affects in vitro, let alone in in vivo, have not been fully documented.

2.10.2 45S5 Bioglass®- derived glass-ceramic scaffolds for bone tissue engineering

One very important aspect of Bioglass® -derived glass-ceramics is its transformation from the

Na2Ca2Si2O9 crystalline phase to an amorphous silicate matrix containing HA crystallites after immersion in simulated body fluid for 28 days. (121) The kinetic of HA formation can be tailored through the fabrication process by changing the sintering conditions. Fine crystals of Na2Ca2Si2O9 grow and nearly complete densification of the scaffolds struts occur when scaffolds are sintered at approximately 1000oC for a couple of hours. These conditions give the scaffold the reasonable compressive and flexural strength, (4) however further work must be carried out to further improve the structural integrity of these scaffolds, for which this project is expected to contribute. The ideal end point with any material that is being tailored for BTE is to provide in situ support, to the surrounding tissue, whilst remaining bioactive. Then in the later stages of the tissue regeneration process the materials can biodegrade at a set rate to provide continued support during this process. (242) To a certain extent this has been achieved with the BBG scaffolds (Chen et al (4)) and because of this they remain a very active research topic, considering the promising properties of these scaffolds in the scope of tissue engineering. There are a number of fabrication techniques including foam replication, (4) freeze casting, (243) sol-gel processing (244) and solid free form fabrication (which includes robocasting, (245) rapid prototyping (246) and 3D printing) and some of these will be discussed in section 2.11.

2.10.3 Other bone tissue engineering scaffolds

Two of the most researched scaffold types are scaffolds made with HA and scaffolds made with calcium phosphates (CaP). HA scaffolds have been produced using a variety of techniques including freeze casting, (247) solid form fabrication, (248) foam replication (249) and sol-gel production. (250) These have produced scaffolds with properties which are sort after in the hunt for a perfect BTE

56 scaffold, including porosity, (251) mechanical strength, (252) cell attachment, (251) lack of a cytotoxic response (251) and bone growth. (248) However, as mentioned in section 2.6, the lack of a degradation of the scaffold does not make it suitable for regeneration therapy; (253) although work is being down to improve the degradation by combining it with more reactive materials to create a composite that is more suitable. (254) (255)

CaP scaffolds of various compositions (such as tricalcium phosphate (beta –TCP) (256) and biphasic calcium phosphate (257)) have been fabricated using a number of techniques such as foam replication, (258) sol-gel processing, (259) solid free-form fabrication (259) and freeze casting. (260) As with the HA scaffolds, described above, these scaffolds have the potential to be used in bone replacement/regeneration therapy as they are observed to have desirable properties such as porosity, (261) mechanical strength, (261) bioactivity (262) (263) and suitable surface topography, (264) lack of a cytotoxic effect (265)and encouraging cell attachment and differentiation. (263) However CaP scaffolds, as is the case with the HA scaffolds, are seen as a stepping stone to composite scaffolds with improved properties (266) (267) and a more suitable degradation/release rate. (253)

Current research has shown that composite scaffolds are the way forward for BTE by taking the materials with the most promise and combining to produce a scaffold which is closer to mimicking bone.

2.11 Scaffold production

There are a variety of techniques being utilized to produce the required three dimensional porous scaffolds, but there are four main types have been identified in the literature as the techniques to use for BTE scaffolds:-

 Foam replication  Sol-gel processing  Solid free-form fabrication  Freeze casting

In this work the BBG scaffolds were produced using the foam replication technique as the BBG scaffold needs to a constant as other variables are being changed in this work and the foam replication technique, according to the literature (4) gives the most consistent scaffold form compared to the others method discussed in this section.

2.11.1 Foam Replication technique to produce scaffolds for bone tissue engineering

The foam replication technique, first described in Chen et al, (4) is a process of dipping a foam template, which range from a polyurethane foam (4) to wood or coral, (268) into a ceramic suspension containing the glass-ceramic, in this case melt-derived 45S5 Bioglass® powder, and a binding agent. (4) (268) The “green body” is then dried and sintered according to the method described in section 4.2.1. The replication technique produces highly porous three-dimensional scaffolds which are easily reproducible, easy to tailor the size and shape of the scaffold, with the added bonus of being mechanically stable, bioactive and biodegradable. (4) (268) When the sacrificial template used in the foam replication process is polyurethane (PU), as used in this work, this produces a highly interconnected pore network due to the nature of the PU foam. (269) (5) (270) In this work the PU foam has 45 pores per inch and a pore size of 350 to 800µm, which led to base 45S5 Bioglass® scaffolds with a pore size of approximately 500µm. (269) (5) (270) (271) The intrinsic permeability for these 45S5 Bioglass® -derived scaffolds has been reported to be 1.96x10-9m2 which is in the range of the values reported for human trabecular bone. (270)

However due to the extremely high porosity, upwards of 90% in some cases, the mechanical strength of the scaffolds is lower than observed for some of the other production methods but this can be improved through the use of polymer coatings. (4) (268) This is shown in Figure 10B) with data obtained from the literature, and shown in appendix A, it can be seen that majority of the scaffolds produced by this technique have a compressional strength similar to that observed for trabecular bone (between 2 and 10 MPa) with porosities of around 75%. It is also observed that the presence of a polymer coating improves the compressional strength whilst not compromising on the porosity.

2.11.2 Sol-gel processing to produce scaffolds for bone tissue engineering

This fabrication technique typically involves the foaming of the gel, as described in section 2.8.2, with the aid of a surfactant. (268) This is then followed by a condensation and gelation process with an ageing process, similar to the one described in section 2.8.2, to strengthen the structure. (268) Finally the liquid by products are removed allowing for the structure to be sintered. (268) This fabrication method produces scaffolds which are highly porous, with both macro-pores (10-500µm) and meso-pores (2-50nm) observed, (268) that are highly bioactive but these sol-gel derived scaffolds suffer from a low mechanical strength. (268) The low mechanical strength, which is also observed with the foam replicated scaffolds and shown in Figure 10C), can be improved through the use of polymer coatings and by producing composite scaffolds. Foam replication is preferred over sol-gel processing

® because of the composition of the base 45S5 Bioglass that is used in this work, , which contains Na2O and compositions with Na2O have only recently been used to form scaffolds using sol-gel processing, as described in section 2.8.2. (2)

Figure 10:-Evaluation of compressive strength against porosity for scaffolds fabricated using A) Freeze casting, B) Foam replication, C) Sol-gel and D) solid free-form fabrication with data for glass-ceramics only, scaffolds coated in a polymer and MSIE samples. Data is shown in appendix A

2.11.3 Freeze drying to produce scaffolds for bone tissue engineering

Freeze casting fabrication (or freeze drying) involves a colloidally stable suspension containing the glass-ceramic which is then rapidly frozen in a mould. (268) The sublimation of the suspensions solvent, at freezing temperatures in a vacuum, leaves behind a porous construct, which after drying is

59 sintered to leave a porous scaffold with improved mechanical strength. (268) However this technique does not produce scaffolds, using 45S5 Bioglass®, which are porous enough to allow cell proliferation and attachment at the present time. However, as shown in Figure 10C), some scaffolds fabricated in this manner have both a suitable porosity (above 75%) and a compressional strength similar to that observed for trabecular bone. Progress is being made, through the use of different suspension solvents, to increase the porosity without losing the increased mechanical strength. (268) This is a technique that once it has established a protocol for 45S5 Bioglass® which balances compressional strength and porosity can be used to produce scaffolds for BTE in the future.

2.11.4 Solid free-form fabricated scaffolds for bone tissue engineering

Solid free-forming fabrication is an umbrella term used for a number of processing techniques including 3D printing, (233) robocasting, (245) fused deposition modelling, (272) ink jet printing, (273) stereolithography, (274) selective laser sintering (234) (235) and rapid prototyping. (246) This technique broadly involves producing scaffolds layer by layer through the use of a paste containing the glass-ceramic. The design of the scaffold, including its shape, size, surface topography, pore structure and pore orientation is controlled by computer aided design rather than traditional moulds (as in sol-gel processing) and foams (as in foam replication). In this way the design of the scaffold is modelled extensively using all the current knowledge of the most suitable and desirable properties before being produced.

This fabrication technique is promising in producing scaffolds that have a desirable balance between porosity and compressional strength, as shown in Figure 10D) but as it is in the early stages of being explored, scaffolds produced using this group of techniques using 45S5 Bioglass® have only just produced the first generation of scaffolds. (275) (276) Since this work is looking at varying the composition through MSIE and polymer coatings an established fabrication technique is required so the scaffolds structure is constant throughout the work. It should be noted that this branch of scaffold fabrication has a lot of potential as these would be easy to reproduce on a large scale because the fabrication is completed by machine rather than by hand, as in the case of foam replication, and the scaffolds can easily be tailored to any size and shape.

2.12 Multifunctional scaffolds for bone tissue engineering

The BBG scaffolds, used in this work, have been developed/manufactured for BTE to have a clear set of functions, which have been described in the previous sections, to encourage and support new bone growth. This scaffold type needs to be developed further incorporating additional functions in order to be a viable option in bone tissue replacement/regeneration therapies. The approach taken in this work is to improve the BBG scaffolds by improving the weak points of this scaffold type, for example the mechanical strength, without disrupting the desirable ones such as the high bioactivity through the use of functional additives which will give the scaffold additional properties, such as antibacterial or angiogenic properties whilst improving pre-existing properties.

One issue with BBG scaffolds is that their mechanical properties, in particular compression strength, does not completely match those of bone making them unsuitable for implantation, as demonstrated for the four main fabrication techniques in Figure 10. One route that researchers have been exploring to improve these mechanical properties is introducing polymers into the scaffold either as a coating or incorporating the polymer into scaffold structure. (121) Researchers have found that these polymer/bioceramic composite scaffolds have improved the scaffolds mechanical properties (7) (121) without compromising on properties such as porosity or biological surface reactivity of the scaffold. Figure 10 shows that the incorporation of polymers into the scaffolds construction improves the compressional strength whilst maintaining the desirable high porosity. The difference in mechanical properties between bone, BBG scaffolds and these polymer/Bioglass® composite scaffolds is illustrated in both Figure 10 and Figure 11. (121) In this project, a number of polymer/Bioglass® composite scaffolds will be investigated to improve the mechanical competence of the scaffolds. From the point of view of making the scaffold have a variety of functions the addition of a polymer coating is advantageous as these polymer coatings have the potential to act as drug delivery vehicles for a range of additives such as antibacterial ions (113)and antibiotics, (224) angiogenic elements, (277) growth factors (278) (279) and vitamins. (280)

With post-operative infections and implant rejection a concern, the ideal scaffold implant would be one that incorporated an antibacterial element to help to prevent these issues and thus reducing the post-operative care and improve the patient’s recovery time. One of the main reasons behind complications in the recovery of a patient after having an implant is bacterial infection. (281) In order to reduce the likelihood of infection an antibacterial element needs to be introduced into the implant and this is one of the main objectives of this work.

Figure 11:- Comparison of elastic modulus against compressive strength for a variety of biomaterials including composites, polymers, bone and bio-ceramics (121)

2.13 Antibacterial scaffolds

One method to reduce the risk of post-operative infection is to incorporate an anti-bacterial element into the implants construction during fabrication rather than over use antibiotics after implantation. In this work the aim is to incorporate an antibacterial element into the scaffold in an effort to produce a scaffold that will help prevent and combat post-operative infection.

As previously stated it is imperative to avoid surgical site infection (SSI) in the tissue after the scaffold has been implanted as infections can be incredibly persistent and difficult to treat without eventually removing the implant. (282) The most common cause of bacterial infection in bone replacement therapy is Staphylococcus aureus (S.aureus) (283) which can be treated with a number of antibiotics including gentamicin (284) and vancomycin. (283) The issue that has arisen in recent years, due to the overuse of antibiotics, is that bacteria have become resistant. (285) With methicillin-resistant S. aureus strains (MRSA) (286) and vancomycin-resistant S. aureus (VRSA) (287) becoming more problematic an alternative therapy is required.

The idea and exploration of metal ions as antibacterial agents has been around for some time with many metals showing promise including silver, (288) (113) copper (289) and zinc. (290) This work is concentrating on incorporating silver and copper into the BBG scaffold in order to produce an antibacterial effect that will reduce the need to use antibiotics after implantation. The release rate of the therapeutic metal ion needs to be such that it is released over a long period (not in a burst) to provide

62 antibacterial coverage that prevents SSI, allows the implant to settle into the implantation site and prevents any other infections developing near the implantation site which ranges between ten to thirty days. (291)

Also it should be noted that the advantage of developing these antibacterial scaffolds would be that they would provide a better localised targeted antibacterial coverage compared to antibiotics delivered by traditional means, such as oral and intravenous. This will allow infections/bacteria to be swiftly dealt with at the source and keeps the surrounding area bacteria/infection free.

2.13.1 Antibacterial properties of 45S5 Bioglass® -derived glass-ceramic scaffolds

Although 45S5 Bioglass® was developed in the late 1960’s, (76) more enthansis was placed on the materials bioactivity and other desirable properties. Now with SSI’s and resistant bacteria becoming more of an issue in bone replacement/regeneration therapy the antibacterial properties are just as important as its other desirable properties.

It has been shown that 45S5 Bioglass® has an antibacterial effect against a wide range of bacteria including both gram-negative and gram-positive bacteria. With 45S5 Bioglass® causing a higher antibacterial effect on gram-negative bacteria, such as E.coli at concentrations between 10 and 50mg/ml. (292) Concentrations above 50mg/ml showed an equal antibacterial performace for both gram-negative and gram-positive bacteria. The difference in antibacterial effect, for the lower concentrations, can be attributed to the difference in the bacteria’s surface properties (the difference in the make up of the bacteria’s cell wall), which are discussed in section 2.18. (292) (293)

A solid theory as to the antibacterial mechanism against bacteria (of either class) has yet to be established. Although there are a number of factors that are being considered as possible contributors to the antibacterial effect of 45S5 Bioglass® including the antibacterial effect of the pH increase casued by the ion exchange at the surface of the 45S5 Bioglass® once it is immersed in a physiological fluid, such as SBF or cell culture media. (292) Another factor to consider that the cell walls are destroyed/damaged by the formation of crystalline structures during the precipatation of HA on the surface of the 45S5 Bioglass® which would lead to cell death. (292)

It should also be noted that due to advances in sterilisation techniques and procedures used in the implantation surgery the antibacterial effect required at the point of implantation will be lower than the effect required after 24 hours for example or in most of the in vitro models used in the literature currently. (294)

2.13.2 Antibacterial properties of silver

Silver has been on the radar of scientists for a long time as an antibacterial element (6) (295) and it has been shown to be a good candidate for the treatment of bone infections (288)with an increase in the research being carried out using silver as an antibacterial element. (288) (113)

Previous research on the exposure of bacteria to silver has been carried out and the following has been observed:-

 The cytoplasm membrane surrounding the cell wall shrinks or detaches (296) (297)  Electron-light region appears in the centre of the bacteria containing the DNA in a condensed form (296)  Electron-dense region around the electron-light region which contains deposited proteins and silver ions (296)  Disrupts ATP production (297) (295)  If the concentration is high enough the cell wall collapses causing bacteria death (296) (295)  Prevents DNA replication meaning that the bacteria cant replicate (296) (297) (295)

There are many theories as to the mechanism of how the silver ions cause an antibacterial effect including attributing it to their small size and therefore their high surface to volume ratio. (295) This allows the metal ions to interact closely with the cytoplasm membranes of the bacteria causing the inactivation of proteins through interactions with thiol groups. (295) (296) This close interaction causes the silver ions to be incorporated into the bacteria’s cell membrane causing the intercellular substances to leak out of the cell leading to cells death. (295) By passing through into the bacteria’s interior the silver disrupts ATP production and DNA replication leading to a bacteriostatic effect as the bacteria can no longer replicate. (296) (297) (295) The concentration of silver used needs to be carefully controlled as high levels of silver causes human cell toxicity. (298) (297) (299) This toxic effect also occurs at a high release rate of silver ions, (300) (301) so when incorporating silver into the scaffold construct the release rate and the concentration both need to be considered. (6) (113)

2.13.3 Copper ions as an antibacterial agent

Copper is also considered to be an antibacterial element, (295) but it also has other properties which are of interest in BTE, such as causing an angiogenic effect, (10) (277) making the addition of copper an ideal choice for a multifunctional BTE scaffold.

Copper ions cause their antibacterial effect in a similar way that silver ions do, in that the positive copper ions are attracted to the negatively charged cell wall, punching holes in the cell wall causing the intercellular substances to leak out leading to cell death. (302) As with the mechanism theorised for the antibacterial effect of silver, more research is required to ascertain a solid theory as to how the copper ions are causing this antibacterial effect. (295) (302) From research carried out with copper ions, extrapolated from work with silver ions, it is clear that the antibacterial effect is dependent on the concentration of metal ions released from the implant and the rate of release and in turn the effect is dependent on the surface concentration of the therapeutic metal ions. It has been found that the higher the surface concentration of the metal ions the greater the antibacterial effect. (6) (281) However the higher the concentration the more likely the surrounding cells are going to succumb to cell death or changes due to toxicity as a result of the presence of the metal ions. (302)

2.14 Angiogenic scaffolds

Another function that could be imparted into the BBG scaffold is the ability to help the surrounding tissue to spontaneously form blood vessels in otherwise to make the scaffolds angiogenic as this would be an appropriate function for a BTE scaffold to have.

In order to facilitate bone regeneration a good vascular network (303) (304) is required to transport the cells (303) and growth factors (305) (306) to the implantation site and to carry away waste away from the implantation site. (307) (308) Angiogenesis is essential in bone regeneration/formation as it regulates bone remodelling, prevents osteoblast cell death, and stimulates blood vessel development and osteogenesis. (277) Bone remodelling and blood vessel development can be triggered by growth factors, such as bone morphogenetic proteins (BMP) for bone remodelling and for angiogenesis vascular endothelial growth factors (VEGF) and fibroblast growth factor-2 (FGF-2). (277) If a material/implant can encourage the vascular network to form faster by being angiogenic whilst maintaining other desirable properties, such as bioactivity, this would produce a suitable implant for BTE. Producing angiogenic scaffolds for BTE is an emerging topic in this field and this project will contribute to producing an angiogenic glass-ceramic scaffold.

The research that has been carried out on the mechanisms that stimulate angiogenesis have found two distinct branches being mechanical stimulation (309) (310) and chemical stimulation. The mechanical stimulation shows that mechanical stress on existent capillaries causes an angiogenic effect. The chemical stimulation of angiogenesis is performed by a number of angiogenic proteins/growth factors such as VEGF, (311) FGF, (312) Ang1, (313) Ang2, (313) TGF-beta (314) and endoglin. (315)

This work, through the use of copper, will be looking to stimulate these growth factors in order to speed up the angiogenesis process.

2.14.1 Angiogenic properties of 45S5 Bioglass®

It has become apparent recently that 45S5 Bioglass® -derived glass-ceramic without any additives exhibits angiogenic properties. It has been observed that 45S5 Bioglass® increases the secretion of VEGF and FGF in vitro (316) (317)as well as increasing the proliferation of micro-vascular endothelial cells which form the base for the new blood vessels in vitro (316) (317) and for formation of endothelial tubules in vitro. (317) It has been shown in vivo that bloods are infiltrating the 45S5 Bioglass® -derived glass-ceramic scaffold to form a fully functional vascular network just three weeks after implantation. (318) Investigation into the angiogenic effects of BBG scaffolds are still in the early stages but are showing promise and in combination with other angiogenic additives, composite scaffolds containing 45S5 Bioglass® would increase the angiogenesis process. (317)

2.14.2 Copper ions as an angiogenic agent

One of the other interesting desirable properties that copper ions have, for tissue engineering applications, is the ability to enhance angiogenesis. (10) (277) As stated previously, growth factors are required to trigger angiogenesis, however incorporating growth factors into a scaffold for use in BTE is problematic as they are require large concentrations to be effective and the bioactivity period that these growth factors is relatively short as they can be unstable. (319) (320) (321)Researchers trying to overcome these problems have stabilised the growth factors in a number of ways such as combing them with polymers, embedding them in microspheres and incorporating them directly into a scaffold. (277) None of these methods are ideal as high doses of the growth factors are required which is not cost effective, safe and there is the problem that they have to be properly preserved until needed.

Researchers are searching for alternative angiogenic agents and this is where copper ions has excelled as they are easier to use as they are less expensive, they do not lose their bioactivity, are stable and they are easier to manipulate. (277) Instead of using the Bioglass® scaffolds as vessels for growth factors, they can be used to delivery copper ions which then go on to stimulate the production of angiogenic growth factors such as VEGF, (9) (289) bFBF (322) and angiogenin (322) and the production of transcription factors important in activating the growth factors such as HIF-1 (289) (322) and NF-kB. (322)

Research has shown that the presence of copper stimulates both the proliferation and migration of essential endothelial cells. (323) (324) (322) It has also been shown to enhance the infiltration of a vascular network into an implanted scaffold. (325) (322) Certain bio-molecules such as heparin, glycl- histidyl-lysine (GHK) and ceruloplasmin that when bound to copper were found to enhance the formation of blood vessels. (322) Initial research which combined CaP scaffolds with copper has shown that in vivo, micro-vessels where formed propagating along the pores of the scaffold. (277)

The challenge that faces researchers, and indeed this work, is controlling the release rate of the copper into the human body as it is cytotoxic at high prolonged or burst doses (326) but if this can be controlled then the addition of copper to a BTE scaffold would produce a composite scaffold capable of inducing angiogenesis. Another issue that has come to light, which has been investigated in terms of intrauterine devices (IUDs), is the problem of a burst release of the copper causing abnormal bleeding and pain in the surrounding tissue. (327)

2.15 Molten salt ion exchange (MSIE)

MSIE is a technique used to introduce metal ions into the crystalline structure of glasses and glass-ceramics. It is a technique that has been used for decades in optics to introduce metal ions such as Rb+, (328) K+, (329) Ag+ (329) (330) and Cu+ (330) into a range of glass optical devices such as wave guides (328) and optical splitters. (329) MSIE has been used to change the refractive index, (331) to chemically strengthen glass (332) and to increase the electrical conductivity of the material. (333) Research has shown that the change in the properties of the material due to MSIE is due to the glass undergoing local re-structuring at the surface at temperatures which are far lower than the glass transition temperature. (334)

The MSIE technique is an established method used to chemically strengthen glass by exchanging a smaller ion within the glass with a larger ion from the molten salt bath. (332) (335) A common example is the exchange of sodium from a soda lime glass with potassium in a potassium nitrate molten salt bath, as demonstrated in Figure 12. (336) (337)

In the present work, the main purpose of the MSIE process was to incorporate specific ions, e.g. silver and copper, into bioactive glass, for enhanced functionality, e.g. antibacterial effect. It is also possible that the MSIE process could contribute to increasing the mechanical properties of the structure (strengthening effect) but this was not the main goal of the MSIE process in this research.

Essentially strengthening by the use of ion exchange works by the larger ion from the salt bath, silver and copper ions in this work, exchanging with the smaller ion from the glass, the sodium ion in

67 this case, and the larger ions force themselves into the space left behind by the smaller ions, essentially “stuffing” the rigid structure at surface of the glass. (332) (335) (336) (337) This stuffing effect leads to a high compression on the surface of the glass and a corresponding balancing tension in the interior of the glass. (332) (335) (336) (337)

Figure 12:- Schematic representation of the chemical strengthening of glass through molten salt ion exchange showing the exchange process between potassium in the melt and sodium in the glass

In the scope of this work a technique is sort to introduce the chosen therapeutic metal ions into the surface of the BBG scaffolds where they will be released soon after implantation. Evidence has shown that MSIE introduces the metal ions only into the surface of the glass if the immersion period is restricted. (330)

Using MSIE to introduce both silver and copper into these optical devices is extremely common and the process is well documented and characterised. (329) (330) (338)However using MSIE to introduce functional therapeutic metal ions (in this work copper and silver have been selected, as discussed in sections 2.13 and 2.14) has not been extensively investigated. (6) (113) With Di Nunzio et al (6) introducing silver into a silicate glass for antibacterial applications and Lopes et al (339) introducing additional calcium into 45S5 Bioglass® -derived glass-ceramic structures to modulate the desirable properties of the 45S5 Bioglass®.

2.15.1 Incorporating Silver ions into 45S5 Bioglass® -derived glass-ceramic scaffolds using molten salt ion exchange

Work has already been carried out using MSIE to introduce silver into silica based glass- ceramics (340) and phosphate based glass-ceramics (341) through the sol-gel technique which was described in section 2.8.2. There are several ways of introducing metal ions into the scaffold, including the traditional melt quenching technique, sol-gel processing and molten salt ion-exchange. Both the sol- gel method and the melt quenching methods are useful if the silver is required throughout the whole structure of the scaffold, (340) however the idea with these scaffolds is that the silver acts as an initial antibacterial element whilst the risk of infection is at its highest. This approach requires that silver ions are only present on the surface of the scaffold; the obvious choice would be to use the molten salt ion- exchange technique. (6)

MSIE has been investigated in the past for use in introducing melt ions into glass waveguides for optical applications. (342) (329) Di Nunzio et al (6) shows that using the MSIE method silver ions can be introduced into bioactive glasses, specifically SiO2-CaO-Na2O. MSIE allows the fabrication of silver-doped scaffolds that can easily be reproduced, with constant diffusion profiles and silver content. This method offers a simple and low cost way to introduce silver into a pre-fabricated scaffold after they have been sintered. (6) The mechanism of silver MSIE, represented in the Equation 2, from Di Nunzio et al (6) was between sodium ions in the glass and silver ions in the salt bath. (6)

Equation 2:- Mechanism of silver-sodium molten salt ion exchange in a glass

Since this project used BBG scaffolds with a composition of SiO2-Na2O-CaO-P2O5, (3) a similar MSIE technique, like those described by Di Nunzio et al, (6) can be used. The molten salt bath needs to contain both silver nitrate and sodium nitrate salts, with the sodium nitrate salt being used to regulate the amount of silver being exchanged into the material; this exchange process is described by Equation 3.

Equation 3:- Molten salt ion exchange process to introduce silver into a pre-fabricated BBG scaffold

2.15.2 Incorporating Copper ions into 45S5 Bioglass® -derived glass-ceramic scaffolds using molten salt ion exchange

As with the silver ions described in section 2.15.1, copper can be introduced into the pre- fabricated scaffolds via the molten salt ion-exchange method and this has been used to introduce copper into waveguides for the past 20 to 30 years. (342) No work has been done, so far, to introduce copper ions into 45S5 Bioglass® scaffolds using the MSIE technique, making the work with copper in this project novel. Oven et al (342) describes the process of copper/sodium MSIE in soda lime glass for wave guides and combined with the silver MSIE processes described in Di Nunzio et al (6) and Newby et al (113) the following exchange equations have been deduced, shown in Equation 4 and Equation 5. Equation 4 shows that the MSIE process occurs between the copper in the salt bath and the sodium in the scaffold and Equation 5 shows the chemical exchange occurring in the salt bath once the BBG scaffold is immersed in the salt bath.

Equation 4:- Mechanism of copper-sodium molten salt ion exchange in a glass

Equation 5:- Molten salt ion exchange process to introduce copper into a pre-fabricated BBG scaffold

The technique is described in the methodology section (section 4.2), with specific details pertaining to the use of MSIE to introduce copper into pre-fabricated BBG scaffolds. The sodium nitrate in the salt bath acts as a regulator to control the amount of copper infiltrating the BBG scaffold’s structure. Controlling the amount of copper that is infiltrating the scaffold is paramount as toxicity issues need to be avoided. Burst release can be avoided if the copper is loaded into the scaffold in small but effective dose or loaded into a polymer carrier. (327)

2.15.3 Dual ion incorporation into 45S5 Bioglass® -derived glass-ceramic scaffolds using molten salt ion exchange

The previous section, sections 2.15.1 and 2.15.2, have described the process behind silver/sodium and copper/sodium MSIE. The last ion-doped scaffold type that is being investigated in this work is to introduce both copper and silver through MSIE. This is a novel application of the MSIE technique especially its use in biomaterials.

Equation 6 shows that the MSIE process occurs between the silver/copper in the salt bath and the sodium in the scaffold and Equation 7 shows the chemical exchange occurring in the salt bath once the 45S5 Bioglass® -derived scaffold is immersed in the salt bath.

Equation 6:- Mechanism of copper/silver-sodium molten salt ion exchange in a glass

Equation 7:- Molten salt ion exchange process to introduce copper and silver into a pre-fabricated BBG scaffold

This technique is described in more detail, with specific details pertaining to using MSIE to introduce silver into pre-fabricated BBG scaffolds, in the methodology section (section 4.2). The sodium nitrate in the molten salt bath acts as a regulator to control the amount of copper/silver infiltrating the BBG scaffold’s structure. As described in previous sections the release mechanism of the silver and copper needs to be controlled in order to prevent a cytotoxic effect occurring once the ion-doped scaffold is introduced to cells.

2.16 Polymer coated scaffolds for bone tissue engineering

As described in section 2.12, one solution to the problem of lack of mechanical strength of the BBG scaffolds is to introduce a polymer into the structure by either coating the scaffold or infiltrating the polymer into the materials structure. (7) (343) Figure 10 shows that previous research has shown regardless of which fabrication method was used to produce the bio-ceramic scaffolds the addition of a polymer into the scaffolds structure improved the compressional competence of the scaffolds. Complete polymer infiltration is not relevant in this work as one of the aims of this study is to produce a highly porous scaffold that should be infiltrated by cells during the regeneration process. This work will investigate the use of polymers to cover the surface of the scaffold struts forming a homogenous layer, but not blocking the macropores.

The fracture toughness of a material is an expression of a materials behaviour, whilst cracked to resist fracturing and crack propagation. (344) Ceramics and glass-ceramics have a low fracture toughness as they are brittle materials and cracks will extend when more force is applied but in their bulk form have a high compressive strength and stiffness. (344) Polymers, however, have a higher fracture toughness due to their elastic nature, ductility and plasticity but have a low compressive strength. (344) By combining the two in a composite structure by coating the highly porous glass-

71 ceramic scaffold with a polymer it is hoped that the fracture toughness will improve due to the polymer coating whilst maintaining a suitable compressive strength. (344) (345)

It is hypothesised that by coating a structure with a polymer it will fill micro-cracks, bridge cracks and forms a polymer-ceramic network, behaving in a way similar to collagen fibres and HA crystals in bone. (65) (77) (344) (346) Collagen fibres help to bridge cracks in bone to increase the fracture strength and toughness. (7) (344) (346)

The polymer acts as a toughening mechanism for the composite in the following ways:-

 Crack bridging  Crack deflection  Crack tip shielding  Pulling out

The polymer acts as a bridge across any micro-cracks in the surface of the scaffold and acts to closure stress to counter act the applied force. (345) Crack deflection occurs when the path of the crack is deflected by the polymer resulting in in more energy being dissipated for a reduced crack propagation. (345) Crack tip shielding occurs at the leading point of a crack and in this case the polymer prevents further propagation of the crack tip. (345) Pulling out occurs when the polymer is pulled out over a crack or defect. (345) All of these mechanisms in combination would protect the glass-ceramic scaffold underneath the polymer coating whilst preventing further propagation of the micro-cracks formed during the sintering process. (345)

Figure 13:- Fracture toughening mechanisms observed for polymer coated ceramics in the literature showing A) crack bridging, B) crack deflection, C) crack tip shielding and D) pull-out. (345)

In selecting a suitable polymer to use as a coating material, the way in which the polymer covers the scaffold and infiltrates the micro-cracks is important as it is this infiltration that improves the mechanical competence of the scaffold. (7) (110) This will be more affective in employing the toughening mechanisms described above compared to if the polymer just forms a layer on top of the scaffold. (344) (345)

Ideally any polymer that is used in this work should infiltrate into the micro-cracks, produced in the sintering process, so the polymer can have a dual purpose, one to improve the mechanical properties of the scaffold and another to act as a drug delivery vehicle, leading thus to multifunctional scaffolds. In the following sections two polymers used in tissue engineering will be discussed in terms of their use in BTE scaffolds as a coating material. The polymers discussed and used in this work are Poly(D,L-lactic acid) (PDLLA) and gelatin.

2.16.1 Poly (D, L-lactic acid) coated scaffolds

A commonly used synthetic polymer group used in tissue engineering are polyesters since the esterification process used to form these polymers is reversible making these polymers easily metabolised by the body making them ideal candidates for use as polymer coatings. (347) Poly(D, L- lactic acid) (PDLLA), structure shown in Figure 14, is one of four isotopes of poly(lactic acid) (PLA) due to its chiral nature with the L and D acid monomer units spread along the polymeric chain which results in a completely amorphous polymer due to its irregular structure. (348) (347) (349) PDLLA has the lowest degradation rate of the four PLA isotopes and for this reason has appealed to researchers in tissue engineering. (348) (347) (349)

Figure 14:- Structure of PDLLA (350)

Recently researchers have used PDLLA to coat bioactive glass scaffolds in order to improve the mechanical properties of the glass-ceramic scaffolds without compromising the desirable properties

73 of the glass-ceramic scaffolds. Various researchers have combined BBG scaffolds and have showed that the addition of the PDLLA coating increases the compressional strength of the scaffold, infiltrates the micro-cracks and overall improves the mechanical competence and stability of the glass-ceramic scaffolds. (110) (351) (352) (353)

It was observed despite the presence of the PDLLA coating the BBG scaffolds maintained the bioactivity. Moreover the PDLLA/Bioglass® scaffold also maintained its mechanical integrity during the immersion period in SBF. (110) In the polymer coated scaffolds the PDLLA retards the transformation of the crystalline phase (Na2Ca2Si3O9) to the amorphous calcium phosphate phase and this ensures mechanical stability of the scaffold over a longer period of time in SBF. (110)

The use of PDLLA as a biodegradable drug delivery vehicle has started to be explored with research into its use to transport antibiotics such as gentamicin (354) (355) and growth factors such as insulin-like growth factor-1 (IGF-1) and transforming growth factor-β1 (TGF-β1). (356) However this falls outside the scope of this work and will not be investigated in this work.

PDLLA coated BBG scaffolds represents the control polymer coated scaffold for this work, as it is already been proven to be a viable coating material for BTE.

2.16.2 Gelatin as a potential coating agent

Gelatin is a natural, biocompatible polymer which is completely dissolvable in vivo and is obtained through the physical/chemical degradation or thermal denaturation of collagen which breaks the triple helix structure. (357) (358) Gelatin contains a mixture of single and multi-stranded polypeptides each with extended left-handed proline helix with a typical structure shown in Figure 15 which contains a number of compounds such as glycine, proline and 4-hydroxyproline. (357) (358)

Since gelatin is a natural polymer, derived from collagen, it is naturally biocompatible and doesn’t induce cell toxicity which can be an issue with synthetic polymers. (359) (360) It comes in two type’s dependant on the preparation of the collagen; type A and type B. (361) (362) Type A gelatin is prepared using an acid pre-treatment and is derived mainly from pig skin and is also known as porcine gelatin. (362) Type B gelatin is prepared using an alkaline pre-treatment and is derived mainly from cow skin and bones and is also known as bovine gelatin. (362) As they are processed in different ways they have slightly different properties. (361) (362)

Figure 15:- Structure of gelatin (357) (358)

The gelatin coating process described by Liu et al (357) is similar to the process to be used in this work and is fully described in the methodology sections (sections 4.2.3.II). Liu et al (357) showed that the presence of gelatin coating on a porous calcium phosphate scaffold improved the compressive strength by a factor of 5 and the elastic modulus by a factor of 3, (357) making gelatin an ideal candidate to improve the mechanical properties of a BBG scaffold. Metze et al (363) which is based on the preliminary studies produced in this work, presented in the results section, showed that coating a highly porous BBG scaffold with gelatin improved the compressional strength without retarding the desirable properties of the BBG scaffold. (363) It has been used in composite materials as a coating agent with HA (364) and CaP, (357) (365) but has only recently been used to coat BBG scaffolds fabricated by the foam replication technique. (363)

Tabata et al (366) discussed the combination of bFGF growth factor with gelatin in a form of a hydrogel. (366) The bFGF is released, caused by the gelatin degrading, and the growth factor goes on to induce angiogenesis. (366) Patel et al (367) describes how gelatin is combined with the growth factor BMP-2 to form a growth factor release system. This work found that the gelatin degraded linearly for up to 28 days and this demonstrated that the release of the growth factor can be controlled by incorporating it into a polymer. (367) This indicates that gelatin is an ideal candidate for being used as a drug delivery system. However there has been no research on incorporating copper or silver with gelatin, so this will also be investigated during the course of this work.

2.17 In vitro cell culture studies to test the behaviour of bone tissue engineering scaffolds

As previously touched upon, in vitro testing of glass-ceramic scaffolds is important to ascertain how the cells interact with the scaffolds. This type of in vitro testing investigates issues with biocompatibility and toxicity that could arise once the scaffold has been implanted. This work subjects the multifunctional scaffolds to three types of cells and how they are used is described in the methodology section:-

 Human periodontal ligamental stromal cells (HPLSC’s)  Rat bone marrow stem cells (rMSC’s)  Human bone marrow stem cells (hMSC’s)

There is limited literature assessing the behaviour of HPLSC’s exposed to BBG scaffolds. However the literature has shown that 45S5 Bioglass® supported the proliferation and attachment of these cells without any toxicity issues. (113) (368) (369) Studies have indicated that the BBG scaffolds have osteogenic potential due to enhanced expression of collagen type I, osteocalcin and alkaline phosphatise. (369) (370) The BBG also showed improvements in stimulation in mineralisation and production of the extracellular matrix and calcium nodules. (370) (371) It should also be noted that when exposed to either rMSC’s or hMSC’s 45S5 Bioglass® was found to support cell proliferation and differentiation and cell survival. (372) (373)In particular induction of the genes responsible for cell adhesion and differentiation have been observed. (374)

When exposed to low levels of silver, HPLSC’s are found to behave normally; there is no difference in their proliferation or attachment compared to the BBG scaffolds and there is no cytotoxic effect observed. (113) Increased exposure to high concentrations of silver has a detrimental effect of the cells due to a cytotoxic effect eventually leading to cell death. (375) Silver has shown dose- dependent cytotoxicity when exposed to both hMSC’s (376) and rMSC’s. (377) The concentration of silver needs to be kept low enough not to have a detrimental effect on the human cells whilst being high enough to have the desired effect on the bacteria.

Similar observations have been made for copper when exposed to HPLSC’s that the HPLSC’s are relatively resistant to the copper (378) although it has been observed that copper does cause a toxic effect on oral cancer cells. (378) It should also be noted that copper released from gold alloys used in dental implants have been observed to damage oral fibroblasts due to the copper elevating the reaction oxygen species. (379) Copper is also observed to have a dose dependant cytotoxic effect on both the rMSC’s (380) and hMSC’s (381) with a detrimental effect at higher concentrations. The concentration

76 of copper also needs to be kept low enough not to have a detrimental effect on the human cells whilst being high enough to have the desired antibacterial and angiogenic effect.

2.18 Bacterial testing on bone tissue engineering scaffolds

As one of the main additional functions that it is hoped that is imparted to the scaffolds is an antibacterial capability this needs to be established in vitro through exposing the multifunctional scaffold to bacteria. This type of in vitro testing investigates whether the scaffold will cause a bacteriostatic or bactericidal effect on the bacteria. A bacteriostatic effect would occur if the bacteria are inhibited but not necessarily killing the bacteria. (382) (383) A bacteriostatic agent merely prevents the growth and/or reproduction of the bacteria whereas a bactericidal agent kills the bacteria. (382) (383) In order to define the antibacterial effect of a material the minimum inhibitory concentration (MIB) and the minimum bactericidal concentration (MBC) need to be determined. These values give the minimum concentration of the material that is required to cause a bacteriostatic or a bactericidal effect, respectively. (382) (383)

This work subjects the multifunctional scaffolds to three different bacteria and how they are used is described in the methodology section:-

 Escherichia coli (E.coli)  Staphylococcus aureus (S.aureus)  Klebsiella pneumoniae (K.pneumoniae)

Bacterial can be split into two classes and depending on which class they belong determines a general behaviour with respect to antibiotics and antibacterial agents. Bacteria are either classed as gram-negative or gram-positive. Gram-negative bacteria do not retain the crystal violet dye when subjected to the gram staining protocol whereas gram-positive bacteria do. (384) Along with the cell shape and structure of the cell wall, this gram staining is the main difference between the two bacterial classes. (384) E.coli and K.pneumoniae are gram-negative rod shaped bacteria whereas S.aureus is a spherical gram-positive bacterium.

A distinct difference between these classes is the structure and thickness of the cell wall, with the cell walls being between 15-80nm for a gram-positive bacteria and only 2nm for a gram-negative bacteria. (385) A gram-positive bacterium has a basic cell wall which is predominately made up of peptidoglycans (386) which makes this bacteria class more mechanically stable than the gram-negative bacteria. (387) Gram-negative bacteria have a complex cell wall made up predominantly of phospholipids, lipopolysaccharides, proteins and a thick layer of peptidoglycans. (386) The thickness

77 of the peptidoglycan layer determines how susceptible the bacterium is to antibacterial agents. (296) (388)

The mechanism of the antibacterial potential of silver and copper was discussed in sections 2.13.2 and 2.13.3 respectively. The literature shows that these metal ions should be more effective, i.e. require a lower concentration of the metal ions to achieve the MIC and the MBC, against the gram- negative bacteria compared to the gram-positive bacteria due to difference in the cell wall structure. (296) (388) (387) (386) (385)

E.coli is a common bacteria which is inexpensive and easy to grow and work with, making it the standard gram-negative bacteria used to ascertain a materials antibacterial potential. S.aureus is the most common cause of post-operative infections and any antibacterial implant material needs to be tested against this bacterium to establish its potential. As S.aureus is a gram-positive bacteria which is causing a lot of problems when it becomes resistant to established antibiotic treatments (MRSA) it is standard for antibacterial testing as if a material has a suitable antibacterial effect on this bacteria it is likely to show an antibacterial effect for other resistant bacteria. K.pneumoniae has also come to notice as a cause of nosocomial infections (hospital-acquired infections) so it is prudent to test an implant material against it to establish the materials antibacterial potential. An antibacterial agent needs to be tested against a range of bacteria as its effect and the MIC and MBC will differ.

2.19 Summary of the literature review

With more people suffering from significant bone trauma which requires extensive reconstructive surgery, the field of BTE has had more of a role to play in the last few decades in becoming a realistic alternative for current bone replacement therapies and treating bone diseases. A large range of materials have been investigated for BTE scaffolds and many of them have been discarded as they do not fulfil the requirements set out by the mandate of BTE, as described in sections 2.6 and 2.9. More recently bioactive glass-ceramics and polymer/bioceramic composites have come to the forefront of the research effort as they can be designed /fabricated to have the desirable properties and structure which are comparable to those exhibited by bone tissue.

In this work BBG scaffolds combined with polymers and/or MSIE to introduce therapeutic metal ions represent a move forward to develop multifunctional scaffolds with enhanced compatibility for bone regeneration. The literature has shown that silver and copper will add additional functionality to the BBG scaffold and can be introduced into the structure of the pre-fabricated scaffold through MSIE. It is hypothesised that the additional copper and silver will improve the base scaffolds

78 antibacterial and angiogenic properties whilst maintaining the desirable properties. However, there is limited literature available on the use of MSIE to introduce therapeutic metal ions into BBG scaffolds.

Biocompatible polymers, such as gelatin and PDLLA, can be used as coatings on BBG scaffolds to improve the mechanical competence by filling up the micro-cracks that occur after the scaffold sinters but to act also as a delivery vehicle for bioactive molecules/therapeutic drugs to introduce an extra function to the scaffold. Combining BBG scaffolds coated with biopolymers is an emerging research area which is slowly expanding and this work will add to the current knowledge by considering gelatin as a suitable biopolymer coating. There is currently no literature on the combination of gelatin coating and metal ion doping as technologies to impart extra functionality to the bioactive glass scaffolds, meaning this work will be the first to produce such ion-exchanged scaffolds coated with gelatin.

This literature review has revealed that the systems selected to this project, including MSIE BBG scaffolds, gelatin coated BBG scaffolds and gelatin coated ion-doped BBG scaffolds are novel and represent a great potential for development of advanced, multifunctional scaffolds for BTE.

Chapter Three

3. Aims and objectives of the project

3.1 Aims

As indicated in chapter 1, the overall aim of this work is to produce a new family of multifunctional Bioglass® based composite scaffolds for BTE. The goal is thus to improve the existing first generation 45S5 Bioglass® scaffolds whilst maintaining the properties that make these scaffolds attractive for BTE, such as high bioactivity, and to incorporate other functions for improved performance. A number of approaches and experimental methods have been attempted during the course of this work including introducing therapeutic metal ions (copper and silver ions) through a MSIE process and coating the base scaffolds with a polymer (gelatin and PDLLA) to improve the structural integrity of the scaffolds.

3.2 Objectives

To achieve the overall aim of this work, stated above, the following tasks were considered:-

 To carry out structural characterisation on the BBG scaffolds using a standard set of characterisation techniques. Bioactivity will be assessed through immersion in simulated body fluid (SBF) and the mechanical competence will be tested though compression strength testing. Further characterisation will be carried out through scanning electron microscopy (SEM), energy dispersive x-ray analysis (EDX), x-ray diffraction (XRD) analysis and Fourier transform infrared spectroscopy (FTIR). The surface topography of the scaffolds will be investigated using white light interferometry (WLI) and the wetting behaviour will be determined by contact angle measurements. The base scaffold will act as a control for the rest of the modified scaffolds.  To carry out a series of cell culture studies on the BBG scaffold. To assess the biocompatibility of the samples using relevant cells investigating cell behaviour when in contact with these scaffolds.  To introduce silver ions and copper ions into the BBG scaffold using the MSIE technique. To optimise the MSIE technique, for the first time here applied to incorporate silver and copper into BBG scaffolds, including control of the concentrations used and, for the combination work

(MSIE introducing simultaneously both silver and copper ions) the ratio between the concentrations of silver and copper ions.  To structurally characterise the MSIE scaffolds. These scaffolds will then be characterised using SBF immersion tests to detect bioactivity, involving a large series of characterisation techniques which includes:- SEM, EDX, FTIR, XRD, mechanical testing, degradation studies, WLI, contact angle studies to measure wettability and depth EDX.  To biologically characterise the MSIE scaffolds. Using different cells to study the biocompatibility and possible toxicity of the MSIE scaffolds to compare their behaviour with that of base scaffolds.  To carry out a range of bacterial studies using the MSIE scaffolds. Using a range of bacteria to study the anti-bacterial properties of the MSIE scaffolds compared to the base scaffolds  To coat the BBG scaffold with PDLLA and both types of gelatin, using PDLLA as a control polymer. To optimise the coating conditions including weight percentage of gelatin in solution used and immersion period used. To compare the two types of gelatin against each other and against the base scaffold coated in PDLLA, including homogeneity of the coating, effect on the scaffolds porosity and mechanical properties  To structurally characterise the polymer coated scaffolds. Using the standard set of structural characterisation techniques, mentioned above, polymer coated base scaffolds are tested and compared to assess the effect of the presence of the polymer coating on scaffold behaviour for BTE.  To coat MSIE scaffolds with gelatin. To optimise the coating conditions for theses scaffolds, including weight percentage of gelatin and type of gelatin used.  To structurally characterise the polymer coated MSIE scaffolds. Using the standard range of characterisation techniques the polymer coated MSIE scaffolds are tested and compared against each other.  To evaluate all results and draw pertinent conclusions regarding knowledge generated from the research and to indicate and suggest avenues for related future research activities.

To facilitate the reading of this thesis Figure 16 gives an illustration of the experimental program indicating how these tasks have been carried out during the course of this work from the starting material, e.g. BBG samples, up to the experimental techniques used to characterise the different scaffold systems developed.

Figure 16:- Overview of tasks and experiments carried out in the framework of the present project

Chapter Four

4. Materials, Methodology and Characterisation techniques

4.1 Materials

The following sections provide the details for the materials used in this work including where they were obtained from and in what capacity they were used in this project.

4.1.1 45S5 Bioglass®

This work utilises the most common and commercially available bioactive glass called 45S5 Bioglass® developed by Hench et al (76) in the 1960’s, as described in chapter two. This bioactive glass

has a composition of 45 wt% SiO2, 24.5 wt% Na2O, 24.5 wt% CaO and 6 wt% P2O5. (3) The 45S5 Bioglass® powder was provided as a gift by Dr I Thompson of King’s College London, UK. The 45S5 Bioglass® powder was fabricated by the melt quenching as described in section 2.8.1. The size of the Bioglass® particles used in this project range from 10 to 20µm. The Bioglass® powder has a density of 2.70gcm-3. (389)

4.1.2 Foam replication technique materials

This project utilises the foam replication technique, described in Chen et al (4) and detailed in section 2.11.1, to produce the standard BBG scaffolds which will then be modified and characterised. The foam that is used as the template is polyurethane foam (PU foam) brought from Recticel, Corby (UK) which has 45 pores per inch (ppi). This template should lead to a trabecular pore structure of the scaffolds with a pore size and interconnectivity suitable for the objectives of this project. (269) (5) (270) (271) The binding agent used is polyvinyl alcohol (PVA) which was obtained from Sigma Aldrich (UK). The foam can be cut to the size and shape required by the characterisation test but the standard size used was 10x10x10mm3.

4.1.3 Molten salt ion exchange techniques

For the silver MSIE process, silver nitrate (AgNO3) (at 0.05mol/l) and sodium nitrate (NaNO3) from VWR (UK) and Sigma Aldrich (UK), respectively, were used. For the copper MSIE process, the same sodium nitrate was used and also copper (II) nitrate (Cu(NO3)2) was used, both from Sigma Aldrich (UK).

4.1.4 Biopolymers for scaffold coating

Gelatin was the biopolymer chosen to coat the scaffolds. The two types of gelatin used in this work are Type A and type B gelatin, with the difference between then described in chapter two. These were both obtained from Sigma Aldrich (UK). Type A gelatin has a pH between four and six, a bloom in the range 240-270 and a density of 0.777gcm-3. (390) Type B gelatin has a bloom of approximately 225, a pH between five and seven and a density of 0.812gcm-3. (391) For this project the two types of gelatin were dissolved in deionised water to produce coatings on the scaffolds. The degradation period depends on the content of the water added which is a measure of the extent of the cross-linking. The higher the water content the faster the gelatin will degrade in vitro. (362)

The other polymer used in this work is poly (D, L lactic acid) (PDLLA) and this has been obtained from PURAC Biomaterials, Holland. PDLLA has a degradation rate of between 12 and 16 months in SBF (211) (205) and a density of 1.26gcm-3. (392) This polymer is dissolved in dimethylcarbonate (DMC) in order to prepare coatings on scaffolds. The DMC was bought from Sigma Aldrich (UK).

4.1.5 Other materials used

The chemicals used in the production of simulated body fluid (SBF), shown in Table 6, were all purchased from Sigma Aldrich (UK). The potassium bromide (KBr) used in the Fourier transform infra-red spectroscopy (FTIR) characterisation was purchased from VWR. 70% Ethanol was used in the sample preparation to fix the samples for X-ray diffraction (XRD) characterisation and this was obtained from VWR (UK). Phosphate buffer solution purchased from Sigma Aldrich (UK) was used in the inductively coupled plasma mass spectroscopy characterisation (MS-ICP). The 70% Ethanol used

84 to remove the excess salt residue after the molten salt processes was purchased from VWR. Sodium dodecylsulfate (SDS) and Extran powder were obtained from Sigma Aldrich (UK).

4.1.6 Materials for the human periodontal ligamental stromal cell study

This part of the research was carried out at the Leeds Dental Institute, UK in collaboration with Dr R El-Gendy.

Teeth used to extract the human periodontal ligamental stromal cells (HPLSC’s) from were obtained with permission from the patients as donations to the Leeds Dental Institute Research Tissue Bank (LREC 07/H1306/93). The alpha modified minimum essential medium (α-MEM) and the foetal bovine serum (FBS) where procured from Lonza (USA), with the L-glutamine and the penicillin/streptomycin from Sigma Aldrich (UK) and Gibco (UK) respectively. The PBS used was obtained Lonza (UK) and was mainly used as a washing agent during this work.

The enzyme Collagense P used to remove the cells from the tooth pulp was provided by Roach (UK) and the 0.25% Trypsin Ethlenediaminetetraacetic acid (EDTA) used to de-attach the cells in the passage process where brought from Sigma Aldrich (UK). The two stains used in the cell viability study, Cell TrackerTM Green CMFDA and Ethidium homodimer-1 (Eh1), were purchased from Invitrogen (UK).

4.1.7 Materials for the rat bone-marrow stromal cell and human bone-marrow stromal cell studies

This part of the research was carried out at the University of Erlangen Medical Centre, Germany in collaboration with Dr L Strobel.

The rat bone-marrow stromal cells (rMSC’s) were obtained from Syngeneic male lewis rats from the Charles River Laboratory, Sulzfeld, Germany with the study approved by the animal care committee of the University of Erlangen-Nuremberg and the government of Mittelfranken, Germany. The human bone-marrow stromal cells (hMSC’s) were obtained from PromoCell Heidelberg, Germany (catalogue number C-12975).

The cell culture medium (CCM) used in this section of the cell biology work was a pre-mixed cell culture medium with Dulbecco’s modified eagle medium (DMEM) and Ham’s F-12 nutrient

85 mixture (Ham’s F-12)(DMEM/F-12), which was bought from Biochrome AG (Germany) with 10% Foetal bovine serum (FBS) from InVitrogen (Germany), 1% L-glutamine from Sigma Aldrich (Germany) and 1% penicillin/streptomycin from Biochrome AG (Germany). Later on 1% Bufferall from Sigma Aldrich (Germany) was added to the CCM. The PBS used in this work was bought from Sigma Aldrich (Germany). Alamar blue (AB) was bought from InVitrogen (Germany) and was used in a 10% solution. The trypsin-EDTA used to detach the cells in the passage stages was brought from PAA (Germany) and the trypin blue used to stain the cells in order to count them was brought from InVitrogen (Germany).

4.1.8 Bacterial study materials

This part of the research was carried out at the Friedrich-Alexander University of Erlangen- Nuremberg, Germany in collaboration with Mrs A Antmann.

The Escherichia coli, Staphylococcus aureus and Klebsiella pneumonia have ATCC (American type culture collection) identifies, which are the internationally recognised strain numbers for the bacteria, of ATCC 23716, ATCC 6538 and ATCC 4352 respectively. The industry standard medium for bacterial work and used in this work was Muller Hinton broth (MHB) and was provided by CARL ROTH (Germany). The other media is based on the one used for the cell biology work described in sections 4.4.4 onwards, with DMEM/F-12 from Biochrome AG (Germany) as the base with 10% FBS from InVitrogen (Germany) and with 1% L-glutamine and 1% Bufferall Sigma Aldrich (Germany). The materials used to produce the Standard Agar I plates used throughout the bacterial work, as described in appendix B.6, were all brought from CARL ROTH (Germany) except for the peptin which was purchased from MERCK (Germany). The metal salts used in determining the minimum inhibitory concentration of the metal salts, CuCl2, AgNO3 and Cu(NO3)2, where all provided by Sigma Aldrich (Germany). Finally the inoculation solutions used in the final bacteria study, described in section 4.4.13, (NaCl and Tween 80) were both brought from CARL ROTH (Germany).

4.2 Methodology

The following sections describe the various methods and processing techniques used in this work, including the foam replication technique to make the scaffolds, the MSIE process and the technique developed to coat scaffolds with the chosen biopolymers.

4.2.1 Foam replication technique

The scaffolds that form the back bone of this research were produced by the foam replication technique first described by Chen et al (4) and explored further in section 2.11.1. This involves producing a bioactive glass slurry mixture, dipping the sacrificial (PU foam) template, drying, burning out the PU foam template and finally sintering to densify the struts.

The slurry was prepared by mixing 6wt% of PVA and 40wt% of 45S5 Bioglass® powder. This slurry composition has been developed over time to be the optimal by members of Boccaccini’s research group at Imperial College to give suitable “green bodies” (Bioglass® coated PU foams) for scaffold production. In the present study, the following steps were followed. The PVA is dissolved in distilled water at 80oC, being constantly stirred on a magnetic heating stirrer. After the PVA granules have dissolved, after at least an hour, the powered 45S5 Bioglass® is added and left to be stirred, for a minimum of one hour, with the heating element switched off. Once the glass powder has been thoroughly mixed in, has cooled sufficiently that it can be handled, the dipping process to produce the “green bodies” can occur.

The PU foam is cut into a shape and size that is suitable for the characterisation technique that it will be subjected to. The standard size of the template foam was 20x20x20mm3 to produce scaffolds of 10x10x10mm3, considering that due to shrinkage during sintering the template foams need to be double of the size required, as a volumetric shrinkage of approximately 50% is anticipated. (4) Once the foam scaffolds are shaped and cut, they then can be dipped into the now cooled slurry mixture, making sure that the foam is thoroughly coated in the slurry mixture but also making sure that excess slurry is removed from the green bodies. Usually the dipping process takes approximately an hour for 50ml of slurry, which would produce between 15-20 20x20x20mm3 green bodies. The green bodies are then dried by either being left in a fume cupboard for 24 hours or in a furnace for 12 hours at 60oC.

Once dry, the green bodies are then sintered according to pre-set conditions which have been investigated and optimised by Bretcanu et al (269) and the sintering profile is shown in Figure 17.

The PVA/PU foam template is burnt out first at 550oC before densification (and crystallisation) can take place at 1100oC. Once the scaffolds have cooled down to room temperature they can be removed from the furnace and used for the rest of the work described in future chapters.

Figure 17:- Sintering profile for the fabrication of BBG scaffolds by the foam replication technique described in Chen at al. (4)

4.2.2 Molten salt ion exchange processing

This technique, as described in section 2.15, has been adapted for use in biomaterials from processing glasses for other purposes such as wave guides (328) (336) (337) and optical splitters (329) (336) (337) and is based on the method which was originally described by Di Nunzio et al (6) for glass- ceramic scaffolds. This technique has been modified, for the first time in this project to work with 45S5 Bioglass® powder to incorporate silver and copper separately and in combination.

I Silver molten salt ion exchange process

As shown in section 2.15.2 the use of MSIE to modify the chemistry of a glass has been applied before in both biomaterials and other applications. Based on the work carried out by Di Nunzio et al (6) silver ions can be introduced into the silicate matrix of a prefabricated scaffold using MSIE. This is possible by using sodium nitrate as a controlling agent in the salt bath. The aim is for a proportion of the sodium in the scaffolds glassy matrix to exchange with the silver in the molten salt bath, as shown by Equation 2 and Equation 3.

o o The melting points of NaNO3 and AgNO3 are 310 C and 210 C, respectively. The molten salt bath temperature will need to be higher than that to ensure that the salt bath is fully molten before the scaffolds are immersed into the salt mixture. In this work the temperature of the furnace was set to

o 400 C throughout the silver MSIE process. Preliminary testing was carried out using the NaNO3 to

AgNO3 ratios suggested by Di Nunzio et al (6) and shown in Table 2.

Table 2:- Concentrations, ratios and quantities of sodium nitrate and silver nitrate used in the preliminary study of the silver MSIE process

Concentration Ratio of NaNO3 Mass of Mass of

to AgNO3 NaNO3 (g) AgNO3 (ml) High 20 to 1 4.25 50 Medium 200 to 1 42.5 50 Low 2000 to 1 42.5 5

The MSIE process used in this work, as described in Newby et al, (113) involves dipping the sintered BBG scaffolds into the molten salt bath for a set immersion period, which for the preliminary study was 15 minutes, removing the scaffold and once cooled washing the scaffold in distilled water before subjecting the silver MSIE scaffolds to a battery of characterisation tests, as out-lined in Figure 18.

Figure 18:- Flow diagram of the MSIE process applied to incorporate silver into Bioglass® - based scaffolds

Based on the preliminary study results, a main MSIE study was carried out. For the main silver MSIE study a number of changes were introduced to the process as the method described previously needed optimising due to several issues identified during the preliminary study. For example, it was found that due to thermal shock the scaffolds could be fractured in the salt bath or in the washing stage

89 or were mechanically failing during handling after cooling and during the washing process. This effect was negated by having the scaffolds in the furnace whilst the salt bath was melting and controlling the rate of temperature increase to 5oC per minute. The heated scaffolds were then dipped into the molten salt bath, left for the required immersion period, removed from the salt bath and placed back into the furnace and cooled at 5oC per minute.

The concentrations used in the preliminary study, listed in Table 2, were only proof of concept concentrations and, as stated also by Di Nunzio et al, (6) the high and medium concentrations from Table 2 were too high and would likely lead to a cytotoxic effect. The silver nitrate concentrations for the main study were then selected from the lowest concentration stated in Table 2 up towards the medium concentration from Table 2, as shown in Table 3.

Table 3:- Concentrations, ratios and quantities of sodium nitrate and silver nitrate used in the main study of the silver MSIE

Concentration Ratio of NaNO3 Mass of Mass of AgNO3 to AgNO3 NaNO3 (g) (ml) High 800 to 1 42.5 12.5 Medium 1400 to 1 42.5 7.14 Low 2000 to 1 42.5 5

In the preliminary study the only immersion period investigated was 15 minutes, however it was shown that for the lowest concentration at this immersion period it was difficult to detect the presence of silver in the scaffold. Longer immersion periods were required for the lower silver nitrate concentrations to allow the silver ions to exchange into the scaffold. Therefore, immersion periods of 15, 30, 45 and 60 minutes were investigated to improve the chances of the MSIE occurring.

The last alteration from the method used in the preliminary study was related to the washing stage after the MSIE process. It was found that a large amount of salt residue was still present on the scaffolds surface despite the scaffolds being washed in distilled water. In this investigation (see also the published study; Newby et al (113)) it was found that introducing an ethanol washing step should help to reduce or even eliminate the salt residue on the surface of the scaffold. The new washing procedure starts with an initial washing step in warm distilled water (at approximately 40oC) for 10 minutes, followed by 10 minutes in 70% ethanol at room temperature and for a further 10 minutes in warm distilled water. Finally the scaffolds are left to dry either in a fume cupboard for 24 hours or for 12 hours at 60oC in a furnace.

This revised process is summarised in Figure 19 and it was used in the main silver MSIE study.

Figure 19:- Flow diagram of the optimised MSIE process for the incorporation of therapeutic ions into Bioglass® -based scaffolds

II Copper molten salt ion exchange process

The copper MSIE process itself followed the same procedure described above for the main silver MSIE study, as shown in Figure 19, with differences in the chemicals used. Once again sodium nitrate is used in the salt melt to act as a controlling mechanism for the MSIE process, as described in chapter two, along with copper(II) nitrate (Cu(NO3)2). The MSIE process is described by Equation 4 and Equation 5 in section 2.15.3.

o The melting point of Cu(NO3)2 is 256 C and as previously stated the melting point of NaNO3 is 310oC making the furnace temperature requirement to be above these values. The furnace temperature

o was set at 400 C. Table 4 shows the concentrations, quantities and ratios of Cu(NO3)2 and NaNO3 used for this work. The copper MSIE process follows the protocol set out in Figure 19 with a limited immersion period of 15 minutes.

Table 4:- Concentrations, ratios and quantities of sodium nitrate and copper nitrate used in the study of the copper MSIE

Concentration Ratio of NaNO3 Mass of Mass of

to Cu(NO3)2 NaNO3 (g) Cu(NO3)2 (g) High 800 to 1 200 0.25 Medium 1400 to 1 175 0.125 Low 2000 to 1 200 0.1

III Silver and copper combined molten salt ion exchange process

The final MSIE process involves having both copper nitrate and silver nitrate in the salt bath along with sodium nitrate to control the MSIE process. This has been described by Equation 6 and Equation 7 shown in section 2.15.4. As stated in sections 4.2.2.I and 4.2.2.II the furnace temperature was set to 400oC to ensure that the salt bath is fully molten when the scaffolds are dipped into the salt bath. The quantities of Cu(NO3)2, AgNO3 and NaNO3 used in this work were 0.25g, 1.56ml and 350g respectively, giving a metal salt ratio of 1:1:1400 in the salt bath. This combination MSIE follows the protocol shown in Figure 19 with immersion periods of 15, 30, 45 and 60 minutes.

4.2.3 Biopolymer coating

The coating of BBG scaffolds by biodegradable polymers to produce scaffolds with enhanced mechanical properties and with additional functions was carried out, as described in section 2.16. In the following sections details of the coating procedure employed for PDLLA and gelatin are provided.

I Poly (D, L-lactic acid) coating

PDLLA coating of BBG scaffolds has been carried out previously; including the work depicted by Chen et al (110) and Bretcanu et al. (352) In the present work, a 5wt% PDLLA solution in an organic solvent (DMC) was required to produce suitable coatings by dipping in solutions. The technique uses one gram of PDLLA to 20ml in DMC to form a solution which is stirred constantly for two hours at room temperature. A simple dipping method was employed to coat the BBG scaffolds with PDLLA, as depicted in Figure 20. The scaffold samples are slowly dipped into the PDLLA solution, left to immerse for five minutes and then scaffolds are left to dry at room temperature in a fume cupboard for 24 hours to undergo further characterisation once they are dry.

Figure 20:- Flow diagram of the coating process used to coat the BBG scaffolds with PDLLA

II Gelatin coating

This coating process was carried out with the two types of gelatin (type A and B), both described in the literature review in section 2.16.3, and the results were compared to determine which gelatin type is more suitable for this particular application. The gelatin solution, for both type A and type B gelatin, is made by adding the gelatin powder (or pellets) to distilled water heated to 50oC stirring continuously until the entire amount of the solid gelatin has dissolved, this took approximately 30 minutes. The solution is then cooled to 40oC for the dipping process to allow the viscous nature of the gelatin to appear and so the scaffolds can be handled during the dipping process. The scaffolds are then slowly immersed into the gelatin solution and left for the required immersion period. The scaffolds are then removed, shaken gently and patted with tissue paper to remove excess gelatin, and left to dry in a fume cupboard for 24 hours in preparation for characterisation. This process is represented in Figure 21.

Figure 21:- Flow diagram representing the process used to coat BBG scaffold samples with gelatin in this work

With very little literature or previous work to give a starting point in terms of which gelatin type, gelatin solution concentration, immersion period in the gelatin solution and the number of coats to use, the first set of studies was designed to determine the concentration of the gelatin solution to be used in the main study. In Liu et al (357) and Bellucci et al (393) a 5wt% gelatin solution was used to coat porous bio-ceramics with gelatin and this was the starting point for the present study. These initial concentrations for both gelatin types are shown in Table 5 with concentrations of 5-15wt% used for type B and 5-10wt% used for type A. To determine which of these concentrations should be considered for optimal results the porosity of the coated scaffold shall be assessed, which was determined using Equation 8. For the preliminary study, immersion periods of one, five, ten and fifteen minutes were used in conjunction with the concentrations stated in Table 5.

As stated in section 5.6.1, the first half of the preliminary study showed that gelatin solutions with concentrations of 5 and 6 wt% combined with an immersion period of one minute gave results that presented the best compromise between the required high porosity and the coating thickness. These concentrations and immersion period were then used to produce multi-layered coated scaffolds, still following the flow chart shown in Figure 21, but repeating the coating process once and twice to produce scaffolds with two and three coats of gelatin, respectively.

Table 5:- Concentrations and quantities of gelatin used in the preliminary studies

Concentration of gelatin Mass of gelatin Mass of distilled water in solution (wt%) (g) (ml) 5 5 95 6 6 94 7 7 93 8 8 92 9 9 91 10 10 90 11 11 89 12 12 88 13 13 87 14 14 86 15 15 85

Since the preliminary studies have shown that a single coat of 5 or 6 wt% gelatin using an immersion period of one minute produces the best compromise between samples porosity and coating thickness, these samples, produced following the procedure outlined in Figure 21, were subjected to a full batch of characterisation techniques, as described in section 4.3, to determine their suitability as scaffolds for BTE.

4.2.4 Combination scaffolds

This set of samples was developed to combine the MSIE process shown in section 4.2.2 with the gelatin coating method described in section 4.2.3. Copper and combination MSIE scaffolds were prepared as described in sections 4.2.2.II and 4.2.2.III, respectively. Only the lowest concentration samples for the copper MSIE and only using samples that are immersed in the molten salt bath for 15 minutes for the combination MSIE scaffolds were considered. After these MSIE scaffolds have been left to dry for 24 hours, as shown in Figure 19, they were coated in gelatin as described in section 4.2.3, following Figure 21, using gelatin solution concentrations of 5 and 6 wt% and an immersion period of one minute for both types of gelatin. Once these coated MSIE scaffolds were dry were submitted to characterisation studies as described in section 4.3.

4.3 Structural Characterisation techniques

The following sections describe the relevant techniques used to characterise the samples. The basic principles behind each of the techniques will be briefly described with any relevant additional information included in the appendices.

4.3.1 Porosity studies

For un-coated scaffolds the porosity is determined by means of Equation 1 shown in section 2.9.1. Only the overall porosity of the scaffolds has been established during this work as the other porosity related characteristics such as pore size and interconnectivity are directly related to this measurement and they have been already established for the base scaffold in previous work. (269) (5) (270) (271) This required the dimensions of the scaffold to be measured (height, width and length) to obtain the volume of the scaffold and the mass of each scaffold is needed in order to calculate the porosity by geometrical methods. For the BBG scaffolds this will produce an accurate measure of the porosity of the scaffold as the density of 45S5 Bioglass® can be used in the calculation. However as the density of the modified glass-ceramic was not determined in this work, the density of 45S5 Bioglass® powder was used, assuming it to be close to the density of the crystallised 45S5 Bioglass® material. This means that the porosity values for the MSIE scaffolds will only be estimation of the porosity and given an appropriate error of 5% on top of the experimental errors. These values will still give a meaningful estimation of the porosity as theoretically the density of the MSIE modified 45S5 Bioglass® powder should not be vastly different from that of the 45S5 Bioglass® powder.

For polymer coated scaffolds a slightly different method is needed in order to work out the coated scaffold porosity. Karageorgiou et al (210) described the equation that is required in order to calculate the porosity of a coated scaffold; this is Equation 8 shown below.

Equation 8:- Porosity of polymer coated scaffolds (210)

The difference between the porosity of uncoated and coated scaffolds will give useful information as to how much the coating is infiltrating into the scaffolds porous structure. With the ideal porosity for BTE scaffold being above 70% (394) and the foam technique producing scaffolds with a

96 porosity of over 90%, (244) the polymer coatings should not be too thick as to significantly decrease the overall porosity of the scaffold.

4.3.2 Simulated body fluid studies

Immersion in simulate body fluid (SBF) is the standard technique currently used to test the acellular bioactivity of biomaterials, as described in chapter two, with the precipitation of HA on the surface being the marker of bioactivity. SBF is prepared according to the standard procedure written by Kokubo et al. (395) (396) It is prepared by dissolving a range of chemicals, shown in Table 6, in 750ml of deionised water at 37oC being stirred constantly. The chemicals are added one at a time in the order shown in and then the solution is then buffered at pH 7.4 using HCL (listed at number 10 in Table 6). Once this pH adjustment has been completed the SBF needs to cool for two to three hours and then it needs to be topped up to one litre using deionised water and stored in a refrigerator until needed.

In a typical experiment, the samples are placed in centrifuge tubes, immersed in 45ml of SBF, placed in an incubator at a constant 37oC for the required immersion period of the study refreshing the SBF every three to four days, as described in Chen et al. (4) The samples are left immersed in the SBF solution for different immersion periods ranging from 24 hours to four weeks. After immersion, the samples are gently washed in deionised water and left to dry in a fume cupboard for 24 hours and then characterised.

Table 6:- Chemicals used for the production of simulated body fluid (SBF)

Order Chemical Quantity number 1 NaCl 7.996g

2 NaHCO3 0.350g 3 KCL 0.224g . 4 K2HPO4 3H2O 0.228g . 5 MgCl2 6H2O 0.305g 6 HCL 40cm3 with a concentration of 1kmol/m3

7 CaCl2 0.278g

8 NaSO4 0.071g

9 (CH2OH)CNH2 6.057g 10 HCL as much as is required to change the pH with a concentration of 1kmol/m3

4.3.3 Scanning electron microscopy

Scanning electron microscopy (SEM) is a technique which makes use of a high energy beam of electrons to interact with the atoms on the surface of a sample. Energy signals that result from these interactions are then detected and converted into high resolution images of the sample’s surface. The images have a 3D appearance because of the large depth of field that the images have, this is useful as it gives a better insight into the structure of the material. (397) For this work the variable pressure JEOL JSM 5610LV instrument has been used. This SEM is fitted with a secondary electron detector as well as a backscattered electron detector and has a magnification range from x50 to x90000. Samples prepared for SEM characterisation have to be dry, mounted on sample stubs using carbon tape and then gold coated (approximately 15nm thick). All MSIE samples were washed in accordance with the washing protocol described in section 4.2.2.I before imaging.

4.3.4 Energy dispersive X-ray spectroscopy

Energy dispersive X-ray spectroscopy (EDX) is a characterisation technique used to determine the elemental composition of a sample. This technique relies on the interaction between a beam of high energy radiation (such as x-rays) and the sample. (398) The elemental composition of a sample can then be identified as each element will give a specific energy signal. (398)

This technique was primarily used to detect the elements present in the tested samples, in particular it was used to determine if silver and copper are present after the MSIE processes, and to examine the change in composition of the scaffold surfaces after the samples were immersed in SBF, as described in section 4.3.2. All MSIE samples were washed in accordance with the washing protocol described in section 4.2.2.I before characterisation. The EDX instrument is an attachment to the JEOL JSM 5610LV SEM described in section 4.3.3. Spot analysis of the sample was carried out at a range of magnifications, between x1000 and x10,000, at a number of sites across the sample to ascertain the elemental composition and that it was consistent across the sample.

4.3.5 X-ray diffraction analysis

X-ray diffraction (XRD) was used to determine the crystalline composition of samples. (399) A Phillips PW 1700 Series automated powder diffractometer was used employing Cu Kα radiation and

98 scanning over an angle range of 5-80o. Pellet samples were analysed straight after they had been produced and modified and subjected to the washing protocol described in 4.2.2.I. Scaffold samples had to first be prepared, modified, washed and then crushed and ground into a fine powder before XRD characterisation.

4.3.6 Fourier transform infra-red spectroscopy

Infrared spectroscopy represents a finger print of a sample with specific peaks (absorption or transmission depending on the equipment) which corresponds to the frequency of vibrations in the sample, related to its bonds and atoms. (400) Since each material has a unique set of atoms and bonds it means that no two materials will produce the same infrared spectrum. (400) This technique is very fast, with one sample being characterised in approximately one minute, and is non-destructive. This technique was used to identify specific bond combinations within the samples which were then used to identify the presence of the modified additions to the BBG construct. This project used a Bruker Vector 22 TGA-IR instrument using the transmission attachment, making all the spectra transmission spectra. The samples were prepared by grinding the washed scaffolds into a fine powder, mixing them with potassium bromide (KBr) and then formed into pellets by using a pellet press, as described in appendix B.1.

4.3.7 White light interferometery

The surface topography of the materials, as described in section 2.9.3, was assessed by white light interferometry using a Zygo® white light interferometer and using MetroPro® 8.1.5 software to analyse the data. Pellets of the different samples types, described in sections 4.2.1, 4.2.2 and 4.2.3, and were prepared as described in appendix B.1 and washed in accordance with the washing protocol described in section 4.2.2.I. The changes in surface topography caused by the MSIE process or the addition of a polymer coating or both (sections 4.2.2, 4.2.3 and 4.2.4 respectively) need to be assessed as well as the surface changes caused by the immersion in SBF for 1, 3, 7, 14, 21 and 28 days as described in section 4.3.2. The use of sintered pellets instead of the sintered scaffolds is because this technique measures the surface topography of a flat substrate (224) and can not be carried out on the scaffolds that have been produced in this work. The substitution of a pellet instead of the scaffold has been used in previous studies. (224) The pellets are sintered and modified in the same manner as the scaffolds meaning that changes to the surface due to the sintering process, MSIE and polymer coating

99 can be extrapolated from the pellet measurements. For future work in order to produce pellets that are closer to the base scaffold, both chemical and physically, the pellets could be produced from a slurry, as described in 4.2.1 for the scaffolds, which when dried is used to make the pellets.

The three quantities that are used to compare different surface topography characteristics are the RMS, the Ra and the PV. RMS is the root-mean-square deviation from the centre line which is selected by the MetroPro® software. (401) RMS is calculated by Equation 9 and is illustrated by Figure 22. (401)

Equation 9:- RMS of a sample being characterised using white light interferometry (401)

Ra is the mathematical average deviation from a central line that is determined by the MetroPro® software and is given by Equation 10, illustrated in Figure 22 and Ra is larger the RMS. (401) PV is the peak to valley distance measured between the lowest and highest points on a sample cross-section. The PV is the maximum roughness height for a sample and it compares the two extremes found on a sample surface as demonstrated in Figure 22. (401)

Equation 10:- Ra of a sample being characterised using white light interferometry (401)

Figure 22:- Definitions of RMS, Ra and PV when applied to the roughness of a sample (401)

100

RMS and Ra give deviations away from the given central line that the software deems the best fit for that particular cross-section of the sample and the PV gives the distance between the tallest peak and the lowest valley. (401)

All three quantities give an overall picture of the surface topography of a sample. Images are also obtained when carrying out this technique and can be used to gain a qualitative assessment of the sample’s topography including a 2D black and white intensity plot, 2D solid colour plot, 3D oblique plot, cross-sections showing the surface profile and 2D coloured surface map. Both the qualitative and the quantitative results will be used to assess the changes in the surface topography due to the different surface modification processes and the immersion in SBF to assess the samples bioactivity.

4.3.8 Wettability

To evaluate the wettability of a sample, as described in section 2.9.2, static contact angle measurements were carried out in collaboration with researchers at the University of Erlangen- Nuremberg, Germany. Pellets were produced as described in appendix B.1 and washed in accordance with the washing protocol described in section 4.2.2.I. As described in section 4.3.7, pellets were used as their surface correlates to a suitable extent to the surface observed for the scaffolds, after sintering and modification. As indicated in appendix B.1, the pellets are sintered and modified in the same manner as the scaffolds meaning that the changes to the surface due to the sintering process, MSIE and polymer coating can be extrapolated from the pellet measurements. It is however understood that this will not be 100% correct due to the slight pressure applied to make the pellets, which is not applied in the scaffold fabrication, which could have an effect on the densification process. Nevertheless such surfaces of sintered pellets can closely resemble the topography and wettability of the porous scaffold.

A 3µl drop of distilled water is placed onto the surface of the sample and the contact angle is measured using a Krüss DSA30 instrument. This was repeated five times and the average was taken for each sample type using a DSA4 program. Wettability information is important because it indicates how well the cells will attach to the surface of the materials after the modifications have been made. In this case an advancing angle was recorded to give contact angle results and as long as the surface of the sample is not significantly very rough, this value will give an accurate contact angle for the sample. In order to get the best measured value for the contact angle both the advancing and the receding angles need to be measured and averaged as the measured contact angle falls between the two angles. This especially needs to be carried out if a significant difference between both contact angles is expected when the sample surface is very rough, which was not the case of the present pellets.

101

4.3.9 Inductively coupled plasma mass spectrometry

Inductively coupled plasma mass spectrometry (ICP-MS) is a mass spectroscopy technique capable of detecting metal ions at very low concentrations within a liquid sample. This is done by ionizing the sample with high temperature plasma and then using a mass spectrometer to separate and quantify those ions. ICP-MS allows the detection of elements within the range of lithium and uranium. ICP-MS can detect a number of elements simultaneously or it can be set to only detect the user required elements. This is done by calibrating the instrument first with control samples of deionised water with the ions that are of interest. In this case control samples of deionised water with copper and silver ions were used to calibrate the instrument as these were of interest. Only changes in the quantity of silver and/or copper ions were detected so the release rate of these ions could be assessed. The sample is vaporised first (before ionisation occurs) and the sample passes through a super-heated plasma (ICP). The ions are separated according to their specific mass to charge ratios by passing the ions through an electromagnetic field. The ions are detected and the signals given off are processed into spectra which is characteristic for the sample.

The washed MSIE samples, prepared as described in sections 4.2.1, 4.2.2 and 4.2.3, were immersed in PBS for immersion periods of 1, 3, 7 and 28 days. The supernatant was removed, with a minimum quantity of 3ml being required, and tested for silver and/or copper using an Optima 7300 DV MS-ICP instrument. The change in the quantity of silver and copper detected in the supernatant was plotted over time to indicate the release rate of silver and the copper from the samples. This release rate was compared between the two immersion media to confirm it is consistent.

4.3.10 Mechanical properties

The mechanical competence of the samples was quantitatively asses by measuring the compressional strength using a Zwick/Roell Z010 instrument with a cross-head speed of 0.5mm per minute and a load of 1kN. The load was applied until the compressive strain (%) reaches 70% and the maximum compressive strength was recorded.

The prismatic sample scaffolds, prepared and washed in accordance with the methods described in sections 4.2.1, 4.2.2 and 4.2.3, have nominal dimensions of 10x10x15mm3 with the load being applied on one of the 10x10mm2 sides. To prevent the sample slipping whilst under the load a 0.05mm Teflon® film was placed onto the plates holding the scaffold. The scaffolds should be visually checked to ensure that the sample has parallel sides to ensure the compressional force being applied will be applied homogenously. The compressional testing was carried out on samples before and after

102 immersion in SBF to investigate the degradation of the sample and the effect of this on the mechanical competence of the sample. Both coated and uncoated samples were tested along with the MSIE samples.

4.3.11 EDX depth studies

Sample pellets for a variety of sample types, including the BBG and the MSIE samples as described in sections 4.2.1 and 4.2.2, were prepared and washed as described in appendix B.1. As described in previous sections, pellets were used as their surface correlates to the surface observed for a scaffold considering that the pellets were sintered and modified in the same manner as the scaffolds and therefore changes to the surface due to the sintering process, MSIE and polymer coating can be directly extrapolated from the pellet measurements.

To ascertain how far the MSIE ions (silver and copper in this work) are penetrating the silicate matrix, the pellets need to be fixed in a resin and polished down to reveal the cross section of the pellet, as shown in Figure 23. The fixing and polishing methods used in this work are described in detail in appendix B.2. EDX, as described in section 4.3.4, was carried out at a number of points across the exposed cross-section, as shown in Figure 23. This gives an indication of the depth of penetration of the ions into the glass matrix.

Figure 23:-Schematic diagram illustrating the fixing, polishing and preparation of samples ready for the EDX depth analysis

103

4.4 Biological Characterisation techniques

This section covers the techniques used to carry out the cell biology and bacterial studies reported in this work. the work detailed in the following sections was carried out in collaboration with the Leeds Dental Institute, UK, for the in vitro studies using human periodontal ligamental stromal cells (HPLSC’s) (sections 4.4.1 to 4.4.3), the Department of Plastic and Hand surgery at the University of Erlangen Medical Centre, Germany, for the in vitro studies using rat bone-marrow stromal cells (rMSC’s) and human bone-marrow stromal cells (hMSC’s) (sections 4.4.4 to 4.4.7). The Institute of Bioprocess Engineering at the Friedrich-Alexander University of Erlangen-Nuremberg, Germany, carried out the bacterial work carried out using Escherichia coli (E.coli) (sections 4.4.8 to 4.4.12.I) and the Institute of Hygiene and Biotechnology at Hohenstein Institute in Bönnigheim, Germany collaborated with the bacterial work carried out using Staphylococcus aureus (S. aureus) and Klebsiella pneumoniae (K. pneumoniae) (sections 4.4.12.II and 4.4.13). All the studies used the BBG scaffold as the control sample to give a comparison.

4.4.1 Cell preparation and extraction for human periodontal ligamental stromal cells

The human periodontal ligamental stromal cells (HPLSC’s) have to be extracted from the external tooth surface of adult wisdom teeth using sterile scalpel as described in Somerman et al. (402) Collagense P suspended in PBS, with a ratio of 5mg:1ml, was used to extract the cells from the periodontal ligament tissue that is obtained from extracted wisdom teeth. The extraction of the tooth periodontal ligament is described in appendix B.3 and in Somerman et al. (402) The enzyme, collagense P, digests the collagen matrix which holds the cells together and the enzyme required 15 to 45 minutes

o at 37 C at 5%CO2 in order to complete this process.

The resulting cells are a mixture of cell types, and they are not purified into one cell type, as a mixture of cells is more accurate biologically and will give more relevant results as a variety of relevant molecular signals remains.

The HPLSC’s are transferred to a cell culture medium of α-MEM with 20% FBS, 2nm of L- glutamine and 100 units/ml of penicillin/streptomycin; this is the standard cell culture medium (CCM) used in this study. (113) This fluid is then centrifuged at 1000 rpm for ten minutes and the excess CCM is removed leaving a cell pellet. Afterwards 5ml of fresh CCM was added to the cell pellet and stirred together. The cell suspension was then passed through a 70µm nylon mesh cell strainer (Falcon) to leave a single cell suspension ready for the cell cultivation.

104

The single cell suspension was placed into a cell culture flask in an incubator at 5% CO2 and 37oC. After 24 hours the cells will have been firmly attached to the bottom of the flask and will have started to elongate. This behaviour can be observed under a light microscope confirming that the cells are alive and in suitable growing conditions with no toxicity issues.

When the cells have almost covered the bottom of the cell culture flask (80-90% confluence) the cells need to be detached using trypsin/EDTA and the resulting cells should be split between two fresh flasks after adding fresh CCM to re-suspend the cells. This process of removing the cells, re- suspending them and splitting them between two new flasks is called passage. More than four passages will cause the cells to differentiate and lose their original character and that is why cells are used as this point in the cell culture studies. This process of obtaining the cell suspension is described by Somerman et al (402) in general and specifically in terms of a study carried out on BBG scaffolds in the work of Newby et al. (113)

4.4.2 In vitro cell seeding using human periodontal ligamental stromal cells

BBG scaffolds prepared by the foam replication technique, as described in section 4.2.1, and silver MSIE scaffolds, as described in section 4.2.2.I, of all three main study concentrations immersed for the four time periods (in total 13 different sample types) with dimensions of 2x2x2mm3 were sterilized using ultra violet radiation at 725nm for 45 minutes.

The sterilized samples were then placed into sample tubes with the lids, which have a hole in the top in order to pipette in 1 ml of the HPLSC’s suspension and top them up with another ml of CCM, which was described in section 4.4.1. This amount equates to approximately 2x105 cells from passage four which equates to a cell density of 500,000 cells. The hole in the tube’s lid is then sealed with breathable film and placed into a rig (developed in-house in the department of Oral Biology, University of Leeds) that will rotate the tubes to keep a constant flow of cell suspension moving over and through the samples; this is known as dynamic cell seeding. This rotating rig was then placed in an incubator at

o 5%CO2 and 37 C for three days. The dynamically seeded scaffolds were then removed from the rotating rig and were statically cultured for a further seven days in 24well well plates in an incubator at 5%CO2 and 37oC with fresh CCM.

105

4.4.3 Biological characterisation using human periodontal ligamental stromal cells

To ascertain the cells viability, after seven days, the seeded scaffolds were stained with cell tracker green and Eh1. The cells were incubated for 45 minutes in plain CCM containing 5µl:5ml of CMFDA 2.5µl:5ml. The seeded scaffolds are washed off for 45 minutes in plain CCM before viewing the samples using a confocal laser microscope (CLSM).

Cell viability assays are classified into four categories:-

 Cytolysis or membrane leaking assays (403) (404)  Mitochondrial activity or caspase assays (405)  Functional assays (406)  Genomic and proteomic assays (407)

The cell tracker green stains the living cells green and is a functional assay aimed at showing the proliferation of the target cells and Eh1 stains the dead cells red by binding to the DNA in dead or dying cells through holes in the cell membrane making it a cytolysis assay. Eh1 was chosen as it detects all dead cells and not just those that have undergone programmed cell death. By complementing it with a dye that stains only living cells it gives a complete picture of which cells are living and dying. This is shown in the results is stacked up images with a penetration depth ranging between 1.06mm to 1.55mm.

4.4.4 Cell preparation and extraction for rat bone-marrow stromal cells and human bone-marrow stromal cells

The cells are prepared and extracted in a similar way to the HPLSC’s described in section 4.4.1 with some exceptions and differences. The CCM used for this cell biology work is DMEM/F-12 with 10% FBS, 1% L-glutamine and 1% penicillin/streptomycin unless otherwise stated.

The rat bone-marrow stromal cells (rMSC’s) and the human bone-marrow stromal cells (hMSC’s) where extracted and prepared as described in appendix B.4 and B.5, respectively, and cultivated up to the fourth passage, as described in section 4.4.1 for the HPLSC’s.

The cells are cultivated in the same manner as described in section 4.4.1 in an incubator with

o 5%CO2 at 37 C until at least 90% confluence is achieved. Cells are then ready to be used in the experiments described in sections 4.4.5 and 4.4.6 after cells have been counted. Cells are counted to make sure that an equal number of cells are present in each well at the start of each study to achieve an accurate comparison between samples.

106

A digital cell counting device was used to count the cells in suspension. For this study, cells were centrifuged at 1000rpm for five minutes to create a cell pellet. The old CCM was removed and 5ml of fresh CCM is added to the cell pellet and mixed thoroughly to “un-make” the pellet. Then 10µl of this cell suspension was pipetted along with 10µl trypin blue (cell stain) into an epidoth tube and the suspension mixed thoroughly. 10µl of this stain cell suspension were pipette into the counting device which will give the quantity of cells in cells per ml of CCM. From this result, the amount of cell suspension per well can be calculated and the required amount can be transferred into each well depending on the study being carried out. In this work the base line of approximately one hundred thousand cells were suspended in 200µl cell suspension.

4.4.5 2D cell cultures using rat bone-marrow stromal cells and human bone-marrow stromal cells

The cells, either rMSC’s or hMSC’s, need to be seeded onto the bottom of wells of 12well well plates before being put in contact with the scaffolds. To give 20 thousand cells per well 40µl of the relevant cell suspension needs to be transferred into each well. This is then topped up with 2ml of CCM. The well plate is then inspected under a light microscope to check the cell suspension and then placed

o in an incubator with 5%CO2 at 37 C for 24 hours to allow the cells to form a monolayer at the bottom of each well.

The BBG scaffolds produced and modified using MSIE, as described in sections 4.2.1 and 4.2.2, were subjected to cell biology tests using rMSC’s and hMSC’s. Including the BBG scaffold, four different sample types were investigated in this part of the work. The dimensions of the scaffolds were between 5x5x5mm3 and 3x3x3mm3. Scaffolds were conditioned using the washing protocol outlined in Table 7. The first four steps, are established in section 4.2.2, were required to remove excess salt residue from the MSIE process and the rest of the steps were required to prepare the scaffolds for the cell biology work.

o The scaffolds were then autoclaved at 120 C for ten minutes in H2O steam for sterilization purposes, immersed in DMEM/F-12 with no additives for an hour, dried in a fume hood for an hour and then used in the preliminary 2D culture study.

107

Table 7:- Washing protocol used for the cell biology work using rMSC’s and hMSC’s

Washing Solution type/drying period Immersion period Quantity used step type (minutes) (ml) 1 Distilled water 15 50 2 70% Ethanol 10 50 3 Distilled water 10 50 4 Drying in a fume cupboard at least 24 hours n/a 5 2.5% SDS 5 25 6 Millipore distilled water 1 50 7 5% Extran 5 25 8 Millipore distilled water 1 50 9 Millipore distilled water on a shaker 15 50 10 Drying in a fume cupboard at least 12 hours n/a

For cell culture, the CCM from the incubated well plate needs to be removed with a pipette and the wells with the monolayer of cells in the bottom are gently washed with PBS before 2ml of fresh CCM is added. A well insert is placed into each well that will contain a sample, with three of each of the four sample types and one well free to act as a well insert control, leaving at least two wells free to act as a media and cell suspension control, respectively. One scaffold was placed into each well before the wells are topped up with another 1ml of CCM. This is depicted in Figure 24.

o The whole well plate ensemble was placed in an incubator with 5%CO2 at 37 C for the duration of the study, replacing the CCM every 24 hours. When fresh CCM is required, the old CCM is pipetted out of the well, the well insert is removed and placed into another sterile well plate for holding, whilst the cell well plate is gently washed in PBS and 2ml of fresh CCM is added. The well inserts were then replaced into the cell well plate and each well was topped up with another 1ml of CCM before being

o placed back into the incubator at 37 C with 5%CO2.

Figure 24:- Preparation of the 2D culture, showing the placement of the well insert and the sample scaffold

108

After that fresh CCM was added to the well plate, but before the well inserts are replaced the monolayer of cells, as shown in Figure 24, can be observed under a light microscope to gauge how the cells are behaving and to assess the cell response to the presences of the scaffolds in the well inserts. The well inserts can then be replaced and the whole ensemble is returned to the incubator, as described in the previous paragraphs. This preliminary 2D culture method is depicted by the flow diagram shown in Figure 25.

Figure 25:- Flow chart showing the initial protocol used for the 2D culture study

The method described in the beginning of this section, shown in Figure 25, was modified after the preliminary study was completed due to the high cell death due to the pH increase caused by the presence of the scaffolds in the well inserts. The pH increase in the area surrounding a BBG sample has frequently been reported in the literature. (182) (408) As described in section 6.3.2, the following changes were made to the protocol to tackle this issue, as outlined in Figure 25:-

 Removal of washing protocol steps 5 to 10 in Table 7 as these steps were considered to hinder the reduction of the pH in the CCM when the scaffolds were added  1% Bufferall was added to the CCM and for the rest of the study to help to buffer the pH to a more suitable level  Reducing the scaffold size to between 2x2x2mm3 and 3x3x3mm3  Increasing the pre-culture stage after autoclaving the scaffolds from an hour in DMEM/F-12 to four days in DMEM/F-12 with the first 24 hours spent being agitated

The final version of the 2D cell culture process is shown in Figure 26.

109

Figure 26:- Flow chart showing the optimised protocol used for the 2D cell culture study

At certain time points AB assays were done to ascertain the change in the number of living cells present in the well. This is done when fresh CCM is due to be added as the well inserts have to be removed in order to carry out the protocol outlined for AB assays in section 4.4.7. The AB assays were completed at 1, 7, 14 and 21 days into the study.

4.4.6 3D cell cultures using rat bone-marrow stromal cells

This 3D cell culture protocol involves statically seeding the rMSC’s directly onto a 3D scaffold meaning the cells will be in direct contact with the scaffold for this study. This protocol follows the work done in the previous 2D culture study and therefore uses the revised conditions, such as the new scaffold dimensions (between 2x2x2mm3 and 3x3x3mm3) and the addition of 1% Bufferall to the CCM.

In this part of the study, BBG scaffolds produced and modified using MSIE, described in sections 4.2.1 and 4.2.2 were subjected to cell biology tests using rMSC’s. Including the BBG scaffold, four different sample types were investigated in this section of the work. The dimensions of the scaffolds were between 3x3x3mm3 and 2x2x2mm3. Scaffolds were conditioned using the washing protocol outlined in the washing and pre-culturing parts of Figure 26.

The cells were prepared as stated in section 4.4.4 and approximately 7.5x104 cells were seeded onto each scaffold. Three of each scaffold type (12 scaffolds in total) were placed into a 12well well plate with one scaffold per well. 2ml of the cell suspension was slowly pipetted onto each scaffold to

110 give the cells the best chance of infiltrating the scaffold and attaching themselves to the inside network of pores of the scaffolds. The well was then topped up with an additional 1ml of the CCM.

o The whole ensemble was then placed into the incubator with 5%CO2 at 37 C for 24 hours. The cells settled and started to attach themselves to the well and the scaffold after three hours so fresh CCM was added after 24 hours and subsequent 24 hours were considered until the end of the study without disturbing the cells.

AB assays, as described in section 4.4.7, are required at set time points in order to assess the changes in the number of living cells present in the well including 1,7 and 14 days. These assays were completed when the CCM was replaced as described in section 4.4.5.

The protocol for the 3D cell culture is summarised in the flow chart shown in Figure 27.

Figure 27:- Flow chart for the 3D cell culture study

4.4.7 Biological characterisation using rat bone-marrow stromal cells and human bone-marrow stromal cells

In this part of the study the main characterisation technique, other than observations under a light microscope, was the assessment using an Alamar Blue (AB) assay.

For this study, the CCM in the well plate was removed and the samples were gently washed in PBS before 1ml of a 10% AB solution was added to each well. Two controls were prepared in the same well plate with one containing just 1ml of the 10% AB solution and the other having 1ml of the 10% AB solution and 1ml of the CCM.

111

The well plate was then covered in foil, as the AB solution is light sensitive, and incubated for

o two to three hours at 37 C and 5%CO2. By the end of this incubation period the solution in the sample wells will have changed from dark blue to a dark pink purple colour.

After the incubation period three 100µl samples of the AB solution were removed from each sample well and the two control wells and they are pipetted into a 96well well plate for the resulting measurements. The sample well plate had fresh CCM media added to the sample wells and it was placed

o back into the incubator at 37 C and 5%CO2 for the rest of the study. The result well plate was wrapped in foil to prevent fluorescent interference from adjacent wells.

The relevant measurement is the absorption of the stained cell suspension from each well at 570nm and 600nm which is measured using a spectromax 190 instrument with the measured absorption results being used in Equation 11 and Equation 12 to give the percentage reduction of the AB.

The number of living cells present in the wells is related to the percentage reduction of the AB. The higher the percentage reduction the more living cells are present in the sample well. Thus this test is only indicative of the number of living cells present at the time of the measurement. The AB assay can be used to assess how the cell population is changing over time, with respect to the sample type and it can give an indication if there are any toxicity issues with the sample.

Equation 11:- AB reduction equation

Equation 12:- Correction factor for the AB reduction equation

112

4.4.8 Bacterial preparation, counting and growth curves

Standard I Agar plates (agar plates) need to be prepared first to give a surface for the bacteria to grow on. The preparation of these plates is described in appendix B.6.

A stock bacteria plate needs to be cultivated to provide the bacteria for the later studies. The bacteria, listed in section 4.1.9, are removed from the freezer and placed in a sterile fume hood which is dedicated to bacterial studies. 100µl of the bacteria are removed once the stock has become liquid, the stock bacteria is then replaced back in the freezer. After that this stock bacteria is pipetted on to an agar plate before using a platinum loop to spread the bacteria around the agar plate. The plate is then labelled, sealed with Parafilm® and placed in an incubator at 37oC for a minimum of 48 hours to cultivate.

The wavelength at which the absorbance is measured needs to be established for the two types of media being used in the bacterial work. The two media that the bacteria will be suspended in:-

 An industry standard, Muller Hinton broth (MHB)  A modified version of the CCM used in the cell biology studies described in sections 4.4.4 onwards containing DMEM/Ham’s F-12 with 10% FBS, 1% L-glutamine and 1% Bufferall (DMEM/F-12).

The absorbance of MHB, DMEM/F-12, pure distilled water and the McFarland standard over a range of 350-800nm was measured using a Perkin Elmer Enspire 2300 multiwave reader coupled with Enspire management software. This will provide spectra of absorbance against wavelength to establish the wavelength that future absorbance values should be measured at. From this result, shown in section 6.4.1, future absorbance values were measured at 650nm to give an expected absorbance for both media types of approximately 0.1.

For the experiments, a pre-culture stock solution is created by adding three to five colony forming units (CFU’s), from the stock bacteria plate to 10ml of media (either MHB or DMEM/F-12).

o Once the bacteria is added to the media the pre-culture stock is incubated with 5% CO2 at 37 C on a shaker for one hour for MBH pre-culture solutions and two hours for DMEM/F-12 pre-culture solutions. The absorbance of the pre-culture suspension and the media is measured at using a Specord 205 photometer at 650nm. The difference in absorption at 650nm before and after the bacteria were added to the media, Δabs, is used to calculate the amount of the pre-culture stock that should be added to the stock solution, using Equation 13. Throughout the studies it will be stated what absorbance difference is used for that particular study. In general the required absorbance difference will either be 0.01 or 0.001.

113

Equation 13:- Calculation of the amount of pre-culture stock solution which is required to be transferred to the stock solution

Subsequently, the required pre-culture stock solution is added to the stock solution and

o incubated with 5%CO2 at 37 C until it is needed.

Growth curves of E.coli have to be established to predict the bacteria’s behaviour in DMEM/F- 12 compared to the industry standard MHB.

The change in absorption at 650nm from media to the stock suspension is measured at a number of time points to establish the growth of the bacteria. This will provide a graph of the change in absorption against time but won’t provide the number of bacteria per ml. This value is evaluated by removing the stock suspension at specific time points, diluting it with NaCl and plating 100µl of these dilutions onto agar plates and transferring into an incubator at 37oC for 24 hours to cultivate.

The dilutions used depend on the number of bacteria found in the bacteria suspension at the time of the absorbance measurement. Using the bacteria counting protocol, outlined in appendix B.7, the number of bacteria in the pre-culture suspension was found to range between 1x108 and 3x108 bacteria per ml. Equation 14 is used to calculate an approximate number of bacteria per ml at the given time point and the number of dilutions needed to produce a plate on which bacteria’s CFU’s can be counted is calculated.

Equation 14:- Number of bacteria per ml to work out dilutions for plating the bacteria

A typical dilution set is shown in Table 8 and gives approximate values that are predicted to be obtained based on the values calculated from Equation 14 and will give the maximum dilution values that are expected to be encountered during these studies. Only the last four dilutions of any dilution set should be plated. The last four dilutions of any sample should be in the magnitude range of 101, 102, 103 and 104 as shown by Table 8 and illustrated in Figure 28.

114

Table 8:- Working dilution table of the dilutions required to get the required number of CFU’s on an agar plate

Dilution Number of CFU's wanted Predicted magnitude of CFU's Predicted dilution factor number on the agar plate on the agar plate per 100µl D8 Below 10 (*) 101 10-7 D7 Between 10 and 100 (*) 102 10-6 D6 Between 100 and 1000 (*) 103 10-5 D5 Between 1000 and 10000 (*) 104 10-4 D4 Not plated 105 10-3 D3 Not plated 106 10-2 D2 Not plated 107 10-1 D1 Not plated 108 100 * 100µl of this dilution is plated, incubated for 24 hours and the results CFU's are counted

Figure 28:- Colony forming units (CFU's) of E.coli showing the process of diluting the E.coli suspension, plating it and then counting the CFU’s after incubation

115

After 24 hours the cultivated plates are removed from the incubator and the CFU’s are counted as shown in Figure 28. Using Equation 15 the number of bacteria per ml present in each plate can be calculated. This can be plotted against time to give another growth curve that can be used to determine the best growing conditions for the bacteria.

Equation 15:- Number of bacteria per mL from counting CFU’s

4.4.9 Minimum inhibitory concentration testing using metal salts against Escherichia Coli

To calculate the minimum concentration of metal salts required to cause either a bacteriostatic or bactericidal effect a stock solution needs to be prepared as well as agar plates as described in section 4.4.8 and appendix B.6, respectively.

Dilutions of several metal salts were prepared using distilled water to a 10mM concentration before using either MBH or DMEM/F-12 (depending on the study), to bring the dilutions to their final concentrations. Table 9 shows the concentrations that were investigated in this work using the protocol outlined in this section. These concentrations were chosen based on previous work in the literature, in particular the work reported in Heidenau et al. (409) The metal salt dilutions will be further diluted when they are added to the bacteria suspension and this was taken into account when making the metal salt dilutions, as shown in Table 9.

Several studies were carried out using the method to be outlined including investigating if

Cu(NO3)2 can be used in the main inhibition study instead of the more commonly used CuCl2, as used

in Heidenau et al (409) and investigating the minimum inhibitory concentration (MIC) of AgNO3,

Cu(NO3)2 and a 1:1 ratio of both salts to represent the three different MSIE scaffolds produced in this work and described in section 4.2.2. This investigation will act as a comparison to the study described in section 4.4.11 where the supernatants from immersed MSIE scaffolds were tested to determine their anti-bacterial potential.

The metal salt dilutions (1ml) were added to 1ml of the stock solution, as described in section

o 4.4.8, and left to cultivate in an incubator with 5%CO2 at 37 C. At certain time points some of the suspension was removed, diluted with NaCl, plated on agar plates and incubated, as described in section 4.4.8. After incubation the CFU’s can be counted and the bacteriostatic effects of the metal salts can be determined for each concentration tested.

116

Table 9:- Metal salts concentrations used in the minimum inhibition study

Metal salt Wanted concentration Working concentration (mM) (mM) 0.05 0.1 0.1 0.2

CuCl2 0.5 1 1 2 0.005 0.01 0.01 0.02 0.02 0.04 0.05 0.1 0.1 0.2

Cu(NO3)2 0.2 0.4 0.5 1 1 2 2 4 5 10 0.005 0.01 0.01 0.02 0.02 0.04

AgNO3 0.05 0.1 0.1 0.2 0.2 0.4 0.5 1 0.01 0.02 Combination of 0.05 0.1

AgNO3 and 0.1 0.2

Cu(NO3)2 0.5 1 1 2

The MIC will be the lowest concentration that produces a bacteriostatic or bactericidal effect on the bacteria. The method described in this section can determine the MIC to cause a bacteriostatic effect on the bacteria as it is not possible to know if the metal salts are killing the bacteria (bactericidal effect) or simply “switching off” the bacteria (bacteriostatic effect) by looking and counting CFU’s. The MIC for a bactericidal effect is described in section 4.4.10.

This study gives thus a narrower focus for the rest of the work by giving a range of concentrations that the metal salts work against the bacteria and the results will be relevant in later studies when the supernatants taken from immersed MSIE scaffolds will be tested, described in section 4.4.11, to confirm if they fall within the minimum inhibitory scale.

117

4.4.10 Testing the bacteriostatic or bactericidal capabilities of metal salts

To test the bacteriostatic/bactericidal capabilities of the concentrations determined through the studies described in section 4.4.9 the following changes to that protocol need to be made. The method used in section 4.4.9 changes at the step where the solutions are due to be plated onto agar plates. The solutions are still diluted and plated as described in section 4.4.9 to produce the series of agar plates used to determine the number of bacteria per ml through counting CFU’s. However an additional plate is made by mixing 10µl of the sample suspension with 100µl sodium thiosulfate in an epidoth tube and transferring this to an agar plate and incubating as normal. Sodium thiosulfate, as used and described by Heidenau et al, (409) deactivates the silver and the copper in the suspensions. This is added to determine if the observed effect on the bacteria is bacteriostatic or bactericidal when compared to the metal salt/bacteria suspensions that did not have sodium thiosulfate added.

If the concentration of the metal salt is bactericidal both the CFU agar plates and the sodium thiosulfate plate will show no bacteria activity. If the concentration of the metal salt is bacteriostatic the CFU plate will show little or no bacteria growth in the form of CFU’s and the sodium thiosulfate plate will show normal bacteria growth coverage as seen in control plates. If the concentration of the metal salts has no effect on the bacteria at all, both the CFU plates and the sodium thiosulfate plate will show normal bacterial growth as seen on a control plate with no metal salts in suspension.

4.4.11 Growth inhibition using supernatants against Escherichia coli

The stock solution used was prepared as described in section 4.4.8 along with the agar plates described in appendix B.6.

This study involves immersing sample scaffolds as described in sections 4.2.1 and 4.2.2 in

o DMEM/F-12 for 1, 3, 7 and 28 days in an incubator with 5%CO2 at 37 C. The samples studied were the BBG scaffold, and the three MSIE scaffold types at the lowest concentrations immersed for 15 minutes. These samples were chosen as they are the same as the ones studied in the cell biology studies, described in sections 4.4.5 and 4.4.6. There were two scaffolds immersed for each sample type meaning that ten samples were tested including two control runs.

After the scaffolds have been immersed for the given time periods 1ml of the supernatant is removed, decanted into fresh tubes, 1 ml of the stock suspension is added and then the tubes are placed

o into an incubator with 9.5%CO2 at 37 C on a shaker for 24 hours.

118

After 24 hours the samples are removed and prepared for dilution and plating, as described in section 4.4.8, to give a CFU count which will give an indication of the bacteriostatic nature of the samples. Sodium thiosulfate plates are also required to ascertain if any changes in the CFU plates for these supernatant samples is due to scaffold supernatant being bacteriostatic or bactericidal as described in section 4.4.9.

4.4.12 Agar diffusion to show growth inhibition of Staphylococcus aureus and Escherichia coli

This section covers the two methods used to determine the growth inhibition of the scaffolds produced (as described in sections 4.2.1 and 4.2.2) to both S.aureus and E.coli through agar diffusion.

I Growth inhibition demonstrated through agar diffusion on Escherichia coli

Agar plates with an E.coli “lawn” need to be prepared in order to test the antibacterial properties of the samples. This is done by preparing the pre-culture stock suspension as described in section 4.4.8, transferring 500µl of this pre-culture stock onto an agar plate as described in appendix B.6. Once this inoculum is on the agar plate it is spread evenly on the surface using a sterile drygalski spatula making sure to distribute the inoculum crosswise in at least three different directions across the plate. These plates are then left to dry in a sterile bacteria specific fume hood for 10 to 15 minute before application of the testing disks.

After the plates have started to dry, sterile testing disks made of filter paper are gently pressed into the surface of the plate using a pair of sterile tweezers, making sure there is plenty of room between each of the disks, meaning there is a maximum of five testing disks per agar plate.

20µl of the sample solution was pipetted onto each disk, taking care to get the inoculum only on the disk, and the plates were then incubated for 24 hours at 37oC. After the incubation period the plates were observed looking for a halo effect around the testing disks. If the samples are having an effect on the bacteria, whether it is bacteriostatic or bactericidal, then there should be a clear E.coli free zone around the testing disk.

The following sample types were tested in this way to ascertain if they were having an antibacterial effect on the E.coli. The supernatants taken from the immersed scaffolds described in section 4.4.11 were tested along with a range of concentrations of metal salt dilutions as described in 4.4.9 for comparison purposes.

119

II Growth inhibition demonstrated through agar diffusion on Staphylococcus aureus

As described in section 4.4.12.I, a bacteria lawn needs to be created first on an agar plate. For this agar diffusion study the bacteria lawn was created by growing S.aureus instead of E.coli.

Once the bacteria lawn had been created and allowed to dry small scaffold pieces were gently pressed into the agar plate. The four samples tested in this study were a BBG scaffold along with the lowest concentration (immersed for 15 minutes) MSIE samples, as described in sections 4.2.1 and 4.2.2, respectively, and in line with the previous bacterial work using the supernatants (section 4.4.9) and the cell biology work (sections 4.4.5 and 4.4.6). The S.aureus lawn plates with the scaffold samples are then incubated for 24 hours at 37oC before the observations of the anti-bacterial effect of the samples on the S.aureus lawn are made. As in section 4.4.12.I a halo around the scaffold is the desired observation as this would indicate that the sample is having an anti-bacterial effect on the S.aureus.

4.4.13 Growth inhibition on 3D scaffolds using Klebsiella pneumoniae

For this part of the study, a 50µl suspension of K.pneumoniae in 0.9%NaCl + 0.05% Tween 80 was pipetted into each sample scaffold. The four samples tested in this study were a BBG scaffold along with the lowest concentration (immersed for 15 minutes) MSIE samples as described in sections 4.2.1 and 4.2.2 respectively and in line with the previous bacterial work using the supernatants (section 4.4.9) and the cell biology work (sections 4.4.5 and 4.4.6). The inoculated scaffolds were then incubated for 18 hours at 37oC.

The scaffolds were eluted using 20ml of a mixture of 0.9%NaCl + 0.20% Tween 80 and the extracted solution was diluted and plated in the same manner as described in section 4.4.8. After these agar plates were incubated for 24 hours at 37oC, the CFU’s were counted and compared to a control sample of a sterile standard textile disk (PES). The number of bacterial cells was determined and compared to the number of bacteria added before the incubation period on the sample. If this number decreases then there has been some anti-bacterial effect caused by the scaffold and if the number of bacteria increases then the presence of the scaffold has had no anti-bacterial effect.

120

4.5 Statistical Analysis

Statistical analysis has been carried out on the collected data where appropriate using Excel. The statistical methods are shown in Appendix E which includes all of the relevant equations used. All data is expressed as the mean ± standard error, as described in appendix E, and unless otherwise stated the number of samples used to determine these values is greater than or equal to five.

To determine if the statistical relevance of the data, analysis of variance (ANOVA) statistical analysis, using Excel, was carried out where p ≤ 0.05 is regarded as a statistically significant result. The method is described in detail in appendix E.

121

Chapter Five

5. Structural results and discussion

5.1 General remarks

This chapter covers the structural characterisation studies carried out on the modified BBG scaffolds developed in this study, namely the MSIE scaffolds, polymer coated scaffolds and BBG samples. This chapter covers the use of a number of techniques which have been introduced in chapter four, which enable the complete structural characterisation of the novel scaffolds developed in this work.

This chapter is split into a number of sub-sections dealing with specific scaffold types and reports the development of the novel scaffold types from conception to fine tuning.

5.2 45S5 Bioglass®- derived glass-ceramic scaffolds

The comprehensive structural characterisation of the BBG scaffold is required as this material will act as a control for the modified samples that have been developed during the course of this work. This section presents the structural characterisation of these samples using the materials and techniques described in chapter four.

Sintered BBG scaffolds produced using the foam replication technique, as described in chapter 4.2.1, gave an average porosity of 92±3% which is consistent with the results found in the literature (4) (410) and as calculated using Equation 1. The un-sintered green body scaffolds were found to have an average porosity of 93±2% indicating that during the binder burn out phase and the subsequent sintering phase the samples are not losing pore volume (within the accuracy of the measurement), through the shrinkage of the scaffold.

From examining the BBG scaffolds using a scanning electron microscope (SEM), as described in chapter 4.3.3, a number of observations can be made. Figure 29 shows a range of SEM images taken of BBG scaffolds showing that the pores of the sintered scaffold range from approximately 200 to 600 microns (Figure 29A) which is consistent with results presented in the literature on this type of scaffolds. (411) (4) (410) (269) (5) (270) (271) Figure 29B shows that the cross sections of the sintered struts are hollow after the PU foam and binder have been burnt out and that the struts have transactional dimension

122 in the range of 25 to 75 microns which is in line with current research. (411) (4) (410) Figure 29 C and D show that the surface topography of the scaffold is not smooth and this rough micro-topography is considered important as it should encourage cell attachment (222) (223) (214) and accelerate the reactions occurring at the surface of the scaffold, (223) which has also been reported in the literature. (222) (223)

Figure 29:- SEM images of BBG scaffolds produced using the foam replication technique at magnifications of A) x50, B) x300, C) x1400 and D) x10000

As described in chapter 4.3.10, the mechanical competence of the BBG scaffolds can be measured by compressive strength testing and this has given rise to a compressive strength of 0.53±0.08MPa which falls into line with results from the literature. (113) (4)

The wettability of BBG samples was determined by measuring the static contact angle on pellets, as described in section 4.3.8. A value of 28±4° was measured, which is in agreement with literature results. (412)

White light interferometry was used to examine the surface roughness of the sintered 45S5 Bioglass® using pellets fabricated using the same process used to make the scaffolds. The measurements were carried out as described in section 4.3.7. The white light interferometry results can be split into

123 quantitative and qualitative results with the qualitative results shown in Figure 30. Figure 30 shows a range of different plots obtained from this technique that can be used to ascertain the topography of the sample. Figure 30a) shows the white light microscopy intensity plot which gives an indication of areas of interest on the samples surface. For the BBG samples little can be seen from this plot and the surface appears fairly smooth but once the settings are changed to show a solid colour plot of the sample area (shown in Figure 30b), the contours of the sample are now highlighted with the peaks and valleys shown in yellow and black, respectively. This image shows that whilst the surface of the pellet appears to be smooth to the naked eye it does actually contain a rough topography, as shown also in both SEM images in Figure 29.

Figure 30c) shows a 3D representation of the surface of the sample showing the change in the surface topography through the use of colour. This data is in agreement with the topography shown in Figure 30b) and in Figure 29. Figure 30d) shows a cross-section of the sample giving an idea of the extent of the surface peaks and valleys which gives rise to the sample PV value. Finally Figure 30e) shows a 2D colour surface map of the sample which shows the peaks and valleys of the sample in terms of a colour scale. Overall Figure 30 shows that a number of different plot types can represent the topography of the sample and in this case it can be seen that apart from one significant peaked area the surface of the pellet appears to have a micro-roughness which is one of the features that is desired for an ideal scaffold for BTE. (222) (223)

From this qualitative data shown in Figure 30 the quantitative results were derived and the average values of the RMS, Ra and PV parameters were found to be 3±1µm, 2.6±0.9µm and 19±9µm, respectively, which is consistent with current data available in the literature on similar bioactive glass- derived scaffolds. (224) (225)

As stated in chapter four, this characterisation technique was carried out on pellets rather than on the foam replicated scaffolds as the measurement can not easily be carried out on a curved surface such as the strut of a scaffold. As an approximation, the pellets can be used as a direct substitute to investigate the surface topography at the micro level because the pellets are prepared and sintered in the same way as the foam replicated scaffolds and thus a similar surface topography is expected.

124

Figure 30:-White light interferometry results for a sintered BBG pellet showing A) a 2D white light microscopy intensity plot, B) a 2D solid plot coloured version of the white light microscopy intensity plot shown in A), C) a 3D oblique plot showing the surface roughness of the sample in terms of colour, D) a sample cross section across the pellet showing the surface profile and E) a 2D coloured surface map of a BBG pellet showing the peaks and valleys in terms of different colours

125

The next set of characterisation techniques investigate the chemical composition and crystalline structure of the BBG samples using EDX, XRD and FTIR as described in sections 4.3.4, 4.3.5 and 4.3.6, respectively. Figure 31 shows the FTIR spectrum of sintered BBG exhibiting the characteristic bond peaks of this material which have been shown in the literature. (413) (414) (415) Relevant peaks include a pair of asymmetric stretching Si-O-Si bonds at 1097 – 1100cm-1 and 1037 – 1050cm-1, a peak showing a stretching Si-O bond at 920 and 940cm-1 which is indicative of a non-bridging oxygen bond, a pair of P-O bending bonds at 720 – 650cm-1 and 625 – 580cm-1 and finally a pair of Si-O-Si bending bonds at 528 – 536cm-1 and 455 – 457cm-1. These four sets of bond peaks should always be present as these are the characteristic peaks for sintered BBG, which is the base material in the present research and the results are supported by previous related studies in the literature. (413) (414) (415)

Figure 31:- FTIR analysis of sintered BBG scaffold including the associated characteristic bond peaks (413) (414) (415)

From the SEM images shown in Figure 29C and D small crystalline grains can be identified which have a diameter of approximately 0.5µm, leading to the conclusion that the sintering process leads to partial crystallisation of the scaffold. This is confirmed by the XRD characterisation shown in Figure 32. Figure 32 confirms the crystalline nature of the sintered BBG samples, the XRD pattern exhibits the characteristic peaks previously recorded for Na2Ca2Si3O9, which in the literature has been shown to be the main crystalline phase of sintered 45S5 Bioglass® after heat treatment at over 100oC. (4) (5) (389) (416)

126

Figure 32:- XRD spectrum of sintered BBG scaffold (at 1100oC for two hours) confirming the crystalline nature of the material (4) (5) (389) (416)

EDX analysis of the BBG scaffold (Figure 33) showed the presence of silicon, sodium, calcium, phosphorus and gold, (the presence of gold is due to the coating applied for SEM observations as described in section 4.3.4). The other elements detected are from the BBG scaffold and are consistent with the literature. (394) (417) (5)

Figure 33:- EDX analysis of a BBG scaffold to determine the elemental composition of the sample

The next stage of the characterisation of the BBG scaffolds is to assess their bioactivity in order to determine if the scaffolds exhibit the expected behaviour in SBF, characterised by the formation of

127

HA on their surface after a short immersion period in SBF. Bioactivity of a material, in the context of bone substituting materials, is assessed by immersing the samples in SBF, as described in section 4.3.2 and by Kokubo et al, (396) for a set time frame. Once removed, the samples are characterised using the same techniques that were used for sample characterisation before immersion. Immersion in SBF for 1, 3, 7, 14, 21 and 28 days is considered to track the development of HA on the surface of the samples, as this is the marker of bioactivity and it is required to promote bone cell attachment as discussed in chapter 2.

Figure 34 shows through SEM images the progression of HA formation on the surface of the BBG scaffold. All six immersion periods show the characteristic HA knobbles on the surface of the scaffold and the increased frequency and density of these knobbles as the immersion period in SBF increases, which is as expected and consistent with the literature. (4) (5) With the presence of characteristic HA formation on the surface of the scaffold it can be determined that the scaffold is in principle suitable for use in BTE. Figure 34 also showed that the samples have a micro-roughness, due to surface reactions leading to the formation of HA, which is one of the desirable features for an ideal scaffold for BTE. (222) (223)

The porosity of the BBG scaffolds, as determined by using Equation 1, declined the longer the scaffolds were immersed in SBF. Decreasing from an average pre-immersion porosity of 93±2% to an average post-immersion porosity of 92.0±0.8%, 91±1%, 89±2%, 84±2%, 80±3% and 79±2% for immersion periods of 1, 3, 7, 14, 21 and 28 days, respectively, which is in line with observations reported in the literature. (241)

The compressive strength of the immersed scaffolds was observed to decrease with increased immersion period, as displayed in Table 10. This is to be expected as the scaffolds will start reacting (ion release) as soon as they are immersed in SBF and the precipitated HA does not provide mechanical reinforcement. (352) (418) The slight reduction in porosity does not counteract the effect of scaffold degradation.

The surface wettability of the BBG pellets was assessed after 14 days of immersion in SBF using the method described in section 4.3.8 and the contact angle was found to be 23±2 degrees, being consistent with available results in the literature. (419) This contact angle is important as hydrophilicity has an effect on the attachment of cells. Contact angles obtained before and after immersion in SBF should be suitable for cell attachment as ideally the contact angle should fall between 48 and 62 degrees in order for the surface to be exhibit the desirable moderately wettable behaviour. (419)

128

Figure 34:- SEM images of BBG scaffolds after immersion in SBF for A) one day (magnification x600), B) three days (magnification of x1500), C) seven days (magnification of x1500), D) 14 days (magnification of x1500), E) 21 days (magnification of x1500) and F) 28 days (magnification of x170) (arrows indicating HA)

129

Table 10:- Average compressive strength of BBG scaffolds after immersion in SBF. * Indicates a statistically significant change due to the immersion period (p≤0.05) using ANOVA analysis.

Immersion Period Compressive Strength (Days) (MPa) 1 0.52±0.06 * 3 0.50±0.04 * 7 0.47±0.05 * 14 0.43±0.05 * 21 0.35±0.08 * 28 0.24±0.03 *

White light interferometry has again been used to examine the surface roughness of the sintered BBG pellets after immersion in SBF for 1, 3, 7, 14, 21 and 28 days. Table 11 shows the RMS, Ra and PV values for the different SBF immersion periods. It is shown that as the immersion period increases so does the RMS, Ra and PV values which is consistent with the current literature. (224) This increase in the roughness values can be attributed to the precipitation of HA onto the surface of the sample, as seen in the SEM images in Figure 34 for the scaffolds.

Table 11:- White light interferometry results for the surface roughness of BBG pellets after immersion in SBF for a number of days. * Indicates a statistically significant change due to the change in immersion period (p≤0.05) using ANOVA analysis.

Immersion period RMS Ra PV (days) (µm) (µm) (µm) 1 4±1 * 3±2 * 20±5 * 3 4±1 * 3±1 * 22±4 * 7 6±2 * 5±1 * 24±7 * 14 6±1 * 6±1 * 26±9 * 21 7±2 * 6±1 * 27±8 * 28 7±1 * 7±2 * 29±5 *

The chemical composition at the scaffolds surface after immersion in SBF has been examined using EDX and the resulting spectra are shown in Figure 35. Compared with Figure 33, Figure 35 shows the lack of sodium, found at approximately 1keV in Figure 33. Increased levels of both calcium and phosphorous are seen in all immersion periods culminating and levelling out after 7 days of immersion in SBF and this indicates the precipitation of HA from as early as 1 day as seen in Figure 35A).

As stated in chapter two the Ca/P ratio can be considered to determine if HA is present on a sample as this ratio for stoichiometric HA should be approximately 1.67 for stoichiometric HA. (420) The Ca/P ratio can be seen to increase in Figure 35, which indicates that HA is formed. The changes to

130 the elemental composition, detected by EDX, are consistent with results in the literature. (421) (422) (420)

Figure 35:- EDX analysis of a BBG scaffolds which have been immersed in SBF for A) 1 day, B) 3 days, C) 7 days, D) 14 days, E) 21 days and F) 28 days showing high levels of Ca and P which confirms the formation of HA

Figure 36 shows the FTIR analysis for BBG scaffolds which have been immersed in SBF for 1, 3, 7, 14, 21 and 28 days. The change in the FTIR spectra, shown in Figure 31 (before SBF immersion), compared to those shown in Figure 36 (after SBF immersion) indicates the exchange of ions from the surface of the scaffold into the SBF and the precipitation of HA onto the scaffolds surface. (423) This continuous exchange of Na, Ca, P and Si leads to the bonds rearranging themselves leading to changes in the FTIR spectra, shown in Figure 31, as the immersion period in SBF increases. (423)

The four important peaks, as indicated in Figure 31, are present in Figure 36 and they still dominate with the Si-O-Si and P-O bonds but a change is observed due to the formation of HA. (423) (424) The asymmetric stretching Si-O-Si bonds are still present between 1100cm-1 and 1037cm-1 but they decrease in intensity over the SBF immersion period due to Si leaching out of the scaffold. (424) (222) The stretching Si-O (the non-bridging oxygen) bond has shifted from 940cm-1- 920cm-1 to 980cm- 1 – 940cm-1and decreases in intensity during the immersion period. This is due to the Si-O bond changing to a bending Si-O-Si bond as the HA forms. This is one of the changes that is indicative of HA formation on the surface of the scaffold and therefore its bioactive capabilities. (423) (424) (222)

131

Moreover a new peak is formed during SBF immersion between 880cm-1 and 900cm-1, which is indicative of an out of plane C-O bending bond which is part of a carbonate group. The presence of this new peak, in Figure 36, suggests the presence of a carbonate apatite on the surface of the scaffold, which is another indicator of the materials bioactivity. (425)

Figure 36:-FTIR analysis of BBG scaffolds immersed in SBF for 1, 3, 7, 14, 21 and 28 days

The P-O bending bonds between 875cm-1 to 825cm-1, 760cm-1 to 650cm-1 and 625cm-1 to 575cm-1, show in Figure 31, are all still present in Figure 36. However the bonds, at 875cm-1 to 825cm- 1, increase in size during the immersion period which indicates the presence of a precipitated HA layer. (423) (424) (222) Moreover peaks indicating a dual P-O bond, between 760cm-1 to 650cm-1, are seen to merge to form one peak which suggests the crystallisation of the precipitated HA. (424) Finally the P- O bond shown between 625cm-1 to 575cm-1 remains constant in its intensity and position in the spectra and is indicative of BBG. The bending Si-O-Si bonds between 536cm-1 and 457cm-1 are still present, in Figure 36 , and they increase in intensity with the immersion period. (423) (424) (222) All of the changes described above are indicative of the precipitation of HA onto the surface of the BBG scaffold after immersion in SBF for a prolonged period and are consistent with previous behaviour reported in the literature. (423) (424) (222) The presence of these surface changes thus confirm the bioactivity of the BBG scaffold and it measures the bioactivity of the modified BBG scaffold produced for this work.

132

Figure 37 shows the XRD analysis for BBG scaffolds after immersion in SBF for 1,3,7,14,21 and 28 days. It can be seen that as the immersion period increases the crystalline structure indicative of BBG, shown in Figure 32, is changing and the characteristic peaks for HA can be seen appearing after 1 day in SBF. However the change in the spectra is most apparent after 14 days of immersion in SBF as the crystalline structure has disappeared to such a point that only HA peaks are observed which is consistent with previous work reported in the literature. (5) (4) (426)

Figure 37:- XRD analysis of BBG scaffolds which have been immersed in SBF for 1, 3, 7, 14, 21 and 28 days

The pH of SBF and the cell culture media (CCM) used in some of the in vitro cell biology work (DMEM/F-12) was measured to investigate the change in pH due to the presence of the BBG scaffold over time. Figure 38 shows that the pH of the SBF and the CCM increased statistically significantly once the scaffold was introduced. This observation agrees with the literature (244) (408) and is caused by the release of ions (for example sodium and calcium) from the surface of the scaffold as soon as the scaffold is immersed. (408) There is some advantage to a slight increase in the pH as it has been found to enhance glycolysis and hence cellular ATP production. (182) (427) However the increases in pH to over 9 will cause a cytotoxic effect when the scaffolds are introduced to cells. (408) Ways to reduce the pH through a specific pre-condition of the scaffolds and its effects are discussed in the in vitro results (Chapter six).

133

Figure 38:- pH changes in supernatants due to the presence of BBG scaffolds with A) pH changes in SBF and B) pH changes in cell culture media (DMEM/F-12). * Indicates a statistically significant change between the SBF and CCM (p≤0.05) using ANOVA analysis.

This section gives an overview of the characteristics of the BBG scaffold that acts as a control for the characterisation of the novel and modified scaffolds developed in this work.

5.3 Silver molten salt ion exchanged scaffolds

This part of the chapter presents the results obtained when the MSIE technique was used to introduce silver ions into the structure of the pre-fabricated BBG scaffolds. The “proof of concept” work is shown in section 5.3.1 which is followed by the more extensive structural characterisation of these novel samples shown in section 5.3.2.

5.3.1 Preliminary processing of silver molten salt ion exchange scaffolds

This “proof of concept” study was designed on the basis of the three concentrations originally proposed by Di Nunzio et al (6) but using sintered BBG constructs in place of the glass-ceramic system investigated by Di Nunzio et al. (6) MSIE was used to incorporate silver ions into sintered BBG pellets, as described in section 4.2.2.I. This preliminary study investigated whether MSIE is a suitable method to introduce therapeutic metal ions into the silicate matrix of pre-fabricated BBG scaffolds. Using XRD and EDX techniques, preliminary information about of the success of the technique was acquired.

134

Figure 39 shows a typical EDX spectra obtained for the preliminary study about the use of MSIE to introduce silver ions into BBG pellets. Figure 39A) shows the characteristic peaks for BBG but no peaks indicating the presence of silver. Silver was detected on the surface of the pellets, as shown in Figure 39B) and C). These results show that the EDX analysis did not pick up the presence of silver in the lowest concentration samples, indicating that silver in samples immersed in this extremely low concentration won’t be detected by the equipment. Since the experiments of the two higher concentrations have produced EDX spectra with silver peaks, it could be extrapolated that the lowest silver concentration was also successful in incorporating silver into the structure but the equipment was not sensitive enough to detect this.

Figure 39:- EDX spectra for silver doped BBG scaffolds with concentrations of silver nitrate in the salt bath; A) low concentration of silver nitrate, B) medium concentration of silver nitrate and C) high concentration of silver nitrate as shown in section 4.2.2.I

To confirm that the silver introduced during the MSIE process is incorporated into the glass structure, and not just forming a residual salt layer on the surface, XRD characterisation has been carried out. Figure 40 shows the XRD patterns for all three concentrations of silver used.

135

Figure 40:- XRD analysis of silver doped scaffolds obtained from different silver concentrations in the molten salt ® bath showing the peaks indicative of sintered/crystallised 45S5 Bioglass (Na2Ca2Si3O9), silver and silver phosphate

Comparing the XRD spectra of the silver MSIE samples and that of an unaltered scaffold, shown in Figure 40 and Figure 32 respectively, it can be seen that the Na2Ca2Si3O9 peaks are still present in the spectrum of the MSIE samples. Figure 40 shows the presence of metallic silver and silver phosphate peaks in all three samples. The XRD characterisation thus confirms the presence of silver in the structure of the BBG samples after the MSIE process. Since the XRD analysis confirmed the presence of silver in the lowest concentration samples, unlike the EDX analysis in Figure 39, it can be conclusively stated that the low concentration MSIE was also effective.

Based on these preliminary results, it was confirmed that the silver MSIE process was successful, indicating that a further more extensive study should be carried out. The conclusions of Di Nunzio et al (6) stated that the most suitable molten salt bath concentration out of the three tested was the lowest, as this still imparted the silver ions into the glass-ceramic structure whilst not having a detrimental effect on the HA formation on the material and not having a toxic effect on cells. This proof of concept study showed therefore that all three molten salt bath concentrations of silver investigated here imparted silver ions into the glass-ceramic structure.

As in Di Nunzio et al paper, (6) this proof of concept study has shown the lowest salt bath concentration has led to the successful incorporation of silver ions into the structure of the pre-fabricated scaffold and the next stage would be to optimize this process including investigating a range of immersion periods (15, 30, 45 and 60 minutes) and other concentrations of salt bath. These new

136 concentrations fall between the lowest and medium concentrations from the proof of concept study, shown in Table 3.

5.3.2 Structural characterisation of silver molten salt ion exchange scaffolds

Based on the results presented in the preliminary study, shown in section 5.3.1, a new set of concentrations of silver nitrate in the salt bath was chosen as shown in Table 3 from section 4.2.2.I. These concentrations were chosen as the best trade-off between being toxic when the scaffolds are in contact with cells and having enough silver in the melt to actually diffuse into the silicate matrix. A range of longer immersion periods in the salt melt were used to enable silver at the low concentrations in the salt bath to diffuse into the structure of the scaffold. This means that there are 12 sample types investigated in this main study, the three concentrations (low, medium and high) over four immersion periods (15, 30, 45,60 minutes).

One of the changes in the production of the silver MSIE samples was the introduction of an updated washing protocol as it was observed that the surface of the scaffold was still covered in a layer of salt residue even after the scaffolds were washed in distilled water, as shown in Figure 41. The new washing protocol, as described in section 4.2.2.I, involved washing the MSIE samples in distilled water, ethanol and distilled water and it was observed that this treatment reduced the salt residue issue as shown in Figure 43. All the MSIE samples, unless stated, in this section have under gone the washing protocol before characterisation.

The other change to the production method shown in Figure 18 was to introduce a controlled rate of heating/cooling to prevent the scaffolds mechanical failure due to thermal shock. This was achieved by heating the scaffolds alongside the salt bath at a rate of 5o/min and after the MSIE process letting them to cool at 5o/min. The optimized production method is described in Figure 19.

The porosity of the silver-doped BBG scaffolds was measured, before MSIE, after MSIE, after being subjected to the updated washing protocol and after immersion in SBF for 14 days, using Equation 1 and results are collectively shown in Figure 42. The results clearly show that the new washing protocol was required as the porosity was reduced from 93±2% for the BBG scaffold to as low as 55±12% which was unacceptable. As explained in the literature review section, an ideal scaffold should have porosity of over 70% in order for the bone regeneration process to occur. After implementing the updated washing protocol the porosity of the silver-doped scaffolds rose to between 74±6% and 81±6% which is a suitable porosity for a BTE scaffold.

137

Figure 41:- BBG scaffolds immersed in a low silver concentration molten salt bath for A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes showing the presence of non-desired salt residue from the salt bath (arrows indicating the salt residue)

Figure 42:- Porosity of sintered BBG scaffolds before being doped with silver through MSIE, after MSIE but before being subjected to the washing protocol and after MSIE and after being subjected to the washing protocol * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

138

It was observed using SEM that there was no apparent physical difference between silver doped scaffolds for the three concentrations used and there was no perceptible difference between the four immersion periods. The results are shown in Figure 43 with the SEM images for the other concentrations shown in appendix C1. These images also show that the silver-doped scaffolds exhibit the same characteristics observed with the BBG scaffolds, shown in Figure 29, in terms of pore size, pore interconnectivity and strut dimensions. The only observed difference is in the surface topography, the silver –doped scaffolds have rougher surface compared to the BBG scaffolds, which is in agreement with observations made by other researchers. (113)

Figure 43:- BBG scaffolds immersed in a low silver concentration molten salt bath for A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes

The silver-doped scaffolds were subjected to the same series of structural characterisation techniques applied on the BBG scaffolds so that a direct comparison can be made, therefore the BBG scaffolds can act as a control throughout this work. XRD, FTIR and EDX and depth EDX techniques were carried out on silver-doped scaffolds to assess the possible chemical changes to the structure due to the MSIE process. Figure 44 shows the XRD characterisation for the all three concentrations of silver nitrate in the salt bath after a 15 minute immersion period, the spectra for the other three immersion periods are shown in appendix C1.

139

Figure 44:- XRD characterisation of BBG scaffold after introducing silver through MSIE. Spectra for the three concentrations with an immersion period of 15 minutes are shown

Figure 44 indicates that all the peaks indicative of sintered/crystallised 45S5 Bioglass® are present, as shown in Figure 32. (4) (5) (389) (416) Some peaks indicate the presence of silver in the crystal structure of the silver-doped samples, as seen in the preliminary study in Figure 40. The silver peaks were present in all of the spectra for all three concentrations over all four immersion periods.

The FTIR spectra presented in Figure 45, along with the two spectra for the other two concentrations shown in appendix C1 show the peaks indicative of BBG including the pair of P-O bending bonds at 720 – 650cm-1 and 625 – 580cm-1 and the pair of Si-O-Si bending bonds at 528 – 536cm-1 and 455 – 457cm-1 as shown in Figure 31. (413) (414) (415) The difference between the spectra

- -1 shown in Figure 31 and in Figure 45 is the presence of a NO 3 bond at 1387cm indicating the presence of a nitrate and the amalgamation of the stretching Si-O bond at 920 and 940cm-1 as well as the pair of asymmetric stretching Si-O-Si bonds at 1097 – 1100cm-1 and 1037 – 1050cm-1 to a single peak at 950- 1000cm-1 which shows the bonds are shifting to lower wavenumbers due to the presence of the silver, with similar observations reported in the literature. (428) (429) (430) The presence of another shifted Si-O peak was detected at 670-680cm-1 which indicates that the structure of the sample is re-arranging itself to accommodate the additional silver present in its structure. (428) (429) (430) These two peaks are considered to be indicative of the inclusion of silver in the structure of a silica based material (428)

140

(429) (430) and are present in all spectra corresponding to all concentrations of the salt bath and for all immersion periods.

Figure 45:- FTIR spectra showing medium concentration of silver MSIE scaffolds with immersion periods of 15, 30, 45 and 60 minutes

Figure 46 shows the EDX analysis of the all the different silver-doped samples, as with Figure 39, there is little indication of the presence of silver at approximately 3keV, (113) (431) for the lowest concentration samples after 15 and 30 minutes (Figure 46A) and B)), but there is a silver peak present for the lowest concentration sample after 45 and 60 minutes MSIE (Figure 46C) and D)). The medium concentration samples at 15 minutes also showed no silver peak but it was found to be present for the three remaining immersion periods. For the highest concentration samples the silver peak at ~3keV was seen at all four immersion periods. (113) (431)

Depth EDX analysis was carried out in accordance with the method described in section 4.3.11 and is shown in Figure 47. This analysis showed that the silver MSIE had a maximum penetration depth ranging between 6±1µm and 30±1µm, which is in agreement with the literature. (432) (433) (434) (435) The maximum intensity count increased with increasing concentration of silver nitrate in the salt bath and the maximum penetration depth increased with both the concentration and the immersion period. This analysis reflects the concentration of silver as a function of the distance from the surface.

141

Figure 47:- Depth EDX analysis of silver doped BBG pellets covering all concentrations of silver nitrate in the molten salt bath and all immersion periods investigated. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

142

The compressive strength of the silver-doped scaffolds is shown in Table 12. Compared to the compressive strength of the BBG scaffold, 0.53±0.08MPa, (113) (4) a slight but significant increase is observed for silver-doped scaffolds. There are difference plausible explanations as to why the ion exchange process increases the mechanical strength of a glass-ceramic and the relevant theories are presented here. The main theory, explained in chapter two, is that the glass-ceramic is strengthened by the mere fact that the ionic radius of the in-bound ion (in this case the silver ion) is larger than the ion that is leaving the glass (in this case the sodium ion), (335) this mechanism is called induced residual stress build up or the stuffing effect. (436) (437) As the larger ions plug the gaps in the atomic structure a high surface compression is introduced as well as a balancing interior tension. (336) (437) This is somewhat supported by the slight change in mass observed in the scaffolds after the ion exchange process. This effect however does not give us the whole picture as to why the ion exchange process increases the compressional strength. A complementary theory that could occur as well as the stuffing effect is that as the BBG scaffolds have a silicate network that contains calcium and sodium as network modifiers, (113) (438) the phosphate is incorporated into the structure as a orthophosphate which is charged, and it is balanced by the calcium in the structure. (113) (438) When silver is exchanged, ions are ionically bonded to non-bridging oxygen’s but the silver also interacts with nearby bridging oxygen’s forming a 2-coordinate system. (113) (438) This interaction is supported by the amalgamation of the Si-O bond at 920 and 940cm-1 and the pair Si-O-Si bonds at 1097 – 1100cm-1 and 1037 – 1050cm- 1 as seen in the FTIR results shown in Figure 45 and appendix C1. (428) (429) The extra interaction between the silver and the bridging oxygen’s depends on how easily silver can shed its outer electron and it is this extra structural interaction that could explain the increased compressive strength due to the silver MSIE. (113) (438) Further research is required to obtain the complete knowledge and to answer the question why the compressional strength of the scaffold is increased by the introduction of the silver ions into the scaffold through the use of MSIE.

Table 12:- Compressive strength of silver MSIE BBG scaffolds. (113) * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Immersion period Compressive strength in MPa for a range of concentrations (minutes) Low concentration Medium concentration High concentration 15 0.62 ± 0.04 * 0.65 ± 0.07 * 0.68 ± 0.06 * 30 0.64 ± 0.03 * 0.66 ± 0.07 * 0.69 ± 0.09 * 45 0.65 ± 0.06 * 0.68 ± 0.04 * 0.72 ± 0.05 * 60 0.67 ± 0.06 * 0.70 ± 0.08 * 0.73 ± 0.04 *

Table 13 shows the RMS, Ra and PV values for the silver-doped samples. It can be seen that these values do not increase with increasing concentration of silver in the salt bath but instead they increase as the immersion period increases. They are however lower than the values seen for the BBG

143 pellets (3±1µm, 2.6±0.9µm and 19±9µm for RMS, Ra and PV, respectively), (224) (225) which indicates that the topography at the surface of the silver-doped pellets is smoother than that of the BBG pellets, which is a result of the MSIE process. Whether this change to the surface topography, compared to the BBG scaffold, will have an effect on the cell behaviour in the cell culture studies will be shown in chapter six.

Table 13:- White light interferometry results for the surface roughness of silver doped 45S5 Bioglass® –derived glass-ceramic pellets * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Concentration of Immersion period RMS Ra PV silver in the salt bath (minutes) (µm) (µm) (µm) Low concentration 15 0.6±0.2 * 0.4±0.1 * 4±1 * Low concentration 30 0.7±0.1 * 0.5±0.1 * 5±1 * Low concentration 45 0.8±0.1 * 0.6±0.1 * 5±1 * Low concentration 60 1.6±0.6 * 1.2±0.4 * 8±1 * Medium concentration 15 0.6±0.1 * 0.5±0.1 * 4±1 * Medium concentration 30 0.7±0.2 * 0.5±0.1 * 6±1 * Medium concentration 45 0.8±0.2 * 0.7±0.2 * 6± 1 * Medium concentration 60 1.6±0.5 * 1.3±0.3 * 9±1 * High concentration 15 0.6±0.1 * 0.5±0.1 * 4±1 * High concentration 30 0.7±0.1 * 0.6±0.1 * 6±1 * High concentration 45 0.9±0.1 * 0.7±0.1 * 6±1 * High concentration 60 1.7±0.2 * 1.3±0.6 * 10±1 *

The next stage is to test the bioactivity of the silver-doped scaffolds by immersing them in SBF for 14 days and characterising them again. An immersion period in SBF of 14 days was chosen as this allows the precipitation of HA, as seen for the BBG scaffold, but will also allow any delay in the H A formation caused by the MSIE process to be observed. The SEM images shown in Figure 48 and the corresponding SEM images in appendix C1 for the other two concentrations all show the characteristic HA knobbles which is consistent with the SEM images in Figure 34 and with results in the literature. (4) (5) These observations indicate that the presence of silver in the structure does not have a detrimental effect on the samples bioactivity. (113) Figure 48 also shows the surface topography has changed from that shown in Figure 43, with a micro-roughness present, as seen in Figure 34, which is due to the MSIE process and the formation of a HA surface layer after immersion in SBF and in this case the release of silver ions. (222) (223) (113)

The porosity of the silver-doped scaffolds, as shown in Figure 42, is seen to decrease by on average 10% after 14 days in SBF which is in line with the porosity decrease seen in the BBG scaffolds as reported in section 5.2. (241)

144

Figure 48:- BBG scaffolds immersed in a medium concentration molten salt bath for A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes and then immersed in SBF for 14 days. Images have a magnification of x1500 (arrows indicating HA)

The XRD analysis shown in Figure 49 and appendix C1 show the characteristic peaks for a BBG scaffold that has undergone SBF immersion, as seen in Figure 37, with the presence of key HA peaks in the crystal structure which is consistent with the literature. (5) (4) (426) The indicative peaks for silver, seen in Figure 32 and Figure 44, are also present indicating that silver is still present in the structure of the silver-doped scaffolds after immersion in SBF for 14 days. (113)

The FTIR analysis shown in Figure 50 and appendix C1 show the results of silver-doped scaffolds after immersion in SBF for 14 days. They show the changes that spectra undergo after immersion in SBF for 14 days, as seen in Figure 36, with the development of the C-O bond at 880cm-1 and 900cm-1 which indicates the presence of carbonate apatite. (425) In addition, the P-O bond at 875cm- 1 to 825cm-1, which increases in size, indicates the presence of a precipitated HA layer, (423) (424) (222) and the dual P-O bond at 760cm-1 to 650cm-1, which indicates the crystallisation of the precipitated HA. (424) Thus, introduction of silver into the crystal structure does not have a detrimental effect on the bioactivity of the BBG scaffold and the formation of HA after 14 days of immersion in SBF is

- -1 confirmed. The FTIR analysis also showed a reduced NO 3 bond at 1387cm and a reduced Si-O bond at 950-1000cm-1. (428) (429) (430) The second Si-O peak detected at 670-680cm-1 in has been reduced

145 to such an extent that it is almost undetectable suggesting that this is bond that breaks first when silver diffuses out of the silicate structure during immersion in SBF.

Figure 49:- XRD characterisation of BBG after introducing silver through MSIE. Spectra for the three concentrations investigated with an immersion period of 15 minutes after immersion in SBF for 14 days are shown

Figure 50:- FTIR spectra showing low concentration silver MSIE scaffolds with immersion periods of 15, 30, 45 and 60 minutes after immersion in SBF for 14 days

146

The EDX analysis of the silver-doped samples after SBF immersion is shown in Figure 51 and appendix C1 showed that the silver peak at ~3keV, as seen in Figure 46, was non-existent in the spectra. (113) (431) However the changes in the calcium and phosphorus peaks, as seen in Figure 35 and discussed in chapter two, which indicate the presence of HA are seen in Figure 51. (421) (422) (420) This result supports the premise that the addition of silver in the crystal structure does not have a detrimental effect on the bioactivity of the sample.

Figure 51:-EDX analysis showing the elemental composition of the silver doped samples over the three concentrations with an immersion period of 60 minutes with A) low concentration of silver in the salt bath, B) medium concentration in the salt bath and C) high concentration of silver in the salt bath

The Depth EDX analysis is shown in Figure 52 and is in agreement with the results shown in Figure 47, in that the maximum penetration depth of silver is between 6±1µm and 30±1µm. (432) (433) (434) (435) Due to the diffusion of some of the silver out of the structure the maximum intensity of the silver decreased. The highest value of relative intensity shown in Figure 47 was at the surface after MSIE as expected, but the highest value of relative intensity shown in Figure 52 after immersion in SBF ranged from 2µm to 10µm. Indicating that during immersion in SBF silver diffused from the surface first, leaving a higher concentration silver below the surface behind. This indicates that incorporation of silver through MSIE gives the scaffold a controlled release mechanism rather than inducing the release of all incorporated silver in one burst. From this result it cannot be ascertained how the formation of HA on the surface of the silver-doped BBG scaffold, but when the ICP-MS results are discussed this should be considered.

147

Figure 52:- Depth EDX analysis of silver doped BBG pellets covering all concentrations of silver nitrate in the salt bath and all immersion periods in the salt bath after 14 days in SBF. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

As with the BBG scaffolds, the compressive strength of the silver-doped scaffolds decreased after immersion in SBF for 14 days, as shown in Table 14. This result is supported by data in the literature as the combination of ion loss and HA precipitation seems to have a detrimental effect on the mechanical competence of the scaffold. (352) (418) The compressive strength values are still all higher than the corresponding value of 0.43±0.05MPa for the BBG scaffold. As discussed previously, this increase in the mechanical competence can be attributed to the stuffing effect (395) (436) (437)(397) of the silver ions and the extra interaction in the silicate network to form a 2-coordinate system. (113) (438)

Table 14-: Compressive strength of silver MSIE BBG scaffolds after immersion in SBF for 14 days. (113) * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Immersion period Compressive strength in MPa for a range of concentrations (minutes) Low concentration Medium concentration High concentration 15 0.50 ± 0.04 * 0.51 ± 0.06 * 0.55 ± 0.05 * 30 0.50 ± 0.05 * 0.53 ± 0.05 * 0.56 ± 0.04 * 45 0.51 ± 0.06 * 0.52 ± 0.06 * 0.55 ± 0.06 * 60 0.49 ± 0.03 * 0.53 ± 0.08 * 0.58 ± 0.07 *

The surface topography, shown in Figure 48, was observed to be marginally changed in comparison to the topography observed in Figure 43 and this is supported by the white light interferometry results shown in Table 15. The RMS, Ra and PV values are marginally higher than those obtained for the BBG pellet after immersion in SBF for 14 days but still fall in the range that is

148 considered suitable for cell attachment. (224) The overall increase in surface roughness can be attributed to the precipitation of HA as is seen in Figure 48. Again whether the change in roughness due to the precipitation of HA on the modified scaffolds surface has an effect on the cell behaviour observed will be discussed in chapter six.

Table 15:- White light interferometry results for the surface roughness of silver doped BBG pellets after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Concentration of Immersion period RMS Ra PV silver in the salt bath (minutes) (µm) (µm) (µm) Low concentration 15 6±1 * 5±1 * 23±7 * Low concentration 30 6±1 * 5±1 * 23±4 * Low concentration 45 7±2 * 5±1 * 23±7 * Low concentration 60 7±2 * 5±1 * 24±6 * Medium concentration 15 7±2 * 6±1 * 24±7 * Medium concentration 30 7±2 * 6±2 * 24±5 * Medium concentration 45 7±2 * 6±2 * 24± 6 * Medium concentration 60 7±2 * 6±2 * 24±7 * High concentration 15 7±2 * 6±2 * 25±7 * High concentration 30 7±2 * 6±1 * 25±8 * High concentration 45 7±2 * 6±2 * 25±4 * High concentration 60 7±2 * 6±2 * 25±9 *

The wettability of the silver-doped samples was obtained in accordance with the method described in section 4.3.8. Results are shown in Figure 53 for samples before immersion in SBF and after immersion in SBF for 14 days. The BBG pellet before immersion in SBF had a contact angle of 28±4o which decreased to 23±2o after SBF immersion. (419) The values for silver-doped samples, shown in Figure 53, are all lower than the corresponding BBG pellet values and they fall outside the lower limit of the range of contact angles, which is considered suitable for cell attachment (between 48 and 62 degrees). (419) As with the white light interferometry results, whether the change in topography has an observable effect on the cells will be shown and discussed in chapter six.

ICP-MS analysis was carried out on the supernatant taken from PBS soaked silver-doped scaffolds and the results are shown in Figure 54. It can be seen that as silver concentration in the salt bath increased so did the released ion concentration and a similar relation was seen for the immersion period, i.e. as the immersion period in the molten salt bath increased so did the released ion concentration. This result indicates release rates of between 27.7±0.05mg per litre per day and 2.43±0.05mg per litre per day. To put that into perspective the release rates extrapolate down to between 0.00115mg per ml per hour and 0.000101mg per ml per hour, which is an extremely small release rate and similar to data reported in the literature. (439) (440)

149

Figure 53:- Wettability data for silver doped BBG scaffolds showing results for pre and post immersion in SBF for 14 days (Data shown are mean values and standard deviation of at least 3 measurements). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Figure 54:-ICP-MS data for silver MSIE scaffolds for all concentrations of silver in the molten salt bath and over all immersion periods. * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis.

150

5.4 Copper molten salt ion exchanged scaffolds

This study investigated copper-doped scaffolds which were produced following the procedure described in section 4.2.2.II and depicted in Figure 19. The investigation focused on three concentrations of copper nitrate in the molten salt bath, shown in Table 4, with a single immersion period of 15 minutes.

The porosity of the copper-doped BBG scaffold was measured after MSIE and after the washing protocol had been implemented using Equation 1. The porosities were found to be 72±6%, 73±7% and 72±7% for the low, medium and high concentration doped samples, respectively. These values are all lower than the porosity of 93±2% for the BBG scaffolds, but are still above the ideal porosity threshold of 70% suggested in the literature. (210)

SEM micrographs, shown in Figure 55, compared with those seen in Figure 29, show that there is marked difference in the surface topography of the copper-doped scaffold compared to the BBG scaffolds. This change in topography is due to the MSIE process and was observed to a lesser extent for the silver-doped scaffolds as discussed in section 5.3.2. However the copper-doped samples still all show the key characteristics that are desirable for a BTE scaffold in terms of pore size, pore interconnectivity and strut placement and dimensions, just with an increase in the surface roughness of the surface.

XRD, FTIR, EDX and depth EDX analysis were carried out to determine the changes to the microstructure, chemistry and crystal structure of the BBG samples that occurred when the samples were subjected to copper MSIE. XRD results, shown in Figure 56, show the characteristic peaks that are associated with sintered/crystallised 45S5 Bioglass®, as seen in Figure 32. (4) (5) (389) (416) There are additional peaks observed in the spectra that are indicative of the possible presence of copper oxide, copper and copper phosphate (441) (442) (443) in the structure and this result indicates that the copper MSIE process was successful in introducing copper into the silicate structure. There were three phases of copper observed, over many samples, and it can not be determined, from the XRD analysis alone, which phase is represented as the indicative peaks for Cu, CuO and CuPO4 overlap each other or are not present in some of the samples. This is in fact a complex glass-ceramic system, but relevant peaks were properly identified, as shown in Figure 56.

151

Figure 55:- BBG scaffolds immersed a variety of different concentrations of copper nitrate molten salt bath for 15 minutes with A) low concentration salt bath (magnification of x150), B) low concentration salt bath (magnification of x1000), C) medium concentration salt bath (magnification of x150) and D) high concentration salt bath (magnification of x150)

Figure 56:- Offset XRD spectra of the three concentrations of copper doped BBG scaffolds

152

The FTIR analysis for the copper-doped samples, shown in Figure 57, indicates the presence of peaks that are associated with BBG scaffolds including the pair of P-O bending bonds at 720 – 650cm-1 and 625 – 580cm-1 and the pair of Si-O-Si bending bonds at 528 – 536cm-1 and 455 – 457cm-1 as shown in Figure 31. (413) (414) (415) As with the silver-doped samples, the difference in the spectra

- -1 is the presence of a NO 3 bond at 1387cm indicating the presence of a nitrate and the amalgamation of the stretching Si-O bond at 920 and 940cm-1 and the pair of asymmetric stretching Si-O-Si bonds at 1097 – 1100cm-1 and 1037 – 1050cm-1 to a single peak at 950-1000cm-1, which indicates that the wavenumber bands are shifting lower showing that the copper is having a depolymerisation effect on the silicate-phosphate framework. (444) (445) (446) A second shifted Si-O peak was detected at 670- 680cm-1 which indicates that the structure accommodates the additional copper present as the CuO is replacing the Na2O. (444) (445) (446) These two shifts in the Si-O peaks are considered to be indicative of the inclusion of copper into the structure of a silica based material (444) (445) (446)

Figure 57:- FTIR spectra of copper doped BBG scaffolds of all concentrations after immersion in the molten salt bath for 15 minutes

Figure 58 shows the EDX analysis for the copper-doped samples and shows the presence of copper at 0.5keV and 8keV which is consistent with the literature. (447) A slight increase in the intensity of the peaks is observed as the concentration of copper in the salt bath increased.

153

Figure 58:- EDX analysis of copper doped BBG scaffolds for all three concentrations of copper in the salt melts with A) low concentration, B) medium concentration and C) high concentration

The depth EDX analysis, shown in Figure 59 , showed that the maximum penetration depth ranged between 9±1µm and 10±1µm which agrees with the literature. (448) (434) The maximum penetration depth increased marginally with increasing copper concentration but the relative intensity measured was observed to increase rapidly with increasing copper concentration.

Figure 59:- Depth EDX analysis of copper doped BBG pellets. * Indicates a significant change between the change in concentration (p≤0.05) using ANOVA analysis.

The compressive strength of the copper-doped scaffolds is shown in Table 16. An increase in the compressive strength of the samples is observed, compared to the value of the BBG scaffolds, (113)

154

(4) which is in agreement with the literature. (449) (450) As discussed earlier for the silver-doped scaffolds, the main mechanism for the increased compressional strength can be attributed to the stuffing effect, (335) (436) (437)(397) as discussed in section 5.3.2. It can also be hypothesized that, in a similar way to the behaviour of the silver ions in the BBG structure, (113) (438) the copper could behave ionically bonded to a non-bridging oxygen’s, but also interacts with other near-by bridging oxygen’s forming a 2-coordinate system which may lead to the increased compressional strength observed. This result is supported by the FTIR analysis, shown in Figure 57, with the amalgamation of the Si-O bond at 920 and 940cm-1 and the pair Si-O-Si bonds at 1097 – 1100cm-1 and 1037 – 1050cm-1.

Table 16:- Compressive strength of copper MSIE BBG scaffolds (samples prepared by MSIE for 15 minutes). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Copper nitrate Compressive strength of concentration (MPa) Low 0.59 ± 0.05 * Medium 0.60 ± 0.04 * High 0.62 ± 0.04 *

The white light interferometry analysis results are shown in Table 17 for the copper-doped samples and it can be seen that the RMS, Ra and PV values have increased with the increasing copper concentration in the salt bath. These values are higher than those observed for the BBG pellets (3±1µm, 2.6±0.9µm and 19±9µm for RMS, Ra and PV, respectively) (224) (225) which indicate that the topography at the surface of the copper-doped pellets is rougher than that of the BBG pellets, which is likely due to the MSIE process. This result is supported by the observations on SEM images of the copper-doped scaffolds compared to the BBG scaffold, seen in Figure 55 and Figure 29, respectively. Whether this increase in roughness compared to the BBG scaffold and the silver-doped scaffolds has an effect on the cell behaviour in the cell culture studies is shown in chapter six.

Table 17:- White light interferometry results for the surface roughness of copper doped BBG pellets. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Concentration of RMS Ra PV copper in the salt bath (µm) (µm) (µm) High concentration 4.1±0.3 * 2.9±0.2 * 22±3 * Medium concentration 4.4±1.0 * 3.1±0.9 * 23±5 * Low concentration 5.1±1.1 * 3.8±0.9 * 26±5 *

As with the previous studies, the bioactivity of the copper-doped samples was assessed by immersion in SBF for 14 days. Figure 60 shows the SEM images of the copper-doped scaffolds after

155 immersion in SBF for 14 days. These images show the presence of the characteristic HA knobbles on the surface of the copper-doped scaffolds, which is consistent with the SEM images in Figure 34 and with results in the literature (4) (5) and indicates that the presence of the copper does not have a detrimental effect on the samples bioactivity. Figure 60 also shows that the topography of the scaffolds has changed due to the presence of the precipitated HA.

Figure 60:- BBG scaffolds immersed in a variety of different concentrations of copper nitrate molten salt bath for 15 minutes with A) low concentration salt bath, B) medium concentration salt bath and C) high concentration salt bath after immersion in SBF for 14days. All images have a magnification of x300

The porosity of the copper-doped scaffolds was measured at 63±7%, 64±6% and 62±7% for the low, medium and high concentration of copper in the salt bath, respectively. The porosity has decreased by on average 10% after 14 days in SBF which is in line with the porosity decrease seen by the BBG scaffolds, discussed in section 5.2. (241)

156

The XRD spectra, shown in Figure 61, exhibit the characteristic peaks for a BBG scaffold that has undergone SBF immersion, as seen in Figure 37, with the presence of HA peaks consistent with the literature. (5) (4) (426) The indicative peaks for the possibility of the presence of copper, seen in Figure 56, are also present in the XRD spectra in Figure 61, indicating that there is still copper present in the structure of the copper-doped scaffolds after immersion in SBF for 14 days. (441) (442) (443) As was observed in Figure 56, copper maybe present in three phases, but it can not be determined with certainty (based only on these results) which phase is predominant as the indicative peaks observed for Cu, CuO and CuPO4 overlap each other or were not always present in all observed spectra shown in Figure 61. It is also seen that the presence of copper does not have a detrimental effect on the precipitation of HA on the surface of the copper-doped samples.

Figure 61:- Offset XRD spectra of the three concentrations of copper doped BBG scaffolds after immersion in SBF for 14 days

The FTIR spectra of the copper-doped scaffolds after immersion in SBF for 14 days are shown in Figure 62. They all show the typical changes that spectra undergo after BBG scaffolds are immersed in SBF for 14 days, as seen in Figure 36, with the development of the C-O bond at 880cm-1 and 900cm- 1, (425) the P-O bond at 875cm-1 to 825cm-1, which increases in size indicating the presence of a precipitated HA layer, (423) (424) (222) the dual P-O bond at 760cm-1 to 650cm-1, which by merging indicates the crystallisation of the precipitated HA. (424) These changes in the spectra show that the

157 introduction of copper does not have a detrimental effect on the bioactivity of the BBG as anticipated

- -1 by the XRD results shown in Figure 61. Figure 62 also shows a reduced NO 3 bond at 1387cm and a reduced Si-O bond at 950-1000cm-1. (444) (445) (446) The second shifted Si-O peak detected at 670- 680cm-1 has been reduced to such an extent that it is almost undetectable suggesting that this is the bond that breaks first when the copper diffuses out of the crystal structure during the immersion period in SBF.

Figure 62:- FTIR spectra for all concentrations of copper doped scaffolds after 15 minutes in the molten salt bath and after immersion in SBF for 14 days

The EDX analysis of the copper-doped samples after SBF immersion shown in Figure 63 confirms the presence of copper at 0.5keV and 8keV which is consistent with the literature (447) but the peaks are reduced compared to the peaks seen in Figure 58. However the change in the calcium and phosphorus peaks, as seen in Figure 35, indicates the presence of precipitated HA, as seen in Figure 63. (421) (422) (420) This result supports the premise that the addition of copper does not have a detrimental effect on the bioactivity of the sample, at least considering the first 14 days of immersion in SBF.

158

Figure 63:- EDX analysis of copper doped BBG scaffolds for all three concentrations of copper in the salt melts with A) low concentration, B) medium concentration and C) high concentration after immersion in SBF for 14 days

The Depth EDX analysis is shown in Figure 64 and it supports the results obtained for the copper-doped scaffolds before immersion in SBF shown in Figure 59, in that the maximum penetration depth ranges between 9±1µm and 10±1µm. (448) (434) Due to the diffusion of some of the copper out of the samples, the maximum intensity of the copper peak decreased. The highest value of relative intensity shown in Figure 59 was recorded at the surface after MSIE, but the highest value of relative intensity shown in Figure 64 after immersion in SBF was at 5µm. This result indicates that during immersion in SBF copper diffused from this region first, leaving some copper behind. This result suggests that incorporation of copper through MSIE into the glass-ceramic matrix of the BBG scaffolds gives the scaffold a controlled release mechanism rather than releasing all of the incorporated copper in one burst. This effect will have implications for the application of the scaffolds as an ion releasing scaffold for antibacterial properties.

The compressive strength of the copper-doped scaffolds decreased after immersion in SBF for 14 days, as was observed with the BBG scaffolds and shown in Table 18. This result is supported by the literature as the combination of ion leaching and HA precipitation usually has a detrimental effect on the mechanical competence of the scaffold. (352) (418) The compressive strength values are still all higher than the corresponding value of 0.43±0.05MPa for the BBG scaffold, indicating that although some of the copper has diffused out of the structure it is still having a positive effect on the scaffold compressive strength. The increased compressional strength can be attributed to the stuffing effect (335) (436) (437)(397) and extra interaction in the silicate network to form a 2-coordinate system, (113) (438) as discussed in previous sections.

159

Figure 64:- Depth EDX analysis of copper doped BBG pellets after immersion in SBF for 14days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Table 18:- Compressive strength of copper MSIE BBG scaffolds after immersion in SBF for 14 days (samples prepared by MSIE for 15 minutes). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Copper nitrate Compressive strength concentration (MPa) Low 0.50 ± 0.04 * Medium 0.49 ± 0.06 * High 0.51 ± 0.06 *

The white light interferometry results, shown in Table 19, support the observations taken from the SEM images in Figure 60, in that the surface topography has changed due to the presence of the precipitated HA. The RMS, Ra and PV values are higher than those obtained for the BBG pellet after immersion in SBF for 14 days but still fall in the range that are considered suitable for cell attachment. (224) The overall increase in surface roughness can be attributed to the precipitation of HA, as seen in Figure 60. The discussion on if this increase in topography will affect the cells used in the cell biology will be discussed in chapter six.

160

Table 19:- White light interferometry results for the surface roughness of copper doped BBG pellets after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Concentration of RMS Ra PV copper in the salt bath (µm) (µm) (µm) High concentration 8.0±0.8 * 7.4±0.3 * 23±5 * Medium concentration 9±1 * 7.7±0.8 * 23±5 * Low concentration 9±1 * 8.0±1.0 * 23±5 *

The wettability of the copper-doped samples is shown in Figure 65 for samples before and after immersion in SBF for 14 days. The BBG pellet before immersion in SBF had a contact angle of 28±4o which was reduced to 23±2o after SBF immersion. (419) The contact angle values of samples with the highest concentration of copper were in the same range as the ones measured for the BBG pellet, but the contact angles for both the medium and low concentration samples were high both before and after immersion in SBF. These results indicate that when the concentration of copper increased so did the measured contact angle. The contact angles for the lowest copper concentration are high enough to fall within the range considered to be suitable for cell attachment, which is between 48o and 62o. (419) Whether the change in topography due to the copper doping process has an effect on the cell behaviour compared to the BBG scaffolds will be shown in chapter six.

Figure 65:- Wettability data for copper doped BBG scaffolds showing results for samples before and after immersion in SBF for 14 days (Data shown are mean values and standard deviation of at least 3 measurements). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

161

ICP-MS analyses were carried out to obtain the amount of copper being released from the copper doped samples after immersion periods in PBS. Results are shown in Figure 66, indicating that as the concentration of copper in the sample increased so did the released amount of copper and the release rate. This data gives release rates of 38.90±0.05mg per litre per day, 118.13±0.05mg per litre per day and 137.67±0.05mg per litre per day for the low, medium and high concentrations of copper in the salt bath, respectively. To put that into perspective the release rates extrapolate down to between 0.0016mg per ml per hour and 0.00571mg per ml per hour, which is an extremely small release rate. (451)

Figure 66:- ICP-MS analysis of copper ion release from copper doped BBG scaffolds immersed in PBS for the three different concentrations of copper nitrate salt used in the MSIE study. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

5.5 Combination molten salt ion exchanged scaffolds

This part of the study investigated silver/copper-doped scaffolds which were produced following the procedure described in section 4.2.2.III and depicted in Figure 19. This study concentrated on a single concentration of silver and copper in the molten salt bath over four immersion periods of 15, 30, 45 and 60 minutes.

The porosity of the silver/copper-doped BBG scaffolds was measured using Equation 1 after MSIE and after the washing protocol had been implemented. The porosities were found to be 72±7%,

162

73±7%, 71±7%, and 73±6% for the immersion periods of 15, 30, 45 and 60 minute samples, respectively. These values are all lower than the porosity of 93±2% for the BBG scaffolds, but are still above the ideal porosity threshold of 70% as described in the literature. (210)

From SEM observations of the silver/copper-doped scaffolds, shown in Figure 67, the surface of the samples appears to be rougher than that of the BBG scaffolds shown in Figure 29, but still having a convenient pore size, strut dimensions and pore interconnectivity.

Figure 67:- SEM images of BBG scaffolds immersed in the combination silver nitrate, copper nitrate and sodium nitrate salt bath for A) 15 minutes (magnification of x400), B) 30 minutes (magnification of x400), C) 45 minutes (magnification of x400) and D) 60 minutes (magnification of x150)

XRD analysis, shown in Figure 68, shows the characteristic peaks that are associated with sintered/crystallised BBG, as seen in Figure 32. (4) (5) (389) (416) There are additional peaks observed in the spectra, which are indicative of the possible presence of copper, (441) (442) (443) as seen in Figure 56, and silver, as seen in Figure 44. As stated in section 5.4 the presence of copper can not be easily attributed to a unique phase due to the fact that in some samples not all the phases were present

163 or overlapping of peaks occurred. This result indicates that the silver/copper MSIE process was successful in introducing both silver and copper into the crystal structure of the BBG.

Figure 68:- XRD patterns of BBG after MSIE introducing both silver and copper for immersion periods of 15, 30, 45 and 60 minutes

The FTIR analysis results, shown in Figure 74, show the indicative peaks associated with the BBG scaffold as seen in Figure 31 including the pair of P-O bending bonds at 720 – 650cm-1 and 625 – 580cm-1 and the pair of Si-O-Si bending bonds at 528 – 536cm-1 and 455 – 457cm-1. (413) (414) (415)

- -1 The difference appears in the form of a NO 3 bond at 1387cm indicating the presence of a nitrate and in the form of the merger of the stretching Si-O bond at 920 and 940cm-1 and the pair of asymmetric stretching Si-O-Si bonds at 1097 – 1100cm-1 and 1037 – 1050cm-1 to a single peak at 950-1000cm-1 which suggest the transformation and shift of the Si-O between 920-940cm-1 to accommodate the presence of the additional copper and silver in the silicate-phosphate framework which has been reported in the literature. (428) (429) (444) (445) (430) (446) With the available results, it is not possible to determine which of the two bonds is present as these bonds occur at the same wavenumber. The presence a second shifted Si-O band was detected at 670-680cm-1 which also indicates that the structure of the sample rearranges to accommodate the additional silver and copper present. (428) (429) (444) (445) (430) (446) The shifting of these two peaks is considered to be indicative of the inclusion of both silver and copper into the structure of a silica based material. (428) (429) (444) (445) (430) (446)

164

Figure 69:- FTIR spectra for silver/copper doped BBG scaffolds obtained by MSIE with different salt bath immersion periods

Figure 70 shows the EDX analysis for the silver/copper-doped samples confirming the presence of silver (at approximately 3keV) and copper (at 0.5keV and 8keV) which is consistent with the results reported above (Figure 46 and Figure 58) and with the literature. (447) (113) (431)

Figure 70:- EDX analysis of the silver/copper doped BBG scaffolds with immersion periods in the salt bath of A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes

The depth EDX analysis, presented in Figure 71, shows that the maximum penetration depth ranged between 6±1µm and 21±1µm for the silver ions and between 11±1µm and 21±1µm for the copper ions which is consistent with results in the literature. (432) (433) (448) (434) (435) The relative amount of silver diffusing into the BBG scaffold is over half that of the copper. This result also shows

165 that the longer the immersion period for the MSIE process, the greater the penetration depth for both silver and copper ions. This analysis would also indicate that copper was favoured in the MSIE process as more copper was detected in the silver/copper-doped scaffolds compared to silver while the ratio of copper to silver in the molten salt bath was equal.

Figure 71:- Depth EDX analysis of silver/copper doped BBG pellets after immersion in salt bath for 15, 30, 45 and 60 minutes. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

The compressive strength of the silver/copper-doped scaffolds, shown in Table 20 shows a moderate increase compared to the value (0.53±0.08MPa) of the BBG scaffolds, (113) (4) which is in line with the literature. (449) (450) (113) (4) It can also be postulated that silver and copper ions affect the mechanical strength of the scaffolds as discussed previously. Namely the increase in the compressive strength can be attributed to the stuffing effect (335) (436) (437)(397) caused by the ion exchange process and to the fact that the ions ionically bond to a non-bridging oxygen’s, replacing a sodium and they also interact with other near-by bridging oxygen’s forming a 2-coordinate system which leads to the increased compressive strength observed. This assumption is supported by the FTIR analysis, shown in Figure 69, with the amalgamation of the Si-O bond at 920 and 940cm-1 and the pair Si-O-Si bonds at 1097 – 1100cm-1 and 1037 – 1050cm-1. (428) (429) (438)

166

Table 20:- Compressive strength of silver/copper-doped MSIE BBG scaffolds obtained by immersion in lower concentration salt bath for the given time periods. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Immersion period Compressive strength (minutes) (MPa) 15 0.61 ± 0.05 * 30 0.62 ± 0.04 * 45 0.64 ± 0.03 * 60 0.65 ± 0.06 *

Table 21 shows the RMS, Ra and PV values for the silver/copper-doped scaffolds indicating that these values have increased compared to the values (3±1µm, 2.6±0.9µm and 19±9µm) for RMS, Ra and PV, respectively for the BBG pellet. (224) (225) This result confirms, the SEM observations in Figure 67 in that the surface of the scaffold is much rougher than that of the BBG scaffold shown in Figure 29. Whether or not the increase in topography observed for the combination ion-doped scaffolds has an effect on the cells is shown and discussed in the cell biology study shown in chapter six.

Table 21:- White light interferometry results for the surface roughness of combination ion doped BBG pellets. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Immersion period RMS Ra PV (minutes) (µm) (µm) (µm) 15 4.1±0.2 * 2.7±0.1 * 20±4 * 30 4.1±0.6 * 2.9±0.3 * 19±4 * 45 4.2±0.6 * 3.3±0.5 * 21±4 * 60 4.2±0.7 * 3.2±0.6 * 19±2 *

The silver/copper-doped scaffolds bioactivity was investigated by immersing them in SBF for 14 days and repeating the characterisation tests mentioned previously. The porosity values of the silver/copper-doped scaffolds after immersion in SBF for 14 days were measured as 63±6%, 62±7%, 63±6% and 64±7% for immersion periods of 15, 30, 45 and 60 minutes, respectively. This is on average 20% lower than the values of the BBG scaffold after the same immersion period and on average 10% lower than the porosity of the doped samples before being immersed in SBF. These decreases of porosity are in line with the observations made for the porosity of the silver-doped and copper-doped samples. Moreover, the decrease in porosity observed after SBF immersion is consistent with the porosity decrease seen by the BBG scaffolds reported in section 5.2. (241)

167

Figure 72 shows SEM images of the silver/copper-doped BBG samples after SBF immersion. The images show characteristic HA knobbles on the surface, consistent with the SEM images in Figure 34 and with reports in the literature, (4) (5) indicating that the presence of either silver or copper ions does not have a detrimental effect on the samples bioactivity, as was reported in the previous sections (sections 5.3.2 and 5.4). Figure 72 also shows that the topography of the scaffolds has changed due to the presence of the precipitated HA.

Figure 72:- BBG scaffolds immersed in the combination silver nitrate, copper nitrate and sodium nitrate salt bath for A) 15 minutes (magnification of x1000), B) 30 minutes (magnification of x1000), C) 45 minutes (magnification of x1000) and D) 60 minutes (magnification of x1500) after immersion in SBF for 14days

The XRD patterns, shown in Figure 73, exhibit the characteristic peaks of a BBG scaffold that has undergone SBF immersion, as seen in Figure 37, with the presence of key HA peaks. (5) (4) (426)

The indicative peaks of both silver and copper (Ag and AgPO4 and possible Cu, CuO and CuPO4 phases for silver and copper, respectively), shown in Figure 68, are still present indicating that silver and copper persist in the structure of the sample. (113) (441) (442) (443) As stated in section 5.4 the presence of copper can not be easily attributed to an unique phase due to the fact that in some samples not all the

168 phases were present or overlapping of peaks occurred The presence of both the silver and the copper along with the characteristic HA peaks indicates that the silver and copper ions have not had a detrimental effect on the bioactivity of the sample.

Figure 73:- XRD analysis of BBG scaffolds after MSIE processing introducing silver and copper for immersion periods of 15, 30, 45 and 60 minutes after immersion in SBF for 14 days

Figure 74 shows the results of the FTIR analysis for the silver/copper doped samples after 14 days in SBF. They all show changes in their spectra that are indicative of the changes in the structure of the BBG sample undergoing SBF immersion. The presence of the C-O bond at 880cm-1 and 900cm- 1, (425) the P-O bond at 875cm-1 to 825cm-1 increasing in size indicate the presence of a precipitated HA layer, (423) (424) (222) and the dual P-O bond at 760cm-1 to 650cm-1 merging suggests the crystallisation of the precipitated HA. (424) Such peaks support the premise that the presence of silver

- -1 and copper ions has no detrimental effect on the bioactivity of the sample. The NO 3 bond at 1387cm is present but it is smaller after SBF immersion, which is consistent with the single ion doped samples, along with the shifted bond at 950-1000cm-1. (428) (429) (430) (444) (445) (446)The second shifted Si- O peak detected at 670-680cm-1 has been reduced to such an extent that it is almost undetectable suggesting that this is the bond that breaks first when copper and/or silver diffuses out of the structure when the sample is immersed in SBF.

169

Figure 74:- FTIR spectra for silver/copper doped BBG scaffolds with different salt bath immersion periods after immersion in SBF for 14 days

Figure 75 shows the EDX analysis of the post SBF immersion silver/copper samples, the presence of copper and silver is detected by the at 0.5keV and 8keV and ~3keV, respectively, which is consistent with Figure 75 and the literature. (113) (447) (431) The intensity of the peaks is reduced compared to those observed in Figure 70. The changes in the phosphate and calcium peaks for the BBG scaffold which indicated the formation of HA, as seen in Figure 35, are also observed in Figure 75. (421) (422) (420) This result supports the premise that the addition of copper and silver does not have a detrimental effect on the bioactivity of the sample.

Figure 75:- EDX analysis of the silver/copper doped BBG scaffolds with immersion periods in the salt bath of A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes after immersion in SBF for 14 days

170

The Depth EDX analysis is shown in Figure 76 and supports the results shown in Figure 71, in that maximum penetration depth ranges between 6±1µm and 21±1µm for silver and between 11±1µm and 21±1µm for copper. (432) (433) (434) (435) (448) Due to the diffusion of some of the copper and silver ions, the maximum intensity of the copper and silver peaks has decreased. The highest value of relative intensity shown in Figure 71 was at the surface after MSIE, but the highest value of relative intensity shown in Figure 76 after immersion in SBF was between 7 to 11µm for copper and 2 to 7µm for silver. These results suggest that during immersion in SBF copper and silver ions diffused from this region first, leaving some copper and silver behind. This indicates that incorporation of copper and silver through MSIE into the glass-ceramic matrix of the BBG scaffolds could give the scaffold a controlled release mechanism, rather than releasing all of the incorporated copper and silver in one burst. This analysis shows that copper was favoured in the MSIE process as more copper was detected in the silver/copper-doped scaffolds, however copper also diffused out faster than silver ion.

Figure 76:- Depth EDX analysis for silver/copper doped BBG pellets after immersion in the salt bath for 15, 30, 45 and 60 minutes followed by immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

As with the BBG scaffolds, the compressive strength of the silver/copper-doped scaffolds decreased after immersion in SBF for 14 days, as shown in Table 22, indicating that the combination of ion loss and HA precipitation has a detrimental effect on the mechanical competence of the scaffold. (352) (418) The compressive strength values are still all higher than the corresponding value of 0.43±0.05MPa for the BBG scaffold.

171

Immersion period in Compressive strength salt bath (minutes) (MPa) 15 0.51 ± 0.06 * 30 0.50 ± 0.06 * 45 0.52 ± 0.05 * 60 0.51 ± 0.05 *

The surface topography, shown in Figure 72, was observed to be marginally rough in comparison to the topography observed in Figure 67 for the BBG scaffold and this result is supported by the white light interferometry results shown in Table 23. The RMS, Ra and PV values are marginally higher than those obtained for the BBG pellet after immersion in SBF for 14 days but still fall in the range that is considered convenient for cell attachment. (224) The overall increase in the surface roughness can be attributed to the precipitation of HA, as is seen in Figure 72. The question of whether the silver/copper MSIE modification has a direct effect on the behaviour of the cells it is exposed to is discussed in chapter six.

Table 23:- White light interferometry results for the surface roughness of combination ion BBG pellets after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Immersion period in salt bath RMS Ra PV (minutes) (µm) (µm) (µm) 15 4.9±0.2 * 4.0±0.1 * 20±6 * 30 4.9±0.6 * 4.0±0.3 * 21±6 * 45 5.1±0.6 * 4.1±0.5 * 22±6 * 60 5.0±0.7 * 4.0±0.6 * 21±7 *

The wettability of the silver/copper-doped samples is shown in Figure 77 for the states before immersion and after immersion in SBF for 14 days. The BBG pellet before immersion in SBF had a contact angle of 28±4o, which changes to 23±2o after SBF immersion. (419) All contact angles measured were above 35 degrees for the silver/copper-doped scaffolds. The contact angle generally increased as the immersion period in the salt bath increased but it did not fall in the range that is deemed most suitable for cell attachment of between 48 and 62 degrees. (419) As with the white light interferometry data for the silver/copper scaffold, the change of topography shown by the wettability

172 results will likely have an effect on the behaviour of the cells that are seeded onto it and this is discussed in chapter six.

Figure 77:- Wettability data for the combination ion doped BBG scaffolds showing spectra for samples before and after immersion in SBF for 14 days (Data shown are mean values and standard deviation of at least 3 measurements). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

ICP-MS analysis was carried out on the silver/copper-doped samples to obtain the amount of copper and silver that is being released from the doped scaffolds during immersion in PBS, shown in Figure 78. This shows that as the concentration of copper and silver in the sample increased so did the released amount of the ions and the release rate. This result indicates far more copper was being released from the samples compared to silver, which supports the depth EDX data shown in Figure 71 and Figure 76. This analysis also showed that the release rate for the silver compared to copper was significantly slower. Silver was released at a rate of approximately 27.11mg per litre per day while copper was being released at a rate between 35.47±0.05mg per litre per day and 52.48±0.05mg per litre per day. To put that into perspective the release rates extrapolate down to approximately 0.0011mg per ml per hour for silver and between 0.0015mg per ml per hour and 0.0022mg per ml per hour for copper.

173

Figure 78:- ICP-MS data for the silver/copper doped samples obtained by different immersion periods in the salt bath, A) showing the release of silver from the samples and B) showing the release of copper from the samples. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

5.6 Polymer coated 45S5 Bioglass®- derived glass-ceramic scaffolds

This section presents the results obtained when gelatin was used to coat pre-fabricated BBG scaffolds. The “proof of concept” work is shown in section 5.6.1 which is followed by the more extensive structural characterisation of these novel composite samples shown in section 5.6.2 with PDLLA coated scaffold acting as a control sample.

5.6.1 Preliminary processing of gelatin coated 45S5 Bioglass®- derived glass-ceramic scaffolds

This study involved coating BBG scaffolds with the two types of gelatin at different concentrations to determine the best concentration leading to uniform coatings without reducing the porosity of the scaffold detrimentally, as described in section 4.2.3II. The concentrations of gelatin used for this preliminary study are listed in Table 5. Four scaffold coating immersion periods were investigated: one, five, ten and 15 minutes. The porosity of the coated scaffolds was calculated using Equation 8, with the density of gelatin being 0.777gcm-3 and 0.812gcm-3 for type A and B, respectively. (390) (391) The porosity results for the coated scaffolds are shown in Table 24 and Table 25 for BBG scaffolds coated in type A gelatin and type B gelatin, respectively.

174

Table 24:- Porosity data for gelatin type A coated BBG scaffolds for a range of polymer immersion periods. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Immersion period in the gelatin suspension Gelatin 1 minute 5 minutes 10 minutes 15 minutes amount Porosity (%) % wt before after before after before after before after 5 94±2 * 93±3 * 88±2 * 87±3 * 95±2 * 93±1 * 92±3 * 90±1 * 6 90±2 * 87±1 * 90±1 * 87±1 * 92±1 * 88±2 * 83±1 * 79±3 * 7 93±2 * 87±2 * 92±1 * 84±1 * 88±2 * 81±2 * 91±1 * 85±3 * 8 93±1 * 85±1 * 92±2 * 82±2 * 89±2 * 77±1 * 84±3 * 76±2 * 9 89±1 * 77±3 * 93±2 * 78±1 * 93±2 * 70±2 * 93±2 * 70±1 * 10 94±3 * 79±3 * 90±1 * 70±1 * 88±2 * 66±3 * 90±2 * 70±1 * 11 95±3 * 84±2 * 93±2 * 79±2 * 79±2 * 69±2 * 92±2 * 76±2 * 12 96±1 * 87±1 * 92±2 * 78±3 * 96±3 * 87±1 * 91±2 * 71±1 * 13 90±1 * 71±2 * 93±1 * 74±3 * 91±3 * 72±2 * 90±1 * 65±1 * 14 88±1 * 80±2 * 88±1 * 70±2 * 90±1 * 70±1 * 92±2 * 71±2 * 15 90±2 * 75±2 * 86±1 * 69±2 * 86±1 * 69±1 * 87±2 * 72±2 *

Table 25:- Porosity data for gelatin type B coated BBG scaffolds for a range of polymer immersion periods. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Immersion period in gelatin suspension Gelatin 1 minute 5 minutes 10 minutes 15 minutes amount Porosity (%) % wt before after before after before after before after 5 90±1 * 90±2 * 84±1 * 84±3 * 90±2 * 88±2 * 87±1 * 85±1 * 6 90±1 * 89±1 * 88±2 * 86±2 * 85±1 * 84±2 * 90±2 * 88±2 * 7 85±1 * 84±2 * 87±2 * 87±1 * 85±2 * 83±2 * 85±2 * 83±1 * 8 92±2 * 88±1 * 87±2 * 86±1 * 92±1 * 90±1 * 89±1 * 86±1 * 9 94±2 * 89±2 * 94±2 * 90±2 * 94±2 * 91±2 * 91±2 * 84±2 * 10 91±2 * 86±2 * 94±1 * 87±1 * 88±2 * 82±2 * 89±1 * 87±2 *

Table 24 indicates that concentrations of 5wt% and 6wt% of type A gelatin with a scaffold immersion period of one minute gave the best trade-off between a desirable scaffold porosity of above 70% and forming a homogeneous coating on the surface of the scaffold. Table 25 also indicates that with an immersion period of one minute concentrations of 5wt% and 6wt% of type B gelatin gave the best trade-off between porosity and formation of a homogeneous layer on the surface of the scaffold. Porosity above 70% is desirable for bone tissue scaffolds and when, in a later study (section 5.7), copper-doped and silver/copper-doped scaffolds are coated in gelatin the overall porosity should still be above this threshold. Since it has been seen, in sections 5.4 and 5.5, that the MSIE process reduces

175 the porosity up to 20% compared to the BBG scaffolds porosity (93±2%) it is important that the addition of a polymer coat does not further significantly reduce the porosity of the modified scaffold.

Figure 79 shows SEM images of 5wt% and 6wt% coated scaffolds with a one minute coating period for both types of gelatin. Compared to the SEM images for the BBG scaffolds, shown in Figure 29, it is observed that gelatin formed a smooth homogeneous layer over the surface whilst not blocking the pore structure. It should be noted that occasionally the polymer (in this work PDLLA and gelatin) will block some of the pores (as seen in Figure 79A)), however it is generally minimal and does not affect the overall porosity to a significant level.

Multiple immersions in the gelatin solution to form a multi-layer coat was investigated for the two types of gelatin using concentrations of 5wt% and 6wt% with an immersion period of one minute, as described in 4.3.II and Figure 21. The measured porosity for these multi-layer gelatin coated 45S5 Bioglass® -derived scaffolds is shown in Table 26. The results show that introducing multiple layers of gelatin reduced the porosity but not by a significant amount with all coated samples remaining above

176 the desirable 70% threshold. (4) It should be noted however, that the double and triple layered samples were more difficult to handle and required a longer drying period in order for the coating to be stable and ready for characterisation.

Table 26:- Porosity data for multiple coats of gelatin on BBG scaffolds. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Coating Original porosity 1st coat porosity 2nd coat porosity 3rd coat porosity type (%) (%) (%) (%) 5% type A 90±1 * 90±2 * 89±3 * 88±3 * 6% type A 91±1 * 88±1 * 87±3 * 86±2 * 5% type B 94±1 * 92±1 * 91±1 * 90±1 * 6% type B 93±1 * 90±2 * 89±2 * 89±2 *

SEM images of multiple gelatin layer coated BBG scaffolds for both gelatin types with concentrations of 5wt% and 6wt% are shown in Figure 80 and Figure 81, respectively. They both show the smooth and homogeneous coatings similar to that observed for the single coated samples, in Figure 79. In particular, Figure 80B) shows the formation of two separate layers and not the merging into one single layer during the coating process.

Figure 80:- BBG scaffolds coated with gelatin type A with A) two coats of 5wt% type A gelatin (magnification of x350), B) three coats of 5wt% type A gelatin (magnification of x350), C) two coats of 6wt% type A gelatin (magnification of x450) and D) three coats of 6wt% type A gelatin (magnification of x450) with a coating immersion time of one minute for each coat (arrows indicate the gelatin coating)

177

From this preliminary study it can concluded that the most suitable concentration of gelatin in the coating process are 5wt% and 6wt% for both gelatin types with a single immersion cycle for one minute. These parameters gave the best trade-off between a desirable porosity and a single homogeneous coating layer. A single coat was preferred because of the favourable crack propagation observed in the SEM images in Figure 79, Figure 80 and Figure 81. In Figure 79D) fraying at the edges of the gelatin coating is observed which indicates that as the crack occurred and the gelatin stretches to keep the two sides of the break together which can contribute to the toughness of the scaffold by a crack bridging mechanism. (452) (453) (454) This behaviour is not present in the multi-layer scaffolds shown in Figure 80 and Figure 81.

Figure 81:- BBG scaffolds coated with gelatin type B with A) two coats of 5wt% type B gelatin (magnification of x700), B) three coats of 5wt% type B gelatin (magnification of x400), C) two coats of 6wt% type B gelatin (magnification of x900) and D) three coats of 6wt% type B gelatin (magnification of x350) with a coating immersion time of one minute (arrows indicate the gelatin coating)

178

5.6.2 Structural characterisation of polymer coated 45S5 Bioglass®- derived glass-ceramic scaffolds

As described in the previous section the parameters used for this study are concentrations of gelatin (both types) in the coating procedure (5wt% and 6wt%) with an immersion period of one minute. In this study the gelatin samples were characterised alongside a 5wt% PDLLA coated BBG scaffold for comparison, given that PDLLA coating on similar bioactive scaffolds has already been investigated. (110) (352)

As shown in the previous section, the porosities of the gelatin coated samples were found to be 93±3% and 90±2% for 5wt% and 6wt% type A gelatin coated samples, respectively, and 90 ±2% and 89±1% for 5wt% and 6wt% type B gelatin coated samples, respectively, and 90±3% for the 5wt% PDLLA coated scaffolds. (110) (352) These results show that the gelatin coated scaffolds have a similar porosity to that obtained when using PDLLA as a coating agent. (110) (352)

Figure 82 shows PDLLA coated scaffolds, similarly to the gelatin coated scaffolds (of all types), shown in Figure 79, the addition of the polymer coating has reduced the surface roughness compared to the uncoated BBG scaffolds, shown in Figure 29, whilst not negatively affecting the pore interconnectivity and size. Although Figure 82A) does show that the PDLLA coating can occasionally block the pores of the scaffold but this was found not to have a significant detrimental effect on the coated scaffolds porosity.

The XRD spectra of the polymer coated samples, shown in Figure 83, exhibit the characteristic peaks for the BBG scaffolds seen in Figure 32, with the exception of the halo between 5o and 17o which corresponds to the presence of the gelatin and PDLLA coatings. (110) (352)

179

Figure 82:- BBG scaffolds coated with PDLLA with A) a 5wt% coating (magnification of x50), B) a 5wt% coating (magnification of x430), C) a 5wt% coating (magnification of x250) and D) a 5wt% coating (magnification of x1000) (arrows indicate the PDLLA coating)

Figure 83:- XRD spectra for gelatin and PDLLA coated BBG scaffolds

180

FTIR analysis was carried out on all four gelatin coated samples and on the PDLLA coated sample and results are presented in Figure 84 and Figure 85, respectively. Figure 84 shows the FTIR spectra for the 5wt% PDLLA coated scaffolds showing the characteristic peaks of certain groups found within PDLLA according to the literature. (455) These include a –OH group at 3502cm-1, a CH3 and – CH group between 2995cm-1 and 2847cm-1 and a carbonyl group at 1756cm-1, which are consistent with literature results. (455) An expanded view of the peaks attributed to PDLLA, from 3750cm-1 to 1500cm- 1, in the spectra shown in Figure 84 can be found in appendix C as some of the peaks are difficult to see when observing the full spectrum to include the peaks attributed to the BBG. Between 1200cm-1 and 455cm-1 the peaks indicative of the BBG scaffold are observed which is consistent with the FTIR spectra in Figure 31 and with literature results. (413) (414) (415)

Figure 84:- FTIR spectra for PDLLA coated BBG scaffolds before and after immersion in SBF for 14 days with A) showing both spectra indicating the peaks corresponding to PDLLA and B) depicting the distinctive peaks characteristic of BBG scaffolds

Figure 85 shows the results of the FTIR analysis for the gelatin coated scaffolds. Spectra exhibit indicative peaks for the BBG scaffold between 1200cm-1 and 455cm-1 consistent with results shown in Figure 31 and with the literature. (413) (414) (415) Peaks that, according to the literature, indicate the presence of gelatin are also observed including an amide A group between 3625cm-1 and 2750cm-1, water at between 2500cm-1 and 2250cm-1 and amide I, II, III between 1700cm-1 and 1100cm-1. (456) (457) An expanded view of the peaks attributed to the gelatin, from 3750cm-1 to 1200cm-1, in the spectra

181 shown in Figure 85 can be found in appendix C as some of the peaks are difficult to see when observing the full spectrum to include the peaks attributed to the BBG.

Figure 85:- FTIR spectra for gelatin coated BBG scaffolds

The EDX analysis of the polymer coated samples is shown in Figure 86, characteristic peaks that are indicative of the BBG scaffolds are observed, as shown also in Figure 33.

Figure 86:- EDX analysis of polymer coated BBG scaffolds including A) Type A 5wt% gelatin coated scaffold, B) Type A 6wt% gelatin coated scaffold, C) Type B 5wt% gelatin coated scaffold, D) Type B 6wt% gelatin coated scaffold and E) PDLLA coated scaffold.

182

The compressive strength of the coated scaffolds was measured and results are presented in Table 27. The values are significantly higher than those measured for the BBG scaffolds at 0.53±0.08MPa. This is due to the presence of the polymer coating confirming that gelatin coatings improved the mechanical competence of the BBG scaffolds to a higher extent than PDLLA.

Table 27:- Compressive strength of polymer coated BBG scaffolds. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Polymer type Compressive strength and wt% (MPa) Gelatin type A 5wt% 1.4 ± 0.1 * Gelatin type A 6wt% 1.4 ± 0.1 * Gelatin type B 5wt% 1.4 ± 0.1 * Gelatin type B 6wt% 1.4 ± 0.1 * PDLLA 5wt% 0.7 ± 0.1 *

The white light interferometry analysis of the gelatin and PDLLA coated samples is shown in Table 28. These RMS, Ra and PV values are all significantly lower than those recorded for the BBG samples at 3±1µm, 2.6±0.9µm and 19±9µm, respectively. This result suggests that the addition of the polymer coating reduces the surface roughness which is supported by the SEM images shown in Figure 79 and Figure 82 .

Table 28:- White light interferometry results for the surface roughness of polymer coated BBG pellets. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Polymer type RMS Ra PV and wt% (µm) (µm) (µm) Gelatin type A 5wt% 0.5±0.1 * 0.4±0.1 * 4±1 * Gelatin type A 6wt% 0.5±0.1 * 0.4±0.1 * 4±1 * Gelatin type B 5wt% 0.5±0.2 * 0.4±0.1 * 4±1 * Gelatin type B 6wt% 0.5±0.1 * 0.4±0.1 * 4±1 * PDLLA 5wt% 0.5±0.1 * 0.3±0.1 * 4.0±0.9 *

The samples bioactivity was assessed by immersing the polymer coated samples in SBF for 14 days and re-characterising them. The porosity of the SBF immersed gelatin coated scaffolds was measured at 72±1% and 71±3% for 5wt% and 6wt% type A gelatin coated samples, respectively and at 70±2% and 69±1for 5wt% and 6wt% type B gelatin coated samples, respectively. The porosity of the PDLLA coated scaffolds was measured to be 71±2%. Compared to the porosity of the BBG scaffold after the same period of immersion in SBF (84±2%), these porosities are lower but they are in line with the reduction in porosity recorded after immersion in SBF for the BBG scaffolds. Whether this

183 topography has an effect on a cells ability to attach itself to the scaffold and proliferate has not been explored in this work but for future studies this needs to be addressed and this will be discussed in detail later in this chapter.

SEM images of gelatin coated scaffolds after immersion in SBF for 14 days are shown in Figure 87 and Figure 88. Figure 87 shows the SBF immersed type A (5wt% and 6wt%) gelatin coated BBG scaffolds and Figure 88 shows the SBF immersed type B (5wt% and 6wt%) gelatin coated BBG scaffolds. As with the BBG scaffolds, shown in Figure 34, after immersion in SBF the presence of the characteristic HA knobbles is detected on the surface of the gelatin coated scaffolds which indicates that the presence of the gelatin coating does not have a detrimental effect on HA precipitation on the surface of the scaffold struts. The presence of the observed HA is observed both on top of the gelatin coating and in areas where the coating has degraded. It’s likely that HA formation has started in an area of the struts not completely covered by gelatin, especially considering the irregular and rough topography of the BBG scaffolds. In the literature, ion diffusion through gelatin has been observed which would explain the presence of HA on top of the gelatin coating as ions released from the scaffold can diffuse through the gelatin coating and interact with the SBF solution to form HA. (458) (459) (460) The time taken for ions to diffuse through the gelatin coating would explain the retarded level of HA present on the gelatin coated samples compared to that on the BBG scaffolds (observed in Figure 34).

Figure 87 and Figure 88 both show the gelatin coatings starting to disintegrate after 14 days immersed in SBF and both show HA formation on top of the gelatin coating and in regions where the coating has lifted up or has degraded HA knobbles are present. Figure 89 shows SEM images of the PDLLA coated scaffolds after immersion in SBF for 14 days. These images also show the presence of the characteristic HA knobbles indicating the samples bioactivity, as seen in the base BBG scaffolds in Figure 34.

184

Figure 87:- BBG scaffolds coated with gelatin type A with A) a 5wt% coating (magnification of x750), B) a 5wt% coating (magnification of x900), C) a 6wt% coating (magnification of x1900) and D) a 6wt% coating (magnification of x900) after immersion in SBF for 14 days (Black arrows indicating the gelatin and white arrows indicating the HA)

Figure 88:- BBG scaffolds coated with gelatin type B with A) a 5wt% coating (magnification of x550), B) a 5wt% coating (magnification of x100), C) a 6wt% coating (magnification of x600) and D) a 6wt% coating (magnification of x450) after immersion in SBF for 14 days (Black arrows indicating the gelatin and white arrows indicating the HA)

185

Figure 89:- BBG scaffolds coated with PDLLA with A) a 5wt% coating (magnification of x50) and B) a 5wt% coating (magnification of x1700) after immersion in SBF for 14 days (Black arrows indicating the PDLLA and white arrows indicating the HA)

These SEM images for the polymer coated scaffolds (Figure 87, Figure 88 and Figure 89) suggest the presence of the characteristic HA knobbles. However, compared to the BBG scaffold after the same immersion period in SBF (shown in Figure 34), the quantity of HA as judged by SEM observation seems to be less on the polymer coated scaffolds, especially for the PDLLA coated samples. It can be hypothesised that the direct contact of the bioactive glass-ceramic surface with SBF, which is required to start the HA formation process is being impaired by the presence of the polymer coating, especially in the case of the PDLLA samples. Only in areas of incomplete coating or as the polymers begin degrade a direct contact between the SBF and the BBG surface occurs leading to HA formation. In the case of gelatin ion diffusion through the coating layer should facilitate the formation of HA on top of the gelatin coating.

XRD analysis of the polymer coated samples after SBF immersion is shown in Figure 90. The presence of a halo between 2ϐ 5 and 17 degrees indicates the presence of the gelatin and PDLLA. (110) (352) The XRD spectra also show the presence of the peak indicative of the formation of HA, as previously shown for the BBG scaffold after SBF immersion (shown in Figure 37). This result indicates that the polymer coated scaffolds are bioactive.

186

Figure 90:- XRD spectra for gelatin coated BBG scaffolds and PDLLA coated 45S5 Bioglass® -derived scaffolds after immersion in SBF for 14 days.

The FTIR analysis carried out on the gelatin coated samples is shown in Figure 91. The indicative gelatin peaks are present, as shown in Figure 85 , including an amide A group between 3625cm-1 and 2750cm-1, water at between 2500cm-1 and 2250cm-1 and an amide I, II, III between 1700cm-1 and 1100cm-1. (456) (457) The indicative peaks for the BBG scaffold are present between 1200cm-1 and 455cm-1 consistent with Figure 36 and the literature. (423) (424) The changes in the spectra recorded before and after SBF immersion are consistent with the precipitation of HA on the surface of the sample (shown from Figure 31 and Figure 36). Including the decrease in intensity of the Si-O-Si bonds at 1100cm-1 and 1037cm-1, (424) (222) the shift and reduction of the Si-O (the non- bridging oxygen) bond from 940cm-1- 920cm-1 to 980cm-1 – 940cm-1 (423) (424) (222) and the formation of the C-O bond between 880cm-1 and 900cm-1. (425) An expanded view of the peaks attributed to gelatin, from 3750cm-1 to 1200cm-1, in the spectra shown in Figure 91 can be found in appendix C as some of the peaks are difficult to see when observing the full spectrum to include the peaks attributed to the BBG.

187

Figure 91:- FTIR spectra for gelatin coated BBG samples after immersion in SBF for 14days showing A) the full spectra and B) the area of the spectra showing the peaks indicative of the BBG

The FTIR spectra for the PDLLA coated samples after immersion in SBF are shown in Figure

-1 84. The peaks indicative of PDLLA are shown, including a –OH group at 3502cm , a CH3 and –CH group between 2995cm-1 and 2847cm-1 and a carbonyl group at 1756cm-1 consistent with the literature. (455) The FTIR spectra also shows between 1200cm-1 and 455cm-1, the peaks indicative of the BBG scaffold after immersion in SBF. (423) (424) (222) (425) The spectra indicate the intensity decrease of the Si-O-Si bonds at 1100cm-1 and 1037cm-1, (424) (222) the movement and reduction of the Si-O (the non-bridging oxygen) bond from 940cm-1- 920cm-1 to 980cm-1 – 940cm-1 (423) (424) (222) and the formation of the C-O bond between 880cm-1 and 900cm-1. (425)

The EDX analysis, shown in Figure 92, was carried out to confirm the possible change in the elemental composition which would indicate the presence of HA in the sample. This is seen as a change in the ratio between calcium and phosphorous peaks to approximately 1.67, as shown also in Figure 35. This result confirms that the polymer coating does not have a significant detrimental effect on the bioactivity of the BBG scaffold.

188

Figure 92:- EDX analysis of polymer coated BBG scaffolds including A) Type A 5wt% gelatin coated scaffold, B) Type A 6wt% gelatin coated scaffold, C) Type B 5wt% gelatin coated scaffold, D) Type B 6wt% gelatin coated scaffold and E) PDLLA coated scaffold after immersion in SBF for 14 days

The compressive strength values of the polymer coated scaffolds after 14 days of immersion in SBF are shown in Table 29. A vast improvement on the BBG scaffolds (at 0.43±0.05MPa) is recorded. The gelatin coated samples outperformed the PDLLA samples but there was no difference between the two types of gelatin or the different concentrations used.

Table 29:- Compressive strength of polymer coated BBG scaffolds after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Polymer type Compressive strength and wt% (MPa) Gelatin type A 5wt% 1 ± 0.1 * Gelatin type A 6wt% 1 ± 0.1 * Gelatin type B 5wt% 1 ± 0.1 * Gelatin type B 6wt% 1 ± 0.1 * PDLLA 5wt% 0.4 ± 0.1 *

The white light interferometry results for RMS, Ra and PV after the polymer coated scaffolds were immersed in SBF for 14 days are shown in Table 28. These values are lower than those observed for the BBG samples after the sample immersion period in SBF, shown in Table 11. This is due to the presence of the polymer coating starting to degrade and the precipitation of HA on the samples surface. If the modification to the surface topography has an effect on the modified scaffolds ability to encourage cell attachment and other desirable cell behaviour is not observed for this type of scaffold but it is something that needs to be researched in the future and will be discussed in comparison to the BBG scaffolds in the summary of this chapter.

189

Table 30:- White light interferometry results for the surface roughness of polymer coated 45S5 Bioglass® –derived glass-ceramic pellets after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Polymer type RMS Ra PV and wt% (µm) (µm) (µm) Gelatin type A 5wt% 5.0±0.1 * 3.2±0.1 * 16±4 * Gelatin type A 6wt% 5.5±0.2 * 4.0±0.1 * 16±4 * Gelatin type B 5wt% 4.7±0.1 * 3.6±0.1 * 17±6 * Gelatin type B 6wt% 5.1±0.1 * 4.0±0.1 * 17±6 * PDLLA 5wt% 6.0±0.2 * 4.9±0.2 * 20±5 *

The wettability values for the polymer coated samples before and after immersion in SBF is shown in Figure 93. Significantly the 5wt% samples of both types of gelatin exhibited lower contact angles compared to those observed for the 6wt% gelatin coated samples and the PDLLA coated samples and are outside the ideal contact angle range of 48 to 65 degrees. (419) The values were similar to those observed for the uncoated scaffolds of 28±4 degrees and 23±2 degrees, before and after SBF immersion, respectively. However the 6wt% gelatin coated samples and the PDLLA coated samples both fall within in the ideal contact angle range pre and post SBF immersion. As with the white light interferometry data for this work, whether the changes in topography and contact angle will have an effect on the scaffolds ability to interact with cells in a positive manner has not been determined for these scaffolds but it is something that needs to be addressed and will be discussed at the end of the chapter.

Figure 93:- Wettability data for the gelatin coated BBG samples showing spectra from before and after immersion in SBF for 14 days (Data shown are mean values and standard deviation of at least 3 measurements). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

190

5.7 Polymer coated molten salt ion exchanged 45S5 Bioglass®- derived glass- ceramic scaffolds

This section presents the results obtained when the MSIE technique was used to produce copper-doped and silver/copper-doped BBG scaffolds, which were then coated in gelatin. The copper- doped and gelatin coated BBG scaffold results are presented in section 5.71 and the results for the silver/copper-doped gelatin coated BBG scaffolds are shown in section 5.7.2.

5.7.1 Structural characterisation of polymer coated copper molten salt ion exchanged scaffolds

This study investigated copper-doped scaffolds, as described in section 4.2.2.II and depicted in Figure 19, which were then coated in gelatin. The lowest concentration of copper in the molten salt bath was used with an immersion period of 15 minutes. The MSIE scaffolds were then coated in one of the four variations of gelatin coating from 5wt% type A, 6wt% type A, 5wt% type B and 6wt% type B.

The porosity values of the gelatin coated copper-doped scaffolds were 70±7%, 70±7%, 69±7% and 68±7% for 5wt% type A, 6wt% type A, 5wt% type B and 6wt% type B, respectively. These are lower than porosities observed for the copper-doped scaffolds (72±6%) and the gelatin coated scaffolds (93±3% and 90±2% for 5wt% and 6wt% type A gelatin coated samples, respectively, and 90 ±2% and 89±1% for 5wt% and 6wt% type B gelatin coated samples, respectively). This reduction of the porosity compared to the BBG scaffolds porosity (93±2%) is mainly due to the MSIE process rather than the polymer coating. These recorded porosities are still above the ideal porosity threshold of 70%, described in the literature review. (210)

SEM images of the four samples types, shown in Figure 94, indicate that the surface of the scaffold has the peaks and troughs typical of BBG scaffolds and copper-doped scaffolds, shown in Figure 29 and Figure 55, respectively, but the surface exhibits the overall smoothness and homogenous structure of the gelatin coated scaffolds, shown previously in Figure 80 and Figure 81. The pore size and dimensions as well as the pore interconnectivity that are desirable for scaffolds for BTE are maintained.

191

Figure 94:- 45S5 Bioglass® –derived glass-ceramic scaffolds doped with copper and then coated with gelatin with A) a 5wt% coating of type A gelatin (magnification of x550), B) a 6wt% coating of type A gelatin (magnification of x250), C) a 5wt% coating of type B gelatin (magnification of x700) and D) a 6wt% coating of type B gelatin (magnification of x900) (arrows indicate the gelatin coating)

The XRD analysis of the copper-doped gelatin coated scaffolds is shown in Figure 95. The characteristic peaks observed for the BBG scaffolds (shown in Figure 32) are observed. (4) (5) (389) (416) The characteristic peaks that were observed in the copper-doped scaffolds (shown in Figure 56) (441) (442) (443) and the peak between 5 and 17 degrees indicating the presence of the amorphous gelatin coating are also present. (110) (352) As stated in section 5.4 the presence of copper can not be easily attributed to a unique crystalline phase due to the fact that in some samples not all the phases where present or overlapping peaks occurred.

192

Figure 95:- XRD analysis of gelatin coated copper-doped 45S5 Bioglass® -derived scaffolds

The results of the FTIR analysis of the copper-doped gelatin coated samples is presented in Figure 96 showing the indicative peaks for copper-doped BBG scaffolds, as seen in Figure 57 between

-1 -1 - -1 1400cm and 455cm . The NO 3 bond at 1387cm indicates the presence of a nitrate, a shifted Si-O bond at 950-1000cm-1, (444) (445) a pair of P-O bending bonds at 720 – 650cm-1 and 625 – 580cm-1, the second shifted Si-O peak was detected at 670-680cm-1 (444) (445) (446) and the pair of Si-O-Si bending bonds at 528 – 536cm-1 and 455 – 457cm-1. Figure 96 also shows the peaks indicative of the presence of gelatin, as seen in Figure 85, with an amide A group between 3625cm-1 and 2750cm-1, water at between 2500cm-1 and 2250cm-1 and an amide I, II, III between 1700cm-1 and 1100cm-1. (456) (457) An expanded view of the peaks attributed to the gelatin, from 3750cm-1 to 1200cm-1, in the spectra shown in Figure 96 can be found in appendix C as some of the peaks are difficult to see when observing the full spectrum to include the peaks attributed to the BBG.

Figure 97 shows the EDX analysis for the copper-doped gelatin coated samples. The presence of copper peak at 0.5keV and 8keV is confirmed which is consistent with the copper-doped only samples shown in Figure 58 and the literature. (447) The spectrum also shows peaks indicative of the elements that make up the BBG scaffold as seen in Figure 33.

193

Figure 96:- FTIR analysis of gelatin coated copper doped BBG scaffolds showing A) the full spectrum and B) the part of the spectrum exhibiting the peaks indicative of the copper-doped scaffold

Figure 97:- EDX analysis of polymer coated copper-doped BBG scaffolds including A) Type A 5wt% gelatin coated scaffold, B) Type A 6wt% gelatin coated scaffold, C) Type B 5wt% gelatin coated scaffold and D) Type B 6wt% gelatin coated scaffold

The compressive strength of the copper-doped gelatin coated scaffolds is shown in Table 31. The values indicate an increase in the compressive strength of the samples compared to the value (0.53±0.08MPa) of the BBG scaffolds. (113) (4) These values are higher than those of the copper-doped scaffolds, shown in Table 16, and the values measured for the gelatin coated scaffolds shown in Table

194

27. This result indicates that the increased compressive strength observed from the copper-doped gelatin coated scaffolds compared to the BBG scaffolds is mainly due to the introduction of the polymer coating but a fraction of it is likely the result of the MSIE process.

Table 31:- Compressive strength of copper doped gelatin coated BBG scaffolds. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Polymer type Compressive strength and wt% (MPa) Gelatin type A 5wt% 1.5 ± 0.1 * Gelatin type A 6wt% 1.5 ± 0.1 * Gelatin type B 5wt% 1.5 ± 0.1 * Gelatin type B 6wt% 1.5 ± 0.1 *

The white light interferometry analysis for the copper-doped gelatin coated scaffolds is shown in Table 32. The RMS, Ra and PV values are all significantly lower than those recorded for the BBG samples at 3±1µm, 2.6±0.9µm and 19±9µm, respectively, and for the copper-doped samples at 5±1µm, 3.8±0.9µm and 26±5µm, respectively. However these values are in line with those observed for the gelatin coated scaffolds shown in Table 28, indicating that the topography of the scaffold was altered by the addition of the polymer coating, as expected. How this change to the topography affects the scaffolds interaction with cells has not been investigated in this work, however it has been discussed in relation to the other types of modified scaffolds later in this chapter.

Table 32:- White light interferometry results for the surface roughness of gelatin coated copper doped BBG pellets. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Polymer type RMS Ra PV and wt% (µm) (µm) (µm) Gelatin type A 5wt% 0.5±0.1 * 0.4±0.1 * 4±1 * Gelatin type A 6wt% 0.5±0.1 * 0.4±0.1 * 4±1 * Gelatin type B 5wt% 0.5±0.1 * 0.4±0.1 * 4±1 * Gelatin type B 6wt% 0.5±0.1 * 0.4±0.1 * 4±1 *

The copper-doped gelatin coated scaffolds bioactivity was determined through characterisation of the samples after immersion in SBF for 14 days. The porosities of the copper-doped gelatin coated scaffolds were measured as 65±7%, 66±7%, 66±6% and 66±7% for type A 5wt%, type A 6wt%, type B 5wt% and type B 6wt% gelatin coated copper-doped scaffolds. The porosity of the samples was observed to fall in line with the decrease in porosity measured for the BBG scaffolds (approximately 10%).

195

SEM images of copper-doped gelatin coated scaffolds are shown in Figure 98. There is limited indication of the presence of the gelatin coating (observed in Figure 94) but the presence of the characteristic HA knobbles is confirmed, which is consistent with the SEM images in Figure 34 and in the literature, (4) (5) indicating that the presence of the copper or gelatin coating does not have a detrimental effect on the samples bioactivity, as seen in section 5.4 and 5.6 for copper-doped scaffolds and gelatin coated scaffolds, respectively. SEM images show an increase in the surface roughness of the struts which can be attributed to the presence of HA crystals precipitated onto the samples surface.

Figure 98:- BBG scaffolds doped with copper and then coated with gelatin with A) 5wt% coating of type A gelatin (magnification of x300), B) 6wt% coating of type A gelatin (magnification of x500), C) 5wt% coating of type B gelatin (magnification of x170) and D) 6wt% coating of type B gelatin (magnification of x250) after immersion in SBF for 14days (Black arrows indicating the gelatin and white arrows indicating the HA)

The XRD analysis of the copper-doped gelatin coated scaffolds is shown in Figure 99. The spectra exhibits the characteristic peaks observed for the BBG scaffold that has undergone SBF immersion, shown in Figure 37. (5) (4) (426) The spectra also show the reduced characteristic peaks that were observed for copper containing phases in the copper-doped scaffolds (shown in Figure 61) (441) (442) (443) and the peak between 5 and 17 degrees indicating the presence of the amorphous gelatin coating (shown in Figure 90). (110) (352) As stated in section 5.4 the presence of copper can not be

196 easily attributed to a unique phase due to the fact that in some samples not all the phases were present or peak overlapping occurred.

Figure 99:- XRD analysis of gelatin coated and copper-doped BBG scaffolds after immersion in SBF for 14 days

The FTIR analysis of the gelatin coated copper-doped scaffolds after immersion in SBF for 14 days is shown in Figure 100. The spectra shows all the characteristic gelatin peaks, as seen also in Figure 85, including an amide A group between 3625cm-1 and 2750cm-1, water at between 2500cm-1 and 2250cm-1 and an amide I, II, III between 1700cm-1 and 1100cm-1. (456) (457) The indicative peaks of a copper-doped scaffold are also present between 1400cm-1 and 455cm-1, as seen in Figure 57. An expanded view of the peaks attributed to the gelatin, from 3750cm-1 to 1200cm-1, in the spectra shown in Figure 100 can be found in appendix C as some of the peaks are difficult to see when observing the full spectrum to include the peaks attributed to the BBG.

The peaks characterising the bioactivity in BBG scaffolds are present in the form of the C-O bond between 880cm-1 and 900cm-1, (425) the P-O bond at 875cm-1 to 825cm-1 increasing in size (423) (424) (222) and the dual P-O bond at 760cm-1 to 650cm-1 merging. (424) The peaks associated with a

- -1 copper-doped sample after SBF immersion include a reduced NO 3 bond at 1387cm , a reduced shifted Si-O bond at 950-1000cm-1 (444) (445) (446) and the second shifted Si-O peak detected at 670-680cm- 1 which has been reduced to such an extent that it is almost undetectable indicating that this bond breaks first when the copper diffuses out of the crystal structure during immersion in SBF.

197

Figure 100:- FTIR analysis of gelatin coated copper doped BBG scaffolds after immersion in SBF for 14 days showing A) the full spectrum and B) the part of the spectrum that presents the peaks indicative of the doped scaffold

The EDX analysis for the copper-doped gelatin coated samples is shown in Figure 101. The change in the composition is observed, which is indicative of HA precipitation with the ratio between the phosphorous and calcium changing to approximately 1.67 (as seen in Figure 35). The presence of copper is also detected at 0.5keV and 8keV, which is consistent with the results shown in section 5.4 and in the literature. (447) This result supports the premise that the addition of copper and coating in gelatin do not have a detrimental effect on the bioactivity of the sample.

Figure 101:- EDX analysis of polymer coated copper-doped BBG scaffolds including A) Type A 5wt% gelatin coated scaffold, B) Type A 6wt% gelatin coated scaffold, C) Type B 5wt% gelatin coated scaffold and D) Type B 6wt% gelatin coated scaffold after immersion in SBF for 14 days

198

The compressive strength of the copper-doped gelatin coated scaffold decreased upon immersion in SBF for 14 days, as shown in Table 33. This result is similar to the previous results and to data in the literature which confirmed that combination of gelatin coating degradation, ion release and HA precipitation has a detrimental effect on the mechanical competence of the scaffold. (352) (418) The compressive strength values are higher than the corresponding value (0.43±0.05MPa) of the BBG scaffold, indicating that copper release and degradation of the gelatin coating which have occurred after 14 days in SBF still have a positive effect on the compressive strength of the scaffold.

Table 33:- Compressive strength of gelatin coated copper doped BBG scaffolds after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Polymer type Compressive strength and wt% (MPa) Gelatin type A 5wt% 1 ± 0.1 * Gelatin type A 6wt% 1 ± 0.1 * Gelatin type B 5wt% 1 ± 0.1 * Gelatin type B 6wt% 1 ± 0.1 *

The white light interferometry results for RMS, Ra and PV after the copper-doped gelatin coated samples were immersed in SBF for 14 days are shown in Table 34. These values are lower than those observed for the BBG samples after the same immersion period in SBF, shown in Table 11. This result is likely due to the presence of the polymer coating which has started to degrade and is also affected by the precipitation of HA on the samples surface. How this compares to the other modified scaffolds in terms of how it would be perform in cell culture tests because of the change in surface roughness has been discussed later in this chapter.

Table 34:- White light interferometry results for the surface roughness of gelatin coated copper doped BBG pellets after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Polymer type RMS Ra PV and wt% (µm) (µm) (µm) Gelatin type A 5wt% 5.1±0.1 * 4.1±0.2 * 16±6 * Gelatin type A 6wt% 5.1±0.2 * 4.1±0.1 * 16±6 * Gelatin type B 5wt% 5.0±0.1 * 4.0±0.1 * 16±4 * Gelatin type B 6wt% 5.0±0.1 * 4.1±0.2 * 16±5 *

The wettability results are shown in Figure 102 indicating that the measured contact angles were all substantially higher than those recorded for the BBG samples of 28±4o and 23±2o, before and after SBF immersion, respectively. In fact these wettability values were higher than those of the separate

199 copper-doped samples and the gelatin coated BBG scaffolds, indicating that the combination of the MSIE with polymer coating increases the contact angle measured. The contact angles for the copper- doped gelatin coated scaffolds were all high enough to fall within the limit of the ideal range of contact angles indicated to be suitable for cell attachment of between 48o and 65o. (419) There is little difference between the contact angles measured for the two types of gelatin coating or the two concentrations of gelatin used. A measure of how this change in contact angle between the different sample types and compared the BBG scaffold had not been determined in this work but it has been touched upon in the summary of this chapter.

Figure 102:- Wettability data for the gelatin coated copper doped BBG scaffolds before and after immersion in SBF for 14 days (Data shown are mean values and standard deviation of at least 3 measurements). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

5.7.2 Structural characterisation of polymer coated combination molten salt ion exchange scaffolds

This study investigated silver/copper-doped scaffolds, as described in section 4.2.2.III and depicted in Figure 19, which were then coated in gelatin. The shortest immersion period of 15 minutes in the molten salt bath containing both silver and copper was used. The MSIE scaffolds were then coated in one of the four variation of gelatin coating from 5wt% type A, 6wt% type A, 5wt% type B and 6wt% type B.

200

The porosity of the gelatin coated silver/copper-doped scaffolds was 71±7%, 70±7%, 68±7% and 68±7% for 5wt% type A, 6wt% type A, 5wt% type B and 6wt% type B, respectively. These values are lower than porosities observed for the silver/copper-doped scaffolds (72±7%) and the gelatin coated scaffolds (93±3% and 90±2% for 5wt% and 6wt% type A gelatin coated samples, respectively, and 90 ±2% and 89±1% for 5wt% and 6wt% type B gelatin coated samples, respectively). This reduction of the porosity compared to the BBG scaffolds porosity (93±2%) is mainly due to the MSIE process rather than for the polymer coating itself. These recorded porosities are still above the ideal porosity threshold of 70% as discussed in the literature review. (210)

SEM images of the four samples types, as shown in Figure 103, reveal that the surface of the scaffold has the peaks and troughs typical of these type of scaffolds shown in Figure 29 and Figure 67, respectively. However the scaffolds exhibit also the overall smoothness and homogenous structure of the gelatin coated scaffolds shown also in Figure 80 and Figure 81. The scaffolds also maintain the pore size, dimensions and pore interconnectivity that is desirable for scaffolds for BTE, as judged by observation of the SEM micrographs.

Figure 103:- BBG scaffolds doped with copper and silver and then coated with gelatin with A) a 5wt% coating of type A gelatin (magnification of x750), B) a 6wt% coating of type A gelatin (magnification of x900), C) a 5wt% coating of type B gelatin (magnification of x700) and D) a 6wt% coating of type B gelatin (magnification of x250) (arrows indicating the gelatin coating)

201

The XRD analysis of the silver/copper-doped gelatin coated scaffolds is shown in Figure 104. The characteristic peaks of the BBG scaffolds (shown in Figure 32) are present. (4) (5) (389) (416) The spectra show also the characteristic peaks that were observed for copper and silver in the silver/copper- doped scaffolds shown previously (Figure 68) (441) (442) (443) and the halo between 2ϐ 5 and 17 degrees indicating the presence of the amorphous gelatin coating, shown previously in Figure 83. (110) (352) As stated in section 5.4 the presence of copper can not be univocally attributed to a single crystalline phase due to the fact that in some samples not all the phases were present or overlapping of peaks occurred.

Figure 104:- XRD analysis of gelatin coated silver/copper-doped BBG scaffolds

The FTIR analysis of the silver/copper-doped gelatin coated samples is shown in Figure 105 showing the indicative peaks for silver/copper-doped BBG scaffolds, as seen in Figure 69 between

-1 -1 - -1 1400cm and 455cm . The following bonds are identified:- NO 3 bond at 1387cm indicates the presence of a nitrate, a single shifted Si-O bond at 950-1000cm-1, (444) (445) a pair of P-O bending bonds at 720 – 650cm-1 and 625 – 580cm-1, the second shifted Si-O peak was detected at 670-680cm-1 (428) (429) (430) (444) (445) (446) and the pair of Si-O-Si bending bonds at 528 – 536cm-1 and 455 – 457cm-1. Figure 105 also shows the peaks indicative of the presence of gelatin, as seen in Figure 85, with an amide A group between 3625cm-1 and 2750cm-1, water at between 2500cm-1 and 2250cm-1 and an amide I, II, III between 1700cm-1 and 1100cm-1. (456) (457) An expanded view of the peaks attributed to the gelatin, from 3750cm-1 to 1200cm-1, in the spectra shown in Figure 105 can be found in appendix C as some of the peaks are difficult to see when observing the full spectrum to include the peaks attributed to the BBG.

202

Figure 105:- FTIR analysis of gelatin coated silver/copper doped BBG scaffolds showing A) the full spectrum and B) the part of the spectrum which shows the typical peaks indicative of the doped scaffold

Figure 106 shows the EDX analysis for the silver/copper-doped gelatin coated samples indicating the presence of silver at approximately 3keV and copper at 0.5keV and 8keV which is consistent with Figure 70 and the literature. (447) (113) The results also show the other peaks that are indicative of the elements that make up the BBG scaffold as expected and as seen in Figure 33.

Figure 106:- EDX analysis of polymer coated silver/copper-doped BBG scaffolds including A) Type A 5wt% gelatin coated scaffold, B) Type A 6wt% gelatin coated scaffold, C) Type B 5wt% gelatin coated scaffold and D) Type B 6wt% gelatin coated scaffold

The compressive strength of the silver/copper-doped gelatin coated scaffolds is shown in Table 35. An increase in the compressive strength of the samples compared to the value (0.53±0.08MPa) of the BBG scaffolds was measured. (113) (4) These values are higher than those of the silver/copper-

203 doped scaffolds, shown in Table 20, and the values measured for the gelatin coated scaffolds shown in Table 27. The result indicates that the increased compressive strength observed from the silver/copper- doped gelatin coated scaffolds compared to the BBG scaffolds is likely due to the introduction of the polymer coating but a fraction of it is contributed by the MSIE process as was seen with the copper- doped gelatin coated scaffolds shown in Table 31.

Table 35: Compressive strength of silver/copper-doped polymer coated BBG scaffolds. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Polymer type Compressive strength and wt% (MPa) Gelatin type A 5wt% 1.54 ± 0.08 * Gelatin type A 6wt% 1.57 ± 0.04 * Gelatin type B 5wt% 1.55 ± 0.06 * Gelatin type B 6wt% 1.54 ± 0.06 *

The white light interferometry analysis is shown in Table 36 for the silver/copper-doped gelatin coated scaffolds. RMS, Ra and PV values are all significantly lower than those recorded for the BBG samples at 3±1µm, 2.6±0.9µm and 19±9µm respectively and slightly lower than those of the silver/copper-doped samples at 0.9±0.2µm, 0.7±0.1µm and 8±4µm respectively. However these values are in line with those observed for the gelatin coated scaffolds shown in Table 28, indicating that the topography of the scaffold was altered by the addition of the polymer coating. How this compares to the other modified scaffolds in terms of its performance when seeded with cells has not been investigated in this work, however it has been discussed later in this chapter.

Table 36:- White light interferometry results for the surface roughness of gelatin coated, silver/copper-doped BBG pellets. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Polymer type RMS Ra PV and wt% (µm) (µm) (µm) Gelatin type A 5wt% 0.5±0.1 * 0.4±0.2 * 4±1 * Gelatin type A 6wt% 0.4±0.2 * 0.3±0.1 * 5±1 * Gelatin type B 5wt% 0.4±0.3 * 0.3±0.3 * 3±2 * Gelatin type B 6wt% 0.5±0.1 * 0.3±0.1 * 4±1 *

The bioactivity of the silver/copper-doped gelatin coated scaffolds was assessed through re- characterisation of the samples after immersion in SBF for 14 days. The porosities of the silver/copper- doped gelatin coated scaffolds were measured as 65±7%, 66±7%, 65±7% and 64±7% for type A 5wt%, type A 6wt%, type B 5wt% and type B 6wt% gelatin coated silver/copper-doped scaffolds. The porosity

204 of the samples was observed to fall in line with the decrease in porosity measured for the BBG scaffolds (approximately 10%).

SEM images of silver/copper-doped gelatin coated scaffolds are shown in Figure 107 confirming the presence of the gelatin coating (observed in Figure 103) and also showing the presence of the characteristic HA knobbles consistent with the SEM images in Figure 34 and the literature, (4) (5) indicating that the presence of the silver, copper or gelatin does not have a significant detrimental effect on the samples bioactivity (as seen in section 5.5 and 5.6 for silver/copper-doped scaffolds and gelatin coated scaffolds, respectively). The SEM images indicate an increase in the roughness of the surface topography which can be attributed to the presence of the HA precipitated onto the samples surface. SEM images also show that the gelatin is present but has partially degraded exposing the surface of the silver/copper-doped scaffold.

Figure 107:- BBG scaffolds doped with copper and silver and then coated with gelatin with A) a 5wt% coating of type A gelatin (magnification of x150), B) a 6wt% coating of type A gelatin (magnification of x2000), C) a 5wt% coating of type B gelatin (magnification of x160) and D) a 6wt% coating of type B gelatin (magnification of x600) after immersion in SBF for 14days

205

The XRD analysis, shown in Figure 108, indicates the characteristic peaks of a 45S5 Bioglass® -derived scaffold that has undergone SBF immersion, as seen in Figure 37, with the presence of key HA peaks consistent with results in the literature. (5) (4) (426) The indicative peaks of both silver and copper, shown in Figure 104, are also present indicating that silver and copper remain in the sample. (113) (441) (442) (443) The presence of a halo between 2ϐ 5 and 17 degrees which indicates that there is an amorphous phase in the sample caused in this case by the presence of the gelatin, as shown in Figure 90. (110) (352) As stated in section 5.4 the presence of copper can not be confidently attributed to a unique phase due to the fact that in some samples not all the phases where present or overlapping of peaks occurred. Therefore it can be stated the presence of silver, copper and the gelatin coating has not had a significant detrimental effect on the bioactivity of the sample under the conditions investigated.

Figure 108:- XRD analysis of gelatin coated silver/copper-doped BBG scaffolds after immersion in SBF for 14 days

Figure 109 shows the FTIR analysis of silver/copper-doped gelatin coated scaffolds after immersion SBF for 14 days. The spectra show the indicative gelatin peaks, as shown in Figure 85, with an amide A group between 3625cm-1 and 2750cm-1, water between 2500cm-1 and 2250cm-1 and an amide I, II, III between 1700cm-1 and 1100cm-1. (456) (457) An expanded view of the peaks attributed to the gelatin, from 3750cm-1 to 1200cm-1, in the spectra shown in Figure 109 can be found in appendix C as some of the peaks are difficult to see when observing the full spectrum to include the peaks attributed to the BBG.

206

The indicative peaks of a silver/copper-doped scaffold are also present between 1400cm-1 and 455cm-1, as seen Figure 69, the spectra also exhibits the typical changes that indicate that the sample has been immersed in SBF, as seen in Figure 74. The peaks indicative of bioactivity in BBG scaffolds are present in the form of the C-O bond between 880cm-1 and 900cm-1, (425) the P-O bond at 875cm-1 to 825cm-1 increasing in size (423) (424) (222) and the dual P-O bond at 760cm-1 to 650cm-1 merging. (424) The peaks associated with a silver/copper-doped sample after SBF immersion include a reduced

- -1 -1 NO 3 bond at 1387cm , a shifted reduced Si-O bond between 950-1000cm (428) (429) (430) (444) (445) (446) and the second shifted Si-O peak detected at 670-680cm-1 which has been reduced to such an extent that it is almost undetectable. This suggests that this bond breaks first when silver and/or copper diffuse out of the silicate structure when immersed in SBF.

The EDX analysis for the silver/copper-doped gelatin coated samples is shown in Figure 110. The elemental change in the composition is observed which is indicative of HA precipitation with the ratio between phosphorous and calcium changing to approximately 1.67, as seen in Figure 35. The presence of copper is also detected at 0.5keV and 8keV and the presence of silver at ~3keV which is consistent with the results shown in section 5.4, section 5.3 and the literature (447) This result supports

207 the premise that the addition of silver and copper and the coating of the scaffold with gelatin do not have a detrimental effect on the bioactivity of the sample.

Figure 110:- EDX analysis of polymer coated silver/copper-doped BBG scaffolds including A) Type A 5wt% gelatin coated scaffold, B) Type A 6wt% gelatin coated scaffold, C) Type B 5wt% gelatin coated scaffold and D) Type B 6wt% gelatin coated scaffold after immersion in SBF for 14 days

The compressive strength of the silver/copper-doped gelatin coated scaffold has decreased upon immersion in SBF for 14 days, as shown in Table 37. This result is in agreement with the previous results and the literature as the combination of the degradation of the gelatin coating, the ion release and HA precipitation has a detrimental effect on the mechanical competence of the scaffold. (352) (418) The compressive strength values are still higher than the corresponding value (0.43±0.05MPa) for the BBG scaffold after immersion in SBF for 14 days, indicating that although some of the copper and silver has diffused out of the structure and the (partial) degradation of the gelatin coating has occurred they are still having a positive effect on the mechanical competence of the scaffold. These compressional values are similar to those observed for the SBF immersed copper-doped gelatin coated scaffolds in Table 33.

Table 37:- Compressive strength of silver/copper-doped polymer coated BBG scaffolds after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Polymer type Compressive strength and wt% (MPa) Gelatin type A 5wt% 1.01 ± 0.04 * Gelatin type A 6wt% 1.00 ± 0.05 * Gelatin type B 5wt% 1.02 ± 0.03 * Gelatin type B 6wt% 1.00 ± 0.07 *

208

The white light interferometry results for RMS, Ra and PV after the silver/copper-doped gelatin coated scaffolds were immersed in SBF for 14 days are shown in Table 38. These values are lower than those observed for the BBG samples after the sample immersion period in SBF, shown in Table 11. This result is likely due to the presence of the polymer coating, which has started to degrade, and due to the precipitation of HA on the samples surface. Whether the change in topography has an effect on the scaffolds ability to positively interact with cells has not been explored in this work for this scaffold type but the change in relation to the other scaffolds produced in this work has been discussed at the end of the chapter.

Table 38:- White light interferometry results for the surface roughness of gelatin coated, silver/copper-doped BBG pellets after immersion in SBF for 14 days. * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

Polymer type RMS Ra PV and wt% (µm) (µm) (µm) Gelatin type A 5wt% 5.2±0.1 * 4.1±0.2 * 16±2 * Gelatin type A 6wt% 5.0±0.2 * 4.4±0.1 * 17±1 * Gelatin type B 5wt% 5.2±0.3 * 4.2±0.3 * 15±2 * Gelatin type B 6wt% 5.1±0.1 * 4.2±0.1 * 16±1 *

Wettability results are summarised in Figure 111, showing that the contact angles obtained were all substantially higher than those recorded for the BBG samples of 28±4o and 23±2o, before and after SBF immersion, respectively. In fact these wettability values were higher than those for the separate silver/copper-doped samples and the gelatin coated base scaffolds, indicating that the combination of MSIE with polymer coating increases the contact angle measured. The contact angles for the silver/copper-doped gelatin coated scaffolds were higher than the upper limit of the ideal range of contact angles suitable for cell attachment of between 48 and 65o. (419) There is little difference between the contact angles measured for the two gelatin types but the 6wt% samples were observed to increase the contact angle more that the 5wt% samples. As with the white light interferometry work for this type of scaffold, it can not be ascertained how the change in topography will affect the cells ability to interact with this type of scaffold, as discussed below.

209

Figure 111:- Wettability data for the gelatin coated silver/copper-doped BBG scaffolds showing spectra from before and after immersion in SBF for 14 days (Data shown are mean values and standard deviation of at least 3 measurements). * Indicates a significant change between the sample types (p≤0.05) using ANOVA analysis.

5.8 Summary of results from the structural characterisation

Overall this chapter investigated the structural properties of the BBG scaffolds along with scaffolds modified by MSIE and polymer coating. The BBG scaffolds were first characterised to ensure that the scaffolds fabricated in this work were consistent with the literature and to provide a control sample for the modified samples. The results obtained for the base scaffolds were consistent with the results obtained in the literature including the porosity of 93±2%, (4) (410) (241) compressive strength of 0.53±0.08MPa (4) (352) (418) and the EDX, (5) (394) (417) (420) (421) (422) XRD (4)(5) (389) (416) (426) (427) and FTIR spectra, (413) (414) (415) (423) (424) (222) (425) as well as contact angle results, (222) (223) (419) white light interferometry data (222) (223) (225) (224) and SEM observations, (4) (5) (222) (223) (410) (411) all matching previous results. The presence of the bioactivity indicator, HA formed on the scaffold surfaces upon immersion in SBF for 14 days, was also confirmed.

The MSIE process used to introduce silver and copper into the structure of the BBG scaffold was found to be successful based on the results presented in this chapter. This was demonstrated by the EDX, XRD and FTIR results all showing changes in the structure that are a direct result of the introduction of silver or copper or both ions. (6) (113) (428) (429) (444) (445) (447) It should be noted, however, that it can not be 100% proved with the techniques used in this work that all of the detected silver and/or copper was incorporated in the structure of the scaffold. It is possible that some salt residue

210 remaining from the MSIE process not removed by the washing protocol could contribute to the measured ions. Another possibility is that salt residue could have infiltrated the micro-cracks present on the scaffold surface. Further work is required to prove the exact concentration of the desired ions which has entered the silicate network. These characterisation techniques also showed that after the MSIE samples were immersed in SBF the addition of copper and/or silver ions did not have a detrimental effect on the bioactivity of the sample as the presence of HA was detected throughout. (6) (113) (428) (429) (444) (445) (447)

The porosity of the MSIE samples was lower, by as much as 20%, than the base scaffolds but still higher than the ideal threshold porosity described in the literature. Both SEM and white light interferometry results have shown that the MSIE process changes the surface topography of the scaffold, increasing the surface roughness of the scaffold struts. This result is supported by the increase in the contact angles measured for the copper-doped and silver/copper-doped samples however these were all still lower than the ideal contact angle value suggested for improved cell attachment. (419) (225) (224)

The MSIE process increased the compressive strength of the scaffold which is likely due to the stuffing effect (335) (436) (437)(397) and due to the addition of the silver and/or copper ion in the silicate matrix by shortening the non-bridging oxygen bond and forming bond like links to other oxygen’s in the matrix. (110) (418)

The depth EDX and ICP-MS results showed that the release rate of silver ions was a gradual process compared to a burst like release of the copper ions out of the scaffold. The silver ions penetrated deeper into the silicate matrix but at a lower concentration than the copper ions. For the silver/copper- doped samples copper was favoured in the MSIE process, in that a greater concentration of copper ion was detected compared to that of silver. However the concentration of copper detected was lower, because of the presence of the silver ions, compared to the copper only doped scaffolds. This is significant as it indicates that the presence of silver in the silver/copper doped samples slows the release rate of copper.

There was an observed change in the surface roughness of the scaffolds which can be attributed to the surface modification due to the MSIE process. For example, the silver-doped scaffolds showed a small reduction in their surface roughness and both the copper and silver/copper-doped scaffolds exhibited a marked increase in their surface roughness values. As stated in the literature review (section 2.9.3) the surface topography has an effect on the scaffolds ability to interact positively with the host tissue as the surface of the biomaterial plays a critical role in the interface interaction between the cells and the scaffold. (220) In the literature, a RMS value of between 0.8µm and 1.9µm is reported to provide a balance between intermediatary cellular responses such as proliferation, protein content and ALP activity and final cellular activity such as bone formation without effecting cell attachment. (461) The

211 silver-doped scaffolds had a measured roughness which was close to this ideal surface roughness, however both the copper and silver/copper-doped samples had a surface roughness significantly higher than this value. Whether this increase will lead to a significant difference in the cell biology results shown in chapter six will be discussed in the relevant section. Certainly, considering that the literature shows that the initial cell response can be influenced by the surface roughness this aspect that needs to be investigates in terms of how the combination of sintering and the MSIE process changes the surface roughness and its effects on cell behaviour. (462)

Taking all results into account the MSIE scaffolds that were taken forward to further optimisation and considered for cell biology work were the lowest concentration 15 minute immersion period silver-doped scaffold, the lowest concentration 15 minute immersion period copper-doped scaffold and the 15 minute immersion period silver/copper-doped scaffold. These scaffolds exhibited the balance between having properties close to those of the base scaffolds, and incorporating therapeutic metal ions into the structure without having a detrimental effect the basic characteristics required for BTE

Coating the BBG scaffolds in gelatin was done to further improve some of the properties of the BBG scaffolds. During the main study the two types of gelatin were used at two concentrations and PDLLA coated BBG were also investigated for comparison. It was observed that the coating procedure had no effect on the chemical composition or crystal structure of the scaffold as observed with EDX and XRD. The FTIR analysis showed characteristic peaks corresponding to both the base scaffold chemical structure, and of the gelatin and PDLLA coating.

The porosity was marginally reduced by the presence of the polymer coating; moreover there was little difference between the polymer types or the different types of gelatin. The addition of a polymer coating did have an effect on the topography of the scaffold, as observed in the SEM images, in that the surface of the struts was smoother and the polymer formed a continuous homogenous layer on the struts surface. This observation was supported by the results from the white light interferometry and wettability analysis. These results indicated reduced RMS, Ra and PV values and an increase in the contact angle which confirmed that the surface of the coated scaffold was smoother than that of the BBG scaffolds potentially making the coated scaffolds more suitable for cell attachment. (419) (225) (224) There was little difference between the two polymer types or the different types or concentration of gelatin in terms of surface roughness.

The other area in which the polymer coating had an integral effect on the base scaffolds was related to mechanical properties, in particular the increase in the compressive strength. The addition of a polymer coating led to a 2x time’s increase of compressional strength compared to the BBG scaffolds. The gelatin coatings improved the compressive strength of the BBG scaffolds marginally higher than the PDLLA coating.

212

In addition, it was observed throughout the study that the inclusion of a polymer coating, whether it be gelatin or PDLLA, did not have a detrimental effect on the BBG scaffolds ability to form a HA layer on the surface after immersion in SBF.

Taking all results into account it can be stated that there is no real advantage between using one type of gelatin over the other and further assessment of the gelatin coated scaffolds of both types is required including cell biology testing. The results also indicate that gelatin improved the BBG scaffolds to a higher extent than PDLLA without impairing the fundamental properties that make the BBG scaffolds particularly attractive for BTE.

Combining ion-exchanged scaffolds with gelatin coating was successful and was found to improve the BBG scaffold. Observations made using EDX, FTIR and XRD showed that MSIE and polymer coating have not led to changes in the bioactive behaviour of the BBG scaffold. After immersion in SBF for 14 days the samples showed precipitation of HA.

The topography of the gelatin coated MSIE scaffolds assessed through SEM observations, white light interferometry and wettability contact angles indicated that the modified scaffolds were smoother than the BBG scaffolds, which confirmed that the polymer coating had reduced the roughness caused by the MSIE process. The wettability improved substantially and the contact angle values were in the range considered suitable for improved cell attachment, indicating that the combination of MSIE and polymer coating changed the strut surface topography in a positive manner. In addition, the compressional strength of the gelatin coated MSIE samples improved compared to the BBG scaffolds.

Summarising, it can be stated that the combination of MSIE and polymer coating has introduced improvements on the BBG scaffolds without compromising the typical characteristics that make these scaffolds attractive for BTE specially the highly interconnective pore structure and bioactivity.

213

Chapter Six

6. Biological Results and Discussion

This chapter covers the cell biology and bacterial studies carried out on the MSIE scaffolds and using the BBG scaffold as a control and to confirm previous biological tests carried out using this material. (373) (463) (464) Unless otherwise stated the concentrations of the molten salt baths for MSIE used in this chapter were the lowest and the immersion period in the molten salt was 15 minutes. (6) The decision to use these parameters was based on previous related work carried out in the literature and the results shown in the previous chapter. (6)

6.1 General remarks

This part of the work was carried out in collaboration with the Leeds Dental Institute, UK, for the in vitro studies using HPLSC’s, the Department of Plastic and Hand surgery at the University of Erlangen Medical Centre, Germany, for the in vitro studies using rMSC’s and hMSC’s. The Institute of Bioprocess engineering at the Friedrich-Alexander University of Erlangen-Nuremberg, Germany collaborated with the bacterial work carried out using E.coli and the Institute of Hygiene and Biotechnology at Hohenstein Institute in Bönnigheim, Germany collaborated with the bacterial work carried out using S. aureus and K. pneumoniae. All the studies used the BBG scaffold as the control sample to give a comparison.

6.2 In vitro cell biology using human periodontal ligamental stromal cells

This set of studies used HPLSC’s and was carried out in collaboration with Dr R. El-Gendy and Prof X.B Yang at the Leeds Dental institute at the University of Leeds, UK. The main purpose of this set of studies was to focus on the silver MSIE scaffolds to investigate the possible toxicity and biocompatibility of the samples and to assess the concentration of molten salt that can give rise to a scaffold that is capable of being implanted without causing cytotoxicity. The cells used in this section were prepared as described in section 4.4.1 and were seeded onto the BBG and silver-doped scaffolds as shown in section 4.4.2. The scaffolds were then stained using a live/dead marker, as described in section 4.4.3, to show the biocompatibility of the samples with the cells. This study used BBG scaffolds

214 as the control samples and three different concentrations of silver MSIE scaffold with an immersion period of 15 minutes in line with previous work. (6)

6.2.1 Cell viability assessment using live/dead markers

As described in section 4.4.3 the scaffolds were stained to show the ratio between live and dead cells on the scaffolds after three days of dynamic seeding and seven days of static cultivation. Figure 112 shows two different BBG scaffolds after being subjected to the two cell stains. Both Figure 112A and Figure 112B show no presence of cells stained red (the stain to show dead cells), but show a vast number of green stained cells attached to the scaffold (the stain to show living cells). This indicates qualitatively that the BBG scaffold did not have detrimental effects on cell attachment and it caused no toxic effect, which is in agreement with previous work undertaken by other researchers on similar scaffolds. (113) (465) (368) The fact that no attached dead cells were detected does not conclusively indicate that the BBG scaffolds did not cause cell death during the dynamic seeding process as the dead cells could have been washed away before they attached. However based on the high number of green dyed living cells, the results give a strong indication that the BBG scaffolds were not toxic to the cells.

Figure 112:- HPLSC’s attachment and viability on the surface of 3D BBG scaffolds with A) and b) two different BBG scaffolds. Viable cells are stained green and dead cells are stained red after a total of 10 days of cell culture

Figure 113 shows low concentration (as described in section 4.2.2) scaffolds that have been immersed in the molten salt bath of NaNO3 and AgNO3 for 15, 30, 45 and 60 minutes after being subjected to the cell stains and viewed under a CLSM. All scaffolds shown in Figure 113 exhibit attached stained living cells (stained green); however the number of living cells has decreased as the

215 immersion period in the molten salt bath increased. Figure 113 also shows the presence of no red dyed cells, as observed for the samples shown in Figure 112. Figure 113A shows a quantity of living (green) cells comparable with those shown in Figure 112 indicating that at this MSIE concentration and immersion period there was no observed detrimental effect of the silver ions on cell attachment. The reduction in the number of attached living cells in the samples with increased immersion periods (observed in Figure 113B, C and D) could be due to dead cells being washed away before they had a chance to attach.

Figure 113:- HPLSC's attachment and viability on the surface of low concentration silver MSIE 3D BBG scaffolds with a ranged of immersion periods in molten salt: A) immersion period of 15 minutes, B) immersion period of 30 minutes, C) immersion period of 45 minutes and D) immersion period of 60 minute. Viable cells are stained green and dead cells are stained red after a total of 10 days of cell culture

Figure 114 and Figure 115 show the stained cells viewed under a CLSM for the medium and high concentrations of silver-doped scaffolds respectively. Both figures show the presence of green living cells and no red dead cells, just as in Figure 112 and Figure 113. However there is a marked reduction in the number of living cells observed. The number of living cells is decreasing as the immersion period in the molten salt bath increases and this was also observed for the low concentration scaffolds shown in Figure 113. The number of observed living cells was also seen to decrease as the concentration of the silver in the salt bath increased.

216

Figure 114:- HPLSC's attachment and viability on the surface of medium concentration silver MSIE 3D BBG scaffolds with a ranged of immersion periods in molten salt: A) immersion period of 15 minutes, B) immersion period of 30 minutes, C) immersion period of 45 minutes and D) immersion period of 60 minutes. Viable cells are stained green and dead cells are stained red after a total of 10 days of cell culture

Figure 115:- HPLSC's attachment and viability on the surface of high concentration silver MSIE 3D BBG scaffolds with a ranged of immersion periods in molten salt: A) immersion period of 15 minutes, B) immersion period of 30 minutes, C) immersion period of 45 minutes and D) immersion period of 60 minutes. Viable cells are stained green and dead cells are stained red after a total of 10 days of cell culture

217

These live/dead markers showed that there is no observed cytotoxic effect caused by the samples once the cells have attached, especially in the lowest concentration of silver-doped scaffolds as the attached cell density was the highest. However it cannot be stated that samples were not causing some toxic effect as most of the cell death would have occurred during the dynamic seeding stage and the dead cells would have been removed by the flow of the media and not observed. This set of results showed that as the concentration of the silver-doped scaffold increased the observed number of living cells decreased. This is useful information for the design of future cell biology and bacterial studies. Only the scaffolds with the lowest concentration of silver will be considered as higher concentrations of silver will have a negative effect on cell attachment. This result supports work that has been carried out previously. (113)

The results do not show whether the surface roughness, ascertained in chapter five, is having an effect on the cells behaviour. The only difference between the cells shown in Figure 112 for the BBG scaffolds and those shown in Figure 113, Figure 114 and Figure 115 for the silver-doped scaffold is a slight decrease in the overall number of cells, which however needs to be supported by other tests to prove conclusive. As this method has a dynamic seeding stage, any cells that might have died will have been taken away meaning that healthy robust cells are left to attach and proliferate. A more detailed discussion on the surface topography in relation to these results is presented at the end of the chapter.

It should be noted that the sterilisation method used for this section of cell biology work will likely have an effect on the chemical and physical properties of the surface of the samples. This is unfortunately unavoidable as any contamination needs to be removed from the scaffolds, which may be due to the fabrication process, MSIE process and general handling of the samples. Little work has been carried out on understanding how the sterilisation process influences the outcome of cell biology studies in this type of scaffolds as it is an important and unavoidable part of the process. This practical aspect needs to be looked at in the future as it could potentially change the interpretation of some studies and this will be discussed in chapter 8.

6.3 In vitro cell biology using rat bone-marrow stromal cells and human bone- marrow stromal cells

The set of studies used rMSC’s and hMSC’s and was carried out in collaboration with Dr L. Strobel from Dr U. Kneser’s research group at the Department of Plastic and Hand Surgery at the University of Erlangen Medical Centre, Germany. The main purpose of this set of studies was to establish the toxicity and biocompatibility of the MSIE scaffolds of all sample types (scaffolds doped with silver, copper and with both silver and copper), as described in section 4.2.2. From the previous

218 qualitative observations in section 6.2 it has been established that the lowest silver concentration sample type at the shortest immersion period would likely yield the most positive reaction from the cells. The cells used in this section were prepared as described in section 4.4.4 and were seeded onto the plain and ion-doped scaffolds as shown in sections 4.4.5 and 4.4.6. The cells were then tested for their viability using an Alamar blue (AB) assay as described in section 4.4.7.

6.3.1 Preliminary toxicity and biocompatibility study of low concentration molten salt ion exchange scaffolds using rat bone-marrow stromal cells

The scaffolds were prepared and seeded for the 2D culture, as described in section 4.4.5. The scaffolds were checked every 24 hours to have the CCM replaced and to have the scaffolds and cells washed gently in PBS for light microscope observations, as described in 4.4.5. This was repeated for three days before confirming that the number of viable cells observed, shown in Figure 116 and Figure 117, were decreasing and the cells were not of a regular shape compared to the cells in the control wells.

Figure 116 shows the light microscopy results after one day of the rMSC’s being exposed to the samples and there is a marked reduction in the number of healthy attached cells shown in the control sample (Figure 116A) compared with not only the cells exposed to the MSIE samples but also with the cells in the BBG scaffolds (Figure 116 B-E). There were also a number of dead cells found, these are the circular orange cells shown in Figure 116, and this behaviour was observed in all the scaffold samples. This result suggests an intrinsic effect of the scaffold which is not due to the influence of the additional ions added. The results shown in Figure 116 do not show whether the cell death is due to a change in the surface roughness affected by the MSIE processing because a number of cells seeded onto the BBG scaffold also suffered from programmed cell death. This hypothesis was further strengthened by the results shown in Figure 117, which indicate that after three days the number of cells in the control sample (Figure 117A) has remained constant and the cells look healthy having started to spread and multiply. However the few remaining cells found on the other samples look irregular and a number of dead cells was also observed.

It is clear that just by observation it can not be concluded that the dead and irregularly shaped cells that are shown in Figure 116 and Figure 117 are part of the normal process of programmed cell death (apoptosis). In order to reach such conclusion a suitable assay would have to be conducted, for example a TUNEL assay which detects DNA fragmentation due to apoptosis. (404) This will be discussed further in the summary of this chapter.

219

Figure 116:- 2D seeded rMSC’s on: A) control well with no sample, B) BBG scaffold, C) low concentration silver- doped BBG scaffold, D) low concentration copper-doped BBG scaffold and E) low concentration combination ion BBG scaffold after one day of contact with the cells

Figure 117:- 2D seeded rMSC’s on: A) control well with no sample, B) BBG scaffold, C) low concentration silver- doped BBG scaffold, D) low concentration copper-doped BBG scaffold and E) low concentration combination ion BBG scaffold after three day of contact with the cells

220

It was observed, despite the pre-seeding washing, that when the scaffolds were immersed in the CCM a colour change from its characteristic orange to a bright pink occurred, because of the presence of phenol red which indicates that there is a pH change towards higher alkali pH’s, as shown in Figure 118.

Figure 118:-Image illustrating an increase in the pH which leads to a change in the colouration of the CCM from A) the typical orange colouration to B) the bright pink colouration which indicates the activation of the phenol red indicator in the media

Indeed, this pH change of the medium is a known problem with cell biology experiments carried out directly on 45S5 Bioglass® samples however there is only limited information in the literature. (408) (182) It is worthwhile noting in this regard, that each laboratory develops its own protocol for the pre-treatment of BBG scaffolds and there is no general protocol established to reduce the pH variation when BBG scaffolds are exposed to the cells. This pH issue was identified in this study and it was addressed before any further in vitro work was carried out so that results that are obtained are relevant and reflect only the effect caused by the additional metal ions added and are not masked by the effects of the chemistry of the base material.

It should be noted that the autoclaving method of sterilisation used for this section of cell biology work has the potential to affect the physical and chemical properties of the surface of the samples. This is unavoidable as any contamination needs to be removed from that scaffolds due to the fabrication process, MSIE process and general handling of the scaffolds. As stated before, this aspect of the cell biology work has not been extensively investigated and it needs to be looked at as it could have an effect on the outcome of cell biology experiments. The need for future studies addressing the effect of sterilisation on the outcome of cell biology studies is discussed in chapter 8.

221

6.3.2 The effect of pH change

The results from the preliminary in vitro work, shown in the previous section, indicated the pH changes of the CCM after the scaffolds have been added (shown in Figure 116 and Figure 117). The pH changes during the washing and sterilization process were observed and then necessary changes to the protocol were introduced. Table 39 shows the pH changes during the washing protocol that was originally in place to prepare the scaffold samples for in vitro experiments. This table shows that the pH of the solutions increased when the BBG scaffold was washed, with an increase of 3.97 from the initial pH of the first washing stage before the sample was added to the end of the washing protocol and an increase of 0.305 from after the first washing stage to the end of the washing protocol. The final pH of 11.71 is far too high for cell culture work. The first three stages of the washing protocol cannot be eliminated as these steps are required to remove any residue salt on the surface of the scaffold after the MSIE process; however the latter stages can be changed for a different washing process as these latter steps contribute to the pH increase.

Table 39:- pH changes throughout the washing procedure

pH Washing solution Before After Difference De-ionised Water 7.74 11.405 3.665 70% Ethanol 11.33 12.15 0.82 De-ionised Water 7.74 11.01 3.27 2.5% SDS 8.18 10.725 2.545 De-ionised Water 7.74 10.62 2.88 5% Extran 12.44 12.465 0.025 De-ionised Water 7.74 11.71 3.97 Overall difference 0.305

After the washing process the scaffolds were then autoclaved in the same conditions as previously stated in sections 4.4.5 and 4.4.6 depending on the experiment being carried out. This process is required to sterilise the scaffolds in preparation for the in vitro work. The final stage of the washing and sterilisation protocol implemented was a final hour of washing in the selected CCM, samples were then patted dry and this was carried out in the sterilised fume hood. The samples were turning the media pink after one hour of immersion but the pH of the solution was reduced from pH 11.71 after the last stage of washing before the autoclaving to pH 11.34.

This result indicated a suitable conditioning of the scaffolds by increasing the post autoclaving immersion period from one hour up to four days. On the first day the samples would be on a shaker to get the CCM infiltrating all of the pores, approaching a dynamic seeding flow through the scaffold, and

222

o the final three days would be in an incubator at 37 C at 5% CO2. This stage needed to be long enough to reduce the pH but obviously not so long that would lead to the metal ions added through MSIE being completely lost. Also the immersion media was changed to determine if this would have an effect in lowering the pH. Two solutions were investigated: DMEM/F-12 (without FBS, L-glutamine or antibiotics) and a 0.9% NaCl solution. After this stage, it was noted that the DMEM/F-12 media had once again turned bright pink but a pH of 8.95 was measured and the 0.9%NaCl had reduced the pH to pH 10.38. Finally, to wash the scaffolds for a final time the medium was replaced with fresh medium and placed on the shaker, this gave way to a further pH reduction to pH 8.23 and pH 9.95 for the DMEM/F-12 and the 0.9% NaCl solution, respectively.

The samples were then used in 2D culture experiments, as described in section 4.4.5, changing the CCM, this was changed back to the standard medium that was used for all cell cultures experiments, every 24 hours and observing the cells under the light microscope. Table 40 and Figure 119 describe observations over the two days that the 2D culture experiment was running. The results show that the pre-conditioning of the scaffolds in DMEM/F-12 had less of a negative effect on the cells compared to the scaffolds that were pre-conditioned in 0.9% NaCl. It also unveiled that the presence of the well insert had a negative effect on the cells (Figure 119A and Figure 119D) compared to the cells that were not restricted by the well insert (Figure 116A). This effect of using the well insert was not investigated further in this work and remains a topic for future studies.

Table 40:- Observations made after introducing a longer pre-conditioning phase based on section 4.4.5

Condition of cells after seeding the cells Sample type Day 1 Day 2 Control well with insert some living cells some living cells Control well with no insert numerous living cells numerous living cells Control well with no insert numerous living cells numerous living cells Pre-conditioned in DMEM/F-12 a few living cells some living cells Pre-conditioned in DMEM/F-12 no living cells no living cells Pre-conditioned in DMEM/F-12 no living cells no living cells Pre-conditioned in DMEM/F-12 a few living cells a few living cells Pre-conditioned in 0.9% NaCl a few living cells no living cells Pre-conditioned in 0.9% NaCl no living cells no living cells Pre-conditioned in 0.9% NaCl no living cells no living cells Pre-conditioned in 0.9% NaCl no living cells no living cells Control well with insert and 1% buffer all numerous living cells numerous living cells Control well with insert and 2% buffer all numerous living cells numerous living cells

223

Figure 119:- rMSC’s seeded onto pre-conditioned scaffolds: A) control well with insert one day after seeding, B) DMEM/F-12 pre-conditioned scaffold one day after seeding, C) 0.9% NaCl pre-conditioned scaffold one day after seeding, D) control well with insert two days after seeding, E) DMEM/F-12 pre-conditioned scaffold two days after seeding and F) 0.9% NaCl pre-conditioned scaffold two days after seeding

An interesting observation was that the two DMEM/F-12 pre-conditioned scaffolds that produced the least negative impact on the cells were at the lower end of the size range of the scaffolds used for the study and this could have affected the number of living cells found. In the preliminary set of in vitro work using rMSC’s the size of the samples ranged from 5x5x5mm3 down to 3x3x3mm3. It is thus likely that the larger size will have contributed to the pH increase issue. This set of experiments also investigated the use of a buffer solution to help to reduce the pH and control the pH after the cells have been seeded. For this process two solutions of Bufferall were used at 1% and 2% solutions. Table 40 shows that the introduction of the solution did not have a detrimental effect on the seeded cells and kept the media at a suitable pH to prevent cell death.

It should be noted that in the studies using HPLSC’s (section 6.2) no pH issues were recorded. This was because of the dynamic seeding system used which kept a constant flow of the CCM flowing over the seeded scaffolds preventing the local pH increasing significantly as it moved the released ions away from the cells. This effect should be considered for future work, certainly dynamic systems should be preferred over static culture tests as they represent a condition which is closer to the in vivo situation.

As with the results shown in Figure 116 and Figure 117, it is important to note that from simple observation it can not be said that any observed cell death is caused by the change in the scaffold type as natural cell death due to apoptosis can occur. (404) In order to prove that the cell death observed is

224 due to experimental conditions and not apoptosis a suitable assay must be used, such as a TUNEL assay. (404) This will be discussed further in the summary of this chapter.

6.3.3 Biocompatibility and toxicity study using low concentration molten salt ion exchange scaffolds

After implementing the new pre-conditioning results explained in section 6.3.2 the samples were subjected to the protocols shown in sections 4.4.5 and 4.4.6, in particular Figure 26 and Figure 27. These samples were then processed using the AB assay, described in section 4.4.7, to give an assessment of the samples biocompatibility. Figure 120 shows the AB assay results for the 2D culture experiments. The results show that the BBG scaffolds had an AB reduction comparable with the control group and for the first week this was the case for the silver-doped samples. The copper-doped and the combination ion-doped samples had no living cells observed after a week and the silver-doped samples had no living cells present after the two week mark.

Figure 120:- 2D rMSC cultures with inserts containing A) control well with no insert and samples, B) wells containing BBG scaffolds, C) wells containing low concentration silver-doped BBG scaffolds, D) wells containing low concentration copper-doped BBG scaffolds and E) wells containing low concentration combination ion- doped BBG scaffolds and then subjected to an AB assay. For copper-doped BBG scaffolds there was no detected metabolic activity after 14 days and for silver/copper-doped BBG scaffolds no metabolic activity could be detected after 7 days. * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis.

225

The results shown in Figure 120 suggest that the metal ions are being released in a fast burst during the first week and this burst release, including sodium and calcium from the BBG scaffold as well as the additional therapeutic ions, could be potentially the reason behind the decrease in the number of observed living cells found in the wells after one week.

This burst release action is well documented for copper ions and this was one of the factors for a low concentration being used in the in vitro work as it was hypothesized that this would reduce the effect of this phenomenon. However from Figure 120 it can be seen that the concentration of copper ions in both the copper-doped and combination ion-doped samples is still high enough to have a significant detrimental effect on the cells.

The results of the AB assay on 3D culture samples are shown in Figure 121, these results are similar to those found in Figure 120. The cells that were seeded onto the BBG scaffold survived through to the end of the two weeks with a slight dip in the number of observed living cells at the end of week one. Regarding the cells seeded onto the MSIE samples, it was observed that only a small number of cells survived the initial seeding process and each sample type showed a low number of living cells after one day. After that there were no living cells detected by the AB assay after one week and two weeks. This result also indicates the effect of the metal ions on the cells which is leading to cell death.

Figure 121:- 3D rMSC cultures containing A) control well with no samples, B) wells containing BBG scaffolds, C) wells containing low concentration silver-doped BBG scaffolds, D) wells containing low concentration copper- doped BBG scaffolds and E) wells containing low concentration combination ion-doped BBG scaffolds and then subjected to an AB assay. * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis.

226

The change in surface roughness could also have an effect on the cell death observed for the modified scaffolds but no independent results are available to prove that in this case. This effect will be discussed in further detail at the end of the chapter. Apoptosis could also be an explanation to the cause of the observed cell death shown both by the images and the AB assays. (404) This can be established by using the appropriate assay to show that any cell death observed is caused by the change in the sample types and not by natural occurrence. (404)

The results for the silver doped scaffolds, shown in Figure 121, are not in agreement with the result shown in section 6.2 using the HPLSC’s, in that the rMSC’s did not survive past the initial seeding process. This result leads to a number of possibilities as to why the same sample parameters led to different results. The main reason for this behaviour is thought to be related to the difference in the seeding processes, with one being dynamic and the second being static. Also the type of cell used could have had an effect on the results with the rMSC’s being more susceptible to the burst release of metal ions than the HPLSC’s. Future work should investigate in detail this behaviour.

These results indicate that two options during the next stage of in vitro work can be explored:-

I). The concentration of the additional metal ions needs to be reduced II). Biodegradable coatings need to be added to the samples to slow down the release of the metal ions into the CCM.

6.3.4 Biocompatibility and toxicity study using extremely low concentration molten salt ion exchange scaffolds

Based on the preliminary results, presented in section 6.3.3, the concentration of the metal ions in the copper and combination ion-doped scaffolds was reduced to a tenth of the low concentration samples. hMSC’s were used for this part of the study to investigate cells behaviour when exposed to the MSIE samples compared with the previous work with HPLSC’s and rMSC’s. The silver-doped samples were not investigated with a reduced concentration as there was sufficient data from the previous studies (sections 6.2 and 6.3.3) to prove that the original low concentration with an immersion period of 15 minutes produced samples that were sufficiently biocompatible to be used in an in vivo set up.

The other change to the experimental method was an increase in the pre-conditioning time that was introduced in section 6.3.2, namely from four days in DMEM/F-12 with no additives to five days in DMEM/F-12 with 1% penicillin/streptomycin and the media being changed every other day.

227

The cells were prepared as described in section 4.4.4 and the samples were prepared as described in sections 4.2.2 and preconditioned as described above and in section 6.3.2. This study used a 2D culture set up as described in section 4.4.5 over a ten day period.

Figure 122A-C shows that after the first day all samples exhibited living cells left in the well. It was also observed that qualitatively the BBG scaffolds (Figure 122A) out-performed the two doped samples (Figure 122B-C). This result could indicate that the pre-conditioning has effectively reduced the pH change caused by the samples as intended. This effect can be confirmed if the BBG samples continue to show no observed cell death after five and ten days. The copper and combination ion-doped samples showed significantly fewer living cells and also showed dead cells present confirmed by the circular pale white cells, Figure 122B-C. It was also observed that there is little difference between the two MSIE samples investigated.

Figure 122:- 2D culture seeded with hMSC’s onto: A) BBG scaffold after one day, B) extremely low concentration copper-doped sample after one day, C) extremely low concentration combination ion-doped sample after one day, D) BBG scaffold after five days, E) extremely low concentration copper-doped sample after five days, F) extremely low concentration combination ion-doped sample after five days, G) BBG scaffold after ten days, H) extremely low concentration copper-doped sample after ten days and I) extremely low concentration combination ion-doped sample after ten days

228

Figure 122D-F shows qualitatively that after five days all three sample types exhibited the presence of living cells and the BBG scaffolds (Figure 122D) had a higher number of living cells, assessed by the visual inspection of the samples, compared to the other two sample types (Figure 122E- F). One observation made from these results is that the number of living cells found in Figure 122D has increased from Figure 122A, indicating that the cells are in suitable conditions to induce cell proliferation. The results show that the pH increase issue which is commonly found in cell cultures on Bioglass® (408) (182) and was discussed in previous sections, has been partially solved. A second observation is that the combination ion-doped samples whilst still showing significantly reduced number of living cells compared to the BBG scaffolds, exhibit a slight increase in the number of living cells ( at least qualitatively assessed) compared to the results after the first day and also this sample shows more living cells compared to the copper-doped samples.

Figure 122G-I shows the samples after ten days of exposure and confirms that the pH increase issue discussed above has been partially resolved for the BBG scaffolds as Figure 122G indicates an increase in the number of living cells present compared to the results shown in Figure 122D and Figure 122A. This result shows that the pre-conditioning and treatment of the BBG scaffolds is suitable to produce samples that will encourage desirable cell attachment and proliferation. Figure 122H and Figure 122I show very few living cells present and it was not possible to determine which of the ion- doped samples performed better in this format. The difference between the two ion-doped samples can be seen when the samples are subjected to an AB assay, which is shown in Figure 123.

Figure 123:- 2D hMSC cultures with inserts containing A) BBG scaffolds, B) extremely low concentration copper- doped samples and C) extremely low concentration combination ion-doped samples and then subjected to an AB assay after one, five and ten days of being seeded with hMSC’s. For copper-doped and silver/copper-doped BBG scaffolds no metabolic activity could be detected after ten days. * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis.

229

The AB assay results, conducted in accordance with section 4.4.7, shown in Figure 123 supports the observations from Figure 122, with all sample types showing some living cells at the first two time points and with the BBG samples outperforming the two ion-doped samples, especially after five days of the cells being exposed to the samples. This results supports the previous statement that the pre- conditioning protocol implemented on the samples has tackled the pH increase issue, discussed previously (caused by the release of the ions as the scaffold starts to degrade) thus allowing the cells to attach and proliferate. This result also shows that there is no statistically significant difference in living cell numbers between the two ion-doped samples over the ten days. Although the combination ion- doped scaffold performed slightly worse than the copper-doped sample, this sample had a higher copper content than the copper only doped sample. It was also confirmed that throughout these in vitro experiments using the hMSC’s the copper containing samples caused more cell death than the BBG samples and the silver-doped samples. It can be therefore concluded that copper ions are showing a cytotoxic effect under the investigated experimental conditions. As the CCM used in this in vitro work contains a small amount of copper in it, the additional copper released by the doped scaffolds seems to cause cell death through over exposure. This result needs to be investigated further. The tests should be repeated using an alternative CCM containing less copper to give an indication of what threshold concentration of copper released by the scaffolds will cause cell death.

Again the change in the surface roughness recorded in chapter five could have attributed to the increase in the cell death and the lack of cell attachment observed in this cell biology work. The fact is that the cell death can not be attributed to any one factor from these results. This will be discussed in detail at the end of the chapter.

6.4 Antibacterial assessment of molten salt ion exchange scaffolds

This set of studies used E.coli and was carried out in collaboration with Mrs A. Amtmann at the Institute of Bioprocess Engineering at the Freidrich-Alexander University of Erlangen-Nuremberg, Germany. The main purpose of this set of studies was to establish the bacteriostatic and bactericidal capabilities of the MSIE scaffolds of all sample types (silver, copper and silver/copper) using the tests described in sections 4.4.8 to 4.4.12.I. To give a direct comparison to the in vitro work shown in sections 6.2 and 6.3 the lowest concentration samples from each of the MSIE obtained at the shortest immersion period were tested. The work conducted using the S.aureus and K.pneumoniae was carried out in collaboration with Dr T Hammer from the Institute of Hygiene and Biotechnology at the Hohenstein Institute in Boennigheim, Germany. The purpose of this set of tests was to give not only a direct comparison with the in vitro work shown in previous sections but also to compare how the MSIE samples behave in contact with different types of bacteria. This work is described in sections 4.4.12.II

230 and 4.4.13. Again samples involving the lowest concentration of silver, copper and silver/copper immersed for the shortest time in the molten salt bath were used. All studies used the BBG scaffolds as the control sample.

6.4.1 Growth curves of Escherichia coli

For comparison purposes with the in vitro work, the studies were carried out using the same medium that was used previously (section 6.3). In addition, for comparison purposes an industry standard medium used in bacterial studies was considered. This was done to have a direct comparison between the cell culture studies and the bacterial work carried out in this section, considering that the tests will evaluate the antibacterial effectiveness of the MSIE samples with respect to their possible cytotoxic effect. The two media are i) DMEM/F-12 with 1% L-glutamine, 10% Foetal bovine serum (FBS) and 1% Buffer all (DMEM/F-12), as used in section 6.3 and ii) Muller-Hinton broth (MHB). The first task was to establish how the E.coli behaves in DMEM/F-12 compared to how it grows in MHB, as described in section 4.4.8.

Figure 124 shows the absorbance due to the presence of E.coli in a variety of different media over a wavelength range of 350nm to 800nm. To produce growth curves the absorbance is measured at a wavelength that shows an absorbance value of approximately 0.1 and that it will not change due to indicators within the media. The significance of the absorbance being 0.1 is that it directly corresponds to the McFarland standard as this is used to adjust the turbidity of a bacterial suspension and gives an industry standard for comparison. (466) (467)

For MHB the absorbance is normally measured between 550nm and 700nm as the value is constant in this range, as shown in Figure 124. However Figure 124 shows that the DMEM/F-12 medium has a large absorbance peak between 475nm and 625nm and this is due to the presence of the indicator phenol red in the medium. The absorbance curve then flattens out which means that an absorbance of approximately of 0.1 can be measured for 630nm and 700nm without interference (as shown by the inserted graph with the highlighted wavelength band in Figure 124). Because of this, the experiments (reported in the next sections) used an absorbance value measured at 650nm for both media.

231

Figure 124:- Absorbance measure over a full visible light spectrum for E.coli grown in a variety of different media, with the yellow band showing the ideal wavelength range at which the absorbance should be measured

After establishing the wavelength at which all absorbance values for the E.coli studies would be considered, the behaviour of E.coli in DMEM/F-12 compared to the MHB was investigated. The results are shown in Figure 125 and Figure 126. Figure 125 shows the change in absorbance, measured at 650nm, over a period of 24 hours for both media types. This result indicates that when the E.coli are grown in DMEM/F-12 the bacteria take approximately twice as long to reach an absorbance value of 1.5 compared to the E.coli grown in the MHB, but once this value is reached the growth curves for both media are significantly similar. Figure 125B in particular shows that E.coli requires more time to acclimatise to the DMEM/F-12 compared to the MHB. This result means that in the rest of the studies on E.coli in DMEM/F-12 should be given double the time in the pre-culture stage in order to have comparable results with the bacteria in MHB.

Figure 126 shows the number of E.coli bacteria after 30, 180 and 300 minutes for both types of media after being plated onto standard I agar plates and incubated for 24 hours, as described in section 4.4.8. These results were obtained using the multiple dilutions technique described in section 4.4.8. It was found that the E.coli that were grown in DMEM/F-12 exhibited a higher number of bacteria plated than those grown in MHB after the first time point. The rate of growth shown by the E.coli in MHB exhibits continuous growth and this result supports the growth curve shown in Figure 124. The rate of growth over the five hour period for the E.coli in DMEM/F-12 is seen to increase continuously and this result also supports the data shown in Figure 124.

232

Figure 125:- Growth of E.coli in both MHB and DMEM/F-12 with A) showing the full incubation period and B) showing the growth curves for the first five hours; the absorbance was measured at 650nm.(The lines connecting the dots are only to help the eye). * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis.

Figure 126:- E.coli bacteria per ml measured using the CFU plating method taking samples at three time points, o plating onto standard I agar plates and incubated for 24 hours at 37 C and 5%CO2. (The lines connecting the dots are only to help the eye) * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis.

233

In summary, the results have shown that the wavelength at which the absorbance should be measured is 650nm for both types of media. It is also concluded that E.coli in DMEM/F-12 needed almost twice as long to start growing as this bacteria needed longer to acclimatise to the media, meaning that a longer pre-culture stage is required for future work.

6.4.2 Copper nitrate vs copper chloride salts in a growth inhibition study using Escherichia coli

As described in section 4.4.9 the growth inhibition of copper nitrate needs to be compared to that of copper chloride to determine if this metal salt can be used in the main metal salt inhibition study, considering that the study is based on the investigation described in Heidenau et al (409) in which copper chloride was used. The method used is described in section 4.4.9 with 10µl of E.coli suspension plated onto standard I agar plates after four and 24 hours of incubation and then incubated for a further 24 hours on the plates ready to have the colony forming units (CFU’s) counted and processed.

Figure 127 shows suspensions of E.coli in DMEM/F-12 (with the two different concentrations of E.coli) with both copper salts. It can be seen that, due to the phenol red in the DMEM/F-12, there is change in colour due to the presence of E.coli going from the normal orange colour (Figure 127Ai-Di) to a cloudy yellow of the control E.coli suspensions (Figure 127Aii-Dii). The image shows that as the concentration of copper salts increases, the turbidity of the suspension decreased and the colour changed towards the orange end of the scale, suggesting that the bacteria were affected by the presence of the copper salts. This is the case for both concentrations of E.coli and for both copper salts, meaning that on a qualitative level there is little difference between how the E.coli behaves in the presence of the two copper salts.

Samples from the suspensions, shown in Figure 127, were taken after four and 24 hours, plated

o and incubated for a further 24 hours at 37 C and 5% CO2 and the number of colony forming units was counted. The results are shown in Table 41. This study was performed for both media types to enable a comparison and to ensure that the next set of experiments would be possible using copper nitrate salt and DMEM/F-12.

Table 41 shows that there is very little difference between the E.coli suspensions containing copper chloride and those containing copper nitrate for the bacteria suspended in DMEM/F-12. The data shown in Table 41 has been collated from the figures shown in appendix C2 and shows that there is a difference in the bacteria’s behaviour in the two media. However since both copper salts behaved the same in the two media the result suggests that copper nitrate can be used to replace copper chloride and that DMEM/F-12 can be used to replace MHB in the metal salt growth inhibition study.

234

Figure 127:- E.coli suspended in DMEM/F-12 with a range of different concentration solutions of copper nitrate o or copper chloride after being incubated for 24 hours at 37 C at 5% CO2: A) copper chloride series with an Δ absorption of 0.001, B) copper chloride series with an Δ absorption of 0.01 C) copper nitrate series with an Δ absorption of 0.001, and D) copper nitrate series with an Δ absorption of 0.01 with concentrations of i) control with no E.coli, ii) control with E.coli, iii) 0.05mM of copper salt, iv) 0.1mM of copper salt, v) 0.5mM of copper salt and vi) 1mM of copper salt

6.4.3 Minimum inhibitory concentration testing using metal salts against Escherichia coli

As described in section 4.4.10, in order to prove whether the metal salts cause a bactericidal or bacteriostatic effect, the E.coli/metal salt suspensions need to be subjected to sodium thiosulfate as this will neutralise the metal salts, as shown by Heidenau et al. (409) This process was carried out using suspensions of silver nitrate and copper nitrate to check that these will be neutralised by the sodium thiosulfate, as indicated by Heidenau et al. (409) This means that in the main metal salt study, it is possible to determine whether a metal salt concentration causes a bacteriostatic effect (the bacteria are “switched off” by the metal ion) or a bactericidal effect (the bacteria are killed by the metal ion).

This study was thus a qualitative test using E.coli/metal salt suspensions which are then plated onto standard I agar plates and compared, as shown in Figure 128. If the concentration causes a bactericidal effect there will be no bacteria present on the plate with and without the additional sodium thiosulfate added to the bacteria/metal salt suspension. On the other hand, if the concentration causes a bacteriostatic effect, then there will no bacteria present on the plate without the additional sodium thiosulfate but there will be bacteria present on the plate with sodium thiosulfate added to the bacteria/metal ion suspension. If the metal salt has no effect on the bacteria then both plates, with and without the sodium thiosulfate added to the bacteria/metal salt suspension, will show the presence of the bacteria

235

Table 41:- CFU counts from copper salt suspensions taken after four and 24 hours using both media types

Time point Concentration Change in 4 hours 4 hours 24 hours 24 hours Sample of copper absorbance of Medium Colony forming units salt (mM) the bacteria (CFU's) Control 0 0.01 DMEM 471 720 TMTC TMTC Control 0 0.001 DMEM 23 35 TMTC TMTC Control 0 0.01 MHB TMTC TMTC TMTC TMTC Control 0 0.001 MHB TMTC TMTC TMTC TMTC

CuCl2 0.05 0.01 DMEM 551 518 TMTC TMTC

CuCl2 0.05 0.001 DMEM 12 71 TMTC TMTC

CuCl2 0.1 0.01 DMEM 698 378 TMTC TMTC

CuCl2 0.1 0.001 DMEM 8 40 TMTC TMTC

CuCl2 0.5 0.01 DMEM TMTC TMTC TMTC TMTC

CuCl2 0.5 0.001 DMEM 630 753 TMTC TMTC

CuCl2 1 0.01 DMEM TMTC TMTC TMTC TMTC

CuCl2 1 0.001 DMEM TMTC TMTC TMTC TMTC

CuCl2 0.05 0.01 MHB TMTC TMTC TMTC TMTC

CuCl2 0.05 0.001 MHB TMTC TMTC TMTC TMTC

CuCl2 0.1 0.01 MHB TMTC TMTC TMTC TMTC

CuCl2 0.1 0.001 MHB TMTC TMTC TMTC TMTC

CuCl2 0.5 0.01 MHB TMTC TMTC TMTC TMTC

CuCl2 0.5 0.001 MHB TMTC TMTC TMTC TMTC

CuCl2 1 0.01 MHB TMTC TMTC TMTC TMTC

CuCl2 1 0.001 MHB TMTC TMTC TMTC TMTC

Cu(NO3)2 0.05 0.01 DMEM 200 916 TMTC TMTC

Cu(NO3)2 0.05 0.001 DMEM 9 29 TMTC TMTC

Cu(NO3)2 0.1 0.01 DMEM 257 872 TMTC TMTC

Cu(NO3)2 0.1 0.001 DMEM 13 91 TMTC TMTC

Cu(NO3)2 0.5 0.01 DMEM TMTC TMTC TMTC TMTC

Cu(NO3)2 0.5 0.001 DMEM 496 477 TMTC TMTC

Cu(NO3)2 1 0.01 DMEM TMTC TMTC TMTC TMTC

Cu(NO3)2 1 0.001 DMEM TMTC TMTC TMTC TMTC

Cu(NO3)2 0.05 0.01 MHB TMTC TMTC TMTC TMTC

Cu(NO3)2 0.05 0.001 MHB TMTC TMTC TMTC TMTC

Cu(NO3)2 0.1 0.01 MHB TMTC TMTC TMTC TMTC

Cu(NO3)2 0.1 0.001 MHB TMTC TMTC TMTC TMTC

Cu(NO3)2 0.5 0.01 MHB TMTC TMTC TMTC TMTC

Cu(NO3)2 0.5 0.001 MHB TMTC TMTC TMTC TMTC

Cu(NO3)2 1 0.01 MHB TMTC TMTC TMTC TMTC

Cu(NO3)2 1 0.001 MHB TMTC TMTC TMTC TMTC * TMTC - too many to count on the agar plate

236

Figure 128A shows a substantial growth of bacteria across the plate compared to Figure 128B that shows that silver nitrate causes an antibacterial effect as no bacteria are observed, indicating that this concentration of silver nitrate caused a bacteriostatic effect. However, it cannot be determined if this effect is bactericidal or bacteriostatic from just the results shown in Figure 128B. Figure 128A shows bacteria growth in the presence of sodium thiosulfate which indicates that the concentration of silver nitrate present in both Figure 128A and Figure 128B had a bacteriostatic effect.

Figure 128:- Standard I agar plates showing; A) E.coli suspension with silver nitrate at a concentration 0.05mM and additional sodium thiosulfate, B) E.coli suspension with silver nitrate at a concentration 0.05mM, C) E.coli suspension with copper nitrate at a concentration 0.05mM and additional sodium thiosulfate and D) E.coli suspension with copper nitrate at a concentration 0.05mM after being incubated in suspension for 24 hours, o plated and further incubated for 24 hours at 5% CO2 and at 37 c using an initial E.coli change in absorption of 0.01

Figure 128C and Figure 128D show that the copper nitrate concentration does not have an antibacterial effect on the E.coli as the plate is observed to be covered in bacteria, which indicates that this concentration of copper nitrate is not high enough to have any antibacterial effect on the E.coli.

These results give an indication for the design of the next set of studies (as shown in section 4.4.11 onwards), but doesn’t prove that the sodium thiosulfate will neutralise the copper nitrate as indicated by Heidenau et al. (409)

237

6.4.4 Metal salts growth inhibition study using Escherichia coli

This initial study used a range of concentrations of metal salts to give an indication of the concentration required to cause not only a bacteriostatic effect but also a bactericidal effect on E.coli for comparison with the study on MSIE scaffolds, which is described in the next section. The experimental process for this study is described in section 4.4.9. The investigation involved adding varying concentrations of silver nitrate or copper nitrate to the DMEM/F-12 with E.coli. This initial study gave a starting point for the final range of concentrations used later on, in particular for the study carried out with the combination of metal salts. This first study only used the change in absorption as the indicator about how bacteria were growing in the metal salt suspension. As shown in section 6.4.1, when bacteria are in ideal growing conditions the change in absorbance will increase as the incubation period increases, meaning that as the absorption change decreases over time the number of living E.coli causing a change in the turbidity of the suspension has decreased.

Figure 129 shows the change in absorption for suspensions of E.coli in DMEM/F-12 with a range of concentrations of silver nitrate. The lower concentrations of silver nitrate (0mM, 0.005mM, 0.01mM and 0.02mM) had little effect on the overall growth of E.coli. These concentrations showed lower changes in absorption than the suspension that contained no silver nitrate which indicates that the presence of silver nitrate, even at these low concentrations, has an effect on the growth of E.coli.

Figure 129:- The change in absorption of the bacterial suspension due the presence of a range of concentrations of silver nitrate over time which gives an indication of the growth rate of the suspended bacteria. * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis.

238

Figure 129 also showed that for the suspensions containing silver nitrate at a concentration of 0.05mM or higher, the change in absorption over time, decreased, indicating that E.coli bacteria cells were not growing. However, whether the antibacterial effect was bacteriostatic or bactericidal cannot be determined from these results. This can only be ascertained by plating the suspensions with sodium thiosulfate, as discussed previously. However these values, in Figure 129, do give an indication of which concentrations should be considered for the next study where the full inhibition concentration of the metal salts will be investigated.

Figure 130 shows the change of absorption due to the presence of varying concentrations of copper nitrate in E.coli suspended in DMEM/F-12. It can be seen that there is no negative effect on the E.coli suspension caused by the presence of the copper nitrate; in fact Figure 130 shows that E.coli grew better in the presence of the copper nitrate compared to the control suspension. Overall this result indicates that the concentration of copper nitrate will have to be significantly higher to have any negative effect on E.coli bacteria.

Figure 130:- The change in absorption of the bacterial suspension due to the presence of a range of concentrations of copper nitrate over time which gives an indication of the growth rate of the suspended bacteria. * Indicates a significant change between the scaffold types (p≤0.05) using ANOVA analysis.

From Figure 129 and Figure 130 it can be determined that the concentrations of silver nitrate and copper nitrate should be at least 0.01mM and 0.1mM, respectively, for the next study to determine the minimum inhibitory concentration of these metal salts. This concentration range will provide a suitable cross section of bacteriostatic, bactericidal and no effect results for each metal salt. Also in the study, reported in the next section, the silver/copper doped scaffolds need to be represented in a metal

239 salt suspension for comparison with the studies in section 6.4.5. Taking the results from Figure 129, the next study investigates concentrations of 0.01mM, 0.05mM, 0.1mM, 0.5mM and 1mM in a ratio of 1:1 of copper nitrate to silver nitrate in the bacteria suspension, as described in section 4.4.9 and 4.4.10.

It was noted that the bacteria were not growing within expected limits in the current conditions and in combination with the metal salts the pH of the bacterial suspension increased as described previously in section 6.3.2 for the cell culture work. This can be partially resolved by increasing the

CO2 levels in the incubator during the bacterial immersion as this will reduce the initial pH in the DMEM/F-12 and keep the pH down to an acceptable level that corresponds to a suitable pH for cell growth, as shown in Figure 131. Using this graph the CO2 levels were increased to 9.5% CO2 in the incubator for 24 hours before the samples were plated. This process reduced the pH down to approximately 7.1 when the level of NaHCO3 was at 2.4mg/ml which is a suitable pH for both cells and bacteria to grow in.

Figure 131:- Graph showing CO2 levels required to change the pH for given levels of NaHCO3 found in DMEM/F- 12 to give a suitable pH value for cells and bacteria to grow

The second study using the metal nitrate salts considered the CFU’s formed after the suspensions had been incubated for 24 hours, plated and the results collected after further 24 hours, as described in sections 4.4.9 and 4.4.10. The concentrations ascertained from the previous study were considered using the change of absorption as the indicator of change in the growth of E.coli.

Figure 132 shows the suspensions of the metal salts in DMEM/F-12 with E.coli and gives an early indication of which concentrations of the metal salts are affecting the E.coli. Figure 132A shows

240 the suspensions with different concentrations of silver nitrate and compare the media control and the E.coli suspension control (Figure 132Ai and Figure 132Aii respectively). The suspensions containing silver nitrate (Figure 132Aiii-vi) all show colouring and a clarity which is similar to the media control rather than the E.coli suspension control. This result indicates that the silver nitrate has a negative effect on the bacteria, as was shown in Figure 127, with the colour and the turbidity of the suspension changing from orange to yellow and from clear to cloudy, respectively.

Figure 132:- DMEM/F-12 and E.coli suspensions with A) silver nitrate with concentrations of iii) 0.01mM iv) 0.05mM v) 0.1mM and vi) 0.5mM, B) copper nitrate with concentrations of iii) 5mM iv) 2mM v) 1mM vi) 0.5mM and vii) 0.1mM, C) combination of silver and copper nitrate in a ratio of 1:1 with concentrations of iii)1 mM iv)0.5mM v)0.1mM vi) 0.05mM and vii) 0.01mM, with i) and ii) in each picture showing a plain media control and an E.coli suspension control, respectively

241

Figure 132B shows the copper nitrate suspensions along with a control medium and an E.coli suspension control (Figure 132Bi and Figure 132Bii respectively). The suspensions with concentrations of copper nitrate of 5mM and 2mM (Figure 132Biii and Figure 132Biv) show the turbidity to be similar to the control medium but with some discolouration compared to the control. This result suggests that these concentrations of copper nitrate have a negative effect on the E.coli and that the discolouration of the media is due to the presence of the dark blue copper nitrate. It is also observed that the suspension with a copper nitrate concentration of 1mM (Figure 132Bv) shows an increase in its turbidity compared to the control. It maintains a similar colour to the control and this would indicate that this concentration has a bacteriostatic effect on the E.coli but it does not exhibit a full bactericidal effect. The lowest two copper nitrate concentrations, 0.5mM and 0.1mM, both showed a similar colour and turbidity to that seen in the E.coli suspension control (Figure 132Bvi, Figure 132Bvii and Figure 132Bii, respectively). This result shows that at these concentrations of copper nitrate E.coli bacteria is not being negatively affected.

Figure 132C shows the combination silver and copper nitrate suspensions along with a control medium and an E.coli suspension control (Figure 132Ci and Figure 132Cii respectively). The lowest concentration suspension showed little difference from the E.coli suspension in terms of its turbidity and colouring (Figure 132Cvii) which indicates that at this concentration the combined metal salts solution has no negative effect on E.coli as the concentration of the metal salts is not sufficiently high. The rest of the concentrations of the combined metal salt suspensions (Figure 132Ciii to Figure 132Cvii) showed colouring and turbidity similar to that seen in the control (Figure 132Ci) indicating that the concentration of metal salts in these suspensions was sufficient to cause a negative effect on E.coli.

Figure 132 gives thus a qualitative representation of the suspensions behaviour after the suspensions have been plated onto agar plates indicating which concentrations of the metal salts have had an effect on the bacteria. However this representation does not show whether the effect on the bacteria will be bacteriostatic or bactericidal.

Table 42, Table 43 and Table 44 present the CFU counts from the plated samples shown in Figure 132 indicating which concentrations of the metal salts have induced a negative effect on the E.coli whether that is bacteriostatic or bactericidal. Each table shows the CFU count for bacteria/metal salt suspensions with and without sodium thiosulfate, therefore observed effects can be categorised as either bacteriostatic, bactericidal or no effect, as described in section 6.4.3.

242

Table 42:- Colony forming units counted after being exposed to varying concentrations of silver nitrate in the E.coli suspension after 24 hours of incubation (TMTC meaning too many to count)

Concentration Colony forming units of silver With Sodium Dilution nitrate Thiosulfate 3 4 5 6 0.01 TMTC TMTC 1289 154 11 3 0.05 526 632 0 0 0 0 0.1 0 0 0 0 0 0 0.5 0 1 0 0 0 0

Table 43:- Colony forming units counted after being exposed to varying concentrations of copper nitrate in the E.coli suspension after 24 hours of incubation (TMTC meaning too many to count)

Concentration Colony forming units of copper With Sodium Dilution nitrate Thiosulfate 3 4 5 6 0.1 TMTC TMTC TMTC TMTC 432 58 0.5 TMTC TMTC TMTC 341 35 5 1 TMTC TMTC 832 172 18 2 2 801 821 6 2 0 0 5 500 621 0 0 0 0

Table 44:- Colony forming units counted after being exposed to varying concentrations of combination of copper nitrate and silver nitrate, in a 1:1 ratio, in the E.coli suspension after 24 hours of incubation (TMTC meaning too many to count)

Concentration Colony forming units of silver and With Sodium Dilution copper nitrate Thiosulfate 3 4 5 6 0.01 TMTC TMTC TMTC 822 114 19 0.05 TMTC TMTC 50 36 1 0 0.1 1022 1011 0 0 0 0 0.5 782 809 0 0 0 0

Table 42 shows the CFU counts for the silver nitrate concentrations indicating that concentrations above 0.01mM cause a bacteriostatic effect and for concentrations above 0.05mM a bactericidal effect on the E.coli. This result indicates that for the silver nitrate to be an effective antibacterial agent the concentration needs to be higher than 0.01mM. This result also supports previous relevant work that has been reported in the literature. (468) (469)

Table 43 shows the CFU counts for different copper nitrate concentrations and this data shows that in order for copper nitrate to produce a bacteriostatic effect the concentration needs to be above

243

2mM and for a bactericidal effect the concentration needs to be above 5mM. These required concentrations are significantly higher than the concentrations required for silver nitrate which indicates for this set of tests that silver nitrate out performs copper nitrate in terms of anti-bacterial potency.

Finally Table 44 shows the CFU count for the combination metal salt data indicating that in order to have a bacteriostatic effect the combined concentration of the two metal salts must be higher than 0.05mM and to produce a bactericidal effect the combined metal salt concentration must be above 0.5mM. This combination of metal salts outperformed copper nitrate but gave similar results to silver nitrate in terms of the bacteriostatic effect. Notably, the bactericidal effect caused by the combination of the two metal salts required a higher concentration compared to that used in silver nitrate only. This result could indicate that the presence of copper nitrate was hampering silver nitrate in causing a bactericidal effect, but the reasons for this behaviour were not further investigated here. These results offer no insight into the mechanisms of how the metal ions cause the bactericidal or bacteriostatic effects on the E.coli, as seen in work carried out by other researchers. (296) (409)

6.4.5 Growth inhibition capabilities of molten salt ion exchange scaffolds using Escherichia coli

The results presented in sections 6.4.1 to 6.4.4 are preliminary optimisation studies based on a similar investigation presented by Heidenau et al (409) and tailored the experimental conditions for the study, which was designed to determine if the MSIE scaffolds will have an antibacterial effect on the E.coli. This study uses the supernatants from immersed doped scaffolds, as described in section 4.4.11 and the results can be compared to those obtained in the previous section.

Figure 133 shows the E.coli suspensions with the supernatants after 24 hours of incubation at

o 37 C at 9.5% CO2 on a shaking plate. The supernatants covered samples of all types of MSIE scaffolds (silver, copper and silver/copper MSIE samples) and the BBG scaffold as a control. Immersion times of 1, 3, 7 and 28 days were investigated. Figure 133 shows a medium control in each picture as well as an E.coli suspension control for comparison. It can be seen that all of the supernatant samples show turbidity and colouring similar to the E.coli suspension control. This result indicates that the concentration of ions released by the scaffolds into the media added to the E.coli suspensions was not high enough to cause any negative effect on the E.coli.

244

Figure 133:-Suspensions of E.coli in DMEM/F-12 with supernatants added from scaffolds immersed in DMEM/F- o 12 for A) 1 day B) 3 days C) 7 days and D) 28 days after 24 hours of incubation at 37 c and 9.5% CO2

The suspensions were then plated onto standard I agar plates with two sets of plates; the first set had additional sodium thiosulfate added to the suspensions and it was not diluted while the second set only used the suspensions and they were diluted, as described in section 4.4.8. Photographs of the plates are shown in appendix C2 and all the collated data shown in the following tables and graphs are taken directly from those figures.

Table 45 to Table 48 show the CFU data for the supernatants taken from the plain, silver, copper and silver/copper samples, respectively, immersed in DMEM/F-12 for 1, 3, 7 and 28 days. These figures show that none of the samples had an effect on the bacteria as CFU’s were present on all the plates and there was no marked reduction during the immersion period, i.e. no statistically meaningful difference between the samples taken after one day of immersion compared to those samples taken after 28 days of immersion.

Table 45:- Colony forming units counted after being exposed to supernatants from immersed plain scaffolds after o 24 hours incubating at 37 C at 9.5% CO2

Immersion Colony forming units period of with sodium Dilution the scaffold thiosulfate 3 4 5 6 Control n/a n/a TMTC 783 102 18 1 day TMTC TMTC TMTC 568 108 20 3 days TMTC TMTC TMTC 691 126 18 7 days TMTC TMTC TMTC 412 78 8 28 days TMTC TMTC TMTC 1034 330 37 * TMTC - too many to count

245

Table 46:- Colony forming units counted after being exposed to supernatants from immersed low concentration o silver MSIE scaffolds after 24 hours incubating at 37 C at 9.5% CO2

Immersion Colony forming units period of with sodium Dilution the scaffold thiosulfate 3 4 5 6 Control n/a n/a TMTC 783 102 18 1 day TMTC TMTC TMTC TMTC 500 120 3 days TMTC TMTC TMTC 606 54 7 7 days TMTC TMTC TMTC 734 135 20 28 days TMTC TMTC TMTC 931 122 24 * TMTC - too many to count

Table 47:- Colony forming units counted after being exposed to supernatants from immersed low concentration o copper MSIE scaffolds after 24 hours incubating at 37 C at 9.5% CO2

Immersion Colony forming units period of with sodium Dilution the scaffold thiosulfate 3 4 5 6 Control n/a n/a TMTC 783 102 18 1 day TMTC TMTC TMTC 682 113 11 3 days TMTC TMTC TMTC 668 142 10 7 days TMTC TMTC TMTC 633 64 11 28 days TMTC TMTC TMTC 171 26 0 * TMTC - too many to count

Table 48:- Colony forming units counted after being exposed to supernatants from immersed low concentration o silver/copper MSIE scaffolds after 24 hours incubating at 37 C at 9.5% CO2

Immersion Colony forming units period of with sodium Dilution the scaffold thiosulfate 3 4 5 6 Control n/a n/a TMTC 783 102 18 1 day TMTC TMTC TMTC 631 150 16 3 days TMTC TMTC TMTC 564 121 18 7 days TMTC TMTC TMTC 352 45 3 28 days TMTC TMTC 718 129 11 1 * TMTC - too many to count

Several interesting aspects arise from analysing these results. Firstly, the BBG scaffold seems to have no anti-bacterial effect on E.coli. Secondly, the concentrations of metal ions introduced through MSIE were not high enough to cause any anti-bacterial effect on the E.coli as no difference was seen between the plates that had had the supernatant from the MSIE scaffolds and the E.coli plates used as a

246 control. The results also show that there was little difference between the types of metal ions used although the longer immersion period did yield slightly better results for the copper and the silver/copper samples, in terms of perceived antibacterial effect on the E.coli.

6.4.6 Agar diffusion studies using Escherichia coli

Agar diffusion tests were carried out, as described in section 4.4.12.I, first using metal salts suspended in DMEM/F-12 and then using the supernatants from low concentration MSIE scaffolds in DMEM/F-12 from a range of immersion periods. This study was done to give a direct comparison between known concentration samples (metal salt solutions) and the unknown concentration samples (scaffolds).

Figure 134 shows agar diffusion using suspensions of silver nitrate in DMEM/F-12 for a range of concentrations to determine which concentrations of silver nitrate would cause an antibacterial effect, but it should be noted that this type of test will not show if the effect on the E.coli is bacteriostatic or bactericidal. The test shows just whether or not the sample is having an effect on the bacteria. It can be seen that the lower concentrations of silver nitrate, shown in Figure 134A, exhibits no visible halo around the sample discs with silver nitrate concentrations of 0.02mM and 0.05mM, indicating that these concentrations had no effect on the E.coli present, however there was a slight halo surrounding the sample with a concentration of 0.1mM of silver nitrate. The concentrations of silver nitrate shown in Figure 134B all exhibit varying sized halos around the sample disc indicating the occurrence of an antibacterial effect. This result confirms the data from section 6.4.4, in that the concentration of silver nitrate needs to be above 0.05mM in order to cause an antibacterial effect on the E.coli.

The same agar diffusion study was carried out using suspensions of copper nitrate in DMEM/F- 12 at a range of concentrations, as described in section 4.4.12. Figure 135 shows the agar diffusion plate of the higher concentration sample discs with concentrations of 0.5mM, 1mM and 2mM of copper nitrate (and a medium control disc). It is observed that none of them show any discernible halo surrounding the disc indicating that the concentration of copper nitrate was not high enough to have any antibacterial effect on the E.coli. This result agrees with the data shown in section 6.4.4 and indicates that for copper to be an effective antibacterial agent the concentration of copper needs to be over 2mM.

247

Figure 134:- Results of the agar diffusion study to show the growth inhibition of silver nitrate solutions with A) showing a plain media disk, 0.1mM, 0.05mM and 0.02mM concentrations of silver nitrate on filter paper disks and B) showing a plain media disk, 1mM, 0.5mM and 0.2mM concentrations of silver nitrate on filter paper disks

Figure 135:- Result of the agar diffusion study to show the growth inhibition of dilutions of copper nitrate inoculated onto filter paper disks. Concentrations of 2mM, 1mM and 0.5mM were used along with a control disks inoculated with media

The second part of the agar diffusion study involved using supernatant from all the sample types to conclude if the concentrations used in the scaffolds are high enough to cause an anti-bacterial effect, as described in section 4.4.12.I. Typical resulting plates are shown in Figure 136. The supernatants

248 taken from the immersed plain scaffolds from the varying immersion periods showed no visible halo around the discs indicating that these had no anti-bacterial effect on the E.coli, as shown Figure 136A. These results thus correspond with the results presented in section 6.4.5.

Figure 136B and Figure 136D show the results for the silver only scaffolds and the combination silver and copper scaffolds along with a metal salt control with a concentration of 1mM. None of the supernatant samples presented in these two figures showed any indication of anti-bacterial activity due to the lack of a halo around the disc; however each of the control samples showed a halo indicating that this concentration was sufficient to cause an anti-bacterial effect on the E.coli. These results are in agreement with the data shown in the previous section.

Figure 136:- Results of the agar diffusion study to show the growth inhibition of: A) BBG scaffold, B) Low concentration silver-doped BBG scaffold, C) Low concentration copper-doped BBG scaffold and D) Low concentration combination ion-doped BBG scaffold in the presence of E.coli bacteria

249

Finally Figure 136C shows the supernatant results for the copper scaffolds with a metal salt control with a concentration of 1mM. None of the sample discs, including the control, showed any antibacterial effect leading to the conclusion that the copper concentration was not high enough in these samples. This data was also in agreement with the results found in the previous work (reported in section 6.4.5).

In conclusion the lack of any visible halo surrounding the supernatant discs confirms the results found in section 6.4.5 indicating that the concentration of metal ions introduced using the MSIE technique is not high enough to cause any antibacterial effect on the E.coli. These agar diffusion tests also support the results shown in section 6.4.4 with respect to the concentration of metal salts required to cause an antibacterial effect on the E.coli.

6.4.7 Agar diffusion studies using Staphylococcus aureus

The effect of the different MSIE scaffolds, at their lowest concentration, and of a control BBG scaffold on S.aureus was tested using agar diffusion, as described in section 4.4.11, to give an indication of the growth inhibition caused by these samples. If the scaffolds had a bacteriostatic or bactericidal effect on S.aureus, a halo effect would be seen on the agar plate surrounding the sample. As mentioned above, the clearer and larger this halo is, the more effective is the sample against bacteria in qualitative terms. The results of the diffusion test carried out using S.aureus are shown in Figure 137.

It can be seen from Figure 137A that BBG scaffolds have no bacteriostatic or bactericidal effect as there is no perceptible halo around the scaffold. Figure 137B-D all show at least a limited amount of bactericidal or bacteriostatic effect as they all exhibited varying degrees of the antibacterial halo. This result is encouraging as it means that the concentration of therapeutic metal ions in the samples is high enough to cause a reaction to S.aureus. The copper-doped sample showed the highest effect against S.aureus as it had the largest antibacterial halo surrounding it compared to the silver-doped sample and the combination ion-doped sample. The combined Cu/Ag-doped sample showed the smallest antibacterial halo but the halo was still present, after the incubation period of 24 hours, indicating that if the sample was left for longer the ions might produce a longer lasting antibacterial effect.

This result must be considered in future work involving MSIE scaffolds as it gives an indication of whether the samples are bacteriostatic or bactericidal against S.aureus. If the halo stayed visible or increased further in size, this would indicate that the bacteria cells were being killed by the scaffold and therefore the sample would be considered bactericidal. However if the halo decreased in size this would indicate that the sample was bacteriostatic.

250

Figure 137:- Results of agar diffusion to show the growth inhibition properties of: A) BBG scaffold, B) Low concentration silver-doped BBG scaffold, C) Low concentration copper-doped BBG scaffold and D) Low concentration combination ion-doped BBG scaffold in the presence of S.aureus bacteria after an incubation period of 24 hours

Comparing this result with the results found using E.coli, it is shown that S.aureus was more susceptible to the MSIE samples than the E.coli. This also showed that whilst the concentration of the MSIE metal ions was high enough to have an antibacterial effect on the S.aureus bacteria it was not high enough to have effect on the E.coli bacteria. The main reason is related to the fact that E.coli is a gram-negative bacteria and S.aureus is a gram-positive bacteria, which means that the bacteria behave differently in terms of how they react to the metal ions present in the incubation medium, as described in the literature review.

251

6.4.8 Growth inhibition studies using Klebsiella pneumoniae

The quantitative bacterial experiment using K.pneumoniae following the method described in section 4.4.13 showed that 18 hours after being seeded with K.pneumoniae there was a complete reduction in the number of CFU’s present, as shown in Table 49, compared with the control sample, which was an industry standard textile (PES). (470) (471) (472)

Table 49:- Results of bacterial experiments carried out on 3D scaffolds loaded with K.pneumoniae and incubated for 18 hours at 36oC

Initial Load Final load Change cfu Increase Sample type Number of cfu’s per sample Percentage (%) or decrease Control textile PES 4.91 x 104 1.24x107 25154.58 Increase 45S5 Bioglass scaffold 4.91 x 104 <20 -99.96 Decrease Silver-doped scaffold (*) 4.91 x 104 <20 -99.96 Decrease Copper-doped scaffold (*) 4.91 x 104 <20 -99.96 Decrease Combination ion-doped scaffold (*) 4.91 x 104 <20 -99.96 Decrease (*) All at the lowest concentration of MSIE samples

All four sample types showed an almost complete reduction in the number of CFU’s, which indicated that all samples had antibacterial effect on the K.pneumoniae bacteria, including the BBG scaffold. Although this study showed that all four sample types affected K.pneumoniae bacteria the results gave no indication as to which of the MSIE samples was more effective, thus it cannot be ascertained which of the released metal ions (whether from the BBG scaffold or from the additional ions for example for silver and copper) gave the most effective antibacterial activity. However this study also indicated, as was the case with the results presented in section 6.4.5 for the S.aureus bacteria, that the concentration of the added metal ions in the MSIE samples was high enough to cause an effect on the K.pneumoniae bacteria, although this effect cannot be attributed to the release of the additional metal ions introduced through MSIE over the ions released from the BBG scaffold. This study did not provide any insight or support to the current theories as to how the bacteria are disrupted by the metal ions, as described in the literature review. Comparing this study to the agar diffusion results shown in section 6.4.6 for the E.coli, it can be stated that the samples were toxic enough to have an effect on the K.pneumoniae but not on the E.coli despite the fact they are both gram-negative bacteria.

252

6.5 Summary of results from the cell biology and bacterial studies

6.5.1 Summary of cell biology work

Overall the cell biology work in collaboration with the Leeds Dental Institute produced relevant results and paved the way forward for the work that was carried out in collaboration with the University of Erlangen-Nuremberg. This work showed that whilst all silver-doped scaffolds obtained by different concentrations and immersion periods produced no observable cytotoxic effect according to the live/dead staining study, the best results were seen on scaffolds obtained by the lowest concentration of silver, and immersed for the shortest time in the salt bath. These results support data published by other researchers on similar scaffold types and parameters. (6) (113)

The studies using rMSC’s showed that the samples needed to go through a more rigorous pre- conditioning before being exposed to the cells in-order to reduce the potential cytotoxic effect caused by the rise in pH. This work showed that once scaffolds were conditioned both the plain and silver doped samples had reduced cytotoxicity compared to that found in preliminary studies. However it was seen that the copper doped samples and silver/copper doped samples created an environment that was cytotoxic, this could suggest that the concentration of the metal ions was too high or that the release rate was excessive and caused an overdose when exposing cells to the samples.

The studies using hMSC’s showed that despite the reduction in concentration in the copper and silver/copper MSIE samples, they still produced a cytotoxic effect on the cells. This result suggests that further optimization of these MSIE scaffolds is required before further in vitro and in vivo work can be carried out with them.

The significant difference between the work carried out at the Leeds Dental Institute and that carried out with the University of Erlangen-Nuremberg is likely related to the seeding system in each case. The HPLSC work used a dynamic seeding system that was constantly moving the cell culture media through and around the scaffolds and therefore washing away dead cells and ions away from the scaffolds. On the other hand the rMSC and hMSC cell culture work was carried out using a static seeding technique which did not allow any movement of the medium around the cells. In terms of this work it can be seen that the dynamic seeding system gave more consistent results and in terms of accuracy it is closer to the actual in vivo environment than the static seeding model. This result also confirms that future studies will benefit from introducing the application of bioreactors into the development of similar scaffolds to optimise scaffold composition in conditions closer to the in vivo environment. (473) (474) (475) It is worth noting that most in vitro work carried out previously on BBG scaffolds was conducted in static conditions. (476) (352) (477) (478) Our results have confirmed,

253 especially in the case of ion-doped scaffolds, the relevance of understanding the in vitro behaviour in dynamic conditions.

As stated in chapter five, the surface of a biomaterial plays a crucial role in the establishment of the interface between the seeded cells and the scaffold and the literature has shown that the quality of the initial cell attachment to a scaffold’s surface determines subsequent cell behaviour including adhesion, spreading, morphology, migration, proliferation and differentiation. (462) However recent findings have shown that not only the roughness is an important parameter but also the surface morphology, in terms of a surfaces texture and overall diversity of observed features, especially for long term cell adhesion compared to the surface chemistry and roughness. (462)

Cells cultures on very rough surfaces, such as those observed for the copper and silver/copper doped scaffolds, tend to exhibit an increase in osteoblasts differentiation, ALP activity and protein production and a decrease in cell proliferation. (461) Cells have been observed to be smaller on rougher surfaces, which was observed in some of the images in this work, with the quality of the cell contact on a rough surface being related to the minimum width of the peaks and troughs on the surface. (462) These observations are a good indicator as to whether the surface roughness affects cell behaviour in the context of the present investigation.

It should be noted that another aspect that makes the assessment of the effect of surface roughness difficult is the MSIE doping process itself. (461) Indeed, the effect of surface roughness on cells maybe be attributed to the actual roughness itself or to the pre-conditioning, sterilisation treatment or to the presence of additional metal ions or a combination of these factors. (461) Clearly, these complex effects need to be investigated further to produce a complete protocol that will take into account some of these parameters to eliminate their additional error, and to be able to ascertain the independence of the different variable involved.

Just as with the changes in surface roughness, the available data do not allow to conclude that the cell death observed was caused by the change in sample type solely as cell death detected could have also been by normal apoptosis. (404) In the future an additional assay needs to be carried out to rule out cells that died due to apoptosis. (404)

6.5.2 Summary of bacterial studies

Overall the bacterial work carried out in this project produced results which confirmed the viability of the use of metal ions in the development of multifunctional Bioglass® -based scaffolds. The growth curves that were first produced showed that whilst the bacteria grew in the CCM that was used

254 in the cell biology studies, the bacteria suspension required more acclimatisation time compared to the more conventional bacteria media (Mueller Hinton broth in this case), but once this period was over the growth exceeded that of the conventional medium.

It was discovered that the optimum growing conditions for the bacteria included a two hour

o acclimatisation period and that the bacteria suspensions were incubated in 9.5%CO2 and 37 C compared

o to standard conditions of 5%CO2 and 37 C. This increase in CO2 levels decreased the pH in the bacteria suspension making the growing conditions in the suspension suitable for the experiments in the content of this work.

The use of sodium thiosulfate was considered to deactivate the silver and copper salts to show whether the addition of these salts had a bactericidal or bacteriostatic effect. The experiment showed that in the DMEM/F-12 medium the presence of sodium thiosulfate deactivated the metal salts which made it possible to determine which concentrations of the metal salts were bactericidal or bacteriostatic.

After 24 hours incubation and 24 hours on agar plates the minimum bactericidal and bacteriostatic concentration for the silver nitrate suspension was found to be 0.05mM. The minimum bacteriostatic concentration for the copper nitrate suspension was found to be 5mM and after using sodium thiosulfate to deactivate the metal nitrate salts the minimum bactericidal concentration was found to be above 5mM. The minimum bacteriostatic concentration for the combined copper and silver nitrate suspension was found to be 0.1mM and after using sodium thiosulfate to deactivate the metal nitrate salts the minimum bactericidal concentration was found to be above 1mM.

A higher concentration of copper nitrate is required to produce the desired antibacterial effect and the presence of the copper nitrate salt appears to have a detrimental effect on the effectiveness of the silver nitrate as suggested by the increased concentration required when the combination of metal nitrate salts was tested. This experiment thus gave insight into the concentration of silver and copper nitrate, separately and in combination, required in the samples in order to be effective as an antibacterial scaffold. It should be noted that although the bactericidal concentrations for the copper and combination nitrate were not determined, the bacteriostatic concentrations were obtained, and this result is important because if a bacteria is held in a bacteriostatic environment for a sufficient period of time, bacteria cells will start to die due to being held in an environment that is toxic.

The experiments then required the use of supernatant that was taken from DMEM/F-12 that had been in contact with MSIE scaffolds for 1,3,7 and 28 days, combined with a bacteria suspension to investigate whether the concentrations of metal ions in the scaffold are sufficient to produce the desired antibacterial effect. The suspensions that were plated with the addition of sodium thiosulfate showed almost complete growth for all sample types over all immersion time points. The suspensions plated without the presence of sodium thiosulfate all showed no effect on the bacterial growth across all sample

255 types and immersion periods in DMEM/F-12. This result indicates that the concentration of the metal salts present in the scaffolds was not high enough to have a bacteriostatic effect or bactericidal effect.

Determining the antibacterial effect of copper and silver doped samples (and nitrate salts) through diffusion gave similar results to the previous set of experiments. Indeed, none of the samples that had been subjected to the various scaffold supernatants showed any inhibition zones in a laboratory setting. This behaviour suggests that the concentration of the metal ions present in the scaffolds was not high enough to cause either a bacteriostatic or bactericidal effect under the experimental conditions investigated. However this diffusion experiment did correlate with the results found in the minimum inhibition experiments in the presence of varying concentrations of metal nitrates, thus enforcing the minimum inhibition values found there.

The minimum inhibition zone experiments carried out using S.aureus (carried out by Dr Timo Hammer from Hohenstein Institut für Textilinnovation e.V, Germany) showed that the presence of the metal ions had an effect on the bacterial growth with the copper scaffold having the largest antibacterial effect on the bacteria out of the four sample types. This result indicates that the concentrations, whilst not being high enough to have an antibacterial effect on the E.coli, are high enough to have at least a bacteriostatic effect. A quantitative test using K.pneumoniae was also carried out to show how the scaffolds performed against this type of bacteria. These preliminary tests showed that all four types of scaffold, including the BBG scaffold, had at least a bacteriostatic effect on this type of bacteria, compared to the control. This finding is useful as it gives an indication that for some types of bacteria the concentration of metal ions present in the scaffolds was high enough to have the desired effect on the bacteria cells and clearly confirmed that the incorporation of metallic ions into bioactive glasses as antibacterial agent needs to be tailored based on the target bacteria.

256

Chapter Seven

7. Concluding remarks

In the scope of this project, the fabrication and characterisation of novel multifunctional three dimensional bioactive composite scaffolds based on 45S5 Bioglass® -derived glass-ceramics intended for BTE applications have been investigated. The overall aim of this work, as described in chapter three, was to improve the existing conventional BBG scaffolds whilst maintaining their desirable intrinsic bioactive properties. The BBG scaffolds were modified using a combination of MSIE processing and polymer coating to produce novel scaffolds with enhanced properties for use in BTE.

The experiments, described in chapter four, were designed and carried out to achieve the objectives set out in chapter three and the results of this work are presented in chapters five and six. In general, the results of this work have demonstrated that the introduction of silver and/or copper through MSIE and the addition of a gelatin coating have improved the mechanical properties, whilst simultaneously introducing additional functionality (antibacterial for example) and maintaining the desirable bioactivity of the BBG scaffolds. The forthcoming sections summaries the key results achieved, presented in chapters five and six. In addition, this section presents a final outlook on how this work contributes to the general field of BTE. This chapter does not explore the scope of future work as this is covered in the next chapter.

7.1 General observations on 45S5 Bioglass®- derived glass-ceramic scaffolds

This work has established that the foam replication technique reliably produces bone tissue scaffolds suitable for further modification. The initial characterisation of the BBG scaffolds was carried out to establish a robust control for comparison with the modified scaffolds and to check that these BBG scaffolds exhibited the established properties discussed in previous research for this class of scaffolds, described in chapter two. It was established that the BBG scaffolds produced using the foam replication technique were similar to those reported previous work as follows:-

 Scaffolds exhibited an average porosity of 93±2% which fell by an average of 10% to 84±2% after immersion in SBF for 14 days  SEM analysis showed that the scaffolds had a pore size in the range of 200-600µm, a strut cross-section of between 25 and 75µm and maintained a pore shape and interconnectivity suitable for BTE, after immersion in SBF for 14 days

257

 Chemical composition and crystalline structure of the BBG scaffolds were investigated using XRD, EDX and FTIR, pre and post SBF immersion  After immersion in SBF, SEM analysis showed the presence of HA which is the marker for bioactivity in the context of bone contacting materials and it has been identified in the literature as the precursor to enable bone attachment  Scaffolds showed a surface topography suitable for cell attachment displaying a contact angle of 28±4o and 23±2o pre and post SBF immersion and white light interferometry RMS, Ra and PV values of 3±1µm, 2.6±0.9µm and 19±9µm pre-SBF immersion and 6±1µm, 6±1µm and 26±9µm post-SBF immersion, respectively  Scaffolds exhibited a compressional strength in the range of 0.4-0.5MPa, which did not significantly after immersion in SBF for 14 days  The BBG scaffolds increased the pH of the SBF and CCM through the rapid release of ions once the scaffolds are immersed, which can cause a cytotoxic effect on any cells or bacteria present  When HPLSC’s were exposed to the BBG scaffolds in CCM in a dynamic cell seeding environment there was no cytotoxic effect on the cells detected and the scaffolds did not have a detrimental effect on the cells ability to attach to the surface of the scaffold  The pH increase caused by the BBG scaffolds caused significant cell death (in the rMSC study) when using a static cell seeding environment  Once the pre-conditioning treatment was introduced, cell death significantly reduced and gave AB results similar to those observed in the control well, confirming that the conditioned BBG scaffolds were suitable to seed rMSC’s and hMSC’s  BBG scaffolds had no anti-bacterial effect on E.coli after the bacteria was added to supernatant CCM or during agar diffusion studies  The BBG scaffolds had no anti-bacterial effect on S.aureus after an agar diffusion study as no perceptible halo around the scaffold was observed  The BBG scaffolds caused a significant reaction when K.pneumoniae was exposed to the scaffolds indicating that there was an anti-bacterial reaction caused by the scaffold

Overall the results of the structural and biological characterisation agree with published literature results and they contribute to the overall knowledge and understanding about these increasingly used BBG scaffolds for BTE.

258

7.2 Molten salt ion exchange scaffolds

This work has established for the first time that the MSIE technique can be used to introduce therapeutic metal ions into prefabricated BBG scaffolds. The introduction of silver and/or copper through MSIE produced the following results:-

 Silver and copper were successfully introduced into BBG scaffolds as confirmed by XRD, FTIR, ICP-MS and EDX and were still present after immersion in SBF for 14 days. Ion doping by MSIE did not alter the scaffold structure significantly so that after immersion in SBF HA was detected on the surface of the scaffold indicating that the introduction of the silver and/or copper did not have a detrimental effect on the desirable bioactivity of the BBG scaffolds  The porosity of the MSIE scaffolds was, on average, 20% lower than that observed for the BBG scaffolds but it was still above the ideal threshold of 75%. The porosity of the ion-doped scaffolds after immersion in SBF, was seen to reduce by approximately 10% which was similar to the reduction of the porosity of the BBG scaffolds after immersion in SBF  The pore size, shape and interconnectivity were observed to be similar to those observed in the base scaffolds indicating that the MSIE processing does not detrimentally effect the overall pore structure of the base scaffold  The surface topography was significantly altered due to the MSIE processing, namely the MSIE process changed the surface roughness of struts. The silver-doped scaffolds exhibited a slightly smoother surface topography with a reduction in the contact angle whilst the copper-doped and silver/copper-doped scaffolds exhibited an increase in their surface roughness along with an increase in the contact angles compared to those observed for BBG scaffolds. This result indicated the possibility of coupling chemical modification and surface roughness control by designing the parameters of the MSIE technique. It can not be conclusively proved that the change in the surface roughness will lead to a change in cell behaviour.  The ion-doped scaffolds had a marginally higher, on average 15% higher, compressional strength than the BBG scaffold. After SBF immersion, the ion-doped samples showed a slight reduction in their compressional strength as the BBG scaffolds  Depth EDX analysis showed that silver penetrated deeper into the scaffolds silicate structure compared to copper but a higher concentration of copper was detected infiltrating the ion-doped scaffolds compared to silver  ICP-MS along with EDX analysis showed that copper was released in a fast intense burst compared to the slow, controlled and low release of the silver within the ion-doped scaffolds  It can not be conclusively proved that the MSIE process has led to a complete incorporation of the ions into the glass-ceramic structure as the silver and copper detected could be also from

259

residue salt that have infiltrated the micro-cracks on the surface of the scaffold. Additional characterisation work is required in this instance  When the silver-doped samples were dynamically seeded with HPLSC’s the silver-doped scaffolds did not have a detrimental effect on the attachment of HPLSC’s and did not cause a cytotoxic effect  Initial cell biology studies with rMSC’s showed that all three ion-doped scaffold types, along with the BBG scaffolds, caused a significant cytotoxic effect, which was mainly attributed to the pH increase caused by the release of ions from the BBG scaffolds. It can not be proved that the change in surface roughness caused by the MSIE process did not have effect on the cell behaviour  After additional pre-conditioning was introduced, the copper-doped and silver/copper-doped scaffolds still had a detrimental cytotoxic effect on the rMSC’s while the silver-doped scaffolds were found to only cause a slight cytotoxic effect. This result indicated that the concentrations used in the copper-doped and silver/copper-doped scaffolds are too high as they are cytotoxic on the rMSC’s. The lowest concentration of silver in the molten salt bath coupled with an immersion period of 15 minutes produces a scaffold that allows cell attachment and does not cause a detrimental cytotoxic effect. The change in topography could be attributed to the change in cell behaviour as well as the introduction of the therapeutic ions  Minimum inhibitory concentrations of silver nitrate and/or copper nitrate to give either a bacteriostatic or bactericidal effect on the E.coli were established as 0.05mM, 5mM and 1mM for silver nitrate in CCM, copper nitrate in CCM and a 1:1 CCM suspension of silver nitrate to copper nitrate, respectively  The concentrations of the therapeutic ions in the silver-doped, copper-doped and silver/copper- doped scaffolds were not sufficient to cause either a bacteriostatic or bactericidal effect on the E.coli  Agar diffusion studies carried out using S.aureus showed that the concentrations of the therapeutic ions in the silver-doped, copper-doped and silver/copper-doped scaffolds was high enough to cause an anti-bacterial effect on the S.aureus  Test carried out using K.pneumoniae also indicated that the concentrations of the therapeutic ions in the silver-doped, copper-doped and silver/copper-doped scaffolds were high enough to cause a decrease in the bacteria population

In conclusion, when balancing the structural results with the biological and bacterial studies, the lowest concentration silver-doped scaffold with a 15 minute immersion period in the molten salt bath gives a scaffold which has the desirable properties of the BBG scaffold, such as bioactivity, whilst improving others, such as slightly the mechanical competence, allowing cell attachment and causing an antibacterial effect on some bacteria types. The copper-doped and silver/copper-doped scaffolds

260 exhibited encouraging results compared to the BBG scaffold by improving the compressional strength, whilst not compromising the bioactivity and having suitable antibacterial properties. However an unacceptable cytotoxic reaction was detected when the scaffolds were exposed to rMSC’s and hMSC’s. The copper-doped and silver/copper-doped scaffolds both require further refinement in the MIB concentration before they are further considered for in vivo studies.

7.3 Polymer coated scaffolds

This work has established that gelatin (of both type A and B) can be used to coat the BBG scaffolds to improve the stability and mechanical competence. The use of gelatin as a novel natural polymer coating agent compared with the established synthetic polymer (e.g. PDLLA) coated scaffolds gave the following outcomes:-

 Through a preliminary porosity study it was established that gelatin coating concentrations of 5wt% and 6wt% for both gelatin types gave the best trade-off between porosity and a suitable continuous homogeneous coating  It is possible to create a multi-layered coating on the BBG scaffolds that did not compromise the porosity significantly  Fraying of the polymers was observed at the edges of cracks during fracture which indicates that the polymer is capable of bridging cracks thus keeping the structure together, which is a desirable property to increase the structural integrity of scaffolds  After immersion in SBF, the presence of HA on the scaffolds struts was also detected, which indicated that the presence of the polymer coating did not hinder the precipitation of HA and therefore did not have a detrimental effect on the BBG scaffolds bioactivity  The polymer coating significantly changed the surface topography of the BBG scaffolds by creating a smooth, relatively continuous, and homogeneous layer on the scaffold. The contact angles for the polymer coated scaffolds were higher than those observed for the BBG scaffolds, with the 6wt% gelatin (of both types) and PDLLA coated scaffolds falling within the ideal contact angle range for promoting cell attachment. It could not be proved whether these values would cause a positive reaction when seeded with cells due to the change in surface roughness  The compressional strength was significantly improved by the presence of the gelatin coating compared to both the BBG scaffolds and the PDLLA coated scaffolds. The PDLLA coated scaffolds improved the compressional strength of the BBG scaffolds by 14% and 1% for pre and post SBF immersion, respectively, and the gelatin coated scaffolds improved the

261

compressional strength of the BBG scaffolds by 87% and 57% for pre and post SBF immersion, respectively

Overall the polymer coated scaffolds exhibited superior mechanical competence in comparison to BBG scaffolds without compromising bioactivity. There was little difference between the PDLLA coated scaffolds and those coated in gelatin except that gelatin coated scaffold had a far higher, and therefore more desirable, improvement in the compressional strength. There was little difference between the two types of gelatin or the two concentrations used. Further work is required to determine the most suitable gelatin coating which will provide the best balance between improvements in mechanical competence by keeping the highest possible bioactivity.

7.4 Composite scaffolds

This work has established that composite scaffolds can be produced by incorporating therapeutic ions through MSIE and then coating the ion-doped scaffolds in gelatin. The following results were obtained:-

 The porosity was approximately 20% lower than that of the BBG scaffolds, in line with the porosities observed for the ion-doped scaffolds, nevertheless the porosity was still in the desirable range. After SBF immersion the ion-doped gelatin coated scaffolds’ porosity dropped by approximately 10% which is the same observation made for the BBG scaffolds  Structurally these ion-doped gelatin coated scaffolds exhibited the same characteristics as their ion-doped and gelatin coated counter parts and the BBG scaffolds. Of importance, the presence of HA after SBF immersion was confirmed indicating that the bioactivity of the BBG scaffold has not been compromised by either the MSIE process or the gelatin coating  The surface topography was changed compared to the BBG scaffolds which is because the MSIE process increased the surface roughness and the gelatin coating formed a smooth and continuous layer over the top. This surface modification increased the contact angles measured into the suggested ideal contact angle required for cell attachment. It can not be concluded that the change in surface roughness will have an effect on the cells behaviour due to the change in surface roughness  The compressional strength was significantly improved compared to the BBG scaffolds with an increase of 102% and 58% pre and post SBF immersion, respectively. There is also improvement over the ion-doped scaffolds and the gelatin coated scaffolds indicating that the combination of MSIE processing and polymer coating is very convenient to improve on the mechanical competence of the BBG scaffolds

262

 It should be noted that as with the scaffolds that were doped but not coated with a polymer, it can not be stated that the ions have entered the silicate matrix without further investigation, which is also discussed in chapter eight. The detected ions could be from any remaining remnants of the salt that may have infiltrated the micro-cracks present on the scaffolds surface due to the sintering process

Overall combining the MSIE process with gelatin coating improved upon the BBG scaffold by increasing the compressive strength without compromising the desirable intrinsic properties of the BBG scaffold. The next step would be to investigate the release rate and release mechanism of the added ions and the degradation rate of the gelatin before completing cell biology work on the ion-doped gelatin coated scaffolds. Indeed the control of ion release by the incorporation of gelatin coatings of different thickness is an intriguing possibility for future studies, as discussed in the next chapter.

7.5 Summary

It is hoped that this work will make a significant contribution to the field of BTE using bioactive glass scaffolds as suggested in Figure 138. This work has aimed to fill the gaps in existing knowledge through the production, optimisation and characterisation of these novel scaffolds and this contributes to the emerging field of multifunctional scaffolds for BTE.

263

Figure 138:- Author’s assessment of the contribution to the field of BTE based on the results of the present work

264

Chapter Eight

8. Future work

This chapter discusses future work that is suggested to be carried out based on the results and conclusion drawn from this project. This includes suitable improvements to the scaffolds used and the application of complementary characterisation experiments that could be carried out to gain further understanding of the samples behaviour under different conditions. This section also shows some of the preliminary work that has been already carried out to assess the scaffolds feasibility in vivo in future projects based on expanding the use of the BBG scaffolds made by the foam replication technique.

8.1 Expanding the use of the foam replication technique

Throughout this work the technology on which scaffold development was based has been the simple, cost effective and reproducible foam replication technique. This traditional scaffold fabrication method is always evolving in order to produce improved scaffolds for BTE. Significant work has already been carried out defining the processing parameters required to use when the starting material used is the standard 45S5 Bioglass® powder, as has been shown in previous chapters. The next possible improvement step for this technique would be to consider reproducible industrial-scale fabrication for commercial uses which should lead to high quality scaffolds that can be tailored to fit individual patients specific bone defect. This technique so far has only been utilised on a relatively small scale and in order to be considered for use in industry it needs to be scaled up for manufacturing. The key advantage of this technique over other scaffold manufacturing methods is that it is very straightforward and reproducible without the requirement of complex and expensive processing equipment. It is, in principle, easy to produce high quality scaffolds of identical 3D, interconnected pore structure and specific dimensions.

The other avenue for future investigations is to use other bioactive glass compositions and to tailor the processing parameters for those compositions. Using this established technique to produce novel scaffolds with alternative bioactive glass compositions gives a base for comparison with the BBG scaffolds developed in this study. The foam replication technique has been used to produce scaffolds using for example borosilicate glass scaffolds (479) (CaO-Na2O-MgO-K2O-SiO2-B2O3-P2O5), calcium

phosphate- based glass-ceramic scaffolds (480) (CaO-P2O5-Na2O-trace element oxide) and silicate phosphate- based glass-ceramic scaffolds (411) (SiO2-P2O5-CaO-MgO-Na2O-K2O). This technique can

265 be expanded into other novel biomaterials, such as titanium phosphate glass-ceramics (481) (P2O5-CaO-

Na2O-TiO2) and magnesium oxide containing glass-ceramics (482) (SiO2-CaO-MgO-P2O5), novel mesoporous glasses (483) (484) (485) or exploring “older” materials in a new light such as HA based scaffolds. (486)

However the performance of bioactive scaffolds fabricated by the foam replica method should be compared, specifically in vivo, with scaffolds of similar chemical composition but obtained by alternative methods, such as sol-gel, (244) (487) (201) freeze drying, (488) (489) (490) evaporation- induced self-assembly (484) (491) (492) and rapid prototyping. (484) (106) (493)

Another possible expansion of the foam replica method to produce improved scaffolds is the application of alternative sacrificial porous templates to create scaffolds with different pore structure, for example using plant-derived sponges such as flax scaffolds, (494) wood-derived scaffolds (495) and herbal scaffolds which incorporated the plant Cissus quadrangularis. (496)

8.2 The use of molten salt ion exchange to introduce multifunctional ions into scaffolds

This work has started to explore the use of MSIE (113) (497) (6) as a technique to introduce therapeutic metal ions into prefabricated BBG scaffolds and has examined two strong candidates namely silver and copper. Future work could take a number of directions including: -

I). The continuation of this project to find the most suitable processing parameters for the dual doping with silver and copper II). Choosing other candidate ions, for example zinc (498) and cobalt, (499) to be introduced using MSIE III). Making the MSIE process suitable for mass producing samples.

With respect to the MSIE work carried out in this project further refinement of the parameters used needs to be carried out in order to obtain samples that are suitable for use in BTE. The concentrations studied for the silver MSIE seem to require only a minor further optimisation effort as the biological studies showed that the lowest concentration at the lowest immersion period gave silver doped samples that were not toxic to the cells, showed the formation of HA on their surface during SBF immersion and exhibited improved mechanical competence. The copper and silver/copper ion doped samples still require refinement before they will be suitable for in vivo studies. The concentration of copper needs to be reduced from a salt bath ratio of 2000:1 and 20000:1, stated in sections 4.2.2.II and 6.3.4, respectively, to values ranging between 50,000:1 and 100,000:1 of sodium nitrate to copper

266 nitrate in the salt bath, as the copper was found to be toxic at the current levels. It should be also investigated if in combination with other ions a suitable copper level will be obtained.

In addition, further research and refinement need to be carried out in order to enhance the MSIE technique for this application to produce samples in a more controllable manner. Even though the depth EDX studies, as described in section 4.3.11 with the results presented in sections 5.3, 5.4 and 5.5, showed an average ion penetration of 8nm, 12nm and 9-10nm for silver, copper and silver/copper doped samples, respectively, the lack of homogeneity of the ion penetration depth is a problem that needs to be addressed.

8.2.1 Other candidates for molten salt ion exchange processing

As mentioned in the last section one of the next steps for the further development of MSIE scaffolds would be to investigate other therapeutic metal ions that could be introduced into the prefabricated BBG scaffolds. In chapter two criteria for candidate ions were set out and this should be followed when selecting other ions to use. There are a number of suitable candidates that should be considered and these are shown in Table 50, which also includes the most relevant functions of those ions. Suitable review articles are available (Hoppe et al, (12) Mourino et al (202) and Boccaccini et al (134)), which provide useful information to guide the design of ion doped scaffolds for BTE.

All ions shown in Table 50, for example, have the potential of being introduced using MSIE albeit not necessarily using the metal nitrate salt, as was used during this work, but using metal chlorides, for example, instead. These different metal ions would provide functional alternatives to the silver and copper ions that were investigated during this work. To our knowledge, this work was the first work that successfully used MSIE technique to introduce simultaneously two therapeutic metal ions into a prefabricated scaffold. The next steps would be to further refine the processing parameters to optimise the production of silver and copper doped samples. This could lead to expanding the technique for wide spread use in industry and to investigate different ions in different combinations. Clearly, the potential and limitations of the MSIE technique to incorporate all or some of the ions listed in Table 50 remain to be investigated and this opens avenues for future interesting research.

267

Table 50:-Relevant metallic ion candidates for MSIE scaffolds

Ion Functions of interest Reference Cobalt Stimulating red blood cell production (500) (501) (502) Increases the expression of pro-angiogenic growth factors (503) (499) Gallium Antibacterial (500) (504) (505) Inhibits bone resorption (506) (507) (508) Treatment of hypercalcaemia associated with tumours in bone (509) (510) (511) Iron Promotion of cell attachment and differentiation (500) (512) (513) Involved in redox reactions of oxygen carrier proteins (514) (515) Magnesium Interacts with phosphate ions (498) (480) (500) Osteoconductive (516) (517) (518) Stimulates the adhesion of osteoblasts (519) (520) (521) Manganese Cofactor for a range of important enzymes (500) (522) (523) Detoxification of free radicals Reduces bone resorption (524) (525) (526) Potassium Increases cell proliferation Strontium Stimulates bone formation (498) (480) (500) Reduction of bone resorption (511) (527) (528) Incrementation of osteoblast differentiation (529) (530) (531) Titanium Osteoconductive (532) (533) Antibacterial (534) (481) Vanadium Promotes mineralization (500) (535) (536) Promotes osteoblast proliferation and differentiation (537) (538) (539) Increases collagen synthesis (540) (541) (542) Inhibits alkaline phosphatise associated osteoblast differentiation (543) (544) Zinc Antibacterial (498) (480) (500) Anti-inflammatory (501) (503) (545) Associated with bone growth hormone (546) (547) (548) Enhances osteoblast differentiation (549) (550) (551) Improves wound healing (552) (553) (554) Increases Runt-related transcription factor 2 (555) (556)

8.3 The use of polymers in combination with bioactive glass scaffolds

The main focus of applying polymer coatings on a BBG scaffold is to improve the scaffold’s mechanical competence as low compressive strength is one of the main disadvantages that BBG scaffolds exhibit, and which must be tackled in order to be realistically considered in in vivo and clinical studies. Any additional properties or functionalities that a polymer coating can provide represent extra advantages of the approach.

The literature review, chapter two, reported on using biodegradable, synthetic polymers derived from a number of sources, including poly(L-lactide-co-D,L-lactide) (PLDLLA), (109) poly(lactide-co- glycolide) (PLGA), (557) poly(DL-lactide) (PDLLA) (558) and poly(3-hydroxybutyrate) (P3HB). (486) These polymers exhibit a range of desirable properties along with improving the mechanical

268 competence of the scaffold but some of the newer novel polymers are difficult to produce and expensive at the current time. (559) (560) (561).

A suitable alternative is to use natural polymers that have not been used extensively to develop coatings on BTE bioceramic or bioactive glass scaffolds. These polymers include gelatin, (562) chitosan, (563) collagen (564) and alginate. (565) As this project has shown, gelatin is a suitable polymer for this application and the increasing literature being published on this topic indicates that there is now a sway towards using some of these cost-effective and highly biocompatible polymers. Another advantage of using natural polymers is that they have in most cases a proven biological track record and the fundamental knowledge on their behaviour in biomedical applications already exists.

Chitosan and alginate are two examples of “older” polymers that have been revisited to be considered as coatings for three dimensional scaffolds and the research community is starting to optimise these polymers for this use. (566) (567) (568) Chitosan in combination with poly(epsilon- caprolactone) (PCL), for example, has shown enhanced cell differentiation (569) in osteoblasts and heparin-functionalized chitosan scaffolds have shown to increase the viability and differentiation of preosteoblasts (570) as well as increasing the mechanical competence of the scaffold in question. (563) (571) (266) Alginate in combination with HA has shown suitable cell attachment properties (572) and a composite of calcium phosphate and alginate has shown osteogenic differentiation (573) whilst once again maintaining suitable mechanical properties. (574) (573) Other significant function of polymer coatings has been considered, which is their use as a drug delivery vehicle, including novel copper ion releasing composite scaffolds containing alginate, (575) calcium phosphate/chitosan composite scaffolds for antibiotic delivery, (576) growth factor delivery by both chitosan and alginate composite materials (577) (578) (579) and novel gallium cross-linked with alginate composite scaffolds. (580) (581) Indeed the field of natural polymer coated bioceramics and bioactive glass scaffolds is receiving an increasing amount of attention, in this context, the present study has contributed knowledge and expertise on the production of the gelatin/Bioglass® system

8.4 Gelatin as a drug delivery vehicle

As discussed in previous sections, the main purpose for the gelatin coating in the present project was to improve the mechanical properties of the BBG scaffolds. After this has been achieved, additional functions of the gelatin coating can be investigated. For example it is interesting to investigate how gelatin coating can be used as a functional drug delivery system. Indeed incorporating drug delivery carriers in the form of a gelatin coating will give the BBG scaffolds another function or local drug

269 delivery device. (500) This will involve introducing a range of additives including growth factors (582) and/or therapeutic drugs, including antibiotics. (583)

8.4.1 Introducing antibiotics and growth factors into gelatin coatings

Introducing a range of functional additives to a polymer is not a novel concept in itself and even introducing additive load polymers onto three dimensional scaffolds has also been investigated. The particular novel concept that it is hoped to expand upon the basis of the present results is the use of an additive loaded gelatin coating on ion-doped BBG scaffolds. A range of additives can be added to the gelatin coating to make the composite scaffold more functional.

Antibiotics are obvious drugs to be introduced into the gelatin composite scaffold to combat possible post-operative infections. Antibiotics, however pose certain challenges when introduced into a system including dosing (584) (585) and degradation issues. (586) (587) Numerous antibiotics have been used in polymer coatings including tetracycline, (588) gentamicin (589) and vancomycin. (583) In addition to some previous work that has been carried out with gelatin loaded with vancomycin, (583) doxorubicin (590) and colistin. (591) One of the main issues is to keep the antibiotic stable before the gelatin degrades and this can only be solved by choosing a suitable antibiotic dose and gelatin concentration. (592) (593) The approach based on therapeutic ion releasing scaffolds presented here, would be radically innovative, as it would combine ion releasing capability with antibiotic release from the degrading gelatin coating.

Growth factors can be incorporated into polymer coatings to promote bone formation (594) (595) and angiogenesis. (596) (597) (598) For example, PDLLA with insulin-like growth factor-1 and transforming growth factor-beta 1, (356) alginate with angiogenic growth factors (177) and P3HB with insulin like growth factors, myostatin and vascular endothelial factors (VEGF) (599) and in gelatin loaded with growth factor-beta 1, (600) fibroblast growth factor 2 (582) and myoglobin (601) have been investigated. The importance of introducing these additives into a biological system during implantation is that they provide extra signalling molecules to the surrounding cells in order to induce a certain function. For example, introducing additional VEGF into a bone scaffold will encourage faster angiogenesis of the surrounding tissue which in turn accelerates the bone healing process.

There are numerous challenges remaining to create composite systems incorporating additive loaded gelatin coatings on three dimensional scaffolds. However this is a rapidly changing field, being investigated extensively and further developments of gelatin coated scaffolds incorporating antibiotics

270 and growth factors is expected and these developments will benefit from the results presented in this thesis.

8.4.2 Therapeutic metal nano-particles – Copper oxide and Iron oxide

The incorporation of metallic nano-particles can be an interesting further aspect for investigation aiming at enhancing the performance of bioceramic and bioactive glass scaffolds including imparting specific electromagnetic properties. (602) (603) (604) (605) Preliminary work was undertaken in the course of this thesis to introduce therapeutic metal nano-particles into both gelatin types on model systems for in vitro studies. Following the processing technique described in section 4.4, but after the gelatin is thoroughly mixed, the metal nano-particles are added to the beaker and the mixture is sonicated to evenly disperse the metal nano-particles throughout the gelatin. The BBG scaffolds are then coated by the gelatin/metal nano-particle composite, the coated samples are left to dry and characterisation studies can then be carried out. The preliminary study carried out included the scaffolds characterisation by SEM, porosity measurement, mechanical properties determination and FTIR studies. The porosity of these scaffolds were comparable with the gelatin only coated scaffolds with a porosity ranging between 77% and 86% and the mechanical competence of these scaffolds was also in the same range as the gelatin only composite scaffolds (approximately 1.30 ±0.05MPa). Figure 139 shows the slight change in colour of the scaffolds after the addition of the metal nano-particles to the gelatin coatings. SEM images, shown in appendix D.3, indicate little visible difference between these scaffolds and those shown in chapter five. Moreover the FTIR results, shown in appendix D.3, mirror the FTIR spectra shown in chapter five with the two amide peaks which are attributed to gelatin, (457) (191) the peaks due to water (457) (191) and the previously identified peaks of the BBG scaffold. It shows no unidentifiable peaks, suggesting that the metal nano-particles were not detected by the FTIR.

Additional characterisation studies, including in vitro studies, should be the next step for these composite scaffolds to determine their suitability for use in BTE. In particular the degradation of the polymer and the release of the metal nano-particles should be extensively studied given the possible cytotoxic effect of released nano-particles. (606) (607) (608)

271

Figure 139:- Metal nano-particle loaded gelatin coated BBG scaffolds: A) gelatin type A coated scaffold, B) gelatin type A coated scaffold loaded with copper oxide nano-particles, C) gelatin type A coated scaffold loaded with iron oxide nano-particles, D) gelatin type B coated scaffold, E) gelatin type B coated scaffold loaded with copper oxide nano-particles, and F) gelatin type B coated scaffold loaded with iron oxide nano-particles.

8.5 Combining molten salt ion exchange scaffolds with polymer coatings

In this work the preliminary studies of combining MSIE scaffolds with polymer coatings have been carried out, proposing these composite scaffolds as promising multifunctional structures for BTE. However these scaffolds are still in the concept stage and any future work will need to characterise, refine and improve these samples before they can be investigated in vivo. Only structural characterisation of these scaffolds has been carried out in this work, the next stage is to carry out initial viability in vitro studies to understand the response of specific relevant cells, such as the HPLSC’s and rMSC’s used in chapter six, to these rather promising but complex scaffolds. Since this work contained the preliminary work on combining ion-doped scaffolds with polymer coatings the suggested route would be for further studies to build upon the results achieved here. One specific experiment that needs to be carried out first is to ascertain the degradation and release mechanisms of the polymer and ions, respectively, as this knowledge will give initial information about the behaviour of the composite scaffold in vitro and eventually in vivo.

This recommendation for future research is also relevant when additives (e.g. bio-molecules or drugs (609) (610) (611)) are added to the gelatin coatings, where it is important to consider the biological and structural performance of the scaffolds containing and releasing these additional factors. There is a great deal of scope for the use of gelatin on ion-doped bioactive glass scaffolds as this is a novel concept for developing advanced scaffolds for BTE and the system represents an interesting field for future research.

272

8.6 Issues brought forward by this work

In the course of this work a number of issues have risen which need to be looked into in greater detail in future investigations. In particular the issue of the pH change due to the presence of the BBG scaffold and the possible changes to the samples due to the sterilisation techniques used as two relevant aspects requiring special attention in future research involving BBG scaffolds.

8.6.1 pH issue with the base scaffolds

As discussed in chapter 6, there is no set protocol that can be adopted to reduce the effect of the change pH occurring due to the presence of the base scaffold. This effect needs to be addressed and more research needs to be carried out to find a suitable protocol for 45S5 Bioglass® -derived glass- ceramic scaffolds. Currently each laboratory has its own method of reducing the pH issue caused by 45S5 Bioglass® -derived glass-ceramic scaffolds. The difference in these methods could account for the different conclusions observed in the literature highlighting the need for a standardised protocol for cell culture studies on 45S5 Bioglass® -derived glass-ceramic scaffolds. A review of the available protocols in the literature needs to be undertaken and a consensus protocol needs to be developed. The work shown in chapter 6 on the pre-conditioning of the scaffolds in order to reduce the cell toxicity due to pH change represents a convenient start point

8.6.2 Sterilisation

As discussed in chapter 6, the sterilisation process of the scaffolds will have an effect on the physical and chemical properties of the sample surface which will possibly impact the results. The surface changes induced by different sterilization techniques need to be discussed and reviewed as this could relate to a difference between different cell biology studies on 45S5 Bioglass® -derived glass- ceramics scaffolds in the literature. It should be noted that surface changes due to the sterilization process on glass-ceramic scaffolds will be relatively small due to their robust inorganic nature but on materials such as polymers the sterilisation technique could change the very nature of the polymer which would have an impact on the results. This effect will be relevant for example in polymer coated BBG scaffolds such as the gelatin coated samples developed in this project. A review of the current processes and results needs to be undertaken to produce a protocol for which future work can be based on in order to achieve reproducible and comparable results between different laboratories.

273

8.7 Further structural characterisation of the scaffolds

In general further structural characterisation will need to be carried out on the modified scaffolds to gain knowledge on their overall behaviour targeting in vivo applications. One suggestion for future research is to confirm the statement made in chapter five in that the improved mechanical competence of MSIE scaffolds was due to the change in atomic structure of the silicate network. (438) (113) Further investigation to confirm this statement could be explored through extended x-ray absorption fine structure (EXAFS) data in combination with Reverse Monte Carlo (RMC) modelling as this study will give an answer as to why the mechanical properties have been improved through molten salt ion exchange and this is discussed further in section 8.7.1.

Other techniques should be used to further explore the scaffolds developed in this work such as Raman spectroscopy and optical emissions work to further characterise the microstructure of the samples in particular the dynamics of the introduced metal ions in the silicate structure of crystallised Bioglass®. (612) In addition NMR spectroscopy (613) (614) will be useful to further investigate the atomic structure of the developed scaffolds.

8.7.1 Molecular modelling using the Reverse Monte Carlo method

As suggested in chapter five, the addition of the larger network modifiers ionically bonded to non-bridging oxygen in the silica network will also interact with the nearby bridging oxygen to form an additional two-coordinate structure. (438) (113) These additional structures could be the reason behind the increase in mechanical strength, as discussed in chapter five. In order to ascertain if this effect is actually occurring in the MSIE material, additional experiments need to be carried out and then used as a basis for a modelling platform to give a pictorial view of the atomic structure of the MSIE material.

Reverse Monte Carlo (RMC) modelling is specifically used to model amorphous and disorganised crystalline structures. (615) Using x-ray absorption spectroscopy (XAS) specifically EXAFS in combination with a RMC modelling platform a three dimensional model can be produced, whilst being subject to certain constraints consistent with available experimental data. (615) A more detailed explanation of how RMC modelling achieves the required three-dimensional model can be found in appendix D.1. Once this has been done, suitable structural models will be available that can be used to determine if the addition of the new network modifiers after MSIE is having structural effects that would lead to the increase in the scaffolds mechanical competence. An example of a basic RMC

274 model of silicon oxide is shown in Figure 140. This RMC model realistically represents the structure of the material and indicates how the bonds and atoms interact.

Figure 140:- RMC model of silicon oxide showing the atomic structure as extrapolated from EXAFS data and other constraints (616)

8.8 Further in vitro characterisation

A main objective of this work has been to produce novel scaffolds which are intended for clinical applications in BTE approaches. These scaffolds thus need to go through a series of rigorous in vitro studies in order to comprehensively understand their behaviour in contact with relevant cells. This work investigated preliminary the interaction of the scaffolds with cells and bacteria and following sections will expand on recommendations for further research in this particular topic.

275

8.8.1 Cell culture characterisation

The main focus of future work concerning the in vitro response of the scaffolds should concentrate on determining a concentration of the therapeutic metal ion that is high enough for it to deliver its desirable function, i.e. be antibacterial in the case of the silver ions, whilst still being biocompatible with the cells. Achieving such optimal ion concentrations has not been completely successful as yet, however the scaffold technology developed in this project is suitable to deliver scaffolds for future studies. The next step should be to fine tune the doped MSIE scaffolds to obtain biocompatible samples which cause no toxic effect to the cells. The design of the study should involve changing the ion concentration and then testing it in vitro, the analysis of cytotoxicity will establish the window of concentrations at which the samples are biocompatible. This process needs to be repeated until the sample has demonstrated full biocompatibility with the target cells.

Once the samples have been deemed biocompatible other in vitro tests can be carried out, as was briefly discussed in chapters 4 and 6, including identifying specific osteogenic markers (617) (618) (619) which are indicative of the formation of new bone tissue and, in the case of the copper doped samples, the presence of angiogenic markers. (620) (621) (622)This would involve RNA extraction, (623) (624) PCR testing (625) (626) and carrying out a range of standard histological / immunohistochemical assays. (627) (628) (629) (630)

In the context of the present project, relevant preliminary work was carried out using BBG scaffolds in collaboration with Dr R El-Gendy from Prof X.B Yang’s research group at the Leeds Dental Institute at the University of Leeds (UK). Using human dental pulp stromal cells (HDPSC’s) in combination with BBG scaffolds a promising combination that promotes bone repair and angiogenesis in vitro has been achieved. (631) (225) These studies used the dynamic seeding technique described in section 4.4.2 which is a model that is closer to in vivo implantation than static seeding techniques (including the ones used in this work), as described in sections 4.4.5 and 4.4.6. These studies then assessed the expression of certain osteogenic (COLIA1, ALP, RUNX-2, OC) and angiogenic (CD34, PECAM1 & VEGFR2) genes and found that in general there was a positive effect on the expression of these genes due to the presence of the BBG scaffolds. (631) (225) Clearly, future work is required following the same experiment protocol, but incorporating ion-doped scaffolds, to increase our current understanding of the scaffolds bone regeneration ability and how the release of copper and silver might affect the expression of relevant genes of HDPSC;s.

276

8.8.2 Bacterial characterisation

The present work used three bacterial strains to test the antibacterial potential of the variety of samples produced. As a suggestion for future research, the bacterial work needs to be carried out in conjunction with the in vitro cell culture, which will lead to identify the effectiveness of a particular or anti-bacterial agent, without causing a toxic effect to the surrounding tissue. Fine tuning of the current crop of ion-doped samples needs to be carried out first to achieve this balance. For the silver doped samples, the optimal concentration window has been approached in this study (as described in chapter six).

Continued work using a range of bacteria that are known to cause post-operative infections should be carried out on the evolving ion-doped samples to further evaluate the samples suitability as an antibacterial agent in more relevant conditions. As stated in section 2.13.1 the most common bacteria involved in post-operative infections is S.aureus especially in terms of its resistance to the common antibiotics which are given if an infection occurs. The agar diffusion study carried out using S.aureus was a convenient starting point and it should be continued to obtain the bacteriostatic and bactericidal concentrations of the antibacterial ion as it was done with E.coli in chapter six of this thesis. It can be suggested that this bacteriostatic/bactericidal study should be carried out for each sample type and for each different bacteria tested. Other relevant bacteria should be investigated, including methicillin- resistant Staphylococcus aureus (MRSA), (632) Clostridium difficile (633) and Staphylococcus epidermidis, (634) to produce scaffolds that are effective against the most common bacteria in post- operative infections.

8.9 In vivo characterisation

Preliminary relevant work has been carried out on the BBG scaffolds produced in this study using different animal models. Since these studies were preliminary and only on one type of scaffolds, the summary of achieved results is presented in this section specifically because they show a clear path for future in vivo studies. Indeed, a wide range of relevant in vivo models are available to explore different conditions, including large cortical bone defects, (635) canine segmental defects (636) and subcutaneous rat implantation model. (637) Specific bone vascularisation in vivo models, such as the A-V loop model, are also available (318) and of relevance in the context of developing new scaffolds with vascularisation potential.

277

8.9.1 In vivo studies using 45S5 Bioglass® –derived scaffolds using human dental pulp stromal cells

In collaboration with Dr R El-Gendy from Prof X.B Yang’s research group at the Leeds Dental Institute at the University of Leeds, UK, the BBG scaffolds produced in this study were combined with human dental pulp stromal cells (HDPSC’s) to form a suitable combination intended to promote bone repair and angiogenesis in vivo. (631) (225)

In these studies HDPSC’s were dynamically seeded, as described in section 4.4.2, onto the BBG scaffolds for five days and then statically cultured for a further two days, sealed in Millipore diffusion chambers and implanted into mice for eight weeks. After removal, the samples were subjected to a range of assays to ascertain the osteogenic and angiogenic potential of the scaffolds. It was seen that the scaffolds are biocompatible in vivo, showed suitable cell attachment, formation of neo-tissue structures within the construct, presence of angiogenic gene expression (225) and evidence of osteogenic differentiation. (631) (225) These results are supported by PCR data and a range of histological staining, including evidence of collagen type I which is characteristic of the presence of mature osteoblasts (631) and the presence of PECAM1 and VEGFR2 markers indicating the angiogenic potential of the construct. (225) Organised capillary-like structures were observed after eight weeks of implantation and this along with the presence of PECAM1 and VEGFR2 markers suggest the angiogenic potential of the BBG scaffolds. (225)

The Millipore diffusion chamber provided a suitable bioreactor for the scaffolds for this preliminary work using HDPSC’s as it provides a stable environment for the scaffold whilst enabling a medium for the exchange of fluids, blood, cells and diffusible products from the host animal. The bioreactor supports studies, in vivo, which provide an environment suitable for angiogenesis and bone remodelling like bone defect models. The application of vascularisation models can be considered the next step for this cell type. (631) Indeed, expansion of these studies incorporating ion-doped scaffolds, particularly the copper-doped scaffolds developed in this work should be carried out. Future studies also should investigate BBG scaffolds which incorporate other angiogenic ions such as cobalt. (500) (501) (502)

278

8.9.2 In vivo study using 45S5 Bioglass® –derived scaffolds in an arteriovenous loop model

In collaboration with Dr A Arkudas from the Department of Plastic and Hand Surgery at the University of Erlangen Medical Centre, Germany, BBG scaffolds, produced in this study, were used for the first time in vivo in an arteriovenous (AV) loop model that was implanted into rats to investigate the angiogenic potential of 45S5 Bioglass®. (318) The full explanation of the AV loop model including materials, techniques and characterisation that relates to the recent publication of the results (318) are shown in appendix D.2 with only a summary of the results and concluding remarks being presented here.

Figure 141 shows the AV loop containing pieces of BBG scaffold (between 1x1x1mm3 and 2x2x2mm3 in size) in fibrin gel during the implantation process. Figure 142 shows the construct removed from the Teflon cage after being implanted for three weeks, indicating that the 45S5 Bioglass® pieces showed no outward signs of significant biodegradation. There was some evidence of vascularisation of the matrix and the patency of the AV loop as the colour change to yellow indicating that functional vessels are being filled with the injected Microfil. (318) Figure 142 does not give a definitive answer to whether the presence of the BBG scaffold pieces has a significant effect on the angiogenesis process in the AV loop. This can only be shown through further characterisation, and results are shown in Figure 143 and Figure 144, showing histological staining and micro CT results, respectively.

Figure 141:- The AV loop was directed through a Teflon isolation chamber which was filled with the BBG scaffold pieces and fibrin gel with a fibrinogen concentration of 10mg/ml and a thrombin concentration of 2 I.U./ml. (318) With A) showing the vascular pedicle in place in the cage before being covered in more BBG pieces and fibrin gel and B) showing the Teflon cage before the lid is placed.

279

Figure 142:- Macroscopic appearance of the inside of the Teflon cage, (shown in Figure 141) after three weeks upon explanation. (318)

Figure 143 shows that the tissue formed in the implantation period of three weeks was localised around the vascular pedicle, indicated by the arrows and shown in Figure 141A). Figure 144 shows the micro CT analysis (carried out by Dr T Fey from the Institute of Glass and Ceramics, University of Erlangen-Nuremberg, Germany) of the material removed from the Teflon cage indicating that a significant level of angiogenesis has occurred during the implantation period. This is particularly evident in Figure 144C) where the full vessel network is shown. This result implies at least on a qualitative basis that the newly formed tissue infiltrates the BBG pieces in the construct, thus the presence of the scaffold has a positive effect on the angiogenesis process occurring in vivo. (318)

Figure 143:- The material removed from the Teflon cage after being subjected to hematoxylin and eosin (H&E) staining to show newly formed tissue in the sample. The arrows are indicate the vascular pedicle that passed through the BBG scaffold and fibrin gel sample with a magnification of x25. (318)

280

Figure 144:- Micro CT analysis giving a 3D representation of the newly formed vessels found within the material removed from the Teflon cage showing the final network structure of the vessels including the original vascular pedicle (318) (Micro CT images obtained by Dr T Fey from the Institute of Glass and Ceramics, University of Erlangen-Nuremberg, Germany)

This in vivo work shows for the first time that axial vascularisation of BBG scaffolds can be achieved using the AV loop model in a rat. Through micro CT and histological staining it was confirmed that after three weeks a new vascular network was created and has infiltrated the implanted BBG scaffold pieces. This infiltration and creation of a new network of vessels will allow transplantation of the whole construct using the AV loop model in the future. (318) Based on these first results future work can be suggested, exploiting the versatility of the AV loop model.

For example, a relevant next step would be to introduce the different types of scaffolds examined during the course of this project to ascertain if the presence of ions, in particular copper ions, improves the blood vessel network formed compared to the BBG scaffolds. These studies should only be carried out in the future investigations after the MSIE samples have been completely optimised in vitro to prevent complications in the in vivo work due to possible toxicity caused by the uncontrolled ion release.

281

References

1. C Hu, C Tercero, S Ikeda, M Nakajima, H Tajima, Y Shen, T Fukuda and F Arai. Biodegradable porous sheet-like scaffolds for soft-tissue engineering using a combined particulate leaching of salt particles and magnetic sugar particles. Journal of Biosciences and Bioengineering. 2013, Vol. 116, 1, pp. 126-131.

2. J R Jones. Review of bioactive glasses: From Hench to hybrid. Acta Biomaterialia. 2013, Vol. 9, 1, pp. 4457-4486.

3. L L Hench. The Story of Bioglass. Journal of Material Science: Materials in Medicine. 2006, 17, pp. 967- 978.

4. Q Z Chen, I D Thompson and A R Boccaccini. 45S5 Bioglass -derived glass-ceramic scaffolds for bone tissue engineering. Biomaterials. 2006, 27, pp. 2414-2425.

5. A R Boccaccini, Q Z Chen, L Lefebvre, L Gremillard and J Chevalier. Sintering, crystallisation and biodegradation behaviour of Bioglass(R)-derived glass–ceramics. Faraday Discussions. 2007, Vol. 136, pp. 27- 44.

6. S Di Nunzio, C Vitale Brovarone, S Spriano, D Milanese, E Verne, V Bergo, G Maina and P Spinelli. Silver containing bioactive glasses prepared by molten salt ion exchange. Journal of the European Ceramic Society. 2004, Vol. 24, pp. 2935–2942.

7. D M Yunos, O Bretcanu and A R Boccaccini. Polymer-bioceramic composites for tissue engineering scaffolds. J. Mater Sci. 2008, 43, pp. 4433-4442.

8. A M Mulligan, M Wilson and J C Knowles. The effect of increasing copper content in phosphate-based glasses on biofilms of Streptococcus sanguis. Biomaterials. 2003, Vol. 23, pp. 1797-1807.

9. H Xie and Y J Kang. Role of Copper in Angiogenesis and Its Medicinal Implications. Current Medicinal Chemistry. 2009, Vol. 16, 10, pp. 1304 -1314 .

10. G J Brewer. Copper in medicine. [ed.] J Broderick and D Coucouvanis. Current Opinion in Chemical Biology. 2003, Vol. 7, pp. 207-212.

11. P M Gullino, M Ziche and G Alessandri. Gangliosides, copper ions and angiogenic capacity of adult tissues. Cancer and Metastasis Reviews. 9, 1990, pp. 239-251.

12. A Hoppe, N S Gueldal and A R Boccaccini. A review of the biological responce to ionic dissolution products from bioactive glasses and glass-ceramics. Biomaterials. 2011, Vol. 32, 11, pp. 2757-2774.

13. B P Chan and K W Leong. Scaffolding in tissue engineering: General approaches and tissue-specific considerations. Eur Spine. 2008, Vol. 17, Suppl 4, pp. 467-479.

14. M Bucheler and A Haisch. Tissue Engineering in Otorhinolaryngology. DNA and Cell Biology. 2003, Vol. 22, 9, pp. 549-564.

15. L D K Buttery and A E Bishop. Chapter 18 - Introduction to Tissue Engineering. [ed.] J R Jones and L L Hench. Biomaterials, Artificial Organs and Tissue Engineering. s.l. : Woodhead Publishing Limited, 2005, 18, pp. 193-200.

16. J J Mao, G Vunjak Novakovic, A G Mikos and A Atala. Translational Approaches in Tissue Engineering and Regenerative Medicine. [ed.] G Vunjak Novakovic, A G Mikos and A Atala J J Mao. 2008. p. 504.

282

17. G E Muschler, C Nakamoto and L G Griffith. Engineering principles of clinical cell-based tissue engineering. Journal of Bone and Joint Surgery - American Volume. 2004, Vol. 86A, 7, pp. 1541-1558.

18. B Denecke, M Woeltje, S Neuss and W Jahnen-Dechent. Tissue engineering - Combinig cells and biomaterials into functional tissues. [ed.] G M Artmann and A Chien. Bioengineering in Cell and Tissue Research. 2008, pp. 193-214.

19. W W Minuth, R Strehl and K Schumacher. Tissue Engineering: From cell biology to artifical organs. s.l. : Wiley-VCH, 2006. p. 324.

20. S J Barnes and L P Harris. Tissue engineering: Roles, materials and applications. s.l. : NOVA, 2008. p. 323.

21. S Fawzi-Grancher, N De Isla, G Faure, J F Stoltz and S Muller. Optimisation of biochemical condition and substrates in vitro for tissue engineering of ligament. Annals of Biomedical Engineering. 2006, Vol. 34, 11, pp. 1767-1777.

22. F H Silver. Biomaterials, Medicine and Tissue engineering: An integrated approach. s.l. : Chapman and Hall, 1994.

23. C T Laurencin, A M A Ambrosio, M D Borden and J A Copper Jr. Tissue Engineering: Orthopedic Applications. Annu Rev Biomed End. 1, 1999, pp. 19-46.

24. M R Rogel, H Qiu and G A Ameer. The role of nanocomposites in bone regeneration. J Mater. Chem. 2008, Vol. 18, 36, pp. 4233-4241.

25. Y Ikada. Challenges in Tissue Engineering. J. R. Soc. Interface. 2006, 3, pp. 589-601.

26. Q Z Chen, S E Harding, N N Ali, H Jawad and A R Boccaccini. Cardiac Tissue engineering. [ed.] A R Boccaccini and J E Gough. Tissue engineering using ceramics and polymers. Cambridge : Woodhead Publishing Limited, 2007, 16, pp. 335-356.

27. J E Gough. Cartilage Tissue Engineering. [ed.] A R Boccaccini and J E Gough. Tissue engineering using ceramics and polymers. Cambridge : Woodhead Publishing Limited, 2007, 26, pp. 566-586.

28. A E Bishop and H J Rippon. Lung Tissue Engineering. [ed.] A R Boccaccini and J E Gough. Tissue engineering using ceramics and polymers. Cambridge : Woodhead Publishing Limited, 2007, 23, pp. 497-507.

29. K Shakesheff. Liver Tissue Engineering. [ed.] A R Boccaccini and J E Gough. Tissue engineering using ceramics and polymers. Cambridge : Woodhead Publishing Limited, 2007, 19, pp. 404-420.

30. D A J Lloyd and S M Gabe. Intestine Tissue Engineering. [ed.] A R Boccaccini and J E Gough. Tissue engineering using ceramics and polymers. Cambridge : Woodhead Publishing Limited, 2007, 24, pp. 508-528.

31. P Kingham and G Terenghi. Nerve Tissue engineering. [ed.] A R Boccaccini and J E Gough. Tissue engineering using ceramics and polymers. Cambridge : Woodhead Publishing Limited, 2007, 22, pp. 466-496.

32. J C Earthman, C G Sheets, J M Paquette, R M Kaminishi, W P Nordland, R G Keim and J C Wu. Tissue engineering in dentistry. Clinics in Plastic surgery. 2003, Vol. 30, 4, pp. 621-639.

33. A Atala. Engineering of tissues and organs. [ed.] A R Boccaccini and J E Gough. Tissue engineering using ceramics and polymers. Cambridge : Woodhead Publishing limited, 2007, 13, pp. 269-293.

34. Formed 3D Bio-Scaffolds via rapid prototyping technology. P S Maher, R P Keatch, K Donnelly and J Z Paxton. [ed.] P Verdonck, M Nyssen and J Haueisen J Vander Stoten. 2009. 4TH EUROPEAN CONFERENCE OF THE INTERNATIONAL FEDERATION FOR MEDICAL AND BIOLOGICAL ENGINEERING. Vol. 22, pp. 2200-2204.

283

35. Y Eitan, U Sarig, N Dahan and M Machluf. Acellular cardiac extracellular matrix tyas a scaffolds for tissue engineering: In vitro cell support, remodelling and biocompatibili. Tissue engineering Part C - Methods. 2010, Vol. 16, 4, pp. 671-683.

36. F Chen, J J Yoo and A Atala. Acellular collagen matrix as a possible "off the shelf" biomaterial for urethral repair. Urology. 1998, Vol. 54, 3, pp. 407-410.

37. S E Dahms, H J Piechota, R Dahiya, T F Lue and E A Tanagho. Composition and biomechanical properties of bladder acellular matrix graft: comparative analysis in rat, pig and human. British journal of urolgy. 1998, Vol. 82, 3, pp. 411-419.

38. J J Yoo, J Meng, F Oberpenning and A Atala. Bladder augmentation using allogenic bladder submucosa seeded with cells. Urology. 1998, Vol. 51, 2, pp. 221-225.

39. Q Z Chen, S E Harding, N N Ali, A R Lyon and A R Boccaccini. Biomaterials in Cardiac Tissue engineering: Ten years of research survey. Materials Science and Engineering R. 2008, 59, pp. 1-37.

40. R Dobson. Number on UK Transplant waiting list reaches new high. BMJ. 2007, 334, pp. 920-921.

41. Transplants, Statistics and Audit Directorate for UK. Transplant Activity in the UK. s.l. : Statistics and Audit Directorate for UK Transplants, 2008.

42. D L Butler, S A Goldstein and F Guilak. Functional Tissue Engineering: The role of Biomechanics. J. Biomed. Eng. 2000, 122, pp. 570-575.

43. P Jevon. Care of the dying and deceased patient: A practical guide for nurses. s.l. : J Wiley and Son, 2009. p. 264.

44. A J Cronin. Transplants save lives, defending the double veto does not: a reply to Wilkinson. Journal of Medicial Ethics. 2007, Vol. 33, 4, pp. 219-220.

45. NHS Blood and Transplant. Transplant activity in the UK: Activity report 2009/2010. s.l. : NHS, 2010.

46. U Kneaser, D J Schaefer, B Munder, C Klemt, C Andree and G B Stark. Tissue Engineering of Bone. Min Invas Ther & Allied Technol. 2002, Vol. 11, 3, pp. 107-116.

47. C Mason, J F Markusen, M A Town, P Dunnill and R K Wang. The potential of optical coherence tomography in the engineering of living tissues. Physics in Medicine and Biology. 2004, Vol. 49, 7.

48. Y Tabata. The importance of drug delivery systems in tissue engineering. Pharmaceutical Science and Technology Today. 2000, Vol. 3, 1, pp. 80-89.

49. A J Kolber. A matter of priority: Transplanting organs preferentially to registered donors. Rutgers Law Review. 2003, Vol. 55, 3.

50. C Quint, Y Kondo, R J Manson, J H Lawson, A Dardik and L E Niklason. Decullularized tissue- engineered blood vessel as an arterial conduit. Proceedings of the National Academy of Sciences of the United States of America. 2011, Vol. 108, 22, pp. 9214-9219.

51. G A Silva, O P Coutinho, P Ducheyne and R L Reis. Materials in particular for Tissue Engineering. 2. Applications to Bone. J Tissue Eng Regen Med. 2007, 1, pp. 97-109.

52. D Noel, F Djouad and C Jorgenese. Regenerative medicine through mesenchymal stem cells for bone and cartilage repair. Current Opinion in Investigational Drugs. 2002, Vol. 3, 7, pp. 1000-1004.

284

53. J L Dragoo, B Samimi, M Zhu, S L Hame, B J Thomas, J R Liberman, M H Hedrick and P Benhaim. Tissue-engineered cartilage and bone using stem cells from human infrapatellar fat pads. Journal of Bone Joint Surgery. 2003, Vol. 85B, 5, pp. 740-747.

54. L A Deane Fattone, R Strauss and J Bruno. The management of periodontal disease in patients who have recieved radiation therapy for head and neck cancer. Special Care in Dentistry. 1987, Vol. 7, 3, pp. 120-123.

55. M A W Merkx, J C Maltha, H-P M Freihofer and A M Kuijpers-Jagtman. Incorporation of particulated bone implants into the facial skeleton. Biomaterials. 1999, Vol. 20, 21, pp. 2029-2035.

56. M Frohlich, W L Grayson, L Q Wan, D Marott, M Drobnic and G Vunjak-Novakovic. Tissue engineered bone grafts: Biological requirements, Tissue culture and Clinical relevance. Current Stem Cell Research Theory. 2008, Vol. 3, 4, pp. 254-264.

57. B J R F Bolland, J M Kanczler, D G Dunlop and R O C Oreffo. Development of in vivo microCT evaluation of neovascularisation in tissue engineered bone construct. Bone. 2008, Vol. 43, 1, pp. 195-202.

58. C S Bartlett, D L Helfet, M R Hausman and E Strauss. Ballistics and gunshot wounds: Effects on musculoskeletal tissues. Journal of American Academy of Orthopaedic Surgeons. 2000, Vol. 8, 1, pp. 21-36.

59. I D Xynos, M V J Hukkanen, J J Batten, L D K Buttery, L L Hench and J M Polak. 45S5 Bioglass(r) stimulates osteoblast turnover and enhances bone formation in vitro: Implications and Applications for Bone Tissue Engineering. Calcif Tissue Int. 2000, 67, pp. 321-329.

60. A O Mekhail, E Abraham, B Gruber and M Gonzalez. Bone transport in the management of posttraumatic bone defects in the lower extremity. Journal of Trauma-Injury, Infection and Critical Care. 2004, Vol. 56, 2, pp. 368-378.

61. D J Hak. The use of osteoconductive bone graft substitutes in orthopaedic trauma. Journal of American Academy of Orthopaedic Surgeons. 2007, Vol. 15, 9, pp. 525-536.

62. A H Gomoll, H Madry, G Knutsen, N van Dijk, R Seil, M Brittberg and E Kon. The subchondral bone in articular cartilage repair: current problems in surgical management. Knee surgery, Sport Traumatology and Arthroscopy. 2010, Vol. 18, 4, pp. 434-447.

63. I Drosse, E Volkmer, R Capanna, P De Biase, W Mutschler and M Schieker. Tissue Engineering for bone defect healing: An update on a multi-component approach. J. Care Injured. 2008, Vol. 39, 2, pp. 9-20.

64. E Wintermantel, J Mayer, J Blum, K-L Eckert, P Luscher and M Mathey. Tissue engineering scaffolds using superstructures. Biomaterials. 1996, Vol. 17, 2, pp. 83-91.

65. P Wozniak and A J El Haj. Bone regeneration and repair using tissue engineering. [ed.] A R Boccaccini and J E Gough. Tissue engineering using ceramics and polymers. Cambridge : Woodhead Publishing limited, 2007, 14, pp. 294-318.

66. R S Behnke. Kinetic Anatomy. s.l. : Human Kinetics, 2006. p. 268.

67. E T Howley and D L Thompson. Fitness Professional's Handbook. 6th edition. s.l. : Human Kinetics, 2012. p. 596.

68. C E Milner. Functional anatomy for sport and exercise. s.l. : Taylor and Francis, 2008. p. 144.

69. E Thorpe. The Pearson General Knowledge Manual 2012. s.l. : Pearson Education India, 2012.

70. J E Shea and S C Miller. Skeletal function and structure: Implications for tissue-targeted therapeutics. Advanced Drug Delivery Reviews. 2005, Vol. 57, 7, pp. 945– 957.

285

71. J P Bilezikian, L G Raisz and T J Martin. Principles of Bone Biology. s.l. : Academic Press, 2008. p. 1900. Vol. 2.

72. B D Browner. Skeletal Trauma: Basic science, management and reconstruction. s.l. : Elsevier Health Sciences, 2003.

73. K Khan, H McKay, P Kannus, D Bailey, J Wark and K Bennell. Physical activity and bone health. s.l. : Human Kinetics, 2001. p. 275.

74. R Baron. [ed.] M J Favus. Primer on the metabolic bone diseases and disorders or mineral metabolism. washington : American society for bone and mineral research, 2003, pp. 1-9.

75. D W Sommerfeldt and C T Rubin. Biology of bone and how it orchestrates the form and function of the skeleton. Eur Spine J (2001) 10 :S86–S95. 2001, Vol. 10, suppl 2, pp. S86-S95.

76. L L Hench. Bioceramics. J. Am. Ceram. Soc. 1998, pp. 1705-1728.

77. R Huiskes and B Van Rietbergen. Biomechanics of bone. [ed.] V C Mow and R Huiskes. Basic orthopaedic biomechanics and mechano-biology. s.l. : Lippincott Williams and Wilkins, 2003, 4, pp. 123-180.

78. S Weiner and H D Wagner. The material bone: Structure-mechanical function relations. Annual Reviews Material Science. 1998, Vol. 28, pp. 271-298.

79. M J Seibel, S P Robins and J P Bilezikian. Dynamics of Bone and Cartilage: Principles and Clinical Applications. s.l. : Academic Press, 2006. p. 920.

80. J Chow and T J Chambers. An assessment of the prevalence of organic material on bone surfaces. Calcif Tissue Int. 1992, Vol. 50, pp. 118-122.

81. A M Parfitt. The cellular basis of the bone remodelling: The quantum concept reexamined in light of recent advances in the cell biology of bone. Calcif Tissue Int. 1984, Vol. 36, pp. s37-s45.

82. A Schindeler, M M McDonald, P Bokko and D G Little. Bone remodeling during fracture repair: The cellular picture. Seminars in Cell & Developmental Biology. 2008, 19, pp. 459-466.

83. C Delloye and G Bannister. Bone grafting in revision hip arthoplasty. s.l. : CRC Press, 2004. p. 400.

84. W Schlickewei and C Schlickewei. The use of bone substitues in the treatment of bone defects - the clinical view and history. Macromol. Symp. 2007, 253, pp. 10-23.

85. L L Hench, R J Splinter, W C Allen and T K Greenlee. Bonding mechanism at the interface of ceramic prosthetic materials. Journal of Biomedical Materials Research Biomedical Materials Symposium. 1971, Vol. 2, pp. 117-141.

86. American Academy of Orthopaedic Surgeons. Bone Graft Substitues. [ed.] C T Laurencin. s.l. : ASTM International, 2003.

87. B D Ratner, A S Hoffman, F J Schoen and J E Lemons. Biomaterials Science: An Introduction to Materials in Medicine. s.l. : Academic Press, 2012. p. 1573.

88. G M Scortecci, C E Misch and K U Benner. Implants and Resorative dentistry. s.l. : Taylor and Francis, 2001. p. 466.

89. C F Carmichael (III). Ridge preservation with Acellular dermal matrix (AlloDerm and AlloDerm GBR) as a membrane prior to dental implant placement: A human and histological study. s.l. : ProQuest, 2008. p. 55.

286

90. U Meyer, T Meyer, J Handschel and H P Wiesmann. Fundamentals of Tissue Engineering and Regenerative Medicine. s.l. : Springer, 2009. p. 1049.

91. B Kaplan, G J Burkhart and F G Lakkis. Immunotherapy in Transplantation: Principles and Practise. s.l. : J Wiley and Sons, 2012. p. 600.

92. M M Malak. Knee Surgery: Complications, Pitfalls and Salvage. s.l. : Springer, 2000. p. 507.

93. M Z Siemionow and M Eisenmann-Klein. Plastic and Reconstructive Surgery. s.l. : Springer, 2010. p. 778.

94. N Karimbux. Clinical Cases in Periodontics. s.l. : J Wiley and Sons, 2011. p. 304.

95. A J Haddad, S A F Peel, C M L Clokie and G K B Sandor. Closure of rabbit calavarial critical-sized defects using protective composite allogenic and alloplastic bone substitutes. Journal of Craniofacial Surgery. 2006, Vol. 17, 5, pp. 926-934.

96. M A Shermak, L Wong, N Inoue and T Nicol. Reconstruction of complex cranial wounds with demineralized bone matrix and bilayer artificial skin. Journal of Craniofacial Surgery. 2000, Vol. 11, 3, pp. 224-231.

97. S J Kim, J W Lim, J J Ryu, J S Ahn, I H Han and S W Shin. Effects of 4 different alloplastic materials on bone regeneration in a rabbit calvarial defects. Tissue Engineering and Regenerative Medicine. 2009, Vol. 6, 1-3, pp. 63-68.

98. D M Kim, M Camelo, M Nevins, A Fateh, P Schupbach and M Nevins. Alveolar ridge reconstruction with a composite alloplastic biomaterial. The International Journal of Periodontics and Restorative Dentistry. 2012, Vol. 32, 6, pp. 204-209.

99. P L Custer, K M Trinkaus and J Fornoff. Comparative motility of hydroxyapatite and alloplastic enucleation implants. Ophthalmology. 1999, Vol. 106, 3, pp. 513-516.

100. G Carotenuto, G Spagnuolo, L Ambrosio and L Nicolais. Macroporous hydroxapatite as alloplastic material for dental applications. Journal of Materials Science - Materials in Medicine. 1999, Vol. 10, 10-11, pp. 671-676.

101. G E Merwin. Bioglass middle-ear prosthesis preliminary-report. Annals of Otology Rhinology and Larynoglogy. 1986, Vol. 95, 1, pp. 78-82.

102. M W Yung. Literature Review of alloplastic materials in ossiculoplasty. Journal of Laryngology and Otology. 2003, Vol. 117, 6, pp. 431-436.

103. S H Liu, R S Yang, R Alshaiki and J M Lane. Collagen in tendon, ligament and bone healing - A current review. Clinical orthopaedics and related research. 1995, 318, pp. 265-278.

104. J Kumagai, K Sarkar, H K Uhthoff, Y Okawara and A Ooshima. Immunohistochemical distribution of type I, type II and type III collagens in the rabbit supraspinatus tenden insertion. Journal of Anatomy. 1994, Vol. 185, Part 2, pp. 279-284.

105. A Inui, T Kokubo, H Fujioka, I Nagura, R Sakata, H Nishimoto, M Kotera, T Nishino and M Kurosaka. Application of layered poly(L-Lactic acid) cell free scaffold in a rabbit rotator cuff defect model. Sports medicine, arthoscopy, rehabilitation, therapy and technology: SMARTT. 2011, Vol. 3, p. 29.

106. K W Luczynski, T Brynk, B Ostrowska, W Swieszkowski, R Reihsner and C Hellmich. Consistent quasistatic and acoustic elasticity determination of poly-L-lactide-based rapid prototyped tissue engineering scaffolds. Journal of Biomedical Materials Research Part A. 2013, Vol. 101A, 1, pp. 138-144.

287

107. A Inui, T Kokubo, Y Mifune, R Sakata, H Nishimoto, K Nishida, T Akisue, R Kuroda, M Satake, H Kaneko and H fujioka. Regeneration of Rotator cuff tear using electrospun poly(D,L-Lactide-co-glycolide) scaffolds in a rabbit model. Arthroscopy - The Journal of Arthroscopic and Related Surgery. 2012, Vol. 28, 12, pp. 1790-1799.

108. N Toyokawa, H Fujioka, T Kokubo, I Nagura, A Inui, R Sakata, M Satake, H Kaneko and M Kurosaka. Electrospun synthetic polymer scaffold for cartilage repair without cultured cells in an animal model. Arthroscopy - The Journal of Arthroscopic and Related Surgery. 2010, Vol. 26, 3, pp. 375-383.

109. C X F Lam, R Olkowski, W Swiezkowski, K C Tan, I Gibson and D W Hutmacher. Composite PLDLLA/TCP Scaffolds for Bone Engineering: Mechanical and In Vitro Evaluations. 2009, Vol. 23, 1-3, pp. 1480-1483.

110. Q Z Chen and A R Boccaccini. Poly(D,L-lactic acid) coated 45S5 Bioglass®-based scaffolds:. Journal of Biomedicall Materials Research. 2006, 77A, pp. 445-457.

111. A R Evans, J Agel, G L DeSilva, T A DeCoster, D R Dirschl, C B Jones, J F Kellam, D W Lundy, L Marsh, D L Sietsema and M K Sen. A New classification scheme for open fractures. Journal of Orthopaedic Trauma. 2010, Vol. 24, 8, pp. 457-464.

112. L Bi, Y Hu, H Fan, G Meng, J Liu, D Li and R Lv. Treatment of contaminated bone defects with clindamycin-reconstructed bone xenograft-composites. J. Biomed Mater Res Part B: Applied Biomater. 2007, Vol. 82B, pp. 418-427.

113. P J Newby, R El-Gendy, J Kirkham, X B Yang, I D Thompson and A R Boccaccini. Ag-doped 45S5 Bioglass(R) -based bone scaffolds by molten salt ion exchange: processing and characterisation. 2011.

114. J C Warning, S A McCracken and J M Morris. A balancing act: mechanisms by which the fetus avoids rejection by the maternal immune system. Reproduction. 2011, Vol. 141, 6, pp. 715-724.

115. J Mansbridge. Transplantation of engineered cells and tissues. [ed.] A R Boccaccini and J E Gough. Tissue engineering using ceramics and polymers. Cambridge : Woodhead Publishing Limited, 2007, 5, pp. 93- 107.

116. H Himmelrich, R Lo-Man, N Winter, P Guermonprez, C Sedlik, M Rojas, D Monnaie, M Gheorghiu, M Lagranderie, M Hofnung, B Gicquel, J M Clement and C Leclerc. Immune responses induced by recombinant BCG strains according to level of production of a foreign antigen: MaIE. Vaccine. 2000, Vol. 18, 24, pp. 2636-2647.

117. W Zimmerli. Orthopaedic device-associated infection. Clinical Microbiology and infection. 2012, Vol. 18, 12, pp. 1160-1161.

118. E T J Rochford, R G Richards and T F Moriarty. Influence of materials on the development of device- associated infections. Clinical Microbiology and Infection. 2012, Vol. 18, 12, pp. 1162-1167.

119. H Behnia and M H K Motamedi. Reconstruction and rehabilitation of short-range high-velocity gunshot injury to the lower face: A case report. Journal of Cranio-Maxillofacial Surgery. 1997, Vol. 25, 4, pp. 220-227.

120. P V Phan, M Grzanna, J Chu, A Polosky, A El-Ghannam, D Van Heerden, D S Hungerford and C G Frondoza. The effect of silica-containing calcium-phosphate particles on human osteoblasts in vitro. Journal of Biomedical Materials Research Part A. 2003, Vol. 67A, 3, pp. 1001-1008.

121. K Rezwan, Q Z Chen, J J Blaker and A R Boccaccini. Biodegradable and bioactive porous polymer/inorganic composite scaffolds for bone tissue engineering. Biomaterials. 2006, Vol. 27, pp. 3413-3431.

122. R B Heimann. Classic and Advanced Ceramics. s.l. : J Wiley and Sons, 2010.

288

123. B Basu and M Kalin. Tribology of Ceramics and Composites: Materials Science Perspective. s.l. : John Wiley and Sons, 2011.

124. D G Poitout. Biomechanics and Biomaterials in Orthopedics. s.l. : Springer, 2004.

125. M J Yaszemski, D J Trantolo, K-U Lewandrowski, V Hasirci, D E Altobelli and D L Wise. Biomaterials in Orthopedics. s.l. : CRC Press, 2004. p. 457.

126. M Navarro, A Michiardi, O Castano and J A Planell. Biomaterials in orthopaedics. Journal of The Royal Society Interface. 2008, 5, pp. 1137-1158.

127. S Hattar, A Asselin, D Greenspan, M Oboeuf, A Berdal and J-M Sautier. Potential of biomimetic surfaces to promote in vitro osteoblast-like cell differentiation. Biomaterials. 2005, Vol. 26, 8, pp. 839-848.

128. E M Bueno and J Glowacki. Biologic Foundations for Skeletal Tussue Engineering. s.l. : Morgan and Claypool Publishers, 2010. p. 222.

129. M Vallet-Regi, M M Garcia and M Colilla. Biomedical Applications of Mesoporous Ceramics: Drug delivery, Smart Materials and Bone Tissue Engineering. s.l. : CRC Press, 2012. p. 231.

130. M M Pereira and L L Hench. Mechanisms of hydroxyapatite formation on porous gel-silica substrates. Journal of Sol-Gel Science and Technology. 1996, Vol. 7, pp. 59-68.

131. L A Pruitt and A M Chakravartula. Mechanics of Biomaterials: Fundamental Principles for Implant design. s.l. : Cambridge University Press, 2011.

132. D B Herren, L Nagy and D Campbell. Osteosynthesis in the hand: Current concepts :FESSH Instructional Course 2008. s.l. : Karger Medical and Scientific Publishers, 2008.

133. R Barbucci. Integrated Biomaterials Science. s.l. : Springer, 2002.

134. A R Boccaccini, X Chatzistavrou, J J Blaker and S N Nazhat. Degradable and Bioactive synthetic composite scaffolds for bone tissue engineering. [book auth.] N Eliaz. Degradation on Implant Materials. s.l. : Springer, 2011, pp. 110-137.

135. S Lee and D Henthorn. Materials in Biology and Medicine. s.l. : CRC Press, 2012. p. 260.

136. R Lanza, R Langer and J P Vacanti. Principles of Tissue engineering. 3rd Edition. s.l. : Academic Press, 2011. p. 1344.

137. H Liu, E B Slamovich and T J Webster. Improved dispersion of nanophase titania in PLGA enhances osteoblast adhesion. [book auth.] M Z Hu and Y Gogotsi S W Lu. Ceramic Nanomaterials and Nanotechnology III: Proceeding of the American Ceramic Society, Indianapolis, Indiana, USA 2004, Ceramic Transactions. s.l. : J Wiley and Sons, 2012, pp. 247-255.

138. N Pallua. Tissue Engineering. s.l. : Springer, 2011. p. 634.

139. P N Bartlett. Bioelectrochemistry: Fundamentals, Experimental techniques and applications. s.l. : J Wiley and Sons, 2008. p. 494.

140. D E Brown and R D Neumann. Orthopedic Secrets. 3rd Edition. s.l. : Elsevier Health Sciences, 2004. p. 450.

141. M Aebi, V Arlet and J K Webb. AOspine Manual: Principle and techniques Volume 1. s.l. : Thiene, 2007. p. 837. Vol. 1.

289

142. T Albrektsson and C Johansson. Osteoinductive, osteoconductive and osseointegration. Europeon Spine Journal. 2001, 10, pp. s96-s101.

143. P A Tran, L Sarin, R H Hurt and T J Webster. Opportunities for nanotechnology-enabled bioactive bone implants. Journal of Materials Chemistry. 2009, Vol. 19, pp. 2653-2659.

144. W P-D Chen. Oculoplastic Surgery: The Essentials. s.l. : Thieme, 2001. p. 499.

145. D J Griffon. Evaluation of Osteoproductive Biomaterials: Allograft, Bone Inducing Agent, Bioactive Glass and Ceramics. 2002.

146. M Zilbermann. Active implants and scaffolds for tissue regeneration. s.l. : Springer, 2011. p. 514.

147. D Y Wong, E E Liao, J C Leveque, H Brumblay, C-Y Lin, F LaMarca, P H Krebsbach and S J Hollister. Engineered Scaffold Architecture Influences Soft Tissue Regeneration. [book auth.] P Bartolo and B Bidanda. Bio-Materials and Prototyping Applications in Medicine. s.l. : Springer, 2008, Chapter 5, pp. 66-77.

148. R Narayan. Biomedical Materials. s.l. : Springer, 2009. p. 566.

149. L J Gibson, M F Ashby and B A Harley. Cellular Materials in Nature and Medicine. s.l. : Cambridge University Press, 2010. p. 320.

150. L A Gonzalez-Torres, M J Gomez-Benito, M Doblare and J M Garcia-Aznar. Influence of the frequency of the external mechanical stimulus on bone healing: A computational study. Medical Engineering and Physics. 2010, Vol. 32, 4, pp. 363-371.

151. L E Claes, C A Heigele, C Neidlinger-Wilke, D Kaspar, W Seidl, K J Margevicius and P Augat. Effects of mechanical factors on the fracture healing process. Clinical Orthopaedics and Related Research. 1998, Vol. 355, pp. S132-S147.

152. A Bailon-Plaza and M C H van der Meulen. Beneficial effects of moderate, early loading and adverse effects of delayed or excessive loading on bone healing. Journal of Biomechanics. 2003, Vol. 36, 8, pp. 1067- 1077.

153. S K Sohaebuddin. Mechanism of Nanoparticle and Nanotube induced Cell Death. s.l. : ProQuest, 2008. p. 75.

154. J Maroothynaden and L L Hench. Bioglass(R) stimulation of embryonic long-bones in altered loading environments. Journal of gravitational physiology. 2001, Vol. 8, 1, pp. 79-80.

155. M Marcacci, E Kon, V Moukhachev, A Lavroukov, S Kutepov and R Quarto. Stem cells associated with macroporous bioceramics for long bone repair: 6- to 7-year outcome of a pilot clinical study. Tissue engineering. 2007, Vol. 13, 5, pp. 947-955.

156. A El-Ghannam. Bone reconstruction: from biocermics to tissue engineering. Expert review of medical devices. 2005, Vol. 2, 1, pp. 87-101.

157. S E Lobo and T L Arinzhek. Biphasic calcium phopshate ceramics for bone regneration and tissue engineering applications. Materials. 2010, Vol. 3, pp. 815-826.

158. J Venkatesan and S K Kim. Chitosan composites for bone tissue engineering- An Overview. Marine Drugs. 2010, Vol. 8, 8, pp. 2252-2266.

159. H Zhou and J Lee. Nanoscale hydroxyapatite particles for bone tissue engineering. Acta Biomaterialia. 2011, Vol. 7, 7, pp. 2769-2781.

290

160. H Yoshikawa and A Myoui. Bone tissue engineering with porous hydroxyapatite. Journal of Artifical Organs. 2005, Vol. 8, 3, pp. 131-136.

161. J A Epinette and M T Manely. Fifteen years of clinical experience with hydroxyapatite coating in joint arthroplasty. s.l. : Springer Science and Buisness Media, 2003. p. 452.

162. S Zhang. Hydroxyapatite coatings for biomedical applications. s.l. : CRC Press, 2013. p. 469.

163. S Samavedi, A R Whittington and A S Goldstein. Calcium phosphate ceramics in bone tissue engineering: A review of properties and their influence on cell behaviour. Acta Biomaterialia. 2013, Vol. 9, pp. 8037-8045.

164. S V Dorozhkin. Amorphous calcium (ortho)phosphate. Acta Biomaterialia. 2010, Vol. 6, 12, pp. 4457- 4475.

165. C Combes and C Rey. Amorphous calcium phosphates: Synthesis, properties and uses in biomaterials. Acta Biomaterialia. 2010, Vol. 6, 9, pp. 3362-3378.

166. M Mastrogiacomo, A Muraglia, V Komdev, F Peyrin, F Rustichelli, A Crovace and R Cancedda. Tissue engineering of bone: Search for a better scaffold. Orthodontics and craniofacial research. 2005, Vol. 8, 4, pp. 277-284.

167. D C Fredericks, J A Bobst, E B Petersen, J V Nepola, J E Dennis, A I Burgess, R J Overby and O H Schulz. Cellular interactions and bone healing responses to a novel porous tricalcium phosphate bone graft material. Orthopedics. 2004, Vol. 21, 1, pp. S167-S173.

168. A S Breitbart, D A Staffenberg, C H M Thorne, P M Glat, N S Cunningham, A H Reddi, J Ricci and G Steiner. Tricalcium phosphate and oeteogenin- A bioactive onlay bone-graft substitute. Plastic and reconstructuve surgery. 1995, Vol. 96, 3, pp. 699-708.

169. J Wojtowics, J Leszczynska, A Chroscicka, A Slosarczyk, Z Paszkiewicz, A Zima, K Rozniatowski, P Jelen and M Lewandowska-Szumiel. Comparative in vitro study of calcium phosphate ceramics for their potency as scaffolds for tissue engineering. Biomedical Materials and Engineering. 2014, Vol. 24, 3, pp. 1609- 1623.

170. S B Sulaiman, T K Keong, C H Cheng, A B Saim and R B H Idrus. Tricalcium phosphate/hydroxyapatite (TCP-HA) bone scaffolds as potential candidate for the formation of tissue engineered bone. The Indian Journal of Medical research. 2013, Vol. 137, 6, pp. 1093-1101.

171. W Holland and G Beal. Glass-ceramic Technology. The American Ceramic Society. Westville, OH, USA : s.n., 2002.

172. I D Xynos, A J Edgar, L D K Buttery, L L Hench and M Polak. Gene expression profiling of human osteoblasts following treatment with the ionic products of Bioglass 45S5 dissolution. Journal of Biomedical Materials Research. 2001, 55, pp. 151-157.

173. M Kobayashi, T Nakamura, Y Okada, A Fukumoto, T Furukawa, H Kato, T Kokubo and T Kikutani. Bioactive bone cement: Comparison of apatite and wollastonite containing glass-ceramic, hydroxyapatite, and beta-tricalcium phosphate fillers on bone-bonding strength. Journal of Biomedical Materials Research. 1998, Vol. 42, 2, pp. 223-237.

174. J R Jones and L L Hench. Regeneration of trabecular bone using porous ceramics. Curr Opin Solid State Mater Sci. 2003, 7, pp. 301-307.

175. L L Hench. Sol-gel materials for bioceramic applications. Curr Opin Solid State Mater Sci. 1997, 2, pp. 604-610.

291

176. K D Lobel, and L L Hench. In-vitro protein interactions with a bioactive gel-glass. J. Sol-gel Sci Tech. 1996, 7, pp. 69-76.

177. H Keshaw, A Forbes and R M Day. Release of angiogenic growth factors from cells encapsulated in alginate beads with bioactive glass. Biomaterials. 2005, 26, pp. 4171-4179.

178. T J Webster, R W Siegel and R Bizios. Osteoblast adhesion on nanophase ceramics. Biomaterials. 1999, Vol. 20, 13, pp. 1221-1227.

179. R L Du, J Chang, S Y Ni, W Y Zhai, J Y Wang. Characterisation and in vitro bioactivity of zinc- containing bioactive glass and glass-ceramics. Journal of Biomaterials applications. 2006, Vol. 20, 4, pp. 341- 360.

180. A Laczka-Osyczka, M Laczka, S Kasugai and K Ohya. Behaviour of bone marrow cells cultured on three different coatings of gel-derived bioactive glass-ceramics at early stages of cell differentiation. Journal of Biomedical Materials Research. 1998, Vol. 42, 3, pp. 433-442.

181. S M Rea, R A Brooks, S M Best, T Kokubo and W Bonfield. Proliferation and differentiation of osteoblast-like cells on apatite-wollastonite/polyethylene composites. Biomaterials. 2004, Vol. 25, 18, pp. 4503- 4512.

182. I A Silver, J Deas and M Erecinska. Interactions of bioactive glasses with osteoblasts in vitro: effects of 45S5 Bioglass(R), and 58S and 77S bioactive glasses on metabolism, intracellular ion concentrations and cell viability. Biomaterials. 2001, Vol. 22, 2, pp. 175-185.

183. L L Hench. Bioceramics: From concept to clinic. Journal of American Ceramic Society. 1991, Vol. 74, 7, pp. 1487-1510.

184. S K Nandi, B Kunda and S Datta. Development and applications of varieties of bioactive glass compositions in dental surgery, third generation tissue engineering, othopaedic surgery and as a drug delivery. [book auth.] R Pignatello. Biomaterials Applications for Nanomedicine. s.l. : InTech, 2011, pp. 69-116.

185. E Ruiz-Hitzky, K Ariga and Y M Lvov. Bio-inorganic hybrid nanomaterials. s.l. : J Wiley and Sons, 2008. p. 521.

186. L L Hench and J Wilson. An introduction to Bioceramics. s.l. : World Scientific, 1993. p. 386.

187. D Shi. Biomaterials and Tissue Engineering. s.l. : Springer, 2004. p. 246.

188. L Sennerby, A Dasmah, B Larsson and M Iverhed. Bone tissue responses to surface-modified zirconia implants: A histomorphometric and removal torque study in the rabbit. Clinical Implant Dentistry and Related Research. 2006, Vol. 7, S1, pp. S13-S20.

189. Y-D Cho, J-C Shin, H-L Kim, M Gerelmaa, H-I Yoon, H-M Ryoo, D-J Kim and J-S Han. Comparison of the osteogenic potential of titanium and modified zirconia-based bioceramics. Intenational Jounal of Molecular Sciences. 2014, Vol. 15, pp. 4442-4452.

190. S Novak, J Druce, Q Z Chen, A R Boccaccini. TiO2 foams with poly-(d,l-lactic acid) (PDLLA) and PDLLA/Bioglass(A(R)) coatings for bone tissue engineering scaffolds. Journal of Material Science. 2009, Vol. 44, 6, pp. 1442-1448.

191. P Niemeyer, U Krause, J Fellenberg, P Kasten, A Seckinger, A D Ho and H G Simank. Evaluation of mineralised collagen and alpha-tricalcium phosphate using human mesenchymal stem cells. Cell Tissue Organs. 2004, Vol. 177, 2, pp. 68-78.

292

192. B Marelli, C E Ghezzi, J E Barralet, A R Boccaccini and S N Nazhat. Three-Dimensional Mineralisation of Dense Nanofibrillar Collagen-Bioglass Hybrid Scaffolds. Biomacromolecules. 2010, Vol. 11, 6, pp. 1470-1479.

193. B Marelli, C E Ghezzi, D Mohn, W J Stark, J E Barralet, A R Boccaccini, S N Nazhat. Accelerated mineralisation of dense collagen-nano bioactive glass hybrid gels increase scaffold stiffness and regulates osteoblastic function. Biomaterials. 2011, Vol. 32, 34, pp. 8915-8926.

194. L Lefebvre, J Chevalier, L Gremillard, R Zenati, G Thollet, D Bernache-Assollant and A Govin. Structural transformations of bioactive glass 45S5 with thermal treatments. Acta Biomaterialia. 2007, Vol. 55, 10, pp. 3305-3313.

195. L Lefebvre, L Gremillard, J Chevalier, R Zenati and D Bernache-Assolant. Sintering behaviour of 45S5 bioactive glass. Acta Biomaterialia. 2008, Vol. 4, 6, pp. 1894-1903.

196. D S Brauer, M N Arjum, M Mneimne, R M Wilson, H Doweidar and R G Hill. Fluoride-containing bioactive glass-ceramics. Journal of Non-crystalline solids. 2012, Vol. 358, 12-13, pp. 1438-1442.

197. A Saranti, I Koutselas and M A Karakassides. Bioactive glasses in the system CaO-B2O3-P2O5: Preparation, structural study and in vitro evaluation. Journal of Non-crystalline solids. 2006, Vol. 352, 2, pp. 390-398.

198. A Goel, R R Rajagopal and J M F Ferreira. Influence of strontium on structure, sintering and biodegradation behaviour of CaO-MgO-SrO-SiO2-P2O5-CaF2 glasses. Acta Biomaterialia. 2011, Vol. 7, 11, pp. 4071-4080.

199. X Chen, Y Meng, Y Li and N Zhao. Investigation on bio-mineralization of melt and sol-gel derived bioactive glasses. Applied Surface Science. 2008, Vol. 255, 2, pp. 562-564.

200. D Arcos, D C Greenspan and M Vallet-Regi. A New quantitative method to evaluate the in vitro bioactivity of melt and sol-gel derived silicate glasses. Journal of Biomedical Materials Research Part A. 2003, Vol. 65A, 3, pp. 344-351.

201. I Cacciotti, M Lombardi, A Bianco, A Ravaglioli and L Montanaro. Sol-gel derived 45S5 bioglass: synthesis, microstructural evolution and thermal behaviour. Journal of Materials Science - Materials in Medicine. 2012, Vol. 23, 8, pp. 1849-1866.

202. V Mourino and A R Boccaccini. Bone tissue engineering therapeutics: controlled drug delivery in three- dimensional scaffolds. Journal of the Royal Society Interface. 2010, Vol. 7, 43, pp. 209-227.

203. X Wang, J S Nyman, X Dong and M Reyes. Biomechanics in Bone Tissue Engineering. s.l. : Morgan and Claypool Publishers, 2010. p. 232.

204. T Clive Lee and P F Niederer. Basic Engineering for Medics and Biologists: An ESEM Primer. s.l. : IOS Press, 2010. p. 369.

205. J R Jones and A G Clare. Bioglasses: An Introduction. s.l. : J Wiley and Sons, 2012. p. 320.

206. N Sultana. Biodegradable polymer-based scaffolds for bone tissue engineering. s.l. : Springer, 2013. p. 70.

207. J M Polak. Advances in Tissues Engineering. s.l. : Imperial College Press, 2008. p. 908.

208. J O Hollinger. An introduction to biomaterials. 2nd edition. s.l. : CRC Press, 2011. p. 624.

209. L Di Silvo. Bone tissue engineering and biomineralisation. [ed.] A R Boccaccini and J E Gough. Tissue engineering using ceramics and polymers. Cambridge : Woodhead Publishing Limited, 2007, 15, pp. 319-334.

293

210. V Karageorgiou and D Kaplan. Porosity of 3D biomaterial scaffolds and osteogenesis. Biomaterials. 2005, Vol. 26, 27, pp. 5474-5491.

211. Q Z Chen, O Bretcanu and A R Boccaccini. Inorganic and composite bioactive scaffolds for bone tissue engineering. [ed.] P K Chu and X Liu. Biomaterials Fabrication and Processing Handbook. s.l. : CRC Press, 2008, pp. 3-44.

212. M Cerruti. Surface characterisation of silicate bioceramics. Philosophical transactions of the royal society A. 2012, Vol. 370, pp. 1281-1312.

213. D O Njobuenwu, E O Oboho and R H Gumus. Determination of contact angle from contact area of liquid droplet spreading on solid substrate. Leonardo Electronic Journal of Practices and technologies. 2007, Vol. 6, 10, pp. 29-38.

214. A N Cormack and A Tilocca. Structure and biological activity of glass and ceramics. 2012, Vol. 370, pp. 1271-1280.

215. J Bico, U Thiele and D Quere. Wetting of textured surfaces. Colloids and Surfaces A. 2002, Vol. 206, pp. 41-46.

216. A Atala and R P Lanza. Methods of tissue engineering. s.l. : Gulf Professional Publishing, 2002. p. 1285.

217. P Ulvskov. Annual Plant reviews, Plant Polysaccharides: Biosynthesis and Bioengineering. s.l. : J Wiley and Sons, 2010. p. 504.

218. A Carre and K L Mittal. Surface and Interfacial Aspects of Cells Adhesion. s.l. : CRC Press, 2011. p. 544.

219. E Nasatzky, J Gultchin and Z Schwartz. The role of surface roughness in promoting osteointegration. Refuʾat ha-peh ṿeha-shinayim. 1993, Vol. 20, 3, pp. 8-19, 98.

220. K Anselme, M Bigerelle, B Noel, E Dufresne, D Judas, A Lost and P Hardouin. Qualitative and quantitative study of human osteoblast adhesion on materials with various surface roughness. Journal of Biomedical Materials Research. 2000, Vol. 49, 2, pp. 155-166.

221. S Adi, H Adi, H K Chan, P M Young, D Traini, R Y Yang and A B Yu. Scanning white-light interferometry as a novel technique to quantify the surface roughness of micron-sized particles for inhalation. Langmuir. 2008, Vol. 24, 19, pp. 11307-11312.

222. J E Gough, I Notingher and L L Hench. Osteoblast attachment and mineralized nodule formation on rough and smooth 45S5 bioactive glass monoliths. 2004, Vol. 68A, 4, pp. 640-650.

223. A Itala, H O Ylanen, J Yrjans, T Heino, T Hentunen, M Hupa and H T Aro. Characterisation of microrough bioactive glass surface: Surface reactions and osteoblast responses in vitro. 2002, Vol. 62, 3, pp. 404-411.

224. L Francis, D Meng, J C Knowles, I Roy and A R Boccaccini. Multifunctional P(3HB) microsphere/45S5 Bioglass(R)- based composite scaffolds for bone tissue engineering. Acta Biomaterialia. 2010, Vol. 6, 7, pp. 2773-2786.

225. R El-Gendy, J Kirkham, P J Newby, Y Mohanram, A R Boccaccini and X B Yang. 45S5 Bioglass® scaffolds enhance angiogenic differentiation of human dental pulp stromal cells used for mineralised tissue engineering.

294

226. B R Prasad, M A Brook, T Smith, S G I Zhao, H Sheardown, R D'souza and Y Rocher. Controlling cellular activity by manipulating silicone surface roughness. Colloids and Surface B - Biointerfaces. 2010, Vol. 78, 2, pp. 237-242.

227. M Kalbacova, B Rezek, V Baresova, C Wolf-Brandstetter and A Kromer. Nanoscale topography of nanocrystalline diamonds promotes differentiation of osteoblasts. Acta Biomaterialia. 2009, Vol. 5, 8, pp. 3076- 3088.

228. N R Washburn, K M Yamada, C G Simon, S B Kennedy and E J Amis. High-throughput investigation of osteoblast responce to polymer crystallinity: Influence of nano-scale roughness on proliferation. Biomaterials. 2004, Vol. 25, 7-8, pp. 1215-1224.

229. H Zhang, X-J Ye and J-S Li. Preparation and biocompatibility evaluation of apatite/wollastonite -derived porous bioactive glass-ceramic scaffolds. Biomedical Materials. 2009, Vol. 4, 4, p. 045007.

230. B Cao, D Zhou, M Xue, G Liu, W Yang, Q Long, L Liu. Study on surface modification of porous apatite-wollastonite bioactive glass ceramic scaffold. Applied surface science. 2008, Vol. 255, 2, pp. 505-508.

231. M Xue, J Ou, D L Zhou, D G Feng, W Z Yang, G D Li and D P Liu. Preparation and properties of porous apatite-wollastonite bioactive glass-ceramics. Bioceramics. 2007, Vol. 19, 330-332, pp. 169-172.

232. P Sooksaen, K Chaithep, T Saliwong and T Duangart. Apatite-wollastonite bioactive glass-ceramic scaffolds for hard tissue engineering. Biomaterials and applications. 2012, Vol. 506, pp. 146-149.

233. J Suwanprateeb, R Sanngam, W Suvannapruk and T Panyathanmaporn. Mechanical and in vitro performance of apatite-wollastonite glass-ceramic reinforced hydroxyapatite composite fabricated by 3-D printing. Journal of Materials Science: Materials in Medicine. 2009, Vol. 20, 6, pp. 1281-1289.

234. D J Wood, J Dyson, K Xiao, K W Dalgarno, P Genever. Processing and characterisation of apatite- wollastonite porous scaffolds for bone tissue engineering. Bioceramics. 2008, Vol. 20, 361-363, pp. 923-926.

235. K Xiao, K W Dalgarno, D J Wood, R D Goodridge and O Ohtsuki. Indirect selective laser sintering of apatite-wollastonite glass-cermic. Proceedings of the institution of mechanical engineers part I - Journal of engineering in medicine. 2008, Vol. 222, H7, pp. 1107-1114.

236. P K Chu and X Liu. Biomaterials fabrication and processing handbook. s.l. : CRC Press, 2008. p. 720.

237. K C R Kolan, M C Leu, G E Hilmas, R F Brown and M Velez. Fabrication of 13-93 bioactive glass scaffolds for bone tissue engineering using indirect slective laser sintering. Biofabrication. 2011, Vol. 3, 2, p. 025004.

238. T S Huang, M N Rahanam, N D Doiphode, M C Leu, B S Bal, D E Day and X Liu. Porous and strong bioactive glass (13-93) scaffolds fabricated by freeze extrusion technique. Materials Science and Engineering C - Materials for biological application. 2011, Vol. 31, 7, pp. 1482-1489.

239. Q Fu, M N Rahaman, B S Bal, R F Brown and D E Day. Mechanical and in vitro performance of 13-93 bioactive glass scaffolds prepared by a polymer foam replication technique. Acta Biomaterialia. 2008, Vol. 4, 6, pp. 1854-1864.

240. R F Brown, D E Day, S Jung, M N Rahaman and Q Fu. Growth and differentiation of osteoblastic cells on 13-93 bioactive glass fibres and scaffolds. Acta Biomaterialia. 2008, Vol. 4, 2, pp. 387-396.

241. Q A Fu, M N Rahaman, H L Fu and X Liu. Silicate, borosilicate and borate bioactive glass scaffolds with controllable degredation for bone tissue engineering applications I: Preparation and in vitro degredation. Journal of Biomedical Materials Research Part A. 2010, Vol. 95A, 1, pp. 164-171.

295

242. J R Jones and A R Boccaccini. Biomedical Applications: Tissue Engineering. [ed.] M Scheffler and P Colombo. Cellular Ceramics. Weinheim : Wiley-VCH Verlay GmbH and co. KgsA, 2005, Vol. 1, 5.8, pp. 547- 570.

243. K K Mallick, J Winnett, W van Grunsven, J Lapworth and G C Reilly. Three-dimensional porous bioscaffolds for bone tissue engineering regeneration: Fabrication via adaptive foam reticulation and freeze casting techniques, characterisation, and cell study. Journal of Biomedical Materials Research Part A. 2012, Vol. 100A, 11, pp. 2948-2959.

244. Q Z Chen and G A Thouas. Fabrication and characterisation of sol-gel derived 45S5 Bioglass -ceramic scaffolds. 2011, Vol. 7, pp. 3616-3626.

245. S Eqtesadi, A Motealleh, P Miranda, A Lemos, A Rebelo and J M F Ferreira. A simple recipe for direct writing complex 45S5 Bioglass(R) 3D scaffolds. Materials Letters. 2013, Vol. 93, pp. 68-71.

246. R Comesana, F Lusquinos, J del Val, M Lopez-Alvarez, F Quintero, A Rivero, M Boutinguiza, A de Carlos, J R Jones, R G Hill and J Pou. Three-dimensional bioactive glass implants fabricated by rapid prototyping based on CO2 laser cladding. Acta Biomaterialia. 2011, Vol. 7, 9, pp. 3476-3487.

247. S Farhangdoust, A Zamanian, M Yasaei and M Khorami. The effect of processing parameters and solid concentration on the mechanical and microstructure properties of freeze-casted macroporous hydroxyapatite scaffolds. Materials Science and Engineering C - Materials for Biological Applications. 2013, Vol. 33, 1, pp. 453-460.

248. J Malmstrom, E Adolfsson, A Arvidsson and P Thomsen. Bone response inside free-form fabricated macroporous hydroxyapatite scaffolds with and without an open microporosity. Clinical Implant Densitry and Related Research. 2007, Vol. 9, 2, pp. 79-88.

249. L R Rodrigues, M A d'Avila, F J M Monteiro and C A D Zavagila. Synthesis and characterisation of nanocrystalline hydroxyapatite gel and its application as scaffold aggregation. Materials research -Ibero- American Journal of Materials. 2012, Vol. 15, 6, pp. 974-980.

250. H Ghomi, M H Fathi and H Edris. Preparation of nanostructure hydroxyapatite scaffold for tissue engineering applications. Journal of Sol-gel science and technology. 2011, Vol. 58, 3, pp. 642-650.

251. Y Gao, W L Cao, X Y Wang, Y D Gong, J M Tian, N M Zhao and X F Zhang. Characterisation and osetoblast-like cell compatibility of porous scaffolds: bovine hydroxyapatite and novel hydroxyapatite artificial bone. Journal of Materials Science - Materials in Medicine. 2006, Vol. 17, 9, pp. 815-823.

252. T Guda, S Oh, M R Appleford and J L Ong. Bilayer hydroxyapatite scaffolds for maxillofacial bone tissue engineering. International Journal of Oral and Maxillofacfial implants. 2012, Vol. 27, 2, pp. 288-294.

253. B Liu and D-X Lun. Current application of beta-tricalcium phosphate composites in orthopaedics. Orthopaedic Surgery. 2012, Vol. 4, 3, pp. 139-144.

254. L Chen, C Y Tang, C P Tsui and D Z Chen. Mechanical properties and in vitro evaluation of bioactivity and degradation of dexamethasone-releasing poly-d-l-lactide/nano-hydroxyapatite composite scaffolds. Journal of the mechanical behaviour of biomedical materials. 2013, Vol. 22, pp. 41-50.

255. B H Yang, M Li, Y Wu, J J Huang, H Zhang and C H Deng. Preparation and characterisation of bone- like hydroxyapatite/poly(methyl methacrylate) composite biomaterials. Science and Engineering of Composite Materials. 2013, Vol. 20, 2, pp. 147-153.

256. K A Hing, L E Wilson and T Buckland. Comparative performance of three ceramic bone graft substitutes. Spine Journal. 2007, Vol. 7, 4, pp. 475-490.

296

257. C E Wilson, C A van Blitterswijk, A J Verbout, W J A Dhert and J D de Bruijn. Scaffolds with a standardized macro-architecture fabricated from several calcium phosphate ceramics using an indirect rapid prototyping technique. Journal of Materials Science - Materials in Medicine. 2011, Vol. 22, 1, pp. 97-105.

258. V Thomas, D R Dean and Y K Vohra. Nanostructured Biomaterials for regenerative medicine. Current Nanoscience. 2006, Vol. 2, 3, pp. 155-177.

259. M Houmard, Q Fu, E Saiz and A P Tomsia. Sol-gel method to fabricate CaP scaffolds by robocasting for tissue engineering. Journal of Materials Science - Materials in Medicine. 2012, Vol. 23, 4, pp. 921-930.

260. Y M Soon, K H Shin, Y H Koh, J H Lee and H E Kim. Compressive strength and processing of camphene-based freeze cast calcium phosphate scaffolds with aligned pores. Materials Letters. 2009, Vol. 63, 17, pp. 1548-1550.

261. K L Lin, L Chen, H Y Qu, J X Lu and J Chang. Improvement of mechanical properties of macroporous beta-tricalcium phosphate bioceramic scaffolds with uniform and interconnected pore structure. Ceramics International. 2011, Vol. 37, 7, pp. 2397-2403.

262. N Nezafati, F Moztarzadeh, S Hesaraki, Z Moztarzadeh and M Mozafari. Biological response of a recently developed nanocomposite based on calcium phosphate cement and sol-gel derived bioactive glass fibers as substitution of bone tissues. Ceramics International. 2013, Vol. 39, 1, pp. 289-297.

263. Y Zhang and M Q Zhang. Three-dimensional macroporous calcium phosphate bioceramics with nested chitosan sponges for load-bearing bone implants. Journal of Biomedical Materials Research. 2002, Vol. 61, 1, pp. 1-8.

264. E Engel, S Del Valle, C Aparicio, G Altankov, L Asin, J A Planell and M P Ginebra. Discerning the role of topography and ion exchange in cell response of bioactive tissue engineering scaffolds. Tissue Engineering Part A. 2008, Vol. 14, 8, pp. 1341-1351.

265. L Xiang, D C Li, B H Lu, Y P Tang, L Wang and Z Wang. Design and fabrication of CAP scaffolds by indirect solid free form fabrication. Rapid Prototyping Journal. 2005, Vol. 11, 5, pp. 312-318.

266. B M Chesnutt, A M Viano, Y L Yuan, Y Z Yang, T Guda, M R Appleford, J L Ong, W O Haggard and J D Burngardner. Design and characterisation of a novel chitosan/nanocrystalline calcium phosphate composite scaffold for bone regeneration. Journal of Biomedical Materials Research Part A. 2009, Vol. 88A, 2, pp. 491-502.

267. L Q Zheng, F Yang, H Shen, X F Hu, C Mochizuki, M Sato, S G Wang and Y D Zhang. The effect of composition of calcium phosphate composite scaffolds on the formation of tooth tissue from human dental pulp stem cells. Biomaterials. 2011, Vol. 32, 29, pp. 7053-7059.

268. Q Fu, E Saiz, M N Rahaman and A P Tomsia. Bioactive glass scaffolds for bone tissue engineering: state of the art and future perspectives. Materials Science and Engineering C. 2011, Vol. 31, 7, pp. 1245-1256.

269. O Bretcanu, C Samaille and A R Boccaccini. Simple methods to fabricate Bioglass (R)-derived glass- ceramic scaffolds exhibiting porosity gradient. Journal of Materiasl Science. 2008, Vol. 43, 12, pp. 4127-4134.

270. I Ochoa, J A Sanz-Herrera, J M Garcia-Aznar, M Doblare, D M Yunos and A R Boccaccini. Permeability evaluation of 45S5 Bioglass(R) - based scaffolds for bone tissue engineering. Journal of Biomechanics. 2009, Vol. 42, pp. 257-260.

271. D M Yunos, Z Ahmad, V Salih and A R Boccaccini. Stratified scaffolds for osteochondral tissue engineering applications: Electrospun PDLLA nanofibre coated Bioglass(R) -derived foams. Journal of Biomaterials Applications. 2013, Vol. 27, 5, pp. 537-551.

297

272. J Korpela, A Kokkari, H Korhonen, M Malin, T Narhi and J Seppala. Biodegradable and bioactive porous scaffold structure prepared using fused deposition modelling. Journal of Biomedical Materials Research Part B - Applied Biomaterials. 2013, Vol. 101B, 4, pp. 610-619.

273. E Vorndran, U Klammert, A Ewald, J E Barralet and U Gbureck. Simultaneous immobilization of bioactives during 3D powder printing of bioceramic drug-releasing matrices. Advanced Functional Materials. 2010, Vol. 20, 10, pp. 1585-1591.

274. L Elomaa, A Kokkari, T Narhi and J V Seppala. Porous 3D modeled scaffolds of bioactive glass and photocrosslinkable poly(epsilon-caprolactone) by stereolithography. Composites Science and Technology. 2013, Vol. 74, pp. 99-106.

275. S Eqtesadi, A Motealleh, P Miranda, A Pajares, A Lemos and J M F Ferreira. Robocasting of 45S5 bioactive glass scaffolds for bone tissue engineering. Journal of the European Ceramic Society. 2014, Vol. 34, 1, pp. 113-124.

276. S Eqtesadi, A Motealleh, P Miranda, A Lemos, A Rebelo, J M F Ferreira. A simple recipe for direct writing complex 45S5 Bioglass(R) 3D scaffolds. Materials Letters. 2013, Vol. 93, pp. 68-71.

277. J Barralet, U Gbureck, P Habibovic, E Vorndran, C Gerard, and C J Doillon. Angiogenesis in calcium phosphate scaffolds by inorganic copper ion release. Tissue engineering: Part A. 2009, Vol. 15, 7, pp. 1601-1609.

278. T H Nguyen, S H Kim, C G Decker, D Y Wong, J A Loo and H D Maynard. A heparin-mimicking polymer conjugate stabilzes basic fibroblast growth factor. Nature Chemistry. 2013, Vol. 5, 3, pp. 221-227.

279. S Mathieu, C Jeanneau, N Sheibat-Othman, N Kalaji, H Fessi and I About. Usefulness of controlled release of growth factors in investigating the early events of dentin-pulp regeneration. Journal of Endodontics. 2013, Vol. 39, 2, pp. 228-235.

280. E Oral, A L Neils, S L Rowell, A J Lozynsky and O K Muratoglu. Increasing irradiation temperature maximizes vitamin E grafting and wear resistence of ultrahigh molecular weigth polyethylene. Journal of Biomedical Materials Research Part B - Applied Biomaterials. 2013, Vol. 101B, 3, pp. 436-440.

281. W Chen, Y Liu, H S Courtney, M Bettenga, C M Agrawal, J D Bumgardner and J L Ong. In vitro anti-bacterial and biological properties of magnetron co-souttered silver-containing hydroxyapatite coating. Biomaterials. 27, 2006, pp. 5512-5517.

282. T J Mäkinen, M Veiranto, J Knuuti, J Jalava, P Törmälä and H T Aro. Efficacy of bioabsorbable antibiotic containing bone screw in the prevention of biomaterial-related infection due to Staphylococcus aureus. Bone. 2005, Vol. 36, 2.

283. P S Loomba, J Taneja and B Mishra. Methicillin and vancomycin resistant S.aureus in hospitilized patients. Journal of Global Infectious Diseases. 2010, Vol. 2, 3, pp. 275-283.

284. A C Stall, E Becker, S C Ludwig, D Gelb and K A Poelstra. Reduction of post operative spinal implant infection using gentamicin microspheres. Spine. 2009, Vol. 34, 5, pp. 479-483.

285. F J Schmitz and M E Jones. Antibiotics for treatment of infections cause by MRSA and elimination of MRSA carriage. What are the choices? International Journal of Antimicrobial Agents. 1997, Vol. 9, pp. 1-19.

286. U Joostena, A Joist, G Gosheger, U Liljenqvist, B Brandt and C Von Eiff. Effectiveness of hydroxyapatite-vancomycin bone cement in the treatment of Staphylococcus aureus induced chronic osteomyelitis. Biomaterials. 2005, Vol. 26, 25.

298

287. P Melichercik, D Johoda, V Bartak, E Klapkova, O Nyc and D Pokarny. In vitro assessment of vancomycin released from bone grafts. Acta Chirurgiae Othopaedicae et Traumatologiae Cechoslocvaca. 2010, Vol. 77, 5, pp. 411-415.

288. R O Becker. Silver ions in the treatment of local infections. Metal-based drugs. 1999, Vol. 6, 4-5, pp. 311- 314.

289. C T Wu, Y H Zhou, M C Xu, P P Han, L Chen, J Chang and Y Xiao. Copper-containing mesoporous bioactive glass scaffolds with multifunctional properties of angiogenesis capacity, osteostimulation and antibacterial activity. Biomaterials. 2013, Vol. 334, 2, pp. 422-433.

290. A H K Chou, R Z LeGeros, Z Chen and Y H Li. Antibacterial effect of zinc phosphate mineralized guided bone regeneration membranes. Implant denistry. 2007, Vol. 16, 1, pp. 89-100.

291. K Saleh, M Olson, S Resig, B Bershadsky, M Kuskowski, T Gioe, H Robinson, R Schmidt and E McElfresh. Predictors of wound infection in hip and knee join replacement: results from a 20 year surveillance program. Journal of Orthopaedic Research. 2002, Vol. 20, 3, pp. 506-515.

292. S Hu, J Chang, M Liu and C Ning. Study on antibacterial effect of 45S5 Bioglass(R). Journal of Materials Science: Materials in Medicine. 2009, Vol. 20, pp. 281-286.

293. A Jacobs, F Lafolie, J M Herry and M Debroux. Kinetic adhesion of bacterial cells to sand: Cell surface properties and adhesion rate. Colloids and Surfaces B: Biointerfaces. 2007, Vol. 59, 1, pp. 35-45.

294. I Allen, M W Wilson and H Newman. Particulate Bioglass(R) reduces viability of bacterial biofilms formed on its surface in an in vitro model. Clinical Oral implants research. 2002, Vol. 13, 1, pp. 53-58.

295. J P Ruparelia, A K Chatterjee, S P Duttagupta and S Mukherji. Strain specificity in antimicrobial activity of silver and copper nanoparticles. Acta biomaterial. 2008, Vol. 4, pp. 707-716.

296. Q L Feng, J Wu, G Q Chen, F Z Cui, T N Kim and J O Kim. A mechanistic study of the antibacterial effect of silver ions on Escherichia coli and Staphylococcus. 2000, Vol. 52, pp. 662-668.

297. C Marambio-Jones and E M V Hoek. A review of the antibacterial effects of silver nanomaterials and potential implications for human health and the environment. 2010, Vol. 12, pp. 1531-1551.

298. C Beer, R Foldbjerg, Y Hayashi, D S Sutherland and H Autrup. Toxicity of silver nanoparticles - Nanoparticle or silver ion? Toxicology Letters. 2012, Vol. 208, 3, pp. 286-292.

299. C Greulich, D Braun, A Peetsch, J Diendorf, B Siebers, M Epple and M Koller. The toxic effect of silver ions and silver nanoparticles towards bacteria and human cells occurs in the same concentration range. RSC Advances. 2012, Vol. 2, 17, pp. 6981-6987.

300. J Y Liu, D A Sonshine, S Shervani and R H Hurt. Controlled release of biological active silver from nanosilver surfaces. ACS Nano. 2010, Vol. 4, 11, pp. 6903-6913.

301. L J Wilkinson, R J White and J K Chipman. Silver and nanoparticles of silver in wound dressings: a review of efficacy and safety. Journal of Wound Care. 2011, Vol. 20, 11, pp. 543-549.

302. Y Z Wan, S Raman, F He and Y Huang. Surface modification of medical metals by ion implantation of silver and copper. Vacuum. 2007, Vol. 81, pp. 1114-1118.

303. T D Fang, A Salim, W Xia, R P Nacamuli, S Guccione, H M Song, R A Carano, E H Filvaroff, M D Bednarski, A J Giaccia and M T Longaker. Angiogenesis is required for successful bone induction during distraction osteogenesis. Journal of Bone and Mineral Research. 2005, Vol. 20, 7, pp. 1114-1124.

299

304. J D Boerckel, B A Uhrig, N J Willet, N Huebsch and R E Guldberg. Mechanical regulation of vascular growth and tissue regeneration in vivo. Proceedings of the national academy of sciences of the united states of america. 2011, Vol. 108, 37, pp. E674-E680.

305. D H R Kempen, L C Lu, A Heijink, T E Hefferan, L B Creemers, A Maran, M J Yaszemski and W J A Dhert. Effect of local sequential VEGF and BMP-2 delivery on ectopic and orthotopic bone regeneration. Biomaterials. 2009, Vol. 30, 14, pp. 2816-2825.

306. M L Chen, K Song, N J Rao, M Q Huang, Z J Huang and Y G Cao. Roles of exogenously regulated bFGF expression in angiogenesis and bone regeneration in rat calvarial defects. International journal of molecular medicine. 2011, Vol. 27, 4, pp. 545-553.

307. P Velasco and B Lange-Asschenfeldt. Dermatological aspects of angiogenesis. British Journal of Dermatology. 2002, Vol. 147, 5, pp. 841-852.

308. O C Velazquez. Angiogenesis and vasculogenesis: Inducing the growth of new blood vessels and wound healing by stimulation of bone marrow-derived progenitor cell mobilization and homing. Journal of Vascular Surgery. 2007, Vol. 45, A, pp. 39A-47A.

309. G Kasper, N Dankert, J Tuischer, M Hoeft, T Gaber, J D Glaeser, D Zander, M Tschirschmann, M Thompson, G Matziolis and G N Duda. Mesenchymal stem cells regulate angiogenesis according to their mechanical environment. Stem Cells. 2007, Vol. 25, 4, pp. 903-910.

310. M D Brown and O Hudlicka. Modulation of physiological angiogenesis in skeletal muscle by mechanical forces: Involvement of VEGF and metalloproteinases. Angiogenesis. 2003, Vol. 6, 1, pp. 1-14.

311. J K Leach, D Kaigler, Z Wang, P H Krebsbach and D J Mooney. Coating of VEGF-releasing scaffolds with bioactive glass for angiogenesis and bone regeneration. Biomaterials. 2006, Vol. 27, 17, pp. 3249-3255.

312. B Behr, P Leucht, M T Longaker and N Quarto. Fgf-9 is required for angiogenesis and osteogenesis in long bone repair. Proceedings of the national academy of sciences of the united states of america. 2010, Vol. 107, 26, pp. 11853-11858.

313. R Maliba, S Lapointe, P E Neagoe, A Brkovic and M G Sirois. Angiopoietins-1 and -2 are both capable of mediating endothelial PAF synthesis: Intracellular signalling pathways. Cellular signalling. 2006, Vol. 18, 11, pp. 1947-1957.

314. L A van Meeteren, M J Goumans and P ten Dijke. TGF-beta receptor signaling pathways in angiogenesis: Emerging targets for anti-angiogenesis therapy. Current Pharmaceutical Biotechnology. 2011, Vol. 12, 12, pp. 2108-2120.

315. N Y Lee, C Golzio, C E Gatza, A Sharma, N Katsanis and G C Blobe. Endoglin regulates PI3- kinase/Akt trafficking and signaling to alter endothelial capillary stability during angiogenesis. Molecular Biology of the Cell. 2012, Vol. 23, 13, pp. 2412-2423.

316. R M Day. Bioactive glass stimulates the secretion of angiogenic growth factors and angiogenesis in vitro. Tissue Engineering. 2005, Vol. 11, 5-6, pp. 768-777.

317. A A Gorustovich, J A Roether and A R Boccaccini. Effect of bioactive glasses on angiogenesis: A review of in vitro and in vivo evidences. Tissue Engineering Part B - Reviews. 2010, Vol. 16, 2, pp. 199-207.

318. A Arkudas, A Balzer, G Buehrer, I Arnold, A Hoppe, R Detsch, P Newby, T Fey, R E Horch, A Boccaccini and U Kneaser. Evaluation of Angiogenesis of Bioactive Glass in the AV Loop Model.

300

319. L C Yu, S C Chen, W C Chang, Y C Huang, K M Lin, P H Lai and H W Sung. Stability of angiogenic agents, ginsenoside Rg, and Re, isolated from Panax ginseng: In vitro and in vivo studies. International Journal of Pharmaceutics. 2007, Vol. 328, 2, pp. 168-176.

320. J Cash, A Korchnak, J Gorman, Y Tandon and G Fraizer. VEGF transcription and mRNA stability are altered by WT1 not DDS(R384W) expression in LNCaP cells. Oncology Reports. 2007, Vol. 17, 6, pp. 1413- 1419.

321. T Davis-Smyth, L G Presta and N Ferrara. Mapping the charged residues in the second immunoglobulin-like domain of the vascular endothelial growth factor placenta growth factor receptor Flt-1 required for binding and structural stability. Journal of Biological Chemistry. 1998, Vol. 273, 6, pp. 3216-3222.

322. L D D'Andrea, A Romanelli, R Di Stasi and C Pedone. Bioinorganic aspects of angiogenesis. Dalton Transactions. 2010, Vol. 39, 33, pp. 7625-7636.

323. A Nasulewiez, A Mazur and A Opolski. Role of copper in tumour angiogenesis - clinical implications. Journal of Trace Elements in Medicine and Biology. 2004, Vol. 18, 1, pp. 1-8.

324. G F Hu. Copper stimulates proliferation of human endothelial cells under culture. Journal of Cellular Biochemistry. 1998, Vol. 69, 3, pp. 326-335.

325. C Gerard, L J Bordeleau, J Barralet and C J Doillon. The stimulation of angiogenesis and collagen deposition by copper. Biomaterials. 2010, Vol. 31, 5, pp. 824-831.

326. C Stahli, N Muja and S N Nazhat. Controlled copper ion release from phosphate-based glasses improves human vein endothelial cell survival in a reduced nutrient environment. Tissue engineering Part A. 2013, Vol. 19, 3-4, pp. 548-557.

327. J Li, J Suo, X Huang, C Ye and X Wu. Release behaviour of copper ion in a novel contraceptive. Contraception. 2007, Vol. 76, pp. 233-237.

328. L L Wang, K M Wang, X Z Liu and Q M Lu. Monomode channel waveguide in KTiOPO4 crystal produced by Rb+ ion exchange combined with Si+ implantation. Applied physics B: Lasers and Optics. 2010, Vol. 99, 1-2, pp. 203-207.

329. O Barkman, V Jerabek and V Prajzler. Optical spliters based on self-imaging effect in multi-mode waveguide made by ion exchange in glass. Radioengineering. 2013, Vol. 22, 1, pp. 352-356.

330. C Maurizio, F D'Acapito, C Sada, E Cattaruzza, F Gonella and G Battaglin. Site transition on metal ions in ion-exchanged metal doped glasses. Materials Science and Engineering B - Advanced functional solid- state materials. 2008, Vol. 149, 2, pp. 171-176.

331. J Spirkova, P Tresnakova, H Malichova and M Mika. Cu+ containing waveguiding layers in a novel silicate glass substrates. Journal of Materials Science - Materials in electronics. 2007, Vol. 18, 1, pp. S375- S378.

332. A I Fu and J C Mauro. Mutual diffusivity, network dilation, and salt bath poisoning effects in ion exchanged glass. Journal of Non-crystalline solids. 2013, Vol. 363, pp. 199-204.

333. J Lee, T Yano, S Shibata, A Nukui and M Yamane. EXAFS study on the local environment of Cu+ ions in glasses of the Cu2-Na2O-Al2O3-SiO2 system prepared by Cu+/Na+ ion exchange. Journal of Non- crystalline solids. 2000, Vol. 277, 2-3, pp. 155-161.

334. A Quaranta, E Catteruzza and F Gonella. Modelling the ion exchange process in glass: Phenomenological approaches and perspectives. Materials Science and Engineering B - Advanced functional solid-state materials. 2008, Vol. 149, 2, pp. 133-139.

301

335. R Gy. Ion Exchange for glass strengthening. Materials Science and Engineering: B. 2008, Vol. 149, 2, pp. 159-165.

336. A K Varshneya. Chemical strengthening of glass: Lessons learned and yet to be learned. International Journal of Applied glass science. 2010, Vol. 1, 2, pp. 131-142.

337. K K Mallick and D Holland. Strengthening of container glasses by ion-exchange dip coating. Journal of Non-crystalline solids. 2005, Vol. 351, pp. 2524-2536.

338. P L Inacio, B J Barreto, F Horowitz, R R B Correia and M B Pereira. Silver migration at the surface of ion-exchanged waveguides: a plasmonic template. Optical Materials Express. 2013, Vol. 3, 3, pp. 390-399.

339. J H Lopes, I O Mazali, R Landers and C A Bertran. Structural investigation of the structure of Bioglass 45S5 enriched with calcium ions. Journal of American Ceramic Society. 2013, Vol. 96, 5, pp. 1464-1469.

340. B Akkopru and C Durucan. Preparation and microstructure of sol-gel derived silver-doped silica. Journal of Sol-Gel Science and Technology. 2007, Vol. 43, 2, pp. 227-236.

341. I Ahmed, E A Abou Neel, S P Valappil, S N Nazhat, D M Pickup, D Carta, D L Carroll, R J Newport, M E Smith and J C Knowles. The structure and properties of silver-doped phosphate-based glasses. Journal of Materials Science. 2007, Vol. 42, 23, pp. 9827-9835.

342. R Oven, M Yin and P A Davies. Characterisation of planar optical waveguides formed by copper-sodium, electric field assisted, ion exchange in glass. Journal of Physics D: Applied Physics. 2004, Vol. 37, pp. 2207- 2215.

343. O Bretcanu, Q Chen, S Misra, A R Boccaccini, I Roy, E Verne and C V Brovarone. Biodegradable polymer coated 45S5 Bioglass-derived glass-ceramic scaffolds for bone tissue engineering. Glass technology- European Journal of glass science and technology part A. 2007, Vol. 48, 5, pp. 227-234.

344. F Bouville, E Maire, S Meille, B Van de Moortele, A J Stevenson and S Deville. Strong, tough and stiff bioinspired ceramics from brittle constituents. Nature. 2014, Vol. 13, pp. 508-514.

345. C Gao, T Liu, C Shuai and S Peng. Enhancement mechanisms of graphene in nano-58S bioactive glass scaffold: mechanical and biological performance. Scientific reports (nature). 4, 2014.

346. M Peroglio, L Gremillard, J Chevalier, L Chazeau, C Gauthier and T Hamaide. Toughening of bioceramics scaffolds by polymer coating. Journal of the European Ceramic Society. 2007, Vol. 27, 7, pp. 2679- 2685.

347. S Dumitriv and V I Popa. Polymeric Biomaterials: Structure and Function. s.l. : Taylor and Francis Group, 2013. p. 920.

348. R A Auras, L T Lim, S E M Selke and H Tsuji. Poly(lactic acid): Synethesis, structures, properties, processing and applications. s.l. : J Wiley and Sons, 2011. p. 528.

349. K Lee and D Kaplan. Tissue engineering I: Scaffold systems for tissue engineering. s.l. : Springer, 2006. p. 243.

350. F Bourya, P Saulnier, J E Proust, I Panaiotov, T Z Ivanova, C Postel and O Abillon. Characterisation of the morphology of poly(alpha-hydroxy acid)s Langmuir-Blodgett films by atom force microscopy measurements. Colliods and Surfaces A: Physicochemical and Engineering Aspects. 1999, Vol. 155, 2-3, pp. 117-129.

302

351. J J Blaker, S N N azhat, V Maquet and A R Boccaccini. Long term in vitro degredation of PDLLA/Bioglass(R) bone scaffolds in acellular simulated body fluid. Acta Biomaterialia. 2011, Vol. 7, 2, pp. 829-840.

352. O Bretcanu, A R Boccaccini and V Salih. Poly-DL-lactic acid coated Bioglass(R) scaffolds: Toughening effects and osteosarcoma cell proliferation. Journal of Material Science. 2012, Vol. 47, pp. 5661-5672.

353. J A Roether, J E Gough, A R Boccaccini, L L Hench, V Maquet and R Jerome. Novel bioresorbable and bioactive composites based on bioactive glass and poly lactide foams for bone tissue engineering. Journal of Materials Science - Materials in Medicine. 2002, Vol. 13, 12, pp. 1207-1214.

354. X Zhang, U P Wyss, D Pichora and M F A Goosen. A mechanistic study of antibiotic release from biodegradable poly(d,l-lactide) cylinders. Journal of Controlled Release. 1994, Vol. 31, pp. 129-144.

355. H Gollwitzer, K Ibrahim, H Meyer, W Mittelmeier, R Busch and A Stemberger. Antibacterial poly(D,L-lactic acid) coating of medical implants using a biodegradable drug delivery technology. Journal of Antimicrobial Chemotherapy. 2003, Vol. 51, pp. 585-591.

356. G Schmidmaier, B Wildemann, H Bail, M Lucke, T Fuchs, A Stemberger, A Flyvbjerg, N P Haas and M Raschke. Local Application of Growth Factors (Insulin-Like Growth Factor-1 and Transforming Growth Factor-b1) From a Biodegradable Poly(D,L-lactide) Coating of Osteosynthetic Implants Accelerates Fracture Healing in Rats. Bone. 2001, Vol. 28, 4, pp. 341-350.

357. B Liu, P Lin, Y Shen and Y Dong. Porous bioceramics reinforced by coating gelatin. Journal of Materials Science: Material in Medicine. 2008, Vol. 19, pp. 1203-1207.

358. G Tronci. Synthesis, Characterisation, and Biological Evaluation of gelatin-based scaffolds. Faculty of Mathematics and Natural Sciences, University of Potsdam. 2010. PhD.

359. R Mihai, I P Florescu, V Coroiu, A Oancea and M Lungu. In vitro biocompatibility testing of some synthetic polymers used for the achievement of nervous conduits. Journal of Medicine and Life. 2011, Vol. 4, 3, pp. 250-255.

360. Y Hong, T V Chirila, S Vijayasekaran, M J Davis and M A L Maley. In vitro cytotoxicity assay of synthetic hydrogels as potential vitreous substitutes. In Vitro Toxicology. 1996, Vol. 9, 1, pp. 73-82.

361. G Zhang, T Liu, Q Wang, L Chen, J Lei, J Luo, G Ma and Z Su. Mass spectrometric detection of marker peptides in tryptic digest of gelatin: A new method to differentiate between bovine and porcine gelatin. Food Hydrocolloids. 2009, Vol. 23, pp. 2001-2007.

362. Gelatin Manufacturers Institute of America. Gelatin Handbook. 2012.

363. A-L Metze, A Grimm, P Nooeaid, J A Roether, J Hum, P J Newby, D W Schubert and A R Boccaccini. Gelatin coated 45S5 Bioglass(R)-derived scaffolds for bone tissue engineering. Key Engineering Materials. 2013, Vol. 541, pp. 31-39.

364. X Liu, L A Smith, J Hu and P X Ma. Biomimetic nanofibrous gelatin/apatite composite scaffolds for bone tissue engineering. Biomaterials. 2009, Vol. 30, pp. 2252-2258.

365. M Y Shie, D C H Chen, C Y Wang, T Y Chiang and S J Ding. Immersion behavior of gelatin- containing calcium phosphate cement. Acta Biomaterialia. 2008, Vol. 4, pp. 646-655.

366. Y Tabata and Y Ikada. Vascularization effect of basic fibroblast growth factor released from gelatin hydrogels with different biodegradabilities. Biomaterials. 1999, Vol. 20, pp. 2169-2175.

303

367. Z S Patel, M Yamamoto, H Ueda, Y Tabata, and A G Mikos. Biodegradable gelatin microparticles as delivery systems for the controlled release of bone morphogenetic protein-2. Acta Biomaterialia. 2008, Vol. 4, pp. 1126-1138.

368. S Srinvasan, R Jayasree, K P Chennazhi, S V Nair and R Jayakumar. Biomcompatible alginate/nano bioglass glass ceramic composite scaffolds for periodontal tissue regeneration. Carbohydrate Polymers. 2012, Vol. 87, 1, pp. 274-283.

369. K T Shaluman, S Sowmya, D Sathish, K P Chennazhi, S V Nair and R Jayakumar. Effect of incorporation of nanoscale bioactive glass and hydroxyapatite in PCL/Chitosan nanofibers for bone and periodontal tissue engineering. Journal of Biomedical Nanotechnology. 2013, Vol. 9, 3, pp. 430-440.

370. V G Varanasi, J B Owyoung, E Saiz, S J Marshall, G W Marshall and P M Loomer. The ionic products of bioactive glass particle dissolution enhance periodontal ligament fibroblast osteocalciun expression and enhance early mineralised tissue development. Journal of Biomedical Materials Research Part A. 2011, Vol. 98A, 2, pp. 177-184.

371. A I Leal, S G Caridade, J L Ma, N Yu, M E Gomes, R L Reis, J A Janson, X F Walboomers and J F Mano. Asymmetric PDLLA membranes containing Bioglass(R) for guided tissue regeneration: Characterisation and in vitro biological behaviour. Dental Materials. 2013, Vol. 29, 4, pp. 427-436.

372. Z G Huan, S Leeflang, J Zhou, W Y Zhai, J Chang and J Duszczyk. In vitro degradation behaviour and bioactivity of magnesium-Bioglass(R) composites for orthopedic applications. Journal of Biomedical Materials Research Part B - Applied Biomaterials. 2012, Vol. 100B, 2, pp. 437-446.

373. X B B Yang, D Webb, J Blaker, A R Boccaccini, V Maquet, C Cooper and R O C Oreffo. Evaluation of human bone marrow stromal cell growth on biodegradable polymer/Bioglass(R) composites. Biochemical and Biophysical Research communications. 2006, Vol. 342, 4, pp. 1098-1107.

374. K F Bombanato-Prado, L S Bellesini, C M Junta, M M Marques, G A Passos and A L Rosa. Microarray-based gene expression analysis of human osteoblasts in response to different biomaterials. Journal of Biomedical Materials Research Part A. 2009, Vol. 88A, 2, pp. 401-408.

375. J F Hernadez-Sierra, O Galicia-Cruz, A Salina-Acosta, F Ruiz, M Pierdant-Perez and A J Pozos- Guillen. In vitro cytotoxicity of silver nanoparticles on human periodontal fibroblasts. Journal of Clinical Pediatric Dentistry. 2011, Vol. 36, 1, pp. 37-41.

376. R L Gamelli, T P Paxton and M Oreilly. Bone-marrow toxicity by silver sulfadiazine. Surgery gynecology and obsetetrics. 1993, Vol. 177, 2, pp. 115-120.

377. Y S Kim, J S Kim, H S Cho, D S Rha, J M Kim, J D Park, B S Choi, R Lim, H K Chang, Y H Chung, I H Kwan, J Jeong, B S Han and J Yu. Twenty-eight day oral toxicity, genotoxicity and gender- related tissue distribution of silver nanoparticles in Sprague-Dawley rats. Inhalation Toxicology. 2008, Vol. 20, 6, pp. 575- 583.

378. N Kinoshita, T Yamamura, H Teranuma, T Katayama, M Tamanyu, T Negoro, K Satoh and H Sakagami. Interaction between dental metals and antioxidants, assessed by cytotoxicity assay and ESR spectroscopy. Anticancer Research. 2002, Vol. 22, 6C, pp. 4017-4022.

379. S M San Miguel, L A Opperman, E P Allen, J E Zielinski and K K H Svoboda. Antioxidant combinations protect oral fibroblasts against metal-induced toxicity. Archives of Oral Biology. 2013, Vol. 58, 3, pp. 299-310.

380. A B Chuklovin, S V Tokalov, A S Yagunov and V D Zharskaya. Acute effect of copper, chromium and manganese upon immature blood cells and macrophages. Trace elements and electrolytes. 1996, Vol. 13, 1, pp. 37-41.

304

381. J P Rodriguez, S Rios and M Gonzalez. Modulation of the proliferation and differentiation of human mesenchymal stem cells by copper. Journal of cellular Biochemistry. 2002, Vol. 85, 1, pp. 92-100.

382. S Cox Gad. Development of Therapeutic Agents Handbook. s.l. : J Wiley and Sons, 2011. p. 1232.

383. G A Pankey and L D Sabath. Clinical relevance of bacteriostatic versus bactericidal mechanisms of action in the treatment of gram-positive bacterial infections. Clinical Infectious Diseases. 2004, Vol. 38, 6, pp. 864-870.

384. M R J Salton and K S Kim. Chapter 2 - Structure. [ed.] S Baron. Medical Microbiology. 1996.

385. S C Parija. Textbook of microbiology and Immunology. s.l. : Elsevier India, 2009. p. 700.

386. J Rundegren, T Sjodin, L Petersson, E Hansson and I Jonsson. Delmopinol interactions with cell walls of gram-negative and gram-positive oral bacteria. Oral microbiology and immunology. 1995, Vol. 10, 2, pp. 102-109.

387. G Seltmann and O Holst. The Bacterial Cell Wall. s.l. : Springer, 2002. p. 280.

388. N Cioffi and M Rai. Nano-Antimicrobials: Progress and prospects. s.l. : Springer, 2012. p. 556.

389. K K Mallick. Development synthesis and characterisation of porous biomaterial scaffold for tissue engineering. [book auth.] P N Kumta and W R Wagner R J Narayon. Advances in Biomaterials and Biomimetic Materials: a collection of papers presented at the 2008 Materials science and technology conference. s.l. : American Ceramic Society , 2009, pp. 115-128.

390. Sigma Aldrich. Manufacturers data for gelatin type A. 2012.

391. —. Manufacturers data for Gelatin type B. 2012.

392. J J Blaker, V Maquet, R J Jerome, A R Boccaccini and S N Nazhat. Mechanical properties of higly porous PDLLA/Bioglass(R) composite foams as scaffolds for bone tissue engineering. Acta Materiallia. 2005, Vol. 1, 6, pp. 643-652.

393. D Bellucci, A Sola, P Gentile, G Ciardelli and V Cannillo. Biomimetic coating on bioactive glass- derived scaffolds mimicking bone tissue. Journal of Biomedical Materials Research Part A. 2012, Vol. 100A, 12, pp. 3259-3266.

394. H Yuan, J D de Brujin, X Zhang, C A van Blitterswijk and K de Groot. Bone induction by porous glass ceramic made from bioglass (45S5). Journal of Biomedical Materials Research. 2001, Vol. 58, 3, pp. 270- 276.

395. M Tanahashi, T Yao, T Kokubo, M Minoda, T Miyamoto, T Nakamura and T Yamamuro. Apatite Coating on Organic Polymers by a Biomimetic Process. Journal of the American Ceramic Society. 1994, Vol. 77, 11, pp. 2805-2808.

396. T Kokubo. A/W glass-ceramic: processing and properties. [ed.] L L Hench and J Wilson. An Introduction to bioceramics. Singapore : World Scientific, 1993.

397. L Reimer. Scanning electron microscopy: physics of image formation and microanalysis. s.l. : Springer, 1998. pp. 1-12.

398. M L Henson and T A Jergovich. Scanning electron microscopy and energy dispersive X-ray spectrometry (SEM/EDS) for the forensic examination of paints and coatings. [ed.] B Caddy. Forensic examination of glass and paint: analysis and interpretation. s.l. : Taylor and Francis, 2001, 11, pp. 243-272.

305

399. C Suryanarayana and M Grant Norton. X-ray Diffraction: A practical approach. s.l. : Plenum Press, 1998. pp. 21-96.

400. B C Smith. Fundamentals of Fourier transform infrared spectroscopy. s.l. : CRC Press LLC, 1996.

401. Zygo(R) corporation. MetroPro Refernce Guide. 2006. Vol. 1, K, p. 517.

402. M J Somerman, S Y Archer, G R Imm and R A Foster. A comparative - study of human periodontal- ligament cells and gingival fibroblasts in vitro. Journal of Dental Research. 1988, Vol. 67, 1, pp. 66-70.

403. L F Cao, L Krymskaya, V Tran, S Mi, M C Jensen, S Blanchard and M Kalos. Development and application of a multiplexable flow cytometry based assay to quantify cell-mediated cytolysis. Cytometry Part A. 2010, Vol. 77A, 6, pp. 534-545.

404. J C Park, Y S Hwang and H Shu. Viability evaluation of engineered tissue. Yonsei Medical Journal. 2000, Vol. 41, 6, pp. 836-844.

405. C Luo, J G Long and J K Liu. An improved spectrophotometric method for a more specific and accurate assay of mitochondrial complex III activity. Clinica Chimera Acta. 2008, Vol. 395, pp. 38-41.

406. C B Zeng, J M Rothfuss, J Zhang, S Vangvervang, W H Chu, S H Li, Z D Tu, J B Xu and R H Mach. Functional assays to define agonists and antagonists of the sigma-2 receptor. Analytical Biochemistry. 2014, Vol. 448, pp. 68-74.

407. L Qian, R S Lin, Y X Ping, M H Sheng and Y Z Hong. A two-dimensional study of electrophoresis protocol suitable for proteomic study of stem in Zizania Iatifolia Turcz. Acta Agricultrae Zhejiangensis. 2010, Vol. 22, 3, pp. 281-286.

408. Dissolution characteristics of bioactive glasses. R S Pryce and L L Hench. [ed.] D Sher and W Walsh B Nissan. Sydney : s.n., 2003. 15th International Symposium on Ceramics in Medicine. Vols. 240-242, pp. 201- 204.

409. F Heidenau, W Mittelmeier, R Detsch, M Haenle, F Stenzel, G Ziegler and H Gollwitzer. A novel antibacterial titania coating: Metal ion toxicity and in vitro surface colonization. 2005, Vol. 16, pp. 883-888.

410. F Baino and C Vitale-Brovarone. Three-dimensional glass-derived scaffolds for bone tissue engineering: Current trends and forecasts for the future. 2011, Vol. 97A, 4, pp. 514-535.

411. C Vitale-Brovarane, F Baino, O Bretcanu and E Verne. Foam-like scaffolds for bone tissue engineering based on a novel couple of silicate-phosphate specular glasses: synthesis and properties. 2009, Vol. 20, pp. 2197-2205.

412. J J Blaker, V Maquet, A R Boccaccini, R Jerome and A Bismarck. Wetting of bioactive glass surfaces by poly(alpha-hydroxyacid) melts: interaction between Bioglass and biodegradable polymers. 2005, Vol. 23.

413. D C Clupper and L L Hench. Bioactive response of Ag-doped tape cast Bioglass(R) 45S5 following heat treatment. Journal of materials science: materials in medicine. 2001, Vol. 12, pp. 917-921.

414. O Bretcanu, X Chatzistavrou, K Paraskevopoulos, R Conradt, I Thompson and A R Boccaccini. Sintering and crystallisation of 45S5 Bioglass powder. 2009, Vol. 29, 16, pp. 3299-3306.

415. M Cerruti, D Greenspan and K Powers. Effects of pH and ionic strength on the reactivity of Bioglass(R) 45S5. 2005, Vol. 26, 14, pp. 1665-1674.

416. A Sola, D Bellucci, M G Raucci, S Zeppetelli, L Ambrosio and V Cannillo. Heat treatment of Na2O- CaO-P2O5-SiO2 bioactive glasses: Densification processes and postsintering bioactivity. 2011, Vol. 100A, 2, pp. 305-322.

306

417. Q Z Chen, K Rezwan, V Francon, D Armitage, S N Nazhat, F H Jones and A R Boccaccini. Surface functionalization of Bioglass(R)- derived porous scaffolds. Acta Biomaterialia. 2007, Vol. 3, 4, pp. 551-562.

418. D Bellucci, V Cannillo, A Sola, F Chiellini, M Gazzarri and C Migone. Macroporous Bioglass(R)- derived scaffolds for bone tissue regeneration. Ceramics International. 2011, Vol. 37, 5, pp. 1575-1585.

419. K L Menzies and L Jones. The impact of contact angle on the biocompatibility of biomaterials. Optometry and vision science. 2010, Vol. 87, 6, pp. 387-399.

420. J A Juhasz and S M Best. Bioactive ceramics: processing, structures and properties. Journal of Material Science. 2012, Vol. 47, pp. 610-624.

421. Z Zhou, J Ruan, J Zou and Z Zhou. Preparation and bioactivity of sol-gel macroporous bioactive glass. Journal of University of Science and Technology Bejing. 2008, Vol. 15, 3, pp. 290-297.

422. D Bellucci, F Chiellini, G Ciardelli, M Gazzarri, P Gentile, A Sola and V Cannillo. Processing and characterisation of innovative scaffolds for bone tissue engineering. Journal of Material Science: Materials in Medicine. 2012, Vol. 23, pp. 1397-1409.

423. G E Stan, A C Popa and D Bojin. Bioreactivity evaluation in simulated body fluid of magnetron sputtered glass and glass-ceramic: A FTIR spectroscopy study. Digest Journal of Nanomaterials and Biostructures. 2010, Vol. 5, 2, pp. 557-566.

424. I Rehman, J C Knowles and W Bonfield. Analysis of in vitro reaction layers formed on Bioglass(R) using thin-film x-ray diffraction and ATR-FTIR microspectroscopy. Journal of Biomedical Materials Research. 1998, Vol. 41, 1, pp. 162-166.

425. E Kontonasaki, T Zorba, L Papadopoulou, E Pavlidou, X Chatzistavrou, K Paraskevopoulous and P Koidis. Hydroxy Carbonate apatite formation on particulate Bioglass in vitro as a function of time. Crystal Research and Technology. 2002, Vol. 37, 11, pp. 1165-1171.

426. Q Z Chen, K Rezwan, D Armitage, S N Nazhat and A R Boccaccini. The surface functionalization of 45S5 Bioglass(R) -based glass-ceramic scaffolds and its impact on bioactivity. Journal of Material Science: Materials in Medicine. 2006, Vol. 17, pp. 979-987.

427. I A Silver and M Erecinska. Interactions of osteoblastic and other cells with bioactive glasses and silica in vitro and in vivo. Muterialwissenschaft und werkstofftechnik. 2003, Vol. 34, 12, pp. 1069-1075.

428. C S Kim, E K Park, S G Kim. Silica-Silver nano structure spheres prepared by spray pyrolysis of sol containing silver precursor. Journal of Sol-gel technology. 2008, Vol. 47, 1, pp. 7-15.

429. R S Varna, D C Kothari and R Tewari. Nano-composite soda lime silicate glass prepared using silver ion exchange. Journal of Non-crystalline solids. Vol. 355, 22-23, pp. 1246-1251.

430. A Vulpoi, L Baia, S Simon and V Simon. Silver effect on the structure of SiO2-CaO-P2O5 ternary system. Materials Science and Engineering: C. 2012, Vol. 32, 2, pp. 178-183.

431. E Verne, S Di Nunzio, M Bosetti, P Appendino, C Vitale Brovarone, G Maina and M Cannas. Surface characterisation of silver-doped bioactive glass. Biomaterials. 2005, Vol. 26, 25, pp. 5111-5119.

432. S Matsusaka and T Watanabe. Micro-machinability of silver-sodium ion exchanged glass by UV nanosecond laser. Journal of Materials Processing Technology. 2008, Vol. 202, 1-3, pp. 514-520.

433. A Berger. Concentration and size depth profile of colloidal silver particles in glass surfaces produced by sodium silver ion exchange. Journal of Non-Crystalline solids. 1992, Vol. 151, 1-2, pp. 88-94.

307

434. S Karlsson, B Jonson, L Wondraczek. Copper, silver, rubidium and caesium ion exchange in soda-lime silica float glass by direct deposition and in line melting of salt pastes. Glass technology - European Journal of Glass technology Part A. 2012, Vol. 53, 1, pp. 1-7.

435. M Martin, J J Videau, L Canioni, F Adamietz, L Sarger and G Le Flen. Planar wavequides formed by Ag+ - Na+ ion exchange in non linear optical glasses: diffusion and optical properties. Applied Optics. 2000, Vol. 39, 3, pp. 435-440.

436. H Fischer, M Brehme, R Telle and R Marx. Effect of ion exchange of glazed dental glass-ceramics on strength parameters. Journal of Biomedical Materials Research; Part A. 2005, Vol. 72A, 2, pp. 175-179.

437. A R Cooper and D A Krohm. Strengthening of glass fibres II: Ion Ex-change. Journal of American Cermic Society. 1969, Vol. 52, 12, pp. 665-669.

438. T Yano, K Azegami, S Shibata and M Yamane. Chemical state of oxygen in Ag+/Na+ ion-exchanged sodium silicate glass. 1997, pp. 94-101.

439. X Wu, J D Li, L Wang, D Huang, Y Zuo and Y B Li. The release properties of silver ions from Ag- nHA/TiO2/PA66 antimicrobial composite scaffolds. Biomedical Materials. 2010, Vol. 5, 4.

440. Y K Chen, X B Zhang, Y T Xie, H Ji, C X Ding, H W Li and K R Dai. Silver release from silver- containing hydroxyapatite coatings. Surface and coatings technology. 2010, Vol. 207, 7, pp. 1892-1896.

441. R Stefan, E Culea and P Pascuta. The effect of copper ions additon on structural and optical properties of zinc borate glasses. Journal of Non-Crystalline Solids. 2012, Vol. 358, 4, pp. 836-846.

442. A A Soliman and I Kashif. Copper oxide content dependence of crystallisation behavior, glass forming ability, glass stability and fragility of lithium borate glasses. Physica B - Condensed Matter. 2010, Vol. 405, 1, pp. 247-253.

443. X Li, X P Wang, D N He and J L Shi. Synthesis and characterisation of mesoporous CaO-MO-SiO(2)- P(2)O(5) (M=Mg, Zn, Cu) bioactive glasses/composites. Journal of Materials Chemistry. 2008, Vol. 18, 4, pp. 4103-4109.

444. P Y Shin, J Y Ding and S Y Lee. 31MAS-NMR and FTIR analysis on the structure of CuO- containing sodium poly- and meta- phosphate glasses. Materials Chemistry and Physics. 2003, Vol. 80, pp. 391-396.

445. A Chahine and M Et-Tabina. Glass transition (Tg) and FTIR study of Na2O-CuO-Bi2O3-P2O5 glasses. Phase Transitions. 2002, Vol. 75, 3, pp. 309-314.

446. J Sulowska and I Walawska. Structural role of Cu in the soil active glasses. Processing and Application of ceramics. 2012, Vol. 6, 2, pp. 77-82.

447. W Su, L Yao, F Yang, P Li, J Chen and L Liang. Electroless plating of copper on surface modified glass substrate. Applied Surface Science. 2011, Vol. 257, 18, pp. 8067-8071.

448. B Macalik, L Krajczyk, J Okal, T Morawska-Kowal, K D Nierzewski and M Suszynska. Preparation and optical properties of soda-lime silicate glasses partially substituted by copper. Radiation effects and defects in solids. 2002, Vol. 157, 6-12, pp. 887-893.

449. J Sulowska, I Wacawska and M Szumera. Effect of copper addition on glass transition of silicate- phosphate glasses. Journal of Thermal Analysis and Calorimetry. 2012, Vol. 109, 2, pp. 705-710.

450. C Maurizio, F D'Acapito, C Sada, E Catteruzza, F Gonella and G Battaglin. Site transition metal ions in ion-exchanged metal-doped glasses. Materials Science and Engineering B: Advanced functional solid-state materials. 2008, Vol. 149, 2, pp. 171-176.

308

451. J R Jones, L M Ehrenfried, P Saravanapavan and L L Hench. Controlling ion release from bioactive glass foam scaffolds with antibacterial properties. Journal of Material Science: Materials in Medicine. 2006, Vol. 17, pp. 989-996.

452. E K Gamstedt and S Ostlund. Fatigue propagation of fibre-bridged cracks in unidirectional polymer- matrix composites. Applied composite materials. 2001, Vol. 8, pp. 385-410.

453. J A De Souza, S Goutianos, M Skovgaard and B F Sorensen. Fracture resistance curves and toughening mechanisms in polymer based dental composites. Journal of the Mechanical Behaviour of Biomedical Materials. 2011, Vol. 4, 4, pp. 558-571.

454. M Seshadri and S Saigal. Crack bridging in Polymer Nanocomposites. Journal of Engineering Mechanics. 2007, Vol. 133, 8, pp. 911-918.

455. X Zheng, S Zhou, Y Xiao, X Yu, X Liu and P Wu. Shape memory effort of poly(D,L-lactide)/Fe3O4 nanocomposites by inductive heating of magnetite particles. Colloids and Surfaces B: Biointerfaces. 2009, Vol. 71, 1, pp. 67-72.

456. J H Muyonga, C G B Cole and K G Duodo. Fourier transform infrared (FTIR) spectroscopic study of acid soluble collagen and gelatin from skins and bones of young and adult Nile perch (lactes Niloticus). Food Chemistry. 2004, Vol. 86, 3, pp. 325-332.

457. S M Paude, K Deshmuki and S Parab. Preparation and characterisation of poly(vinyl alcohol) and gelatin blend films. Journal of Applied Polymer Science. 2008, Vol. 109, 2, pp. 1328-1337.

458. M M Rahman, S Pervez, B Nesa and M A Khan. Preparation and characterisation of porous scaffold composite films by blending chitosan and gelatin solutions for skin tissue engineering. Polymer International. 2013, Vol. 62, 1, pp. 79-86.

459. G K Hunter, S C Nyburg and K P H Pritzker. Hydroxyapatite formation in collagen, gelatin and agarose gels. Collagen and related research. 1986, Vol. 6, 3, pp. 229-238.

460. M C Chang and R DeLong. Calcium phosphate formation in gelatin matrix using free ion precursors of Ca2+ and phosphate ions. Dental Materials. 2009, Vol. 25, 2, pp. 261-268.

461. A L Rosa and M M Beloti. Effect of cpTi surface roughness on human bone marrow cell attachment, proliferation and differentiation. Brazilian Dental Journal. 2003, Vol. 14, 1, pp. 16-21.

462. H Schweikl, R Muller, C Englert, K-A Hiller, R Kujat, M Nerlich and G Schmalz. Proliferation of osteoblasts and fibroblasts on model surfaces of varying roughness and surface chemistry. Journal of Material Science: Materials in Medicine. 2007, Vol. 18, pp. 1895-1905.

463. S Proum and M Weinberg. Bioglass bone grafting material in the treatment of human periodontal defects. Journal of Dental Research. 1997, Vol. 76.

464. X Y Yang, Y L Gan, X Gao, L Zhao, C Y Gao, X L Zhang, Y B Feng, K Ting and Z R Gou. Preparation and characterisation of trace elements-multidoped injectable biomimetic materials for minimally invasive treatment of osteoporotic bone trauma. Journal of Biomedical Materials Research Part A. 2010, Vol. 95A, 4, pp. 1170-1181.

465. E Kontonasaki, A Sivropoulou, L Papadopoulou, P Garefis, K Paraskevopoulous and P Koidis. Attachment and proliferation of human periodontal ligament fibroblasts on bioactive glass modified cermics. 2007, Vol. 34, pp. 57-67.

466. P J Doolin. Medical assisting made incredibly easy: Lab competencies. s.l. : Lippincott Williams and Wilkins, 2007.

309

467. V Kothari, B Darji, M Doshi and J Ratani. Antimicrobial plant products for management of phytopathogens. s.l. : GRIN, 2012.

468. T N Kim, Q L Feng, J O Kim, J Wu, H Wang, C H Chen, and F Z Cui. Antimicrobial effects of metal ions (Ag+, Cu2+, Zn2+) in hydroxyapatite. 1998, Vol. 9, pp. 129-134.

469. P Lalueza, M Monzon, M Arruebo and J Santamaria. Bactericidal effects of different silver containing materials. 2011, Vol. 46, p. 2070=2076.

470. I Sawada, R Fachrul, T Ito, Y Ohmukai, T Maruyama and H Matsuyama. Development of a hydrophilic polymer membrane containing silver nanoparticles with both organic antifouling and antibacterial properties. Journal of Membrane science. 2012, Vol. 387, pp. 1-6.

471. X M Tang, S H Flint, R J Bennett, J D Brooks and R H Morton. Biofilm growth of individual and dual strains of Klebsiella oxytoca from the dairy industry on ultrafiltration membranes. Journal of Industrial Microbiology and Biotechnology. 2009, Vol. 36, 12, pp. 1491-1497.

472. X Tang, S H Flint, R J Bennett and J D Brooks. The efficacy of different cleaners and sanitisers in cleaning biofilms on UF membranes used in the dairy industry. Journal of Membrane science. 2010, Vol. 352, 1-2, pp. 71-75.

473. N Siti-Ismail, A Samadikuchaksaraei, A E Bishop, J M Polak and A Mantalaris. Development of a novel three-dimensional, automatable and integrated bioprocess for differentiation of embryonic stem cells into pulmonary aveolar cells in a rotating vessel bioreactor system. Tissue engineering part C: Methods. 2012, Vol. 18, 4, pp. 263-272.

474. A Mantalaris. Bioprocess systems engineering: bridging the 'scales' between 'molecules, cells and processes'. Microbial biotechnology. 2011, Vol. 4, 2, pp. 121-124.

475. B Carpentier, P Layrolle and C Legallais. Bioreactors for bone tissue engineering. International Journal of Artificial organs. 2011, Vol. 34, 3, pp. 259-270.

476. X Chatzistavrou, O Tsigkou, H D Amin, K M Paraskevopoulos, V Salih and A R Boccaccini. Sol-gel based fabrication and characterisation of new bioactive glass-ceramic composites for dental applications. Journal of the European Ceramic Society. 2012, Vol. 32, 12, pp. 3051-3061.

477. C M Revell, J A Dietrich, C C Scott, A Luttage, L S Baggett and K A Athanasiou. Characterisation of fibroblast morphology on bioactive surfaces using vertical scanning interferometry. Matrix Biology. 2006, Vol. 25, 8, pp. 523-533.

478. V C Modglin, R F Brown, S B Jung and D E Day. Cytotoxicity assessment of modified bioactive glasses with MLO-A5 osteogenic cells in vitro. Journal of Materials Science: Materials in Medicine. 2013, Vol. 24, 5, pp. 1191-1199.

479. H Fu, Q Fu, N Zhou, W Huang, M N Rahaman, D Wang and X Liu. In vitro evaluation of borate-based bioative glass scaffolds prepared by a polymer foam replication method. 2009, Vol. 29, pp. 2275-2281.

480. J Li, S Cai, G Xu, X Li, W Zhang and Z Zhang. In vitro biocompatibility study of calcium phosphate glass ceramic scaffolds with different trace element doping. 2012, Vol. 32, pp. 356-363.

481. N J Lakhkar, J H Park, N J Mordan, V Salih, I B Wall, H W Kim, S P King, J V Hanna, R A Martin, O Addison, J F W Mosselmans and J C Knowles. Titanium phosphate glass microspheres for boen tissue engineering. 2012, Vol. 8, pp. 4181-4190.

482. Y Zhou, H Li, K Lin, W Zhai, W Gu and J Chang. Effect of heat treatment on the properties of SiO2- CaO-MgO-P2O5 bioactive glasses. 2012, Vol. 23, pp. 2101-2108.

310

483. T P Xiu, Q Liu and J C Wang. Comparison between surfactant-templated mesoporous and conventional sol-gel-derived CaO-B2O3-SiO2 glasses; Compositional, textural and in vitro bioactive properties. Journal of Solid State Chemistry. 2008, Vol. 181, 4, pp. 863-870.

484. S Shruti, A J Salinas, G Lusvardi, G Malavasi, L Menabue and M Vallet-Regi. Mesoporous bioactive scaffolds prepared with cerium-, gallium and zinc-containing glasses. Acta Biomaterialia. 2012, Vol. 9, 1, pp. 4836-4844.

485. C Vaid, S Murugavel, R Kashayap and R P Tandon. Synthesis and in vitro bioactivity of surfactant templated mesoporous sodium silicate glasses. Microporous and Mesoporous Materials. 2012, Vol. 159, pp. 17- 23.

486. M R Foroughi, S Karbasi and R Ebrahimi-Kahrizangi. Physical and mechanical properties of poly-3- hydroxybutyrate- coated nanocrystalline hydroxyapatite scaffolds for bone tissue engineering. 2012, Vol. 19, pp. 667-675.

487. Q Z Chen, Y A Li, L Y Jin, J M W Quinn and P A Komesaroff. A new sol-gel process for producing Na2O-containing bioactive glass ceramics. Acta Biomaterialia. 2010, Vol. 6, 10, pp. 4143-4153.

488. P Gentile, M Mattioli-Belmonte, V Chiono, C Ferretti, F Baino, C Tonda-Turo, C Vitale-Brovarone, I Pashkuleva, R L Reis and G Ciardelli. Bioatcive glass/polymer composite scaffolds mimicking bone tissue. Journal of Biomedicial Materials Research Part A. 2012, Vol. 100A, 10, pp. 2654-2667.

489. R R Hao, D P Wang, A H Yao and W H Huang. Preparation and Characterisation of beta- TCP/CS scaffolds by Freeze-extraction and Freeze-gelation. Journal of Wuhan University of technology- Materials Science edition. 2011, Vol. 26, 2, pp. 372-376.

490. X Wu, Y Liu, X Li, P Wen, Y Zhang, Y Long, X Wang, Y Guo, F Xing and J Gao. Preparation of aligned porous gelatin scaffolds by unidirectional freeze-drying method. Acta Biomaterialia. 2010, Vol. 6, 3, pp. 1167-1177.

491. M Zhu, J L Shi, Q J He, L X Zhang, F Chang and Y Chen. An emulsification-solvent evapouration route to mesoporous bioactive glass microspheres for bisphosphonate drug delivery. Journal of Materials Science. 2012, Vol. 47, 5, pp. 2256-2263.

492. C H Huang and R A Doong. Sugercane bagasse as the scaffold for mass production of hierarchically porous carbon monoliths by surface self-assembly. Microporous and Mesoporous Materials. 2012, Vol. 147, 1, pp. 47-52.

493. B Derby. Printing and Prototyping of Tissues and Scaffolds. Science. 2012, Vol. 338, 6109, pp. 921-926.

494. T Gredes, M Wrobel-Kwiatkowska, M Dominiak, T Gedrange and C Kunert-keil. Osteogenic capacity of transgenic flax scaffolds. Biomedical Engineering. 2012, Vol. 57, 1, pp. 53-58.

495. A Ruffini, S Sprio and A Tampieri. Study of hydrothermal transformation of wood-derived calcium carbonate into 3D hierarchically organised hydroxyapatite. Chemical Engineering Journal. 2013, Vol. 217, pp. 150-158.

496. S Soumya, K M Sajesh, R Jayakumar, S V Nair and K P Chennazhi. Development of phytochemical scaffold for bone tissue engineering using Cissus quadrangularis extract. Carbohydrate polymers. 2012, Vol. 87, 2, pp. 1787-1795.

497. S Cazalbou, D Eichert, X Ranz, C Drouet, C Combes, M F Harmand and C Rey. Ion exchanges in apatites for biomedical application. Journal of Materials Science - Materials in Medicine. 2005, Vol. 16, 5, pp. 405-409.

311

498. M Zhang, C Wu, K Lin, W Fan, L Chen, Y Xiao and J Chang. Biological responses of human bone marrow mesenchymal stem cells to Sr-M-Si (M= Zn, Mg) silicate bioceramics. 2012, Vol. 100A, pp. 2979- 2990.

499. S Ohnishi, T Yasuda, S Kitamura and N Nagaya. Effect of hypoxia on gene expression of bone marrow- derived mesenchymal stem cells and mononuclear cells. 2007, Vol. 25, 5, pp. 1166-1177.

500. V Morino, J P Cattalini and A R Boccaccini. Metallic ions as therapeutic agents in tissue engineering scaffolds: an overview of their biological applications and strategies for new developments. 2012, Vol. 9, pp. 401-419.

501. H Kawamura, A Ito, S Miyakawa, P Layrolle, K Ojima, N Ichinose and T Tateishi. Stimulatory effect of Zinc-releasing calcium phosphate implant on bone formation in rabbit femora. 2000, Vol. 50, 2, pp. 184-190.

502. W Fan, R Crawford and Y Xiao. Enhancing in vivo vascularized bone formation by colbalt chloride- treated bone marrow stromal cells in tissue engineered periosteum model. 2010, Vol. 31, pp. 3580-3589.

503. X Wu, N Itoh, T Taniguchi, T Nakanishi, Y Tatsu, N Yumoto and K Tanaka. Zinc- induced sodium- dependant vitamin C transport 2 expression: potent roles in osteoblast differentiation. 2003, Vol. 420, pp. 114- 120.

504. R P Warrell, B Bosco, S Weinerman, B Levine, J Lane, R Bockman. Gallium nitrate for advanced paget disease of bone - Effectiveness and dose-response analysis. 1990, Vol. 113, 11, pp. 847-851.

505. R P Warrell, M Isaacs, C J Coonly, N W Alcock and R S Bockman. Metabolic effects of gallium nitrate adminstered by prolonged infusion. 1985, Vol. 69, 6, pp. 653-655.

506. R P Warrell, R S Bockman, C J Coonly, M Isaacs and H Staszewski. Gallium nitrate inhibits calcium resportion from bone and is effective treatment for cancer related hypercalcemia. 1984, Vol. 73, 5, pp. 1487- 1490.

507. L R Bernstein. Mechanisms of therapeutic activity for gallium. 1998, Vol. 50, 4, pp. 665-682.

508. E Verron, M Masson, S Khoshniat, L Duplomb, Y Wittrant, M Baud'huin, Z Badran, B Bujoli, P Janvier, J C Scimeca, J M Bouler and J Guicheux. Gallium modulates osteoblastic bone resorption in vitro without affecting osteoblasts. 2010, Vol. 159, 8, pp. 1681-1692.

509. E Verron, A Loubat, G F Carle, C Vignes-Colombeix, I Strazic, J Guicheux, N Rochet, J M Bouler and J C Scimeca. Molecular effects of gallium on osteoclastic differentiation of mouse and human monocytes. 2012, Vol. 83, 5, pp. 671-679.

510. S P Valappil, D Ready, E A Abou Neel, D M Pickup, L A O'Dell, W Chrzanowski, J Pratten, R J Newport, M E Smith, M Wilson, J C Knowles. Controlled delivery of antimicrobial gallium ions from phosphate-based glasses. 2012, Vol. 5, 4, pp. 1198-1210.

511. E A Abou Neel, W Chrzanowski, D M Pickup, L A O'Dell, N J Mordan, R J Newport, M E Smith and J C Knowles. Structure and properties of strontium-doped phosphate-based glasses. 2009, Vol. 6, 34, pp. 435-446.

512. M Nanami, T Ookawara, Y Otaki, K Ito, R Moriguchi, K Miyagawa, Y Hasuike, M Izumi, H Eguchi, K Suzuki and T Nakanishi. Tumor necrosis factor-alpha-induced iron sequestration and oxidatine stress in human endothelial cells. 2005, Vol. 25, 12, pp. 2495-2501.

513. T A Rouault. How mammals acquire and distribute iron needed for oxygen-based metabolism. 2003, Vol. 1, 3, pp. 326-328.

312

514. E A Abou Neel, I Ahmed, J Pratten, S N Nazhat and J C Knowles. Characterisation of antibacterial copper releasing degradable phosphate glass fibres. 2005, Vol. 26, 15, pp. 2247-2254.

515. I Ahmed, C A Collins, M P Lewis, I Olsen and J C Knowles. Processing, characterisation and biocompatibility of iron-phosphate glass fibres for tissue engineering. 2004, Vol. 25, 16, pp. 3223-3232.

516. R K Rude, H E Gruber, H J norton, L Y Wei, A Frausto and B G Mills. Bone loss induced by dietry magnesium reduction to 10% of the nutrient requirement in rats is associated with increased release of substance P and tumor necrosis factor-alpha. 2004, Vol. 134, 1, pp. 79-85.

517. R K Rude, H E Gruber, L Y Wei, A Frausto, and B G Mills. Magnesium deficiency: Effect on bone and mineral metabolism in the mouse. 2003, Vol. 72, 1, pp. 32-41.

518. A Hartwig. Role of magnesium in genomic stability. 2001, Vol. 475, 1-2, pp. 113-121.

519. Y Yamasaki, Y Yoshida, M Okazaki, A Shimazu, T Uchida, T Kubo, Y Akagawa, Y Hamada, J Takahashi and N Matsuura. Synthesis of functionally graded MgCO3 apatite accelerating osteoblast adhesion. 2002, Vol. 62, 1, pp. 99-105.

520. Y Yamasaki, Y Yoshida, M Okazaki, A Shimazu, T Kubo and Y Akagawa. Action of FgMgCO3- Ap- collagen composite in promoting bone formation. 2003, Vol. 24, pp. 4913-4920.

521. H Zreiqat, C R Howlett, A Zannettino, P Evans, G Schulze-Tanzil, C Knabe and M Shakibaei. Mechanisms of magnesium-stimulated adhesion of osteoblastic cells to commonly used orthopaedic implants. 2003, Vol. 62, 2, pp. 175-184.

522. N A Law, M T Caudle and V L Pecoraro. Manganese redox enzymes and model systems: Properties, structures and reactivity. [ed.] A G sykes. 1999, Vol. 46, pp. 305-440.

523. A Takeda. Manganese action in brain function. 2003, Vol. 41, 1, pp. 79-87.

524. D Bellucci, V Cannillo and A Sola. A new potassium-based bioactive glass: Sintering behaviour and possible applications for bioceramic scaffolds. 2011, Vol. 37, 1, pp. 145-157.

525. M Marangella, M Di Stefano, S Casalis, S Berutti, P D'Amelio and G C Isaia. Effects of potassium citrate supplementation on bone metabolism. 2004, Vol. 74, 4, pp. 330-335.

526. W Song, Q G Wang, C X Wan, T Shi, D Markel, R Blaiser and W P Ren. A novel alkali metals/strontium co-substituted calcium polyphosphate scaffolds in bone tissue engineering. 2011, Vol. 98B, 2, pp. 255-262.

527. G Boivin, P Deloffre, B Perrat, G Panczer, M Boudeulle, Y Mauras, P Allain, Y Tsoudevos and P J Meunier. Strontium distubution and interactions with bone mineral in monkey iliac bone after strontium salt (S12911) adminstration. 1996, Vol. 11, 9, pp. 1302-1311.

528. S G Dahl, P Allain, P J Marie, Y Mauras, G Boivin, P Ammann, Y Tsouderous, P D Delmas and C Christiansen. Incorporation and distribution of strontium in bone. 2001, Vol. 28, 4, pp. 446-453.

529. M D Grynpas, E Hamilton, R Cheung, Y Tsouderous, P Deloffre, M Holt and P J Marie. Strontium increases vertebral bone volume in rats at low dose that does not induce detectable mineralization defect. 1996, Vol. 18, 3, pp. 253-259.

530. P J Marie, P Ammann, G Boivin and C Rey. Mechanisms of action and therapeutic potential of stontrium in bone. 2001, Vol. 69, 3, pp. 121-129.

313

531. L L Zhu, S Zaidi, Y Z Peng, H Zhou, B S Moonga, A Blesius, I Dupin-Roger, M Zaidi and L Sun. Induction of a program gene expression during osteoblast differentiation with strontium ranelate. 2007, Vol. 355, 2, pp. 307-311.

532. C E Wen, Y Yamada, K Shimojima, Y Chino, H Hosokawa and M Mabuchi. Novel titanium foam for bone tissue engineering. 2002, Vol. 17, 10.

533. J C Wohlfahrt, M Monjo, H J Ronold, A M Aass, J E Ellingsen, S P Lyngstadaas. Porous titanium granules promote bone healing and growth in rabbit tibia peri-implant osseous deffects. 2010, Vol. 21, pp. 165- 173.

534. M Yoshinari, Y Oda, T Kato and K Okuda. Influence of surface modifications to titanium on antibacterial activity in vitro. 2001, Vol. 22, 14, pp. 2043-2048.

535. K H Orvig and C Thompson. Boon and bane of metal ions in medicine. 2003, Vol. 300, 5621, pp. 936- 939.

536. A S Tracey and D C Crans. Vanadium compounds: Chemistry, Biochemistry and Therapeutic applications. 1998, Vol. 711, pp. 82-103.

537. J Wang, V G Yuen and J H McNeill. Effect of vanadium on insulin sensitivity and appetite. 2001, Vol. 50, 6, pp. 667-673.

538. R K Upreti. Membrane vanadium interaction: A toxicokinetic evaluation. 1995, Vol. 153, 1-2, pp. 167- 171.

539. D A Barrio, E R Cattaneo, M C Apezteguia and S B Etcheverry. Vanadyl(IV) complexes with saccharides. Bioactivity in osteoblast-like cells in culture. 2006, Vol. 84, 7, pp. 765-775.

540. D Barrio and S B Etcheverry. Vanadium and bone development: putative signalling patheways. 2006, Vol. 84, 7, pp. 677-686.

541. A M Cortizo, M S Molinuero, D A Barrio and L Bruzzone. Osteogenic activity of vandyl(IV)-ascorbate complex: Evaluation of its mechanisms of action. 2006, Vol. 38, 7, pp. 1171-1180.

542. D A Barrio, P A M Williams, A M Cortizo and S B Etcheverry. Synthesis of a new vanadyl(IV) complex with trehalose (TreVO): insulin-mimetic activities in osteoblast-like cells in culture. 2003, Vol. 8, 4, pp. 459-468.

543. S B Etcheverry, D A Barrio, A M Coritzo and P A M Williams. Three new vanadyl(IV) complexes with non-steroidal anti-inflammatory drugs (Ibuprofen, Naproxen and Tolmetin). Bioactivity on osteoblast-like cells in culture. 2002, Vol. 88, 1, pp. 94-100.

544. S B Etcheverry and D A Barrio. Potential use of Vanadium compounds in therapeutics. 2010, Vol. 17, 31, pp. 3632-3642.

545. E Munthe, J Aaseth and E Jellum. Trace elements and rheumatoid arthritis(RA)- pathogentic and therapeutic aspects. 1986, Vol. 59, 7, pp. 365-373.

546. M Yamaguchi, H Oishi and Y Suketa. Zinc stimulation of bone protein synthesis in tissue culture: activation of aminoacyl-tRNA sythetase. 1988, Vol. 37, 21, pp. 4075-4080.

547. J Ovesen, B Moller-Madsen, J S Thomsen, G Damscher and L Mosekilde. The positive of zinc on skeletal strength in growing rats. 2001, Vol. 29, 6, pp. 565-570.

548. S Haimi, G Garianc, L Moimas, B Lindroos, H Huhtala, S Raty, H Kuokkanen, G K Sandor, C Schmid, S Miettinen and R Suuronen. Characterisation of zinc-releasing three dimensional bioactive glass

314 scaffolds and their effect on human adipose stem cell proliferation and osteogenic differentiation. 2009, Vol. 5, 8, pp. 3122-3131.

549. D Chen, L C Waite and W M Pierce. In vitro effects of zinc on markers of bone formation. 1999, Vol. 68, 3, pp. 225-234.

550. S L Hall, H P Dimai and J R Farley. Effects of zinc on human skeletal alkaline phosphatase activity in vitro. 1999, Vol. 64, 2, pp. 163-172.

551. H Kawamura, A Ito, T Muramatsu, S Miyakawa, N Ochiai and T Tateishi. Long-term implantation of zinc-releasing calcium phophate ceramics in rabbit femora. 2003, Vol. 65, 4, pp. 468-474.

552. M Yamaguchi, H Oishi and Y Suketa. Stimulatory effect of zinc on bone formation in tissue engineering. 1987, Vol. 36, 22, pp. 4007-4012.

553. I S Kwun, Y E Cho, R A Lomeda, H I Shin, J Y Choi, Y H Kang and J H Beattie. Zinc defficiency supresses matrix mineralization and retards osteogenesis with catch up possibly through RUNX 2 modulation. 2009, Vol. 46, 3, pp. 732-741.

554. C Lang, C Murgia, M Leong, L W tan, G Perozzi, D Knight, R Ruffin and P Zalewski. Anti- inflammatory effects of zinc and alterations in zinc transporter mRNA in mouse models of allergic inflammation. 2007, Vol. 292, 2, pp. L577-L584.

555. Y H Cho, S J Lee, J Y Lee, S W Kim, C B Lee, W Y lee and M S Yoon. Antibacterial effect of intraprostatic zinc injection in a rat model of chronic bacterial prostatitis. 2002, Vol. 19, 6, pp. 576-582.

556. A B Lansdown, U Mirastschijski, N Stubbs, E Scanlon and M S Agren. Zinc in wound healing: theorectical, experimental and clinical aspects. 2007, Vol. 15, 1, pp. 2-16.

557. S S Kim, M S Park, O Jeon, C Y Choi and B S Kim. Poly(lactide-co-glycolide)/hydroxyapatite composite scaffolds for bone tissue engineering. 2006, Vol. 27, 8, pp. 1399-1409.

558. P P Chen, S M Zhang, L Cheng, S L Huang, J Liu, W Zhou, H Gong and Q M Luo. Adhesion and proliferation of chondrocytes on nano- HA/PDLLA scaffolds for bone tissue engineering. 2006, Vols. 309-311, pp. 919-922.

559. J G C Gomez, V Fontolan, R C P Alli, M F A Rodrigues, C L B Netto, L F Silva and D A Simoes. Production of P3HB-co-3HV by soli ioslated bacteria able to use sucrose. Revista De Microbiologica. 1997, Vol. 28, 1, pp. 43-48.

560. L F Silva, M K Taciro, M E M Ramos, J M Carter and J G C Pradella and J G C Gomez. Poly-3- hydroxybutyrate (P3HB) production by bacteria from xylose, glucose and sugarcane bagasse hydrolsate. Journal of Industrial Microbiological and Biotechnology. 2004, Vol. 31, 6, pp. 245-254.

561. L F Silva, J G C Gomez, M S Oliveira and B B Torres. Propionic acid metabolism and poly-3- hydroxybutyrate-co-3-hydroyvalerate (P3HB-co-3HV) production by Burkholideria sp. Journal of Biotechnology. 2000, Vol. 76, 2-3, pp. 165-175.

562. X P Qi, F P He and J D Ye. Microstructure and Mechanical Properties of Calcium Phosphate Cement/Gelatine Composite Scaffold with Oriented Pore Structure for Bone Tissue Engineering. 2012, Vol. 27, 1, pp. 92-95.

563. N Nwe, T Furuike and H Tamura. The Mechanical and Biological Properties of Chitosan Scaffolds for Tissue Regeneration Templates Are Significantly Enhanced by Chitosan from Gongronella butleri. 2009, Vol. 2, 2, pp. 374-398.

315

564. A L Andrade, S P Andrade and R Z Domingues. In vivo performance of a sol-gel glass-coated collagen. Journal of Biomedical Materials Research Part B- Applied Biomaterials. 2006, Vol. 79B, 1, pp. 122-128.

565. K Y Lee and D J Mooney. Alginate: Properties and biomedical applications. 2012, Vol. 37, 1, pp. 106- 126.

566. A C Jayasuriya and A Bhat. Optimization of scaled-up chitosan microparticles for bone regeneration. Biomaterials Materials. 2009, Vol. 4, 5.

567. P B Malafaya and R L Reis. Optimization of chitosan-based composite and bi-layered scaffolds produced by particles aggregation for osteochondral tissue engineering; influence of hydroxylapatite. Tissue Engineering Part A. 2008, Vol. 14, 5.

568. N Yalamanchi, M B Klein, A S Dosanjh, H M Pham and J Chang. Optimization of an in vitro calcium- alginate chondrocyte construt by rabbit serum and ascorbic acid. Journal of Investigative Medicine. 2002, Vol. 50, 1.

569. H T Sasmazel. Novel hybrid scaffolds for the cultivation of osteoblast cells. 2011, Vol. 49, 4, pp. 838-846.

570. M Gumusderelioglu and S Aday. Heparin-functionalized chitosan scaffolds for bone tissue engineering. 2011, Vol. 346, 5, pp. 606-613.

571. W W Thein-Han and R D K Misra. Biomimetic chitosan-nanohydroxyapatite composite scaffolds for bone tissue engineering. 2009, Vol. 5, 4, pp. 1182-1197.

572. H R Lin and Y J Yeh. Porous alginate/hydroxyapatite composite scaffolds for bone tissue engineering: Preparation, characterization, and in vitro studies. 2004, Vol. 71B, 1, pp. 52-65.

573. G S Lee, J H Park, U S Shin and H W Kim. Direct deposited porous scaffolds of calcium phosphate cement with alginate for drug delivery and bone tissue engineering. 2011, Vol. 7, 8, pp. 3178-3186.

574. O Jeon, K H Bouhadir, J M Mansour and E Alsberg. Photocrosslinked alginate hydrogels with tunable biodegradation rates and mechanical properties. 2009, Vol. 30, 14, pp. 2724-2734.

575. M M Erol, V Mourino, P J Newby, X Chatzistavrou, J A Roether, L Hupa and A R Boccaccini. Copper-releasing, boron-containing bioactive glass-based scaffolds coated with alginate for bone tissue engineering. 2012, Vol. 8, 2, pp. 792-801.

576. Y Zhang and M Q Zhang. Calcium phosphate/chitosan composite scaffolds for controlled in vitro antibiotic drug release. 2002, Vol. 62, 3, pp. 378-386.

577. O Jeon, C Powell, L D Solorio, M D Krebs and E Alsberg. Affinity-based growth factor delivery using biodegradable, photocrosslinked heparin-alginate hydrogels. 2011, Vol. 154, 3, pp. 258-266.

578. S Y Im, S H Cho, J H Hwang and S J Lee. Growth factor releasing porous poly (epsilon-caprolactone)- chitosan matrices for enhanced bone regenerative therapy. 2003, Vol. 26, 1, pp. 76-82.

579. S E Kim, J H Park, Y W Cho, H Chung, S Y Jeong, E B Lee and I C Kwon. Porous chitosan scaffold containing microspheres loaded with transforming growth factor-beta 1: Implications for cartilage tissue engineering. 2003, Vol. 91, 3, pp. 365-374.

580. V Mourino, P J Newby, F Pishbin, J P Cattalini, S Lucangioli and A R Boccaccini. Physicochemical, biological and drug-release properties of gallium crosslinked alginate/nanoparticulate bioactive glass composite films. 2011, Vol. 7, 14, pp. 6705-6712.

316

581. V Mourino, P J Newby and A R Boccaccini. Preparation and Characterization of Gallium Releasing 3-D Alginate Coated 45S5 Bioglass (R) Based Scaffolds for Bone Tissue Engineering. 2010, Vol. 12, 7, pp. B282- B291.

582. A Narita, M Takahara, T Ogino, S Fukushima, Y Kimura and Y Tabata. Effect of gelatin hydrogel incorporating fibroblast growth factor 2 on human meniscal cells in an organ culture model. 2009, Vol. 16, 4, pp. 285-289.

583. D Tigani, C Zolezzi, F Trentani, A Ragaini, M Iafisco, S Manara, B Palazzo and N Roveri. Controlled release of vancomycin from cross-linked gelatine. 2008, Vol. 19, 3, pp. 1325-1334.

584. B Szomolay, I Klapper and M Dindos. Analysis of adaptive response to dosing protocols for biofilm control. Siam Journal on applied mathematics. 2010, Vol. 70, 8, pp. 3175-3202.

585. K K K Ho, N Cole, R X Chen, M D P Willcox, S A Rice and N Kumar. Immobilization of antibacterial dihydropyrrol-2-ones on functional polymer supports to prevent bacterial infections in vivo. Antimicrobial Agents and Chemotherapy. 2012, Vol. 56, 2, pp. 1138-1141.

586. P B Petratos, J Chen, D Felsen and D P Poppas. Local pharamaceutical release from a new hydrogel implant. Journal of Surgical research. 2002, Vol. 103, 1, pp. 55-60.

587. C Vilos, L Constandil, N Herrera, P Solar, J Escobar-Fica and L A Velasquez. Ceftiofur-loaded PHBV microparticles: A potential formulation for a long-acting antibiotic to treat animal infections. Electronic Journal of Biotechnology. 2012, Vol. 15, 4.

588. H W Kim, J C Knowles and H E Kim. Porous scaffolds of gelatin-hydroxyapatite nanocomposites obtained by biomimetic approach: Characterization and antibiotic drug release. 2005, Vol. 74B, 2, pp. 686-698.

589. O Goeau-Brissonniere, V Leflon, M Letort and M H Nicolas. Resistance of antibiotic-bonded gelatin- coated polymer meshes to Staphylococcus aureus in a rabbit subcutaneous pouch model. 1999, Vol. 20, 3, pp. 229-232.

590. B Gaihre, M S Khil and H Y Kim. In vitro anticancer activity of doxorubicin-loaded gelatin-coated magnetic iron oxide nanoparticles. 2011, Vol. 28, 4, pp. 286-293.

591. M Shi, J D Kretlow, P P Spicer, Y Tabata, N Demian, M E Wong, F K Kasper and A G Mikos. Antibiotic-releasing porous polymethylmethacrylate/gelatin/antibiotic constructs for craniofacial tissue engineering. 2011, Vol. 152, 1, pp. 196-205.

592. Z H Hao, B H Qu, Y L Wang, S S Tang, G Wang, M Qiu, R L Zhang, Y Y Liu and X L Xiao. Preparation and characterization of lung-targeting ceftiofur-loaded gelatin microspheres. 2011, Vol. 37, 12, pp. 1422-1428.

593. H C Yu, G R Liu, Q B Yie, J L Chen, J P Yuan and N Y Zhang. A novel method of immobilized enzyme-granular gelatin for absorbing and crosslinking penicillin acylase. 1988, Vol. 13, 3, pp. 177-180.

594. F Yang, J Wang, J Hou, H Guo and C Liu. Bone regeneration using cell-mediated responsive degradable PEG-based scaffolds incorporating with rhBMP-2. Biomaterials. 2013, Vol. 34, 5, pp. 1514-1528.

595. C Wang, K L Lin, J Chang and J Sun. Osteogenesis and angiogenesis induced by porous beta- CaSiO3/PDLGA composite scaffold via activation of AMPK/ERK1/2 and PI3K/Akt pathways. Biomaterials. 2013, Vol. 34, 1, pp. 64-77.

596. Z Zhou, J Lu, J Dou, Z Lv, X Qin and J Lin. FHL1 and Smad4 synergistically inhibit vascular endothelial growth factor expression. Molecular Medicine reports. 2013, Vol. 7, 2, pp. 649-653.

317

597. J S De Jong, P J Van Diest, P Van der Valk and J P A Baak. Expression of growth factors, growth- inhibiting factors, and their receptors in invasive breast cancer. II : Correlations with proliferation and angiogenesis. Journal of Pathology. 1998, Vol. 184, 1, pp. 53-57.

598. N Ferrara. The role of vascular endothelial growth factor in pathological angiogenesis. Breast cancer research and treatment. 1995, Vol. 36, 2, pp. 127-137.

599. T Gredes, A Spassov, R Mai, H Mack, B W Loster, M Mazurkiewicz-Janik, J Fanghanel and T Gedrange. Changes in insulin like growth factors, myostatin and vascular endothelial growth factor in rat musculus latissimus dorsi by poly 3-hydroxybutyrate implants. 2009, Vol. 60, pp. 77-81.

600. T Guo, J N Zhao, J B Chang, Z Ding, H Hong, J N Chen and J F Zhang. Porous chitosan-gelatin scaffold containing plasmid DNA encoding transforming growth factor-beta 1 for chondrocytes proliferation. 2006, Vol. 27, 7, pp. 1095-1103.

601. H L Jiang and K J Zhu. Preparation and release characteristics of protein-loaded polyanion/gelatin complex. 2001, Vol. 6, 2, pp. 231-240.

602. Y G Lv, Z S Deng and J Liu. 3-D numerical study on the induced heating effects of micro/nanoparticles on human body subject to external medical electromagnetic field. IEEE Transactions on Nanobioscience. 2005, Vol. 4, 4, pp. 284-294.

603. D Wesselinova. Current major cancer targets for nanoparticles systems. Current cancer drug targets. 2011, Vol. 11, 2, pp. 164-183.

604. N Bock, A Riminucci, C Dionigi, A Russo, A Tampieri, E Landi, V A Goranov, M Marcacci and V Dediu. A novel route in bone tissue engineering; Magnetic biomimetic scaffolds. Acta Biomaterialia. 2010, Vol. 6, 3, pp. 786-796.

605. S E Barry. Challenges in the development of magnetic particles for therapeutic applications. International Journal of Hyperthermia. 2008, Vol. 24, 6, pp. 451-466.

606. D D Guo, C H Wu, H Jiang, X M Wang and B A Chen. Synergistic cytotoxic effect of different sized ZnO nanoparticles and daunorubicin against leukemia cancer cells under UV irradiation. Journal of Photochemistry and Photobiology. 2008, Vol. 93, 3, pp. 119-126.

607. G F Schneider, V Subr, K Ulbrich and G Decher. Multifunctional Cytotoxic Stealth Nanoparticles. A Model Approach with Potential for Cancer Therapy. Nano Letters. 2009, Vol. 9, 2, pp. 636-642.

608. J J Corbalan, C Medina, A Jacoby, T Malinski and M W Radomski. Amorphous silica nanoparticles trigger nitric oxide/peroxynitrite imbalance in human endothelial cells; inflammatory and cytotoxic effects. International Journal of Nanomedicine. 2011, Vol. 6, pp. 2821-2835.

609. A N Generalova, V A Oleinikov, A Sukhanova, M V Artemyev, V P Zubov and I Nabiev. Quantum dot-containing polymer particles with thermosensitive fluorescence. Biosensors and Bioelectronics. 2013, Vol. 39, 1, pp. 187-193.

610. Y Chen, P A Wilbon, J H Zhou, M Nagarkatti, C P Wang, F X Chu and C B Tang. Multifunctional self-fluorescent polymer nanogels for label-free imaging and drug delivery. Chemical Communications. 2013, Vol. 49, 3, pp. 297-299.

611. P S Spuhler, L Sola, X Zhang, M R Margo, J T Greenspun, M Chiari and M S Unlu. Precisely controlled smart polymer scaffold for nanoscale manipulation of biomolecules. Analytical chemistry. 2012, pp. 10593-10599.

318

612. A Quaranta, A Rahman, G Mariotto, C Maurizio, E Trave, F Gonella, E Catteruzza, E Gibaudo and J E Broquin. Spectroscopic investigation of structural rearrangements in silver ion-exchanged silicate glasses. 2012, pp. 3757-3764.

613. A Pedone, T Charpentier, G Malavasi and M C Menziani. New insights into the atomic structure of 45S5 Bioglass by means of solid-state NMR spectroscopy and accurate first-principles simulations. Chemistry of Materials. 2010, Vol. 22, 19, pp. 5644-5652.

614. S Hossain and R Oki . NMR and FTIR investigation of sol-gel derived bioglass. Abstract of Papers American Chemical Society. 2003, Vol. 226, 1-2, p. INOR 471.

615. D A Keen, M G Tucker and M T Dove. Reverse Monte Carlo modelling of crystalline disorder. Vol. 17, pp. S15-S22.

616. P J Newby. The atomic-scale structure of a bioactive glass. Department of Physical Sciences, University of Kent, Canterbury. 2007. 3rd year Mphys project.

617. D Shouhed, H T Kha, J A Richardson, C M Amantea, T J Hahn and F Parhami. Osteogenic oxysterols inhibit the adverse effects of oxidative stress on osteogenic differentiation of marrow stronal cells. Journal of Cellular Biochemistry. 2005, Vol. 95, 6, pp. 1276-1283.

618. F G Quiroz, O M Posada, D Gallego-Perez, N Higuita-Castro, C Sarassa, D J Hansford, P Agudelo- Florez and L E Lopez. Housekeeping gene stability influences the quantification of osteogenic markers during stem cell differentiation to the osteogenic lineage. Cytotechnology. 2010, Vol. 62, 2, pp. 109-120.

619. X J Luo, J Chen, W X Song, N Tang, J Y Luo, Z L Deng, K A Sharff, G He, Y Bi, B C He, E Bennett, J Y Huang, Q Kang, W Jiang, X Y Su, G H Zhu, H Yin, Y Wang, J S Souris, L Chen, G W Zuo, A G Montag, R R Reid, R C Haydon, H H Luu and T C He. Osteogenic BMPs promote tumour growth of human osteosarcomas that habor differentiation defects. Laboratory Investigation. 2008, Vol. 88, 12, pp. 1264-1277.

620. T D Shanafelt, S Geyer, N Bone, S Schwager, Y K Lee, G Nowakowski, T G Call, C S Zent and N E Kay. Evaluation of the prognostic implications of pro- and anti-angiogenic cytokines in 311 patients with untreated chronic lymphocytic leukemia. Blood. 2005, Vol. 106, 11, pp. 211A-211A, Part 1, Meeting Abstract - 712.

621. R T P Poon, S T Fan and J Wong. Clinical implications of circulating angiogenic factors in cancer patients. Journal of Clinical Oncology. 2001, Vol. 19, 4, pp. 1207-1225.

622. S Silvan and S Manoharan. Apigenin prevents deregulation in the expression pattern of cell-proliferative, apoptotic, inflammatory and angiogenic markers during 7, 12-dimethylbenz[a]anthrecene-induced hamster buccal pouch carcinogenesis. Archives of oral biology. 2013, Vol. 58, 1, pp. 94-101.

623. R E Farrell Jr. RNA Methodologies: A laboratory guide for isolation and characterisation. [ed.] 3rd Edition. s.l. : Elsevier, 2005.

624. R R Rapley and D L Manning. RNA isolation and characterisation protocols. [ed.] R R Rapley and D L Manning. s.l. : Humana Press, 1998. Vol. 86.

625. C R Newton and A Graham. PCR. [ed.] 2nd edition. s.l. : Springer, 1997.

626. M McPherson and S Moller. PCR: The basics. [ed.] 2nd edition. s.l. : Taylor and Francis, 2006.

627. C Crowder and S D Stout. Bone histology: An anthropological perspective. s.l. : CRC Press - Taylor and Francis, 2012.

319

628. Y H An and K L Martin. Handbook of histology methods for bone and cartilage. s.l. : Humana Press, 2010.

629. C D Adler. Bone diseases: Macroscopic, Histological and radiological diagnosis of structural changes in the skeleton. [ed.] Forword written by F Feldman and D C Dahlin. s.l. : Springer, 2000.

630. P Chu and L Weiss. Modern Immunohistochemistry. s.l. : Cambridge Medicine, 2009.

631. R El-Gendy, X B Yang, P J Newby, A R Boccaccini and J Kirkham. Osteogenic differentiation of human dental pulp stromal cells on 45S5 Bioglass® based scaffolds in vitro and in vivo. Tissue Engineering Part A. 2013, Vol. 19, 5-6, pp. 707-715.

632. P J Sanderson. The Role of Methicillin-Resistant Staphylococcus aureus in Orthopaedic Implant Surgery. 2001, Vol. 13, 1, pp. 89-95.

633. P J Jenkins, K Teoh, P M Simpson, J Dave, A H W R Simpson and S Breusch. Clostridium difficile in patients undergoing primary hip and knee replacement. 2010, Vol. 92B, 7, pp. 994-998.

634. W Y Matar, S M Jafari, C Restrepo, M Austin, J J Purtill and J Parvizi. Preventing Infection in Total Joint Arthroplasty. 2010, Vol. 92A, 2, pp. 36-46.

635. T Livingston, P Ducheyne and J Garino. In vivo evaluation of a bioactive scaffold for bone tissue engineering. 2002, Vol. 62, 1, pp. 1-13.

636. M R Appleford, S Oh, N Oh and J L Ong. In vivo study on hydroxyapatite scaffolds with trabecular architecture for bone repair. 2009, Vol. 89A, 4, pp. 1019-1027.

637. Q A Fu, M N Rahaman, B S Bal, K Kuroki and R F Brown. In vivo evaluation of 13-93 bioactive glass scaffolds with trabecular and oriented microstructures in a subcutaneous rat implantation model. 2010, Vol. 95A, 1, pp. 235-244.

638. X Liu, M N Rahaman, Q Fu and A P Tomsia. Porous and strong bioactive glass (13-93) scaffolds prepared by unidirectional freezing of camphene-based suspensions. Acta Biomaterialia. 2012, Vol. 8, 1, pp. 415-423.

639. X Liu, M N Rahaman and Q Fu. Oriented bioactive glass (13-93) scaffold with controllable pore size by unidirectional freezing of camphene-based suspensions: Microstructure and mechanical response. Acta Biomaterialia. 2011, Vol. 7, pp. 406-416.

640. Q Fu, M N Rahaman, B S Bal and R F Brown. Preparation and in vitro evaluation of bioactive glass (13- 93) scaffolds with orientated microstructures for repair and regeneration of load bearing bones. Journal of Biomedical Research Part A. 2010, Vol. 93A, 4, pp. 1380-1390.

641. Q Fu, M N Rahaman, F Dogan and B S Bal. Freeze-cast hydroxyapatite scaffolds for bone tissue engineering. Biomedical Materials. 2008, Vol. 3, p. 025005.

642. —. Freeze casting of porous hydroxyapatite scaffolds II. Sintering, microstructure and mechanical behaviour. Journal of Biomedical Materials Research Part B: Applied Materials. 2008, Vol. 86B, 2, pp. 514- 522.

643. D Y Li and M S Li. Preparation of porous alumina ceramic with ultra-high porosity and long straight pores by freeze casting. Journal of Porous Materials. 2012, Vol. 19, 3, pp. 345-349.

644. S Deville, E Saiz and A P Tomsia. Freeze casting of hydroxyapatite scaffolds for bone tissue engineering. Biomaterials. 2006, Vol. 27, 32, pp. 5480-5489.

320

645. A Zamanian, S Farhangdoust, M Yasaei, Mkhorami and M Abbasabadi. The effect of sintering temperature on the microstructural and mechanical characteristics of hydroxyapatite macroporous scaffolds prepared via freeze-casting. Bioceramics. 2013, Vols. 529-530, 24, pp. 133-137.

646. B H Yoon, Y H Koh, C S Park and H E Kim. Generation of large pore channels for bone tissue engineering using camphene-based freeze casting. Journal of American Ceramic Society. 2007, Vol. 90, 6, pp. 1744-1752.

647. S Deville. Freeze-casting of porous biomaterials: Structure, properties and opportunities. Materials. 2010, Vol. 3, 3, pp. 1913-1927.

648. S W Yook, H E Kim, B H Yoon, Y M Soon and Y H Koh. Improvement of compressive strength of porous hydroxyapatite scaffolds by adding polystynene to camphene-based slurries. Materials Letters. 2009, Vol. 63, 11, pp. 955-958.

649. J H Kim, J H Lee, T Y Yan, S Y Yoon, B K Kim and H C Park. TBA-based freeze/gel casting of porous hydroxyapatite scaffolds. Ceramics International. 2011, Vol. 37, 7, pp. 2317-2322.

650. D Zhong, Y Zhang, R Xie and K Zhou. Freeze gel casting of aqueous alumina suspensions for porous ceramics. Ceramic International. 2012, Vol. 38, 7, pp. 6063-6066.

651. E J Lee, Y H Koh, B H Yoon, H E Kim and H W Kim. Highly porous hydroxyapatite bioceramics with interconnected pore channels using camphene-based freeze casting. Materials Letters. 2007, Vol. 61, 11-12, pp. 2270-2273.

652. A Macchetta, I G Turner and C R Bowen. Fabrication of HA/TCP scaffolds with a graded and porous structure using a camphene based freeze-casting method. Acta Biomaterialia. 2009, Vol. 5, 4, pp. 1319-1327.

653. Y Suetsugu, Y Hotta, M Iwasashi, M Sakane, M R Kuchic, T Ikoma, T Higaki, N Ochai and J Tanaka. Structural and tissue reaction properties of novel hydroxyapatite ceramics with unidirectional pores. Bioceramics. 2007, Vol. 19 Parts 1 and 2, 330-332, pp. 1003-1006.

654. Z Y Wu, R G Hill, S Yue, D Nightingale, P D Lee and J R Jones. Melt-derived bioactive glass scaffolds produced by a gel-cast foaming technique. Acta Biomaterialia. 2011, Vol. 7, 4, pp. 1807-1816.

655. F Baino, M Ferraris, O Bretcanu, E Verne and C Vitale-Brovarone. Optimisation of composition, structure and mechanical strength of bioactive 3-D glass-ceramic scaffolds for bone substitution. Journal of Biomaterials Applications. 2013, Vol. 27, 7, pp. 872-890.

656. E A Aguiler-Reyes, C A Leon-Patino, B Jacinto-Diaz and L P Lefebvre. Structural characterisation and mechanical evaluation of bioactive glass 45S5 foams obtained by a powder technology approach. Journal of the American Ceramic Society. 2012, Vol. 95, 2, pp. 3776-3780.

657. Q Fu, E Saiz and A P Tomsia. Direct ink writing of highly porous and strong glass scaffolds for load bearing bone defects repair and regeneration. Acta Biomaterialia. 2011, Vol. 7, pp. 3547-3554.

658. E Varndran, M Klarner, U Klammert, L M Grover, S Patel, J E Barralet and U Gbureck. 3D powder printing of beta-tricalcium phosphate ceramics using different strategies. Advanced Engineering Materials. 2008, Vol. 10, 12, pp. B67-B71.

659. A Butscher, M Bohner, S Hofmann, L Gauckler and R Muller. Structural and materials approaches to bone tissue engineering in powder based three dimensional printing. Acta Biomaterialia. 2011, Vol. 7, 3, pp. 907-920.

660. K Kato, S Ochiai, A Yamamoto, Y Daigo, K Honma, S Matano and K Omari. Novel multi layer Ti foam with cortical bone strength and cytocompatibility. Acta Biomaterialia. 2013, Vol. 9, 3, pp. 5802-5809.

321

661. X Liu, M N Rahaman, E Gregory and B S Bal. Mechanical properties of bioactive glass (13-93) scaffolds fabricated by robotic deposition for structural bone repair. Acta Biomaterialia. 2013, Vol. 9, 6, pp. 7025-7034.

662. M Genet, M Houmard, E Saiz and A P Tomsia. A two-scale Weibull approach to the failure or porous ceramic structure made by robocasting: Possibilities and limits. Journal of European Ceramic Society. 2013, Vol. 33, 4, pp. 679-688.

663. J R Jones, L M Ehrenfried and L L Hench. Optimising bioactive glass scaffolds for bone tissue engineering. Biomaterials. 2006, Vol. 27, 7, pp. 964-973.

664. J Zhao, K Duan, J W Zhang, L Y Guo and J Weng. Preparation of highly interconnected porous hydroxyapatite scaffold by chitin gel-casting. Materials Science and Engineering C. 2011, Vol. 31, 3, pp. 697- 701.

665. H Ghomi, M Fathi and H Edris. Preparation of nanostructure hydroxyapatite scaffold for tissue engineering applications. Journal of Sol-gel science and technology. 2011, Vol. 58, 3, p. 642.

666. H Ghomi, M H Faith and H Edris. Effect of the composition of hydroxyapatite/bioactive glass nanocomposite foams on their bioactivity and mechanical properties. Materials Rsearch Bulletin. 2012, Vol. 47, 11, pp. 3523-3532.

667. R Singh, P D Lee, J R Jones, G Podogasundarampaillai, T Post, T C Lindley and R J Dashwood. Hieraichically structured titanium foams for tissue scaffold application. Acta Biomaterialia. 2010, Vol. 6, 12, pp. 4596-4604.

668. J H Lee, H E Kim and Y H Koh. Highly porous titanium (Ti) scaffolds with bioactive microporous hydroxyapatite/TiO2 hybrid coating layer. Materials Letters. 2009, Vol. 63, 23, pp. 1995-1998.

669. J R Jones, S Lin, S Yue, P D Lee, J V Hanna, M E Smith and R J Newport. Bioactive glass scaffolds for bone regeneration and their hierarchical characterisation. Proceedings of the institution of mechanical engineers part H- Journal of Engineering in Medicine. 2010, Vol. 224, H12, pp. 1373-1387.

670. N Li and R Wang. Macroporous sol-gel bioglass scaffold with high compressive strength. Ceramics International. 2012, Vol. 38, 8.

671. G Turco, E Marsich, F Bellomo, S Semerano, I Donati, F Brun, M Grandolfo, A Accardo and S Paoletti. Alginate/Hydroxyapatite biocomposite for bone ingrowth: A trabecular structure with high and isotropic connectivity. Biomacromoles. 2009, Vol. 10, 6, pp. 1575-1583.

672. E Landi, F Valentini and A Tampieri. Porous hydroxyapatite/gelatine scaffolds with ice-designed channel-like porosity for biomedical applications. Acta Biomaterialia. 2008, Vol. 4, 6, pp. 1620-1626.

673. S M Kim, S A Yi, S H Choi, K M Kim and Y K Lee. Gelatin-layered and multi-sized porous beta- tricalcium phosphate for tissue engineering scaffold. Nanoscale research letters. 2012, Vol. 7, 78, pp. 1-5.

674. P M Hunger, A E Donius and U G K Wegst. Structure-properly processing correlations in freeze-cast composite scaffolds. Acta Biomaterialia. 2013, Vol. 9, 5, pp. 6338-6348.

675. M Erol, A Ozyuguran, O OZarpat and S Kucukbayrak. 3D composite scaffolds using strontium containing bioactive glasses. Journal of European Ceramic Society. 2012, Vol. 32, 11, pp. 2747-2755.

676. L Liverani, J A Roether, P Nooeaid, M Trombetta, D W Shubert and A R Boccaccini. Simple fabrication technique for multilayered stratified composite scaffolds suitable for interface tissue engineering. Materials Science and Engineering A. 2012, Vol. 557, pp. 54-58.

322

677. W G Bian, D C Li, Q Lian, X Li, W J Zhang, K Z Wang and Z M Jin. Fabrication of a bio-inspired beta-tricalcium phosphate/collagen scaffold based on ceramic stereolithography and gel casting for osteochondral tissue engineering. Rapid Prototyping Journal. 2012, Vol. 18, 1, pp. 68-80.

678. T Seria, J A Planell and M Navarro. High resolution PLA-based composite scaffolds via 3-D printing technology. Acta Biomaterialia. 2013, Vol. 9, 3, pp. 5521-5530.

679. F J Martinez-Vazquez, F H Perea, P Miranda, A Pajares and F Guiberteau. Improving the composite strength of bioceramic robocast scaffolds by polymer infiltration. Acta Biomaterialia. 2010, Vol. 6, 11, pp. 4361-4368.

680. M T Arafat, C X F Lam, A K Ekaputra, S Y Wong, X Li and I Gibson. Biomimetic composite coating on a rapid prototyped scaffold for bone tissue engineering. Acta Biomaterialia. 2011, Vol. 7, 2, pp. 809-820.

681. L Zhou, C J Hu, G P Xu and J L Yan. In vitro and in vivo study of calcium polyphosphate fibre/calcium phosphate cement/micromorselized bone composite for bone defect repair. Journal of Biomedicial Materials Research Part B - Applied Biomaterials. 2012, Vol. 100B, 5, pp. 1190-1197.

682. D Nadeem, M Kiamehr, X Yang and B Su. Fabrication and in vitro evaluation of a sponge-like bioactive-glass/gelatin composite scaffold for bone tissue engineering. Materials Science and Engineering C. 2013, Vol. 33, 1, pp. 2669-2678.

683. O Mahony, O Tsigkou, C Ionescu, C Minelli, L Ling, R Hanly, M E Smith, M M Stevens and J R Jones. Silca-gelatin hybrids with tailorable degradation and mechanical properties for tissue regeneration. Advanced Functional Materials. 2010, Vol. 20, 22, pp. 3835-3845.

684. C Vitale-Brovarone, M Miola, C Balagna and E Verne. 3D glass-ceramic scaffolds with antibacterial properties for bone grafting. Chemical Engineering Journal. 2008, Vol. 137, 1, pp. 129-136.

323

Conference papers and publications

Papers

Submitted/In progress

 45S5 Bioglass® scaffolds enhance angiogenic differentiation of human dental pulp stromal cells used for mineralised tissue engineering. R El-Gendy, J Kirkham, P J Newby, Y Mohanram, A R Boccaccini, and X B. Yang.  45S5 Bioglass®-derived glass-ceramic scaffolds modified using molten salt ion exchange to introduce antibacterial metal ions: processing and characterisation. P J Newby, E Saiz, I D Thompson and A R Boccaccini  Combining polymer coating and molten salt ion exchange to produce multifunctional 45S5 Bioglass®-based scaffolds for bone tissue engineering: processing and characterisation. P J Newby, E Saiz, I D Thompson and A R Boccaccini

Published

 Stiffness improvement of 45S5 Bioglass®– based scaffolds through PCL and collagen coatings : an ultrasonic study. J Hum, K W Luczynski, P Nooeaid, P J Newby, O Lahayne, C Hellmich and A R Boccaccini. Strain (2013) [Epub ahead of print]  Osteogenic differentiation of human dental pulp stromal cells on 45S5 Bioglass® based scaffolds in vitro and in vivo. R El-Gendy, X B Yang, P J Newby, A R Boccaccini and J Kirkham. Tissue Engineering Part A 19(5-6) (2013) 707-715  Gelatin coated 45S5 Bioglass®-derived scaffolds for bone tissue engineering. A Metze, A Grimm, P Nooeaid, J A Roether, J Hum, P J Newby, D W Schubert and A R Boccaccini. Key Engineering Materials 541(2013) 31-39  Evaluation of Angiogenesis of Bioactive Glass in the AV Loop Model. A Arkudas, A Balzer, G Buehrer, I Arnold, A Hoppe, R Detsch, P J Newby, T Fey, P Greil, R E Horch, A R Boccaccini and U. Kneser. Tissue Engineering Part C Methods. (2013) [Epub ahead of print]  Copper-releasing, boron-containing bioactive glass-based scaffolds coated with alginate for bone tissue engineering. M M Erol, V Mouriño, P J Newby, X. Chatzistavrou, J A Roether, L Hupa and A R Boccaccini. Acta Biomaterialia 8 (2012) 792–801  Ag-doped 45S5 Bioglass®-based bone scaffolds by molten salt ion exchange: processing and characterisation. P J Newby, R El-Gendy, J Kirkham, X B Yang, I D Thompson and A R Boccaccini. J Mater Sci: Mater Med (2011) 22:557–569

324

 Physicochemical, biological and drug-release properties of gallium cross linked alginate/nanoparticulate bioactive glass composite films. V Mouriño, P J Newby, F Pishbin, J P Cattalini, S Lucangioli and A R Boccaccini. Soft Matter, 2011, 7, 6705  Preparation and Characterization of Gallium Releasing 3-D Alginate Coated 45S5 Bioglass® Based Scaffolds for Bone Tissue Engineering. V Mouriño, P J Newby and A R Boccaccini. Advanced Engineering Materials, Special Issue: Advanced Materials in Transportation / Materials For Health Care. July 2010. Volume 12, Issue 7, pp B283 – B291

Conferences

 TERMIS AP 2011- Singapore. Therapeutic metal MSIE scaffolds for bone tissue engineering. P J Newby, A R Boccaccini. Presented by P J Newby  24th European Conference on Biomaterials 2011 – Dublin. Silver MSIE scaffolds for bone tissue engineering. P J Newby, A R Boccaccini. Presented by P J Newby  TERMIS EU 2010 – Galway, Ireland. A New Gallium Releasing Composite Scaffold for Bone Tissue Engineering. V Mouriño, P J Newby, A R. Boccaccini. Presented by V Mouriño  Euromat 2009 - Glasgow, UK. Preparation and Characterization of Gallium Releasing 3-D Alginate Coated 45S5 Bioglass® Based Scaffolds for Bone Tissue Engineering. V Mouriño, P J Newby, A R. Boccaccini. Presented by A R. Boccaccini

Book chapters

In press

Bioactive Glass Scaffolds for Bone Tissue Engineering (Chapter 5) Contribution in book "BIOACTIVE GLASSES: MATERIALS, PROPERTIES AND APPLICATIONS" edited by Professor Heimo Ylänen, Tampere University of Technology, Finland, Woodhead Publishing Limited (in press, 2010) P J Newby, X Chatzistavrou, A R. Boccaccini

325

Appendices

Appendix A – Background theory

A.1 Multifunctional scaffolds

Table 51:- Collated data for a variety of materials fabricated by one of the four fabrication methods discussed in chapter two, showing the change in porosity against compressional strength with FC meaning freeze casting, FR meaning foam replication, SFF meaning solid free-form replication and SG meaning sol-gel processing

Sample Fabrication Porosity Compressive Reference material method (%) Strength (MPa) 13-93 FC 50±4 35±11 (638) 13-93 FC 50±4 47±5 (638) 13-93 FC 20 180 (639) 13-93 FC 60 16 (639) 13-93 FC 57 25±3 (640) 13-93 FC 57 10±2 (640) HA FC 50 12±1 (641) (642) HA FC 50 5±1 (641) (642) HA FC 57 13 (642) HA FC 52 93 (642) Alumina ceramic FC 63 2.6 (643) Alumina ceramic FC 82 37 (643) 13-93 FC 55 25 (2) 13-93 FC 60 16 (2) 13-93 FC 50 47 (2) HA FC 68 20 (247) HA FC 48 60 (247) HA FC 76±2.75 5.26±0.52 (247) HA FC 47 145 (644) HA FC 56 65 (644) HA FC 75 8 (645) HA FC 76 16.7 (646) (647) HA FC 55 2.5 (646) (647) HA FC 71 1.1±0.2 (647) (648) HA FC 72 1.5±0.5 (647) (648) HA FC 73 2.3±0.5 (647) (648) HA FC 73 1.6±0.5 (647) (648) HA FC 79.3 2.7 (649) HA FC 41.9 35.1 (649) Alumina ceramic FC 75 11 (650) Alumina ceramic FC 65 12 (650) Alumina ceramic FC 60 13 (650)

326

Alumina ceramic FC 45 15 (650) CaP FC 62 93±1.6 (647) (260) CaP FC 65 6.2±1.3 (647) (260) HA FC 75 0.94 (647) (651) HA FC 56 17 (647) (651) TCP/HA FC 72.5 2.3 (647) (652) TCP/HA FC 31.4 36.4 (647) (652) HA FC 75 13.1 (647) (653) 13-93 FR 80±5 11±1 (241) 13-93 FR 78±2 7±0.5 (241) 13-93 FR 82±3 5±0.5 (241) boron modified 45S5 FR 72±3 6.4±1 (479) 13-93 FR 85±2 11±1 (239) HA FR 86 0.21 (642) HA FR 74 2.5 (642) 45S5 Bioglass® FR 90 0.35 (642) SCNPK FR 84 1.48 (654) 45S5 Bioglass® FR 89 0.41 (4) 45S5 Bioglass® FR 89 0.31 (4) 45S5 Bioglass® FR 90 0.25 (4) 45S5 Bioglass® FR 91 0.37 (4) 45S5 Bioglass® FR 91 0.35 (4) 45S5 Bioglass® FR 92 0.3 (4) 45S5 Bioglass® FR 92 0.25 (4) HA FR 86 0.21 (4) HA FR 88 0.1 (4) HA FR 90 0.12 (4) HA FR 93 0.05 (4) 45S5 Bioglass® FR 90 0.13 (224) 45S5 Bioglass® FR 90 0.53 (113) 45S5 Bioglass® FR 70 3 (655) 45S5 Bioglass® FR 90 0.4 (2) 13-93 FR 85 11 (2) 45S5 Bioglass® FR 90 0.4 (410) 45S5 Bioglass® FR 45 6 (410) BioK FR 75 0.16 (410) SCNM FR 40 6 (410) SCK FR 61 4 (410) FaGC FR 35 40 (410) FaGC FR 75 2 (410) CEL2 FR 48 7 (410) CEL2 FR 65 3.5 (410) 13-93 FR 85 11 (410) 13-93 FR 75 7 (410) ICEL2 FR 85 0.4 (410)

327

Beta-TCP FR 72 4 (410) 45S5 Bioglass® FR 65 1.7 (656) 6P53B SFF 60 136±22 (657) HA SFF 35 30 (642) HA SFF 41 34 (642) CaP SFF 73 6 (259) 6P53B SFF 60 150 (2) 13-93 SFF 50 140 (2) 13-93 SFF 50 20 (2) Bioactive glass SFF 60 16 (2) CaP SFF 54.45 5.8±0.8 (265) Beta-TCP SFF 56 1.33±0.03 (658) Beta-TCP SFF 60 1.20±0.02 (658) Beta-TCP SFF 61 1.19±0.04 (658) CaP SFF 35 4 (659) CaP SFF 40 3 (659) CaP SFF 77 0.4 (659) CaP SFF 79 0.2 (659) CaP SFF 30 31 (659) CaP SFF 3 45 (659) CaP SFF 7 25 (659) CaP SFF 27 4 (659) CaP SFF 39 0.8 (659) CaP SFF 36 4.3 (659) CaP SFF 35 34 (659) Titanium foam SFF 80 150 (660) Titanium foam SFF 17 240 (660) 13-93 SFF 47 250 (661) HA SFF 45 10 (662) HA SFF 45 15 (662) HA SFF 25 50 (662) HA SFF 35 35 (662) HA SFF 45 25 (662) HA SFF 55 20 (662) HA SFF 65 5 (662) HA SFF 35 45 (662) HA SFF 65 10 (662) TCP/HA SFF 55 17 (662) TCP/HA SFF 60 10 (662) TCP/HA SFF 65 8 (662) TCP/HA SFF 70 4 (662) TCP/HA SFF 80 2 (662) SCNPK SG 82 1.22 (654) SCNPK SG 81 1.79 (654) SCNPK SG 82 2 (654)

328

SCNPK SG 75 2.31 (654) SCNPK SG 74 2.82 (654) SCNPK SG 80 1.67 (654) SCNPK SG 74 1.75 (654) SCNPK SG 79 1.88 (654) SCNPK SG 75 2.07 (654) 70S30C SG 88 0.36 (663) 70S30C SG 86 0.51 (663) 70S30C SG 82 2.26 (663) 70S30C SG 82 2.25 (663) 45S5 Bioglass® SG 85 2.3 (244) 45S5 Bioglass® SG 87 1.6 (244) 45S5 Bioglass® SG 90 1.1 (244) 45S5 Bioglass® SG 92 1 (244) HA SG 87.2±2.1 0.75 (664) HA SG 86.8±2.8 0.9 (664) HA SG 86.4±2.6 1.1 (664) HA SG 87 1.5 (665) 13-93 SG 82 2.5 (2) 70S30C SG 85 1.8 (410) CaP SG 73 6 (259) Bioglass/HA SG 86 2.78 (666) Bioglass/HA SG 84±1 1.36±0.09 (666) Bioglass/HA SG 85±1 2.78±0.11 (666) Bioglass/HA SG 86±1.5 1.95±0.16 (666) Bioglass/HA SG 86±2 1.21±0.05 (666) Bioglass/HA SG 88±2 0.92±0.09 (666) Bioglass/HA SG 881 1.54±0.08 (666) Bioglass/HA SG 91±1.7 0.87±0.12 (666) Titanium foam SG 81±2 7 (667) Titanium foam SG 70 18±0.3 (668) 58S SG 70 2.4 (669) 70S30C SG 70 0.25 (669) 58S SG 55 30 (670) 58S SG 54 15 (670) 58S SG 59 14 (670)

329

Table 52:- Collated data for a variety of materials fabricated by one of the four fabrication methods discussed in chapter two, with the addition of a polymer coating, showing the change in porosity against compressional strength with FC meaning freeze casting, FR meaning foam replication, SFF meaning solid free-form replication and SG meaning sol-gel processing

Sample Fabrication Additives Porosity Compressive Reference material method (%) Strength (MPa) HA FC Coated with Alginate 88 6.3 (671) HA FC Coated with Alginate 88 8.3 (671) HA FC Coated with gelatin 56±2 13.8±1.3 (647) (672) HA FC Coated with gelatin 48±2 11.5±1.3 (647) (672) Beta-TCP FC Coated with gelatin 78.04±1.58 0.15±0.03 (673) Beta-TCP FC Coated with gelatin 82.65±4.17 0.11±0.01 (673) Beta-TCP FC Coated with gelatin 77.29±0.68 0.78±0.03 (673) Beta-TCP FC Coated with gelatin 85.83±1.02 0.53±0.05 (673) Alumina ceramic FC Coated with chitosan/gelatin 91.8±0.1 0.244±0.044 (674) Alumina ceramic FC Coated with chitosan/gelatin 90.2±0.1 0.346±0.023 (674) 45S5 Bioglass® FR Coated in P(3HB) microspheres 70 0.27 (224) Sr-doped Bioglass® FR Coated in gelatin 79±1.9 1.4±0.09 (675) 45S5 Bioglass® FR Coated in gelatin 87 1 (363) 45S5 Bioglass® FR Coated in PDLLA 90 0.3 (2) 45S5 Bioglass® FR Coated in PHB 90 0.3 (2) HA FR P(3HB) 90 1.46 (486) 45S5 Bioglass® FR coated with chitosan 75 4±2 (676) 45S5 Bioglass® FR coated with chitosan 75 15±5 (676) 45S5 Bioglass® FR Coated with Alginate 75 12±4 (676) 45S5 Bioglass® FR coated with chitosan 75 1.2±0.9 (676) Beta-TCP SFF coated in collagen 60 12 (677) CaP SFF composite with PLA/PEG 75±0.86 92.32±2.18 (678) CaP SFF composite with PLA/PEG 75±0.86 28.38±3.99 (678) CaP SFF composite with PLA/PEG/G5 70±1.2 99.81±3.55 (678) CaP SFF composite with PLA/PEG/G5 70±1.2 44.19±2.67 (678) Beta-TCP SFF coated with PLA 49 60±10 (679) Beta-TCP SFF coated with PCL 49 130±20 (679) Beta-TCP SFF PCL/TCP composite 65 14 (680) Beta-TCP SFF PCL/TCP composite with CHA 65 17 (680) PCL/TCP composite with Beta-TCP SFF CHA/gelatin 65 16 (680) CaP SG composite with CPP fibre 72.1 9.28 (681) CaP SG composite with CPP/bone 58.2 6.21 (681)

330

70S30C SG composite with gelatin 75 0.32±0.03 (682) 70S30C SG composite with gelatin 75 0.24±0.04 (682) Silica glass SG hybrid with gelatin 90 1 (683) Silica glass SG hybrid with gelatin 80 4 (683) Silica glass SG hybrid with gelatin 85 2 (683) Silica glass SG hybrid with gelatin 75 5 (683) Silica glass SG hybrid with gelatin 70 6 (683) Silica glass SG hybrid with gelatin 65 8 (683)

Table 53:- Collated data for a variety of materials fabricated by one of the four fabrication methods discussed in chapter two, subjected to MSIE coating, showing the change in porosity against compressional strength with FR meaning foam replication

Sample Fabrication Additives Porosity Compressive Reference material method (%) Strength (MPa) 45S5 Bioglass® FR Silver MSIE 75 0.63 (113) 45S5 Bioglass® FR Silver MSIE 75 0.67 (113) 45S5 Bioglass® FR Silver MSIE 75 0.7 (113) Fa-GC FR Silver MSIE 75 2 (684)

331

Appendix B – Experimental methods

B.1 Pellet production

This method will make pellets that have a thickness of approximately 1mm and they will have a diameter that depends on which pellet dye is used. Before starting the pellet making press needs to be thoroughly cleaned with acetone and dried. When this is carried out the following things are needed:-

 Tissue paper  2 weighing boats (one for the weighing and one to hold the pellets)  Gloves  Bioglass powder  Spatula

Method

1) Take the dye out of its box and clean everything with acetone and dry everything thoroughly. 2) Take the large cylinder and place the bottom cap in its correct position, as shown below. 3) Slot into the hole on of the small metal disc’s making sure that the shiny side is facing towards the top of the hole as this is the side that will be in contact with the Bioglass® powder. 4) Weigh out 0.03-0.05g of the Bioglass powder and pour it into the top of the large cylinder. Give the cylinder a sharp tap and make sure that the bottom disc is completely covered. 5) Slot the second metal disc into the tube making sure that the shiny side is facing down when you place it into the dye. 6) Place the whole ensemble into the press machine and tighten the top screw until it feels tight. 7) Close the side screw and gradually increase the pressure, by using the lever at the side of the machine, until it is very hard to move the lever. 8) Leave for five minutes or until the dial on the front of the machine stabilises and release the pressure by turn the screw at the side. 9) Take the whole ensemble out of the press machine and turn it upside down. 10) Take the bottom cap off and replace it with the large collar. 11) Place the whole ensemble back into the press machine making sure that the rod is pointing down. 12) Tighten the top screw and put a little pressure on the powder press. When a small hissing sound is heard, or you can feel the pellet has come loose, remove the whole ensemble from the press machine. 13) Remove the newly made pellet and place in the second weighing boat.

332

14) Clean the powder press with acetone and repeat until all the pellets that are required have been made. 15) Sinter in the same manner as a Bioglass scaffold, described in chapter four.

B.2 Preparation of samples for EDX depth studies

This method describes how the samples were prepared ready for depth EDX analysis. This technique requires prefabricated pellets that are to be characterised and these are prepared in accordance with the methods described in chapter four. When this is carried out the following things are needed:-

 Prefabricated sample pellets  Bakelite  Mounting press  Polishing machines with 500, 400, 300, 200 and 100 grit values  70% Ethanol

Method

1) The sample has to be embedded in Bakelite to make this characterisation possible. The prefabricated tablet is placed into the mounting press with its edge pointing up. 2) Add 20 to 30 ml of Bakelite powder into the press insuring that the pellet stays upright and is completely covered. 3) Place the upper ram back into the mounting press and seal it up. The press will then heat and press the Bakelite to mount the pellet. 4) Once cooled, remove the mounted pellet from the press. Wash the mounted pellet with 70% Ethanol and then water. Leave to dry thoroughly before the grinding and polishing process. 5) The mounted pellet now needs to be ground down to expose the pellet ready for EDX characterisation across the pellets cross-section. Gently apply pressure to the mounted pellet as you start polishing using the 500 grit values. Continue until a suitable pellet area has been exposed. 6) Now polish the mounted pellet using progressively finer grit values on the polishing machines. 7) Wash again in 70% Ethanol and then water and leave to dry. 8) Coat the whole top surface, the surface with the exposed pellet cross-section, with a layer of gold, as you would if you were going to carry out normal EDX or SEM, before characterising the mounted pellet.

333

B.3 Extraction of tooth pulp to obtain human periodontal ligamental stromal cells

This method describes the extraction of human periodontal ligamental stromal cells from teeth which were used to assess the suitability of the silver ion exchanged scaffolds for implantation. All the work is done in a tissue culture hood with all the utensils being sterilised using heat or watered down ethanol before being placed under the hood. The teeth used are donated adult wisdom teeth or premolars. This process removes the pulp from the centre of the tooth. When this process is carried out the following equipment is required:-

 Collagenese P  Lab standard gloves  Phosphate buffer solution  Samples tubes  Syringe and 0.45µm cellulose acetate filter  Petri dishes  70% Ethanol  Scalpel  Wire cutters or vice  Suction air pump  Incubator  α-MEM with 20% FBS, 2nm of L-glutamine and 100 units/ml of penicillin/streptomycin

Method

The first process is to prepare the enzyme which is used to digest the collagen matrix to delink the cells from the tooth pulp.

1) Collagenese P needs to be suspended in phosphate buffer solution (PBS). Suspend the enzyme in a ratio of 5mg of the powdered enzyme to 1ml of PBS. 40mg of enzyme to 8ml of PBS is enough enzyme to digest the pulp of two teeth 2) Place the enzyme and PBS in a 45ml sample tube and mix thoroughly. The suspension will change from a clear solution to pale yellow when the enzyme is fully suspended. 3) The suspension is now filtered to remove contamination and bacteria using a 0.45µm cellulose acetate filter and syringe. The next stage is to prepare the teeth by removing the pulp

334

1) Each tooth comes with a corresponding serial number which should be noted as these will correspond to the patient that they were removed from as this will have some bearing on the results. As younger teeth generally give better results. 2) Wash the tooth in ethanol to start the sterilisation process and removed any excess ethanol using an air suction pump. Wash the tooth in PBS and then removed any excess using an air suction pump. 3) Scrape any excess tissue from the outside of the tooth with a scalpel and wash again in PBS and remove any excess PBS using the air suction pump. The tooth is now ready to have the pulp removed. 4) Place the tooth into the thumb of a sterilised glove and is cracked open using a pair of wire cutters. If the wire cutters do not work, place the glove wrapped tooth in a vice until it cracks open. 5) Scrape out the pulp from the centre of the tooth and place into a petri dish with PBS to keep it moist. Discard the rest of the tooth and chop the pulp up into very small pieces to aid the enzyme digestion process 6) Transfer the tooth pulp into sample tubes that contain the enzyme only half closing the lids, to

o allow the cells to breathe, and place into an incubator at 37 C at 5%CO2 for a minimum of 10 minutes but no longer 30 minutes. This allows the enzyme to digest the collagen matrix which holds the cells together in the pulp. If the enzyme has not had enough time to work there will be wisps of collagen visible in the sample tube. If the enzyme is left too long the enzyme will start to digest the cell membranes which will kill the cells that are required in the work. 7) The cells can now be suspended in the CCM and are ready for use in the cell culture work described in chapter four

B.4 Extraction of rat bone-marrow stromal cells

This method describes the extraction of rat bone marrow cells from rat femurs used for this work. All the work is done in a tissue culture hood with all the utensils being sterilised using heat or watered down ethanol before being placed under the hood. In the cells are extracted from 12 week old rats and the extraction protocols have been approved by the CPCSEA. This process is carried out the following equipment is required:-

 Phosphate buffer solution  Samples tubes  1 ml Syringe and needle  0.45µm cellulose acetate filter

335

 70% Ethanol  DMEM/F-12 with 10% FBS, 1% L-glutamine and 1% penicillin/streptomycin

Method

1) The rats have to be humanely sacrificed and are sterilised by placing them in 70% ethanol for 10 minutes. 2) The femurs are removed and stripped of adherent muscle and other connective tissue. 3) Insert a needle into the bone and aspirate the cells out of the bone using a 1ml syringe containing the CCM until all the bone marrow has been removed. 4) The cell suspension is then passed through the 0.45µm filter to remove any contamination and bacteria. This will also form a single cell suspension which is then ready for use in the experiments described in chapter four.

B.5 Extraction of human bone-marrow stromal cells

This method describes the extraction of human bone marrow cells from femoral heads used for this work. All the work is done in a tissue culture hood with all the utensils being sterilised using heat or watered down ethanol before being placed under the hood. This process is carried out the following equipment is required:-

 Phosphate buffer solution  Samples tubes  0.45µm cellulose acetate filter  DMEM/F-12 with 10% FBS, 1% L-glutamine and 1% penicillin/streptomycin

Method

1) The human bone marrow stem cells where extracted from the femoral heads that were removed during total hip arthoplasty procedures. The femoral heads where segmented into two hemispheres to expose the trabecular bone whilst still in the sterile conditions of the operating theatre. 2) When the femoral heads are received in the laboratory they are placed into the tissue hood and the cells are they aspirated from the heads using PBS. 3) The cells suspension is then filtered using a 0.45µm filter to remove contamination and bacteria. 4) The cells are then suspended in the CCM and are then ready to be used once they are filtered to form a single cell suspension

336

B.6 Standard I Agar plate production

Standard I agar, also known as Standard methods agar, Plate count agar and Tryptone Glucose yeast agar is a growth media used for the enumeration of bacteria from a variety of sources. It is a translucent pale yellow colour when agar plates are formed using this media, this makes the bacteria growing on the surface easy to spot and count. The peptone, yeast extract, and glucose in the media provides the nutritional requirements needed for growth of the E.Coli bacteria This media can be brought pre-mixed although in this work won’t be using pre-mixed media. In order to make one litre of the media you need the following quantities:-

 3g of Yeast extract (CARL ROTH, Germany)  1.1g of α-D(+)-Glucose monohydrate (CARL ROTH, Germany)  15g of Peptone from casein (pancreatically digested)(Merck, Germany)  5.85g of Sodium chloride (CARL ROTH, Germany)  20g of Agar – Kobe I (CARL ROTH, Germany)  800ml of Deionised water

Using these quantities approximately 60 to 65 agar plates can be produced, assuming that 15ml of the solution is used per plate. These chemicals are used in the following method to produce the Standard I agar ready for use in agar plates:-

1) Weigh out the chemicals stated above, in that order, into a glass bottle with a plastic seal using a glass funnel. 2) When adding the first of the deionised water wash the funnel to ensure all of the weighed out chemicals make it into the solution, also make sure that none of the powder clings to the side of the bottle by swirling it gently whilst adding the deionised water. 3) Once all of the deionised water has been added place a magnetic stirrer bar into the bottle and place on a magnetic stirrer to dissolve the chemicals. 4) The pH of the solution needs to be approximately pH 7.4; the pH of the solution is adjusted by adding Sodium Hydroxide to the solution. 5) Once the pH has been adjusted, the bottle and its contents need to be autoclaved for sterilisation purposes. Indicator tape must be placed on the bottle to prove that the autoclaving process has worked. The bottle needs to be autoclaved for 15minutes at 121oC at a pressure of 2bar. 6) Remove the bottle once it is cool enough to handle and use as indicated in appendix B.11 to produce agar plates.

If too much of the Standard I agar solution has been produced it can be stored for up to 6 months at 4-8oC, but it is important to test the pH before its use as it might be better to make a fresh batch.

337

This protocol can be followed for most types of media but for this work it is for producing Endo Agar plates and Standard I Agar plates. All this work is to be carried out under a sterilised fume cabinet hood in order not to contaminate or spread the bacteria to other sources. At the bacteria lab in Erlangen, where this work was carried out, no gloves where to be worn whilst using the hood, instead from elbow to fingers were washed in a sticky disinfection solution to prevent the bacteria from being spread. Before starting the agar solutions need to be liquid, which isn’t an issue if they have been freshly made, but if they have been stored then they need to be heated up for a short period to return them to their liquid form. It should also be noted when handling anything or opening anything within the hood it should always been done coming from the side to prevent contamination.

To start the autoclaved solutions and sterile Petri dishes should be brought to the hood. Using a gas lamp run the mouth of the agar flasks in the flame to prevent any contamination getting into the flasks. Then it is just a case of pouring approximately 15ml of the agar solution into each Petri dish and then replacing the lid. These need to be left for about 10 minutes to start solidifying with their lids on and then a further 15 minutes without their lids to prevent condensation that might have built up from the slightly warm agar. Check if the plates have completely solidified, if not leave them a little longer, and then replace the lids. These are then ready to use straight away. These plates can also be stored at 4-8oC for up to six months but if they show signs of cracking then it is advisable to make some new plates.

B.7 Bacteria counting

During the course of the cell culture work and the bacteria studies the number of cells and bacteria need to be counted in order to progress with the experiment. In general the way cells and bacteria are counted is very similar with only a few minor differences. Each counting plate will have at least one counting chamber, as shown in Figure 145.

The counting chamber is divided into a number of counting squares or cells and it is in these that the bacteria and cells are counted underneath a light microscope. The bacteria don’t need a stain in order to be seen but the cells need to be stained with Trypin blue in order to be counted. A small amount of the bacteria or cell suspension is pipette into the counting chamber and a glass plate is placed on top of the counting chamber as shown Figure 146.

338

Figure 145:- Schematic of a counting plate showing the areas that cells and bacteria are counted in

Figure 146:- Counting chamber used in bacteria and cell counting

This ensemble is placed under the microscope and the bacteria or cells are then counted manually. The number of bacteria or cells is counted per little square, ideally at least four of the smallest counting chambers need to be counted in order for a more accurate count to be measured. It should be noted that the bacteria can swim in and out of focus of the light microscope and therefore a larger number of counting chambers should be looked at to take into account this movement. When counting pick two adjacent sides of the counting area and any bacteria or cells straddling them should be count and any that are straddling the other two sides should be disregarded for that particular counting area, as shown in Figure 147.

339

Figure 147:- Schematic of which cells/bacteria can be included in the counting process

Once an average number of bacteria or cells per counting area has been taken this can be scaled up in order to calculate the number of bacteria or cells per millilitre and then this number can be used in further calculates used in the experiment. To do this the volume of the counting chambers needs to be calculated using the values given on the plate, as shown in Figure 146, and then using this, the number of bacteria can be scaled up accordingly.

For example, a chamber has a height of 0.1mm and a small square area of 0.0025mm2, meaning each small square has a volume of 2.5x10-4mm3. If an average of 24 bacteria per small counting square were counted then it could be said that there are 24 bacteria per 2.5x10-4mm3 or per 2.5x10-4µl. This is scaled up to work out how many bacteria per µl and then per ml, eventually it is calculated that there are 9.6x107 bacteria per ml.

340

Appendix C – Results

C.1 Silver molten salt ion exchange scaffolds

Figure 148:- 45S5 Bioglass® –derived glass-ceramic scaffolds immersed in a medium concentration molten salt bath for A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes Images have a magnification of x150.

Figure 149:- 45S5 Bioglass® –derived glass-ceramic scaffolds immersed in a high concentration molten salt bath for A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes Images have a magnification of x150.

341

Figure 150:- Low concentration of silver MSIE scaffolds with immersion periods of 15, 30, 45 and 60 minutes

Figure 151:- High concentration of silver MSIE scaffolds with immersion periods of 15, 30, 45 and 60 minutes

342

Figure 152:- 45S5 Bioglass® –derived glass-ceramic scaffolds immersed in a low concentration molten salt bath for A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes and then immersed in SBF for 14 days. Images have a magnification of x1500

Figure 153:- 45S5 Bioglass® –derived glass-ceramic scaffolds immersed in a high concentration molten salt bath for A) 15 minutes, B) 30 minutes, C) 45 minutes and D) 60 minutes and then immersed in SBF for 14 days. Images have a magnification of x1500.

343

Figure 154:- High concentration of silver MSIE scaffolds with immersion periods of 15, 30, 45 and 60 minutes after immersion in SBF for 14 days

Figure 155:- Medium concentration of silver MSIE scaffolds with immersion periods of 15, 30, 45 and 60 minutes after immersion in SBF for 14 days

344

C.2 Bacterial assessment of molten salt ion exchange scaffolds

Figure 156:- Control samples taken after four and 24 hours of exposure E.coli and being plated onto standard I o agar plates and incubated for 24 hours at 37 c at 5% CO2: A) initial bacterial absorbance of 0.01 and 0.001 after four hours of exposure, B) initial bacterial absorbance of 0.001 after 24 hours of exposure and C) initial bacterial absorbance of 0.01 after 24 hours of exposure using DMEM/F-12 as the suspension media

Figure 157:- Control samples taken after four and 24 hours of exposure E.coli and being plated onto standard I o agar plates and incubated for 24 hours at 37 c at 5% CO2: A) initial bacterial absorbance of 0.01 and 0.001 after four hours of exposure, B) initial bacterial absorbance of 0.001 after 24 hours of exposure and C) initial bacterial absorbance of 0.01 after 24 hours of exposure using MHB as the suspension media

Figure 158:- Samples taken after four hours of exposure to varying concentrations of copper chloride salt and o being plated onto standard I agar plates and incubated for 24 hours at 37 c at 5% CO2: A) concentrations of copper chloride of 1mM and 0.5mM with an initial bacterial absorbance of 0.01, B) concentrations of copper chloride of 0.1mM and 0.05mM with an initial bacterial absorbance of 0.01, C) concentration of copper chloride of 1mM with an initial bacterial absorbance of 0.001, D) concentration of copper chloride of 0.5mM with an initial bacterial absorbance of 0.001, E) concentration of copper chloride of 0.1mM with an initial bacterial absorbance of 0.001 and F) concentration of copper chloride of 0.05mM with an initial bacterial absorbance of 0.001 using DMEM/F-12 as the suspension media

345

Figure 159:- Samples taken after four hours of exposure to varying concentrations of copper chloride salt and o being plated onto standard I agar plates and incubated for 24 hours at 37 c at 5% CO2: A) concentrations of copper chloride of 1mM with an initial bacterial absorbance of 0.01, B) concentrations of copper chloride of 0.5mM with an initial bacterial absorbance of 0.01,C) concentrations of copper chloride of 0.1mM with an initial bacterial absorbance of 0.01, D) concentrations of copper chloride of 0.05mM with an initial bacterial absorbance of 0.01E) concentration of copper chloride of 1mM with an initial bacterial absorbance of 0.001, F) concentration of copper chloride of 0.5mM with an initial bacterial absorbance of 0.001, G) concentration of copper chloride of 0.1mM with an initial bacterial absorbance of 0.001 and H) concentration of copper chloride of 0.05mM with an initial bacterial absorbance of 0.001 using MHB as the suspension media

Figure 160:- Samples taken after four hours of exposure to varying concentrations of copper nitrate salt, an E.coli suspension in DMEM/F-12 and being plated onto standard I agar plates and incubated for 24 hours at 37oc at 5% CO2: A) concentration of copper nitrate of 1mM with an initial bacterial absorbance of 0.001, B) concentration of copper nitrate of 0.5mM with an initial bacterial absorbance of 0.001, C) concentration of copper nitrate of 0.1mM with an initial bacterial absorbance of 0.001, D) concentration of copper nitrate of 0.05mM with an initial bacterial absorbance of 0.001, E) concentration of copper nitrate of 1mM with an initial bacterial absorbance of 0.01, F) concentration of copper nitrate of 0.5mM with an initial bacterial absorbance of 0.01, G) concentration of copper nitrate of 0.1mM with an initial bacterial absorbance of 0.01, and H) concentration of copper nitrate of 0.05mM with an initial bacterial absorbance of 0.01 using DMEM/F-12 as the suspension media

346

Figure 161:- Samples taken after four hours of exposure to varying concentrations of copper nitrate salt, an E.coli o suspension in MHB and being plated onto standard I agar plates and incubated for 24 hours at 37 c at 5% CO2: A) concentration of copper nitrate of 1mM with an initial bacterial absorbance of 0.001, B) concentration of copper nitrate of 0.5mM with an initial bacterial absorbance of 0.001, C) concentration of copper nitrate of 0.1mM with an initial bacterial absorbance of 0.001, D) concentration of copper nitrate of 0.05mM with an initial bacterial absorbance of 0.001, E) concentration of copper nitrate of 1mM with an initial bacterial absorbance of 0.01, F) concentration of copper nitrate of 0.5mM with an initial bacterial absorbance of 0.01, G) concentration of copper nitrate of 0.1mM with an initial bacterial absorbance of 0.01, and H) concentration of copper nitrate of 0.05mM with an initial bacterial absorbance of 0.01

Figure 162:- Samples taken after 24 hours of exposure to varying concentrations of copper chloride salt, an E.coli suspension in DMEM/F-12 and being plated onto standard I agar plates and incubated for 24 hours at 37oc at 5% CO2: A) concentration of copper chloride of 1mM with an initial bacterial absorbance of 0.01, B) concentration of copper chloride of 0.5mM with an initial bacterial absorbance of 0.01, C) concentration of copper chloride of 0.1mM with an initial bacterial absorbance of 0.01, D) concentration of copper chloride of 0.05mM with an initial bacterial absorbance of 0.01, E) concentration of copper chloride of 1mM with an initial bacterial absorbance of 0.001, F) concentration of copper chloride of 0.5mM with an initial bacterial absorbance of 0.001, G) concentration of copper chloride of 0.1mM with an initial bacterial absorbance of 0.001, and H) concentration of copper chloride of 0.05mM with an initial bacterial absorbance of 0.001

347

Figure 163:- Samples taken after 24 hours of exposure to varying concentrations of copper chloride salt, an E.coli o suspension in MBH and being plated onto standard I agar plates and incubated for 24 hours at 37 c at 5% CO2: A) concentration of copper chloride of 1mM with an initial bacterial absorbance of 0.01, B) concentration of copper chloride of 0.5mM with an initial bacterial absorbance of 0.01, C) concentration of copper chloride of 0.1mM with an initial bacterial absorbance of 0.01, D) concentration of copper chloride of 0.05mM with an initial bacterial absorbance of 0.01, E) concentration of copper chloride of 1mM with an initial bacterial absorbance of 0.001, F) concentration of copper chloride of 0.5mM with an initial bacterial absorbance of 0.001, G) concentration of copper chloride of 0.1mM with an initial bacterial absorbance of 0.001, and H) concentration of copper chloride of 0.05mM with an initial bacterial absorbance of 0.001

348

Figure 165:- Samples taken after 24 hours of exposure to varying concentrations of copper nitrate salt, an E.coli o suspension in MHB and being plated onto standard I agar plates and incubated for 24 hours at 37 c at 5% CO2: A) concentration of copper nitrate of 1mM with an initial bacterial absorbance of 0.01, B) concentration of copper nitrate of 0.5mM with an initial bacterial absorbance of 0.01, C) concentration of copper nitrate of 0.1mM with an initial bacterial absorbance of 0.01, D) concentration of copper nitrate of 0.05mM with an initial bacterial absorbance of 0.01, E) concentration of copper nitrate of 1mM with an initial bacterial absorbance of 0.001, F) concentration of copper nitrate of 0.5mM with an initial bacterial absorbance of 0.001, G) concentration of copper nitrate of 0.1mM with an initial bacterial absorbance of 0.001, and H) concentration of copper nitrate of 0.05mM with an initial bacterial absorbance of 0.001

Figure 166:- Supernatents plated after an immersion period of 28 days in DMEM/F-12, added to an E.coli suspension for 24 hours and then plated onto standard I agar plates with sodium thiosulfate; A) plates showing the supernatents taken from sample 1 with samples including a plain 45S5 Bioglass® -glass-ceramic scaffold, the lowest concentration MSIE samples of each type and B) plates showing the supernatents taken from sample 2 with samples including a plain 45S5 Bioglass® -glass-ceramic scaffold, the lowest concentration MSIE samples of each type

349

Figure 167:- Supernatents plated after an immersion period of 7 days in DMEM/F-12, added to an E.coli suspension for 24 hours and then plated onto standard I agar plates with sodium thiosulfate; A) plates showing the supernatents taken from sample 1 with samples including a plain 45S5 Bioglass® -glass-ceramic scaffold, the lowest concentration MSIE samples of each type and B) plates showing the supernatents taken from sample 2 with samples including a plain 45S5 Bioglass® -glass-ceramic scaffold, the lowest concentration MSIE samples of each type

Figure 168:- Supernatents plated after an immersion period of 3 days in DMEM/F-12, added to an E.coli suspension for 24 hours and then plated onto standard I agar plates with sodium thiosulfate; A) plates showing the supernatents taken from sample 1 with samples including a plain 45S5 Bioglass® -glass-ceramic scaffold, the lowest concentration MSIE samples of each type and B) plates showing the supernatents taken from sample 2 with samples including a plain 45S5 Bioglass® -glass-ceramic scaffold, the lowest concentration MSIE samples of each type

350

Figure 169:- Supernatents plated after an immersion period of 1 day in DMEM/F-12, added to an E.coli suspension for 24 hours and then plated onto standard I agar plates with sodium thiosulfate; A) plates showing the supernatents taken from sample 1 with samples including a plain 45S5 Bioglass® -glass-ceramic scaffold, the lowest concentration MSIE samples of each type and B) plates showing the supernatents taken from sample 2 with samples including a plain 45S5 Bioglass® -glass-ceramic scaffold, the lowest concentration MSIE samples of each type

C.3 Polymer coated 45S5 Bioglass® -derived glass-ceramic scaffolds

Figure 170:- FTIR spectra for PDLLA coated BBG scaffolds before and after immersion in SBF for 14 days from Figure 84 showing just the section of the spectra relating to the PDLLA peaks

351

Figure 171:- FTIR spectra for gelatin coated BBG scaffolds from Figure 85 just showing the peaks relating to the gelatin coating

Figure 172:- FTIR spectra for gelatin coated BBG scaffolds after immersion in SBF for 14 days from Figure 91 just showing the peaks relating to the gelatin coating

352

C.4 Polymer coated molten salt ion exchanged 45S5 Bioglass® -derived glass-ceramic scaffolds

Figure 173:- FTIR spectra for gelatin coated copper-doped BBG scaffolds from Figure 96 just showing the peaks relating to the gelatin coating

Figure 174:- FTIR spectra for gelatin coated copper-doped BBG scaffolds after immersion in SBF for 14 days from Figure 100 just showing the peaks relating to the gelatin coating

353

Figure 175:- FTIR spectra for gelatin coated silver/copper-doped BBG scaffolds from Figure 105 just showing the peaks relating to the gelatin coating

Figure 176:- FTIR spectra for gelatin coated silver/copper-doped BBG scaffolds after immersion in SBF for 14 days from Figure 109 just showing the peaks relating to the gelatin coating

354

Appendix D – Future Work

D.1 Reverse Monte Carlo modelling

Reverse Monte Carlo (RMC) modelling is a technique for generating a 3D computer model based on experimental data for amorphous and disordered crystalline structures. (616) The experimental data is now not considered in isolation, but can be treated as complementary to the process of producing a 3D model of the materials’ structure. (616)

A “box” is set up and a set of atoms are placed in the box where a calculation takes place based on the current atom configuration and chemical formula. Then one atom is randomly selected and is moved a random amount and the calculation is re-done based on the new configuration. (616) This movement will either be accepted or rejected based on whether the atom overlaps another atom or whether the value is similar to the starting experimental value. (616) If the move is rejected the atom is moved back to its original starting position and the process is repeated. If the move is accepted then another atom is selected at random and the process of moving the atoms can start again. (616) As the number of successfully moved atoms increases the calculated values will start to get closer towards the given experimental value until they reach a certain point (the point where the calculated value equals the experimental value) and at this point the calculated values will oscillate slightly. (616) This will then lead to an atomic configuration that will give rise to a model that is plausible and consistent with the experimental data. (616)

The atomic structure produced by RMC modelling isn’t unique; it’s just the model that is consistent with the given experimental data and any other given constraints. (616) Other modelling techniques produce 3D models which are valid with the given data but there is no way, at present, of determining which the correct model is without referring to additional information. (616)

RMC is independent of potential, so it is not affected by potential parameters. Most models are compared to the experimental data at the radial distribution function stage (G(r)), which describes how on average the atoms in a system are packed around each other, because the configurations are just too small for the models to be compared to experimental data at the structure factor stage (F(Q)). (616) For a model that fits to the experimental data the comparison needs to be made whilst the data is at the F(Q) stage. (616)

As RMC models a 3D atomic structure, the real radial distribution and the real structure factor (Gc(r) and Fc(Q) respectively) must correspond to a possible valid physical structure. However the experimental radial distribution function (Ge(r)), may contain errors which could potentially mean that the atomic model would not correspond to a possible valid physical structure, meaning the model is

355 inconsistent with the given experimental data. (616) In RMC, the partial radial distribution functions (PRDF’s) have to be consistent with each other and they have to be valid in order to produce an atomic model that is consistent with the given experimental data. This constraint in RMC helps to improve the partials (partials meaning only looking at one particular pair of atoms) separation and means that information can be obtained from undetermined cases, where no PRDF’s have been determined by conventional methods. (616)

In principle a configuration of atoms with a Boltzmann distribution of energies needs to be produced. To produce configurations like this the Markov chain (weighted sampling procedure) is used to generate and samples configurations. (616) The Markov chain satisfies a number of requirements that are needed for producing a model: -

 The first being that variables x(n) are found by using a rule that requires the (n + 1)nt to have a probability distribution of x(n+1) that is dependent on the distribution, of the nth element, of x(n).  The second requirement is that if P(x => y) is the probability of getting to the y state from the x state, then P(x =>y) must allow the movement to every state within the configuration.  The last requirement states that microreversibilty must be satisfied, such that xP(x => y) = yP(y => x) when the system of atoms is in equilibrium.

This can be achieved by the following process, but can only be done for a system of atoms where the number of atoms, volume and temperature are kept constant. Firstly the a number of atoms (let’s call this number N) are placed inside a “box” which is surrounded by other identical boxes, in most cases cubes are used as the box shape, these cubes have a side length of L. The atom number density can be described as ρ = N/L3 and this atom number density must be equal to the required density of the system. The probability of this configuration is described in Equation 16. This shows the original probability of the configuration before anything was done to the system. (616)

Equation 16 :- MC modelling – the probability of the original configuration

Po  expU o kT

The total potential energy (Uo) can be calculated on the basis of the specified form of inter- atomic potential. (616)

Next one atom is moved at random and meaning the probability changes from that calculated in

Equation 16. Equation 17 shows the new probability after one atom is moved at random. Where Un is the new total potential energy at that point. (616)

356

Equation 17 :- MC modelling – the probability of the new configuration after one atom move

Pn  expU n kT

If Equation 16 and Equation 17 are combined, the constraint for whether the atom stays in that new position or whether it is moved back to its original position can be derived. Equation 18 shows the probability of the new configuration divided by the probability of the old configuration. (616)

Equation 18 :- MC modelling – Constraint of whether an atom stays in the new position or moved back to its original position

Pn Po  exp U n U o  kT  expU kT

If ΔU is less than zero the new configuration of atoms is accepted and the atom that was moved stays put in that new position and this becomes the new starting point. (616) If ΔU is greater than zero then the move is rejected and the atom is moved back to its original position. This process of moving atoms and checking there probabilities is repeated until ΔU decreases such that it reaches an equilibrium value and then it oscillates about this value. That was a brief description of the theory behind the Monte Carlo modelling (MC) which is needed as a starting point for RMC modelling. (616)

In experiments it is expected for the data collected to have some errors within it, we can assume

E that the experimental structure factor (F (Qi)) only has statistical errors that have normal distributions

C on them. The difference between the real structure factor (F (Qi)) and the structure factor that is found experimentally (Ei) is expressed by Equation 19. (616)

Equation 19 :- RMC modelling –Experimental Structure factor relationship

c e Ei  F Qi  F Qi 

The probability of this difference can be expressed in terms of the standard deviation of the normal distribution (σ(Qi)), shown in Equation 20. Where σ is the error and Qi is the scattering vector. (616)

Equation 20 :- RMC modelling – Probability of the difference in experimental structure factors in terms of the standard deviation of the normal distribution

1  E 2  pE   exp i  i  2  2 Qi   2 Qi  

Knowing this probability (the probability of the difference between the real structure factor and the experimental structure factor) the probability for the real structure factor can be calculated, shown in Equation 21. (616)

357

Equation 21 :- RMC modelling – Probability of the real structure factor

m m  1   m E 2  P  pE     exp i   i     2  i1  2    i1 2 Qi  

E Where m is the number of Qj points in F and is shown in Equation 22. (616)

Equation 22 :- RMC modelling –Standard deviation in terms of the scattering vector

1  m  m

   Qi   i1 

In order to model a system using the real structure factor (FE(Q)) a statistical configuration of atoms whose F(Q)’s need to satisfy the probability distribution shown in Equation 19. Then the exponent (and this is shown in Equation 23. (616 2א can be written in terms of

Equation 23:- RMC modelling - Constraint of whether an atom stays in the new position or moved back to its original position

m 2 2 C E 2   F Qi  F Qi   Qi  i1

and from this it can be seen (2/2א-)Then it can be seen that the probability is proportional to exp (in RMC is the equivalent of U/kT in the MC method, shown in Equation 17. (616 2/2א that

Like the MC method, a starting configuration is needed, the positions of the N atoms may be chosen at random, and they may be obtained from either another model or from a known crystal structure. The next step is to calculate the radial distribution function for the initial configuration of atoms, shown in Equation 24. (616)

Equation 24 :- RMC modelling – Radial distribution function for the initial configuration of atoms

nC r g C r  o o 4r 2r

C Where no (r) is the number of atoms at a distance r and r+Δr from a random central atom, this is averaged over when all the atoms that were used as the central point and ρ is the number density. (616) The length of the box is once again L and this should be large enough so that g(r > L2) = 1. The radial distribution function, g(r) is only ever calculated for r < L/2 and the nearest image convention is used to determine the atomic separations. The next thing that should be calculated is the transformation of the radial distribution into the total structure factor; this is shown in Equation 25, where Q is the momentum transfer in this case. (616)

358

Equation 25 :- RMC modelling – Transformation of the radial distribution into the total structure factor

4  F 2 Q 1  r g C r 1 sin Qrdr o     o    Q 0

Now the difference between the experimental measured structure factor [FE(Q)] and the

C 2 o andא structure factor that is obtained from the starting configuration can be found [F (Q)]. This gives this is shown in Equation 26. (616)

Equation 26 :- RMC modelling – Difference between the experimental structure factor and the starting structure factor

2 2 C E 2 o  Fo Qi  F Qi   Qi  i

Where the sum is over the m experimental points and σ(Qi) is the experimental error on those points. In practical situation a uniform σ is normally used because the distribution of the systematic errors is unknown. (616) Next is the part were one atom is moved and a new radial distribution needs to be calculated as well as a new total structure factor and this is done by combining Equation 21and (is given in Equation 27. (616 2א Equation 22. The new

Equation 27 :- RMC modelling - Constraint of whether an atom stays in the new position or moved back to its original position

2 C E 2 2  n  Fn Qi  F Qi   Qi 

2 2 o then the atom’s move is accepted and the new configuration becomes the newא n less thanא If

2 2 o , then theא n is great thanא starting configuration and the atom is left in its new position. If however

2 2 (o )/2), otherwise the move is rejected. (616א- nא)-)move can be accepted as long as the probability is exp

will decrease until it reaches an equilibrium point and 2א As with the MC method, the value of the function will oscillate around that point. The resulting configuration file that is produced will then (once run through the proper plotting program) produced a 3D structure model that is consistent with the given experimental structure factor within an experimental error. (616)

There is one thing that should be considered and this is the idea of the constraints. There are two very important constraints within the RMC method:-

 The constraint on the closest distance of approach of two atoms  The constraint on the coordination of the atoms.

The first constraint that shall be discussed is the constraint on the distance of closest approach of two atoms. Due to errors on the experimental data and the limited range of the data, the data would

359 not forbid some atoms getting close to each other. (616) Realistic values for closest approach distances can be obtained by the use of the direct Fourier transformation of the experimental total structure factor. (616) One dominant effect of determining the structure of a material is the packing of the atoms. To include the information of the different atom sizes in the model would limit the number of structures that would be consistent with the given data. (616)

The second constraint is the one that is concerned with the coordination of the atoms. A coordination number nαβ, this is defined as the number of atoms of type β between two fixed distances of one atom of type α. (616) The lower fixed distance is the same as the closest distance of approach of the two atom types (or zero). (616) If the proportion of α type atom’s in the configuration with coordination of fRMC and the required proportion of α atoms as freq then an additional term can be added to Equation 23, shown in Equation 28. (616)

Equation 28 :- RMC modelling - Constraint of whether an atom stays in the new position or moved back to its original position plus the new coordination factor

m 2 C E 2 2 2   F Qi  F Qi   Q  f req  f RMC   C  i1

Multiple coordination constraints can be added to the model by adding additional term to

Equation 26. (616) σc acts as a weighting function of the coordination constant relative to the given data.

If σc is approximately equal to zero then it is impossible for atoms with the constrained coordination to change it, this is used to mimic covalent bonding. (616) In a lot of cases hard sphere Monte Carlo simulation with such constraints, this is RMC with no data and is used to produce a structure with suitable topology (it’s an extension of geometry) before experimental data is fitted to it. (616)

D.2 The Arteriovenous loop

The arteriovenous loop model (AVL) is based on a loop created in vivo in the medial thigh of a rat with a vein graft from the contralateral leg. This model has shown to enhance neovascularisation. The cage prevents vascularisation occurring from surrounding tissue and therefore the observed vascularisation is exclusively from the construct matrix.

The in vivo model consists of a heat resistant medical grade Teflon® cage with an inner diameter of 10mm, height of 10mm and a lid. Four green plastic tubes where evenly spaced out in the cage to keep the loop in shape and from touching.

360

The construct matrix consisted of pieces of BBG ranging in size between 1mm to 2mm and a fibrin gel. The fibrin gel had a thrombin concentration of 2 I.U./ml, a fibrinogen concentration of 10mg/ml and an aprotinin concentration of 1500 K.I.E/ml.

The model is then surgically implanted into the rat and left for three weeks before being removed ready for characterisation to determine the vascular network. Using histology and micro computed topography the vascular network established in the cage can be established.

D.3 Therapeutic metal nano-particles – Copper oxide and Iron oxide

Figure 177:- FTIR analysis of BBG scaffolds coated with a) 6wt% type A gelatin with copper nano-particles, b) 6wt% type A gelatin doped with iron oxide nano-particles, c) 6wt% type B gelatin doped with copper nano- particles and d) 6wt% type B gelatin doped with iron oxide nano-particles

361

Figure 178:- BBG scaffolds coated with gelatin with A) a type A 5wt% coating with iron oxide nano-particles (magnification of x500), B) a type B 5wt% coating with iron oxide nano-particles (magnification of x250), C) a type A 5wt% coating with copper oxide nanoparticles (magnification of x 300) and D) a type B 5wt% coating with copper oxide nanoparticles (magnification of x160)

362

Appendix E – Statistical Analysis

The mean of a data set is the sum of the data (xn) devided by the number of samples in the data set (n), as shown by Equation 29.

Equation 29:- Mean of a data set

Data values are described as the mean ± standard error (SE) where the SE is shown in Equation 30.

Equation 30:- Standard error on the mean

To ascertain the ANOVA value (F) a number of steps must be taken, first being working out the total mean of all of the data collect, as shown in Equation 31. With n being the number of samples in total and a being the number of data sets.

Equation 31:- Group mean of multiple data sets

In its simplest form, ANOVA analysis is the variance (the standard deviation squared) between the data sets divided by the variance within a data set, as shown by Equation 32.

Equation 32:- ANOVA analysis of multiple data sets

This is calculated by working out the mean square value between the data sets divided by the mean square value within a data set, as shown in Equation 33.

363

Equation 33:- ANOVA analysis of multiple data sets

The mean square between the data sets is calculated by dividing the sum of the squares between the data sets by the number of data sets minus one and the mean square within a group is calculated by dividing the sum of the squares within a data set by the number of samples from that data set minus 1. This is represented in Equation 34 and Equation 35.

Equation 34:- ANOVA analysis of multiple data sets

Equation 35:- ANOVA analysis of multiple data sets

The value of observed F is then compared to a critical value of F from the literature derived according to the significance wanted, which in this case is p ≤0.0. If the observed value for F is greater than or equal to the critical value of F, then the difference between the data sets is due to the change in parameters and not chance or a random error.

This can be compared to the student T Test looking at the difference between two data sets to assure that the ANOVA analysis is correct as shown in Equation 36.

Equation 36:- Student T Test of two data sets

364

Appendix F – Copyright permissions

Table 54 showing the status of copyright permissions obtained for this work.

Table 54:- Permission status of copyright of figures, tables and data used in this work

Page Figure or Original Copyright owner Year Status number table number source 239 Figure 137 Dr T Hammer Dr T Hammer 2011 Permission granted 240 Table 49 Dr T Hammer Dr T Hammer 2011 Permission granted 208 Figure 116 Dr L Strobel Dr L Strobel 2011 Permission granted 208 Figure 117 Dr L Strobel Dr L Strobel 2011 Permission granted 212 Figure 119 Dr L Strobel Dr L Strobel 2011 Permission granted 213 Figure 120 Dr L Strobel Dr L Strobel 2011 Permission granted 214 Figure 121 Dr L Strobel Dr L Strobel 2011 Permission granted 216 Figure 122 Dr L Strobel Dr L Strobel 2011 Permission granted 217 Figure 123 Dr L Strobel Dr L Strobel 2011 Permission granted 204 Figure 112 Dr R El-Gendy Dr R El-Gendy 2010 Permission granted 205 Figure 113 Dr R El-Gendy Dr R El-Gendy 2010 Permission granted 206 Figure 114 Dr R El-Gendy Dr R El-Gendy 2010 Permission granted 206 Figure 115 Dr R El-Gendy Dr R El-Gendy 2010 Permission granted 265 Figure 141 Dr A Arkudas Dr A Arkudas 2013 Permission granted 266 Figure 142 Dr A Arkudas Dr A Arkudas 2013 Permission granted 266 Figure 143 Dr A Arkudas Dr A Arkudas 2013 Permission granted 267 Figure 144 Dr A Arkudas Dr A Arkudas 2013 Permission granted 62 Figure 12 John Wiley and Sons John Wiley and Sons 2010 Permission granted 56 Figure 11 Elsevier Elsevier 2006 Permission granted 44 Figure 4 Springer Springer 2001 Permission granted 35 Figure 7 Springer Springer 2006 Permission granted

365

Permission granted for use of the images and data shown in Figure 137 and Table 49 from Dr Timo Hammer is shown in Figure 179.

Figure 179:- Permission granted for the use of the images and data shown in Figure 137 and Table 49 from Dr Timo Hammer

Permission granted for use of the images and data shown in Figure 116, Figure 117, Figure 119, Figure 120, Figure 121, Figure 122 and Figure 123 from Dr Leonie Strobel is shown in Figure 180.

Figure 180:- Permission granted for the use of the images shown in Figure 116, Figure 117, Figure 119, Figure 120, Figure 121, Figure 122 and Figure 123 from Dr Leonie Strobel

366

Permission granted for use of the images and data shown in Figure 112, Figure 113, Figure 114 and Figure 115 from Dr Reem El-Gendy is shown in Figure 181.

Figure 181:- Permission granted for the use of the images shown in Figure 112, Figure 113, Figure 114 and Figure 115 from Dr Reem El-Gendy

Permission granted for use of the images and data shown in Figure 141, Figure 142, Figure 143 and Figure 144 from Dr Andreas Arkudas is shown in Figure 182.

Figure 182:- Permission granted for the use of the images shown in Figure 141, Figure 142, Figure 143 and Figure 144 from Dr Andreas Arkudas

367

Permission granted for the use of the image shown in Figure 12 from the publisher John Wiley and Sons is shown in Figure 183.

Figure 183:- Permission granted for the use of the image shown in Figure 12 from the publisher John Wiley and Sons

Permission granted for the use of the image shown in Figure 11 from the publisher Elsevier is shown in Figure 184.

Figure 184:- Permission granted for the use of the image shown in Figure 11 from the publisher Elsevier

368

Permission granted for the use of the image shown in Figure 4 from the publisher Springer is shown in Figure 185.

Figure 185:- Permission granted for the use of the image shown in Figure 4 from the publisher Springer

Permission granted for the use of the image shown in Figure 7 from the publisher Springer is shown in Figure 186.

Figure 186:- Permission granted for the use of the image shown in Figure 7 from the publisher Springer

369