Developing Functionalized Polymer Systems to Promote

Zachary K. Zander

DEVELOPING FUNCTIONALIZED POLYMER SYSTEMS TO PROMOTE

SPECIFIC INTERACTIONS AND PROPERTIES

A Dissertation

Presented to

The Graduate Faculty of The University of Akron

In Partial Fulfillment

of the Requirements for the Degree

Doctor of Philosophy

Zachary K. Zander

February, 2018

DEVELOPING FUNCTIONALIZED POLYMER SYSTEMS TO PROMOTE

SPECIFIC INTERACTIONS AND PROPERTIES

Zachary K. Zander

Dissertation

Approved: Accepted:

______Advisor Department Chair Dr. Matthew L. Becker Dr. Coleen Pugh

______Committee Member Dean of College Dr. Ali Dhinojwala Dr. Eric J. Amis

______Committee Member Dean of the Graduate School Dr. Bryan Vogt Dr. Chand Midha

______Committee Member Date Dr. Li Jia

______Committee Member Dr. Hazel Barton

ABSTRACT

Post-Fabrication, QAC-Functionalized Thermoplastic Polyurethanes for Contact-

Killing Catheter Applications. Catheter-related infections are an estimated $2.3 billion annual burden to the U.S. healthcare system, and result in approximately 28,000 deaths per year.1-2 To combat these infections, a TPU containing an allyl ether side-chain functionality (allyl-TPU) that allows for rapid and convenient surface modification with antimicrobial QACs is explored. A series of quaternary ammonium thiol compounds (Qx-

SH) possessing various hydrocarbon tail lengths (8 – 14 carbons) are synthesized and attached to the allyl-TPU surface using thiol-ene “click” chemistry, and antimicrobial testing of the QAC-functionalized TPUs reveal that Q8-SH is most effective against various bacteria. A prototype catheter is extruded and functionalized (post-fabrication) with Q8-

SH, and biofilm formation tests demonstrate its ability to inhibit biofilm accumulation.

Ionomers for Tunable Softening of Thermoplastic Polyurethane. Plasticizer migration and leaching leads to changes in material properties over time and produces environmental concerns. Thermoplastic polyurethane (TPU) sulfonate ionomers with quaternary ammonium counterions are synthesized to achieve soft TPUs without the use of low molecular weight plasticizers. The incorporation of a functional sulfonate monomer containing bulky ammonium counterions along the polymer backbone reduces the durometer hardness of the TPU by interfering with the polar interactions of the hard

iii segment, and disrupting the crystallinity. The synthetic procedure allows for facile tuning of the mechanical properties by increasing both the steric bulk of the counterion the feed ratio of ionic monomer, which decreases the durometer hardness of the TPU.

Control of Mesh Size and Modulus by Kinetically Dependent Cross-linking in

Hydrogels. Suitable control substrates are needed to probe the effects of mechano- transduction on stem cell differentiation. To address this, a hydrogel platform that utilizes tetra-polyethylene glycol (PEG) with modified chain ends for control of cross-linking kinetics and affords a range of substrate elasticities while maintaining the chemical composition of the gel is explored. Rheology and SANS experiments are performed to demonstrate how variations in cross-linking kinetics can be used to precisely control the modulus and microstructure of a gel, and the results indicate that increased structural heterogeneity results in lower moduli hydrogels.

DEDICATION

To my wife, Stephanie Zander, who has stood unwaveringly by my side in all facets of life, especially throughout my education. You are my most loyal supporter, my toughest critic, and my best friend.

ACKNOWLEDGEMENTS

The completion of this dissertation was made possible by the support of many individuals. First, I would like to thank my advisor, Dr. Matthew Becker, for his guidance and enthusiasm over the years. I am especially grateful for the level of respect and room for creative freedom he granted me while pursuing this degree. I would also like to thank my committee members, Dr. Ali Dhinojwala, Dr. Li Jia, Dr. Bryan Vogt, and Dr.

Hazel Barton, for creating a stimulating work environment through both collaboration and conversation. This work would not have been completed without their input, suggestions, and challenging inquiries.

Special thanks go to all the members of Dr. Becker’s group, past and present, whom I’ve had the honor of working with. Through many engaging conversations, brainstorming sessions, and occasional mischief, they made this monumental undertaking possible and much more enjoyable. I will always cherish the memories and companionship!

Last, but definitely not least, I would like to thank my family and friends. Their continuous love and support has kept me positive, grounded, and focused. Each of these individuals has shaped me into who I am, and helped me obtain my goals.

TABLE OF CONTENTS

Page

LIST OF TABLES ...... ix

LIST OF FIGURES ...... x

LIST OF SCHEMES ...... xvii

CHAPTER

I. INTRODUCTION ...... 1

1.1. Developing Functionalized Polymer Systems ...... 1

II. MATERIALS AND INSTRUMENTATION ...... 7

2.1. Materials ...... 7

2.2. Instrumentation ...... 9

III. POST-FABRICATION, QAC-FUNCTIONALIZED THERMOPLASTIC POLYURETHANES FOR CONTACT-KILLING CATHETER APPLICATIONS ...... 15

3.1. Abstract ...... 15

3.2. Introduction ...... 17

3.3. Experimental ...... 37

3.4. Results ...... 58

3.5. Conclusion ...... 73

3.6. Acknowledgment ...... 75

IV. IONOMERS FOR TUNABLE SOFTENING OF THERMOPLASTIC POLYURETHANE ... 76

4.1. Abstract ...... 77

vii

4.2. Introduction ...... 77

4.3. Experimental ...... 82

4.4. Results ...... 89

4.5. Conclusion ...... 105

4.6. Acknowledgement ...... 106

V. CONTROL OF MESH SIZE AND MODULUS BY KINETICALLY DEPENDENT CROSS- LINKING IN HYDROGELS ...... 107

5.1. Abstract ...... 107

5.2. Introduction ...... 108

5.3. Experimental ...... 111

5.4. Results ...... 119

5.5. Conclusion ...... 129

5.6. Acknowledgement ...... 129

VI. CONCLUSION ...... 130

6.1. Developing Functionalized Polymer Systems ...... 130

REFERENCES ...... 136

APPENDICES ...... 154

APPENDIX A. SUPPORTING FIGURES ...... 155

APPENDIX B. SUPPORTING SCHEMES & TABLES ...... 200

viii

LIST OF TABLES

Table Page

3.1. Contact-killing assay (ISO 22196) results ...... 68

4.1. List of ionic and non-ionic TPUs and their properties ...... 91

4.2. Tensile Properties of Non-ionic and Ionic TPUs with same wt.% Hard Segment . 99

5.1. Summary of the mechanical and structural properties determined for the kinetically-controlled hydrogels...... 125

6.1. Reagent quantities and yields for Qx-OH precursors...... 203

6.2. Reagent quantities and yields for Qx-S-S series...... 203

6.3. Reagent quantities and yields for Qx-SH series...... 203

6.4. Reagent table for compounds used in various TPU polymerizations...... 204

6.5. Molecular weight and physical properties for TPUs synthesized in this study. . 204

6.6. Quantification of rhodamine-SH and QAC present on UV treated and phys. ads. allyl-TPU samples as determined by fluorescence spectroscopy and XPS experiments, respectively...... 205

6.7. Additional contact-killing results for Qx-SH modified allyl-TPU samples...... 206

LIST OF FIGURES

Figure Page

3.1. Commonly used polymeric or polymer-coated medical devices for temporary implantation...... 19

3.2. The onset of biofilm formation is often facilitated by fouling of the device surface with proteins and other biological compounds (i.e. a conditioning film)...... 21

3.3. Classification of common strategies used to achieve antimicrobial and antifouling polymeric constructs for the prevention of DA-HAIs...... 24

3.4. Current commercial processing techniques for achieving antimicrobial or antifouling properties...... 31

3.5. Depiction of antifouling (repel) and antimicrobial (kill) surfaces and proposed dual-functional systems that either focus on killing incoming microbes while possessing an underlying repelling mechanism (kill-first), or repelling microbes while safeguarding with an underlying killing mechanism (repel-first)...... 35

3.6. Fluorescence data for the untreated control, phys. ads., and UV treated allyl-TPU samples modified using “click” reaction conditions with rhodamine-SH...... 63

3.7. (A) XPS high-resolution N1s spectra overlay of an untreated control, phys. ads., and UV treated sample demonstrating the appearance of a quaternary ammonium peak (400 – 402 eV). The solid lines represent raw data interpolated with a cubic b-spline curve, while the dashed lines represent the total curve fits + for each sample. (B) The % NR4 relative to urethane N is shown for UV treated and phys. ads. samples modified with the Qx-SH series...... 65

3.8. The live/dead contact-killing assay results for S. aureus after 5 min (top), and E. coli after 10 min (bottom) for control, phys. ads. and UV treated allyl-TPU blade- coated samples modified with Q8-SH...... 69

3.9. Brightfield microscopy images of catheter cross-sections (3.0 mm segments) from the 48 h biofilm assay were taken for (A) CC1, (B) untreated control, (C) phys. ads., and (D) UV treated samples modified with Q8-SH. (E) The % biofilm blockage was determined using Olympus VS-Desktop software and the averages

and standard deviations are displayed (n=3). (F) Photograph of the untreated control, phys. ads., and UV treated catheters following completion of the 48 h biofilm assay...... 71

3.10. SEM images demonstrating the appearance of bacterial EPS on catheter cross- sections from the 48 h biofilm assay were taken for (A) CC1, (B) untreated control, (C) phys. ads., and (D) UV treated samples modified with Q8-SH at 45× and (E-H) 300× magnification, respectively...... 72

4.1. Diagram representing the plasticization of TPU using bonded sulfonate groups with bulky quaternary ammonium counterions...... 81

+ 4.2. Structure of BES, where X is a quaternary ammonium group (NR4 ), and R = CH3 or (CH2)nCH3 with n = 6, 10, or 12...... 81

4.3. 1H-NMR overlay for BES (top) and TDA-BES (bottom)...... 90

4.4. 1H-NMR spectrum for TPU30(PE)-4.1TD...... 93

4.5. Time dependence of Shore A durometer values for a non-ionic TPU (■, TPU25(PE)Sn), and TPU ionomers with ca. 4% (●, TPU25(PE)-4.4TD) and 8% (▲, TPU25(PE)-7.6TD) TDA-BES monomer...... 95

4.6. Effect of hard segment content (wt%), ionic monomer concentration (mol%), and ammonium countercation on the durometer of non-ionic TPUs (■), nominal TPU(PE)-5TH (●), nominal TPU(PE)-5TD (▲), nominal TPU(PE)-8TD (♦)...... 96

4.7. Example DSC thermograms for TPU30(PE) (bottom), TPU30(2)-4.7TH (middle) and TPU30-4.1TD (top): (a) cooling following first heating scan; (b) second heating scan...... 97

4.8. (a) Representative stress vs. strain data for TPU25(PE) (■), TPU25(PE)-4.4TD (●) and TPU25(PE)-7.6TD (▲). (b) Shore A durometer vs. secant modulus at 50% strain (E50) for TPU25(PE), TPU25(PE)-4.4TD and TPU25(PE)-7.6TD...... 100

4.9. DMA data for (a) TPU30(PE), (b) TPU30(PE)-4.7TH, and (c) TPU30(PE)-4.1TD: E’ (■), E” (●), and tan δ (▲) are plotted as a function of temperature...... 103

4.10. Frequency dependence of the complex viscosity (η*) at 140 °C for TPU30(PE) (■), TPU30(PE)-4.7TH (●), and TPU30(PE)-4.1TD (▲)...... 104

5.1. Rheology data demonstrating tunable gelation times and moduli...... 122

5.2. The absolute scattering profiles of hydrogels fabricated at pH 5.7...... 124

5.3. Correlation between the mechanical properties (G’ = ■) and structural properties of mesh size (ξm = ●), and phase correlation length (δp = ▲)...... 128

6.1. 1H-NMR spectrum of Q14-OH demonstrates a 1:1 molar ratio of peaks e and g, which indicates the formation of the desired compound...... 155

6.2. 1H-NMR spectrum of Q12-OH demonstrates a 1:1 molar ratio of peaks e and g, which indicates the formation of the desired compound...... 156

6.3. 1H-NMR spectrum of Q8-OH demonstrates a 1:1 molar ratio of peaks e and g, which indicates the formation of the desired compound...... 157

6.4. 1H-NMR spectrum of 3,3’-dithiodipropanoyl chloride displays two triplets which confirms the purity of the compound, and demonstrates quantitative conversion to the acid chloride...... 158

6.5. 13C-NMR spectrum of 3,3’-dithiodipropanoyl chloride confirms the purity of the compound, and demonstrates quantitative conversion to the acid chloride. ... 159

6.6. 1H-NMR spectrum of Q14-S-S shows the appearance of peaks d and e, which are equimolar to peaks c, f, and g from the corresponding Q14-OH, indicating complete conversion to the desired disulfide...... 160

6.7. 1H-NMR spectrum of Q12-S-S shows the appearance of peaks d and e, which are equimolar to peaks c, f, and g from the corresponding Q12-OH, indicating complete conversion to the desired disulfide...... 161

6.8. 1H-NMR spectrum of Q8-S-S shows the appearance of peaks d and e, which are equimolar to peaks c, f, and g from the corresponding Q8-OH, indicating complete conversion to the desired disulfide...... 162

6.9. 1H-NMR spectrum of Q14-SH shows the proton resonances α and ß to the carbonyl (peak d) converge, and are equimolar to peaks c, e, and f from the corresponding Q14-S-S, indicating complete conversion to the desired thiol. .. 163

6.10. 1H-NMR spectrum of Q12-SH shows the proton resonances α and ß to the carbonyl (peak d) converge, and are equimolar to peaks c, e, and f from the corresponding Q12-S-S, indicating complete conversion to the desired thiol. .. 164

6.11. 1H-NMR spectrum of Q8-SH shows the proton resonances α and ß to the carbonyl (peak d) converge, and are equimolar to peaks c, e, and f from the corresponding Q8-S-S, indicating complete conversion to the desired thiol. .... 165

6.12. 13C-NMR spectra overlay of Q14-S-S and Q14-SH shows the shifting of the carbon α to the carbonyl downfield and the ß carbon upfield, indicating conversion to the desired thiol...... 166

xii

6.13. 13C-NMR spectra overlay of Q12-S-S and Q12-SH shows the shifting of the carbon α to the carbonyl downfield and the ß carbon upfield, indicating conversion to the desired thiol...... 167

6.14. 13C-NMR spectra overlay of Q8-S-S and Q8-SH shows the shifting of the carbon α to the carbonyl downfield and the ß carbon upfield, indicating conversion to the desired thiol...... 168

6.15. 1H-NMR spectrum of rhodamine B base...... 169

6.16. 1H-NMR spectrum of rhodamine B piperazine amide shows the introduction of peaks b, d, and i from piperazine, and the integration of peaks a and b indicate the amidation reaction was successful. HDO overlaps with peaks c and d...... 170

6.17. ESI-MS of rhodamine B piperazine amide shows the molecular ion [M]+ = 511.3 Da (calculated = 511.31 Da), as well as the doubly charged ion, [M]2+ at m/z = 255.7...... 170

6.18. 1H-NMR spectrum of rhodamine B 4-(3-hydroxylpropyl) piperazine amide demonstrates the appearance of peaks b and d, as well as the upfield shifting of the piperazine proton resonances (peaks e and c)...... 171

6.19. ESI-MS of rhodamine B 4-(3-hydroxylpropyl) piperazine amide shows the molecular ion [M]+ = 569.4 Da (calculated = 569.35 Da)...... 171

6.20. 1H-NMR spectrum of rhodamine B disulfide...... 172

6.21. ESI-MS of rhodamine B disulfide shows the doubly charged ion [M]2+ = 656.4 m/z which is 1312.8 Da (calculated = 1312.68 Da), as well as the triply charged ion, [M]3+ at m/z = 438.0...... 172

6.22. 1H-NMR spectrum of rhodamine-SH shows the proton resonances α and ß to the carbonyl (peaks e and f, respectively) converge, and are equimolar to all other peaks from the corresponding rhodamine B disulfide, indicating complete conversion to the desired thiol...... 173

6.23. ESI-MS of rhodamine-SH shows the molecular ion [M]+ = 657.4 Da (calculated = 657.35 Da) with minimal impurities...... 173

6.24. 13C-NMR spectra overlay of rhodamine B disulfide and rhodamine-SH shows the shifting of the carbon α to the carbonyl downfield and the ß carbon upfield, indicating conversion to the desired thiol...... 174

6.25. Normalized absorbance and emission spectra for rhodamine-SH. The λabs = 568 nm and the λem = 592 nm...... 174

xiii

6.26. Fluorescence standard curve for various concentrations of rhodamine-SH in DMSO...... 175

6.27. 1H-NMR spectrum of LAP...... 175

6.28. (A) UV-vis absorption spectra for LAP photoinitiator at several concentrations and (B) the absorbance at λ = 365 nm vs. concentration for determination of the molar absorptivity (ε) of LAP...... 176

6.29. 1H-NMR spectrum of a 30 wt.% (50 mol%) HMDI control TPU...... 176

6.30. 1H-NMR spectrum of 8% alloc-TPU shows the appearance of peaks “n” and “o” relative to the control TPU, which correspond to the protons from the allyl functional group...... 177

6.31. 1H-NMR overlay of various batches of 8% allyl-TPU normalized to the propylene peak of the polyether soft segment with an inlay displaying the allylic proton resonances (δ = 5.22 and 5.85 ppm)...... 178

6.32. SEC molecular weight plot for control TPU and 8% allyl-TPU batches...... 178

6.33. (A) Thermogravimetric analysis (TGA) was performed to determine the degradation profile for control and 8% allyl-TPU. (B) Expanded region showing the onset degradation temperature of 8% allyl-TPU is ca. 10 °C lower than the control...... 179

6.34. (A) Differential scanning calorimetry (DCS) first heating cycle and (B) cooling cycle...... 179

6.35. High resolution N1s XPS spectra of inner lumen of (A) phys. ads. and (B) UV treated catheter tubing (longitudinal sections)...... 180

6.36. (A) Diagram of the biofilm formation test, displaying the ordering of catheter segments (CC = Cook® Beacon® Tip Torcon NB® Advantage Catheter segments). (B) The eluent from biofilm formation testing 48 h post-inoculation was spread plated and displayed only P. aeruginosa colonies...... 180

6.37. 1H-NMR Spectrum for TDA-BES. Integration confirms 1:1 substitution and ionic monomer purity...... 181

6.38. 1H-NMR Spectrum for DDA-BES...... 182

6.39. 1H-NMR Spectrum for THA-BES...... 183

6.40. ESI-MS was utilized to confirm the purity of the ionic diol (THA-BES)...... 183

xiv

6.41. FT-IR spectra of ionomer product obtained after 2 and 3 h reaction time...... 184

6.42. 1H-NMR spectrum for 30 wt% hard segment control TPU with PE diol...... 185

6.43. 1H-NMR spectra overlay for various wt% control TPUs with PE diol...... 186

6.44. 1H-NMR spectrum for TPU30(PC)-3.8DD...... 187

6.45. 1H-NMR spectra overlay for various compositions of DDA-BES TPU ionomers with PC diol...... 188

6.46. 1H-NMR spectrum for TPU30(PE)-4.7TH...... 189

6.47. 1H-NMR spectrum for TPU30(PE)-7.6TD...... 190

6.48. SAXS data for TPU30(PE) control (■), TPU30(PE)-4.1TD (●) and TPU30(PE)-4.7TH (▲)...... 191

6.49. The boc-protected, 4-arm aminooxy crosslinker 1H-NMR spectrum shows complete substitution of the (boc-aminooxy)acetic acid to pentaerythritol; the ratio of peaks b and c are 1:1, indicating quantitative conversion...... 192

6.50. The 1H-NMR spectrum demonstrates successful deprotection of the boc- protecting group by the disappearance of the peak at 1.48 ppm, and provides the spectrum for the final 4-arm aminooxy crosslinker product...... 193

6.51. The ESI-MS reveals the molecular ion peak [M+H]+ at 429.1 Da (calculated 429.15 Da), as well as higher molecular weight species corresponding to sodiated molecular ions [M+Na]+ and hydration complexes, which are expected with a hygroscopic material...... 194

6.52. The 1H-NMR spectrum for keto-PEG shows a major peak (e) resulting from the ethylene glycol repeat unit...... 195

6.53. 1H-NMR spectrum expanded region (1.5 – 4.5 ppm) for keto-PEG reveals the end-group peaks, and demonstrates successful substitution of the levulinic acid to the 4-arm PEG by proton integrations...... 196

6.54. The MALDI-MS spectrum shows that a single distribution for the keto-PEG exists, confirming only tetra-substitution occurred with no minor species/series...... 197

6.55. Expanded MALDI-MS spectrum (10,360 – 10,480 Da) indicates a repeat unit of 44 Da which corresponds to PEG...... 197

6.56. The FT-IR spectrum shows the disappearance of the C=O stretch from keto-PEG (precursor) at 1718 cm-1, as the C=O stretch from the cross-linker (hydrogel) at 1765 cm-1 becomes more prominent...... 198

6.57. (A) Strain sweep was conducted to determine the LVR (0.1 – 10%) and 1% strain was used for subsequent frequency and time sweeps. (B) Frequency sweep on pre-formed hydrogel demonstrates minimal frequency dependence; 10 Hz was selected for reporting...... 198

6.58. The absolute scattering profiles of hydrogels fabricated at pH 6.8...... 199

6.59. The absolute scattering profiles of hydrogels fabricated at pH 7.1...... 199

xvi

LIST OF SCHEMES

Scheme Page

3.1. Post-fabrication, surface functionalization of allyl-TPU with Qx-SH reagents was carried out in DI water at room temperature using LAP photoinitiator and UV light (365 nm, I = 1.2 mW∙cm-2)...... 63

5.1. Tunable hydrogels were fabricated by mixing 4-arm crosslinker and keto-PEG precursors in phosphate-citrate buffer over a range of pH (5.7 – 7.1) and buffer concentrations (10 – 100 mm)...... 120

6.1. The various QAC compounds (Qx-OH) were produced via neat quaternization reactions of DOA, DDA, and DTDA (m = 6, 10, 12) with 8-chloro-1-octanol...... 200

6.2. 3,3’-dithiopropionic acid was treated with excess thionyl chloride and refluxed overnight to produce 3,3’-dithiopropanoyl chloride...... 200

6.3. The QAC disulfide reagents (Qx-S-S) were synthesized by esterification of the corresponding Qx-OH compounds with 3,3’-dithiopropanoyl chloride...... 200

6.4. The Qx-S-S reagents were reduced with TCEP to generate the corresponding Qx- SH compounds used for thiol-ene surface functionalization...... 200

6.5. Rhodamine-SH synthetic scheme beginning with the formation of the lactone, amidation with piperazine, nucleophilic substitution of 3-bromo-1-propanol, esterification with 3,3’-dithiopropanoyl chloride, and reduction to thiol using TCEP...... 201

6.6. LAP was synthesized using a Michaelis-Arbuzov reaction between the acid chloride and alkyl phosphonite to generate the acyl phosphinate, followed by treatment with LiBr...... 202

6.7. Allyl-TPU synthetic scheme...... 202

xvii

CHAPTER I

INTRODUCTION

1.1. Developing Functionalized Polymer Systems

Precise control of material properties and interactions can be achieved by incorporating chemical functionality into the polymer system. As polymer applications become more specialized and/or problems arise with existing material platforms, the need for functional materials becomes more apparent. For example: delamination of fluoropolymer coatings on catheter materials has led to widespread product recalls and spurred research into grafting neutral, hydrophilic polymers (such as polyethylene glycol, polyoxazolines, and polysulfobetaines) to material surfaces.3-5 Similarly, the leaching of hormone mimicking plasticizers from packaging materials, such as bisphenol A (BPA), has led researchers to explore alternative methods for plasticization.6-8 In addition, specialized polymer systems such as chemically identical hydrogels with mechanically tunable properties and peptide functionalized hydrogel microarrays can provide important biological insights.9-11 In this body of work, the development of several polymer platforms integrated with functional reagents and their ability promote specified properties and/or interactions were examined. A cursory introduction to each polymer system explored is provided here, with more extensive background information provided in subsequent chapters. 1

1.1.1. Contact-Killing Thermoplastic Polyurethanes

Estimates indicate that 5 million central venous catheters (CVCs) and >30 million urinary catheters are inserted annually in the U.S. with an incidence of infection between

3 – 8% and 10 – 30%, respectively.12-14 In many cases, these infections are biofilm- associated, which complicates their treatment and often necessitates device removal.12,

15-17 To prevent infection, the majority of current antimicrobial materials focus on releasing biocides and have demonstrated marginal success, with their primary drawback being the inevitable loss of activity once the anti-infective compound has been released.18-20 In addition, sub-lethal doses of antibiotics have been shown to accelerate resistance pathways and biofilm formation.21-22 To address this, researchers have turned attention to “contact-active” approaches. Contact-active materials feature monomers, functionalized side chains, or surface grafted moieties that are lethal to incoming bacteria upon contact. The majority of contact-active materials employ some form of quaternary ammonium compounds (QACs) as the biocidal component, which have demonstrated ability to kill bacteria while remaining non-cytotoxic to human cells.23 In contrast to biocide release methods, contact-actives are ideally non-leaching and should retain their activity for extended durations, limiting pathways for developing bacterial-resistance.24-

In this study, we explore a thermoplastic polyurethane containing an allyl ether side-chain functionality (allyl-TPU) that allows for rapid and convenient surface modification with antimicrobial reagents, post-processing. A series of quaternary

2 ammonium thiol compounds (Qx-SH) possessing various hydrocarbon tail lengths (8 – 14 carbons) were synthesized and attached to the allyl-TPU surface using thiol-ene “click” chemistry. A quantitative assessment regarding the amount of Qx-SH available on the surface following the “click” reactions is performed using fluorescence spectroscopy and

X-ray photoelectron spectroscopy (XPS). Contact-killing assays on functionalized TPUs were performed to screen a series of Qx-SH compositions for optimal antimicrobial activity against several microbes linked to catheter infections, and live/dead fluorescence staining examined their contact-killing efficiency. Scale-up, extrusion, and post- fabrication functionalization of allyl-TPU were performed with the most promising Qx-SH candidate, and the catheter prototype was tested for biofilm formation resistance to

Pseudomonas aeruginosa, a biofilm-forming species commonly associated with infections from indwelling medical devices.26

1.1.2. Soft Thermoplastic Polyurethane Ionomers

Plasticizer migration and leaching limits rubber longevity and produces environmental concerns.27-28 As a result, increasing restrictions on the use of traditional plasticizers have created a demand for alternative methods to achieve soft TPUs (shore A durometer < 50). The research described herein provides an alternative approach for softening TPUs (i.e., lowering the durometer and the melt viscosity), which involves incorporating bonded sulfonate groups with quaternary ammonium counterions into the

TPU backbone, similar to previous work to internally plasticize sulfonated polystyrene ionomers.29-30

Ionomers are polymers that contain a small concentration of covalently bonded ionic species, such as carboxylate, sulfonate or phosphonate groups.31 In most cases,

“hard” counterions, such as metal ions, are used to form the ion-pair, which in ionomers is condensed because of the relatively low dielectric constant of the polymer matrix.32

The interest in ionomers stems from the large property changes that result from interchain supramolecular bonding of the contact ion-pairs. These interactions represent transient, reversible crosslinks that generally increase the modulus, strength, and toughness of the ionomer, though some extensibility of the parent polymer is lost due to the formation of a physical network.33-35 Less common, is the addition of ionic functionality to a polymer for the purposes of internal plasticization. This can be achieved by using bulky counterions, e.g., alkyl ammonium or phosphonium ions that weaken the ionic, dipole-dipole, or ion-dipole interactions responsible for the mechanical and physical property changes.29-30

In this work, thermoplastic polyurethane (TPU) sulfonate ionomers with quaternary ammonium cations are synthesized to achieve soft TPUs without using conventional low molecular weight plasticizers. The sulfonated monomer N,N-bis(2- hydoxyethyl)-2-aminoethane-sulfonic acid (BES) neutralized with bulky ammonium counterions was incorporated as a chain extender to internally plasticize the TPU. The effect of increasing the steric bulk of the counterion and the concentration of the ionic species were examined and utilized to produce a range of soft TPUs. Characterization of the thermo-mechanical properties of the resulting TPU ionomers was examined by differential scanning calorimetry (DSC), dynamic mechanical analysis (DMA), shore A durometer hardness, and tensile testing.

1.1.3. Kinetically-Controlled Hydrogels

The favorable properties of PEG, including its hydrophilic nature, ease of end group modification, and limited immunogenicity and antigenicity, contribute to its widespread use in biomedical applications.36-41 In addition, reports have shown that human mesenchymal stem cells (hMSCs) sense the rigidity of their microenvironment and respond by altering their genomic signaling as well as differentiation and proliferation processes.42-49 Consequently, there has been significant interest in controllably perturbing the mechanical properties of synthetic scaffolds to direct the maturation of hMSCs; in particular, hydrogel systems with tunable mechanical properties that mimic the properties of native tissues and provide a favorable culture enviroment.38, 49-53 However, these systems generally require chemical or structural modification in order to tune the

5 moduli, which challenges the interpretation and control of cell studies that aim to examine solely the influence of substrate elasticity. Strategies to produce hydrogels with tunable elastticity for cell studies have included altering the precursor concentration or stoichiometry to modulate the effective cross-link density, adjusting precursor chain lengths to vary the molecular weight between cross-links, incorporating an additional or modified cross-linking agent, and changing the scaffold chemistry altogether.38, 43, 46, 49-58

To the best of our knowledge, few reports have demonstrated a covalently cross-linked hydrogel system that employs invariant precursor chemistry, concentration and stoichiometry to produce gels with tunable mechanical properties.10, 49 In this work, a

PEG-based hydrogel system that employs a kinetically-controlled oxime reaction for cross-linking, and allows for the production of mechanically distinguishable hydrogels without any alteration in precursor chemistry or composition will be explored. The resulting mechanical and structure properties of the gels will be examined using small- amplitude oscillatory shear (SAOS) rheology and small-angle neutron scattering (SANS).

The relationship between network heterogeneity and substrate elasticity will be analyzed to demonstrate how the influence of cross-linking kinetics on nanoscale structural details can be employed to induce changes in macroscopic mechanical properties.

CHAPTER II

MATERIALS AND INSTRUMENTATION

2.1. Materials

All commercial reagents and solvents were used as received without further purification unless otherwise noted. All solvents were reagent grade with purities ≥ 99%.

The chloroform-d (CDCl3) and dimethyl sulfoxide-d6 (DMSO-d6) were purchased from

Cambridge Isotopes Laboratories, Inc. Dimethyl sulfoxide (DMSO) was purchased from

i J.T. Baker, and diethyl ether (Et2O) and isopropyl alcohol ( PrOH) were purchased from

EMD Millipore. Calcium chloride (CaCl2), tryptic soy broth (TSB), Luria-Bertani (LB) agar, and Mueller-Hinton broth (MH) were purchased from VWR. Anhydrous methylene chloride (CH2Cl2), anhydrous tetrahydrofuran (THF), anhydrous dimethylformamide

(DMF), ethyl acetate (EtOAc), dimethylformamide (DMF), 2-butanone, tetrahydrofuran

(THF), methanol (MeOH), hexanes, didodecyldimethylammonium bromide (DDA-Br), tetrahexylammonium bromide (THA-Br), terakis(decyl)ammonium bromide (TDA-Br),

N,N-diisopropylethylamine (DIPEA), N,N-dimethyloctylamine (DOA), N,N- dimethyldodecylamine (DDA), N,N-dimethyltetradecylamine (DTDA), 3,3’-dithiopropionic acid, thionyl chloride (SOCl2), tris(2-carboxyethyl)phosphine hydrochloride (TCEP), 2,4,6- trimethylbenzoyl chloride, dimethyl phenylphosphonite, rhodamine B, trimethylaluminum solution (2.0 M in toluene), piperazine, 3-bromo-1-propanol, sodium 7 chloride (NaCl), sodium hydroxide (NaOH), sodium sulfate (Na2SO4), sodium bicarbonate

(NaHCO3), lithium bromide (LiBr), ammonium sulfate ((NH4)2SO4), magnesium chloride

(MgCl2), ethylenediaminetetraacetic acid iron(III) sodium salt (Fe-EDTA), 3-allyloxy-1,2- propanediol, tin(II) 2-ethylhexanoate (stannous octoate), 4,4’-methylenebis(phenyl isocyanate) (MDI), 4,4’-methylenebis(cyclohexyl isocyanate) mixture of isomers (HMDI),

Amberlite® IRN-78 hydroxide form, N,N-bis (2-hydoxyethyl)-2-aminoethane-sulfonic acid

(BES), 1,4-butanediol (BDO), 4-(dimethylamino)pyridine, p-toluenesulfonic acid monohydrate, (boc-aminooxy)acetic acid, levulinic acid, and diisopropylcarbodiimide

(DIC), and 4.0 M hydrogen chloride solution in dioxane (4M HCl/dioxane) were all purchased from Sigma-Aldrich. The 8-chloro-1-octanol was purchased from Alfa Aesar, and sodium phosphate dibasic (Na2HPO4), potassium phosphate monobasic (KH2PO4), sodium citrate, sodium citrate dihydrate, and casamino acids were purchased from Fisher

Scientific. The 4-arm, 10K polyethylene glycol (hydroxy-terminated) was purchased from

Creative PEGWorks. Silica (porosity = 60 Å, surface area = 450 – 550 m2∙g-1, bulk density

= 0.5 g∙mL-1, pH = 6.0 – 7.0) was purchased from Sorbent Technologies, Inc. Arcol-E351

-1 polyol (2,800 푀̅w, 38.5 – 41.5 mg KOH∙g ), and 2,000 푀̅푤 polycarbonate (PC) and polyester (PE) diols were kindly donated by Covestro and dried under high vacuum to remove residual solvents. The 4-(dimethylamino)-pyridinium-4-toluene sulfonate (DPTS) was prepared by adding molar equivalents of dimethylaminopyridine and p- toluenesulfonic acid monohydrate separately in THF, and collecting the precipitate by filtration.

The bacterial strains used in this study included Staphylococcus epidermidis (ATCC

12228), Staphylococcus aureus (25923), Escherichia coli (ATCC 25922), Pseudomonas aeruginosa (ATCC 27853), Enterococcus faecalis (ATCC 29212), methicillin-resistant

Staphylococcus aureus (MRSA) (ATCC BAA-41).

2.2. Instrumentation

Nuclear Magnetic Resonance (NMR): The 1H-NMR spectra were obtained using a

Varian NMRS 300 MHz spectrometer, using a relaxation time ranging between 1 s – 3 s and averaging over 32 – 64 scans. Chemical shifts were reported in ppm (δ), and

1 referenced to the chemical shifts of residual solvent resonances ( H-NMR: CDCl3 = 7.26

13 ppm, D2O = 4.79 ppm, DMSO-d6 = 2.50; C-NMR: CDCl3 = 77.16 ppm, DMSO-d6 = 39.52).

The relaxation time for 13C-NMR spectra was 2 s and 128 scans were taken.

Fourier-Transform Infrared Spectroscopy (FT-IR): FT-IR spectra were recorded by a

Digilab Excalibur Series FTS3000, with a scanned wavenumber range from 400 to 4000 cm-1. The spectra were recorded for 64 scans and the baseline was deducted and normalized to the C-H stretch peak intensity.

Electrospray Ionization Mass Spectrometry (ESI-MS): ESI-MS was performed using a HCT Ultra II quadrupole ion trap mass spectrometer (Bruker Daltonics, Billerica, MA) equipped with electrospray ionization (ESI) source. Samples were dissolved in MeOH and diluted to 0.01 µg∙mL-1 prior to injection. The sample solutions were injected into the ESI source by direct infusion, using a syringe pump, at a flow rate of 3 µL∙min-1. The tip of the

ESI needle was grounded, and the entrance of the capillary, through which ions enter in

9 the vacuum system of the mass spectrometer, was held at 3.5 kV. The pressure of the nebulizing gas (N2) was set at 10 psi, and the flow rate and temperature of the drying gas

-1 (N2) was 8 L∙min and 300 °C, respectively. Data collection was performed on positive mode and the ESI-MS data was analyzed by Bruker Daltonik’s DataAnalysis v4.0 software.

Matrix-assisted laser desorption Mass Spectrometry (MALDI-MS): MALDI-MS was performed using a Bruker UltraFlex III MALDI tandem time-of-flight (TOF/TOF) mass spectrometer (Bruker Daltonics, Billerica, MA, USA) equipped with a Nd:YAG laser emitting at 355 nm. The matrix and cationization salt were DCTB (2-[(2E)-3-(4-tert- butylphenyl)-2-methylprop-2-enylidene]malonitrile) and sodium trifluoroacetate, respectively. Solutions of the matrix (20 mg∙mL-1) and cationizing salt (10 mg∙mL-1) were prepared in THF. The matrix and cationizing agent solutions were mixed in 10:1 (v/v) ratio and applied to the target. After drying, a spot of the sample was applied, followed by an additional drop of the matrix/cationizing agent.

pH Meter: The pH values of buffers were tested using an Orion® 350 PerpHecT® benchtop pH meter with an Orion® ROSS® Sure-Flow pH electrode at room temperature.

Small-Amplitude Oscillatory Shear (SAOS) Rheology: SAOS measurements were performed using a TA Instruments ARES G2 Rheometer (TA Instruments, New Castle, DE) equipped with 8 mm parallel plate geometry. The linear viscoelastic response region was determined with a strain sweep conducted at a frequency of ω = 1 rad∙s-1. Frequency sweeps were conducted at 1% strain from 100 rad∙s-1 to 0.1 rad∙s-1. Time sweeps were

-1 performed at 1% strain and 1 rad∙s using 25 mm parallel plates and a set gap height of

0.80 mm.

Small-Angle Neutron Scattering (SANS): SANS measurements were taken at the

National Institute of Standards and Technology Center for Neutron Research (NCNR) using the instrument NGB30. The scattering “wavevector” q was measured, where 푞 =

4휋 휃 ( ) ∗ sin ( ), λ is neutron beam wavelength and θ is scattering angle. Three detector 휆 2 distances (13.2 m, 4 m, and 1.3 m), and the use of lenses for the 13.2 m detector distance, were examined to provide a measured q range of 0.001 Å-1 to 0.5 Å-1. The scattering results were circularly averaged over the 2D detector to attain the 1D scattering of q versus intensity. The data was fit with using IGOR Pro software.

Shore A Durometer: Durometer measurements were performed using a Folwer®

Shore A Portable Durometer (Folwer High Precision, Auburndale, MA), following the procedure described in ASTM D2240. The instrument was calibrated using a standardized

Shore A 50 material prior to each measurement. A type A-2 Shore A durometer hardness tester (Shore Instrument & MFG Co. New York) was also used and calibrated with a standardized Shore A 60 material prior to each measurement.

Differential Scanning Calorimetry (DSC): DSC was performed using a TA

Instruments Q2000 DSC on sample sizes of ca. 10 – 15 mg using temperature ramps for heating between 20 – 30 °C∙min-1 and a cooling rate between 20 – 30 °C∙min-1.

Thermogravimetric Analysis (TGA): Thermogravimetric analysis was performed using a TA Instruments TGA 2950 (TA Instruments – Waters L.L.C, New Castle, DE) on sample sizes of ca. 10 mg using a heating ramp of 10 °C∙min-1, after holding temperature for 5 min at 110 °C to remove water.

Dynamic Mechanical Analysis (DMA): DMA was performed using a TA Instruments

Q800 Dynamic Mechanical Analyzer. Dynamic tensile measurements were made using strain amplitudes (ε) = 0.1 - 5 % depending on temperature, and the temperature was scanned from -80 °C to 140 °C using a heating rate of 2 °C∙min-1 (ε = 0.2%, ω = 1 rad·s-1).

Small-Angle X-ray Scattering (SAXS): SAXS experiments were performed using a

Rigaku MicroMax 002+ equipped with a 2D multiwire area detector and a sealed copper tube (CuKα radiation, λ = 1.54 Å). The voltage and current for the X-ray tube were 45 kV and 0.88 mA, respectively.

Tensile Testing: Uniaxial tensile tests were carried out on an Instron Universal

Testing Machine (Model 5567) in accordance with ASTM D638 for dumbbell shape Type

V using an extension rate of 500 mm∙min-1 and a 1 kN load cell.

Size Exclusion Chromatography (SEC): Size exclusion chromatography (SEC) was performed using an EcoSEC HLC-8320GPC (Tosoh Bioscience LLC, King of Prussia, PA) equipped with a TSKgel GMHHR-M mixed bed columns and refractive index (RI) detector.

Molecular weights were calculated using a calibration curve determined from poly(styrene) standards (PStQuick MP-M standards, Tosoh Bioscience LLC) with THF as eluent flowing at 1.0 mL∙min-1, and a sample concentration of 4 mg∙mL-1.

UV-vis and Fluorescence Spectroscopy: UV-vis and Fluorescence spectroscopy was carried out using a BioTek SynergyTM Mx Microplate Reader (BioTek, Vermont) with Gen

5TM reader control and data analysis software.

Fluorescence Microscopy: Fluorescence microscopy was performed using an IX81 inverted microscope (Olympus, Center Valley, PA) with FITC and TRITC filters at 400× magnification. Brightfield microscopy was performed at 40× magnification. Images were processed and analyzed/quantified using ImageJ and Olympus VS-desktop software.

X-ray Photoelectron Spectroscopy (XPS): The XPS spectra were obtained using a

VersaProbe II Scanning XPS Microprobe from Physical Electronics (PHI), under ultrahigh vacuum conditions with a pressure of 2.0 µPa. Automated dual beam charge neutralization was used during the analysis of the samples to provide accurate data. The analyzer pass energy was 117.4 eV for the survey spectra and 23.5 eV for the high- resolution scans in the N1s regions. Survey scans in the range of 0 – 700 eV were used to evaluate the percentage of different atoms present on the surface of the samples. Atomic concentrations were calculated with PHI MultiPak software. The XPS high-resolution spectra of N1s were decomposed into two components by using the curve fitting routine in MultiPak. A goodness of fit (χ2) better than 1.5 was achieved for each fit. Each spectrum was collected using a monochromatic (Al Kα) x-ray beam (E = 1486.6 eV) over a

100 μm × 1400 μm probing area with a beam power of 100 W.

Scanning Electron Microscopy (SEM): SEM was performed on gold sputter-coated samples using a JEOL-7401 Field Emission Scanning Electron Microscope (JEOL USA, Inc.,

Peabody, MA) at an accelerating voltage of 2.0 kV under 45× and 300× magnification.

CHAPTER III

POST-FABRICATION, QAC-FUNCTIONALIZED THERMOPLASTIC POLYURETHANES FOR

CONTACT-KILLING CATHETER APPLICATIONS

In part, this work has been reprinted with permission from Zander, Z. K.; Becker,

M. L., Antimicrobial and Antifouling Strategies for Polymeric Medical Devices. ACS

In part, this work has been submitted for publication in Biomaterials as Zander,

Z.K; Chen, P.; Hsu, Y.; Dreger, N.Z.; Savariau, L.; McRoy, W.C.; Cerchiari, A.E.; Chambers,

S.D.; Barton, H.A.; Becker, M.L., Post-Fabrication, QAC-Functionalized Thermoplastic

Polyurethanes for Contact-Killing Catheter Applications. Biomaterials 2018, submitted.

3.1. Abstract

Estimates indicate that 5 million central venous catheters (CVCs) and >30 million urinary catheters are inserted annually in the U.S. with an incidence of infection between

3 – 8% and 10 – 30%, respectively.12-14 In many cases, these infections are biofilm- associated, which complicates their treatment and often necessitates device removal.12,

15-17 This study examines a thermoplastic polyurethane containing an allyl ether side- chain functionality (allyl-TPU) that allows for rapid and convenient surface modification with antimicrobial reagents, post-processing. A series of quaternary ammonium thiol compounds (Qx-SH) possessing various hydrocarbon tail lengths (8 – 14 carbons) are 15 synthesized and attached to the surface using thiol-ene “click” chemistry. A quantitative assessment regarding the amount of Qx-SH available on the surface is performed using fluorescence spectroscopy and X-ray photoelectron spectroscopy (XPS). Contact-killing assays suggest the Q8-SH composition has the highest antimicrobial activity, and a live/dead fluorescence assay reveals the rapid killing of S. aureus (>75% in 5 min) and E. coli (90% in 10 min) inocula. Scale-up and extrusion of the allyl-TPU provides catheter tubing for biofilm formation testing with P. aeruginosa, and surface-functionalized catheters modified with Q8-SH demonstrate their ability to reduce biofilm formation.

3.2. Introduction

3.2.1. Device-Associated Infections

The Centers for Disease Control and Prevention (CDC’s) National Healthcare Safety

Network revealed that 722,000 hospital-acquired infections (HAIs) occurred in U.S. acute care facilities during 2011.59 This corresponds to 1 in 25 inpatient admissions developing an HAI which ultimately led to more than 75,000 patient deaths. A significant portion of these HAIs (>25%) were directly associated with implanted medical devices.59 These statistics do not include HAIs that occur in intensive care units (ICU) where the overall number of patients is less, but the risk of mortality from infection is increased greatly as a consequence of the patient’s diminished health status. For example: there are nearly

80,000 central venous catheter (CVC) associated bloodstream infections that occur each year in U.S. ICUs, with a 12-25% mortality rate.2 The medical costs associated with these patient complications alone are estimated between $296 million to $2.3 billion, annually.1-2

In addition, urinary tract infections (UTIs) are believed to account for 30-40% of

HAIs world-wide, and 80% are directly linked to catheterization, i.e. catheter-associated

UTIs (CA-UTIs).13, 60 Although the mortality rate for CA-UTIs is <5%, their usage is 6 times greater than CVCs, making CA-UTIs one of the leading causes for nosocomial bloodstream infections (BSIs).61 In developing countries, it is expected that the device-associated infection rates in ICUs are 3 to 5 times higher.62 Device-associated HAIs (DA-HAIs) are not always caused by host response to the device and/or patient bacteria; research has

17 suggested HAI prevention programs that educate hospital personnel on infection-control practices can reduce DA-HAIs by as much as 50-70%.2, 63 The conclusions from these reports, however, also suggest that achieving 100% prevention may not be possible, even with comprehensive guidelines in place. Hence, there is a need for antimicrobial and antifouling devices to assist medical professionals when a lapse in infection-control guidelines occurs. To reduce or eliminate DA-HAIs, the role of antimicrobial or antifouling devices would be to inhibit the growth and reduce/eliminate the bacterial contamination, or to prevent the initial attachment and subsequent colonization of the device.

According to the U.S. Food and Drug Administration (FDA), medical devices can range from tongue depressors to programmable pacemakers, and are intended to (i) diagnose, cure, mitigate, treat, or prevent disease; (ii) or to repair, improve, or replace some essential structure or function of the body.64 This definition encompasses a large body of devices. However, the focus of our research will be primarily on indwelling polymer/polymer-coated devices (Figure 3.1). The emphasis on these devices is appropriate when considering their annual usage, infection rates, and attributable mortality.12, 65-66 Notably, it is estimated that >30,000,000 urinary catheters and

5,000,000 CVCs are inserted annually, with infection rates between 10-30% and 3-8%, respectively.12-13 Accordingly, this introduction will focus primarily on the development of DA-HAIs, active research and current commercial approaches for antimicrobial and antifouling polymeric devices/coatings, as well as a future outlook on preventing medical device infections.

Figure 3.1. Commonly used polymeric or polymer-coated medical devices for temporary implantation. This figure includes (A) endo-tracheal tubes, (B) peritoneal catheters, (C) urinary catheters, (D) fracture fixation devices with polymer coated pins, and (E) and central venous catheters.

3.2.2. Pathogenesis

Biofilm formation has been identified as the critical event in the pathogenesis of catheter-related infections (CRIs) and other DA-HAIs.16-17, 67 A biofilm is a community of microbial cells that have attached to a substrate and secreted an exopolysaccharide matrix to form the structure of the film.68 The primary reason biofilms are problematic is because they are significantly more resistant to antimicrobial treatment.16, 21, 67 The resistance of biofilm dwelling bacteria is believed to be a result of several mechanisms, including delayed antimicrobial uptake through the biofilm matrix, physiological changes such as decreased growth rates and the development of persister cells, as well as rapid

19 plasmid exchange between microorganisms within the film.68-69 Biofilms that host multiple bacterial species have a mixed-species fitness advantage; in most cases, antibacterial chemotherapy is not sufficient to treat infections caused by these biofilms and device removal is required.12, 15-16, 70

Interestingly, the process of biofilm formation most commonly begins with the adhesion of non-pathogenic or commensal bacteria to the device surface, such as

Staphylococcus. epidermidis and Staphylococcus aureus (Figure 3.2).15-16, 69 The initial contamination likely comprises a small population of microorganisms that attach to the surface either during device insertion through contact with the patient’s/medical professional’s skin flora, or by bacterial migration from the insertion site.13, 17 Once the bacteria have attached to the surface, they begin to multiply, co-adhere, and produce an insoluble matrix of gelatinous exopolymers to form microcolonies. Further maturation of the biofilm, and potentially infiltration by secondary microbial species, leads to the formation of a mature biofilm. In a mature biofilm, the bacteria can detach and become planktonic or portions of the film may slough off creating a bolus of infection, which presents a potentially life-threatening situation for the patient. A multitude of reports are available that describe in detail the process and signaling involved with each step of biofilm formation on a device surface.61, 71-74 Preventing biofilm formation has relied on techniques that aim to prevent microbial adhesion (antifouling) or to eliminate adhering microbes (antimicrobial). In both approaches, however, the role of a conditioning film is often overlooked.

Figure 3.2. The onset of biofilm formation is often facilitated by fouling of the device surface with proteins and other biological compounds (i.e. a conditioning film). The process of biofilm formation begins with 1) bacterial attachment, followed by 2) co- adhesion and matrix production, 3) microcolony formation, 4) further maturation, and 5) results in a mature biofilm that can disperse planktonic bacteria. There is also potential for infiltration by a secondary species during this process. (This figure was modified with permission from the Annual Review of Microbiology, Volume 56 © 2002 by Annual Reviews, http://www.annualreviews.org).75

Upon insertion of a medical device, the device surface becomes rapidly coated with proteins and other host molecules forming what is known as a conditioning film.73,

76 Even if a polymeric device or coating is initially inhospitable for microbial attachment, the accumulation of a conditioning film can present favorable binding sites which will eventually lead to microbial attachment and biofilm formation (Figure 3.2). This issue was highlighted in a recent commentary from an FDA and NSF cosponsored workshop in which Phillips, et al. state: “There is ample evidence to support in vitro biofilm prevention capabilities of antibiofilm agents and technologies. However, the most current in vitro test methodologies usually do not incorporate in vivo device conditions…”77 Thus, it is important for researchers developing said technologies to consider the microenvironment of the location intended for device use, and attempt to simulate or 21 screen for non-specific adsorption of the corresponding host molecules. For example: the conditioning film in vascular catheters is typically an accumulation of platelets and plasma proteins such as albumin, fibrinogen and fibronectin, whereas urinary catheters usually become coated with proteins and electrolytes from the patient’s urine.61, 76, 78 It has been shown that one of most common microbial species responsible for DA-HAIs (S. aureus) can adhere to various proteins via site-specific adhesion receptors and produce coagulase enzymes that further promote bacterial adhesion and thrombogenesis.79-83 Alternatively, significant adsorption of plasma proteins alone may cause platelet adhesion and activation, leading to surface-induced thrombogenesis.84-86 Therefore, while it is necessary for a device or coating to thwart microbial attachment, a more sustainable biofilm prevention method would also prohibit non-specific adsorption of proteins and other biological compounds.

To reduce the frequency of DA-HAIs, many physicians resort to administering antibiotics, either orally or intravenously, prior to or immediately following device implantation.87 Biofilm-associated infections that develop on medical devices and in surrounding tissues are often treated with antimicrobial therapy to no avail due to the nature of the biofilm, and device removal/revision surgery is required.88-89 In addition, antibiotics used to target biofilm infections have limited activity towards bacteria infecting peri-implant tissues, requiring additional antibiotic treatment (e.g. rifampicin) that functions as a monotherapy towards these pathogens and results in high risk for developing resistance.90 Thus, biofilms contribute to a multi-fold problem as a

22 consequence of the infections they cause and the method by which they are treated. The looming fear associated with bacterial-resistance and a lack of new antibiotics to fight them has highlighted our need to prevent and care for infections using alternative methods.91

3.2.3. Active Research

Active research and technology in current polymeric devices or coatings that aim to prevent DA-HAIs have been reviewed extensively.3, 20, 60, 70, 92-93 The strategies employed can be divided into two categories, antimicrobial and antifouling (Figure 3.3): antimicrobial materials are inhibitory or lethal to approaching microorganisms, whereas antifouling materials prevent the adhesion of microbes and proteins. The primary mode of action for antimicrobial activity is achieved through (i) biocide release or (ii) the presence of contact-killing moieties near the surface. Antifouling strategies generally employ one of several mechanisms to prevent adhesion including (i) steric repulsion/hydration, (ii) specific protein interactions, or (iii) low surface energy.60, 92

Henceforth, the active research pertaining to these strategies and their application in current medical devices will be discussed.

Figure 3.3. Classification of common strategies used to achieve antimicrobial and antifouling polymeric constructs for the prevention of DA-HAIs. This figure was adapted from Siedenbiedel and Tiller in Polymers 2012.92

3.2.3.1. Antimicrobial Strategies

Antimicrobial materials that employ biocide release approaches have demonstrated marginal success, with their primary drawback being the inevitable loss of activity once the anti-infective compound has been released or is no longer available at lethal concentrations. In addition, sub-lethal doses of antibiotics have been shown to accelerate resistance pathways and biofilm formation.21-22 For a comprehensive review of release strategies, the reader is referred elsewhere.18-20 However, several alternative release-based strategies that aim to prevent biofilm formation are gaining significant attention. These methods include the use of bacteria-specific antibodies for opsonization, gallium and iron complexes to interfere with bacteria/biofilm metabolism of iron, and nitric oxide release as a mimic of the natural immune response exerted by macrophages.18, 21, 70, 94 24

The alternative to release-based strategies involves a “contact-active” approach.

Contact-active materials feature monomers, functionalized side chains, or surface grafted moieties that are lethal to incoming bacteria upon contact. The majority of contact-active materials employ quaternary ammonium compounds (QAC), host defense peptides or mimics thereof, and other cationic moieties as the biocidal component. The mechanism of their activity is believed to be the disruption of the microbial cell wall/membrane; this is achieved either by opposite charge attraction and subsequent penetration of the active group leading to the disruption of the phospholipid bilayer, or by creating a charge imbalance which breaks down the membrane potential.23 In contrast to biocide release methods, these biocides are non-leaching and should not lose their activity nor present pathways for developing bacterial-resistance.24-25 A large degree of literature has already explored various synthetic methods for producing contact-active materials based on cationic monomers, side chains, and covalent surface grafts.23, 60, 92-93 Accordingly, the use of synthetic mimics of antimicrobial peptides (SMAPs) for modification of polymer surfaces and backbones is also being intensely explored, and suggests that these are potent contact-active compounds. Several studies have shown that SMAPs and amphiphilic quaternary ammonium compounds, can retain bactericidal activity when anchored to a surface, and are non-cytotoxic towards mammalian cells.95-97 The primary drawback of this strategy, however, is the accumulation of intracellular components/debris from dead bacteria that eventually mask the contact-active surface properties. Further work in this area should focus on incorporating a release mechanism,

25 or utilize the contact-active as a complementary component to existing antifouling technologies. Examples of such release mechanisms include the use of hydrolysis to reversibly switch between contact-active QAC and low-fouling zwitterionic surface moieties, enzyme degradable layer-by-layer assemblies of alternating biocidal and repelling polymers, and pH triggered collapse of grafted antifouling chains to expose an underlying antimicrobial motif.98-102

3.2.3.2. Antifouling Strategies

Antifouling polymers and coatings utilize several strategies for resisting bacteria and protein attachment (Figure 3.3). The most widely investigated mechanism employs hydration forces and/or steric repulsion to generate fouling-resistant surfaces. Steric repulsion is included here because grafted polyethylene glycol (PEG) accounted for the majority of the original systems examined to divulge whether entropically unfavorable chain compression or hydration forces were responsible for the observed antifouling properties.103 While steric repulsion may play a role, it is now understood that hydrophilicity alone can impart surfaces that possess a tightly correlated water layer, which creates a physical and energetic barrier and causes interactions with approaching proteins or bacteria to become thermodynamically unfavorable.104-106 Although PEG has been considered the gold standard in antifouling materials, it is known to oxidize under biologically relevant conditions in vitro, which could lead to the destruction of the hydration layer.107-109 However, claims regarding the oxidative degradation of PEG in vivo are rarely proven, and are even contradicted by some reports.107 Nevertheless, efforts to

26 find alternatives to PEG with higher oxidative stability have led many researchers to explore polyglycerols, polyoxazolines, and a variety of zwitterionic polymers, side chains, and surface grafts.3, 110 Neutral, hydrophilic PEG alternatives such as poly(glycerol) and poly(2-methyl-2-oxazoline) have demonstrated comparable protein resistance to PEG controls, and improved oxidative stability.111-113 In addition, systems which incorporate zwitterionic moieties have demonstrated efficacy at preventing fouling through the formation of a tightly bound (ionically solvated) hydration layer.104, 114-115 One such system utilized a graft from redox polymerization process to modify peripherally inserted central catheter (PICC) surfaces with zwitterionic polymeric sulfobetaine (polySB). The polySB modified PICCs demonstrated successful reduction of microbial attachment and thrombosis in vitro and in vivo (canine model).116 Because of their synthetic flexibility and orthogonality toward other functional groups, zwitterionic compounds and polymers could be the next generation of antifouling materials for medical device applications.

Additional antifouling strategies utilize specific protein interactions to prevent bacterial adhesion and non-specific adsorption of other proteins. A classic example of this method involves the passivation of surfaces with albumin, a non-adhesion protein that hinders cell attachment and blocks non-specific protein adsorption.117 For example, albumin coated tympanostomy tubes have been shown to reduce adhesion of foreign materials within the ear canal.118-119 However, in blood-contacting devices albumin is eventually replaced by adhesive proteins with higher substrate affinity and leads to device fouling.120-121 Surfaces that are modified by adsorption or covalent attachment of heparin

27 also elicit specific protein interactions that effectively reduce device fouling.122-123

Although heparin is primarily employed to prevent thrombus formation by binding antithrombin and changing its structure to accelerate antithrombin-mediated inhibition of clotting factors, it has demonstrated efficacy at preventing bacterial colonization.124-126

The mechanism by which heparin reduces bacterial adhesion to device surfaces is not known. Since heparin contains the highest negative charge density of any known biological macromolecule, it is generally assumed that bacteria are repelled through electrostatic interactions. However, electrostatic repulsion of bacteria has only been observed under controlled environments, and some negatively charged surfaces have demonstrated higher protein adsorption in conjunction with reduced bacterial adhesion suggesting that protein interactions may be responsible.127-129 Furthermore, heparin- coated surfaces have demonstrated selective plasma protein adsorption wherein lower amounts of fibrinogen and fibronectin were adsorbed, and studies have shown that heparin inhibits binding of bacterial adhesins to fibronectin in vitro.130-131 The vast landscape of solid substrates being explored for medical device applications and their specific interactions with serum proteins is poorly understood. Additional computational modeling and experimentation that examines general surface features (charge density, surface energy, topography, etc.) and their influence on the binding and conformational changes of common plasma proteins could prove useful in this field.132-134

Low surface energy materials exploit the principals of surface free energy and the work of adhesion to achieve low fouling. When the surface free energy is reduced and the interfacial tension between the liquid and substrate is high, the work of adhesion is minimized.135 The majority of low surface energy materials are hydrophobic polymers; fluoropolymers such as poly(tetrafluoroethylene) (PTFE) and silicones such as poly(dimethylsiloxane) (PDMS) have been explored extensively because they possess among the lowest of surface energies (<25 mN/m), chemical and thermal stability, and are presumably bioinert.117, 136-137 Although, in vitro and in vivo testing of these materials for medical device applications have demonstrated their tendency to suffer from irreversible, non-specific absorption of proteins.117, 138 In fact, the perceived bioinert properties of fluoropolymers demonstrated in clinical applications may actually be a result of surface passivation with high proportions of albumin, hindering cell attachment and thrombosis.117, 139 Alternatively, recent discoveries in the quest for bio-inspired materials have shown that surface topography/roughness can impart super-hydrophobic properties, which may prove useful in reducing the non-specific adhesion of proteins and bacteria.140 Technologies that employ surface topography to prevent bacterial adhesion, such as those developed by SharkletTM, are in a unique position to play a larger role in the future of polymeric medical devices and coatings, since they are theoretically unimpeded by bacterial resistance.141 Although this technology has demonstrated efficacy at inhibiting the growth of E. coli and S. aureus in vitro, its ability to prevent infection under physiologically relevant conditions is yet to be determined.142-143

3.2.4. Current Technology

Existing medical device and medical-grade coating manufacturers have found several useful methods for incorporating antimicrobial and antifouling functionalities into materials. For antimicrobial polymers, commercialized products have primarily focused on releasing biocides, which can be physically adsorbed or impregnated into the material

(Figure 3.4). The most commonly used biocides include silver compounds or ions, chlorhexidine, and antibiotics such as rifampin and minocycline.20, 60 Physical adsorption can be achieved by simply submerging the device into a biocide-containing solution and allowing the biocidal compounds to bind through non-covalent interactions. This process is often facilitated by pre-conditioning the surface with surfactant compounds to improve the binding affinity and slow the release of the biocide. As an example, Cook Spectrum®

CVCs contain a mixture of minocycline and rifampin antibiotics that are physically adsorbed to the catheter surface using a surfactant coating.144-145 For impregnation, thermostable biocides may be melt processed with a base polymer so that the biocidal entity is homogenously dispersed throughout the material. This method has been primarily utilized for catheters impregnated with silver compounds or ions, such as Agion® and Oligon CVCs.60, 144 However, for temperature sensitive biocides, impregnation can also be achieved by swelling the material in a suitable solvent containing the desired compound.146 An additional method that has been more recently developed, and has been active in the literature for some time, is the bulk incorporation of a biocidal unit

(Figure 3.4).92-93, 147 This approach involves the end-functionalization, side chain

30 functionalization, or polymerization of a biocidal monomer (such as quaternary ammonium salts and biguanides) into the polymer backbone. As opposed to biocide release, this method provides a contact-active moiety. Technologies of this type appear to be gaining some ground, and may already be employed in DSM’s ComfortCoatTM non- silver based antimicrobial coatings.148

Figure 3.4. Current commercial processing techniques for achieving antimicrobial or antifouling properties. These methods incorporate an active moiety through physical or covalent interactions with the base polymer.

Additional polymer devices and coatings available on the market exploit antifouling strategies to aid in the prevention of DA-HAIs. Most notably, a variety of polyelectrolyte and PEG based coatings that utilize principles of steric repulsion and hydration forces to prevent device fouling are employed. These materials are typically applied to the device/base polymer surface via spray-coating or dip-coating, followed by a curing step.149-151 The coatings are often imbedded with biocides or used as an adsorbent for biocidal moieties.152 Other manufacturing methods used for incorporating 31 antifouling moieties include covalent attachment and migrating additives (Figure 3.4).

The covalent attachment of heparin using PhotoLink® technology has been applied in

SurModics® hemocompatible coatings. An additional method that allows for covalent

“end point” attachment of herapin through the use of a priming layer is known as CBAS®

Heparin Surface, which is currently applied in Gore® vascular devices.153-154 Lastly, migrating additive techniques have been primarily focused on fluoropolymers.

AngioDynamics licenses EndexoTM technology developed by Interface Biologics in their venous catheters, which utilizes low molecular weight fluoropolymers that bloom to the surface much like a plasticizer. This method provides the favorable low surface energy properties of fluorinated polymers, while reducing the likelihood of delamination or flaking that has prompted FDA recalls for multiple devices containing (PTFE) and other lubricous coatings.5 The use of migrating additives also negates challenges in characterizing coating defects/delamination, since the additive is considered an integral part of the base material. Of the existing products and processing methods discussed, there is clearly potential for crossover potential regarding how antimicrobial and antifouling moieties may be incorporated into polymeric devices and coatings.

Researchers are encouraged to consider how they might incorporate their system into existing manufacturing processes to achieve materials that exhibit both antimicrobial and antifouling properties.

3.2.5. Outlook on Combating Device-Associated Hospital Acquired Infections

The cost of device-associated infections from a mortality and financial perspective emphasizes the need for preventative measures. It has become apparent that HAIs resulting from the implantation of polymeric devices cannot be prevented using aseptic techniques.2, 63 In addition, biofilm formation has been identified as a critical event leading to DA-HAIs.16-17, 67 Upon insertion, the conditioning of a device with proteins and other biological compounds facilitates microbial adhesion and subsequent biofilm formation.73, 76 Thus, it is imperative that in vitro testing of developing technologies aims to simulate the conditions of the target implant site in vivo to improve clinical translation.

This could include the development of protocols and utilization of human serum ex vivo for monitoring protein adsorption, and testing antimicrobial activity in the presence of serum proteins.

Active research and current technologies are exploring the effectiveness of various functionalized polymeric devices and coatings, which has provided a multitude of active moieties and methods for their incorporation. One problem with existing antimicrobial and antifouling technologies is their loss in effectiveness over time.21, 155

Covalent attachment and bulk incorporation of contact-active moieties are an intriguing alternative for sustained antimicrobial activity, although concerns regarding fouling with biomolecules and deceased microbes exists. In addition, neutral, hydrophilic materials and zwitterionic functionalities are demonstrating efficacy at preventing bacterial attachment and non-specific adsorption of proteins. 104, 111, 113 However, more studies are

33 needed to determine their effectiveness in vivo, in addition to reliable methods for attaching such moieties to the surface of medical devices. To date, antifouling and antimicrobial polymer-based strategies have struggled to prevent DA-HAIs in clinical settings, as evidenced by high infection rates.2, 12, 59 Therefore, consideration towards the use of dual-functional materials should be taken while understanding the necessity for a translationally relevant solution.

As stated by Zou, et al., “It takes walls and knights to defend a castle.”155 This underscores the concept of utilizing antifouling techniques to prevent bacteria and protein adhesion in conjunction with antimicrobial moieties that are lethal towards bacteria (Figure 3.5). As depicted, the antifouling mechanism could be a hydration layer, and the antimicrobial could be contact-active. However, any variation or combination of the methods and moieties discussed in this viewpoint may apply. For example, the release of nitric oxide from a nanopatterned surface, or the immobilization of QACs within a hydrophilic surface coating. Initial work pertaining to the concept of dual- functionalization is already underway.101, 156-159 What is yet to be seen is whether a “kill- first” or “repel-first” mechanism is more effective for sustainable prevention of DA-HAIs.

However, one significant challenge towards achieving such dual-functional materials is the need for reliable and orthogonal surface functionalization methods.

Figure 3.5. Depiction of antifouling (repel) and antimicrobial (kill) surfaces and proposed dual-functional systems that either focus on killing incoming microbes while possessing an underlying repelling mechanism (kill-first), or repelling microbes while safeguarding with an underlying killing mechanism (repel-first).

3.2.6. Approach

Techniques used to modify polymeric surfaces for functionalization in the past have involved harsh acidic or basic conditions and methods that are limited in scope, difficult to reproduce, or challenging to implement in a continuous process.160-162 The advent of “click” chemistry has provided a toolbox of orthogonal functionalities that enables the production of multi-functional surfaces post-fabrication.163-168 Click reactions are categorized as rapid, high yielding reactions that can be performed using mild reagents and conditions (aqueous solutions, room temperature, or physiological conditions) with minimal by-products or purification. Examples of such reactions including Cu(I)-catalyzed or strain-promoted azide-alkyne cycloadditions, thiol-ene and thiol-yne reactions, oxime ligations, and Diels-Alder reactions have been reviewed extensively.168-170 In short, click chemistry provides a handle for covalent surface

35 modification in an efficient and reliable manner. The development of platform technologies which incorporate conjugate functional groups into polymers that react selectively and rapidly with corresponding bioactive motifs provides the opportunity to create translationally relevant, dual-functional materials. With strategic synthetic planning, many varieties and combinations of antifouling and antimicrobial moieties can be conjugated to polymer surfaces using click chemistry, which should enable researchers to move on from testing of ideal substrates (e.g. silica) and into more relevant materials and conditions for implantation.

In this work, a series of quaternary ammonium thiol compounds (Qx-SH) possessing various hydrocarbon tail lengths (8 – 14 carbons) will be synthesized and attached to the surface using thiol-ene “click” chemistry. A quantitative assessment will be performed to elucidate the amount of Qx-SH available on the surface using fluorescence spectroscopy and X-ray photoelectron spectroscopy (XPS). Contact-killing assays will be performed to determine the optimal Qx-SH composition, and the contact- killing efficiency of functionalized surfaces will be explored. Extruded catheter tubing will be modified post-fabrication with the optimal Qx-SH compound and tested for biofilm formation resistance to P. aeruginosa.

3.3. Experimental

3.3.1. Materials

All commercial reagents and solvents were used as received without further purification. The chloroform-d (CDCl3) and dimethyl sulfoxide-d6 (DMSO-d6) were purchased from Cambridge Isotopes Laboratories, Inc (Tewksbury, MA). Diethyl ether

i (Et2O) and isopropyl alcohol ( PrOH) were purchased from EMD Millipore (Burlington,

MA). Calcium chloride (CaCl2), tryptic soy broth, and Mueller-Hinton broth were purchased from VWR (Radnor, PA). Anhydrous methylene chloride (CH2Cl2), ethyl acetate

(EtOAc), dimethylformamide (DMF), 2-butanone, tetrahydrofuran (THF), methanol

(MeOH), N,N-diisopropylethylamine (DIPEA), N,N-dimethyloctylamine (DOA), N,N- dimethyldodecylamine (DDA), N,N-dimethyltetradecylamine (DTDA), 3,3’-dithiopropionic acid, thionyl chloride (SOCl2), tris(2-carboxyethyl)phosphine hydrochloride (TCEP), 2,4,6- trimethylbenzoyl chloride, dimethyl phenylphosphonite, rhodamine B, trimethylaluminum solution (2.0 M in toluene), piperazine, 3-bromo-1-propanol, sodium chloride (NaCl), sodium hydroxide (NaOH), sodium sulfate (Na2SO4), sodium bicarbonate

(NaHCO3), lithium bromide (LiBr), ammonium sulfate ((NH4)2SO4), magnesium chloride

(MgCl2), ethylenediaminetetraacetic acid iron(III) sodium salt (Fe-EDTA), 3-allyloxy-1,2- propanediol, tin(II) 2-ethylhexanoate (stannous octoate), 4,4’-methylenebis(cyclohexyl isocyanate) mixture of isomers (HMDI), and 1,4-butanediol (BDO) were all purchased from Sigma-Aldrich (St. Louis, MO). The 8-chloro-1-octanol was purchased from Alfa

Aesar (Haverhill, MA), and dimethyl sulfoxide (DMSO), sodium phosphate dibasic

(Na2HPO4), potassium phosphate monobasic (KH2PO4), sodium citrate, and casamino acids were purchased from Fisher Scientific (Hampton, NH). Arcol-E351 polyol (2,800 푀̅w,

38.5 – 41.5 mg KOH∙g-1) was kindly donated by Covestro.

3.3.2. Methods

3.3.2.1. Quaternary Ammonium Alcohol “Qx-OH” Synthesis

To obtain a series of Qx-OH compounds with various alkyl chain lengths (x = 8, 12,

14), 8-chloro-1-octanol and several tri-substituted amines were reacted neat. The general procedure is exemplified by the following: for Q14-OH, 28.4 mL (93.4 mmol, 1.05 eq) of

DTDA was added to a 100 mL round bottom flask and stirred under N2 purge. Then, 15.0 mL (88.9 mmol, 1.00 eq) of 8-chloro-1-propanol was injected dropwise via syringe and the temperature was gradually increased to 100 °C. The reaction was allowed to stir overnight, which afforded a viscous yellow solution. After precipitation (3×) in Et2O, a pure white solid was obtained (23.3 g, 64.6% yield). Reagent quantities for the syntheses of the various Qx-OH compounds are recorded in Appendix B (Table 6.1).

1 3 Q14-OH. H-NMR (300 MHz, 303K, DMSO-d6): δ = 0.86 (t, JH-H = 6.7 Hz, 3H), 1.25

3 (m, 30H), 1.41 (m, 2H), 1.63 (m, 4H), 2.99 (s, 6H), 3.16 – 3.28 (m, 4H), 3.38 (q, JH-H = 6.3

3 Hz, 2H), 4.35 (t, JH-H = 5.1 Hz, 1H) ppm.

1 3 Q12-OH. H-NMR (300 MHz, 303K, DMSO-d6): δ = 0.86 (t, JH-H = 6.7 Hz, 3H), 1.27

3 (m, 26H), 1.40 (m, 2H), 1.63 (m, 4H), 2.99 (s, 6H), 3.28 – 3.19 (m, 4H), 3.38 (q, JH-H = 6.4

3 Hz, 2H), 4.37 (t, JH-H = 5.1 Hz, 1H) ppm.

1 3 Q8-OH. H-NMR (300 MHz, 303K, DMSO-d6): δ = 0.86 (t, JH-H = 6.6 Hz, 3H), 1.28

3 (m, 18H), 1.46 – 1.37 (m, 2H), 1.63 (m, 4H), 3.01 (s, 6H), 3.32 – 3.20 (m, 4H), 3.37 (q, JH-H

3 = 6.4 Hz, 2H), 4.44 (t, JH-H = 5.1 Hz, 1H) ppm.

3.3.2.2. 3,3’-Dithiodipropanoyl Chloride Synthesis

In an oven dried, two-neck 250 mL round bottom flask fixed with a condenser, alkaline scrubber between the condenser and nitrogen line, and an addition funnel, 34.6 g (164.6 mmol, 1.00 eq) of 3,3’-dithiodipropionic acid was added. With stirring under N2 at 23 °C, 60.0 mL (827.1 mmol, 5.00 eq) of SOCl2 was added dropwise via addition funnel over a period of 30 min. The suspension was gradually brought to reflux, and allowed to stir for 16 h or until the solution turned clear yellow. Subsequently, the excess thionyl chloride and gaseous by-products were removed by vacuum transfer, while maintaining anhydrous conditions. The resulting yellow oil product, quantitative conversion by 1H-

NMR and 13C-NMR, was used directly for further synthetic steps.

1 3 3,3’-dithiopropanoyl chloride. H-NMR (300 MHz, 303 K, CDCl3): δ = 2.95 (t, JH-H =

3 13 7.0 Hz, 4H), 3.32 (t, JH-H = 7.0 Hz, 4H) ppm. C-NMR (300 MHz, 303 K, CDCl3): δ = 32.00,

46.05, 172.09 ppm.

3.3.2.3. Disulfide-QAC “Qx-S-S” Synthesis

The Qx-S-S reagents were produced using freshly prepared 3,3’-dithiodipropanoyl chloride and the desired Qx-OH. Anhydrous techniques were utilized to preserve the acid chloride functionality. The procedure for Q14-S-S is provided as an example: 9.44 g (23.2 mmol, 2.01 eq) of Q14-OH was dissolved in ca. 50-75 mL of anhydrous CH2Cl2 in a 250 mL

2-neck flask fixed with a reflux condenser, and an alkaline scrubber between the condenser and nitrogen line. Then, 2.00 mL (11.6 mmol, 1.00 eq) of 3,3’-dithiopropanoyl chloride was injected dropwise at room temperature, and the reaction was heated at reflux for at least 16 h. The conversion was quantitative by 1H-NMR, and the reaction solution was neutralized by evaporating the solvent, and re-dissolving in sat. NaHCO3. The water was removed via rotary evaporation, and the product was extracted from the salts by re-dissolving in CH2Cl2 and filtering. The filtrate was then stirred over Na2SO4, filtered, and dried under vacuum (8.51 g, 74.3% isolated yield). Reagent quantities and yields for the synthesis of the Qx-S-S series are recorded in Appendix B (Table 6.2).

1 3 Q14-S-S. H-NMR (300 MHz, 303 K, DMSO-d6): δ = 0.84 (t, JH-H = 6.6 Hz, 6H), 1.13

3 3 – 1.44 (m, 60H), 1.51 – 1.72 (m, 12H), 2.69 (t, JH-H = 6.8 Hz, 4H), 2.90 (t, JH-H = 6.8 Hz, 4H),

3 13 3.03 (s, 12H), 3.22 – 3.32 (m, 8H), 4.03 (t, JH-H = 6.6 Hz, 4H) ppm. C-NMR (300 MHz, 303

K, DMSO-d6): δ = 14.36, 22.15, 22.57, 25.70, 26.20, 26.28, 28.51, 28.86, 28.99, 29.21,

29.33, 29.45, 29.52, 29.56, 31.78, 33.18, 33.86, 50.27, 63.04, 64.52, 171.53 ppm.

1 3 Q12-S-S. H-NMR (300 MHz, 303 K, DMSO-d6): δ = 0.85 (t, JH-H = 6.7 Hz, 6H), 1.13

3 3 – 1.45 (m, 52H), 1.49 – 1.73 (m, 12H), 2.70 (t, JH-H = 6.8 Hz, 4H), 2.91 (t, JH-H = 6.8 Hz, 4H),

3 13 3.01 (s, 12H), 3.21 – 3.30 (m, 8H), 4.03 (t, JH-H = 6.6 Hz, 4H) ppm. C-NMR (300 MHz, 303

K, DMSO-d6): δ = 14.39, 22.13, 22.56, 25.69, 26.18, 26.25, 28.50, 28.85, 28.95, 29.18,

29.28, 29.41, 29.48, 31.76, 33.18, 33.85, 50.29, 63.11, 64.53, 171.57 ppm.

1 3 Q8-S-S. H-NMR (300 MHz, 303K, DMSO-d6): δ = 0.86 (t, JH-H = 6.7 Hz, 6H), 1.13 –

3 3 1.43 (m, 36H), 1.52 – 1.70 (m, 12H), 2.69 (t, JH-H = 6.8 Hz, 4H), 2.91 (t, JH-H = 6.8 Hz, 4H),

3 13 3.02 (s, 12H), 3.22 – 3.31 (m, 8H), 4.03 (t, JH-H = 6.6 Hz, 4H) ppm. C-NMR (300 MHz, 303

K, DMSO-d6): δ = 14.39, 22.15, 22.51, 25.68, 26.18, 26.26, 28.50, 28.85, 28.92, 28.94,

31.62, 33.19, 33.85, 50.28, 55.45, 63.13, 64.54, 171.58 ppm.

3.3.2.4. Thiol-QAC “Qx-SH” Synthesis

The Qx-SH reagents were obtained via reduction of Qx-S-S using TCEP. The general procedure is exemplified by the following: 1.50 g of Q14-S-S (1.52 mmol, 1.00 eq) was added to a 150 mL round bottom flask and kept under a flow of Ar. Separately, 0.88 g of TCEP (3.07 mmol, 2.00 eq) was dissolved in 50 mL of Ar purged DI water and the pH was adjusted to ca. 6 using 1 M NaOH. The TCEP solution was added directly to the flask containing Q14-S-S, and allowed to stir for 4 h at 23 °C. The reaction was then saturated with NaHCO3 and stirred for an additional 30 min, then lyophilized for 24 h to remove water. The product was extracted out from the salts by dissolving in CH2Cl2 and filtering.

The filtrate was stirred over Na2SO4, filtered, vacuum dried, and 1.22 g (81.3% yield) was recovered as a yellow semi-solid. Reagent quantities and yields for the synthesis of the

Qx-SH series are recorded in Appendix B (Table 6.3).

1 3 Q14-SH. H-NMR (300 MHz, 303 K, DMSO-d6): δ = 0.85 (t, JH-H = 6.5 Hz, 3H), 1.17

– 1.37 (m, 30H), 1.52 – 1.69 (m, 6H), 2.62 (m, 4H), 3.02 (s, 6H), 3.20 – 3.32 (m, 4H), 4.03

3 13 (t, JH-H = 6.5 Hz, 2H) ppm. C-NMR (300 MHz, 303 K, DMSO-d6): δ = 14.39, 19.74, 22.11,

22.55, 25.69, 26.16, 26.23, 28.51, 28.83, 28.94, 29.18, 29.27, 29.40, 29.48, 29.52, 31.75,

38.39, 50.33, 55.41, 63.17, 64.36, 171.72 ppm. ESI-MS, m/z theoretical: [M]+ = 458.40 Da, observed: [M]+ = 458.5 Da.

1 3 Q12-SH. H-NMR (300 MHz, 303 K, DMSO-d6): δ = 0.85 (t, JH-H = 6.5 Hz, 3H), 1.17

– 1.35 (m, 26H), 1.50 – 1.70 (m, 6H), 2.62 (m, 4H), 3.02 (s, 6H), 3.20 – 3.31 (m, 4H), 4.03

3 13 (t, JH-H = 6.5 Hz, 2H) ppm. C-NMR (300 MHz, 303 K, DMSO-d6): δ = 14.40, 19.74, 22.12,

22.55, 25.69, 26.16, 26.23, 28.51, 28.83, 28.93, 29.17, 29.27, 29.39, 29.47, 31.75, 38.39,

50.32, 55.41, 63.18, 64.36, 171.73 ppm. ESI-MS, m/z theoretical: [M]+ = 430.37 Da, observed: [M]+ = 430.4 Da.

1 3 Q8-SH. H-NMR (300 MHz, 303K, DMSO-d6): δ = 0.86 (t, JH-H = 6.5 Hz, 3H), 1.14 –

1.42 (m, 18H), 1.50 – 1.70 (m, 6H), 2.62 (m, 4H), 3.02 (s, 6H), 3.22 – 3.33 (m, 4H), 4.03 (t,

3 13 JH-H = 6.5 Hz, 2H) ppm. C-NMR (300 MHz, 303 K, DMSO-d6): δ = 14.39, 19.75, 22.14,

22.51, 25.69, 26.16, 26.25, 28.51, 28.82, 28.91, 28.93, 31.62, 38.40, 50.30, 55.43, 63.18,

64.36, 171.74 ppm. ESI-MS, m/z theoretical: [M]+ = 374.31 Da, observed: [M]+ = 374.4 Da.

3.3.2.5. Rhodamine B Thiol “Rhodamine-SH” Synthesis

Rhodamine-SH was synthesized by esterification of rhodamine B 4-(3- hydroxylpropyl) piperazine amide with 3,3’-dithiopropanoyl chloride, followed by TCEP reduction to provide the corresponding thiol. The rhodamine B 4-(3-hydroxylpropyl) piperazine amide was obtained using a multistep procedure adapted from Nguyen and

Francis.171

Rhodamine B base: 15 g of rhodamine B (31.3 mmol) was dissolved in 1 M NaOH and extracted with multiple portions of EtOAc. The combined organic layers were washed with 1 M NaOH (3×) and brine (3×), then stirred over Na2SO4 and dried under vacuum.

The product was isolated as a pink foam (12.5 g, 90.2% yield).

1 3 Rhodamine B base. H-NMR (300 MHz, 303K, CDCl3): δ = 1.18 (t, JH-H = 7.0 Hz,

3 3 3 12H), 3.37 (q, JH-H = 7.0 Hz, 8H), 6.35 (dd, JH-H = 8.9, 2.5 Hz, 2H), 6.46 (d, JH-H = 2.4 Hz,

3 3 3 2H), 6.59 (d, JH-H = 8.9 Hz, 2H), 7.21 (d, JH-H = 7.5 Hz, 1H), 7.53 – 7.68 (m, 2H), 8.01 (d, JH-

H = 6.9 Hz, 1H) ppm.

Rhodamine B piperazine amide: 11.1 g (25.1 mmol, 1.00 eq) of rhodamine B base was dissolved in an oven dried schlenk flask with 20 mL of anhydrous CH2Cl2. In a separate oven dried 2-neck, 250 mL flask fixed with a condenser, 8.63 g of piperazine (100.2 mmol,

4.00 eq) was dissolved in 35 mL of anhydrous CH2Cl2 under N2. Using air-free techniques,

25 mL of a 2.0 M solution of trimethylaluminum in toluene (50.0 mmol, 2.00 eq) was added dropwise to the piperazine solution. Gas evolution occurred, and after one hour of stirring a white precipitate formed in the flask. The rhodamine B base solution was

43 added dropwise to the heterogenous mixture, and the reaction was gradually heated to reflux and stirred for 24 h. To terminate the reaction, 0.1 M HCl was added dropwise

(slowly) until gas evolution was no longer observed. The solution was filtered and rinsed with CH2Cl2, and the solvent was removed. The crude product was dissolved in dilute

NaHCO3, and washed with multiple portions of EtOAc to remove excess starting material.

The aqueous layer was saturated with NaCl, acidified with 1 M HCl, and extracted (3×)

i with 2:1 PrOH/CH2Cl2. The combined organic layers were stirred over Na2SO4, filtered and dried under vacuum. The resulting purple solid was dissolved in a minimal amount of MeOH and precipitated into Et2O, centrifuged at 5000 RPM for 2 min and decanted, then re-dissolved in CH2Cl2 and vacuum dried. A dark purple pearlescent solid was obtained (7.8 g, 56.8% yield).

1 Rhodamine B piperazine amide. H-NMR (300 MHz, 303K, DMSO-d6): δ = 1.22 (t,

3 3 JH-H = 6.9 Hz, 12H), 2.94 (m, 4H), 3.39 – 3.81 (m, overlaps with HDO, 12H), 6.95 (d, JH-H =

3 3 2.0 Hz, 2H), 7.09 (dd, JH-H = 9.6, 2.0 Hz, 2H), 7.16 (d, JH-H = 9.5 Hz, 2H), 7.49 – 7.59 (m,

1H), 7.70 – 7.85 (m, 3H), 9.71 (s, 1H) ppm. ESI-MS, m/z theoretical: [M]+ = 511.31 Da, observed: [M]+ = 511.3 and [M]2+ = 255.7 Da.

Rhodamine B 4-(3-hydroxylpropyl) piperazine amide: 4.00 g of rhodamine B piperazine amide (7.31 mmol, 1.00 eq) was dissolved in 15 mL of DMF, and 2.00 mL of 3- bromo-1-propanol (22.1 mmol, 3.03 eq) and 4.46 mL of DIPEA (25.6 mmol, 3.50 eq) were added. The reaction was stirred under N2 for 24 h at 23 °C, whereupon an additional 2.00 mL of 3-bromo-1-propanol (22.1 mmol, 3.03 eq) and 4.46 mL of DIPEA (25.6 mmol, 3.50

44 eq) were added and stirred for 24 h. The reaction solution was diluted with sat. NaHCO3 and washed with EtOAc (3×) to remove DIPEA and excess 3-bromo-1-propanol. The

i aqueous layer was then extracted with 1:3 PrOH/CH2Cl2 and the organic layers were combined, stirred over Na2SO4, filtered, and dried under vacuum. A dark purple solid was obtained (3.99 g, 90.2% yield).

Rhodamine B 4-(3-hydroxylpropyl) piperazine amide. 1H-NMR (300 MHz, 303K,

3 3 DMSO-d6): δ = 1.21 (t, JH-H = 6.9 Hz, 12H), 1.47 (m, 2H), 2.07 (br, 4H), 2.19 (t, JH-H = 7.1

3 Hz, 2H), 3.17 – 3.45 (m, overlaps with HDO, 6H), 3.66 (q, JH-H = 7.0 Hz, 8H), 4.39 (s, 1H),

3 3 6.96 (br, 2H), 7.13 (m, 4H), 7.53 (dd, JH-H = 5.9, 2.9 Hz, 1H), 7.64 (m, 1H), 7.74 (dd, JH-H =

5.3, 3.6 Hz 2H) ppm. ESI-MS, m/z theoretical: [M]+ = 569.35 Da, observed: [M]+ = 569.4

Da.

Rhodamine B disulfide: using anhydrous techniques, 3.50 g of rhodamine B 4-(3- hydroxylpropyl) piperazine amide (5.78 mmol, 2.00 eq) was added to a 150 mL, 2-neck flask fixed with a condenser and dissolved in ca. 40 mL of anhydrous CH2Cl2 with stirring under N2. At room temperature, 0.50 mL of 3,3’-dithiopropanoyl chloride (2.90 mmol,

1.00 eq) was added dropwise and the reaction was gradually brought to reflux. After 16 h, the reaction was diluted with sat. NaHCO3, stirred for 30 min, and extracted (3×) with

CH2Cl2. The organic layers were collected, stirred over Na2SO4, filtered, and dried under vacuum. A dark purple solid was obtained (3.95 g, 99.3% yield).

1 3 Rhodamine B disulfide. H-NMR (300 MHz, 303K, DMSO-d6): δ = 1.20 (t, JH-H = 6.6

3 3 Hz, 24H), 1.65 (m, 4H), 2.07 (br, 8H), 2.20 (t, JH-H = 7.1 Hz, 4H), 2.68 (t, JH-H = 6.6 Hz, 4H),

3 3 2.89 (t, JH-H = 6.5 Hz, 4H), 3.11 – 3.48 (m, overlaps with HDO, 8H), 3.65 (q, JH-H = 7.0 Hz,

3 3 16H), 4.01 (t, JH-H = 5.9 Hz, 4H), 6.95 (br, 4H), 7.13 (m, 8H), 7.53 (dd, JH-H = 5.2, 3.3 Hz,

3 13 2H), 7.64 (m, 2H), 7.73 (dd, JH-H = 5.2, 3.3 Hz 4H) ppm. C-NMR (300 MHz, 303 K, DMSO- d6): δ = 12.94, 25.67, 33.07, 33.79, 45.84, 54.49, 55.40, 62.98, 96.31, 113.44, 114.75,

127.85, 130.04, 130.25, 130.73, 130.97, 132.35, 135.94, 155.49, 156.03, 157.46, 166.57,

171.60 ppm. ESI-MS, m/z theoretical: [M]2+ = 656.34 Da, observed: [M]2+ = 656.4 and

[M+H]3+ = 438.0 Da.

Rhodamine-SH: 346 mg of rhodamine B disulfide (0.25 mmol, 1.00 eq) was dissolved in 2 mL of DMF in a round bottom flask and kept under a flow of Ar. Separately,

150 mg of TCEP (0.52 mmol, 2.10 eq) was dissolved in 10 mL of Ar purged DI water and the pH was adjusted to ca. 6 using 1 M NaOH. The TCEP solution was added to the flask containing rhodamine B disulfide, and allowed to stir for 4 h at 23 °C. The reaction was then saturated with NaHCO3 and stirred for an additional 30 min. The resulting solution was extracted with CH2Cl2 (3×) and the organic layers were combined and washed with equal portions of sat. NaHCO3 (2×) and sat. NaCl solution (3×), then stirred over Na2SO4, filtered, and dried under vacuum. A dark purple solid with notable odor was recovered

(290 mg, 83.7% yield).

1 3 Rhodamine-SH. H-NMR (300 MHz, 303K, DMSO-d6): δ = 1.20 (t, JH-H = 6.6 Hz,

12H), 1.67 (m, 2H), 2.11 (br, 4H), 2.23 (br, 2H), 2.61 (m, 4H), 3.15 – 3.47 (m, overlaps with

3 3 HDO, 4H), 3.66 (q, JH-H = 7.0 Hz, 8H), 4.02 (t, JH-H = 6.1 Hz, 2H), 6.95 (br, 2H), 7.13 (m,

3 3 4H), 7.53 (dd, JH-H = 5.7, 2.9 Hz, 1H), 7.64 (m, 1H), 7.74 (dd, JH-H = 5.1, 3.5 Hz 2H) ppm.

13 C-NMR (300 MHz, 303 K, DMSO-d6): δ = 12.93, 19.72, 25.65, 38.36, 45.85, 54.47, 54.49,

62.78, 96.33, 113.45, 114.75, 127.85, 130.05, 130.26, 130.71, 130.95, 132.33, 135.96,

155.52, 156.01, 157.47, 166.60, 171.72 ppm. ESI-MS, m/z theoretical: [M]+ = 657.35 Da,

+ observed: [M] = 657.4 Da. UV-vis (DMSO), λabs = 568 nm, λem = 592 nm.

3.3.2.6. Lithium Phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) Photoinitiator

Synthesis

LAP was synthesized as previously reported.172 Briefly, 2.80 mL of dimethyl phenylphosphonite (17.6 mmol, 1.00 eq) was added to an oven dried flask under Ar at 23

°C. While stirring, 2.94 mL of 2,4,6-trimethylbenzoyl chloride (17.6 mmol, 1.00 eq) was added dropwise and allowed to react for 18 h. Then, a four-fold excess of LiBr (6.1 g) in

100 mL of 2-butanone was added to the reaction mixture and heated to 50 °C for 10-15 min with stirring until a white precipitate formed. The solution was cooled to room temperature and set for 2 h, then suction filtered and rinsed generously with 2-butanone to remove excess LiBr. The solid white precipitate (4.45 g, 85.9% yield) was dried under vacuum and analyzed by 1H-NMR and UV-vis spectroscopy. The molar absorptivity (ε) was determined and compared to the literature.

1 LAP. H-NMR (300 MHz, 303K, D2O): δ = 2.01 (s, 6H), 2.23 (s, 3H), 6.88 (s, 2H),

7.41 – 7.51 (m, 2H), 7.51 – 7.61 (m, 1H), 7.70 (m, 2H) ppm. UV-vis (H2O), λabs = 372 nm, ε

= 179 ± 3 M-1cm-1 (lit. value = 218 M-1cm-1).

3.3.2.7. Control and Allyl-TPU Polymerizations

TPU polymerizations were performed in bulk with mechanical stirring at 100 °C.

The following is provided as an example procedure: for a 100 g batch of 50 mol% (30 wt.%)

HMDI control TPU, 61.7 g (22.0 mmol, 1.0 eq) of Arcol E-351 and 8.2 g (92.3 mmol, 4.2 eq) of BDO were preheated to 100 °C in a porcelain enamel-lined tin can with overhead mechanical stirring. Then, 28.1 mL (114.3 mmol, 5.2 eq) of HMDI was added, immediately followed by 2-3 drops of stannous octoate catalyst. The mixture was stirred for 2 – 5 min, or until the mixture became too viscous to stir. The resulting TPU was oven cured at 100

°C for 24 h. To produce a TPU containing 8 mol% (2.4 wt.%) 3-allyloxy-1,2-propanediol, denoted as “allyl-TPU”, the molar ratio of BDO to 3-allyloxy-1,2-propanediol was modified, while maintaining the molar ratio of HMDI to Arcol-E351. Additionally, 3- allyloxy-1,2-propanediol was added directly with the mixture of diols and preheated to

100 °C before adding HMDI. Reagent quantities used for TPU polymerizations are recorded in Appendix B (Table 6.4).

1 Control TPU. H-NMR (300 MHz, 303 K, DMSO-d6): δ = 0.97 (m, 10H), 1.02 – 1.20

(m, 120H), 1.26 (m, 10H), 1.37 – 1.81 (m, 100H), 1.99 (m, 10H), 3.16 – 3.89 (m, 135H),

4.07 (m, 16H), 4.20 (m, 5H), 4.51 (m, 6H), 4.65 (m, 4H), 4.79 (m, 6H), 4.91 (m, 4H) ppm.

SEC (THF): 푀̅푛 = 68 kDa, 푀̅푤 = 175 kDa, and Đm = 2.6. DSC: Tm = 72 °C and 119 °C, Tg = -

60.5 °C. TGA: Td = 255 °C

1 Allyl-TPU. H-NMR (300 MHz, 303 K, DMSO-d6): δ = 0.96 (m, 25H), 1.07 – 1.16 (m,

145H), 1.21 (m, 12.5H), 1.24 – 1.78 (m, 86H), 1.85 (m, 16H), 1.99 (m, 12.5H), 3.23 – 3.83

(m, 194H), 4.05 (m, 16H), 4.20 (m, 5H), 4.54 (m, 2.5H), 4.67 (m, 2H), 4.82 (m, 3.5H), 4.93

3 (m, 2H), 5.22 (dd, JH-H = 24.2, 13.8 Hz, 2H), 5.85 (m, 1H) ppm. SEC (THF): 푀̅푛 = 92 kDa, 푀̅푤

= 269 kDa, Đm = 2.9. DSC: Tm = 72 °C and 115 °C, Tg = -62.5 °C. TGA: Td = 245 °C.

3.3.2.8. Characterization

NMR spectra were obtained using a Varian Mercury 300 MHz NMR spectrometer operated at 303 K. All chemical shifts are reported in ppm (δ) and referenced to the

1 chemical shifts of residual solvent resonances ( H-NMR: CDCl3 = 7.26 ppm, D2O = 4.79

13 ppm, DMSO-d6 = 2.50; C-NMR: CDCl3 = 77.16 ppm, DMSO-d6 = 39.52). Mass spectrometry was performed using a HCT Ultra II quadrupole ion trap mass spectrometer

(Bruker Daltonics, Billerica, MA) equipped with electrospray ionization (ESI) source.

Samples were dissolved in MeOH and diluted to 0.01 µg∙mL-1 prior to injection. The sample solutions were injected into the ESI source by direct infusion, using a syringe pump, at a flow rate of 3 µL∙min-1. The tip of the ESI needle was grounded, and the entrance of the capillary, through which ions enter in the vacuum system of the mass

49 spectrometer, was held at 3.5 kV. The pressure of the nebulizing gas (N2) was set at 10

-1 psi, and the flow rate and temperature of the drying gas (N2) was 8 L∙min and 300 °C, respectively. Data collection was performed on positive mode and the ESI-MS data was analyzed by Bruker Daltonik’s DataAnalysis v4.0 software. Differential scanning calorimetry (DSC) was performed using a TA Instruments Q2000 DSC (TA Instruments –

Waters L.L.C., New Castle, DE) on sample sizes between 5 – 10 mg using temperature ramps for heating of 20 °C∙min-1 and a cooling rate of 20 °C∙min-1. Thermogravimetric analysis was performed using a TA Instruments TGA 2950 (TA Instruments – Waters L.L.C,

New Castle, DE) on sample sizes of ca. 10 mg using a heating ramp of 10 °C∙min-1, after holding temperature for 5 min at 110 °C to remove water. Durometer measurements were performed on compression molded, cylindrical samples (stacked thickness >6.5 mm). A Folwer® Shore A Portable Durometer (Folwer High Precision, Auburndale, MA) was used for durometer hardness testing, following the procedure described in ASTM

D2240. The instrument was calibrated using a standardized Shore A 50 material prior to each measurement. Size exclusion chromatography (SEC) was performed using an EcoSEC

HLC-8320GPC (Tosoh Bioscience LLC, King of Prussia, PA) equipped with a TSKgel GMHHR-

M mixed bed columns and refractive index (RI) detector. Molecular weights were calculated using a calibration curve determined from poly(styrene) standards (PStQuick

MP-M standards, Tosoh Bioscience LLC) with THF as eluent flowing at 1.0 mL∙min-1, and a sample concentration of 4 mg∙mL-1. The molecular weight data and other polymer properties are recorded in Appendix B (Table 6.5).

3.3.2.9. Polymer Processing

TPU films were produced by doctor blade coating a 30 wt.% solution of allyl-TPU in THF onto polyethylene terephthalate (PET) using a gap height of 1.0 mm, and line speed of 15 cm∙min-1. The blade-coated films were dried overnight at 23 °C, and placed in a vacuum oven for 48 h at 25 °C. The films were then punched to 2.0 cm in diameter using a manual punch set to produce cylindrical samples ca. 250 μm thick, which were measured using a digital caliper. Compression molded control and allyl-TPU samples were produced using a TMP 35-ton vacuum molding press by heating to 120 °C and pressing at

140 MPa for 15 min, followed by water cooling to room temperature. The molds were then punched to 2 cm in diameter using a manual punch set to produce cylindrical samples (1.0 mm thick, 2.0 cm diameter). Catheter tubes of allyl-TPU were extruded by

Cook Polymer Technology using a custom designed, single screw extruder (general purpose 19.05 mm screw) with a 2.18 mm die and a 1.40 mm mandrel. The screw speed was 20 rpm and the line speed was 9.75 m∙min-1. Heating zones were ramped from 165

°C to 177 °C, and the head pressure was 2.76 MPa (8.62 MPa behind the screens). The material was dried at 82 °C overnight.

3.3.2.10. Allyl-TPU Surface Functionalization

For proof of concept and a quantitative estimate of the amount of Qx-SH that attaches to the surface via thiol-ene chemistry, a rhodamine-SH dye containing the same synthetic core as the Qx-SH compounds was reacted with allyl-TPU using thiol-ene “click” conditions. In a 12-well plate, blade-coated films of allyl-TPU (250 µm thick, 2.0 cm

51 diameter, ca. 110 mg, 20 μmol allyl) were submerged in 2 mL of an Ar purged (30 min) aqueous solution containing rhodamine-SH (10.0 mM) and LAP (5.0 mM), and allowed to pre-soak for 30 min under Ar. The samples were then treated with UV light (λ = 365 nm,

I = 1.2 mW∙cm-2) or kept in the absence of UV light for 30 min to control for physical adsorption of the dye (denoted as “phys. ads.” samples). Following treatment, the dye/photoinitiator solutions were drawn up and discarded, and the samples were rinsed with 5 mL of DI water (3×) then submerged in 5 mL of DI water. The samples were further rinsed with 10% EtOH (3×) and soaked in 10% EtOH for 15 min (3×), rinsing with 10% EtOH in between each soak. The samples were blown dry with N2 and dissolved in DMSO for fluorescence studies.

To functionalize the allyl-TPU blade-coated samples with Qx-SH reagents, thiol- ene reactions were performed using the same procedure described for rhodamine-SH.

Briefly, blade-coated films of allyl-TPU (250 µm thick, 2.0 cm diameter, ca. 110 mg, 20

μmol allyl) were submerged in 2 mL of an Ar purged (30 min) aqueous solution containing

Qx-SH (10.0 mM) and LAP (5.0 mM), and allowed to pre-soak for 30 min under Ar. UV treated samples were irradiated for 30 min (λ = 365 nm, I = 1.2 mW∙cm-2) while phys. ads. samples were kept in the absence of UV light for 30 min. Following treatment, the dye/photoinitiator solutions were drawn up and discarded, and the samples were rinsed with 5 mL of DI water (3×) then submerged in 5 mL of DI water. The samples were further rinsed with 10% EtOH (3×) and soaked in 10% EtOH for 15 min (3×), rinsing with 10% EtOH in between each soak.

For post-fabrication functionalization of the inner lumen of allyl-TPU catheter tubing with Qx-SH reagents, 0.5 mL of an Ar purged solution containing Qx-SH (10.0 mM) and LAP (5.0 mM) was flowed through 25.0 cm segments of catheter tubing every 7.5 min for the duration of a 30 min pre-soak and 30 min UV treatment (or absence of UV for phys. ads. control). Following treatment, the catheter tubes were rinsed continuously for

30 s with DI water (3x), then rinsed continuously for 30 s with 10% EtOH (3x). All samples were blown dry with N2 and placed under vacuum for 24 h before X-ray photoelectron spectroscopy (XPS) analysis or ethylene oxide (EtO) sterilization.

3.3.2.11. Surface Quantification and Analysis

For surface quantification, fluorescence studies were carried out using a BioTek

SynergyTM Mx Microplate Reader (BioTek, Vermont) with Gen 5TM reader control and data analysis software. An excitation wavelength of λex = 568 nm was used while scanning the emission range of 586 – 700 nm at a step of 1 nm∙s-1. A standard curve was constructed by serially diluting a stock solution of rhodamine-SH in DMSO in triplicate, pipetting 300

µL of each solution into a quartz 96-well plate, measuring the emission intensity, and plotting the maximum emission (λem = 592 nm) intensity vs. concentration. To quantify the dye present on UV treated, phys. ads., and untreated allyl-TPU blade-coated samples, the films were dissolved in DMSO (5 mL) and diluted as necessary to achieve fluorescence intensities within the standard curve. Following dilution, 300 µL of each solution was pipetted into a quartz 96-well plate and the intensity at λem for each sample was measured using the plate reader. The dye concentration was determined from the λem intensity

53 using the standard curve, and the molar quantity of dye present on each sample (n =3) was calculated (accounting for dilutions) and reported in terms of mol∙cm-2 based on the surface area of the samples (Appendix B, Table 6.6).

XPS was employed to characterize the surface composition of the allyl-TPU films and the inner lumen of catheter tubing (longitudinal sections) treated with Qx-SH reagents. The XPS spectra were obtained using a VersaProbe II Scanning XPS Microprobe from Physical Electronics (PHI), under ultrahigh vacuum conditions with a pressure of 2.0

µPa. Automated dual beam charge neutralization was used during the analysis of the samples to provide accurate data. The analyzer pass energy was 117.4 eV for the survey spectra and 23.5 eV for the high-resolution scans in the N1s regions. Survey scans in the range of 0 – 700 eV were used to evaluate the percentage of different atoms present on the surface of the samples. Atomic concentrations were calculated with PHI MultiPak software. The XPS high-resolution spectra of N1s were decomposed into two components by using the curve fitting routine in MultiPak. A goodness of fit (χ2) better than 1.5 was achieved for each fit. Each spectrum was collected using a monochromatic

(Al Kα) x-ray beam (E = 1486.6 eV) over a 100 μm × 1400 μm probing area with a beam

+ power of 100 W. The quaternary nitrogen peak (NR4 , eV = 401 – 402) was integrated and its quantity relative to urethane nitrogen (N, eV = 398.5) was reported (Appendix B, Table

6.6).

3.3.2.12. Antimicrobial Testing

For antimicrobial testing, blade-coated samples and catheter tubing of allyl-TPU were sterilized by EtO sterilization using an Anprolene benchtop sterilizer (Anderson

Products, Inc., Haw River, NC) following the manufacturer’s protocol to deliver approximately 0.5 cc∙L-1 of EtO gas over a 12 h sterilization cycle at 35% humidity and room temperature, followed by a 48 h purge under vacuum. The bacterial strains used in this study included Staphylococcus epidermidis (ATCC 12228), Staphylococcus aureus

(25923), Escherichia coli (ATCC 25922), Pseudomonas aeruginosa (ATCC 27853),

Enterococcus faecalis (ATCC 29212), methicillin-resistant Staphylococcus aureus (MRSA)

(ATCC BAA-41).

The contact-killing assay was performed on physically adsorbed (phys. ads.), and

UV-treated blade-coated samples modified with a series of Qx-SH reagents using a method adapted from ISO 22196.173 Briefly, the samples were inoculated with same day cultures of select bacteria (e.g. E. coli, S. epidermidis) in dilute nutrient media (0.2% tryptic soy broth (TSB) in 1× phosphate buffered saline (PBS)) at approximately 150 colony- forming units (CFU)∙mm-2. Generally, 35 μL of 8.5×105 CFU/mL cultures were dispersed across the surface of the sample using sterile polypropylene cover films (area = 198 mm2). Inoculated samples were incubated in a humidified room air incubator (36 °C, ca.

80% rel. humidity) for 20 h. Following incubation, surviving cells were recovered from the samples using vigorous agitation (i.e. vortexing for 20 s in PBS). Surviving cells were enumerated via 10-fold series dilution and plating 0.1 mL onto tryptic soy agar (TSA),

55 incubating overnight at 37 °C and counting CFU. The mean CFU recovered per sample

(CFU/sample) were calculated by accounting for dilutions. Internal controls samples of polypropylene and chlorhexidine treated with polypropylene were included in the assay.

A live/dead assay was performed using overnight cultures of E. coli grown in TSB and S. aureus grown in Mueller-Hinton broth (MH). The cells were washed with PBS (3×), resuspended in PBS, and diluted to obtain an OD600 = 0.15 (measured on a Hach DR2800

Spectrophotometer, λ = 600 nm). The bacterial suspensions (1.0 mL) were stained with

2.0 μL of a dye mixture containing equal portions of 3.34 mM SYTO 9 and 20 mM propidium iodide (L7012 LIVE/DEAD® BacLightTM Bacterial Viability Kit, ThermoFisher

Scientific) and allowed to incubate at room temperature in the dark for 15 min. Then, 10

μL of stained bacterial suspension was placed on a glass slide and covered with either untreated, phys. ads., or UV-treated samples modified with Q8-SH for 5 min (S. aureus) or 10 min (E. coli) before imaging. Live/dead microscopy was performed using an IX81 inverted microscope (Olympus, Center Valley, PA) and images were processed and quantified using ImageJ software.

Biofilm formation was analyzed using overnight cultures of P. aeruginosa grown in TSB and adjusted to an OD600 = 0.10. An assembly of catheter tubing was prepared using aseptic techniques, and included Cook® Beacon® Tip Torcon NB® Advantage catheter segments (25.0 cm in length) denoted as “CC” and numbered 1 – 4 from upstream (1) to downstream (4). CC segments were located before, after, and in between the untreated, phys. ads., and UV-treated allyl-TPU catheter segments modified with Q8-

SH to account for any downstream influence of the experimental catheters. With a peristaltic pump flowing at 1.5 mL∙min-1, the bacterial inoculum was streamed through the catheter assembly for 2 h, followed by FAB medium (0.10 mM CaCl2, 0.01 mM Fe-

EDTA, 0.15 mM (NH4)2SO4, 0.33 mM Na2HPO4, 0.20 mM KH2PO4, 0.50 mM NaCl, 0.50%

(wt/vol) casamino acids, 1.0 mM MgCl2, and 10 mM sodium citrate) for 48 h. Samples were fixed with 50 mL of a 4% paraformaldehyde solution, rinsed with 200 mL of PBS, and submerged in PBS for further testing.

Photographs of the catheter tubing were taken using a camera with a 16- megapixel Sony Exmor RS IMX240 sensor and f/.19 lens. Catheter cross-sections were cut

(ca. 2.5 mm in length) from each catheter segment, rinsed thoroughly with DI water, and lyophilized for 24 h for scanning electron microscopy (SEM) analysis. SEM was performed on gold sputter-coated samples using a JEOL-7401 Field Emission Scanning Electron

Microscope (JEOL USA, Inc., Peabody, MA) at an accelerating voltage of 2.0 kV under 45× and 300× magnification. In addition, lyophilized cross-sections (3.0 mm in length) from randomly selected locations along each catheter segment were placed vertically on a glass slide and imaged under brightfield microscope at 4× magnification. Using Olympus

VS-Desktop software, the % biofilm blockage was determined by measuring the inner luminal area of untreated catheters compared to the area of the biofilm on the interior of the contaminated catheters (Equation 1). Results are reported as averages with standard deviations (n = 3) for each catheter segment.

푏푖표푓푖푙푚 푎푟푒푎 % 푏푖표푓푖푙푚 푏푙표푐푘푎푔푒 = [ ] ∙ 100 (1) 푖푛푛푒푟 푙푢푚푖푛푎푙 푎푟푒푎

3.4. Results

3.4.1. Characterization of Thiol-ene Reagents

The QAC reagents used for surface functionalization “Qx-SH” were produced by first generating the corresponding Qx-OH compounds (where x = 8, 12, or 14 carbons) via quaternization reactions performed in bulk (Appendix B, Scheme 6.1). A series of quaternary ammonium alcohols with tail lengths (x) were produced by this method, and

1H-NMR confirmed the purity of the compounds (Appendix A, Figures 6.1 – 6.3). The 1H-

NMR spectra revealed a sharp singlet peak (δ = 3.00) corresponding to the methyl groups of the quaternary amine, which was integrated and compared to aliphatic peaks of the 8- carbon spacer and various carbon tail lengths for each Qx-OH compound. To achieve the corresponding disulfides, 3,3’-dithiopropionic acid was converted to a diacid chloride

(Appendix B, Scheme 6.2) and subsequently reacted with the Qx-OH compounds to yield the Qx-S-S series (Appendix B, Scheme 6.3). The conversion of 3,3’-dithiopropionic acid to 3,3’-dithiopropanoyl chloride was quantitative, as confirmed by 1H-NMR and 13C-NMR

(Appendix A, Figure 6.4 and Figure 6.5, respectively), and esterification afforded the Qx-

S-S reagents, which were also characterized by 1H-NMR (Appendix A, Figures 6.6 – 6.8).

Notably, the 1H-NMR spectra for all Qx-S-S compounds demonstrated the appearance of two triplets (δ = 2.70 and 2.90 ppm) corresponding to the methylene protons α and β to the disulfide, which were integrated and compared to the methylene protons (δ = 4.03 ppm) adjacent to the newly formed ester, as well as the aliphatic protons of the hydrocarbon spacer and tail. The Qx-S-S compounds were reduced using TCEP (Appendix

B, Scheme 6.4) to generate the desired Qx-SH reagents, which were also analyzed by 1H-

NMR (Appendix A, Figures 6.9 – 6.11). The 1H-NMR demonstrated the coalescence of the

Qx-S-S triplets (δ = 2.70 and 2.90 ppm) into a multiplet (δ = 2.62 ppm) for the Qx-SH series.

More convincingly, the 13C-NMR spectra overlay of Qx-S-S and Qx-SH demonstrated significant shifting of the carbon α to the carbonyl downfield (from 33.85 to 38.39 ppm) and the β carbon upfield (from 33.18 to 19.74 ppm) (Appendix A, Figures 6.12 – 6.14).

The rhodamine-SH was achieved through esterification of rhodamine B 4-(3- hydroxylpropyl) piperazine amide with 3,3’-dithiopropanoyl chloride, followed by TCEP reduction (Appendix B, Scheme 6.5). 1H-NMR and ESI-MS were used to confirm each synthetic step towards producing rhodamine B 4-(3-hydroxylpropyl) piperazine amide

(Appendix A, Figures 6.15 – 6.19). Following esterification with 3,3’-dithiopropanoyl chloride, 1H-NMR revealed the appearance of the expected triplets (2.68 ppm and 2.89 ppm) and ESI-MS exhibited a doubly charged ion [M]2+ = 656.34 Da, which corresponds to the mass of rhodamine-S-S (1312.68 Da) (Appendix A, Figure 6.20 and Figure 6.21, respectively). TCEP reduction of rhodamine-S-S afforded the desired rhodamine-SH, and

1H-NMR demonstrated the merging of the triplets (δ = 2.68 and 2.89 ppm) into a multiplet

(δ = 2.61 ppm) (Appendix A, Figure 6.22). ESI-MS confirmed the mass of the rhodamine-

SH (molecular ion [M]+ = 657.4 Da) (Appendix A, Figure 6.23). In addition, 13C-NMR reveals the shifting of the carbon α to the carbonyl downfield (from 33.79 to 38.36 ppm) and the

β carbon upfield (from 33.07 to 19.72 ppm) (Appendix A, Figure 6.24). UV-vis and fluorescence spectroscopy of rhodamine-SH in DMSO provided the λabs = 568 nm and the

λem = 592 nm (Appendix A, Figure 6.25), and a standard curve of the fluorescence intensity at λem vs. concentration for rhodamine-SH in DMSO was constructed, yielding a slope of

(168.7 ± 0.1)×109 M-1 with an R2 = 0.99 (Appendix A, Figure 6.26).

The LAP photoinitiator was synthesized using a Michaelis-Arbuzov reaction

(Appendix B, Scheme 6.6) and 1H-NMR confirmed the purity of the compound (Appendix

A, Figure 6.27). The integration of peaks a – c were approximately equimolar to the integrations for peaks d – f, indicating a 1:1 substitution occurred. LAP photoinitiator was used for subsequent thiol-ene reactions due to its water solubility and substantially higher

ε at λ = 365 nm (observed = 179 ± 3 M-1cm-1, lit. value = 218 M-1cm-1) compared to other commercially available water-soluble photoinitiators, such as Irgacure® 2959 (ε = 4 M-

1cm-1).174 The UV-vis absorption spectra for LAP at several concentrations were recorded and a linear plot of the absorbance at λ = 365 nm vs. concentration was constructed to determine the molar absorptivity of LAP (Appendix A, Figure 6.28).

3.4.2. Characterization of Control and Allyl-TPU

A control TPU consisting of an aliphatic diisocyanate (HMDI) and a mixture of diols including BDO and Arcol-E351 was synthesized to mimic a medical grade TecoflexTM TPU with shore A hardness = 90 (Appendix B, Scheme 6.7). 1H-NMR was used to determine the resulting composition by integration of peaks f, g, and h, providing the molar composition of HMDI:Acrol-E351:BDO = 0.5:0.1:0.4. (Appendix A, Figure 6.29). The allyl-

TPU was synthesized in the same manner as the control, except the feed ratio of BDO was reduced to include 3-allyloxy-1,2-propanediol to the mixture of diols. 1H-NMR of the allyl-

3 TPU shows the appearance of a doublet of doublets (δ = 5.22 ppm, JH-H = 24.2, 13.8 Hz) and a multiplet (δ = 5.85 ppm) corresponding to the allylic protons (Appendix A, Figure

6.30). The resulting composition was determined by integration of peaks f, g, h, and n, which provided the molar composition of HMDI:Acrol-E351:BDO:allyl = 0.5:0.1:0.32:0.08.

(Appendix B, Table 6.4). An additional 2.27 kg of allyl-TPU was synthesized in 0.454 kg batches for extrusion of the catheter tubes, and batches were designated numbers 1 through 5. 1H-NMR spectra were integrated for each batch to confirm the allyl content

(Appendix A, Figure 6.31) and SEC was performed to monitor molecular weight consistency between batches (Appendix A, Figure 6.32); the data are recorded in

Appendix B (Table 6.5). In addition, the durometer hardness of the allyl-TPU was the same as the control (shore A durometer = 90), and the thermal properties were examined to gauge the extrusion conditions; TGA demonstrated the onset degradation temperature

(Td) for allyl-TPU was 245 °C while the control TPU was 255 °C (Appendix A, Figure 6.33), and DSC thermograms revealed that the melting temperature (Tm) and glass transition temperature (Tg) were practically unaffected by the introduction of 3-allyloxy-1,2- propanediol into the TPU (Appendix A, Figure 6.34, Appendix B, Table 6.5). The resulting allyl-TPU blade-coated samples and extruded tubes were also optically clear.

3.4.3. Post-Fabrication Surface Functionalization and Quantification

Surface modification of allyl-TPU blade-coated samples and catheter tubing was achieved using thiol-ene “click” chemistry (Scheme 3.1). The thiol-ene reaction provides an efficient and convenient method to modify surfaces containing alkene functionalities,

61 and can be performed in water with the assistance of a photoinitiator and UV light. For a quantitative assessment of the amount of Qx-SH that attaches to the surface via thiol-ene reactions, a rhodamine-SH dye containing the same synthetic core as the Qx-SH compounds was reacted with allyl-TPU blade-coated samples and analyzed by fluorescence spectroscopy (Figure 3.6). Notably, the rhodamine-SH dye shares an identical chemical structure with the Qx-SH reagents up to 7 atoms from the thiol functionality, which should provide a reasonable comparison from a reactivity standpoint.

As shown, the untreated control does not exhibit fluorescence over the scanned emission range, while the UV treated and phys. ads. samples achieved fluorescence intensities corresponding to 5.5 ± 0.1 and 3.6 ± 0.1 nmol∙cm2 of rhodamine-SH per sample surface area, respectively (Appendix B, Table 6.6). The observed increase in dye quantity for UV treated samples compared to phys. ads. controls may be attributed to an additional quantity of covalently attached rhodamine-SH; however, since the disappearance of the allyl functional groups located near the surface could not be resolved from the bulk signal due to instrumental resolution, the proportion of physically adsorbed dye that becomes covalently linked to the surface could not be elucidated. As a quantitative result, the subtraction of the phys. ads. dye quantity from the UV treated dye quantity provides a minimum of attached rhodamine-SH, while the UV treated sample alone provides the potential maximum quantity of attached rhodamine-SH. Therefore, a range regarding the expected quantity of covalently attached thiol compounds was reasoned to be between

1.9 ± 0.1 to 5.5 ± 0.1 nm∙cm2 for allyl-TPU samples.

Scheme 3.1. Post-fabrication, surface functionalization of allyl-TPU with Qx-SH reagents was carried out in DI water at room temperature using LAP photoinitiator and UV light (365 nm, I = 1.2 mW∙cm-2).

Figure 3.6. Fluorescence data for the untreated control, phys. ads., and UV treated allyl- TPU samples modified using “click” reaction conditions with rhodamine-SH. Emission scans were taken from λ = 586 – 700 nm at an excitation wavelength of λex = 568 nm, which provided the intensity at λmax (592 nm) for each sample. Experiments were performed in triplicate and the average fluorescence intensities with standard deviations are plotted (n=3).

Additional characterization of surface-functionalized samples was performed using XPS. High-resolution N1s spectra were obtained for Qx-SH modified allyl-TPU samples and catheter tubing to confirm the presence of QACs on the surface, and to evaluate the proportion of Qx-SH present on UV treated samples relative to phys. ads. controls (Figure 3.7). High-resolution N1s XPS reveals a major peak at 398.4 eV corresponding to nitrogen (N) contained in the urethane bonds throughout the TPU backbone, as well as a minor peak between 401 – 402 eV corresponding to quaternary

+ nitrogen (NR4 ) introduced by the Qx-SH compounds. As shown, the UV treated samples

+ demonstrated a more pronounced NR4 peak compared to the phys. ads. and untreated controls (Figure 3.7A). For the Qx-SH series, XPS was performed in triplicate on independent batches of post-fabrication functionalized allyl-TPU blade-coated samples,

+ and the average % NR4 relative to urethane N was determined (Figure 3.7B, Appendix B,

Table 6.6). The results indicated that the UV treated samples contained significantly higher QAC content than their respective phys. ads. controls for each Qx-SH group.

Furthermore, the average proportion (x)̄ of physically absorbed Qx-SH across all groups was found to be x ̄ < 1/3 of the total Qx-SH present in UV treated allyl-TPU blade-coated samples (Appendix B, Table 6.6). Additional high-resolution N1s spectra of the inner lumen of phys. ads. and UV treated allyl-TPU catheter tubing (longitudinal sections)

+ modified with Q8-SH were obtained. The phys. ads. sample did not exhibit an NR4 peak,

+ while the UV treated tubing contained 14.4% NR4 relative to urethane N (Appendix A,

Figure 6.35).

Figure 3.7. (A) XPS high-resolution N1s spectra overlay of an untreated control, phys. ads., and UV treated sample demonstrating the appearance of a quaternary ammonium peak (400 – 402 eV). The solid lines represent raw data interpolated with a cubic b-spline curve, + while the dashed lines represent the total curve fits for each sample. (B) The % NR4 relative to urethane N is shown for UV treated and phys. ads. samples modified with the Qx-SH series. XPS measurements were taken on three separate batches of allyl-TPU blade-coated samples and the averages with standard deviations are displayed. Statistical significance is indicated by *, **, and *** with p values <0.05 between UV treated and phys. ads. samples for each Qx-SH compound.

Comparison of the XPS and fluorescence data suggests the relative proportion of physically adsorbed Qx-SH compounds is less than that shown for rhodamine-SH (Figure

3.6 and Figure 3.7). This may be contributed to differences in non-covalent interactions between the respective thiols and the allyl-TPU; while the Qx-SH may hydrogen bond via ester and thiol functional groups, the rhodamine-SH possesses tertiary amine, amide, ester and thiol functionalities, as well as potential π-π interactions between the rhodamine core and the allyl groups of the TPU. Overall, combining the quantitative results of the fluorescence assay with the XPS data suggests that the quantity of Qx-SH available on the surface of allyl-TPU samples post-functionalization is likely between 1.9

± 0.1 to 5.5 ± 0.1 nm∙cm2, of which less than 1/3 is physically adsorbed (Appendix B, Table

6.6).

3.4.4. Antimicrobial Testing

3.4.4.1. Contact-Killing Assay

Initial screening for antimicrobial activity was performed on phys. ads. and UV- treated allyl-TPU blade-coated samples modified with the Qx-SH series using a 24 h contact-killing assay adapted from ISO 22196. The results of the assay demonstrated a 6- log reduction in E. coli compared to the negative control for Q8-SH (UV-treated and phys. ads.) and Q12-SH (UV-treated) samples (Table 3.1). Less notable reductions were observed for the Q12-SH (phys. ads.) and Q14-SH (UV-treated and phys. ads.) samples, which demonstrated ca. 3-log reductions or less compared to the negative control. In addition, complete reductions (5-log) in S. epidermidis compared to the negative control

66 were observed for all Qx-SH compositions (UV-treated and phys. ads). Further contact- killing assays with MRSA, E. faecalis, and P. aeruginosa were performed (Appendix B,

Table 6.7), demonstrating similar results (i.e. complete reduction of MRSA and E. faecalis for nearly all samples, and complete reduction of P. aeruginosa for Q8-SH phy. ads., Q8-

SH UV-treated, and Q12-SH UV-treated samples). The contact-killing assay highlighted the antimicrobial efficacy of both phys. ads. and UV-treated samples modified with Q8-

SH, suggesting this composition was the most potent of the Qx-SH series, and prompting further investigation into its antimicrobial properties.

Table 3.1. Contact-killing assay (ISO 22196) results

Mean CFU/Sample Recovered a

Sample E. coli S. epidermidis

Polypropylene b 2.04 (± 0.06)×106 0.93 (± 0.03)×105

Chlorhexidine c 0.00 ± 0.00 0.00 ± 0.00

Q8-SH 0.00 ± 0.00 0.00 ± 0.00

phys. Ads. Q12-SH 1.17 (± 0.41)×103 0.00 ± 0.00

Q14-SH 6.03 (± 0.03)×105 0.00 ± 0.00

Q8-SH 0.00 ± 0.00 0.00 ± 0.00

UV-treated Q12-SH 0.00 ± 0.00 0.00 ± 0.00

Q14-SH 6.61 (± 0.01)×105 0.00 ± 0.00 a Mean CFU/sample data were determined by serial dilution, performed in duplicate (n = 2). b Negative control for assay. c Positive control for assay (chlorhexidine treated polypropylene).

3.4.4.2. Live/Dead Fluorescence Assay

To distinguish the antimicrobial activity of phys. ads. and UV-treated blade-coated samples modified with Q8-SH, and to gain an understanding of the contact-killing efficiency, a live/dead fluorescence assay was performed. The results demonstrate that the majority of the S. aureus and E. coli were killed with 5 – 10 min of exposure (Figure

3.8). Although the Q8-SH phys. ads. samples killed a comparable portion of the S. aureus

(ca. 50%) to the UV-treated samples (ca. 75%) within 5 min, they were unable to match the killing efficiency of the UV-treated samples for E. coli at 10 min; UV-treated samples killed 90% of E. coli compared to 5% for the phys. ads. samples.

Figure 3.8. The live/dead contact-killing assay results for S. aureus after 5 min (top), and E. coli after 10 min (bottom) for control, phys. ads. and UV treated allyl-TPU blade-coated samples modified with Q8-SH. Pie charts are representative of the cell count enumerated using ImageJ software (green indicates live cells, while red indicates cell death).

3.4.4.3. Biofilm Formation Testing

A biofilm formation assay was also performed to evaluate the antimicrobial effectiveness of Q8-SH functionalized catheter tubing, and its potential to prevent biofilm formation. P. aeruginosa, a particularly problematic biofilm forming species, was used to inoculate an assembly of catheter segments in the following order: CC1, untreated control, CC2, phys. ads., CC3, UV-treated, and CC4 (Appendix A, Figure 6.36). The untreated, phys. ads, and UV-treated allyl-TPU catheters were post-fabrication modified on the inner lumen with Q8-SH, and the intermixing of CC segments served as a control to monitor downstream effects of the experimental group. Brightfield microscopy images of catheter cross-sections taken after 48 h of growth demonstrated a significant biofilm had formed on the interior of CC1 and untreated allyl-TPU catheters, while the phys. ads.

69 and UV-treated catheters contained notably less material (Figure 3.9A-D). The % biofilm occlusion was quantified from brightfield imaging (n =3) (Figure 3.9E). It was evident that

CC1 contained the highest % biofilm blockage (ca. 90% internal volume occlusion), while the UV-treated catheter contained the least (ca. 15%). Interestingly, the untreated control allyl-TPU suffered approximately 50% less biofilm blockage than CC1, and the performance of the phys. ads. and UV-treated samples was comparable (77% and 85% reductions from CC1; 53% and 69% reductions from the untreated control, respectively).

CC2 – CC4 were not notably different from CC1, indicating that the experimental group did not affect the downstream assay (data not shown). In addition, a photograph of the experimental catheters provided a visual observation of the biofilm (Figure 3.9F); a notable biofilm formed on the untreated control, while the phys. ads. and UV-treated allyl-TPU modified with Q8-SH remained relatively clear, with the phys. ads. catheter slightly more turbid than the UV-treated catheter.

SEM imaging was also performed to confirm that the blockage was created through biofilm growth. The presence of bacterial extracellular polymeric substance (EPS) is quite notable on CC1 and the untreated control catheter segments, but much less for the phys. ads. and UV-treated catheters modified with Q8-SH (Figure 3.10A-D). At higher magnification (300×), SEM reveals the 3-dimensional architecture of the EPS material, which is more mature on CC1 and the untreated control than the phys. ads. and UV- treated catheters (Figure 3.10E-H). Qualitatively, the biofilm appears to be sparser on the

UV treated catheter than the phys. ads. control.

Figure 3.9. Brightfield microscopy images of catheter cross-sections (3.0 mm segments) from the 48 h biofilm assay were taken for (A) CC1, (B) untreated control, (C) phys. ads., and (D) UV treated samples modified with Q8-SH. (E) The % biofilm blockage was determined using Olympus VS-Desktop software and the averages and standard deviations are displayed (n=3). (F) Photograph of the untreated control, phys. ads., and UV treated catheters following completion of the 48 h biofilm assay.

Overall, blade-coated samples of allyl-TPU modified with the Qx-SH series demonstrated variable antimicrobial properties (adapted ISO 22196 assay), with Q8-SH proving to be the most effective. A live/dead fluorescence assay performed on allyl-TPU samples modified with Q8-SH revealed rapid contact-killing properties; nearly all the S. aureus inoculum was killed within 5 min and the E. coli within 10 min for UV-treated samples. In addition, a biofilm formation assay with P. aeruginosa showed that the catheter tubing functionalized with Q8-SH was more resistant to biofilm formation than a nylon-based Cook angiographic catheter (Cook® Beacon® Tip Torcon NB® Advantage), as well as untreated and phys. ads. control catheters.

Figure 3.10. SEM images demonstrating the appearance of bacterial EPS on catheter cross-sections from the 48 h biofilm assay were taken for (A) CC1, (B) untreated control, (C) phys. ads., and (D) UV treated samples modified with Q8-SH at 45× and (E-H) 300× magnification, respectively.

3.5. Conclusion

Incorporation of the commercially available 3-allyloxy-1,2-propanediol monomer into TPU provided a functional handle (alkene) that allowed for rapid and convenient surface modification (post-processing) using thiol-ene “click” chemistry, while maintaining relatively benign conditions (water, room temperature, UV light). Blade- coated samples of allyl-TPU were surface-functionalized with a series of Qx-SH compounds containing an 8-carbon spacer between the ester and ammonium head group, and various hydrocarbon tail lengths (8 – 14 carbons). A semi-quantitative assessment of the amount of Qx-SH available on the surface was performed using a fluorescence assay with a structurally comparable rhodamine-SH dye, and a series of XPS measurements on Qx-SH functionalized samples. The results suggested that quantity of

Qx-SH on the surface of allyl-TPU samples post-modification was likely between 1.9 ± 0.1 to 5.5 ± 0.1 nm∙cm2, of which < 1/3 is physically adsorbed.

It was determined via contact killing assays that surfaces modified with Q8-SH possessed the highest antimicrobial activity against both Gram-negative and Gram- positive bacteria, and this composition was scaled-up for further antimicrobial studies. A live/dead assay demonstrated that the UV-treated allyl-TPU samples modified with Q8-

SH killed the majority of S. aureus and E. coli inocula (OD600 = 0.15) within 10 min. In addition, it was evident that the UV-treated surfaces exhibited more rapid contact-killing than their respective phys. ads. controls. Biofilm formation testing also demonstrated that the accumulation of P. aeruginosa biofilm on UV-treated allyl-TPU catheters modified with Q8-SH was less compared to phys. ads., untreated, and CC controls as evidenced by brightfield microscopy and SEM imaging.

Future work will include hemolytic activity and cytotoxicity assays to evaluate compatibility for use in patients, and treatment of the substrates with bodily fluids ex vivo to quantify the adsorption of proteins or other biomolecules on the surface using quartz crystal microbalance (QCM). Additional testing, prior to in vivo studies, will involve passivating the functionalized surfaces with various proteins, and evaluating their contact-killing efficacy post-passivation.

3.6. Acknowledgment

The authors gratefully acknowledge financial support from Cook Medical and The

W. Gerald Austen Endowed Chair in Polymer Science and Polymer Engineering. The XPS spectra were obtained with the assistance of Dr. Zhorro Nikolov on VersaProbe II XPS

Microprobe in the Surface and Optical Analysis Facility of the National Polymer Innovation

Center at The University of Akron. Mass spectrometry was performed by Selim Gerislioglu in the lab of Dr. Chrys Wesdemiotis at The University of Akron Mass Spectrometry Center.

CHAPTER IV

IONOMERS FOR TUNABLE SOFTENING OF THERMOPLASTIC POLYURETHANE

This work has been reprinted with permission from Zander, Z. K.; Wang, F.;

Becker, M. L.; Weiss, R.A., Ionomers for Tunable Softening of Thermoplastic

Chemical Society.

4.1. Abstract

Thermoplastic polyurethane (TPU) sulfonate ionomers with quaternary ammonium cations were synthesized to achieve soft TPUs without using conventional low molecular weight plasticizers. The sulfonated monomer N,N-bis(2-hydoxyethyl)-2- aminoethane-sulfonic acid (BES) neutralized with bulky ammonium counterions was incorporated as a chain extender to internally plasticize the TPU. Increasing the steric bulk of the counterion and the concentration of the ionic species produced softer TPUs with improved melt processability. The incorporation of the sulfonate species suppressed crystallinity of the TPU hard block, which was mainly responsible for the softening of the polymer. The synthetic procedure developed allows for facile tuning of the mechanical properties of the TPU by simply switching the counterion and/or increasing the feed ratio of ionic monomer. The precursors in this study were synthesized and analyzed via 1H-

NMR, and the thermo-mechanical properties of the resulting TPU ionomers were characterized by differential scanning calorimetry, dynamic mechanical analysis, Shore A hardness and static mechanical testing.

4.2. Introduction

Ionomers are polymers that contain a small concentration of covalently bonded ionic species, such as carboxylate, sulfonate or phosphonate groups.31 In most cases,

“hard” counterions, such as metal ions, are used to form the ion-pair, which in ionomers is condensed because of the relatively low dielectric constant of the polymer matrix.32

This characteristic, in addition to the relatively low ion density, distinguishes ionomers

77 from polyelectrolytes, which are highly charged polymers that are usually water soluble.

The interest in ionomers stems from the large property changes that result from interchain supramolecular bonding of the contact ion-pairs. These interactions represent transient, reversible crosslinks that generally increase the modulus, strength, and toughness of the ionomer, though some extensibility of the parent polymer is lost due to the formation of a physical network. The presence of the ionic groups and phase- separated ionic nanodomains, often termed ionic clusters, that form in most ionomers also affect the glass transition temperature and the transport properties of the material.33-35

Less common is the addition of ionic functionality to a polymer for the purpose of internally plasticizing the polymer. This can be achieved by using bulky counterions, e.g., alkyl ammonium or phosphonium ions that weaken the ionic, dipole-dipole, or ion-dipole interactions responsible for the mechanical and physical property changes. For example,

Weiss et al.30 and Weiss and Stamato29 used alkyl ammonium cations with varying alkyl chain lengths to lower the glass transition and melt viscosity of sulfonated polystyrene ionomers.

Thermoplastic polyurethanes (TPU) are linear segmented block copolymers that possess polar hard segments derived from diisocyanates, such as methylene diphenyl diisocyanate (MDI), and relatively non-polar soft segments formed from oligomeric diols, such as polyesters and polycarbonates. The disparity in the polarity of these two segments and the crystallizability of the hard segment produces microphase separation

78 of the hard segments into nanodomains that provide physical crosslinks responsible for the desirable properties of TPUs, such as excellent elasticity, abrasion resistance, and toughness.175-176 Crystallization of the hard segments also increases the modulus and hardness (durometer) of a TPU. TPUs with lower durometer are normally achieved by adding low molecular weight plasticizers, such as dipropylene glycol dibenzoate or benzoate esters. However, those compositions tend to be tacky and are often difficult to process for use in common TPU applications. In addition, plasticizer leaching and migration is a major industrial challenge that eventually leads to a decline in the thermal and mechanical properties of plastics, and has also brought about serious health and environmental concerns.28, 177-186 As a result, increasing restrictions on the use of traditional plasticizers have created a demand for alternative methods for softening TPUs.

The objective of the research described herein is to demonstrate an alternative approach to the addition of plasticizers for softening TPUs, i.e., lowering the durometer and the melt viscosity. This approach involves incorporating bonded sulfonate groups with quaternary ammonium counterions into a TPU (Figure 4.1), similar to what was previously done to internally plasticize sulfonated polystyrene ionomers.29-30 In this report, the softening of TPU is achieved by incorporating an ionic diol, N,N-bis (2- hydoxyethyl)-2-aminoethane-sulfonic acid (BES), coupled with various bulky alkyl ammonium cations (Figure 4.2) during the chain extension step of the TPU synthesis. The hypothesis was that steric hindrance of the bulky quaternary ammonium groups would weaken the dipole-dipole interactions of the sulfonate groups and/or lower crystallinity

79 of the hard block and create additional free volume that softens the polymer and lowers the melt viscosity. Unlike the problem with softening TPUs with conventional low molecular weight plasticizers that can diffuse out of the polymer, the ammonium counterions in these compositions should be non-fugitive and non-extractable, because of the strong Coulombic interactions with the fixed counterion and the large enthalpic penalty that would accrue if two free ions were created in the non-polar continuous phase.

Figure 4.1. Diagram representing the plasticization of TPU using bonded sulfonate groups with bulky quaternary ammonium counterions. The blue rectangular portions represent hard segments, while the black lines represent the soft segments. The bis-hydroxyl functionality of the BES-ammonium monomer directs its placement into the hard segment, where it disrupts the polar interactions and crystallinity of the hard segments and softens the TPU.

+ Figure 4.2. Structure of BES, where X is a quaternary ammonium group (NR4 ), and R = CH3 or (CH2)nCH3 with n = 6, 10, or 12.

Another potential advantage of this approach is that one could conceivably develop a family of TPUs with varying durometer using a single ionic diol (BES) concentration, by simply varying the cation used. For example, hard cations, such as metal ions, could be incorporated in order to promote strong intermolecular attractive interactions between polymer chains that would harden the TPU. The amount of hardness could also be tuned by altering the size and/or the charge of the cation, which

81 influences the Coulomb energy of the ion-pair and the functionality of the physical crosslinks that arise from the supramolecular bonds.33-34 Conversely, the use of soft ions

(e.g., bulky quaternary ammonium counterions) may be used to develop softer TPU, and the degree of softening may be controlled by altering the length of the alkyl chain(s) or number of alkyl chains on the ammonium ion. Using this approach would offer the advantages of internal plasticization by the addition of bulky monomers, yet reduce the synthetic burden required to formulate a series of soft TPUs. This report describes the synthesis of soft TPU ionomers by incorporating quaternary ammonium sulfonate groups into the hard blocks and the characterization of their thermal, mechanical, and rheological properties.

4.3. Experimental

4.3.1. Materials

All commercial reagents and solvents were used as received without further purification, unless noted otherwise. Anhydrous tetrahydrofuran (THF), anhydrous dimethylformamide (DMF), 4,4’-methylenebis(phenyl isocyanate) (MDI), 1,4-butanediol

(BDO), stannous octoate, Amberlite® IRN-78 hydroxide form, didodecyldimethylammonium bromide (DDA-Br), tetrahexylammonium bromide (THA-

Br), terakis(decyl)ammonium bromide (TDA-Br), and N,N-bis (2-hydoxyethyl)-2- aminoethane-sulfonic acid (BES) were purchased from Sigma-Aldrich. The 2,000 푀̅푤 polycarbonate (PC) and polyester (PE) diols were donated by Bayer MaterialScience and dried under high vacuum to remove any residual solvent.

4.3.2. Preparation of Ionic Monomers

The ammonium bromide salts (DDA-Br, THA-Br and TDA-Br) were dissolved in a

75:25 (v/v) methanol/water mixture to produce a 20.0 g∙L-1 solution. For example: 5.00 g

(7.60 mmol) of TDA-Br was dissolved in 250 mL of a methanol/water mixture. The hydroxide exchange column was packed with 15.0 mL (1.10 meq) of Amberlite® IRN-78 residue suspended in the methanol/water mobile phase. The ammonium bromide salt solutions were slowly passed through the exchange column, with a flow rate of ca. 1 drop∙sec-1. The basicity of the eluent was monitored using pH paper to ensure that the bromide anion was being exchanged for hydroxide. The resulting ammonium hydroxide solutions were neutralized by adding solid BES with sonication to ensure sufficient dissolution and mixing. Once the solutions were completely neutralized, the methanol/water solvent was removed via rotary evaporation, followed by lyophilization.

The resulting crude products contained clear to off-white, viscous liquids as well as some white solids. The viscous liquids were extracted by dissolving the mixtures in THF and filtering off the insoluble salts. The THF was then removed via rotary evaporation, and the purified ammonium salt-BES monomers (DDA-BES, THA-BES and TDA-BES) were obtained. 1H-NMR confirmed the successful syntheses and purity of these ionic monomers, with typical yields between 75-90%. Ionic monomers (Appendix A, Figures

1 6.37 – 6.39) e.g., TDA-BES, H-NMR (300 MHz, DMSO-d6) δ: 0.91 (t, 12H), 1.29 (m, 56H),

1.51-1.68 (br, 8H), 2.56 (t, overlaps with DMSO-d6, 2H), 2.50 (t, overlaps with DMSO-d6,

4H), 2.79 (t, 2H), 3.11-3.26 (br, 8H), 3.40 (m, 4H), 4.41 (t, 2H).

4.3.3. Solution Polymerization of TPU and TPU Ionomers

Reactions using PC diols were carried out in DMF, whereas PE diol reactions were performed in THF as a result of the diol solubility. The following procedure describes the pre-polymer technique187 and is designed to produce a 30 wt% hard segment TPU, with and without ionic monomer. In a separate container, 8.00 g (4.00 mmol, 1.00 eq) of PC or PE diol was dissolved in 8-9 mL of DMF or THF, respectively, to produce a ca. 50 wt% solution of the diol. A 250 mL three-neck reaction flask equipped with a condenser, injection port, and mechanical stirrer was preheated to 65 °C. Then, 3.10 g (12.4 mmol,

3.10 eq) of MDI (solid) was added to the pre-heated reaction flask and purged with dry

N2 in order to melt the MDI, yet the temperature was low enough to limit dimerization.

The diol solution (50 wt%) was added drop-wise to the reaction flask over a period of 10-

15 minutes using a syringe pump while stirring the solution at 250 rpm. The slow addition of diol solution allowed for sufficient heat dissipation to control the temperature of the exothermic reaction that produced an isocyanate-terminated pre-polymer. After 2-3 h, an aliquot was taken from the reaction solution, and the isocyanate (NCO) concentration was measured by titration method using ASTM Standard D2572-97. When the desired

NCO content was achieved, the pre-polymer was chain-extended.

For a non-ionic TPU, 745 μL (8.40 mmol, 2.10 eq) of BDO was dissolved in 1-2 mL of DMF/THF to produce a 30 wt% solution, along with 2-3 drops (ca. 0.1 eq, < 0.2 wt%) of stannous octoate (if catalyst was utilized). The BDO/catalyst mixture was added dropwise via syringe pump into the reactor over a period of 5 min, with continuous stirring at 250

84 rpm and N2 purge. Non-ionic TPUs (Appendix A, Figure 6.42 and Figure 6.43), e.g. 30 wt%

1 hard segment content, H-NMR (300 MHz, DMSO-d6) δ: 1.54 (m, 44H), 1.71 (m, 4H), 2.31

(m, 44H), 3.79 (s, 2H), 4.10 (m, 4H), 4.21 (t, overlaps with previous peak, 48H), 7.22 (dd, J

= 7.94 Hz, 8H), 9.49-9.61 (m, 2H).

For a TPU ionomer, a mixture (totaling 8.40 mmol, 2.10 eq) of BDO and the desired ionic monomer (DDA-BES, THA-BES, or TDA-BES) was dissolved in 3-4 mL of DMF/THF to produce a 30 wt% solution, along with 2-3 drops of stannous octoate (if catalyst was utilized). The targeted weight percent of ionic monomer was calculated based on the molecular mass of the ionic monomer and the molar equivalents used, e.g. if a TPU containing 5 wt% THA-BES (566.92 g/mol) is desired, then 0.680 g (1.20 mmol, 0.30 eq) of

THA-BES and 638 μL (7.20 mmol, 1.80 eq) of BDO are used. The BDO/ionic monomer/catalyst mixture was added dropwise via syringe pump into the reactor over a period of 5 min, with continuous stirring at 250 rpm and N2 purge. After 2 – 3 h, an aliquot was taken and FT-IR was performed to confirm the absence of NCO. The resulting reaction solution was precipitated into cold methanol and filtered. If necessary, the product was re-dissolved in minimal THF and precipitated again into diethyl ether. The product was dried in an oven vacuum at 50 °C for 48 h. TPU ionomers (Appendix A, Figures

6.44 – 6.47), e.g. 4.1% TDA-BES TPU with 32.7 wt% hard segment content using PE diol,

1 “TPU30(PE)-4.1TD,” H-NMR (300 MHz, DMSO-d6) δ: 0.85 (t, 12H), 1.25 (m, 56H), 1.52 (m,

44H), 1.69 (m, 4H), 2.30 (m, 44H), 3.77 (s, 2H), 4.09 (m, 4H), 4.19 (t, overlaps with previous peak, 48H), 7.21 (dd, J = 7.96 Hz, 8H), 9.47-9.59 (m, 2H).

4.3.4. Bulk Polymerization of TPU and TPU Ionomers

For the purposes of achieving higher molecular weight (MW) polyurethanes and to ensure scalability, the TPU syntheses were also performed under bulk conditions. To produce 100 g of non-ionic TPU, 80.0 g (40.0 mmol, 0.210 eq) of PE diol and 5.00 mL (56.0 mmol, 0.290 eq) of BDO were preheated in a C-enamel lined tin can at 100 °C with mechanical stirring at 2000 RPM. In a separate container, a 5% excess of MDI was heated in an oven at 60 °C until melted. The excess MDI was to account for loss during transfer, so that the desired amount of MDI was achieved (24.0 g, 96.0 mmol, 0.50 eq). The molten

MDI was then poured into the reaction mixture containing diol species, and 3-5 drops of catalyst was added over a period of 10 s. After ca. 1 min of stirring at 250 RPM, the mixture became too viscous to stir and was poured/transferred onto Teflon paper. The resulting product was cured in an oven for 24 h at 90 °C.

To produce 100 g of 5% TDA-BES TPU ionomer, 80.0 g (40.0 mmol, 0.208 eq) of PE diol, 4.32 mL (48.8 mmol, 0.254 eq) of BDO, and 5.74 g (7.30 mmol, 0.0480 eq) of TDA-

BES were preheated in a C-enamel lined tin can at 100 °C with mechanical stirring at 200

RPM. In a separate container, a 5% excess of MDI was heated in an oven at 60 °C until melted. The excess MDI was to account for loss during transfer, so that the desired amount of MDI was achieved (24.0 g, 96.0 mmol, 0.500 eq). The molten MDI was then poured into the reaction mixture containing diol species, and 3-5 drops of catalyst was added over a period of 10 s. Qualitatively, a notable decrease in the bulk viscosity could be seen for the TPU ionomers compared to the control TPU, allowing for longer reaction

86 times. After ca. 2-3 min of stirring at 250 RPM, the mixture became too viscous to stir and was poured/transferred onto Teflon paper. The resulting product was cured in an oven for 24 h at 90 °C.

4.3.5. Characterization

The 1H-NMR spectra were obtained using a Varian NMRS 300 MHz spectrometer.

All chemical shifts are reported in ppm (δ), and referenced to the chemical shifts of

1 residual solvent resonances ( H-NMR: CDCl3 = 7.26 ppm, D2O = 4.79 ppm, DMSO-d6 = 2.50 ppm). Abbreviations for multiplicities are listed as following: s = singlet, d = doublet, t = triplet, br = broad singlet, m = multiplet. FT-IR spectra were recorded by a Digilab

Excalibur Series FTS3000, with a scanned wavenumber range from 400 to 4000 cm-1.

Samples were prepared by adding several drops of the reaction mixture to KBr pellets and spectra were recorded after 64 scans. The baseline was deducted and normalized to the

C-H stretch peak intensity. Durometer measurements were performed on compression molded, cylindrical samples (thickness: 6.35 mm, diameter: 14.0 mm). A type A-2 Shore

A durometer hardness tester (Shore Instrument & MFG Co. New York) was used, following the test procedure described in ASTM 2240. The instrument was calibrated using a standardized Shore A 60 material prior to each measurement. Differential scanning calorimetry (DSC) was performed using a TA Instruments Q2000 DSC on sample sizes of ca. 10 – 15 mg using temperature ramps for heating of 30 °C∙min-1 and a cooling rate of

30 °C∙min-1. The dynamic mechanical analysis (DMA) was performed using a TA

Instruments Q800 Dynamic Mechanical Analyzer. Thin sheets of compression molded

87 sample (thickness: 0.5 mm) were cut into rectangular pieces (width: 3.0 mm) for all measurements. Small-angle X-ray scattering (SAXS) experiments were performed on a

Testing Machine (Model 5567) in accordance with ASTM D638 for dumbbell shape Type

V and an extension rate of 500 mm∙min-1 using a 1 kN load cell. The samples containing

5 mol% ionic species or less were injection molded at ca. 185 °C using a HAAKE MiniJet

Injection Molding System. The sample with 7.6 mol% ionic species was injection molded at ca. 155 °C. After molding, the samples were aged in the laboratory at room temperature and ambient conditions for 48 h prior to testing. At least 4 specimens were used to calculate the tensile properties for each material. Oscillatory shear rheology measurements were performed on thin sheets of compression molded samples

(thickness: 0.5 mm) using a TA Instruments ARES-G2 Rheometer, equipped with 8 mm parallel plates. The linear viscoelastic response region was determined with a strain sweep conducted at a frequency of ω = 1 rad∙s-1. Dynamic mechanical analysis (DMA) was performed using a TA Instruments Q800 dynamic mechanical analyzer. Samples were cut from compression molded films (thickness ca. 0.5 mm). Dynamic tensile measurements were made using strain amplitudes, ε = 0.1 - 5 % depending on temperature, and temperature was scanned from -80 °C to 140 °C using a heating rate of 2 °C∙min-1 (ε =

0.2%, ω = 1 rad·s-1).

4.4. Results

4.4.1. Characterization of Ionic Monomers

1H-NMR was used to confirm the purity and successful syntheses of the ionic monomers. For example, a spectral overlay of BES and TDA-BES displayed significant differences in chemical shifts by the upfield shifting of each peak belonging to BES (peaks

4, 5, 5’, 7 and 8) upon coupling with the alkyl ammonium counterion (Figure 4.3). In addition, the TDA-BES spectrum shows the appearance of peaks corresponding to the aliphatic chains of the ammonium counterion (peaks 1-3, and 6), and integration of these peaks reveals a 1:1 molar ratio of BES:TDA, further indicating successful conversion

(Appendix A, Figure 6.37). The 1H-NMR integrations for DDA-BES and THA-BES (Appendix

A, Figures 6.38 and Figure 6.39, respectively) also demonstrate 1:1 molar ratios of

BES:ammonium counterion, indicating successful conversion. Several peaks for the ionic monomers are obscured by the DMSO-d6 residual solvent resonance (δ = 2.50). As a result, the BES:ammonium counterion stoichiometry was calculated by integrating the protons alpha to the hydroxyl groups (t, δ = 3.40) and alpha to the sulfonate anion (t, δ =

2.80) of BES, and comparing their ratio with peaks (1-3 and 6) from the various ammonium counterions. Mass spectrometry in negative mode was used to confirm the absence of bromine to provide further evidence that substitution occurred (Appendix A, Figure 6.40).

All ionic monomers were found to be stable viscous liquids at room temperature, which classifies them as ionic liquids, and they ranged from clear to slightly yellow in appearance.

Figure 4.3. 1H-NMR overlay for BES (top) and TDA-BES (bottom). A comparison of these spectra reveals that all of the peaks from BES (4, 5, 5’, 7, and 8) demonstrate a notable upfield shift as a result of the coupling. In addition, the disappearance of the sulfonate proton (peak 9) provides further evidence of successful conversion. It should be noted that peaks 4 and 5 in the TDA-BES spectrum overlap with the DMSO-d6 residual solvent peak.

4.4.2. Characterization of TPUs and TPU Ionomers

A variety of TPUs with varying hard segment (15, 25, and 30 wt%) and ionic monomer (1-8 mol%) contents were synthesized (Table 4.1). During TPU synthesis, FT-IR was used to measure the disappearance of the NCO peak at 2270 cm-1 (Appendix A, Figure

6.41). The expanded region between 2000 – 2500 cm-1 shows that the reaction was complete or nearly complete after 3 h; the intensity of the residual “peak” for the 3 h spectrum is comparable to the noise. In order to determine the hard segment and ionic

90 monomer content, the 1H-NMR singlet peak (s, δ = 3.78) from the methylene group of

MDI was used as a reference for the 1H-NMR integrations. This TPUs series containing varying hard segment content (Appendix A, Figures 6.42 – 6.43) and ionic monomer content (Appendix A, Figures 6.44 – 6.47) was analyzed by 1H-NMR, and their compositions were recorded (Table 4.1).

Table 4.1. List of ionic and non-ionic TPUs and their properties

Soft Hard Segment Ionic Monomer Molecular Weight Shore A Sample ̅ Segment Content (wt.%) Content (mol%) (g/mol) (푀̅푛/푀̅푤) Durometer

TPU30(PC) PC 30.0 0.0 / 21,500 71

TPU30(PC)-0.9DD PC 32.5 0.9 / 49,400 70

TPU30(PC)-2.5DD PC 32.4 2.5 / 26,700 70

TPU30(PC)-3.8DD PC 32.2 3.8 / 21,200 58

TPU30(PE) PE 29.0 0.0 11,400 / 20,300 93

TPU30(PE)-4.7TH PE 32.0 4.7 7,200 / 18,600 71

TPU30(PE)-4.1TD PE 32.7 4.1 7,900 / 19,500 67

TPU25(PE)Sn c PE 24.7 0 18,900 / 45,400 65

TPU25(PE)-4.3TD c PE 27.2 4.3 13,300 / 33,500 58

TPU30(PE)Sn c PE 29.4 0 12,900 / 37,000 86

TPU15(PE)Sn c PE 15.2 0 14,000 / 28,100 27

TPU25(PE)Sn c, d PE 24.8 0 55,500 / 93,500 72

TPU25(PE)-4.4TD c, d PE 24.6 4.4 - 58

TPU30(PE)-7.6TD c, d PE 29.0 7.6 - 37 a The hard segment and ionic monomer content were calculated by 1H-NMR. b The hard segment content includes the MDI, BDO and BES-ammonium monomer. c Stannous octoate catalyst was used to increase molecular weight. d Reactions were performed using bulk conditions, 100 °C with mechanical stirring. e Durometer measurements were performed 24 hours after compression molding. 91

A sample spectrum of 4.1 wt% TDA-BES TPU with PE diol “TPU30(PE)-4.1TD” is provided in Figure 4.4. An example calculation of the TPU composition is as follows: the peak for the methylene resonance in MDI (peak 5) was integrated and the ratio of the integration and the actual number of protons (2 for methylene) was set to a value of 1.

The relative concentration of the PE diol was then calculated from the ratio of the integration of peak 3 (17.25) and the actual number of protons in the diol (44). Thus, for the spectrum in Figure 4.4, the ratio of diol to MDI was (17.25/44) = 0.39. Similarly, the relative concentration of the ionic species was calculated from the ratio of the integration of peak 2 (4.60) and the actual number of protons in TDA-BES (6), i.e., (4.60/56) = 0.082.

Since the BDO resonances are not clearly resolved due to overlap with other peaks, the relative BDO concentration was assumed to be the difference between the MDI and the concentrations (normalized to MDI) of the PE diol and the ionic monomer. For the

TPU30(PE)-4.1TD spectrum, that calculation [BDO] = [MDI] – [Desmophen] – [DD-BES] is:

1 – 0.39 – 0.082 = 0.528.

Figure 4.4. 1H-NMR spectrum for TPU30(PE)-4.1TD. The peaks labeled 1 and 2 correspond to the protons on the aliphatic chains (2) and chain ends (1) for TDA-BES. Peaks 2, 5, and 6 were used for determining the hard segment and ionic monomer content.

The initial TPU series using DDA-BES was used to estimate the ionic monomer concentration required to lower the durometer (i.e., measured as Shore A hardness) of a

TPU containing 30 wt% hard segment with a PC diol soft segment. In addition to a control,

TPU30(PC), ionomers containing 0.9, 2.5, and 3.8% DDA-BES were synthesized

(TPU30(PC)-xDD where x = the mol% ionic monomer in product and DD denotes the DDA cation). The durometer values listed in Table 4.1 for this series indicate that an ionic monomer concentration of just 3.8 mol% produced significant softening (13 durometer points). A change from DDA to tetra-substituted ammonium salts (THA, TDA) with more

93 steric hindrance improved the softening effect. The PC diol was also switched to a PE diol to eliminate any crystallinity contributed by the soft segment. The data in Table 4.1 compare the effect of increasing the chain length of the ammonium salt from hexyl to decyl, and the effect of decreasing the hard segment content for a non-ionic TPU.

Increasing the alkyl chain length of the quaternary ammonium counterion from hexyl (n

= 6) to decyl (n = 10) produced softer TPU ionomers, as the Shore A durometer decreased from 93 for the non-ionic TPU30(PE) to 71 for TPU30(PE)-4.7TH and 67 for TPU30(PE)-

4.1TD, even though the latter TPU had a lower ionic monomer concentration (4.1 vs. 4.7 mol%, see Table 4.1). Decreasing the hard segment concentration also lowered the durometer, as would be expected by the lower MDI crystallinity in the TPU.

4.4.3. TPU Durometer and Softening

Shore A durometer measurements were performed following ASTM Standard D-

2240 (Standard Test Method for Rubber Property—Durometer Hardness), and the values for the TPUs are listed in the last column of Table 4.1. The one exception to the D-2240 procedure was that the specimens were aged in the laboratory at about 23 °C for only 24 h, as opposed to the 40 h prescribed by the ASTM Standard. The error involved with the shorter aging time is considered small, based on the time-dependent durometer data shown in Figure 4.5 for TPU25(PE), TPU25(PE)-4.4TD and TPU30(PE)-7.6TD, which shows that after aging the samples for 24 h (1440 min), the durometer appears to have reached equilibrium. The time dependence is most likely due to the slow crystallization of the hard segments, though that was not confirmed experimentally.

Figure 4.5. Time dependence of Shore A durometer values for a non-ionic TPU (■, TPU25(PE)Sn), and TPU ionomers with ca. 4% (●, TPU25(PE)-4.4TD) and 8% (▲, TPU25(PE)-7.6TD) TDA-BES monomer. The durometer was measured over a 24 h period (t = 0 was immediately following compression molding cooling cycle to room temperature), and indicates changes in the crystallization rate/behavior upon increasing incorporation of ionic monomer.

In general, decreasing the hard segment concentration and increasing the ionic monomer concentration produced softer TPUs (Figure 4.6). Also shown in Figure 4.6, the size of the two counterion used (THA vs. TDA) had only a minor effect on the durometer

(several point decrease); however, changing the substitution on the ammonium from two aliphatic chains to four decreased the durometer an additional ca. 10 points. Overall, the objective of this investigation to soften the TPU by the incorporation of sulfonate groups was achieved, though in addition to the sulfonate concentration, the amount of softening depended on the hard segment content of the TPU, the number of alkyl chains on the ammonium counterion and the size of the counterion. The lowest durometer reported in this study was a Shore A of 37 with TPU25(PE)-7.6TD, i.e., a TPU containing 25 wt% hard segment with the PE diol and 7.6 mol% of TDA-BES ionic monomer, which was a 35-point decrease from the control non-ionic TPU). 95

Figure 4.6. Effect of hard segment content (wt%), ionic monomer concentration (mol%), and ammonium countercation on the durometer of non-ionic TPUs (■), nominal TPU(PE)- 5TH (●), nominal TPU(PE)-5TD (▲), nominal TPU(PE)-8TD (♦). Shore A durometer values were measured after 24 h. The dashed line is the linear least squares fit of the data for the non-ionic TPUs. The ionic concentration of the nominally 5 mol% samples varied from 4.1 to 4.9 mol% and the nominally 8 mol% sample was 7.6 mol%.

4.4.4. Thermal, Mechanical and Viscoelastic Properties of the TPUs

Several techniques were employed to characterize the thermal, mechanical, and viscoelastic properties of the TPUs prepared with the PE diol. Typical DSC thermograms of the TPUs display the heating and cooling behavior of TPU30(PE), TPU30(PE)-4.7TH and

TPU30(PE)-4.1TD (Figure 4.7). In general, the introduction of the ionic monomer had little effect on the glass transition, but it suppressed crystallization, as evidenced by the absence of the exotherm seen in the cooling curve for the non-ionic TPU at ca. 155 °C for the ionic TPUs. The heating scan for the non-ionic TPU shows a melting point at 185 °C and a weak endotherm near 100 °C, the origin of which is not known with certainty; it may be melting of some small hard block crystals. The very small endothermic peak in all of the cooling scans near 80 °C is not likely due to the TPU, and is probably an artifact due to loss in control of the temperature of the DSC during cooling. It is not the hard segment 96 glass transition, which occurs at a lower temperature based on the dynamic mechanical data described later in this report. The ionic TPUs exhibited significant cold-crystallization between 50 and 100 °C (i.e., the broad exotherm in the heating scan), which was followed by a broad melting endotherm between 100 and 150 °C.

Figure 4.7. Example DSC thermograms for TPU30(PE) (bottom), TPU30(2)-4.7TH (middle) and TPU30-4.1TD (top): (a) cooling following first heating scan; (b) second heating scan. Exothermic behavior is up in these scans, and the curves have been vertically displaced for clarity. The crystallization peak for TPU30(PE) in the cooling scan near 155 °C is suppressed in the ionic TPUs, and the corresponding melting peak at 185 °C in the heating scan is broadened and shifted to lower temperature for the ionomers. In addition, the ionic TPUs exhibited significant cold-crystallization between 50 and 100 °C (i.e., the broad exotherm in the heating scan), which was followed by a broad melting endotherm between 100 and 150 °C.

The large suppression of the melting point compared to the non-ionic TPU (185

°C) is most likely due to the formation of very small crystallites. The change in the crystallization behavior and lowering of the melting point is consistent with the introduction of a non-crystallizable component into the TPU.188-189 The presumed small crystallites may also be a consequence of the ionic species nucleating crystallization, as has been observed with other semi-crystalline ionomers.190-192 However, at this time

97 those comments are strictly speculation, since the microstructure and crystallization behavior were not studied in detail. The lower crystallinity of the ionic TPUs is, at least, partly responsible for the softening effect of incorporating the ionic species into the polymer. The size of the cation may also have an effect, though it appears that the softening imparted by the ionic monomer is primarily a result of disrupting the hard segment crystallinity (as the Tg remains unaffected). Hence, the data in Figure 4.7 and

Table 4.1 suggest that the dominant influence for the softening is the reduction in the crystallization rate and the amount of crystallinity achieved.

Typical tensile stress strain data for TPU25(PE), TPU25(PE)-4.4TD and TPU25(PE)-

7.6TD are shown in Figure 4.8a, and the tensile properties are summarized in Table 4.2.

Because the linear region of the stress-strain curve was difficult to resolve for these elastomers, the modulus was reported as a secant modulus at 50% strain. The introduction of the ionic species decreased the modulus, presumably due to its suppression of crystallization as previously described. In addition, the durometer for these three TPUs tracked linearly with the 50% strain secant modulus (E50) (Figure 4.8b).

The linear fit had a coefficient of determination (r2) that was nearly unit y, which indicates the fit is excellent. However, the large 95% confidence intervals shown by the dashed lines indicate that more data are needed to validate the linear correlation. The ionomers also had lower stress at break than the non-ionic TPU, Table 4.2, which is consistent with the lower hard segment crystallinity. The shape of the stress-strain curves for TPU25(PE) and TPU25(PE)-4.4TD were similar and the stress levels achieved were also similar, though

98 the strain at break of was much lower for the ionomer. The lower strain at break may be due to a rougher surface of the ionomer specimen, which produced premature fracture.

That explanation is consistent with the large standard deviation for all of the samples and the much higher elongation at break for the TPU25(PE)-7.6TD sample, which is what would be expected as the hard segment crystallinity decreased. However, one cannot dismiss the possibility that the trend of decreasing strain to break for low sulfonate concentration and higher strain to break at higher concentration, because of the complex microstructure of these TPUs. Because of the large cations used, which typically suppress the ionic aggregation observed in metal-neutralized ionomers, it is doubtful that the lower strain to break is due to the formation of nano-scale ionic clusters. However, SAXS data for TPU30(2), TPU30(2)-4.7TH and TPU30(2)-4.1TD (see Appendix A, Figure 6.48) indicated that the domain spacing, do, for the TPU microstructure increased from do = 14 nm for the non-ionic TPU to closer to 20 nm when the ionic groups were introduced into the TPU.

Table 4.2. Tensile Properties of Non-ionic and Ionic TPUs with same wt.% Hard Segment

a b c Sample E50 MPa σu MPa εu %

TPU25(PE) 6.59 ± 1.20 21.2 ± 0.67 560 ± 130

TPU25(PE)-4.4TD 5.92 ± 0.51 18.8 ± 1.41 397 ± 94

TPU25(PE)-7.6TD 4.57 ± 0.97 6.85 ± 1.45 650 ± 125 a b c E50 = secant modulus at 50% strain, σu = breaking stress, εu = breaking strain.

Figure 4.8. (a) Representative stress vs. strain data for TPU25(PE) (■), TPU25(PE)-4.4TD (●) and TPU25(PE)-7.6TD (▲). (b) Shore A durometer vs. secant modulus at 50% strain (E50) for TPU25(PE), TPU25(PE)-4.4TD and TPU25(PE)-7.6TD. The solid, red line is the linearly least squares fit to the data (r2 = 0.997) and the dashed lines represent the 95% confidence limit for the linear correlation.

100

The effect of the alkyl chain length for the TPU ionomers neutralized with quaternary ammonium cations is shown by the dynamic mechanical properties of

TPU30(PE), TPU30(PE)-4.7TH and TPU30(PE)-4.1TD (Figure 4.9). The non-ionic TPU

(TPU30(PE)) showed a glass transition for the soft phase characterized by peaks in the loss modulus (E”) at -25 °C and tan δ at -22 °C (Figure 4.9a). A second mechanical transition occurred above room temperature, as indicated by the large drop in E’ and the peak in tan δ at 36 °C. The origin of this transition is not clear. It may be a glass transition associated with the hard segment nanodomains or it could be due to melting of smaller, imperfect crystals in the hard nanodomains. The glassy dynamic modulus (E’) of

TPU30(PE) was about 2 GPa and E’ dropped to ca. 200 MPa at the soft phase glass transition, and to ca. 10 MPa during the hard segment glass transition (ca. 20-40 °C). At

100 °C, E’ was ca. 2 MPa. The persistence of the rubber-like modulus at high temperature is consistent with the crystalline hard segments that act as physical crosslinks in the TPU.

For TPU30(PE)-4.7TH, with a C6 chain-length for the alkyl groups of the quaternary ammonium cation, the DMA curves (Figure 4.9b) show that the temperatures of the peaks in E” and tan δ for the soft phase glass transition, -25 °C and -20 °C, respectively, were similar to those for the non-ionic polymer. However, the decrease of E’ from ca. 2 GPa to ca. 30 MPa was much larger than that of the non-ionic polymer. This result is a consequence of the lower crystallinity in the ionomer, as evidenced by the DSC data in

Figure 4.7 that shows the crystallization rate was much slower in the ionomer. The drop in E’ at the transition above room temperature occurred over a broader temperature

101 interval, -25 – 50 °C, than for the non-ionic TPU, but peaks in E” and tan δ were not clearly resolved for the ionomer. E’ decreased much more rapidly above the hard segment glass transition compared to the non-ionic TPU, as E’ at 100 °C was less than 0.1 MPa for the ionomer.

The DMA curves for TPU30(PE)-4.1TD (Figure 4.9c), with a C10 chain-length for the alkyl groups of the quaternary ammonium cation, reveal only a single, very broad glass transition. Low temperature peaks for E” and tan(δ) were observed at -21 °C and -11 °C, which were lower than for the non-ionic TPU and TPU30(PE)-4.4TH, despite having a lower ionic monomer content. E’ decreased from ca. 2 GPa to 2 MPa over a temperature range from -5 °C to 40 °C. The much larger decrease of E’ for this ionomer indicates that either the crystallization rate of the hard segments was even slower for the TDA- neutralized ionomer than for the THA-neutralized ionomer, or that the bulkier cation was more effective at suppressing the crystallinity. In fact, the shape of the E’ curve in Figure

4.9c and the lack of any peak in E” and tan(δ) above 20 °C suggest that TPU30(PE)-4.1TD was not phase separated, i.e., that there was only a single, mixed phase. The SAXS data

(Appendix A, Figure 6.48) clearly indicate microphase separation, but the SAXS data were obtained at ca. 23 °C so they do not rule out the possibility of phase mixing above an upper critical solution temperature. As seen with the TPU30(PE)-4.7TH ionomer, E’ of

TPU30(PE)-4.1TD decreased relatively rapidly above room temperature, again consistent with little or no hard segment crystallinity, and E’ at 100 °C was only ca. 0.2 MPa.

102

Figure 4.9. DMA data for (a) TPU30(PE), (b) TPU30(PE)-4.7TH, and (c) TPU30(PE)-4.1TD: E’ (■), E” (●), and tan δ (▲) are plotted as a function of temperature.

103

The complex viscosity functions, η*(ω), for TPU30(PE), TPU30(PE)-4.7TH and

TPU30(PE)-4.1TD at 140 °C are shown in Figure 4.10. For the non-ionic polymer, the melting point of the hard segment phase was 186 °C, so the TPU was a semi-crystalline solid at 140 °C. As a result, there was no melt flow (η*→∞) and the data were consistent with yielding of a solid (slope = ca. -1). In contrast, the ionomers did exhibit melt flow with relatively low viscosities, ca. 300-400 Pa·s at higher frequencies, which would be more typical for conventional polymer processing operations. The improved processability of the ionomers is a result of the suppression of crystallinity in these materials, and is perhaps a result of some plasticization imparted by the bulky quaternary ammonium sulfonate (QAS) groups.

Figure 4.10. Frequency dependence of the complex viscosity (η*) at 140 °C for TPU30(PE) (■), TPU30(PE)-4.7TH (●), and TPU30(PE)-4.1TD (▲). This series of non-ionic TPU and ionomers possess similar hard segment concentration of nominally 30 wt%.

104

4.5. Conclusion

The objective of the project was to soften TPU via internal plasticization, i.e., without using a potentially fugitive plasticizer, and to demonstrate the potential for creating a family of TPU ionomers with a range of hardness by simply varying the counterion used in the ionic internal plasticizer. This was achieved using a conventional

TPU manufacturing process by adding an additional monomer, a commercial sulfonate- containing diol that was modified to a quaternary ammonium sulfonate (QAS) diol prior to polymerization. The Shore A durometer values for the TPU ionomers were decreased from non-ionic TPU controls (i.e., TPUs with comparable hard segment content) by as much as 35 durometer points, depending on the counterion and ionic monomer content

(see Table 4.1). The hypothesis was that the presence of the QAS in the polymer backbone would internally plasticize the TPU. This may have been, in part, responsible for achieving low durometer TPUs, but it appeared that the primary reason for increasing the softness of the TPU was suppression of crystallization. The crystallization suppression was either due to slower crystallization kinetics or disruption of the hard segment crystallinity, caused by the bulky ammonium counterion.

In addition to lowering the durometer of the TPU, the incorporation of the QAS entity lowered the melt viscosity and the temperature at which melt flow occurred for the TPUs, the latter which was a result of the lower crystallinity of the TPUs. The tensile modulus decreased, the strain at break increased, and the ultimate stress of the TPU decreased (Table 4.2) as the sulfonation level increased.

105

There are several questions regarding the details of changes in the TPU microstructure, crystallization kinetics, and rheological behavior remaining. More thorough investigation of those characteristics is in progress in order to understand how they are affected by the incorporation of sterically hindered sulfonate species– i.e., the development of a quantitative understanding of the relationships between the sulfonate concentration, the structure of the TPU, and the material properties. Developing this understanding and exploring a wider range of counterions (both hard and soft) will provide a platform that allows for precise control over the mechanical properties of TPU ionomers, achievable through facile ion exchange reactions. The method of plasticization described here is just one possible alternative to conventional plasticizers; however, this system suppresses crystallinity as opposed to suppressing the Tg, which is potentially useful for applications where maintaining the Tg of the material is critical. On the other hand, if the crystallization kinetics and/or crystalline content of the material are essential, this method would not be advised. However, there is potential to utilize a mixture of hard and soft counterions, which could possibly remedy the retardation of the crystallization rate and still achieve the desired softening.193

4.6. Acknowledgement

This work was supported by a grant from Bayer MaterialsScience (now Covestro

LLC). We thank Mr. Chongwen Huang, Chao Wang, Tyler Tommey and Dibyendu Debnath for their help with the rheology and mechanical property measurements.

106

CHAPTER V

CONTROL OF MESH SIZE AND MODULUS BY KINETICALLY DEPENDENT CROSS-LINKING IN

HYDROGELS

This work has been reprinted with permission from Zander, Z. K.; Hua, G.;

Wiener, C. G.; Vogt, B. D.; Becker, M. L., Control of Mesh Size and Modulus by Kinetically

Dependent Cross-Linking in Hydrogels. Advanced Materials 2015, 27 (40), 6283-6288.

4270790872744).

5.1. Abstract

Polyethylene glycol (PEG) based hydrogels have demonstrated their efficacy as synthetic scaffolds for tissue engineering.194 The favorable properties of PEG, including its hydrophilic nature, versatility in end group modification, and limited immunogenicity and antigenicity contribute to its wide use in hydrogel research for biomedical applications.37 In recent years, studies have shown that human mesenchymal stem cells

(hMSCs) can sense the rigidity of their microenvironment, and respond by altering their differentiation and proliferation processes.43 These studies have stimulated a great deal of interest in controlling the mechanical properties of synthetic scaffolds to direct the maturation of hMSCs. In particular, many researchers are focused on creating hydrogel systems with tunable mechanical properties in order to mimic the properties of various 107 native tissues; although, the majority of these systems require some chemical modification in order to achieve a range of moduli. Herein, we report a PEG-based hydrogel system that employs a kinetically-controlled oxime reaction for cross-linking, and allows for the production of mechanically distinguishable hydrogels using identical precursor chemistry. For an oxime cross-linking reaction, the kinetics of network formation are influenced by pH and buffer strength, and may be utilized to intrinsically control the degree of heterogeneity within the scaffold microstructure. Small-amplitude oscillatory shear (SAOS) rheology and small-angle neutron scattering (SANS) were used to determine how mechanical properties are influenced by structural features for this hydrogel system.

5.2. Introduction

Polyethylene glycol (PEG) based hydrogels are used widely for tissue engineering.194 The favorable properties of PEG, including its hydrophilic nature, ease of end group modification, and limited immunogenicity and antigenicity, contribute to its widespread use in biomedical applications.36-41 Many reports have shown that human mesenchymal stem cells (hMSCs) sense the rigidity of their microenvironment and respond by altering their genomic signaling as well as differentiation and proliferation processes.42-49 Consequently, there has been significant interest in controllably perturbing the mechanical properties of synthetic scaffolds to direct the maturation of hMSCs; in particular, hydrogel systems with tunable mechanical properties that mimic the properties of native tissues.38, 49-53 However, these systems generally require chemical or

108 structural modification in order to tune the moduli, which challenges the interpretation and control of cell studies. Strategies have included altering the precursor concentration or stoichiometry to modulate the effective cross-link density, adjusting precursor chain lengths to vary the molecular weight between cross-links, incorporating an additional or modified cross-linking agent, and changing the scaffold chemistry altogether.38, 43, 46, 49-58

To the best of our knowledge, very few reports have demonstrated a covalently crosslinked hydrogel system that employs invariant precursor chemistry, concentration and stoichiometry to produce gels with tunable mechanical properties.10, 49

Herein, we report a PEG-based hydrogel system that employs a kinetically- controlled oxime reaction for cross-linking, and allows for the production of mechanically distinguishable hydrogels without any alteration in precursor chemistry or composition.

The kinetics of network formation, which are influenced by pH and buffer strength, intrinsically regulate the degree of heterogeneity in the scaffold microstructure and produce the range of mechanical properties observed in this system. Prior research has shown that cross-linker reactivity greatly influences the degree of heterogeneity for in situ polymerized hydrogels.57-58 As network defects (dangling ends and elastically ineffective loops) become more prevalent, they dictate the physical properties of the gels, i.e. modulus and mesh size. Likewise, the cross-linker reactivity is expected to influence the properties of hydrogels fabricated from methods that do not involve radical cross- linking reactions. For example, the gelation time for hydrogel systems with acid/base driven cross-linking is dependent on cross-linking kinetics under various pH conditions,

109 but the resulting mechanical and structural properties were not explored.195-197 Here, we demonstrate how kinetically-controlled, oxime hydrogel systems can be used to modulate the mechanical properties and these property differences are directly linked to minor changes in structural features.

The oxime ligation involves a “click” condensation between aminooxy and carbonyl functional groups; the pH and buffer strength dictate the kinetics of oxime formation.195-196, 198 This reaction offers a robust method for hydrogel formation and use in biomaterial applications, as no metal catalyst is required, water is produced as a by- product, and it can be performed under ambient to physiological temperatures. In addition, the oxime reaction is an orthogonal method, allowing for multiple “click” reactions to be carried out in the same system.10 In general, the oxime reaction is acid- catalyzed under mild conditions (pH 4-6). At lower pH, the conjugate acid formation of the aminooxy component impedes the reaction, while at higher pH the hydrolysis becomes the rate-limiting step.199-200 Interestingly, the kinetics of oxime formation are also dependent on buffer concentration as neutrality is approached, which is demonstrated by the non-monotonic variation in gelation times with buffer concentration for the pH range (pH 5.7 – 7.1) selected in this study. In order to demonstrate how the rate of oxime formation/cross-linking influences the network structure and mechanical properties of the hydrogels, small-amplitude oscillatory shear

(SAOS) rheology and small-angle neutron scattering (SANS) were used to measure the gelation kinetics, the storage modulus (G’), and phase correlation length (δp) as a function

110 of both pH and buffer concentration. Equilibrium swelling studies enabled calculations of the mesh size (ξm) according to Flory-Rehner theory, providing additional evidence of these structural changes.201

5.3. Experimental

5.3.1. Materials

All commercial reagents and solvents were used as received without further purification. The 4-arm, 10K polyethylene glycol (hydroxy-terminated) was purchased from Creative PEGWorks, and 4-(dimethylamino)pyridine, p-toluenesulfonic acid monohydrate, (boc-aminooxy)acetic acid, levulinic acid, and diisopropylcarbodiimide

(DIC), 4.0 M hydrogen chloride solution in dioxane (4M HCl/dioxane), anhydrous methylene chloride (CH2Cl2), ethyl acetate (EtOAc), dimethylformamide (DMF), tetrahydrofuran (THF), methanol (MeOH), and hexanes were purchased from Sigma-

Aldrich. Diethyl ether (Et2O) was purchased from EMD Millipore. Potassium phosphate monobasic (KH2PO4) and sodium citrate dihydrate were purchased from Fisher Scientific.

Silica (porosity = 60 Å, surface area = 450 – 550 m2∙g-1, bulk density = 0.5 g∙mL-1, pH = 6.0

– 7.0) was purchased from Sorbent. The 4-(dimethylamino)-pyridinium-4-toluene sulfonate (DPTS) was prepared by adding molar equivalents of dimethylaminopyridine and p-toluenesulfonic acid monohydrate separately in THF, and collecting the precipitate by filtration.

111

5.3.2. Instrumentation

The 1H-NMR spectra were obtained using a Varian NMRS 300 MHz spectrometer.

All chemical shifts are reported in ppm (δ), and referenced to the chemical shifts of

1 residual solvent resonances ( H-NMR: CDCl3 = 7.26 ppm, D2O = 4.79 ppm). Abbreviations for multiplicities are listed as following: s = singlet, d = doublet, t = triplet, br = broad singlet, m = multiplet. FT-IR spectra were recorded by a Digilab Excalibur Series FTS3000, with a scanned wavenumber range from 400 to 4000 cm-1. Samples were prepared by grinding KBr powder with dried sample powder into pellets. The spectra were recorded after 64 scans and the baseline was deducted and normalized to the same reference peak intensity. Mass spectrometry for the 4-arm keto-PEG was performed using a Bruker

UltraFlex III MALDI tandem time-of-flight (TOF/TOF) mass spectrometer (Bruker

Daltonics, Billerica, MA, USA) equipped with a Nd:YAG laser emitting at 355 nm. The matrix and cationization salt were DCTB (2-[(2E)-3-(4-tert-butylphenyl)-2-methylprop-2- enylidene]malonitrile) and sodium trifluoroacetate, respectively. Solutions of the matrix

(20 mg∙mL-1) and cationizing salt (10 mg∙mL-1) were prepared in THF, and the polymer sample (10 mg∙mL-1) was prepared in water. The matrix and cationizing agent solutions were mixed in 10:1 (v/v) ratio and applied to the target. After drying, a spot of the sample was applied, followed by an additional drop of the matrix/cationizing agent. For the 4- arm aminooxy crosslinker, the spectrum was collected using a Bruker HCTultra II quadrupole ion trap (QIT) mass spectrometer (Billerica, MA) equipped with an ESI source.

The pH values of buffers were tested using an Orion® 350 PerpHecT® benchtop pH meter

112 with an Orion® ROSS® Sure-Flow pH electrode at room temperature. The moduli of the hydrogels were determined using an ARES G2 Rheometer (TA Instruments, New Castle,

DE) equipped with 8 mm parallel plate geometry. For time sweeps, 25 mm parallel plate geometry with a set gap of 0.80 mm was employed. SANS was used to investigate the structural features of the hydrogels. All measurements were taken at the National

Institute of Standards and Technology Center for Neutron Research (NCNR). Using the

4휋 instrument NGB30, the scattering “wavevector” q was measured, where 푞 = ( ) ∗ 휆

휃 sin ( ), λ is neutron beam wavelength and θ is scattering angle. Three detector distances 2

(13.2 m, 4 m, and 1.3 m), and the use of lenses for the 13.2 m detector distance, were examined to provide a measured q range of 0.001 Å-1 to 0.5 Å-1. The scattering results were circularly averaged over the 2D detector to attain the 1D scattering of q versus intensity. The data was fit with the Broad Peak model in the q range of 0.001 to 0.5 Å-1, and the phase correlation length (δp) of the gel nanostructure was determined.

5.3.3. Synthesis of 4-arm Aminooxy Cross-linker

In a 250 mL round bottom flask, 0.712 g (5.23 mmol, 1 eq) of pentaerythritol was dissolved in 20 mL of DMF, then 5.0 g (26.15 mmol, 5 eq) of (boc-aminooxy)acetic acid and 0.770 g (2.62 mmol, 0.5 eq) of DPTS were added. The reaction flask was purged with

N2 and stirred vigorously until all reagents were dissolved. The reaction flask was cooled to 0 °C in an ice bath for 15 min, followed by the injection of 4.05 mL (26.15 mmol, 5 eq) of DIC. The reaction was allowed to gradually warm to room temperature with stirring for 24 h. The reaction mixture was filtered to remove the urea by-product, and rotary 113 evaporated to remove solvent. The crude product was re-dissolved in EtOAc, cooled in liquid nitrogen and centrifuged to further remove urea by-product. The solution was concentrated and purified using silica column chromatography with a mobile phase of 5:3

(EtOAc:hexanes). After rotary evaporation, the white solid intermediate was obtained.

1 H-NMR (300 MHz, CDCl3): δ = 1.48 (s, 36H, (CH3)3), 4.24 (s, 8H, CH2), 4.45 (s, 8H, CH2),

7.90 (s, 4H, NH). The intermediate product was dissolved in ca. 25 mL of 4M HCl/dioxane to remove the protecting group, and stirred for 24 h at room temperature under N2 purge.

The resulting reaction solution was precipitated in cold diethyl ether, centrifuged, and vacuum dried. A fluffy white solid was obtained (1.75 g, 79.6% yield). 1H-NMR (300 MHz,

+ D2O): δ = 4.45 (s, 2H, CH2), 4.84 (s, 2H, CH2). ESI-MS m/z: [M+H] = 429.1 Da (calculated =

429.1 Da). Other species include sodiated and hydrated products: 451.1 [M+Na]+, 469.2

+ + [M+Na] ∙ H2O, 483.1 [M+H] ∙ 3H2O.

5.3.4. Synthesis of 4-arm Keto-PEG

In a 250 mL round bottom flask, 8.0 g (0.8 mmol, 1.0 eq) of 4-arm PEG (10K) was dissolved in 20 mL of CH2Cl2 with stirring and N2 purge. Then, 0.46 g (4.0 mmol, 5.0 eq) of levulinic acid was added to the flask, followed by 0.12 g (0.4 mmol, 0.5 eq) of DPTS.

The reaction flask was cooled in an ice bath for 15 minutes, and then injected with 0.62 mL (4.0 mmol, 5.0 eq) of DIC using a micropipette. The reaction flask was allowed to gradually come to room temperature with stirring under nitrogen for 24 hours. Using a

Buchner funnel, the reaction mixture was filtered to remove solid DIC-urea, and washed with CH2Cl2. The filtrate was concentrated via rotary evaporation and re-dissolved in a

114 minimal amount of CH2Cl2 for precipitation into cold methanol, followed by centrifugation. The precipitate was then re-dissolved in a minimal amount of solvent again, and precipitated into cold diethyl ether, followed by centrifugation at 5000 RPM for 2 minutes. The precipitate was then vacuum dried, and a white solid was obtained

1 (7.10 g, 85.0% yield). H-NMR (300 MHz, CDCl3): δ = 2.15 (s, 3H, CH3), 2.57 (t, 2H, CH2),

2.71 (t, 2H, CH2), 3.38 (br, 2H, CH2), 3.61 (m, 224H, [OCH2CH2]56), 3.84 (t, 2H, CH2), 4.19 (t,

+ 2H, CH2). MALDI-MS m/z: [M+Na] calculated for 56-mer C13H24N4O12 = 10,388.12 Da; observed = 10,390.64 Da. End group analysis for the 56-mer provided a calculated mass

= 532.13 Da; observed = 534.64 Da, which is approximately the mass of the C5H8 pentaerythritol core (68.13 Da) + four levulinic acid groups (4 × 116.05 = 464.2 Da).

5.3.5. Phosphate-Citrate Buffer Preparation

Stock solutions of 0.1 M citric acid and 0.1 M disodium phosphate were prepared ahead with ultrapure water. The pH values and buffer concentrations for different gel systems were adjusted by diluting with additional water and mixing different portions of the desired buffer concentration solutions to obtain the desired pH. Precise pH values of the buffer solutions were tested with a pH meter.

5.3.6. Hydrogel Fabrication

In general, hydrogels were fabricated using a precursor mixing method: the precursor solutions were prepared by dissolving keto-PEG (138.2 mg, 0.013 mmol) in the desired buffer (800 μL), and 4-arm cross-linker (7.6 mg, 0.013 mmol) in the desired buffer

(200 μL). The precursor solutions were then mixed with sufficient initial shaking to

115 provide a balanced stoichiometric solution of keto-PEG and 4-arm crosslinker, and gelation ensued. For oscillatory shear measurements and swelling studies, the hydrogel precursor solutions were rapidly mixed and cast into silicone molds, then covered and placed in a reduced airflow environment to minimize evaporation. The gels were allowed to set for at least 8 hours to ensure complete gelation, and punched to 8.0 mm diameter

(ca. 2 mm thick) using a biopsy punch. For SANS measurements, the hydrogel precursor solutions were rapidly mixed and injected into a 0.5 mL titanium cell (25.0 mm diameter,

1.0 mm path length), then allowed to react for at least 8 hours. All hydrogels were fabricated at room temperature, and were designed to produce 12 wt.% PEG hydrogels.

The mass fraction of total precursors was calculated using Equation 2:

m(PEG)+m(cross−linker) wt% = (2) m(4PEG)+m(cross−linker)+m(buffer)

The mass fraction (wt.%) of PEG was determined by removing the mass of the cross-linker from the numerator of Equation 2. All of the gels cast for the purposes of this paper were 12 wt.% PEG (vol. fraction φ = 0.135). FT-IR was used to confirm hydrogel formation, monitoring the dissapearance of the carbonyl from keto-PEG as the appearance of the carbonyl from the crosslinker. Inverted tube tests, as well as rheology time sweeps also confirm the gelation process.

5.3.7. Swelling Studies

The hydrogels were cast in a silicon mold and punched to 8.0 mm diameter (ca. 2 mm thick) using a biopsy punch, then swollen in a glass vial with deionized water (2.5 mL) for 48 h. The excess water was gently removed by blotting with a soft tissue, and the

116 samples were weighed using an analytical balance (±0.01 mg) to obtain the swollen mass

(Ms). The hydrogel samples were then lyophilized for 48 hours and weighed again to obtain the dry mass (Md). After determining the swollen ratio (Q), the Flory-Rehner equations (Equation 3 – Equation 6) were utilized to calculate the average mesh size (ξm)

10 for hydrogels formed under each pH and buffer concentration condition (n = 5). For this series of equations, v2 is the swollen polymer volume fraction, 푣̅ is the specific volume of

-1 3 -1 PEG (0.893 cm ∙g), V1 is the molar volume of water (18 cm ∙mol ), 푀̅푛is the number-

-1 average molecular weight (10,330 g∙mol ), 푀̅푐 is the average molecular weight between

2 1/2 crosslinks, χ1 is the polymer-solvent interaction parameter (0.426 for PEG in water), (푟표̅ ) is the root-mean-square end to end distance, l is the bond length (1.46 Å), and Cn is the characteristic ratio for PEG (4.0).10, 51, 194

1 푀 −푀 = 푄 = 푠 푑 (3) 푣2 푀푑

2 1 2 푣̅/푉1[ln(1−푣2)+푣2+휒1푣2 ] = − 1/3 푣2 (4) 푀̅̅̅푐̅ 푀̅̅̅̅푛̅ 푣 − 2 2

1/2 2 1/2 1/2 2푀푐 (푟표̅ ) = 푙 ∙ 퐶푛 ∙ ( ) (5) 푀푟

−1/3 2 1/2 휉푚 = 푣2 ∙ (푟표̅ ) (6)

5.3.8. Hydrogel Rheology

The moduli of the hydrogels were determined using an ARES G2 Rheometer (TA

Instruments, New Castle, DE) equipped with 8 mm parallel plate geometry. The hydrogel samples were centered on the test geometry, and the gap height was set to ca. 1.6 mm with a constant normal force of ca. 2 N. Initially, a strain sweep was performed to

117 determine the linear viscoelastic regime (LVR), and 1% strain was selected for frequency sweeps. The frequency sweeps were conducted from 100 rad∙s-1 to 0.1 rad∙s-1; all frequencies between 1-100 rad∙s-1 demonstrated linear behavior, and 10 rad∙s-1 was selected for reporting (n = 3). It should be noted that no notable drying of the hydrogels nor slippage was observed during testing. Time sweeps were performed at 1% strain and

-1 1 rad∙s to determine the gelation time. The precursor solutions were mixed and shaken for 5 seconds before injecting 400 μL of the hydrogel solution into the rheometer equipped with 25 mm parallel plates and a set gap height of 0.80 mm, and the response was measured.

5.3.9. Small-Angle Neutron Scattering

SANS was used to investigate the structural features of the hydrogels. All measurements were taken at the National Institute of Standards and Technology Center for Neutron Research (NCNR). Using the instrument NGB30, the scattering “wavevector”

4휋 휃 q was measured, where 푞 = ( ) ∗ 푠푖푛 ( ), λ is neutron beam wavelength and θ is 휆 2 scattering angle. Three detector distances (13.2 m, 4 m, and 1.3 m), and the use of lenses for the 13.2 m detector distance, were examined to provide a measured q range of 0.001

Å-1 to 0.5 Å-1. The scattering results were circularly averaged over the 2D detector to attain the 1D scattering of q versus intensity. The data was fit with the Broad Peak model

-1 in the q range of 0.001 to 0.5 Å , and the phase correlation length (δp) of the gel nanostructure was determined.

118

5.4. Results

5.4.1. Precursor Synthesis and Hydrogel Formation

The 4-arm cross-linker used for hydrogel formation was syntheiszed in two steps by esterification of the boc-protected hydroxylamine with pentaerythritol, followed by deprotection, and the structures were confirmed by 1H-NMR and ESI-MS (Appendix A,

Figures 6.49 – 6.51). Keto-PEG was synthesized by esterification of levulinic acid with hydroxy-terminated tetra-PEG, which was confirmed by 1H-NMR and MALDI-MS

(Appendix A, Figures 6.52 – 6.55). A series of hydrogels were prepared by mixing separate equimolar precursor solutions of levulinic acid-functionalized tetra-PEG (keto-PEG, 10 kDa) and 4-arm aminooxy cross-linker, each dissolved in phosphate-citrate buffer of the desired pH and buffer concentration. FT-IR confirmed hydrogel formation as the carbonyl from keto-PEG disappears and the carbonyl from the crosslinker become more prevalent

(Appendix A, Figure 6.56). Scheme 5.1 outlines the conditions used for hydrogel formation and depicts how the observed heterogeneity effects the mechanical and structural properties of these hydrogels. The pH range and buffer concentrations selected were designed to alter the cross-linking reaction kinetics with increasing pH and buffer strength, yet remain within various physiological tissue conditions.

119

For the rheology and swelling studies, the keto-PEG and 4-arm cross-linker solutions were mixed and subsequently cast into rectangular silicon molds. The molds were placed in a humidifed container to minimize evaporation, and allowed to react for at least 8 hours at room temperature to ensure complete gelation. Cylindrical hydrogel samples for the rheology and swelling studies were punched from this mold to generate samples with 8 mm diameter and 2 mm thick.

Scheme 5.1. Tunable hydrogels were fabricated by mixing 4-arm crosslinker and keto- PEG precursors in phosphate-citrate buffer over a range of pH (5.7 – 7.1) and buffer concentrations (10 – 100 mm). The kinetics of network formation are influenced by pH and buffer strength, and can be exploited to control the structural properties of the hydrogel network, which in turn dictate the mechanical properties (G’) for the resulting hydrogel. The cross-linking reaction involves oxime formation, which produces water as a by-product and does not require a metal catalyst.

5.4.2. Mechanical Properties

The mechanical properties and gelation times were examined using an Ares G2

Rheometer equipped with 8 mm parallel plate geometry. Initially, a strain sweep (γ) was conducted to determine the linear viscoelastic region (LVR) followed by a frequency

120 sweep (ω) at the fixed strain (γ = 1%) (Appendix A, Figure 6.57). All frequencies between

1-100 rad∙s-1 demonstrated linear behavior and ω = 10 rad∙s-1 was selected for reporting

G’ (Figure 5.1A). Strain and frequency sweeps show the majority of the data is within the

LVR and non-linear behavior is only observed at large strains and low frequencies

(Appendix A, Figure 6.58). To illustrate the effect of pH and buffer strength on gelation kinetics for this oxime system, time sweeps were also conducted (γ = 1%, ω = 1 rad∙s-1) by rapidly mixing the precursors for 5 seconds and injecting 400 μL of the solution into the rheometer and measuring the response in parallel plate geometry with a gap of 0.80 mm.

The time sweeps shown in Figure 5.1B and Figure 5.1C demonstrate the gelation kinetics dependence on pH and buffer strength, respectively. In general, the lower pH (5.7) hydrogels formed faster due to acid catalysis; however, the buffer strength significantly influenced cross-linking kinetics as neutrality was approached for high buffer concentration (100 mm) samples, which displayed much longer gelation times (Figure

5.1C). Since the kinetics of the oxime reaction are known to vary with pH and buffer concentration and all other variables were held constant, the variations in the resulting storage moduli of the hydrogels (Figure 5.1A) indicated that a kinetically controlled cross- linking reaction may be utilized to produce mechanically distinguishable hydrogels from identical precursors. To gain a deeper understanding of how the rate of cross-linking influences mechanical strength, a detailed investigation of the network microstructure was pursued.

121

Figure 5.1. Rheology data demonstrating tunable gelation times and moduli. (A) The resulting G’ for hydrogels formed at each pH (5.7 = ●, 6.8 = ■, 7.1 = ▲) were plotted against the various buffer concentrations (10, 20, 50, 100 mm) to show the production of mechanically distinguishable gels, ranging from G’ = 20.0 – 31.5 kPa. In addition, time sweeps were performed at 1% strain and 1 rad∙s-1 to monitor the gelation point for hydrogels formed at different pH (B) and also at different buffer concentration (C). The pH time sweeps shown were performed on hydrogel precursor solutions of 100 mm 122 buffer concentration to demonstrate the effect of pH on cross-linking kinetics (n=3). The buffer strength time sweeps were performed at pH = 5.7 to demonstrate the influence of buffer concentration on reaction kinetics (n=3). All hydrogels were prepared either from the same batch, or different batches with weight variations ≤ 0.2 mg. Standard conditions include precursors mixed under stoichiometric balance (1:1, aminooxy:ketone) at room temperature, and 12 wt.% PEG (vol. fraction, φ = 0.135).

5.4.3. Structural Properties

Neutron scattering and swelling studies were used to quantify the structural properties of the hydrogels and establish a correlation between the mechanical properties and nanoscale structural features. The absolute scattering from SANS was fit using the Broad Peak model (Equation 7), where qo is related to the phase correlation length by the relationship qo = 2π/δp, q is the scattering vector, n is the Porod exponent, m is the Lorentzian exponent, ξL is the Lorentzian screening length, and the remaining variables (A, B, C) are scaling constants.202-203

퐴 퐶 퐼(푞) = 푛 + 푚 + 퐵 (7) 푞 1+(|푞−푞표|휉퐿)

The Broad Peak model can be used to characterize amorphous soft materials with

204 scattering length density correlations. Typically, the d-spacing (do) is provided by qo for phase separated systems (qo = 2π/do), but is interpreted here as the distance between scattering inhomogeneities related to cross-linking defects and denoted as the phase correlation length, δp. Figure 5.2A illustrates the scattering profiles for different buffer concentrations at pH 5.7. At low q (0.001 – 0.01 Å-1) associated with large length scales, there are statistically significant differences in the scattering, however, the scattering is almost invariant with buffer concentration in the high q regime (0.05 – 0.5 Å-1). This

123 behavior is observed at each pH (Appendix A, Figures S10 – S11). The differences seen at low q (0.001 – 0.01 Å-1) suggests changes in the large-scale (102 – 103 Å) structural features, such as network defects and cross-link density fluctuations, while the similarities at high q (0.05 – 0.5 Å-1) reveal that the hydrogel structures remain the same on a molecular scale (< 102 Å). Figure 5.2B illustrates the high quality fits of these scattering data to the Broad Peak model (Equation 7). The raw scattering profiles and corresponsing fits for pH 6.8 and 7.1 are also shown in Appendix A (Figure 6.58 and Figure 6.59, respectively). From the fits, δp provided an indication of the distance between scattering length density fluctuations, which can be related to the relative degree of inhomogeneity resulting from cross-link/network defects in the hydrogel structure.

Figure 5.2. The absolute scattering profiles of hydrogels fabricated at pH 5.7. (A) The series of buffer concentrations exhibit differences at low q (10-3 – 10-2 Å-1), but are similar in the high q regime (>10-1 Å-1). The intensity variations seen at low q suggests changes in the large-scale structural features, such as network defects and cross-link density fluctuations, while the similarities at high q demonstrate that the structures are identical on a molecular level (1 – 10 Å). (B) The scattering curves (pH 5.7) for each buffer concentration were fit using the Broad Peak model (solid lines) to determine the the phase correlation lengths (δp). The scattering curves are offset vertically for clarity.

124

To further support the SANS data for structural characterization, swelling studies were performed on the hydrogel samples. After determining the swelling ratio (Q), Flory-

Rehner theory was used to obtain ξm for the hydrogels at each pH and buffer concentration, as shown in Table 5.1.

Table 5.1. Summary of the mechanical and structural properties determined for the kinetically-controlled hydrogels. The association between mechanical and structural properties generally reveals an inverse relationship; where G’ is at a maximum, ξm and δp is at a minimum.

a b c pH [Buffer] (mM) G’ (kPa) ξm (Å) δp (Å)

5.7 10 22.24 (±0.02) 107.2 (±0.3) 144 (±2)

20 27.9 (±0.3) 100.9 (±0.4) 129 (±1)

50 23.1 (±0.1) 104.6 (±0.8) 157 (±4)

100 20.1 (±0.8) 107.2 (±0.6) 239 (±17)

6.8 10 25.2 (±0.6) 103 (±1) 162 (±4)

20 28.02 (±0.05) 99 (±1) 150 (±3)

50 31.1 (±0.5) 95.9 (±0.4) 191 (±10)

100 23.9 (±0.9) 104 (±3) 172 (±3)

7.1 10 21.6 (±0.4) 104.8 (±0.5) 178 (±7)

20 26.0 (±0.2) 100.7 (±0.8) 149 (±3)

50 31.55 (±0.03) 96.4 (±0.8) 146 (±2)

100 29.9 (±0.4) 100 (±2) 187 (±6) a Storage moduli (G’) were obtained from SAOS frequency sweep data at 10 rad∙s-1 (n = b 3). Mesh sizes (ξm) were determined from swelling studies using the series of equations c described by Flory-Rehner theory (n = 5). Phase correlation length (δp) results were derived using the Broad Peak model to fit SANS scattering curves, error corresponds to the standard deviation resulting from general fit error.

125

5.4.4. Correlating Structural and Mechanical Properties

The phase correlation length was compared to the mesh size, and their relationship with the resulting storage modulus was analyzed. Figure 5.3A shows the δp,

ξm, and G’ for hydrogels formed at pH 5.7, and indicates that δp and ξm follow a similar trajectory with an apparent inverse relationship to G’ for these length scales. In general, when G’ is maximized δp and ξm appear at a minimum, and when G’ is at a minimum δp and ξm are maximized. This correlation can also be seen for pH 6.8 (Figure 5.3B) and pH

7.1 (Figure 5.3C), as δp and ξm appear to be nearly inverse that of G’ with increasing buffer concentration. These results are intuitive because δp and ξm provide an indication of the relative size and distance between network defects, and it is understandable that a more flawed network (with larger δp and ξm) would produce an elastically weaker hydrogel. The

ξm from swelling studies assumes a homogeneous network and provides a mesh size based on the molecular weight between cross-links; network defects increase the molecular weight between cross-links and yield an increased ξm, allowing for an interpretation of the relative size of defects in the network. The differences seen between δp and ξm are likely related to the homogeneous network assumption, as well as the difference between the distance of scattering length density fluctuations measured by δp and the actual size of the mesh. In addition, ξm follows the inverse relationship with

G’ more closely than δp, since the swelling ratio is directly influenced by elastically effective chain entanglements, which are difficult to identify by SANS. The relevant physical quantities measured by rheology, SANS, and swelling studies are displayed in

126

Table 5.1. Overall, the data indicates that smaller mesh sizes/correlation lengths are associated with higher moduli gels, whereas low moduli gels correspond with larger δp and ξm. This relationship is a direct consequence of decreased or increased structural heterogeneities within the hydrogel network, respectively.205-206 More importantly, the data suggests that nanoscale structural details can be influenced by reaction kinetics to induce changes in macroscopic mechanical properties.

127

Figure 5.3. Correlation between the mechanical properties (G’ = ■) and structural properties of mesh size (ξm = ●), and phase correlation length (δp = ▲). For hydrogels produced at pH 5.7 (A), δp from SANS trends similarily with ξm derived from swelling studies, and both are inversely related to the storage modulus of the hydrogels. These trends are also seen for pH 6.8 (B) and pH 7.1 (C), as ξm and δp demonstrate an approximately inverse relationship with storage modulus. 128

5.5. Conclusion

In conclusion, it was shown that small changes in the structural features, resulting from variations in the kinetics of cross-linking, are capable of producing mechanically distinguishable hydrogels from chemically identical precursors. For an oxime cross- linking reaction, the kinetics of network formation are influenced by pH and buffer strength, and may be utilized to intrinsically control the degree of heterogeneity within the scaffold microstructure. Further developing and understanding this and other kinetically-driven systems should enable a facile route for precise control over the mechanical properties of hydrogels. Overall, these PEG-based hydrogels could provide definitive insight into the effect of substrate elasticity on human mesenchymal stem cell

(hMSC) differentiation because the scaffolds employ the same precursor chemistry, yet exhibit a range of mechanical properties.

5.6. Acknowledgement

This work was funded by RESBIO “Integrated Technology Resource for Polymeric

Biomaterials” (NIH-NIBIB & NCMHD P41EB001046) and the National Science Foundation

(DMR-1105329).

129

CHAPTER VI

CONCLUSION

6.1. Developing Functionalized Polymer Systems

In conclusion, it was demonstrated that the incorporation of various functionalities into polymer systems can provide precise control over the resulting material properties and/or specific interactions. The addition of 3-allyloxy-1,2- propanediol into a TPU provided an alkene functional handle that allowed for rapid and convenient surface modification, post-processing. The TPU surfaces were functionalized with QAC and exhibited contact-killing activity; a specific interaction with bacteria cell membranes that causes cell death. Similarly, ionic monomers of N,N-bis(2-hydoxyethyl)-

2-aminoethane-sulfonate coupled with a variety of bulky ammonium counterions were incorporated into the backbone of a TPU, resulting in significant mechanical property changes that decreased the shore A durometer. Lastly, tetra-PEG chain ends were modified with levulinic acid to provide ketone functional groups and reacted with a 4-arm aminooxy cross-linker in a series of different buffer conditions to alter the cross-linking kinetics. The variations in cross-linking kinetics resulted in the formation of chemically identical hydrogels with distinguishable mechanical and structural properties. Additional conclusions and suggested future work for each of the polymer platforms developed in this dissertation are provided in the following sections. 130

6.1.1. Contact-Killing Thermoplastic Polyurethanes

The successful incorporation of 3-allyloxy-1,2-propanediol into a TPU provided an alkene functional handle that allowed for efficient surface modification of blade-coated and extruded catheter samples via thiol-ene “click” chemistry. The “click” reactions proved to be effective in relatively benign conditions (water, room temperature, UV light), as evidenced by XPS and fluorescence spectroscopy. Blade-coated samples of allyl-TPU were functionalized with a series of Qx-SH compounds containing various hydrocarbon tail lengths (8 – 14 carbons), and a quantitative assessment of the amount of Qx-SH available on the surface was performed using a fluorescence assay and a series of XPS measurements. The results indicated that the quantity of Qx-SH on the surface of allyl-

TPU samples post-modification was likely between 1.9 ± 0.1 – 5.5 ± 0.1 nm∙cm2, of which less than 1/3 was physically adsorbed.

Overall, blade-coated samples of allyl-TPU modified with the Qx-SH series demonstrated higher antimicrobial activity with shorter alkyl tail lengths (i.e. 8 > 12 > 14 carbon tail). From the contact-killing assay (ISO 22196), it was determined that surfaces modified with Q8-SH possessed the highest antimicrobial activity towards gram-negative and gram-positive microorganisms (Table 3.1). Although the contact-killing assay did not reveal a difference between Q8-SH modified UV treated samples phys. ads. controls, differences were observed for Q12-SH modified UV treated samples phys. ads. controls were observed. This led to the conclusion that the Q8-SH must be a potent antimicrobial, prompting a scale up of this composition. A live/dead fluorescence assay performed on

131 allyl-TPU samples modified with Q8-SH revealed rapid contact-killing properties, in which nearly all S. aureus and E. coli inocula (OD600 = 0.15) were killed within 5 min and 10 min, respectively. In addition, it was evident from this assay that the UV treated surfaces exhibited more rapid contact-killing than their respective phys. ads. controls (Figure 3.8).

Lastly, biofilm formation testing with P. aeruginosa showed that the catheter tubing functionalized with Q8-SH was more resistant to biofilm formation than a Cook® Beacon®

Tip Torcon NB® Advantage catheter, as well as untreated and phys. ads. control catheters, evidenced by brightfield microscopy and SEM imaging (Figure 3.9 and Figure 3.10, respectively).

Future work will include hemolytic activity and cytotoxicity assays to evaluate compatibility in vitro. In addition, of the substrates with bodily fluids ex vivo should be performed to quantify the adsorption of proteins or other biomolecules on the surface using quartz crystal microbalance (QCM). This data would provide an understanding of whether the Q8-SH modified allyl-TPU surfaces will foul with biomolecules present at the intended implant site. Additional testing, prior to in vivo studies, should involve passivating the functionalized surfaces with various proteins (albumin and fibrinogen), and evaluating their contact-killing efficacy post-passivation.

6.1.2. Soft Thermoplastic Polyurethane Ionomers

132 counterion coupled to the sulfonate monomer. This was achieved using a conventional

TPU manufacturing process by adding an additional monomer, a commercial sulfonate diol that was modified to a quaternary ammonium sulfonate (QAS) diol prior to polymerization. The Shore A durometer values for the TPU ionomers decreased from non-ionic TPU controls (i.e., TPUs with comparable hard segment content) by as much as

35 durometer points, depending on the counterion and ionic monomer content (see Table

4.1). The hypothesis was that the presence of the QAS in the polymer backbone would internally plasticize the TPU. This may have been, in part, responsible for achieving low durometer TPUs, but it appeared that the primary reason for the softening of the TPU was suppression of crystallization. The crystallization suppression was either due to slower crystallization kinetics or disruption of the hard segment crystallinity, caused by the introduction of the QAS entity.

In addition to lowering the durometer of the TPU, the incorporation of the QAS lowered the melt viscosity and the temperature at which melt flow occurred for the TPUs, the latter of which was a result of lower crystallinity in the TPUs. The tensile modulus decreased, the strain at break increased, and the ultimate stress of the TPU decreased

(Table 4.2) as the QAS content increased.

Additional work regarding the details of changes in the TPU microstructure, crystallization kinetics, and rheological behavior is needed to understand how they are affected by the incorporation of sterically hindered sulfonate species (i.e., the development of a quantitative understanding of the relationships between the sulfonate

133 concentration, the structure of the TPU, and the material properties). Developing this understanding and exploring a wider range of counterions (both hard and soft) will provide a platform that allows for more precise control over the mechanical properties of

TPU ionomers. This work would be readily achievable through facile ion exchange reactions. Overall, the method of plasticization described here is just one possible alternative to conventional plasticizers; however, this system suppresses crystallinity as opposed to suppressing the Tg, which is potentially useful for applications where maintaining the Tg of the material is critical. On the other hand, if the crystallization kinetics and/or crystalline content of the material are essential, this method would not be advised. However, there is potential to utilize a mixture of hard and soft counterions, which could possibly remedy the retardation of the crystallization rate and still achieve the desired softening.193

6.1.3. Kinetically-Controlled Hydrogels

A hydrogel platform that provides a range of substrate elasticities while maintaining the chemical composition of the gel was developed by modifying PEG chain ends to employ variations in cross-linking kinetics that control defect density within the hydrogel network. Rheology and SANS experiments demonstrated how variations in cross-linking kinetics can be used to precisely control the modulus and microstructure of a hydrogel. The data indicated that smaller mesh sizes/correlation lengths were associated with higher moduli gels, whereas low moduli gels corresponded with larger structural defects (Figure 5.3). This relationship was a direct consequence of decreased

134 and increased structural heterogeneities within the hydrogel network, respectively.205-206

More importantly, the data suggested that nanoscale structural details can be influenced by reaction kinetics to induce changes in macroscopic mechanical properties.

In conclusion, it was shown that small changes in the structural features, resulting from variations in the kinetics of cross-linking, are capable of producing mechanically distinguishable hydrogels from chemically identical precursors. For an oxime cross-linking reaction, the kinetics of network formation are influenced by pH and buffer strength, and may be utilized to intrinsically control the degree of heterogeneity within the scaffold microstructure. Further developing and understanding this and other kinetically-driven systems should enable a facile route for precise control over the mechanical properties of hydrogels. Overall, these PEG-based hydrogels could provide definitive insight into the effect of substrate elasticity on human mesenchymal stem cell (hMSC) differentiation because the scaffolds employ the same precursor chemistry, yet exhibit a range of mechanical properties. Future work should aim to encapsulate hMSCs within the hydrogel and evaluate the effects of elasticity on stem cell fate/changes in cell metabolic activity.

135

REFERENCES

1. O'Grady, N. P.; Alexander, M.; Dellinger, E. P.; Gerberding, J. L.; Heard, S. O.; Maki, D. G.; Masur, H.; McCormick, R. D.; Mermel, L. A.; Pearson, M. L.; Raad, I. I.; Randolph, A.; Weinstein, R. A., Guidelines for the Prevention of Intravascular Catheter–Related Infections. Clin. Infect. Dis. 2002, 35 (11), 1281-1307.

2. Pronovost, P.; Needham, D.; Berenholtz, S.; Sinopoli, D.; Chu, H.; Cosgrove, S.; Sexton, B.; Hyzy, R.; Welsh, R.; Roth, G.; Bander, J.; Kepros, J.; Goeschel, C., An Intervention to Decrease Catheter-Related Bloodstream Infections in the ICU. New Engl. J. Med. 2006, 355 (26), 2725-2732.

3. Ngo, B. K. D.; Grunlan, M. A., Protein Resistant Polymeric Biomaterials. ACS Macro Letters 2017, 6 (9), 992-1000.

4. Zander, Z. K.; Becker, M. L., Antimicrobial and Antifouling Strategies for Polymeric Medical Devices. ACS Macro Letters 2018, 7 (1), 16-25.

5. U.S. Food and Drug Administration. 2016 Medical Device Recalls. https://www.fda.gov/MedicalDevices/Safety/ListofRecalls/ucm480134.htm.

6. Chiellini, F.; Ferri, M.; Morelli, A.; Dipaola, L.; Latini, G., Perspectives on alternatives to phthalate plasticized poly(vinyl chloride) in medical devices applications. Prog. Polym. Sci. 2013, 38 (7), 1067-1088.

7. Rogers, J. A.; Metz, L.; Yong, V. W., Review: Endocrine disrupting chemicals and immune responses: A focus on bisphenol-A and its potential mechanisms. Mol. Immunol. 2013, 53 (4), 421-430.

8. Zander, Z. K.; Wang, F.; Becker, M. L.; Weiss, R. A., Ionomers for Tunable Softening of Thermoplastic Polyurethane. Macromolecules 2016, 49 (3), 926-934.

9. Jia, J.; Coyle, R. C.; Richards, D. J.; Berry, C. L.; Barrs, R. W.; Biggs, J.; James Chou, C.; Trusk, T. C.; Mei, Y., Development of peptide-functionalized synthetic hydrogel microarrays for stem cell and tissue engineering applications. Acta Biomater. 2016, 45, 110-120.

10. Lin, F.; Yu, J.; Tang, W.; Zheng, J.; Defante, A.; Guo, K.; Wesdemiotis, C.; Becker, M. L., Peptide-Functionalized Oxime Hydrogels with Tunable Mechanical Properties and Gelation Behavior. Biomacromolecules 2013, 14 (10), 3749-3758. 136

11. Zander, Z. K.; Hua, G.; Wiener, C. G.; Vogt, B. D.; Becker, M. L., Control of Mesh Size and Modulus by Kinetically Dependent Cross-Linking in Hydrogels. Adv. Mater. 2015, 27 (40), 6283-6288.

12. Kojic, E. M.; Darouiche, R. O., Candida Infections of Medical Devices. Clin. Microbiol. Rev. 2004, 17 (2), 255-267.

13. Maki, D. G.; Tambyah, P. A., Engineering out the risk for infection with urinary catheters. Emerging Infect. Dis. 2001, 7 (2), 342-347.

14. Weinstein, R. A.; Darouiche, R. O., Device-Associated Infections: A Macroproblem that Starts with Microadherence. Clin. Infect. Dis. 2001, 33 (9), 1567-1572.

15. Mack, D.; Rohde, H.; Harris, L. G.; Davies, A. P.; Horstkotte, M. A.; Knobloch, J. K., Biofilm formation in medical device-related infection. Int. J. Artif. Organs 2006, 29 (4), 343-59.

16. von Eiff, C.; Jansen, B.; Kohnen, W.; Becker, K., Infections associated with medical devices: pathogenesis, management and prophylaxis. Drugs 2005, 65 (2), 179-214.

17. Donlan, R. M., Biofilms and device-associated infections. Emerging Infect. Dis. 2001, 7 (2), 277-281.

18. Hetrick, E. M.; Schoenfisch, M. H., Reducing implant-related infections: active release strategies. Chem. Soc. Rev. 2006, 35 (9), 780-789.

19. Wu, P.; Grainger, D. W., Drug/device combinations for local drug therapies and infection prophylaxis. Biomaterials 2006, 27 (11), 2450-67.

20. Brooks, B. D.; Brooks, A. E.; Grainger, D. W., Antimicrobial Medical Devices in Preclinical Development and Clinical Use. In Biomaterials Associated Infection: Immunological Aspects and Antimicrobial Strategies, Moriarty, T. F.; Zaat, S. A. J.; Busscher, H. J., Eds. Springer New York: New York, NY, 2013; pp 307-354.

21. Bryers, J. D., Medical biofilms. Biotechnol. Bioeng. 2008, 100 (1), 1-18.

22. Bagge, N.; Schuster, M.; Hentzer, M.; Ciofu, O.; Givskov, M.; Greenberg, E. P.; Hoiby, N., Pseudomonas aeruginosa biofilms exposed to imipenem exhibit changes in global gene expression and beta-lactamase and alginate production. Antimicrob. Agents Chemother. 2004, 48 (4), 1175-87.

23. Kenawy, E.-R.; Worley, S. D.; Broughton, R., The Chemistry and Applications of Antimicrobial Polymers: A State-of-the-Art Review. Biomacromolecules 2007, 8 (5), 1359-1384. 137

24. Kohanski, M. A.; Dwyer, D. J.; Collins, J. J., How antibiotics kill bacteria: from targets to networks. Nat Rev Micro 2010, 8 (6), 423-435.

25. Walsh, C., Molecular mechanisms that confer antibacterial drug resistance. Nature 2000, 406 (6797), 775-781.

26. Cole, S. J.; Records, A. R.; Orr, M. W.; Linden, S. B.; Lee, V. T., Catheter-Associated Urinary Tract Infection by Pseudomonas aeruginosa Is Mediated by Exopolysaccharide-Independent Biofilms. Infect. Immun. 2014, 82 (5), 2048-2058.

27. Halden, R. U., Plastics and Health Risks. Annu. Rev. Public Health 2010, 31 (1), 179- 194.

28. Rahman, M.; Brazel, C. S., The plasticizer market: an assessment of traditional plasticizers and research trends to meet new challenges. Prog. Polym. Sci. 2004, 29 (12), 1223-1248.

29. Weiss, R.; Stamato, H., Development of an ionomer tracer for extruder residence time distribution experiments. Polym. Eng. Sci. 1989, 29 (2), 134-139.

30. Weiss, R. A.; Agarwal, P. K.; Lundberg, R. D., Control of ionic interactions in sulfonated polystyrene ionomers by the use of alkyl-substituted ammonium counterions. J. Appl. Polym. Sci. 1984, 29 (9), 2719-2734.

31. Tant, M. R.; Mauritz, K. A.; Wilkes, G. L., Ionomers: Synthesis, structure, properties and applications. Springer Netherlands: 1997.

32. Bazuin, C. G.; Eisenberg, A., Dynamic mechanical properties of plasticized polystyrene-based ionomers. I. Glassy to rubbery zones. J. Polym. Sci., Part B: Polym. Phys. 1986, 24 (5), 1137-1153.

33. Fitzgerald, J. J.; Weiss, R. A., Synthesis, Properties, and Structure of Sulfonate Ionomers. Journal of Macromolecular Science, Part C 1988, 28 (1), 99-185.

34. Lefelar, J. A.; Weiss, R. A., Concentration and counterion dependence of cluster formation in sulfonated polystyrene. Macromolecules 1984, 17 (6), 1145-1148.

35. Tudryn, G. J.; Liu, W.; Wang, S.-W.; Colby, R. H., Counterion Dynamics in Polyester−Sulfonate Ionomers with Ionic Liquid Counterions. Macromolecules 2011, 44 (9), 3572-3582.

36. Alcantar, N. A.; Aydil, E. S.; Israelachvili, J. N., Polyethylene glycol–coated biocompatible surfaces. J. Biomed. Mater. Res. 2000, 51 (3), 343-351.

138

37. Elisseeff, J.; McIntosh, W.; Anseth, K.; Riley, S.; Ragan, P.; Langer, R., Photoencapsulation of chondrocytes in poly(ethylene oxide)-based semi- interpenetrating networks. J. Biomed. Mater. Res. 2000, 51 (2), 164-171.

38. Jeon, O.; Bouhadir, K. H.; Mansour, J. M.; Alsberg, E., Photocrosslinked alginate hydrogels with tunable biodegradation rates and mechanical properties. Biomaterials 2009, 30 (14), 2724-2734.

39. Lee, J. H.; Lee, H. B.; Andrade, J. D., Blood compatibility of polyethylene oxide surfaces. Prog. Polym. Sci. 1995, 20 (6), 1043-1079.

40. Liu, S. Q.; Rachel Ee, P. L.; Ke, C. Y.; Hedrick, J. L.; Yang, Y. Y., Biodegradable poly(ethylene glycol)–peptide hydrogels with well-defined structure and properties for cell delivery. Biomaterials 2009, 30 (8), 1453-1461.

41. Liu, S. Q.; Tian, Q.; Hedrick, J. L.; Po Hui, J. H.; Rachel Ee, P. L.; Yang, Y. Y., Biomimetic hydrogels for chondrogenic differentiation of human mesenchymal stem cells to neocartilage. Biomaterials 2010, 31 (28), 7298-7307.

42. Bellas, E.; Chen, C. S., Forms, forces, and stem cell fate. Curr. Opin. Cell Biol. 2014, 31, 92-97.

43. Engler, A. J.; Sen, S.; Sweeney, H. L.; Discher, D. E., Matrix Elasticity Directs Stem Cell Lineage Specification. Cell 2006, 126 (4), 677-689.

44. Gilbert, P. M.; Havenstrite, K. L.; Magnusson, K. E. G.; Sacco, A.; Leonardi, N. A.; Kraft, P.; Nguyen, N. K.; Thrun, S.; Lutolf, M. P.; Blau, H. M., Substrate Elasticity Regulates Skeletal Muscle Stem Cell Self-Renewal in Culture. Science 2010, 329 (5995), 1078.

45. Khetan, S.; Guvendiren, M.; Legant, W. R.; Cohen, D. M.; Chen, C. S.; Burdick, J. A., Degradation-mediated cellular traction directs stem cell fate in covalently crosslinked three-dimensional hydrogels. Nature Materials 2013, 12, 458.

46. Lee-Thedieck, C.; Rauch, N.; Fiammengo, R.; Klein, G.; Spatz, J. P., Impact of substrate elasticity on human hematopoietic stem and progenitor cell adhesion and motility. J. Cell Sci. 2012, 125 (16), 3765.

47. Murphy, W. L.; McDevitt, T. C.; Engler, A. J., Materials as stem cell regulators. Nature Materials 2014, 13, 547.

48. Wingate, K.; Bonani, W.; Tan, Y.; Bryant, S. J.; Tan, W., Compressive elasticity of three-dimensional nanofiber matrix directs mesenchymal stem cell differentiation

139

to vascular cells with endothelial or smooth muscle cell markers. Acta Biomater. 2012, 8 (4), 1440-1449.

49. Young, D. A.; Choi, Y. S.; Engler, A. J.; Christman, K. L., Stimulation of adipogenesis of adult adipose-derived stem cells using substrates that mimic the stiffness of adipose tissue. Biomaterials 2013, 34 (34), 8581-8588.

50. Dadsetan, M.; Szatkowski, J. P.; Yaszemski, M. J.; Lu, L., Characterization of Photo- Cross-Linked Oligo[poly(ethylene glycol) fumarate] Hydrogels for Cartilage Tissue Engineering. Biomacromolecules 2007, 8 (5), 1702-1709.

51. Herrick, W. G.; Nguyen, T. V.; Sleiman, M.; McRae, S.; Emrick, T. S.; Peyton, S. R., PEG-Phosphorylcholine Hydrogels As Tunable and Versatile Platforms for Mechanobiology. Biomacromolecules 2013, 14 (7), 2294-2304.

52. Missirlis, D.; Spatz, J. P., Combined Effects of PEG Hydrogel Elasticity and Cell- Adhesive Coating on Fibroblast Adhesion and Persistent Migration. Biomacromolecules 2014, 15 (1), 195-205.

53. Zustiak, S. P.; Leach, J. B., Hydrolytically Degradable Poly(Ethylene Glycol) Hydrogel Scaffolds with Tunable Degradation and Mechanical Properties. Biomacromolecules 2010, 11 (5), 1348-1357.

54. Akagi, Y.; Gong, J. P.; Chung, U.-i.; Sakai, T., Transition between Phantom and Affine Network Model Observed in Polymer Gels with Controlled Network Structure. Macromolecules 2013, 46 (3), 1035-1040.

55. Akagi, Y.; Matsunaga, T.; Shibayama, M.; Chung, U.-i.; Sakai, T., Evaluation of Topological Defects in Tetra-PEG Gels. Macromolecules 2010, 43 (1), 488-493.

56. DeForest, C. A.; Sims, E. A.; Anseth, K. S., Peptide-Functionalized Click Hydrogels with Independently Tunable Mechanics and Chemical Functionality for 3D Cell Culture. Chem. Mater. 2010, 22 (16), 4783-4790.

57. Grube, S.; Oppermann, W., Inhomogeneity in Hydrogels Synthesized by Thiol–Ene Polymerization. Macromolecules 2013, 46 (5), 1948-1955.

58. Lindemann, B.; Schröder, U. P.; Oppermann, W., Influence of the Cross-Linker Reactivity on the Formation of Inhomogeneities in Hydrogels. Macromolecules 1997, 30 (14), 4073-4077.

59. Magill, S. S.; Edwards, J. R.; Bamberg, W.; Beldavs, Z. G.; Dumyati, G.; Kainer, M. A.; Lynfield, R.; Maloney, M.; McAllister-Hollod, L.; Nadle, J.; Ray, S. M.; Thompson,

140

D. L.; Wilson, L. E.; Fridkin, S. K., Multistate Point-Prevalence Survey of Health Care–Associated Infections. New Engl. J. Med. 2014, 370 (13), 1198-1208.

60. Francolini, I.; Donelli, G.; Crisante, F.; Taresco, V.; Piozzi, A., Antimicrobial Polymers for Anti-biofilm Medical Devices: State-of-Art and Perspectives. In Biofilm-based Healthcare-associated Infections: Volume II, Donelli, G., Ed. Springer International Publishing: Cham, 2015; pp 93-117.

61. Trautner, B. W.; Darouiche, R. O., Catheter-associated infections: Pathogenesis affects prevention. Arch. Intern. Med. 2004, 164 (8), 842-850.

62. Rosenthal, V. D.; Maki, D. G.; Graves, N., The International Nosocomial Infection Control Consortium (INICC): Goals and objectives, description of surveillance methods, and operational activities. Am. J. Infect. Control 2008, 36 (9), e1-e12.

63. Umscheid, C. A.; Mitchell, M. D.; Doshi, J. A.; Agarwal, R.; Williams, K.; Brennan, P. J., Estimating the proportion of healthcare-associated infections that are reasonably preventable and the related mortality and costs. Infect. Control Hosp. Epidemiol. 2011, 32 (2), 101-14.

64. U.S. Food and Drug Administration. Medical Devices. https://www.fda.gov/MedicalDevices/default.htm.

65. Kalanuria, A. A.; Zai, W.; Mirski, M., Ventilator-associated pneumonia in the ICU. Crit Care 2014, 18 (2), 208-208.

66. Finkelstein, E. S.; Jekel, J.; Troidle, L.; Gorban-Brennan, N.; Finkelstein, F. O.; Bia, F. J., Patterns of infection in patients maintained on long-term peritoneal dialysis therapy with multiple episodes of peritonitis. Am. J. Kidney Dis. 2002, 39 (6), 1278- 1286.

67. Francolini, I.; Donelli, G., Prevention and control of biofilm-based medical-device- related infections. FEMS Immunol. Med. Microbiol. 2010, 59 (3), 227-238.

68. Donlan, R. M.; Costerton, J. W., Biofilms: Survival Mechanisms of Clinically Relevant Microorganisms. Clin. Microbiol. Rev. 2002, 15 (2), 167-193.

69. Percival, S. L.; Suleman, L.; Vuotto, C.; Donelli, G., Healthcare-associated infections, medical devices and biofilms: risk, tolerance and control. J. Med. Microbiol. 2015, 64 (Pt 4), 323-34.

70. Harding, J. L.; Reynolds, M. M., Combating medical device fouling. Trends Biotechnol. 2014, 32 (3), 140-6.

141

71. Watnick, P.; Kolter, R., Biofilm, City of Microbes. J. Bacteriol. 2000, 182 (10), 2675- 2679.

72. Bryers, J. D.; Ratner, B. D., Bioinspired implant materials befuddle bacteria. ASM News-American Society for Microbiology 2004, 70 (5), 232-232.

73. Donlan, R. M., Biofilm Formation: A Clinically Relevant Microbiological Process. Clin. Infect. Dis. 2001, 33 (8), 1387-1392.

74. Davies, D. G.; Parsek, M. R.; Pearson, J. P.; Iglewski, B. H.; Costerton, J. W.; Greenberg, E. P., The Involvement of Cell-to-Cell Signals in the Development of a Bacterial Biofilm. Science 1998, 280 (5361), 295.

75. Stoodley, P.; Sauer, K.; Davies, D. G.; Costerton, J. W., Biofilms as Complex Differentiated Communities. Annu. Rev. Microbiol. 2002, 56 (1), 187-209.

76. Raad, I., Intravascular-catheter-related infections. Lancet 1998, 351 (9106), 893- 8.

77. Phillips, K. S.; Patwardhan, D.; Jayan, G., Biofilms, medical devices, and antibiofilm technology: key messages from a recent public workshop. Am. J. Infect. Control 2015, 43 (1), 2-3.

78. Tambyah, P. A.; Halvorson, K. T.; Maki, D. G., A Prospective Study of Pathogenesis of Catheter-Associated Urinary Tract Infections. Mayo Clin. Proc. 1999, 74 (2), 131- 136.

79. Herrmann, M.; Vaudaux, P. E.; Pittet, D.; Auckenthaler, R.; Lew, P. D.; Perdreau, F. S.; Peters, G.; Waldvogel, F. A., Fibronectin, Fibrinogen, and Laminin Act as Mediators of Adherence of Clinical Staphylococcal Isolates to Foreign Material. J. Infect. Dis. 1988, 158 (4), 693-701.

80. Raad, II; Luna, M.; Khalil, S. M.; Costerton, J. W.; Lam, C.; Bodey, G. P., The relationship between the thrombotic and infectious complications of central venous catheters. JAMA 1994, 271 (13), 1014-1016.

81. Busscher, H. J.; Weerkamp, A. H., Specific and non-specific interactions in bacterial adhesion to solid substrata. FEMS Microbiol. Lett. 1987, 46 (2), 165-173.

82. Palmer, J.; Flint, S.; Brooks, J., Bacterial cell attachment, the beginning of a biofilm. J. Ind. Microbiol. Biotechnol. 2007, 34 (9), 577-588.

83. Menzies, B. E., The role of fibronectin binding proteins in the pathogenesis of Staphylococcus aureus infections. Curr. Opin. Infect. Dis. 2003, 16 (3), 225-9. 142

84. Jass, J.; Surman, S.; Walker, J. T., Microbial Biofilms in Medicine. In Medical Biofilms, John Wiley & Sons, Ltd: 2005; pp 1-28.

85. Xu, L.-C.; Bauer, J.; Siedlecki, C. A.; Group, A. C. f. t. H. a. B. I. R., Proteins, Platelets, and Blood Coagulation at Biomaterial Interfaces. Colloids Surf., B 2014, 124, 49- 68.

86. Goodman, S. L.; Cooper, S. L.; Albrecht, R. M., Integrin receptors and platelet adhesion to synthetic surfaces. J. Biomed. Mater. Res. 1993, 27 (5), 683-95.

87. Cunha, B. A., Oral Versus IV Treatment for Catheter-related Bloodstream Infections. Emerging Infect. Dis. 2007, 13 (11), 1800-1801.

88. Grainger, D. W.; van der Mei, H. C.; Jutte, P. C.; van den Dungen, J. J. A. M.; Schultz, M. J.; van der Laan, B. F. A. M.; Zaat, S. A. J.; Busscher, H. J., Critical factors in the translation of improved antimicrobial strategies for medical implants and devices. Biomaterials 2013, 34 (37), 9237-9243.

89. Stewart, P. S.; William Costerton, J., Antibiotic resistance of bacteria in biofilms. The Lancet 2001, 358 (9276), 135-138.

90. Busscher, H. J.; van der Mei, H. C.; Subbiahdoss, G.; Jutte, P. C.; van den Dungen, J. J. A. M.; Zaat, S. A. J.; Schultz, M. J.; Grainger, D. W., Biomaterial-Associated Infection: Locating the Finish Line in the Race for the Surface. Sci. Transl. Med. 2012, 4 (153), 153rv10.

91. Piddock, L. J. V., The crisis of no new antibiotics—what is the way forward? Lancet Infect. Dis. 2012, 12 (3), 249-253.

92. Siedenbiedel, F.; Tiller, J. C., Antimicrobial Polymers in Solution and on Surfaces: Overview and Functional Principles. Polymers 2012, 4 (1), 46.

93. Timofeeva, L.; Kleshcheva, N., Antimicrobial polymers: mechanism of action, factors of activity, and applications. Appl. Microbiol. Biotechnol. 2011, 89 (3), 475- 92.

94. Minandri, F.; Bonchi, C.; Frangipani, E.; Imperi, F.; Visca, P., Promises and failures of gallium as an antibacterial agent. Future Microbiol. 2014, 9 (3), 379-397.

95. Kuroda, K.; Caputo, G. A., Antimicrobial polymers as synthetic mimics of host- defense peptides. Wiley Interdiscip. Rev.: Nanomed. Nanobiotechnol. 2013, 5 (1), 49-66.

143

96. Som, A.; Vemparala, S.; Ivanov, I.; Tew, G. N., Synthetic mimics of antimicrobial peptides. Pept. Sci. 2008, 90 (2), 83-93.

97. Takahashi, H.; Caputo, G. A.; Vemparala, S.; Kuroda, K., Synthetic Random Copolymers as a Molecular Platform To Mimic Host-Defense Antimicrobial Peptides. Bioconjugate Chem. 2017, 28 (5), 1340-1350.

98. Bieser, A. M.; Thomann, Y.; Tiller, J. C., Contact-Active Antimicrobial and Potentially Self-Polishing Coatings Based on Cellulose. Macromol. Biosci. 2011, 11 (1), 111-121.

99. Cao, Z.; Mi, L.; Mendiola, J.; Ella-Menye, J.-R.; Zhang, L.; Xue, H.; Jiang, S., Reversibly Switching the Function of a Surface between Attacking and Defending against Bacteria. Angew. Chem. Int. Ed. 2012, 51 (11), 2602-2605.

100. Mi, L.; Jiang, S., Integrated Antimicrobial and Nonfouling Zwitterionic Polymers. Angew. Chem. Int. Ed. 2014, 53 (7), 1746-1754.

101. Vaterrodt, A.; Thallinger, B.; Daumann, K.; Koch, D.; Guebitz, G. M.; Ulbricht, M., Antifouling and Antibacterial Multifunctional Polyzwitterion/Enzyme Coating on Silicone Catheter Material Prepared by Electrostatic Layer-by-Layer Assembly. Langmuir 2016, 32 (5), 1347-1359.

102. Yan, S.; Shi, H.; Song, L.; Wang, X.; Liu, L.; Luan, S.; Yang, Y.; Yin, J., Nonleaching Bacteria-Responsive Antibacterial Surface Based on a Unique Hierarchical Architecture. ACS Appl. Mater. Interfaces 2016, 8 (37), 24471-24481.

103. Jeon, S. I.; Lee, J. H.; Andrade, J. D.; De Gennes, P. G., Protein—surface interactions in the presence of polyethylene oxide. J. Colloid Interface Sci. 1991, 142 (1), 149- 158.

104. Chen, S.; Li, L.; Zhao, C.; Zheng, J., Surface hydration: Principles and applications toward low-fouling/nonfouling biomaterials. Polymer 2010, 51 (23), 5283-5293.

105. Nurioglu, A. G.; Esteves, A. C. C.; de With, G., Non-toxic, non-biocide-release antifouling coatings based on molecular structure design for marine applications. J. Mater. Chem. B 2015, 3 (32), 6547-6570.

106. Ostuni, E.; Chapman, R. G.; Holmlin, R. E.; Takayama, S.; Whitesides, G. M., A Survey of Structure−Property Relationships of Surfaces that Resist the Adsorption of Protein. Langmuir 2001, 17 (18), 5605-5620.

144

107. Browning, M. B.; Cereceres, S. N.; Luong, P. T.; Cosgriff-Hernandez, E. M., Determination of the in vivo degradation mechanism of PEGDA hydrogels. Journal of Biomedical Materials Research Part A 2014, 102 (12), 4244-4251.

108. Herold, D. A.; Keil, K.; Bruns, D. E., Oxidation of polyethylene glycols by alcohol dehydrogenase. Biochem. Pharmacol. 1989, 38 (1), 73-76.

109. Ulbricht, J.; Jordan, R.; Luxenhofer, R., On the biodegradability of polyethylene glycol, polypeptoids and poly(2-oxazoline)s. Biomaterials 2014, 35 (17), 4848- 4861.

110. Chen, S.; Jiang, S., An New Avenue to Nonfouling Materials. Adv. Mater. 2008, 20 (2), 335-338.

111. Konradi, R.; Acikgoz, C.; Textor, M., Polyoxazolines for Nonfouling Surface Coatings — A Direct Comparison to the Gold Standard PEG. Macromol. Rapid Commun. 2012, 33 (19), 1663-1676.

112. Siegers, C.; Biesalski, M.; Haag, R., Self-Assembled Monolayers of Dendritic Polyglycerol Derivatives on Gold That Resist the Adsorption of Proteins. Chemistry – A European Journal 2004, 10 (11), 2831-2838.

113. Weinhart, M.; Becherer, T.; Schnurbusch, N.; Schwibbert, K.; Kunte, H.-J.; Haag, R., Linear and Hyperbranched Polyglycerol Derivatives as Excellent Bioinert Glass Coating Materials. Adv. Eng. Mater. 2011, 13 (12), B501-B510.

114. Krishnamoorthy, M.; Hakobyan, S.; Ramstedt, M.; Gautrot, J. E., Surface-Initiated Polymer Brushes in the Biomedical Field: Applications in Membrane Science, Biosensing, Cell Culture, Regenerative Medicine and Antibacterial Coatings. Chem. Rev. 2014, 114 (21), 10976-11026.

115. Krishnan, S.; Weinman, C. J.; Ober, C. K., Advances in polymers for anti-biofouling surfaces. J. Mater. Chem. 2008, 18 (29), 3405-3413.

116. Smith, R. S.; Zhang, Z.; Bouchard, M.; Li, J.; Lapp, H. S.; Brotske, G. R.; Lucchino, D. L.; Weaver, D.; Roth, L. A.; Coury, A.; Biggerstaff, J.; Sukavaneshvar, S.; Langer, R.; Loose, C., Vascular Catheters with a Nonleaching Poly-Sulfobetaine Surface Modification Reduce Thrombus Formation and Microbial Attachment. Sci. Transl. Med. 2012, 4 (153), 153ra132.

117. Liu, F.; Grainger, D. W., Fluorinated Biomaterials. In Comprehensive Biomaterials II, Ducheyne, P., Ed. Elsevier: Oxford, 2017; pp 648-663.

145

118. Kinnari, T. J.; Jero, J., Experimental and Clinical Experience of Albumin Coating of Tympanostomy Tubes. Otolaryngology - Head and Neck Surgery 2005, 133 (4), 596-600.

119. Kinnari, T. J.; Salonen, E.-M.; Jero, J., Durability of the binding inhibition of albumin coating on tympanostomy tubes. Int. J. Pediatr. Otorhinolaryngol. 2003, 67 (2), 157-164.

120. Andrade, J. D.; Hlady, V., Protein adsorption and materials biocompatibility: A tutorial review and suggested hypotheses. In Biopolymers/Non-Exclusion HPLC, Springer Berlin Heidelberg: Berlin, Heidelberg, 1986; pp 1-63.

121. Leonard, E. F.; Vroman, L., Is the Vroman effect of importance in the interaction of blood with artificial materials? J. Biomater. Sci. Polym. Ed. 1992, 3 (1), 95-107.

122. Begovac, P. C.; Thomson, R. C.; Fisher, J. L.; Hughson, A.; Gällhagen, A., Improvements in GORE-TEX® vascular graft performance by Carmeda® bioactive surface heparin immobilization. Eur. J. Vasc. Endovasc. Surg. 25 (5), 432-437.

123. Lindholt, J. S.; Gottschalksen, B.; Johannesen, N.; Dueholm, D.; Ravn, H.; Christensen, E. D.; Viddal, B.; Flørenes, T.; Pedersen, G.; Rasmussen, M.; Carstensen, M.; Grøndal, N.; Fasting, H., The Scandinavian Propaten® Trial - 1-Year Patency of PTFE Vascular Prostheses with Heparin-Bonded Luminal Surfaces Compared to Ordinary Pure PTFE Vascular Prostheses - A Randomised Clinical Controlled Multi-centre Trial. Eur. J. Vasc. Endovasc. Surg. 41 (5), 668-673.

124. Nomura, S.; Lundberg, F.; Stollenwerk, M.; Nakamura, K.; Ljungh, Å., Adhesion of staphylococci to polymers with and without immobilized heparin in cerebrospinal fluid. J. Biomed. Mater. Res. 1997, 38 (1), 35-42.

125. Portolés, M.; Refojo, M. F.; Leong, F.-L., Reduced bacterial adhesion to heparin- surface-modified intraocular lenses. J. Cataract Refract. Surg. 1993, 19 (6), 755- 759.

126. Ruggieri, M. R.; Hanno, P. M.; Levin, R. M., Reduction of Bacterial Adherence to Catheter Surface with Heparin. The Journal of Urology 1987, 138 (2), 423-426.

127. Han, D. K.; Park, K. D.; Ryu, G. H.; Kim, U. Y.; Min, B. G.; Kim, Y. H., Plasma protein adsorption to sulfonated poly(ethylene oxide)-grafted polyurethane surface. J. Biomed. Mater. Res. 1996, 30 (1), 23-30.

128. Park, K. D.; Kim, Y. S.; Han, D. K.; Kim, Y. H.; Lee, E. H. B.; Suh, H.; Choi, K. S., Bacterial adhesion on PEG modified polyurethane surfaces. Biomaterials 1998, 19 (7), 851-859. 146

129. Razatos, A.; Ong, Y.-L.; Sharma, M. M.; Georgiou, G., Molecular determinants of bacterial adhesion monitored by atomic force microscopy. Proc. Natl. Acad. Sci. U. S. A. 1998, 95 (19), 11059-11064.

130. Arciola, C. R.; Bustanji, Y.; Conti, M.; Campoccia, D.; Baldassarri, L.; Samorı,̀ B.; Montanaro, L., Staphylococcus epidermidis–fibronectin binding and its inhibition by heparin. Biomaterials 2003, 24 (18), 3013-3019.

131. Weber, N.; Wendel, H. P.; Ziemer, G., Hemocompatibility of heparin-coated surfaces and the role of selective plasma protein adsorption. Biomaterials 2002, 23 (2), 429-439.

132. Kubiak-Ossowska, K.; Jachimska, B.; Mulheran, P. A., How Negatively Charged Proteins Adsorb to Negatively Charged Surfaces: A Molecular Dynamics Study of BSA Adsorption on Silica. The Journal of Physical Chemistry B 2016, 120 (40), 10463-10468.

133. Kubiak-Ossowska, K.; Tokarczyk, K.; Jachimska, B.; Mulheran, P. A., Bovine Serum Albumin Adsorption at a Silica Surface Explored by Simulation and Experiment. The Journal of Physical Chemistry B 2017, 121 (16), 3975-3986.

134. Michael, K. E.; Vernekar, V. N.; Keselowsky, B. G.; Meredith, J. C.; Latour, R. A.; García, A. J., Adsorption-Induced Conformational Changes in Fibronectin Due to Interactions with Well-Defined Surface Chemistries. Langmuir 2003, 19 (19), 8033- 8040.

135. Callow, M. E.; Fletcher, R. L., Special Issue Marine Biofouling and CorrosionThe influence of low surface energy materials on bioadhesion — a review. Int. Biodeterior. Biodegrad. 1994, 34 (3), 333-348.

136. Colas, A.; Curtis, J., Silicone biomaterials: history and chemistry. Biomaterials science: an introduction to materials in medicine 2004, 2, 80-85.

137. Owen, M. J., Fluorosilicones. In Advances in Silicones and Silicone-Modified Materials, American Chemical Society: 2010; Vol. 1051, pp 99-108.

138. Zhang, H.; Chiao, M., Anti-fouling Coatings of Poly(dimethylsiloxane) Devices for Biological and Biomedical Applications. Journal of Medical and Biological Engineering 2015, 35 (2), 143-155.

139. Godek, M. L.; Michel, R.; Chamberlain, L. M.; Castner, D. G.; Grainger, D. W., Adsorbed serum albumin is permissive to macrophage attachment to perfluorocarbon polymer surfaces in culture. Journal of Biomedical Materials Research Part A 2009, 88A (2), 503-519. 147

140. Zhang, Y.-L.; Xia, H.; Kim, E.; Sun, H.-B., Recent developments in superhydrophobic surfaces with unique structural and functional properties. Soft Matter 2012, 8 (44), 11217-11231.

141. Kesel, A.; Liedert, R. Antifouling coating. Int. Pat. Appl. WO2008025538A1, Mar. 6, 2008.

142. Chung, K. K.; Schumacher, J. F.; Sampson, E. M.; Burne, R. A.; Antonelli, P. J.; Brennan, A. B., Impact of engineered surface microtopography on biofilm formation of Staphylococcus aureus. Biointerphases 2007, 2 (2), 89-94.

143. Reddy, S. T.; Chung, K. K.; McDaniel, C. J.; Darouiche, R. O.; Landman, J.; Brennan, A. B., Micropatterned surfaces for reducing the risk of catheter-associated urinary tract infection: an in vitro study on the effect of sharklet micropatterned surfaces to inhibit bacterial colonization and migration of uropathogenic Escherichia coli. J. Endourol. 2011, 25 (9), 1547-52.

144. Cooper, S. L.; Guan, J., Advances in Polyurethane Biomaterials. Woodhead Publishing: 2016.

145. Raad, I.; Darouiche, R.; Hachem, R.; Sacilowski, M.; Bodey, G. P., Antibiotics and prevention of microbial colonization of catheters. Antimicrob. Agents Chemother. 1995, 39 (11), 2397-400.

146. Liu, J.-L.; DeMars, B. J. Methods of manufacturing drug-loaded substrates. U.S. Patent 8,673,388 B2, Mar. 18, 2014.

147. Tiller, J. C.; Liao, C.-J.; Lewis, K.; Klibanov, A. M., Designing surfaces that kill bacteria on contact. Proc. Natl. Acad. Sci. U. S. A. 2001, 98 (11), 5981-5985.

148. Wang, S.; Ward, R. S.; Tian, Y.; Li, L.; McCrea, K.; Parakka, J.; Jones Jr, R. L. Antimicrobial polymers and their uses. U.S. Pat. Appl. 2011/0124772 A1, May 26, 2011.

149. Dias, A. J. A. A.; Hensen, G. J. E.; Belt, J. W.; Rooijmans, M.; De Bont, N. H. M.; Currie, E. P. K. Hydrophilic coating comprising a polyelectrolyte. U.S. Patent 8,512,795 B2, Aug. 20, 2013.

150. Fan, Y.-L.; Marlin, L.; Sahatjian, R. A.; Schultz, S. A. Medical device with lubricious coating. U.S. Patent 5,509,899, Apr. 23, 1996.

151. Opolski, M. P. Medical apparatus having protective, lubricious coating. U.S. Patent 5,272,012, Dec. 21, 1993, 1993.

148

152. Stevens, K. N. J.; Croes, S.; Boersma, R. S.; Stobberingh, E. E.; van der Marel, C.; van der Veen, F. H.; Knetsch, M. L. W.; Koole, L. H., Hydrophilic surface coatings with embedded biocidal silver nanoparticles and sodium heparin for central venous catheters. Biomaterials 2011, 32 (5), 1264-1269.

153. Begovac, P. C.; Thomson, R. C.; Fisher, J. L.; Hughson, A.; Gällhagen, A., Improvements in GORE-TEX® vascular graft performance by Carmeda® bioactive surface heparin immobilization. Eur. J. Vasc. Endovasc. Surg. 2003, 25 (5), 432-437.

154. Verhoeven, M.; Cahalan, L. L.; Hendriks, M.; Fouache, B.; Cahalan, P. T. Method for making heparinized biomaterials. U.S. Patent 5,679,659, Oct. 21, 1997.

155. Zou, P.; Hartleb, W.; Lienkamp, K., It takes walls and knights to defend a castle - synthesis of surface coatings from antimicrobial and antibiofouling polymers. J. Mater. Chem. 2012, 22 (37), 19579-19589.

156. Glinel, K.; Jonas, A. M.; Jouenne, T.; Leprince, J.; Galas, L.; Huck, W. T. S., Antibacterial and Antifouling Polymer Brushes Incorporating Antimicrobial Peptide. Bioconjugate Chem. 2009, 20 (1), 71-77.

157. Jiang, J.; Zhu, L.; Zhu, L.; Zhang, H.; Zhu, B.; Xu, Y., Antifouling and Antimicrobial Polymer Membranes Based on Bioinspired Polydopamine and Strong Hydrogen- Bonded Poly(N-vinyl pyrrolidone). ACS Appl. Mater. Interfaces 2013, 5 (24), 12895- 12904.

158. Liu, S. Q.; Yang, C.; Huang, Y.; Ding, X.; Li, Y.; Fan, W. M.; Hedrick, J. L.; Yang, Y.-Y., Antimicrobial and Antifouling Hydrogels Formed In Situ from Polycarbonate and Poly(ethylene glycol) via Michael Addition. Adv. Mater. 2012, 24 (48), 6484-6489.

159. Zhi, Z.; Su, Y.; Xi, Y.; Tian, L.; Xu, M.; Wang, Q.; Padidan, S.; Li, P.; Huang, W., Dual- Functional Polyethylene Glycol-b-polyhexanide Surface Coating with in Vitro and in Vivo Antimicrobial and Antifouling Activities. ACS Appl. Mater. Interfaces 2017, 9 (12), 10383-10397.

160. Chan, C. M.; Ko, T. M.; Hiraoka, H., Polymer surface modification by plasmas and photons. Surf. Sci. Rep. 1996, 24 (1), 1-54.

161. Goddard, J. M.; Hotchkiss, J. H., Polymer surface modification for the attachment of bioactive compounds. Prog. Polym. Sci. 2007, 32 (7), 698-725.

162. Mittal, K. L., Polymer surface modification: relevance to adhesion. CRC Press: 2004; Vol. 3.

149

163. Lin, F.; Zheng, J.; Yu, J.; Zhou, J.; Becker, M. L., Cascading “Triclick” Functionalization of Poly(caprolactone) Thin Films Quantified via a Quartz Crystal Microbalance. Biomacromolecules 2013, 14 (8), 2857-2865.

164. Nebhani, L.; Barner-Kowollik, C., Orthogonal Transformations on Solid Substrates: Efficient Avenues to Surface Modification. Adv. Mater. 2009, 21 (34), 3442-3468.

165. Schenk, F. C.; Boehm, H.; Spatz, J. P.; Wegner, S. V., Dual-Functionalized Nanostructured Biointerfaces by Click Chemistry. Langmuir 2014, 30 (23), 6897- 6905.

166. Zheng, J.; Hua, G.; Yu, J.; Lin, F.; Wade, M. B.; Reneker, D. H.; Becker, M. L., Post- Electrospinning “Triclick” Functionalization of Degradable Polymer Nanofibers. ACS Macro Letters 2015, 4 (2), 207-213.

167. Zheng, J.; Kontoveros, D.; Lin, F.; Hua, G.; Reneker, D. H.; Becker, M. L.; Willits, R. K., Enhanced Schwann Cell Attachment and Alignment Using One-Pot “Dual Click” GRGDS and YIGSR Derivatized Nanofibers. Biomacromolecules 2015, 16 (1), 357- 363.

168. Tang, W.; Becker, M. L., "Click" reactions: a versatile toolbox for the synthesis of peptide-conjugates. Chem. Soc. Rev. 2014, 43 (20), 7013-7039.

169. Azagarsamy, M. A.; Anseth, K. S., Bioorthogonal Click Chemistry: An Indispensable Tool to Create Multifaceted Cell Culture Scaffolds. ACS Macro Letters 2013, 2 (1), 5-9.

170. Moses, J. E.; Moorhouse, A. D., The growing applications of click chemistry. Chem. Soc. Rev. 2007, 36 (8), 1249-1262.

171. Nguyen, T.; Francis, M. B., Practical Synthetic Route to Functionalized Rhodamine Dyes. Org. Lett. 2003, 5 (18), 3245-3248.

172. Majima, T.; Schnabel, W.; Weber, W., Phenyl-2,4,6-trimethylbenzoylphosphinates as water-soluble photoinitiators. Generation and reactivity of O=Ṗ(C6H5)(O−) radical anions. Die Makromolekulare Chemie 1991, 192 (10), 2307-2315.

173. International Organization for Standardization. Measurement of antibacterial activity on plastics and other non-porous surfaces (ISO 22196:2011). https://www.iso.org/home.html.

174. Fairbanks, B. D.; Schwartz, M. P.; Bowman, C. N.; Anseth, K. S., Photoinitiated polymerization of PEG-diacrylate with lithium phenyl-2,4,6-

150

trimethylbenzoylphosphinate: polymerization rate and cytocompatibility. Biomaterials 2009, 30 (35), 6702-6707.

175. Kwei, T. K., Phase separation in segmented polyurethanes. J. Appl. Polym. Sci. 1982, 27 (8), 2891-2899.

176. Lee, H. S.; Wang, Y. K.; Hsu, S. L., Spectroscopic analysis of phase separation behavior of model polyurethanes. Macromolecules 1987, 20 (9), 2089-2095.

177. Cadogan, D. F.; Howick, C. J., Plasticizers. In Kirk-Othmer Encyclopedia of Chemical Technology, John Wiley & Sons, Inc.: 2000.

178. Clarke, M. L.; Wang, J.; Chen, Z., Sum Frequency Generation Studies on the Surface Structures of Plasticized and Unplasticized Polyurethane in Air and in Water. Anal. Chem. 2003, 75 (14), 3275-3280.

179. Horn, O.; Nalli, S.; Cooper, D.; Nicell, J., Plasticizer metabolites in the environment. Water Res. 2004, 38 (17), 3693-3698.

180. Jaeger, R. J.; Rubin, R. J., Plasticizers from Plastic Devices: Extraction, Metabolism, and Accumulation by Biological Systems. Science 1970, 170 (3956), 460-462.

181. Kovacic, P., How dangerous are phthalate plasticizers? Integrated approach to toxicity based on metabolism, electron transfer, reactive oxygen species and cell signaling. Med. Hypotheses 2010, 74 (4), 626-628.

182. Vieira, M. G. A.; da Silva, M. A.; dos Santos, L. O.; Beppu, M. M., Natural-based plasticizers and biopolymer films: A review. Eur. Polym. J. 2011, 47 (3), 254-263.

183. Zhou, X.-M.; Lü, W.-J.; Su, L.; Shan, Z.-J.; Chen, X.-G., Binding of Phthalate Plasticizers to Human Serum Albumin in Vitro: A Multispectroscopic Approach and Molecular Modeling. J. Agric. Food. Chem. 2012, 60 (4), 1135-1145.

184. Teramoto, Y.; Nishio, Y., Biodegradable Cellulose Diacetate-graft-poly(l-lactide)s: Thermal Treatment Effect on the Development of Supramolecular Structures. Biomacromolecules 2004, 5 (2), 397-406.

185. Teramoto, Y.; Nishio, Y., Biodegradable Cellulose Diacetate-graft-poly(l-lactide)s: Enzymatic Hydrolysis Behavior and Surface Morphological Characterization. Biomacromolecules 2004, 5 (2), 407-414.

186. Vázquez-Torres, H.; Cruz-Ramos, C. A., Blends of cellulosic esters with poly(caprolactone): Characterization by DSC, DMA, and WAXS. J. Appl. Polym. Sci. 1994, 54 (8), 1141-1159. 151

187. Król, P., Synthesis methods, chemical structures and phase structures of linear polyurethanes. Properties and applications of linear polyurethanes in polyurethane elastomers, copolymers and ionomers. Prog. Mater Sci. 2007, 52 (6), 915-1015.

188. Flory, P. J., Thermodynamics of Crystallization in High Polymers II. Simplified Derivation of Melting‐Point Relationships. The Journal of Chemical Physics 1947, 15 (9), 684-684.

189. Flory, P. J., Thermodynamics of Crystallization in High Polymers. IV. A Theory of Crystalline States and Fusion in Polymers, Copolymers, and Their Mixtures with Diluents. The Journal of Chemical Physics 1949, 17 (3), 223-240.

190. Gorda, K. R.; Peiffer, D. G., Properties of sulfonated poly(butylene terephthalate). J. Polym. Sci., Part B: Polym. Phys. 1992, 30 (3), 281-292.

191. Orler, E. B.; Moore, R. B., Influence of Ionic Interactions on the Crystallization of Lightly Sulfonated Syndiotactic Polystyrene Ionomers. Macromolecules 1994, 27 (17), 4774-4780.

192. Sauer, B. B.; McLean, R. S., AFM and X-ray Studies of Crystal and Ionic Domain Morphology in Poly(ethylene-co-methacrylic acid) Ionomers. Macromolecules 2000, 33 (21), 7939-7949.

193. Orler, E. B.; Calhoun, B. H.; Moore, R. B., Crystallization Kinetics as a Probe of the Dynamic Network in Lightly Sulfonated Syndiotactic Polystyrene Ionomers. Macromolecules 1996, 29 (18), 5965-5971.

194. Slaughter, B. V.; Khurshid, S. S.; Fisher, O. Z.; Khademhosseini, A.; Peppas, N. A., Hydrogels in Regenerative Medicine. Adv. Mater. 2009, 21 (32-33), 3307-3329.

195. Grover, G. N.; Braden, R. L.; Christman, K. L., Oxime Cross-Linked Injectable Hydrogels for Catheter Delivery. Adv. Mater. 2013, 25 (21), 2937-2942.

196. Grover, G. N.; Lam, J.; Nguyen, T. H.; Segura, T.; Maynard, H. D., Biocompatible Hydrogels by Oxime Click Chemistry. Biomacromolecules 2012, 13 (10), 3013- 3017.

197. Lutolf, M. P.; Hubbell, J. A., Synthesis and Physicochemical Characterization of End-Linked Poly(ethylene glycol)-co-peptide Hydrogels Formed by Michael-Type Addition. Biomacromolecules 2003, 4 (3), 713-722.

152

198. Stempel, G. H.; Schaffel, G. S., A Comparative Study of the Kinetics and Mechanisms of Formation of the Phenylhydrazone, Semicarbazone and Oxime of d-Carvone1,2. J. Am. Chem. Soc. 1944, 66 (7), 1158-1161.

199. Clayden, J.; Greeves, N.; Warren, S., Organic Chemistry. OUP Oxford: 2012.

200. Jencks, W. P., Studies on the Mechanism of Oxime and Semicarbazone Formation1. J. Am. Chem. Soc. 1959, 81 (2), 475-481.

201. Flory, P. J., Principles of Polymer Chemistry. Cornell University Press: 1953.

202. Glinka, C. J.; Barker, J. G.; Hammouda, B.; Krueger, S.; Moyer, J. J.; Orts, W. J., The 30 m Small-Angle Neutron Scattering Instruments at the National Institute of Standards and Technology. J. Appl. Cryst. 1998, 31 (3), 430-445.

203. Kline, S., Reduction and analysis of SANS and USANS data using IGOR Pro. J. Appl. Cryst. 2006, 39 (6), 895-900.

204. Hammouda, B. Probing Nanoscale Structures - The SANS Toolbox. http://www.ncnr.nist.gov/staff/hammouda/the_SANS_toolbox.pdf (accessed February).

205. Neuburger, N. A.; Eichinger, B. E., Critical experimental test of the Flory-Rehner theory of swelling. Macromolecules 1988, 21 (10), 3060-3070.

206. Patel, S. K.; Malone, S.; Cohen, C.; Gillmor, J. R.; Colby, R. H., Elastic modulus and equilibrium swelling of poly(dimethylsiloxane) networks. Macromolecules 1992, 25 (20), 5241-5251.

153

APPENDICES

154

APPENDIX A.

SUPPORTING FIGURES

Figure 6.1. 1H-NMR spectrum of Q14-OH demonstrates a 1:1 molar ratio of peaks e and g, which indicates the formation of the desired compound.

155

Figure 6.2. 1H-NMR spectrum of Q12-OH demonstrates a 1:1 molar ratio of peaks e and g, which indicates the formation of the desired compound.

156

Figure 6.3. 1H-NMR spectrum of Q8-OH demonstrates a 1:1 molar ratio of peaks e and g, which indicates the formation of the desired compound.

157

Figure 6.4. 1H-NMR spectrum of 3,3’-dithiodipropanoyl chloride displays two triplets which confirms the purity of the compound, and demonstrates quantitative conversion to the acid chloride.

158

Figure 6.5. 13C-NMR spectrum of 3,3’-dithiodipropanoyl chloride confirms the purity of the compound, and demonstrates quantitative conversion to the acid chloride.

159

Figure 6.6. 1H-NMR spectrum of Q14-S-S shows the appearance of peaks d and e, which are equimolar to peaks c, f, and g from the corresponding Q14-OH, indicating complete conversion to the desired disulfide.

160

Figure 6.7. 1H-NMR spectrum of Q12-S-S shows the appearance of peaks d and e, which are equimolar to peaks c, f, and g from the corresponding Q12-OH, indicating complete conversion to the desired disulfide.

161

Figure 6.8. 1H-NMR spectrum of Q8-S-S shows the appearance of peaks d and e, which are equimolar to peaks c, f, and g from the corresponding Q8-OH, indicating complete conversion to the desired disulfide.

162

Figure 6.9. 1H-NMR spectrum of Q14-SH shows the proton resonances α and ß to the carbonyl (peak d) converge, and are equimolar to peaks c, e, and f from the corresponding Q14-S-S, indicating complete conversion to the desired thiol.

163

Figure 6.10. 1H-NMR spectrum of Q12-SH shows the proton resonances α and ß to the carbonyl (peak d) converge, and are equimolar to peaks c, e, and f from the corresponding Q12-S-S, indicating complete conversion to the desired thiol.

164

Figure 6.11. 1H-NMR spectrum of Q8-SH shows the proton resonances α and ß to the carbonyl (peak d) converge, and are equimolar to peaks c, e, and f from the corresponding Q8-S-S, indicating complete conversion to the desired thiol.

165

Figure 6.12. 13C-NMR spectra overlay of Q14-S-S and Q14-SH shows the shifting of the carbon α to the carbonyl downfield and the ß carbon upfield, indicating conversion to the desired thiol.

166

Figure 6.13. 13C-NMR spectra overlay of Q12-S-S and Q12-SH shows the shifting of the carbon α to the carbonyl downfield and the ß carbon upfield, indicating conversion to the desired thiol.

167

Figure 6.14. 13C-NMR spectra overlay of Q8-S-S and Q8-SH shows the shifting of the carbon α to the carbonyl downfield and the ß carbon upfield, indicating conversion to the desired thiol.

168

Figure 6.15. 1H-NMR spectrum of rhodamine B base.

169

Figure 6.16. 1H-NMR spectrum of rhodamine B piperazine amide shows the introduction of peaks b, d, and i from piperazine, and the integration of peaks a and b indicate the amidation reaction was successful. HDO overlaps with peaks c and d.

Figure 6.17. ESI-MS of rhodamine B piperazine amide shows the molecular ion [M]+ = 511.3 Da (calculated = 511.31 Da), as well as the doubly charged ion, [M]2+ at m/z = 255.7.

170

Figure 6.18. 1H-NMR spectrum of rhodamine B 4-(3-hydroxylpropyl) piperazine amide demonstrates the appearance of peaks b and d, as well as the upfield shifting of the piperazine proton resonances (peaks e and c). The proton integrations indicate successful substitution.

Figure 6.19. ESI-MS of rhodamine B 4-(3-hydroxylpropyl) piperazine amide shows the molecular ion [M]+ = 569.4 Da (calculated = 569.35 Da).

171

Figure 6.20. 1H-NMR spectrum of rhodamine B disulfide. The shifting of peak i downfield, as well as the emergence and integration of peaks e and f indicate the esterification was successful.

Figure 6.21. ESI-MS of rhodamine B disulfide shows the doubly charged ion [M]2+ = 656.4 m/z which is 1312.8 Da (calculated = 1312.68 Da), as well as the triply charged ion, [M]3+ at m/z = 438.0. The peaks 569.4 and 775.4 are singly charged ions produced in the mass spectrometer by methanolysis, resulting in rhodamine B 4-(3-hydroxylpropyl) piperazine amide (569.4 Da) and the corresponding methyl ester (775.4 Da) cleavage products.

172

Figure 6.22. 1H-NMR spectrum of rhodamine-SH shows the proton resonances α and ß to the carbonyl (peaks e and f, respectively) converge, and are equimolar to all other peaks from the corresponding rhodamine B disulfide, indicating complete conversion to the desired thiol.

Figure 6.23. ESI-MS of rhodamine-SH shows the molecular ion [M]+ = 657.4 Da (calculated = 657.35 Da) with minimal impurities. The peak at 569.4 is the rhodamine B 4-(3- hydroxylpropyl) piperazine amide fragment resulting from ester cleavage.

173

Figure 6.24. 13C-NMR spectra overlay of rhodamine B disulfide and rhodamine-SH shows the shifting of the carbon α to the carbonyl downfield and the ß carbon upfield, indicating conversion to the desired thiol.

Figure 6.25. Normalized absorbance and emission spectra for rhodamine-SH. The λabs = 568 nm and the λem = 592 nm.

174

Figure 6.26. Fluorescence standard curve for various concentrations of rhodamine-SH in DMSO. The linear fit yielded an R2 value = 0.99 with a slope of (168.7 ± 0.1)×109 M-1.

Figure 6.27. 1H-NMR spectrum of LAP. The integrations of peaks a – c compared to peaks d – f confirm a 1:1 substitution occurred.

175

Figure 6.28. (A) UV-vis absorption spectra for LAP photoinitiator at several concentrations and (B) the absorbance at λ = 365 nm vs. concentration for determination of the molar absorptivity (ε) of LAP. The linear fit yielded an R2 value = 0.99 with a slope (ε) = 179 ± 3 M-1∙cm-1.

Figure 6.29. 1H-NMR spectrum of a 30 wt.% (50 mol%) HMDI control TPU. The proton integrations confirm the molar composition of HMDI:Acrol-E351:BDO is approximately 0.5:0.1:0.4.

176

Figure 6.30. 1H-NMR spectrum of 8% alloc-TPU shows the appearance of peaks “n” and “o” relative to the control TPU, which correspond to the protons from the allyl functional group. The integrations indicate an incorporation of ca. 8 mol%, wherein the overall molar composition of HMDI:Acrol-E351:BDO:allyl is approximately 0.5:0.1:0.32:0.08.

177

Figure 6.31. 1H-NMR overlay of various batches of 8% allyl-TPU normalized to the propylene peak of the polyether soft segment with an inlay displaying the allylic proton resonances (δ = 5.22 and 5.85 ppm).

Figure 6.32. SEC molecular weight plot for control TPU and 8% allyl-TPU batches.

178

Figure 6.33. (A) Thermogravimetric analysis (TGA) was performed to determine the degradation profile for control and 8% allyl-TPU. (B) Expanded region showing the onset degradation temperature of 8% allyl-TPU is ca. 10 °C lower than the control.

Figure 6.34. (A) Differential scanning calorimetry (DCS) first heating cycle and (B) cooling cycle. Exothermic behavior is up in these scans, and the curves have been vertically displaced for clarity. The first heating scans reveal a main melting transition near 120 °C for both the control and allyl-TPU. The cooling scans show that the TPUs have a Tg between -60 and -65 °C, but did not exhibit any significant crystallization on this time scale.

179

Figure 6.35. High resolution N1s XPS spectra of inner lumen of (A) phys. ads. and (B) UV treated catheter tubing (longitudinal sections). The raw data is interpolated with a cubic + b-spline (solid lines) and the individual fits of the N, NR4 , and total fit are represented by dashed lines. The phys. ads. control did not exhibit QACs on the surface, while the UV + treated sample contained 14.4% NR4 relative to urethane N following modification with Q8-SH.

Figure 6.36. (A) Diagram of the biofilm formation test, displaying the ordering of catheter segments (CC = Cook® Beacon® Tip Torcon NB® Advantage Catheter segments). (B) The eluent from biofilm formation testing 48 h post-inoculation was spread plated and displayed only P. aeruginosa colonies.

180

Figure 6.37. 1H-NMR Spectrum for TDA-BES. Integration confirms 1:1 substitution and ionic monomer purity. Peaks 4 and 5 overlap which each other, as well as DMSO-d6.

181

Figure 6.38. 1H-NMR Spectrum for DDA-BES. Integration confirms 1:1 substitution and ionic monomer purity.

182

Figure 6.39. 1H-NMR Spectrum for THA-BES. Integration confirms 1:1 substitution and ionic monomer purity.

Figure 6.40. ESI-MS was utilized to confirm the purity of the ionic diol (THA-BES). (a) In negative mode, the absence of bromine in the spectrum shows the ion exchange was efficient, and the mass of the BES anion (212.0 Da) was detected. (b) In addition, positive mode shows the mass of the tetrahexylammonium counterion (354.5 Da).

183

Figure 6.41. FT-IR spectra of ionomer product obtained after 2 and 3 h reaction time. Disappearance of the isocyanate (NCO) absorption at 2270 cm-1 would correspond to complete reaction. The expanded spectrum showing the spectral region 2000 – 2500 cm- 1 shows that after 3 h the reaction is essentially complete. The red arrow identifies the residual signal for the NCO resonance, which is comparable to the noise in the spectrum.

184

Figure 6.42. 1H-NMR spectrum for 30 wt% hard segment control TPU with PE diol. The peaks labeled 2 and 5 are from BDO, 4 is the methylene of MDI, 7 and 8 are the aromatic protons from MDI, 9 are the amine protons in urethane linkages, and the remainder belong to the PE diol. The ratio of peak 4:3 provided an accurate calculation of the hard segment content.

185

Figure 6.43. 1H-NMR spectra overlay for various wt% control TPUs with PE diol. The intensity was normalized to the leading aliphatic peak (δ = 1.54) and a notable decrease in the intensity of hard segment components (BDO, MDI-methylene, and MDI-aromatic) was observed with decreasing MDI and BDO feed ratio. Each spectra was integrated separately to calculate the composition.

186

Figure 6.44. 1H-NMR spectrum for TPU30(PC)-3.8DD. The peaks labeled 1 and 5 correspond to the DDA-BES (-CH3) protons on the end of the aliphatic chains and the methyl groups, respectively. Peaks 4, 5, and 6 were used for determining the hard segment and ionic monomer content.

187

Figure 6.45. 1H-NMR spectra overlay for various compositions of DDA-BES TPU ionomers with PC diol. The intensity was normalized to the PC diol triplet (δ = 2.26) and a gradual decrease in the intensity of the DDA methyl peak was observed with decreasing ionic monomer feed ratio. Each spectrum was integrated separately to calculate the composition.

188

Figure 6.46. 1H-NMR spectrum for TPU30(PE)-4.7TH. The peaks labeled 1 and 2 correspond to the protons on the aliphatic chains (2) and chain ends (1) for THA-BES. Peaks 2, 5, and 6 were used for determining the hard segment and ionic monomer content.

189

Figure 6.47. 1H-NMR spectrum for TPU30(PE)-7.6TD. The peaks labeled 1 and 2 correspond to the protons on the aliphatic chains (2) and chain ends (1) for TDA-BES. Peaks 2, 5, and 6 were used for determining the hard segment and ionic monomer content.

190

Figure 6.48. SAXS data for TPU30(PE) control (■), TPU30(PE)-4.1TD (●) and TPU30(PE)- 휃 4휋푠푖푛( ) 4.7TH (▲). The scattering vector was defined as 푞 = 2 , where λ is the wavelength 휆 of the scattering radiation (λCuKa = 0.154 nm) and θ is the scattering angle. The do was calculated from the Bragg spacing associated with the peak in scattering intensity, i.e., 2휋 푑표 = , where qmax is the scattering vector value at the maximum intensity of the 푞푚푎푥 peak.

191

Figure 6.49. The boc-protected, 4-arm aminooxy crosslinker 1H-NMR spectrum shows complete substitution of the (boc-aminooxy)acetic acid to pentaerythritol; the ratio of peaks b and c are 1:1, indicating quantitative conversion.

192

Figure 6.50. The 1H-NMR spectrum demonstrates successful deprotection of the boc- protecting group by the disappearance of the peak at 1.48 ppm, and provides the spectrum for the final 4-arm aminooxy crosslinker product.

193

Figure 6.51. The ESI-MS reveals the molecular ion peak [M+H]+ at 429.1 Da (calculated 429.15 Da), as well as higher molecular weight species corresponding to sodiated molecular ions [M+Na]+ and hydration complexes, which are expected with a hygroscopic material.

194

Figure 6.52. The 1H-NMR spectrum for keto-PEG shows a major peak (e) resulting from the ethylene glycol repeat unit. The integrations confirm tetra-substitution of levulinic acid to keto-PEG. MALDI-MS was performed for further evidence of complete conversion.

195

Figure 6.53. 1H-NMR spectrum expanded region (1.5 – 4.5 ppm) for keto-PEG reveals the end-group peaks, and demonstrates successful substitution of the levulinic acid to the 4- arm PEG by proton integrations.

196

Figure 6.54. The MALDI-MS spectrum shows that a single distribution for the keto-PEG exists, confirming only tetra-substitution occurred with no minor species/series.

Figure 6.55. Expanded MALDI-MS spectrum (10,360 – 10,480 Da) indicates a repeat unit of 44 Da which corresponds to PEG. End group analysis was performed to confirm tetra- substitution: the peak at 10,413.64 Da is the 56-mer (4-arms ∙ 56 ∙ 44 Da = 9856 Da), with an end group mass of 534.64 Da (10,413.64 – 9856 – Na), which is approximately the mass of the C5H8 pentaerythritol core + the mass of 4 levulinic acid groups.

197

Figure 6.56. The FT-IR spectrum shows the disappearance of the C=O stretch from keto- PEG (precursor) at 1718 cm-1, as the C=O stretch from the cross-linker (hydrogel) at 1765 cm-1 becomes more prominent. The precursor sample was prepared using dried precursor powders (prior to cross-linking), and the hydrogel sample was prepared by lyophilizing a hydrogel sample (pH 7.1, 10 mM), followed by grinding with KBr powder into crystal plates (n = 64 scans).

Figure 6.57. (A) Strain sweep was conducted to determine the LVR (0.1 – 10%) and 1% strain was used for subsequent frequency and time sweeps. (B) Frequency sweep on pre- formed hydrogel demonstrates minimal frequency dependence; 10 Hz was selected for reporting. All frequency sweeps were conducted using 8 mm parallel plate geometry.

198

Figure 6.58. The absolute scattering profiles of hydrogels fabricated at pH 6.8. (A) The series of buffer concentrations exhibit differences in the scattering intensity at low q (10- 3 – 10-2 Å-1), and the similarities of the scattering curves in the high q regime (>10-1 Å-1). (B) The scattering curves (pH 6.8) for each buffer concentration were fit using the Broad Peak model (solid lines), from which the the phase correlation lengths (δp) were determined. The relatively small fit error associated with the Broak Peak model constitutes its use for modeling this hydrogel system.

Figure 6.59. The absolute scattering profiles of hydrogels fabricated at pH 7.1. (A) The series of buffer concentrations exhibit differences in the scattering intensity at low q (10- 3 – 10-2 Å-1), and the similarities of the scattering curves in the high q regime (>10-1 Å-1). (B) The scattering curves (pH 7.1) for each buffer concentration were fit using the Broad Peak model (solid lines), from which the the phase correlation lengths (δp) were determined. The relatively small fit error associated with the Broak Peak model constitutes its use for modeling this hydrogel system.

199

APPENDIX B.

SUPPORTING SCHEMES & TABLES

Scheme 6.1. The various QAC compounds (Qx-OH) were produced via neat quaternization reactions of DOA, DDA, and DTDA (m = 6, 10, 12) with 8-chloro-1-octanol.

Scheme 6.2. 3,3’-dithiopropionic acid was treated with excess thionyl chloride and refluxed overnight to produce 3,3’-dithiopropanoyl chloride.

Scheme 6.3. The QAC disulfide reagents (Qx-S-S) were synthesized by esterification of the corresponding Qx-OH compounds with 3,3’-dithiopropanoyl chloride.

Scheme 6.4. The Qx-S-S reagents were reduced with TCEP to generate the corresponding Qx-SH compounds used for thiol-ene surface functionalization.

200

Scheme 6.5. Rhodamine-SH synthetic scheme beginning with the formation of the lactone, amidation with piperazine, nucleophilic substitution of 3-bromo-1-propanol, esterification with 3,3’-dithiopropanoyl chloride, and reduction to thiol using TCEP.

201

Scheme 6.6. LAP was synthesized using a Michaelis-Arbuzov reaction between the acid chloride and alkyl phosphonite to generate the acyl phosphinate, followed by treatment with LiBr.

Scheme 6.7. Allyl-TPU synthetic scheme. To incorporate a functional moiety, the feed ratio of BDO was reduced, while maintaining the molar ratio of HMDI:Arcol E-351 used for the control TPU snythesis. To produce allyl-TPU, diol A containing the allyloxy functionality (OCH2-CH=CH2) was incorporated into the backbone. For a 30 wt.% HMDI TPU containing 8 mol% of 3-allyloxy-1,2-propanediol, the molar ratios are reported in terms of the repeat unit, denoted as nR; where nA = 0.16, nB = 0.64, nC = 0.20, and n is an integer representing the total number of repeat units in the polymer.

202

Table 6.1. Reagent quantities and yields for Qx-OH precursors.

DTDA DDA DOA 8-chloro-1-octanol Qx-OH Yield g (%) mL (mmol) mL (mmol) mL (mmol) mL (mmol)

Q14-OH 28.4 (93.4) - - 15.0 (88.9) 23.3 (64.6)

Q12-OH - 20.3 (74.7) - 12.0 (71.1) 17.9 (66.7)

Q8-OH - - 15.4 (74.7) 12.0 (71.1) 15.0 (65.4)

Table 6.2. Reagent quantities and yields for Qx-S-S series.

3,3’-dithiopropanoyl Q14-OH Q12-OH Q8-OH Yield Qx-S-S chloride g (mmol) g (mmol) g (mmol) g (%) mL (mmol)

Q14-S-S 9.44 (23.2) - - 2.00 (11.6) 8.51 (74.3)

Q12-S-S - 8.78 (23.2) - 2.00 (11.6) 7.90 (73.2)

Q8-S-S - - 7.48 (23.2) 2.00 (11.6) 6.72 (70.6)

Table 6.3. Reagent quantities and yields for Qx-SH series.

Q14-S-S g Q12-S-S g Qx-S-S Q8-S-S g (mmol) TCEP g (mmol) Yield g (%) (mmol) (mmol)

Q14-SH 1.50 (1.52) - - 0.88 (3.07) 1.22 (81.3)

Q12-SH - 1.41 (1.52) - 0.88 (3.07) 1.17 (82.8)

Q8-SH - - 1.25 (1.53) 0.88 (3.07 1.05 (83.8)

203

Table 6.4. Reagent table for compounds used in various TPU polymerizations.

3-allyloxy-1,2- Stannous Acrol-E51 g TPU HMDI mL (mmol) BDO mL (mmol) propanediol mL octoate mL (mmol) (mmol) (mmol)

Control 28.1 (114.3) 61.7 (22.0) 8.2 (92.3) - 0.15 (0.46)

8% allyl 139.6 (567.4) 306.1 (109.3) 32.6 (367.3) 11.23 (90.8) 0.40 (1.24)

Table 6.5. Molecular weight and physical properties for TPUs synthesized in this study.

Molecular Weight Data Physical Properties

푀̅푛 푀̅푤 TPU Đm % allyl a Tg (°C) Tm (°C) Td (°C) Durometer b (kDa) (kDa)

Control 68 175 2.6 0.0 -60.5 72, 119 255 90

8% allyl 92 269 2.9 8.0 -62.5 72, 115 245 90

8% allyl - 1 37 87 2.4 8.1 - - - -

8% allyl - 2 43 108 2.5 8.0 - - - -

8% allyl - 3 42 113 2.7 7.8 - - - -

8% allyl - 4 35 82 2.3 7.9 - - - -

8% allyl - 5 41 103 2.5 7.8 - - - - a Determined by 1H-NMR integration. b Shore A durometer measurements were taken on compression molded samples in accordance with ASTM D2240.

204

Table 6.6. Quantification of rhodamine-SH and QAC present on UV treated and phys. ads. allyl-TPU samples as determined by fluorescence spectroscopy and XPS experiments, respectively.

-9 -2 + Fluorescence Data / 10 mol∙cm XPS Data (% NR4 relative to N)

Portion of Physical UV Physical UV Thiol Covalent Range a Physically Adsorption Treated Adsorption Treated Adsorbed QAC b

Rhodamine-SH 3.6 ± 0.1 5.5 ± 0.1 1.9 ± 0.1 - 5.5 ± 0.1 - - -

Q14-SH - - - 4.1 ± 0.6 12.4 ± 1.5 33.2 ± 6.1%

Q12-SH - - - 4.1 ± 0.2 14.1 ± 0.4 29.1 ± 1.6%

Q8-SH - - - 2.6 ± 2.2 9.4 ± 0.8 27.1 ± 23.8% a Errors are reported as propagated standard deviations after accounting for standard error in the calibration curve. b Values are the quotient of physical adsorption divided by UV treated, errors are propagated standard deviations. All experiments were performed in triplicate and averages and standard deviations are reported (n = 3).

205

Table 6.7. Additional contact-killing results for Qx-SH modified allyl-TPU samples.

Mean CFU/Sample Recovered a

Sample MRSA E. faecalis P. aeruginosa

Polypropylene b 1.51×105 1.70×105 7.94×106

Chlorhexidine c 0.00 0.00 0.00

Q8-SH 0.00 0.00 0.00

phys. Ads. Q12-SH 0.00 0.00 7.24×105

Q14-SH 0.00 0.00 4.79×106

Q8-SH 0.00 0.00 0.00

UV-treated Q12-SH 0.00 0.00 0.00

Q14-SH 1.00×102 0.00 5.25×106 a Mean CFU/sample data were determined by drop plating (n = 1). b Negative control for assay. c Positive control for assay (chlorhexidine treated polypropylene).

206