Investigation of the Molecular Organisation of Antibacterial Insect Wing Surfaces

Investigation of the molecular organisation of antibacterial insect wing surfaces

Submitted in total fulfilment of the requirements for the degree of

Doctor of Philosophy

Song Ha Nguyen

Department of Chemistry and Biotechnology

Faculty of Science, Engineering, and Technology

Swinburne University of Technology

May 2015

______Abstract

It has been known for some time that the surfaces of many plants and insects exhibit superhydrophobicity and self-cleaning properties. Plant leaves have been intensively studied for over a decade since the discovery of the “lotus effect”, where water droplets readily roll off the surfaces of lotus leaves. The water droplets, once landed on the surfaces of the leaves, retain their spherical shape. This arises from the superhydrophobicity of the surface. When this surface is tilted on a small angle (θtilt < 10°), the water droplets roll off the surface of the leaves, collecting all the dirt and contaminants as they go, affording the leaves the ability to ‘self-clean’. A large amount of information has been obtained regarding the surface properties of lotus leaves; however the surface properties of insects have not received as much attention in this context, despite the fact that they often present similarly effective self-cleaning properties. This work was undertaken in an effort to fill this gap in scientific knowledge.

Insect wings surfaces are covered by an epicuticle layer that is composed of hydrophobic substances, which are responsible for imparting the superhydrophobic and self-cleaning properties to some species. The surfaces of dragonfly wings are covered by an array of nanopillars, which were recently discovered to have bactericidal activity. In this work, the molecular organisation of dragonfly wings was investigated to gain an insight into how the individual nanopillars were constructed and behave when coming into contact with attaching bacterial species. Over 60 different substances were identified as comprising the insect wing surfaces, including aliphatic hydrocarbons and fatty acids. The precise chemical compositions not only directly contribute to the superhydrophobic properties of the wings, but also determine their nano-architectures. A new sub-layer of the dragonfly wing epicuticle is proposed, the meso-epicuticle, which is composed only of aliphatic hydrocarbons. This meso-epicuticle is distinct from the outer- and inner-epicuticular layers.

The bactericidal properties of dragonfly wings have been reported to arise purely due to physical means, and hence the surface morphologies and topographies of the dragonfly wings were characterised to investigate the nature of the mechanism(s) responsible for the bactericidal effect. The wing surfaces of two species of dragonfly, Hemianax papuensis and Diplacodes bipunctata, were investigated for this purpose. It

ii was suggested that mechano-bactericidal and self-cleaning properties of the wings are two distinct mechanisms that dragonflies have developed in order to cope with bacterial contamination. Wetting and adhesion are affected by both the surface topography and surface chemistry with the presence of precisely defined topographical features resulting in bactericidal properties. Moreover, a subtle change in the morphological parameters may instead enhance the ability for the surface to self-clean.

This knowledge about the relative contribution of molecular organisation and surface topology to the overall surface properties of dragonfly wings was used to inform the synthetic stage of the project. Eicosane, docosane, palmitic acid and stearic acid, all of which were found to be major components present in the epicuticular layer of dragonfly wings, were allowed to self-assemble on the surface of highly-ordered pyrolytic graphite (HOPG) to produce an ordered structure, which may replicate the antibacterial properties of dragonfly wings. Both of the alkanes assembled into ‘micro- rodlet’ type structures, while the fatty acids recrystallised into sharper ‘micro-blades’. In the case of the alkane micro-rodlets, bacterial cells coming into contact with the substrate surface containing these rodlets were found to align along the alkane crystallites. The mechanism responsible for this behaviour was determined to be a result of the air-pockets trapped between the surface structures, and hence the bacteria were limited in their colonisation to the areas of the surface that could be accessed by water. Bacterial alignment on the fatty acid interfaces was less obvious; however these surfaces were capable of killing the bacteria. The structure of the micro-blades was integral in this case; it appeared that the high radius of curvature of the sharp edge were able to rupture the bacterial cells through high stress upon contact and adhesion.

The results of this study contribute to the general knowledge available on the molecular organisation of the epicuticle of dragonfly wings, with specific contribution to the surface properties of the wings, such as their superhydrophobicity, bactericidal behaviour and their ability to self-clean. The key mechanisms responsible for each of these properties were proposed to be closely related to the interplay between the wettability of the surface and the ability of the surface to retain air between the topographical structures. This study also introduced a fast and facile method of producing artificial surfaces, being the self-assembly of lipids onto ordered substrata. The resulting surfaces were found to be successful in exerting antibacterial effects both before and after bacteria attachment. iii

Acknowledgements

First and foremost, I would like to acknowledge and send my greatest appreciation to my primary supervisor, Professor Elena Ivanova for all of her advice and support during the course of my study. Just like a mother figure, her inspiration, motivation, and encouragement were the driving forces pushing me forward to completion of this work. She does not mind to stand up for her students whenever they are treated unfairly. Despite her busy schedule, she has always taken time to talk with me whenever I stumbled across problems, and I’m really grateful for that.

Similarly, I would also like to thank Professor Russell Crawford, Professor David Mainwaring, and Dr. Peter Mahon for co-supervising the project, and for spending time helping me on various manuscripts and papers that I prepared. Being supervisors of a non-English speaking student would not be an easy job and they have always been patient with me, I really appreciate that. Special thank you to Dr. Peter Mahon, who is my supervisor and also supporter, for talking to me caringly as a friend rather than simply a teacher and I’m very thankful for that.

To Vy and Hiệp, sometimes I questioned our friendships, and felt doubtful, but when I look around I always see you guys right there, and I think that’s what matters the most. I’m thankful for this confusing friendship because when I need your supports the most, you guys never fail me, and I really appreciate that.

To Nguyên and Chris, you two are my rocks. Thanks for being here with me, for your co-operation in numerous impossible-sounding missions, and for your emotional support. When things are down, you guys have always brought me back to my ground.

I would like to send my gratitude to Dr. Jafar Hasan, Dr. Mohammad Al Kobaisi, Dr. Vi Khanh Truong, and Dr. Hooi Jun Ng for their knowledge and assistance in operating various instruments, and techniques. Also, thank you to Veselin ‘Vanya’ Boshkovikj, you are one of the funniest guys I have ever met. You guys made my PhD life more colourful and much easier.

Thank you to Dr. James Wang for his assistance in performing SEM experiments. Similarly, I would like to thank Dr. Mark Tobin and Dr. Ljiljana Puskar for helping with our work at the Australian Synchrotron, FTIR beamline; without their expertise, I would not be able to become a regular there.

A special thanks and gratitude to chú Ngân, Soula, Savithri, Chris, Rebecca and all the laboratory technicians who have helped me throughout the course of this work. Talking to you all made me sometimes forget how tough PhD life can be.

To all my friends here at Swinburne, you guys have made my every day more fun and enjoyable. I will never forget our little office where the temperature is always above 30°, with the white board where everyone writes their thoughts. It’s a pity that around the time I write all this, you guys are not around. I would rather say my thankful words out loud individually, but I’m afraid that I’ll in tears before I complete my sentences, so I send my special gratitude here, to Yến, Matt, Kaylass, Shanthi, Phong and Qudsia.

To our lovely and handsome dog, Sammy. You are my mascot and on my wallpaper screen; looking at you make my days go by much easier, so if you can understand, I would like to say thanks to you.

Especially, thank you to my new and current family. The word by itself says a lot how much you all mean to me. To my new family, I came along and you all took me in with your kindest hearts, treated me as part of the family, giving me support and advice which always touch my heart. To my husband, Dr. Hayden Webb, I don’t know how many words are enough to express my gratitude to you for your endless emotional support, thanks for always being by my side sharing my happiness and motivating me through the toughest times despite my difficult nature. Before being my friend, my husband, you are my senior, my teacher who does not mind to be harsh to me for my improvement.

Con muốn gửi những lời cám ơn sâu sắc nhất tới ba, mẹ và bé Ngân. Con khô khan, nhưng con hy vọng mọi người hiểu được là con mang ơn mọi người như thế nào. Sự hy sinh thầm lặng của mẹ luôn là động lực vực con đứng dậy trong ư4ng tháng ngày khó khăn. Cha ít nói chuyện với con, nhưng con biết khi con thành công, cha sẽ là người nói nhiều nhất, khoe thành tích con gái cha với mọi người mà cha gặp. Nếu không có hai người, con không có mặt trên đời để giờ đây có thể ngồi đây, tự hào viết ra những dòng chữ ý nghĩa cho đời. Chị cám ơn em gái của chị, chị đi ròng rã 6-7 năm trời, mọi gánh nặng, trách nhiệm làm con đều để em gánh hết. Chị em mình tuy xa mặt, nhưng hy vọng sẽ không cách lòng.

Declaration

I, Song Ha Nguyen, declare that this thesis is original work and contains no material that has been accepted for the award of Doctor of Philosophy, or any other degree or diploma, except where due reference is made.

I declare that to the best of my knowledge this thesis contains no material previously published or written by any other person except where due reference is made. Wherever contributions of others were involved every effort has been made to acknowledge the contributions of the respective workers or authors.

Signature ______

List of Publications

Book Chapters

Song Ha Nguyen, Hayden K. Webb, Russell J. Crawford & Elena P. Ivanova (2014). Natural superhydrophobic surfaces. In: Elena P. Ivanova, and Russell J. Crawford, Superhydrophobic surfaces (Elsevier).

Hayden K. Webb, Song Ha Nguyen, Russell J. Crawford & Elena P. Ivanova (2014), Biological interactions with superhydrophobic surfaces. In: Elena P. Ivanova, and Russell J. Crawford, Superhydrophobic surfaces (Elsevier).

Song Ha Nguyen, Hayden K. Webb & Elena P. Ivanova (2014) Natural antibacterial surfaces. In: Elena P. Ivanova, and Russell J. Crawford, Antimicrobial surfaces (Springer).

Hayden K. Webb, Song Ha Nguyen & Elena P. Ivanova (2014) Introduction to antibacterial surfaces. In: Elena P. Ivanova, and Russell J. Crawford, Antimicrobial surfaces (Springer).

Peer-reviewed article

Song Ha Nguyen, Hayden K. Webb, Peter J. Mahon, David Mainwaring, Russell J. Crawford, and Elena P. Ivanova, “Graphite-templated self-organized alkane microcrystal interfaces regulate bacterial attachment”, Biofouling, (2015), [In press]

Song Ha Nguyen, Hayden K. Webb, Jafar Hasan, Mark J. Tobin, David E. Mainwaring, Peter J. Mahon, Richard Marchant, Russell J. Crawford, and Elena P. Ivanova, “Wing wettability of Odonata species as a function of epicuticular waxes spatial area”, Vibrational Spectroscopy, 75 (2015), p. 173-177

Song Ha Nguyen, Hayden K. Webb, Peter J. Mahon, Russell J. Crawford, Elena P. Ivanova, “Natural insect and plant micro-/nanostructured surfaces: An excellent selection of valuable templates with superhydrophobic and self-cleaning properties”, Molecules, 19 (2014), p. 13614-13630.

Song Ha T. Nguyen, Hayden K. Webb, Jafar Hasan, Mark J. Tobin, Russell J. Crawford, Elena P. Ivanova, “Dual roles of outer epicuticular lipids in determining the wettability of dragonfly wings", Colloids and Surfaces B: Biointerfaces, 106 (2013), p.126-134.

vii

Elena P. Ivanova, Song Ha Nguyen, Hayden K. Webb, Vi Khanh Truong, Jafar Hasan, Robert N. Lamb, Alex Duan, Mark J. Tobin, Peter J. Mahon, Russell J. Crawford, “Molecular organisation of the nanoscale surface structures of the dragonfly Hemianax papuensis wing epicuticle", PLoS One, 8 (2013), e67893.

Song Ha Nguyen, Hayden K. Webb, Jafar Hasan, Mohammed Al Kobaisi, Robert N. Lamb, Alex H.-F. Wu, Peter J. Mahon, David Mainwaring, Russell J. Crawford, Elena P. Ivanova, “Nanomorphology of dragonfly wings - finely-tuned through evolution”, PNAS [Under preparation]

Song Ha Nguyen, Hayden K. Webb, Peter J. Mahon, David Mainwaring, Russell J. Crawford, and Elena P. Ivanova, “Mechanobactericidal fatty acid microblades assembled on highly-ordered pyrolytic graphite”, Small [Under preparation]

Conference presentations with published abstracts

Song Ha Nguyen, Hayden K. Webb, Peter J. Mahon, David E. Mainwaring, Russell J. Crawford, & Elena P. Ivanova, “Antimicrobial activity of self-assembled carboxylic acid crystals on graphite”, III International Conference on Antimicrobial Research (ICAR) 2014

Song Ha Nguyen, Hayden K. Webb, Jafar Hasan, Mark J. Tobin, David E. Mainwaring, Peter J. Mahon, Richard Marchant, Russell J. Crawford, Elena P. Ivanova, “Correlation between the proportion of epicuticular waxes and wettability among various species of dragonfly wing”, 7th International Workshop on Infrared Microscopy and Spectroscopy with Accelerator Based Sources (WIRMS) 2013.

Conference posters with published abstracts

Song Ha Nguyen, Hayden K. Webb, Peter J. Mahon, David Mainwaring, Russell J. Crawford, & Elena P. Ivanova, “Regulation of bacterial attachment through graphite- templated self-organized alkane microcrystal interfaces”, Australian Society of Microbiology Meeting 2014.

Song Ha Nguyen, Hayden K. Webb, Mark J. Tobin, David E. Mainwaring, Peter J. Mahon, Russell J. Crawford, & Elena P. Ivanova, “FTIR detection of different phases of fatty acids forming 3D-assemblies”, Australia Synchrotron User Meeting 2014.

Song Ha Nguyen, Hayden K. Webb, Jafar Hasan, Robert N. Lamb, Mark J. Tobin, Peter J. Mahon, David E. Mainwaring, Russell J. Crawford, & Elena P. Ivanova, viii

“Dissimilar bactericidal activities of nanopatterned dragonfly wing surfaces”, Nano Today 2013.

Song Ha Nguyen, Hayden K. Webb, Jafar Hasan, Vi Khanh Truong, Robert N. Lamb, Xiaofei Duan, Mark J. Tobin, Peter J. Mahon, Russel J. Crawford, Elena P. Ivanova, “Hierarchical construction of the surfaces of dragonfly wings on the micro- and nano-scale”, Australian Society of Microbiology Meeting 2013.

Song Ha Nguyen, Jafar Hasan, Hayden K. Webb, Mark J. Tobin, Russell J. Crawford, Elena P. Ivanova, “Physico-Chemical characterisation of dragonfly wing nanostructures using Synchrotron Infrared Microscopy”, Australia Synchrotron User Meeting 2012

Table of Contents

Abstract ...... ii

Acknowledgement ...... iv

Declaration ...... vi

List of Publications ...... vii

Book Chapters ...... vii

Peer-reviewed article ...... vii

Conference presentations with published abstracts ...... viii

Conference posters with published abstracts ...... viii

Table of Contents ...... x

List of Abbreviations ...... xv

List of Figures ...... xvii

List of Tables ...... xxviii

Chapter 1. Introduction ...... 1

1.1. Overview ...... 2

1.2. Aims of the study ...... 3

Chapter 2. Literature review ...... 5

2.1. Overview ...... 6

2.2. Arthropod cuticle ...... 6 2.2.1. Morphology of insect wings ...... 6 2.2.2. Aerodynamics of insect flight ...... 7

2.3. The surface of insect wings ...... 9 2.3.1. Concept of self-cleaning ...... 9 2.3.2. Wettability theory ...... 10

2.4. Surface chemistry ...... 13

2.5. Hierarchical structure induced self-cleaning ...... 22 2.5.1. Superhydrophobicity and self-cleaning surfaces in nature ...... 22 x

2.5.1.1. Plant surface structures ...... 23 2.5.1.2. Insect surface structures ...... 25 2.5.2. Superhydrophobicity with additional properties ...... 31 2.5.2.1. Superhydrophobic surfaces with high adhesive force ...... 31 2.5.2.2. Anisotropic superhydrophobic surfaces ...... 33 2.5.3. Hierarchical structure of superoleophobic surfaces ...... 35

2.6. Bacteria ...... 36 2.6.1. Bacterial physiology ...... 36 2.6.2. Biofilm formation ...... 39

2.7. New approaches for controlling biofilm formation ...... 41 2.7.1. Antibiofouling ...... 41 2.7.2. Bactericidal activity of insect wings ...... 43

Chapter 3. Materials and methods ...... 46

3.1. Overview ...... 47

3.2. Materials ...... 47

3.3. Protocols for handing of insect wings ...... 47 3.3.1. Fresh and aged insect wings ...... 47 3.3.2. Extraction of the lipids from dragonfly wings ...... 48

3.4. Surface fabrication ...... 48

3.5. Surface characterisation ...... 48 3.5.1. Atomic force microscopy ...... 48 3.5.2. Scanning electron microscopy ...... 49 3.5.2.1. Biological samples ...... 49 3.5.2.2. Synthetic surfaces ...... 49 3.5.2.3. Quantification of alignment using fast Fourier transforms (FFT) analysis ...... 50 3.5.3. Raman spectroscopy ...... 50 3.5.4. Wettability ...... 50

3.6. Chemistry characterisation techniques ...... 51 3.6.1. X-ray photoelectron spectroscopy ...... 51

3.6.2. Synchrotron radiation Fourier Transform Infrared spectroscopy (SR- FTIR) ...... 51 3.6.2.1. FTIR in tranmission mode ...... 51 3.6.2.2. FTIR in attenuated total reflection (ATR) mode ...... 52 3.6.3. Gas chromatography – mass spectrometry ...... 52

3.7. Bacterial strains, growth, and sample preparation ...... 53

3.8. Biological activity assays ...... 54 3.8.1. Plate counting assay ...... 54 3.8.2. Quantitation of bactericidal activity using confocal images for fabricated surfaces ...... 54

3.9. Confocal laser scanning microscopy (CLSM) ...... 55

3.10. Statistical analysis ...... 55

Chapter 4. Surface characterisation and functionality of the epicuticle of insect wings ...... 56

4.1. Overview ...... 57

4.2. Surface topology and wettability of dragonfly wings ...... 57 4.2.1. Scanning electron microscopy ...... 57 4.2.2. Atomic force microscopy ...... 58 4.2.3. Surface wettability ...... 60

4.3. Correlation between surface chemistry and the wettability of a surface ...... 62 4.3.1. Variations in the surface wettability of dragonfly wings ...... 62 4.3.2. Relationship between the quantity of surface epicuticular lipids and wettability ...... 64

4.4. Dual roles of epicuticular lipids in determining the wettability of dragonfly wings ...... 69 4.4.1. Wettability of dragonfly wings as a function of aliphatic hydrocarbon content ...... 69 4.4.2. The chemical composition of untreated and chloroform extracted wings ...... 70 4.4.3. Changes in surface morphology on removal of surface aliphatic hydrocarbons ...... 74 xii

4.5. Summary ...... 79

Chapter 5. Molecular organisation of dragonfly wings ...... 81

5.1. Overview ...... 82

5.2. Physical effects on dragonfly wings resulting from extraction with chloroform ...... 82 5.2.1. Changes in surface morphologies ...... 82 5.2.2. Changes in surface topographies ...... 85

5.3. The chemical composition of the Hemianax papuensis dragonfly wing epicuticle ...... 89

5.4. Molecular organisation of H. papuensis dragonfly wing epicuticle ...... 95

5.1. Summary ...... 98

Chapter 6. Nanostructural effects on antibacterial activity and wettability of dragonfly wings ...... 100

6.1. Overview ...... 101

6.2. Bactericidal activities of dragonfly wings ...... 102

6.3. Comparative surface chemistry of the dragonfly wings ...... 105

6.4. Variation in surface structure ...... 108 6.4.1. Surface topologies ...... 108 6.4.2. Surface morphologies ...... 113

6.5. Structure-dependent wettability ...... 115

6.6. Summary ...... 122

Chapter 7. Systematic bacterial patterning using self-assembled microtextured alkanes ...... 123

7.1. Overview ...... 124

7.2. Thermodynamics of alkane recrystallisation on HOPG and silicon ...... 125

7.3. Patterns of the self-assembled crystals ...... 127 7.3.1. Crystal morphology ...... 127 7.3.2. Eicosane and docosane crystal topographies ...... 130

xiii

7.4. Bacterial attachment to HOPG template-mediated self-organised alkane microcrystallites ...... 132 7.4.1. Local hydrophobicity for bacterial attachment and alignment ...... 135

7.5. Summary ...... 139

Chapter 8. Synthetic mechanobactericidal surfaces composed of dragonfly fatty acids ...... 141

8.1. Overview ...... 142

8.2. Crystal patterns of carboxylic acids on graphite ...... 142

8.3. Thermodynamics of crystallisation ...... 146 8.3.1. Directional crystals formed by the two fatty acids ...... 146

8.4. Biological activities ...... 150

8.5. Gold-coating of the fatty acids crystals ...... 153

8.6. Spatial wettability ...... 156

8.7. Summary ...... 160

Chapter 9. General Discussion ...... 162

9.1. Overview ...... 163

9.2. Molecular organisation of the surface of insect wings ...... 164

9.3. Assessment of the antibacterial potential of superhydrophobic insect wing surfaces ...... 166

9.4. Fabrication and evaluation of synthetic surfaces with the ability to control bacterial attachment ...... 168

Chapter 10. Conclusions and future directions ...... 171

10.1. Summary and conclusions ...... 172

10.2. Future directions ...... 173

10.3. Final remarks ...... 174

Bibliography ...... 175

xiv

List of Abbreviations

AFM Atomic Force Microscopy

SEM Scanning Electron Microscopy

GCMS Gas Chromatography and Mass Spectrometry

XPS X-ray Photoelectron Spectroscopy

FTIR Fourier Transform Infrared Microscopy

ATR Attenuated total reflection

CLSM Confocal Laser Scanning Microscopy

CA Contact angle

WCA Water contact angle

CAH Contact angle hysteresis

CI Composite interface

SPME Solid-phase Micro-extraction

PDMS Polydimethylsiloxane

HCs Hydrocarbons

FA Fatty acid

Alcs Alcohols

Alds Aldehydes

ST Setae

SP Spatula

BR Branched

RO Radial outward

HOPG Highly ordered pyrolytic graphite

FFT Fast Fourier Transforms

UHV Ultra high vacuum

TPL Three phase line xv

MD Molecular dynamics

Au Gold

Sa Staphylococcus aureus

Pa Pseudomonas aeruginosa

xvi

List of Figures

Figure 2.1. The role of topography and surface chemistry in determining the superhydrophobicity of a surface, according to the Cassie-Baxter model. The surface presented in A exhibits a relatively low water contact angle (WCA), as it is composed of a material that is hydrophilic, and does not possess a surface structure that can trap significant quantities of air. Changing the chemical composition of the surface (B), or the physical structure of the surface can both lead to an increased WCA, in accordance with the Cassie-Baxter equation. Surface D exhibits the highest WCA, as it combines a low surface energy material with a physical structure that can trap substantial quantities of air in the surface. Examples of plant and insect surfaces belonging to each of these surface types are presented in the inset images...... 12

Figure 2.2. Images of some superhydrophobic plant surfaces, and their corresponding high resolution SEM images: (A) lotus leaves; (B) Indian Canna leaves; (C) taro leaves; (D) perfoliate knotweed leaves. (Adapted with permission from Elsevier (Guo and Liu, 2007)) ...... 24

Figure 2.3. The surface morphologies of the Hemianax papuensis dragonfly wing. Credit photograph to © Arthur Chapman...... 26

Figure 2.4. Insect wings and their corresponding high resolution SEM images. (A-A2) Isoptera Nasutitermes sp.; (B-B2) Orthoptera Acrida cinerea cinerea; (C-C2) Hemiptera Meimuna opalifera. (Adapted with permission from ACS publication and Elsevier (Byun, Hong, Saputra et al., 2009; Watson, Cribb and Watson, 2010) and Encyclopedia of Life (2014) (eol.org, license agreements can be found at http://creativecommons.org/licenses/by-nc-sa/2.0/ and http://creativecommons.org/licenses/by-nc/2.0/), photographs taken by Kenpei and Tomomarusan)...... 30

xvii water adheres strongly to the surface. These structures also enable the geckoes to adhere to solid surfaces, enabling them to climb walls. ST: setae, SP: spatula, BR: branched. (Reproduced with permission from ACS publication and Elsevier (Gao, Wang, Yao et al., 2005; Feng, Zhang, Xi et al., 2008; Bhushan, 2012)...... 32

Figure 2.6. The anisotropic surface structures of (A) rice leaves and (B) butterfly wings. The superhydrophobic structures present on both surfaces are directionally ordered, resulting in a variable wettability that is dependent on the direction of movement. (Reproduced with permission from Elsevier (Guo and Liu, 2007; Peng, Hu and Zhang, 2011), photographs taken by Anne Ten Donkelaar and Stepanka Nemcova)...... 34

Figure 2.7. General structure of bacterial cells (Reproduced with permission for single use from ‘Molecular expression’ http://micro.magnet.fsu.edu/cells/bacteriacell.html). 37

Figure 2.9. Schematic of the two stages biofilm formation on surfaces. Reproduced with permission from American Chemical Society (Lichter, Van Vlietpa and Rubner, 2009)...... 41

Figure 2.10. Bacterial cells were punctured by the surface of (A, A1) cicada wing Psaltoda claripennis, and (B-B3) dragonfly wing Diplacodes bipunctata. Cells were clearly punctured by nanopillars presented on the surfaces of these insect wings. (Reproduced with permissions from Encyclopedia of Life (eol.org, license agreements can be found at http://creativecommons.org/licenses/by-nc/2.0/), photographs taken by Graham Wise)...... 44

Figure 3.1. Schematic diagram of Synchrotron the FTIR in ATR mode...... 52

Figure 4.1. High resolution scanning electron micrograph of the H. papuensis and H. tau dragonfly wings. (A) The side view of the H. papuensis wing and (B) its corresponding top view; (C) and (D) are side view and top view of the H. tau wing. ... 58

Figure 4.2. Surface topography of (A) Hemianax papuensis and (B) Hemicordulia tau dragonfly wings with their corresponding line profiles. Inset scale bar = 500 nm...... 59

xviii

Figure 4.3. A water droplet coming into contact with the H. papuensis dragonfly wing surface. The surface is sufficiently hydrophobic for the droplet to bounce after it makes initial impact...... 61

Figure 4.4. Average static water contact angle (grey) and average contact angle hysteresis (black), with example water droplet contact angles (above). The shaded area at the top of the graph represents the minimum water contact angle threshold required for self-cleaning ability. Error bars indicates standard deviation, n=15...... 63

Figure 4.5. High resolution scanning micrographs of (A) H. papuensis, (B) H. tau, (C) D. melanopsis dragonfly wings and their corresponding 3D image...... 64

Figure 4.6. Infrared absorbance spectra of the wings of three species of dragonfly. Most bands in the spectra were very similar for all three species, however the increased intensity of the CH2 stretching bands in the spectrum of the H. papuensis dragonfly provides evidence that there is a greater amount of aliphatic hydrocarbons present on the wing surface...... 65

Figure 4.7. Infrared absorbance spectra of two species of damselfly wings, Xanthagrion erythroneurum and Ischnura heterosticta, and their corresponding WCA. The intensity of the CH2 stretching bands provides an indication of the amount of aliphatic hydrocarbons present on the wing surface...... 68

Figure 4.8. Loss of the superhydrophobic and self-cleaning properties of dragonfly wings after extraction with chloroform. The average water contact angles are presented in grey and average contact angle hysteresis presented in black. The shaded area at the top of the graph represents the minimum water contact angle threshold required for classification as superhydrophobic surfaces. Error bars indicates standard deviations, n=20...... 70

Figure 4.9. Infrared spectra of the H. tau dragonfly wings before and after chloroform extraction. The decrease in aliphatic hydrocarbons present on the surface of the dragonfly wings occurs with increasing chloroform extraction time. The intensity of the

CH2 stretching bands is a good approximation of the concentration of aliphatic hydrocarbons, i.e. lipids (waxes) present on the wing surface. The spectra were acquired at the Australian Synchrotron...... 71

xix

Figure 4.10. Relative proportion of all identified chemical classes present in the epicuticular layer of the H. tau dragonfly wing at three extraction times. The epicuticle was dominated by n-alkanes...... 72

Figure 4.11. Cross-section (A, C, E, G) and surface (B, D, F, H) morphology of Hemicordulia tau dragonfly wing surfaces before and after chloroform extraction over 10, 20 and 30 s. Decreased water contact angles were observed with increased extraction time (inset). Scale bars = 400 nm ...... 75

Figure 4.12. Surface topography of Hemicordulia tau dragonfly wings after various periods of chloroform extraction, inferred from typical two-dimensional AFM scans (5 µm × 5 µm areas) and corresponding cross-sectional profiles. The inset images are AFM scans of 1 m × 1 m areas. Scale bar = 1 m...... 77

Figure 5.1. Cross-sectional images of H. papuensis dragonfly wing membranes prior to, and post-extraction with chloroform, and corresponding images of the WCA. Extraction for 10 s had only a minor effect on the surface morphology, however the WCA decreased by more than 20°. After 1 h extraction, there was no trace of the surface nanostructures, indicating the complete removal of the epicuticular lipids. The WCA on the 1 h extracted surface was 93°, nearly 70° lower than the original, untreated wing. Scale bars = 400 nm...... 83

Figure 5.2. Top view micrographs and 3D images of Hemianax papuensis dragonfly wings prior to, and after 10 s and 1 h of chloroform extraction. The general structure of the dragonfly wings surface largely remained intact after 10 s of chloroform extraction, however, the structure disappeared completely after a 1 h extraction period...... 85

Figure 5.3. Surface topography of Hemianax papuensis dragonfly wings after various periods of extraction with chloroform. Typical two-dimensional AFM scans (1.0 µm × 1.0 µm areas) and cross-sectional profiles of the untreated (A), 10 s (B) and 1 h (C) chloroform-extracted wing surfaces are presented. The surface features became broader and more spatially distributed after 10 s of extraction, however after 1 h of extraction the nanostructures were no longer visible. Scale bars = 100 nm...... 87

Figure 5.4. Representative infrared spectra of untreated and chloroform-extracted wing membranes of the Hemianax papuensis dragonfly. The intensity of the CH stretching bands (shaded), and carbonyl peak (indicated by arrows) decreased successively with extended extraction time. Spectra were acquired in transmission mode...... 90 xx

Figure 5.5. Relative proportions of the major compound classes (top) and chain length (bottom) of dragonfly wing epicuticle components, extracted in 10 s and 1 h...... 94

Figure 5.6. Atomic proportions of carbon, oxygen and nitrogen in the wings of Hemianax papuensis...... 96

Figure 5.7. Proposed model of the epicuticle of the Hemianax papuensis wing membranes. Three layers are contained within the epicuticle: the outer epicuticle, the mesoepicuticle and the inner epicuticle...... 98

Figure 6.1. Bactericidal effects of H. papuensis (top) and D. bipunctata (bottom) dragonfly wings against various bacterial strains. The dead cells in CLSM images were stained with propidium iodide, indicated in red, while live cells were stained with SYTO 9, indicated in green (Scale bar = 5µm). Yellow fluorescence indicated the binding of both dyes, which indicates that sufficient propidium iodide had passed through the membrane, indicating cell rupture and death...... 103

Figure 6.2. Representative scanning electron micrographs of various bacterial cells on the surface of H. papuensis (left column) and D. bipunctata (right column) dragonfly wings, demonstrating the attachment behaviour of P. aeruginosa (A, B), S. aureus (C, D), vegetative B. subtilis (E, F) cells and its spores (G, H). All cell types were deformed when adhered to D. bipunctata wings, however only P. aeruginosa cells were similarly affected on the H. papuensis wings. Scale bars = 1 µm, inset scale bars = 200 nm. ... 104

Figure 6.3. Bactericidal efficiency of D. bipunctata and H. papuensis dragonfly wing surfaces against four different bacterial cell types. The values presented are the number of cells killed per cm2 of wing surface per minute of incubation, over the first three hours...... 105

Figure 6.4. Representative infrared spectra of the wing membranes of H. papuensis (red) and D. bipunctata (blue) dragonflies. The spectra of two dragonfly wings showed their highly similarities in fundamental chemical compositions...... 106

Figure 6.5. High resolution XPS spectra of H. papuensis and D. bipunctata dragonfly wings. High-resolution scans were performed in approximately 20 eV intervals across the O 1s and C 1s peaks. The C 1s peaks of both species are dominated by C–C and C– H bonds, whereas there is an approximately 3:4 ratio between C–O and C=O components within the O 1s peaks...... 107

xxi

Figure 6.6. Photographs of (A) H. papuensis and (B) D. bipunctata dragonflies and corresponding scanning electron micrographs of their wing surfaces (C, D). According to SEM analysis, their structures appeared to be highly similar, however the H. papuensis wings appeared to have more connecting structures between the pillars and the peaks of D. bipunctata wings appeared slightly sharper and thinner. Insets are 3D representations of sections of the micrographs. (A, and B are reproduced with permission from the Encyclopedia of Life (2014))...... 109

Figure 6.7. Fast Fourier transform analysis of (A) H. papuensis and (B) D. bipunctata wing surface images. The grayscale plots of cross-sections of the 2D FFT at 0° show a longer range directional periodicity up to 2.5 µm for D. bipunctata and 4 µm for H. papuensis wings. The area % and population distribution of the wings surface features demonstrated that the feature size of H. papuensis wings is about 70 nm and 100 nm for D. bipunctata wings...... 110

Figure 6.8. Cross-sections of the wings of H. papuensis and D. bipunctata dragonflies. The nanopillar arrays on the surface of D. bipunctata wings appeared relatively consistent in height, whilst those on the surface of H. papuensis wings appeared to vary more in their height...... 111

Figure 6.9. Typical two-dimensional AFM scans (2.5 µm × 2.5 µm areas) and the cross-sectional profiles of (A) H. papuensis and (B) D. bipunctata wings. The surface of the H. papuensis wing appeared highly irregular, with an increased degree of hierarchy of the nanopillars, whilst the line profile of D. bipunctata wings was comprised of regular peaks with a similar height and relatively consistent spatial distribution...... 112

Figure 6.10. Demonstration of bearing analysis. The bearing ratio of a surface is defined as the proportion of the projected area that is occupied by the surface at a given depth from the highest point on the surface (demonstrated in A-C, shaded areas). (D) a demonstration of how three statistical parameters were extracted directly from the plot, i.e. peak height (Rpk), core roughness (Rk) and valley depth (Rvk)...... 114

Figure 6.11. Fourier-transform analysis of H. papuensis dragonfly wing structures at various magnification levels. The inflection points (indicated) demonstrate the dominant feature sizes on the wings of H. papuensis wing at each magnification. The feature size indicated for the 15000× magnification corresponds to the negative space and provides an indication of the spacing of the nanopillars, whilst the size indicated at 35000× and xxii

70000× magnification levels correspond to the nanopillars themselves. The presence of two different nanopillar heights is indicative of the presence of a hierarchical surface structure...... 116

Figure 6.12. Fourier-transform analysis of D. bipunctata dragonfly wing structures at various magnification levels. The inflection points (indicated) demonstrate the dominant feature size on the wings of D. bipunctata at each magnification. The feature size indicated for 15000× magnification corresponds to the negative space and provides an indication of the spacing of the nanopillars, whilst the size indicated at 35000× and 70000× magnification levels correspond to the nanopillars themselves. The nanopillars appeared to be relatively similar at both the larger magnifications...... 117

Figure 6.13. Cassie-Baxter wettability of H. papuensis and D. bipunctata dragonfly wings based on their bearing profiles. If the surface is assumed to adhere to the Cassie- Baxter model of wettability, the theoretical wettability of the surface can also be plotted as a function of the depth of impingement of a water droplet between the surface features of the wing...... 119

Figure 7.1. The process of making ordered microcrystalline interfaces ...... 124

Figure 7.2. Topographical analysis of the surfaces of silicon wafer and HOPG. Scans of 5µm x 5µm areas of (a) HOPG, and (b) silicon with corresponding line profiles. Both substrata have highly smooth surfaces...... 125

Figure 7.3. Scanning electron micrographs of (A, C) eicosane and (B, D) docosane self- assembled on the surface of (A, B) HOPG and (C, D) silicon wafer. Rodlet directionality was observed for both lipids on the surface of the HOPG, suggesting an epitaxial relationship; whilst on the surface of silicon wafer, the lipids appeared as irregular clumps...... 127

Figure 7.4. HOPG template-mediated recrystallisation patterns of eicosane and docosane. Scanning electron micrographs of (a) eicosane and (b) docosane and their corresponding crystallite orientations demonstrate the tendency of the crystals to align in certain directions, often forming triangular shapes. This pattern is attributable to the alignment of alkane molecules with the molecular order of the HOPG substratum. ... 128

Figure 7.5. Schematic representation of alkane alignment when they are adsorbed on graphite in both zigzag and armchair orientations. Two alkane molecules in the same

xxiii orientation will align an angle of 60°, while molecules in the opposite orientations will be aligned at 30° or 90°...... 129

Figure 7.6. 2D-FFT analysis of micrographs of (A) eicosane and (B) docosane on HOPG, Preferential orientation of the crystals can be observed, with predominant directionalities 60° apart, suggesting that most of the initially adsorbed molecules adopted either the zigzag or armchair configurations...... 130

Figure 7.7. Topography of eicosane (A) and docosane (B) alkane microcrystals self- organised on the graphite templates and their corresponding cross-sectional profiles. The directionality of the alkane microcrystals are indicated with arrows...... 131

Figure 7.8. Bacterial attachment to the self-organised alkane microcrystallites. Bacterial cells attached only along the interfaces formed by the microcrystallites. The white ovals indicate the alignment of bacterial cells in straight lines, following the same patterns that were evident in the SEM and AFM images of the alkane microcrystals (see Figure 7.4 and 7.6). SYTO®9 fluorescent dye was used to stain the live bacteria, whilst non-viable cells were labelled red using propidium iodide...... 133

Figure 7.9. Attachment of S. aureus and P. aeruginosa cells on the HOPG surface. Bacterial cells attached randomly on the surface without any apparent pattern...... 134

Figure 7.10. Alkane microcrystallite-directed bacterial cell attachment. (A) Raman microspectroscopic mapping indicated air retention on the surface of eicosane microcrystallites when immersed in water. Lighter areas indicate a high intensity of - OH bands (attributable to higher quantities of water) and darker areas indicate regions lacking water (i.e. air pockets) (B) Schematic representation of spherical bacterial cells attaching to the microcrystallite-water-air interface. The force of attraction to alkanes is characterised by van der Waals forces, resulting in an effective radial adhesion force

Fadh. The contact angle θ and the resultant capillary force Fcap established the depth of penetration at the water-air interface. The net force between Fadh and Fcap as represented by the penetration angle ψ illustrates the interplay between bacterial hydrophilicity and lipid adhesion. The rod-shaped bacteria at this TPL can add additional complexity due to further deformations induced in the fluid-fluid interface...... 136

Figure 7.11. (A) Density of attached bacterial cells as inferred from confocal analysis and (B) Surface wettability of eicosane, docosane, S. aureus and P. aeruginosa. The S. xxiv aureus cells appear to more readily attach onto the eicosane, whereas the P. aeruginosa cells more readily attach onto the docosane...... 138

Figure 8.1. The molecular structures of palmitic (left) and stearic (right) acids, and the SEM morphologies of their corresponding microcrystallites. Palmitic acid self- assembled to form thin microblades, whilst stearic acid appeared to form interjoined pillars. The differences in the sharpness of the features formed by the two fatty acids are apparent in the 3D images...... 143

Figure 8.2. The topography of palmitic acid (A) and stearic acid (B) self-organised microcrystals on HOPG templates, and their corresponding cross-sectional profiles. From the AFM images it can be seen that palmitic acid crystallised into relatively parallel features whilst the stearic acid self-assembled into structures pointing in various directions. The line profile of palmitic acid demonstrated the sharp edges of each crystallite, whilst stearic acid features consisted of flattened high peaks...... 144

Figure 8.3. Morphological analysis of the crystallites of (A) palmitic acid, and (B) stearic acid. These data were obtained from particle size analysis using Gwyddion data processing software (Nečas and Klapetek, 2012). Any particles below 200 nm in width were excluded to minimise the effects of noise...... 146

Figure 8.4. FFT analysis of scanning electron micrographs of palmitic acid and stearic acid microcrystallites. Some directionality was observed in palmitic acid (white ovals) but the stearic acid microcrystallites appeared to orient somewhat randomly...... 147

Figure 8.5. Adsorption alignment of straight chain carboxylic acid molecules self- assembled into monolayers on graphite. The alignments shown are all in zigzag configuration. The adjacent chains within one domain will have their –COOH group pointing either in the same direction or opposite directions. Reproduced with permission from AIP Publishing LLC (Yang, Berber, Liu et al., 2008)...... 148

Figure 8.6. Chemical composition of FA interfaces. (A) FTIR spectra of palmitic acid and stearic acid obtained in mapping mode, scan area 50 µm × 50 µm; and the 2D spectral maps of the C-H stretching bands of (B) palmitic acid, (C) stearic acid, and the C=O stretching band of stearic acids...... 149

xxv

Figure 8.7. Peak shift of the C-H stretching band of stearic acid. (A) Peak shift spectra, (B) the 2D map of C-H stretching band, (C) the SEM image of stearic acid, and (D) the 2D map indicating the relative position of the peak maximum...... 150

Figure 8.8. Bactericidal activities of palmitic and stearic fatty acids crystallite interfaces. Confocal Laser Scanning Microscopy (CLSM) images show the viability of bacterial cells on each of the fatty acids. Dead cells are indicated in red, viable cells in green. There was relative large number of red signals in all cases (particularly S. aureus on palmitic acid). Scale bars = 5µm...... 151

Figure 8.9. Bactericidal efficiency of the two surfaces against P. aeruginosa and S. aureus cells. In most cases, the surfaces succeeded in killing greater than 60% of the bacteria, except in the case of S. aureus on the surface of stearic acid, which killed 49%...... 152

Figure 8.10. Bactericidal activities of the gold-coated surfaces. Both surfaces showed strong bactericidal activities against both P. aeruginosa and S. aureus. Dead cells are indicated in red, viable cells in green. Scale bar = 5µm...... 154

Figure 8.11. Killing efficiency of palmitic and stearic acid interfaces against Pseudomonas aeruginosa (Pa) and Staphylococcus aureus (Sa). In most cases, the bactericidal efficiency increased noticeably on the gold-coated surface compared to the native surfaces...... 155

Figure 8.12. Attachment rates of Pseudomonas aeruginosa (Pa) and Staphylococcus aureus (Pa) on palmitic and stearic acid interfaces. The number of attached cells increased substantially as observed under confocal microscopy. This increase in number of cells is most likely due to enhancement of fluorescent signals on metal surface (gold-coated)...... 155

Figure 8.13. Water contact angles on native fatty acids and after coating with a thin layer of gold...... 157

Figure 8.14. Raman spectra of palmitic acid and stearic acid crystallites immersed under water for air imaging...... 159

Figure 8.15. Image scans of fatty acid (FA) crystallites and water using Raman microspectroscopy. The regions, which are darker in both images, indicated the presence of air pockets...... 159

xxvi

Figure 8.16. Schematic representation of forces applied to a S. aureus cell ‘riding’ the ridge of a palmitic acid microblade. The wetting force of the three phase lines (Ftpl) on either side of the microblade pull the cell in opposing directions, however the component of each force normal to the microcrystal interface (Fnor) are cumulative. The lateral force components (Flat) are 180° opposed to each other, and may aid in tearing the cell...... 160

Figure 9.1. Proposed model of the epicuticle of dragonfly wing membranes. Three layers are contained within the epicuticle: the outer-epicuticle, the meso-epicuticle, and the inner-epicuticle...... 166

Figure 9.2. Photographs of H. papuensis (left) and D. bipunctata (right) dragonflies in their natural habitats, and their corresponding geographic distributions. The yellow spots in the world map indicate their general global distributions whilst colour coding of the Australian map indicates their common and preferable habitats, based on surveys of the number of dragonflies present at the sampled locations. Reproduced with permission from the Encyclopedia of Life (2014) and Atlas of Living Australia (National Research Infrastructure for Australia (NCRIS), 2014)...... 168

xxvii

List of Tables

Table 2.1. Different types of solvents used for extraction of the lipid components from the insect body and insect wings surfaces...... 15

Table 2.2. Micro- and nano-scale wax crystal morphologies on the epidermal cells of insect wing surfaces and their WCA ...... 28

Table 4.1. Assignment of chemical components to the major absorption bands present in the IR spectra of the wing membranes ...... 66

Table 4.2. Wax components of the Hemicordulia tau wing epicuticle isolated and identified by GCMS. Components that could not be unambiguously identified are grouped together...... 73

Table 4.3. Roughness analysis of the untreated and chloroform extracted H. tau wing surfaces...... 78

Table 5.1. Roughness analysis of the untreated and chloroform extracted Hemianax papuensis wing surfaces ...... 88

Table 5.2. Individual lipid components of the Hemianax papuensis wing epicuticle isolated by extraction with chloroform, identified by GCMS. Components that could not be unambiguously identified are grouped together...... 91

Table 6.1. Roughness analysis data for the two dragonfly wing surfaces, H. papuensis and D. bipunctata dragonflies...... 113

Table 6.2. Bearing statistics of the H. papuensis and D. bipunctata dragonfly wing surfaces...... 114

Table 7.1. AFM roughness analysis of the graphite and silicon wafer surfaces ...... 126

Table 7.2. AFM roughness analysis of the eicosane and docosane crystals on HOPG surfaces ...... 131

Table 8.1. AFM roughness analysis of the eicosane and docosane crystals on HOPG surfaces ...... 145

xxviii

Chapter 1. Introduction

______1

1.1. Overview

It is well-documented that biofilm formation on medical implants by pathogenic bacteria can have dramatic consequences, leading to the failure of the device, often resulting in the necessity to surgically remove the implant. Subsequent complications that can then arise include systemic infection, loss of organ or limb function, amputation or death (Ong, Razatos, Georgiou et al., 1999; Donlan, 2001; Diaz, Cortizo, Schilardi et al., 2007). The problems are exacerbated by the enormous number of resistant strains that have emerged as the result of the heavy chemical input in current sterilisation methods. For these reasons, there is a need for innovative approaches and alternatives for controlling bacterial colonisation.

Natural occurring surfaces have always provided great inspiration for innovative technologies that can be used in a wide range of potential applications (Watson, Cribb and Watson, 2010; Watson, Cribb and Watson, 2011). These surfaces have evolved and have been optimised for millions of years in order to adapt to ever-changing environments (Barthlott and Neinhuis, 1997; Barthlott, Neinhuis, Cutler et al., 1998; Cheng, Rodak, Wong et al., 2006; Byun, Hong, Saputra et al., 2009; Koch, Bhushan, Ensikat et al., 2009; Boeve, Voigt and Gorb, 2011; Hu, Watson, Cribb et al., 2011). Nature is therefore an excellent resource for the development of innovative technologies that can be applied widely in the related fields of chemistry, medicine and material science. During the past few decades, the ‘lotus effect’ has arisen as the archetype for surfaces that possess highly important properties, such as superhydrophobicity, and the ability to self-clean (Yoshimitsu, Nakajima, Watanabe et al., 2002; Sun, Feng, Gao et al., 2005; Dani, 2006; Watson, Cribb and Watson, 2010).

The term ‘self-cleaning’ originates from the combination of superhydrophobicity and a low water droplet sliding angle (Sun, Feng, Gao et al., 2005). Furthermore, superhydrophobicity arises as a result of the combination of the surface chemistry and the surface roughness, with small changes in either property having the potential to significantly influence the extent of surface wettability (Barthlott and Neinhuis, 1997; Bhushan and Jung, 2008; Bhushan, Jung and Koch, 2009b; Ensikat, Ditsche-Kuru, Neinhuis et al., 2011; Watson, Cribb and Watson, 2011). A great deal of effort has been made in obtaining a better understanding of these phenomena in order to adapt them to dealing with current issues surrounding the formation of biofilms (Bico, Marzolin and

Quéré, 1999; Herminghaus, 2000; Öner and McCarthy, 2000; Nakajima, Hashimoto and Watanabe, 2001; Lafuma and Quéré, 2003; Patankar, 2004; Ma and Hill, 2006; Watson, Cribb and Watson, 2010). The concepts of self-cleaning and superhydrophobicity, whilst being closely related, are distinct phenomena that are not usually successfully distinguished in the literature. This shortcoming has led to confusion between the separate concepts of superhydrophobicity and self-cleaning. Fürstner and Barthlott were the first to assert that superhydrophobicity and self-cleaning were two different phenomena (Fürstner, Barthlott, Neinhuis et al., 2005). A self-cleaning surface must possess superhydrophobicity but a self-cleaning surface may not necessarily exhibit superhydrophobicity. Therefore, when designing self-cleaning materials, superhydrophobicity is an important consideration.

A number of insects and plants contain surfaces that exhibit both superhydrophobicity and self-cleaning properties. While a number of studies have been conducted on the surface properties of plant leaves to obtain an insight into the mechanisms responsible for these properties, this is not the case for the surfaces of insects. For example, while the surface chemistry of insects has been investigated for many years from an agricultural perspective, these studies did not have a focus on surface wettability. Many of these insect surfaces are also bactericidal, the mechanisms of which are purely physical. This recently obtained knowledge warrants further study into the ways by which these surfaces kill bacteria and may provide an insight into the mechanisms by which the formation of biofilms on surfaces may be controlled. It may then be viable to mimic these environmentally-friendly processes that do not require the use of chemical synthesis or fabrication techniques.

1.2. Aims of the study

The ultimate aim of this work is to successfully construct an artificial surface that is based on a naturally occurring insect wing that can kill bacterial cells via a process of mechanical stresses. As mentioned above, controlling biofilm formation and bacterial infections would be of great benefit to society. In order to achieve this aim, three intermediate objectives are identified as key stages toward making such a surface capable of acting to prevent bacterial proliferation and infection.

The first objective is to physically and chemically characterise the surfaces of superhydrophobic various insect wings. This involves a number of complementary

3 methods including atomic force microscopy (AFM), scanning electron microscopy (SEM), gas chromatography and mass spectrometry (GCMS), x-ray photoelectron spectroscopy (XPS) and Fourier transform infrared microscopy (FTIR) to investigate the composition, architecture and morphology of the wings of several species of Odonata in order to obtain a general overview of the interspecies consistency. This information provides clearer insight into how the surfaces of dragonfly wings are structured.

The second objective is to assess the bactericidal potential of different dragonfly wings. Several different strains of pathogenic bacteria will be used for this assessment. Primarily, three techniques will be used, including SEM, confocal laser scanning microscopy (CLSM) and plate counting to quantify the efficiency of each wing in controlling the extent of bacterial attachment. This is an important consideration as it provides an insight into the mechanisms by which the bacteria are able to be killed when coming into contact with the surface. This will be of great benefit when designing antibacterial surfaces.

The third objective is to design, fabricate and evaluate synthetic surfaces that are constructed according to the knowledge generated from the previous stages. The intermediate aim is then to prepare surfaces that contained ordered structures that can control bacterial attachment, using fast and facile methods such as self-assembly. A number of the surface characteristics that were identified in the initial stages of this research will be used to produce self-assembled surfaces that are able to limit the extent of bacterial adhesion. The knowledge generated regarding the mechanisms by which the bactericidal action takes place represent an important guide for designing surfaces that can control the formation of biofilms.

In the following chapters, the current state of knowledge regarding the fields of superhydrophobicity, self-cleaning, insect wings and anti-bacterial surfaces will be discussed. Following this discussion, chapters discussing the surface characteristics and properties of dragonfly wings will be presented, along with an analysis of the mechanisms whereby these surfaces are able to control bacterial attachment. Finally, the possible impact of the knowledge generated here and the future perspectives of this work will be addressed.

Chapter 2. Literature review

______5

2.1. Overview

Surfaces with superhydrophobic and self-cleaning properties have gained much attention during the past few decades because of their high potential as templates in the design of biomimetic artificial surfaces and in material sciences where products and devices that can exhibit similar properties (Yoshimitsu, Nakajima, Watanabe et al., 2002; Sun, Feng, Gao et al., 2005; Dani, 2006; Watson, Cribb and Watson, 2010). Such surfaces have applications that range from daily life applications, such as self-cleaning glass, anti-fouling coatings and oil/water separation, to high-technology devices, such as anti-wetting and lab-on-a-chip devices. Insect wings are one of many examples of natural surfaces that possess superhydrophobic and self-cleaning properties, and as such they have attracted great attention from the scientific community interested in this wide range of potential applications.

Surfaces found in nature have been optimised through the process of evolution for millions of years. Insects have evolved the ability to fly for at least 400 million years (Grimaldi and Engel, 2005) while they represent approximately half of all the living macro-organisms on earth. Insect wings are composed of lightweight materials, which can withstand extremely heavy loadings with minimal resource requirements (Wan, Cong, Wang et al., 2008). In order to adapt to an environment that provides little protection against wetting by rain and other aqueous sources that they may encounter, insects evolved wings into geometric, non-smooth structures (Wan, Cong, Wang et al., 2008; Watson, Cribb and Watson, 2011). These structures are comprised of a range of different waxes, which impart anti-wetting and self-cleaning properties as the result of their superhydrophobic properties (Buckner, Nelson and Mardaus, 1994; Gibbs, 1998; Buckner, Hagen and Nelson, 1999; Nelson, Freeman and Buckner, 2000; Arsene, Schulz and Van Loon, 2002; Nelson and Charlet, 2003; Boeve, Voigt and Gorb, 2011). It is now generally accepted that the impenetrability of an insect to water is due to the presence of a thin, superficial layer of wax in the cuticle as originally reported by Ramsay in 1935 (Ramsay, 1935).

2.2. Arthropod cuticle

2.2.1. Morphology of insect wings

Insects were the first organism that developed the ability to fly and took to the skies at least 90 million years prior to the earliest winged vertebrates (Grimaldi and Engel, 6

2005). Currently, flying insects are organised in the superorder Pterygota. Insect cuticles consist of two main components: a crystalline polymer, chitin, embedded in fibrillar form in a matrix (Kreuz, Arnold and Kesel, 2001). This arrangement builds up three distinct layers of insect wing cuticle: the epicuticle, the endocuticle and the exocuticle (Xiao, Bai, Wang et al., 2007; Wan, Cong, Wang et al., 2008). The cuticle is secreted by a single layer of epidermal cells, the hypodermis (Lockey, 1988), and transported to the cuticle surface by a system of pore canals (Gorb, 1997). The epicuticle is the thinnest and outermost layer at the dorsal and ventral surfaces, being less than 1 µm in thickness. According to Lockey, the epicuticle may consist of two thinner layers: outer epicuticle and inner epicuticle (Lockey, 1988). The outer epicuticle (cuticulin layer) is about 10 to 18 nm thick while the inner layer is where the waxy components are formed. The components of these waxes vary depending on the species and their living environment. Generally, cuticle waxes are composed of varying proportions of fatty alcohols, fatty acids, fatty aldehydes, esters, steroids, long chain hydrocarbons and sometimes aromatic compounds (Lapointe, Hunter and Alessandro, 2004). The endocuticle and the exocuticle together can be considered as one layer, known as the procuticle (Wootton, 1992; Wagner, Neinhuis and Barthlott, 1996; Nelson, Freeman, Buckner et al., 2003; Vukusic, Wootton and Sambles, 2004; Jarrold, Moore, Potter et al., 2007; Gołebiowski, Maliński, Boguś et al., 2008; Gorb, Tynkkynen and Kotiaho, 2009; Gołebiowski, Bogus, Paszkiewicz et al., 2011). The endocuticle (middle layer of the cuticle) is where proteins are sclerotized whereas the exocuticle, the inner most layer, consists of chitin and occupies about 25 % - 40 % of insect cuticle by dry-weight (Wan, Cong, Wang et al., 2008).

2.2.2. Aerodynamics of insect flight

Insects are one of the most wonderful organisms on Earth. They are abundant yet very elusive, they are fascinating to watch but difficult to obtain. Their flying mechanisms are combinations of various motions that are assisted by the sophisticated architecture of their wings. Their wings are framed by a system of veins which aid in stabilising the wing as a whole (Kreuz, Arnold and Kesel, 2001). Its basic framework consists of chitin, in the form of a long chain crystalline polymer providing support for the membrane and bearing the forces applied to the wings during flight (Kesel, Philippi and Nachtigall, 1998; Rajabi, Moghadami and Darvizeh, 2011). The junctions between the vein and the wing membrane are comprised of the protein resilin, which enhances 7 the flexibility of the wing (Andersen and Weis-Fogh, 1964; Neville, Parry and Woodhead-Galloway, 1976; Gorb, 1999). The vein system of the wings not only provides support for the membrane but also bears a majority of the bending and twisting forces applied to the wings. There are three different types of veins existing in the wings, including ambient, longitudinal and cross veins. The ambient veins function as the stabilising framework and surround the wing, while longitudinal veins act as girders which divide the wings into small cells (Rajabi, Moghadami and Darvizeh, 2011). This venation system, together with their light-weight building materials supports routine flights as well as longer colonisation flights (Song, Xiao, Bai et al., 2007).

The main features of the flapping motions seen in most insects include back and forth strokes along the horizontal plane, an incline stroke plane or up and down strokes against wind (Wang, 2005). In order to support their body weight, insect wings typically produce 2-3 times more lift than conventional aerodynamics (Ellington, 1999). To minimise the materials needed to form the wing but still be able to protect themselves from wetting and pollutants, insect wing surface membranes are optimised to display a disordered, rough microstructure, composed of numerous nanometre-scale columns (Song, Xiao, Bai et al., 2007; Wan, Cong, Wang et al., 2008). Researchers have been long fascinated by the excellent flying abilities of insect wings despite their limitations in size and weight. Among all species of insects, dragonflies and damselflies belonging to the Order Odonata are unsurpassed in the insect world for flying abilities. Just as the wings of the airplane are the determining factor in their aerodynamics, dragonfly wings determine their flying ability (Zhao, Yin and Zhong, 2011).

Odonata represents the Order that contains one the oldest forms of insect flight in which their two pairs of wings can beat independently. For this reason, the wings provide the insect the availability of various motions which mark their wonderful signature in nature. Even though damselflies and dragonflies are very similar in size, their wings are constructed completely differently. This enables dragonflies to become predators rather than prey, which is also the case for some species of damselflies (Wakeling and Ellington, 1997). Dragonfly wings were designed in a way that allows their stroke planes to be nearly normal to the direction of the resultant force, which makes their velocities and accelerations greater than those of damselflies (Wakeling and Ellington, 1997).

Studies of the aerodynamic behaviour of insect wings have provided valuable ideas for micro-air vehicles. These flying abilities also provide new instrumental advances at the nano- and micro-scale enables characterisation of more valuables properties of insect wings. For example, the intricate hierarchical nanoarchitecture of insect wings can be used to increase the water contact angle of incoming droplets and/or reduce adhesion on the surface during rainy conditions that enhance their resistance against the wetness of the semi-aquatic environments in which some insects reside (e.g. cranefly and water-striders) (Cheng, 1973; Suter, Rosenberg, Loeb et al., 1997; Hu, Chan and Bush, 2003; Watson, Myhra, Cribb et al., 2008; Rothstein, 2010; Su, Ji, Zhang et al., 2010; Watson, Cribb and Watson, 2010; Hu, Watson, Cribb et al., 2011; Watson, Cribb and Watson, 2011). These structures can often contribute to colouration exhibited by some insects, such as the iridescence in the hindwings of the damselfly Neurobasis chinensis chinensis, the blue iridescence in the wings of the giant tropical wasp Megascolia procer javanensis, and the transparent and translucent wings of butterflies (Vukusic, Wootton and Sambles, 2004; Hooper, Vukusic and Wootton, 2006; Sarrazin, Vigneron, Welch et al., 2008; Perez Goodwyn, Maezono, Hosoda et al., 2009; Shevtsova, Hansson, Janzen et al., 2011). This colourisation phenomenon is caused by the unique photonic properties of insect wings arising as a result of the combination of ‘chemical’ and ‘physical’ colours (Huang, Wang and Wang, 2006).

2.3. The surface of insect wings

2.3.1. Concept of self-cleaning

During the past decade, lotus leaves have become the benchmark for superhydrophobic and self-cleaning surfaces (Ensikat, Ditsche-Kuru, Neinhuis et al., 2011). The ‘lotus effect’ is the phenomenon in which water droplets form pearl-like shapes and move freely across the surface of the leaf, collecting dirt and contaminants as they roll off the surface (Barthlott and Neinhuis, 1997; Riedel, Eichner and Jetter, 2003; Koch, Neinhuis, Ensikat et al., 2004; Sun, Feng, Gao et al., 2005; Cheng, Rodak, Wong et al., 2006; Koch and Ensikat, 2008; Koch, Bhushan, Ensikat et al., 2009; Rothstein, 2010; Buschhaus and Jetter, 2011; Ensikat, Ditsche-Kuru, Neinhuis et al., 2011). According to concept of the ‘lotus effect’, self-cleaning is explained as the result of the presence of superhydrophobicity and low sliding contact angle (Feng, Li, Li et al., 2002; Sun, Feng, Gao et al., 2005). For example, Nelumbo nucifera (lotus) and

Colocasia esculenta (taro) leaves exhibit water contact angles (WCA) of θ ≈ 160º. These leaves have been the subject of various applications in the fabrication of biomimetic artificial surfaces.

The majority of species of the Odonata Order inhabit environments which are both wet and dirty, therefore needed to evolve strategies for coping with potential fouling particles, in the form of superhydrophobic and self-cleaning surfaces (Wagner, Neinhuis and Barthlott, 1996; Simonsen, 2001; Koch, Dommisse and Barthlott, 2006; Byun, Hong, Saputra et al., 2009; Sun, Watson, Zheng et al., 2009; Wan, Cong, Jin et al., 2009; Watson, Watson, Hu et al., 2010; Hu, Watson, Cribb et al., 2011; Watson, Cribb and Watson, 2011). Recently, cranefly and water-striders have been receiving increased attention, as they have the ability to walk freely on water without penetrating the surface due to the presence of fine structures of their cuticle (Hu, Watson, Cribb et al., 2011). Other biological surfaces, such as duck feathers and butterfly wings, contain corrugated surfaces that form air pockets that prevent water from touching the surface (Nosonovsky and Bhushan, 2005; Nosonovsky and Bhushan, 2007b).

2.3.2. Wettability theory

When a surface exhibits a high water contact angle (WCA > 150°) and low contact angle hysteresis (CAH < 10°), the surfaces are regarded as being not only superhydrophobic but also self-cleaning (Shirtcliffe, McHale, Atherton et al., 2010; Su, Ji, Zhang et al., 2010; Bhushan, 2012). CAH is the difference between the advancing and receding contact angles (Nosonovsky and Bhushan, 2008; Ferrari and Ravera, 2010). Superhydrophobicity occurs when the water droplets form an almost spherical shape on the surface and therefore will readily move when subjected to low tilting angles, resulting in self-cleaning. Any dirt or contaminants on the surface are then collected by the droplets and are removed as the droplet rolls off the surface. This characteristic is usually associated with low-drag (Bhushan, 2012) and low adhesion properties (Mei, Luo, Guo et al., 2011). Depending on the structure and chemical composition of the surface, antibiofouling (Webb, Truong, Hasan et al., 2012), anisotropic wetting (Nishimoto and Bhushan, 2013), anti-reflection (Chattopadhyay, Huang, Jen et al., 2010) and anti-icing (Mei, Luo, Guo et al., 2011) properties may also be present.

Whilst Wenzel’s theory explains that surface hydrophobicity is a function of surface roughness as well as chemical composition (Wenzel, 1936), it is the theory developed by Cassie and Baxter (Cassie and Baxter, 1944) that is most often applicable for superhydrophobic surfaces. The Cassie-Baxter model also considers the topography and chemistry of a surface in determining the hydrophobicity. According to this model, surface hydrophobicity is a function of surface chemical heterogeneity according to the following equation: cos θ = f1cos θ1 + f2cos θ2 Equation 2.1. where θ is the observed contact angle of the heterogeneous surface, f1 and f2 are the area fractions of surface components 1 and 2, with θ1 and θ2 being their respective contact angles. In the case of superhydrophobic surfaces, one of the components is typically air, which exhibits a water contact angle of 180°. This allows the simplification of equation (1) to: cos θ = f1(cos θ1 + 1) – 1 Equation 2.2.

Superhydrophobicity arises from the combination of hierarchical surface structures (i.e. surfaces with multiples scales of roughness) that enable the entrapment of air on low surface energy materials. The sliding angle, the other parameter that is important in determining hydrophobicity, is defined as the critical angle at which the water droplets start to slide along a tilted surface (Jung and Bhushan, 2006; Bhushan, Jung and Koch, 2009b; Yan, Gao and Barthlott, 2011). A superhydrophobic surface would effectively form a composite interface (CI) with air residing between the asperities on the surface. The CI can be destabilised and irreversibly transformed into a homogeneous interface, for example, by the application of sonication (Nosonovsky and Bhushan, 2007a). There are three factors which can destroy a CI; (i) a capillary wave formed at the liquid-air interface, (ii) nanodroplet accumulation in the valleys on the surface and (iii) hydrophilic surface regions arising from the chemical heterogeneity of the surface. Hierarchical roughness or ‘Cassie-Baxter structure’ needs to exist in order for the CI to be stabilised (Neinhuis and Barthlott, 1997; Nosonovsky and Bhushan, 2007a).

According to the Cassie-Baxter expression, any variation in the chemical composition or the topographical structure of a surface will affect the values of θ1 and f1 respectively, which will result in a change in the observed composite water contact angle (Figure 2.1). The surface illustrated in (B) possesses a different chemical composition to that of (A) and exhibits a higher water contact angle θ1. The surface shown in (C) is composed of the same material as (A), however the altered surface topography enables it to trap a greater quantity of air, decreasing f1, and causing the observed angle to approach that of air, i.e. 180°. Surface (D) is composed of a low energy material in addition to possessing the Cassie-Baxter structure, which results in a surface that exhibits an even greater water contact angle.

2.4. Surface chemistry

Three quarters of all insect species on Earth are Arthropoda (Grimaldi and Engel, 2005) and their existence has always directly or indirectly affected farming and food production. Therefore, studies on insect chemistry have progressed for a long period of time. Recent advances in the knowledge of insect cuticular wax chemistry have revealed that the waxes not only provide a significant barrier to protect insects from desiccation and the adverse effects of microorganisms but also function as communication cues. This chemical knowledge is relevant for a number of applications in medicine and biomedicine. Unlike plants, insects can only produce cuticular waxes once in their life time and they can neither be renewed nor added at a later date (Kreuz, Arnold and Kesel, 2001; Wan, Cong, Jin et al., 2009). The border between the outer epicuticle and the inner epicuticle is thin, so that the chemical information of individual layers is difficult to distinguish. However, with many advances in technologies, alternative methods are being developed to study epicuticle waxes.

Traditionally, insect cuticles have been subjected to extraction using various organic solvents in order to remove the waxes for analysis. There have been many studies conducted on the chemical composition of insect cuticle based on solvent extraction and these studies are summarised in Table 2.1. Depending on polarity of the targeted waxes, the solvent can be varied from mild, i.e., acetone, ethanol, hexane and heptane, to very strong solvents such as benzene and chloroform (Wan, Cong, Wang et al., 2008). Therefore, the use of different solvents will result in the extraction of different components of the waxes. For example, chloroform was described as the most

13 suitable solvent for extraction of aliphatic hydrocarbons and wax esters on the epicuticular layer of the wing surfaces. A newer technique has been developed recently for surface wax extraction, known as solid-phase micro-extraction (SPME). This method involves a single SPME fibre coated with polydimethylsiloxane (PDMS), which is rubbed directly on the surface for approximately 5 minutes to remove the waxes of the insect cuticles (Zhang, Yang and Pawliszyn, 1994; Moneti, Dani, Pieraccini et al., 1997; Liebig, Peeters, Oldham et al., 2000; Dani, 2006). The fibre is then inserted directly into injection port of the GCMS for chemical analysis. Another technique that has been suggested to be one of the best methods for surface wax extraction is silica- rubbing, since it allows the greatest recovery of a majority of the epicuticular waxes (Choe, Ramírez and Tsutsui, 2012). The targeted surface, which is placed with silica gel and subsequently vortex for 30 s, is removed from the gel and the gel is then extracted with hexane (Choe, Ramírez and Tsutsui, 2012).

Table 2.1. Different types of solvents used for extraction of the lipid components from the insect body and insect wings surfaces.

Comments Dielectric Extraction Solvents Insect body Insect wings References constant time Chemistry Physical General

Damselflies Pyrrhosoma 1,2- (Gorb, Kesel nymphuls and Coenagrion No specific chemistry was Destroyed up to Propylene 1h 70% area dissolved and Berger, puella; Dragonfly Aeshna identified 30% surface pillae oxide 2000) cyanea

(Geiselhardt, Leaf beetle HCs* – n- and branched To remove pillae No morphology change Geiselhardt n-Pentanea 1.84 Gastrophysa 15 m alkanes, alkenes (C16 - Odd number of C in was reported and Peschke, viridula C37) backbone skeleton 2009)

Goldenrod gall Branched alkane (C28- To extract HCs* on (Nelson and fly Eurosta 1 m C32). Long chain Alds† epicuticular waxes Lee Jr, 2004) solidaginis (C30, C32)

Saturated homologous n- Identified Booklouse (C21-C34), mono- (C28-C42) epicuticular lipids: (Howard and Liposcelis 30 s and dimethyl-alkanes (C31- HCs* and FA# Lord, 2003) bostrychophila C43) amides (stearoyl No morphology change Hexanea 1.89 # amides) FA amides (C16-C22) was reported

Homologous alkanes (C21- C45) Ectoparasitoid Wax ester: FAs# - short To extract (Howard and Habrobracon 3 m chain, unbranched, even epicuticular HCs* Baker, 2003) hebetor numbered carbons (C8- and wax esters ‡ C20), with secondary Alcs moieties (C22-C25)

Comments Dielectric Extraction Solvents Insect body Insect wings References constant time Chemistry Physical General

To dissolve (Wan, Cong, Dragonfly Pantala No specific chemistry was Dissolved 30-40% of 2 h insoluble Wang et al., flavescens fabricius identified surface area compounds 2008) Heptane 1.92 Damselflies Pyrrhosoma (Gorb, Kesel nymphuls and Coenagrion No specific chemistry was Dissolved 20% of area, Affected 70% wax 1 h and Berger, puell; Dragonfly Aeshna identified 10% areas looked intact covering 2000) cyanea

Toxic organic solvent dissolve ester Dragonfly Pantala No specific chemistry was Dissolved 50-60% of Benzene 2.28 2 h flavescens fabricius identified surface structure Very non-polar Be able to remove HCs*

Damselflies Pyrrhosoma (Gorb, Kesel nymphuls and Coenagrion 1 h and Berger, puella; Dragonfly Aeshna Very strong solvent. 2000) cyanea No specific chemistry was Chloroform 4.81 100 % of area dissolved Treated for 2 h was identified too long. (Wan, Cong, Dragonfly Pantala 2 h Wang et al., flavescens fabricius 2008)

Periplaneta To extract brunnea * (Saïd, P. americana HCs* ranged epicuticular HCs Dichloro- No morphology change Costagliola, d 9.08 P. fuliginosa 2 m C -C while other three * methane 24 43 was reported C19-C25 HCs Leoncini et al., species ranged C -C . P. australasiae, 21 41 profile is species 2005) P. american specific

Comments Dielectric Extraction Solvents Insect body Insect wings References constant time Chemistry Physical General

Damselflies Pyrrhosoma (Gorb, Kesel 1,2-Dichloro nymphuls and Coenagrion No specific chemistry was 5% of area dissolved, Can destroy up to 10.42 1 h and Berger, ethane puella; Dragonfly Aeshna identified 40% area looked intact 50% surface pillae 2000) cyanea

Damselflies Pyrrhosoma (Gorb, Kesel nymphuls and Coenagrion Pillae appeared 1 h and Berger, puella; Dragonfly Aeshna amorphous Least destroying 2000) cyanea effects. No specific chemistry was Acetone 20.7 (25) Relative polar, not identified Subtle changes on very effective for (Wan, Cong, Dragonfly Pantala surface non-polar 2 h Wang et al., flavescens fabricius compounds Pillae appeared non- 2008) uniform

Damselflies Pyrrhosoma (Gorb, Kesel nymphuls and Coenagrion 40% of area dissolved, Relative polar, not 1 h and Berger, puella; Dragonfly Aeshna 10% looked intact very effective to non-polar 2000) cyanea No specific chemistry was Ethanol 24.6 compounds. identified Subtle changes Unmatched results (Wan, Cong, Dragonfly Pantala 2 h obtained with Wang et al., flavescens fabricius Pillae appeared different species amorphous 2008)

Damselflies Pyrrhosoma 30% of area looked More polar solvent (Gorb, Kesel nymphuls and Coenagrion No specific chemistry was intact Methanol 32.6 (25) 1 h Unmatched results and Berger, puella; Dragonfly Aeshna identified 70% was substantially obtained with 2000) cyanea destroyed different species

(Wan, Cong, Dragonfly Pantala 15-20% of the area 2 h Wang et al., flavescens fabricius dissolved 2008)

Comments Dielectric Extraction Solvents Insect body Insect wings References constant time Chemistry Physical General

Saturated and unsaturated Silverleaf # whitefly FA with even numbered Hexanec/CH To remove Bemisia 1-2/ carbons( C14-C20) No morphology change (Buckner and Cl :MeOH epicuticular lipids 3 argentifolii was reported Hagen, 2003) (2:1) 30 m Phosphatidylethanolamine and surface pillae Belows and (PE), phosphatidylcholine Perring (PC)

Saturated wax ester (C44- (Nelson, Giant whitefly C60) Freeman and Aleurodicus † ‡ Buckner, dugesii Long chain Alds and Alcs (C30) 2000)

1.5/0.5 m Alds† and Alcs‡ with even Hexane used for * numbered carbons (C26- HCs and wax esters Giant whitefly C32) on epicuticular layer (Nelson, Aleyrodes Guershon and Hexanec/ Wax esters: even numbered No morphology change Chloroform singularis Gerling, 1998) Chloroformb carbons FAs# (C16-C24), was reported removed and Alcs‡ (C26, C28) intracuticular waxes, i.e., long chain Alds† and Even carbon numbered wax ‡ esters (C38-C64) Alcs Silverleaf (Buckner, whitefly Odd numbered carbons n- 1/1 m Hagen and Bemisia alkanes (C25-C35) Nelson, 1999) argentifolii Even carbon numbered Alds† and Alcs‡ (C28-C34)

Comments Dielectric Extraction Solvents Insect body Insect wings References constant time Chemistry Physical General

Adult whitefly Even numbered carbons † ‡ (Buckner, Bemisia tabaci Alds and Alcs (C30 - Nelson and and 1/1-2 m C34) Mardaus, Trialeurodes Even numbered carbons 1994) vaporariorum wax esters (C38-C54) Hexane used for * Free FAs# (C14-C28) HCs and wax esters on epicuticular layer (Nelson, n-, and branched alkanes Dustywing Freeman, Hexanec/ 1.5/ 0.5 m (C23-C33) No morphology change Chloroform Semidalis flinti Buckner et al., Chloroformb was reported removed Even carbon numbered wax intracuticular 2003) esters (C34-C44) waxes, i.e., long chain Alds† and Methyl-branched alkanes Alcs‡ (C25-C53) Potato beetles (Yocum, Leptinotarsa 2 m/ 45 s Triacylglycerol (C50, C52 Buckner and decemlineata and C54) with even carbon- Fatland, 2011) numbered acid isomers (C16, C18)

Comments Dielectric Extraction Solvents Insect body Insect wings References constant time Chemistry Physical General

Dried-bean Even carbons numbered (Gołebiowski, beetle FAs#, Alds†, Alcs‡ (C16- Malinski, 10 s/ 5 m Acanthoscelides C18); triacylglycerols and Nawrot et al., obtectus sterol Petroleum ether 2008) used to minimise dissolution of Calliphora intracuticular lipid vicina, (free FAs#, (Gołebiowski, Petroleum Dendrolimus Saturated and unsaturated e glycerides) which Maliński, ether / pini and 10 s/ 5 m C5-C20 in alkyl chain of No morphology change removed by Boguś et al., Dichloro- Galleria FAs#. was reported f dichloromethane 2008) methane mellonella larvae High contents of FA# were explained # FAs C15-C34 found in as the result of anti- (Gołebiowski, Dendrolimus petroleum ether fraction fungal infections. Bogus, 10 s/ 5 m pini FAs# C8-C34 found in Paszkiewicz et dichloromethane fraction al., 2010) a: 0.5mL-1mL, b: 4-8mL, c: 5-10mL, d: 50mL, e: 150mL, f: 200mL *HCs: Hydrocarbons, #FAs: Fatty acids, †Alds: Aldehydes, ‡Alcs: Alcohols The polarity of a solvent reflects the balance between the polar and non-polar components of a single molecule, and solvents ranging in polarity can be used to extract the corresponding wax components from the wing surfaces. High polarity solvents will dissolve polar compounds and low polarity solvents dissolve low polarity compounds. There are different parameters for classifying the polarity of a solvent, i.e., dipole moment, dielectric constant and miscibility with water. The greater the dielectric constant, the greater the solvent polarity (water = high, gasoline = low).

Insect cuticular waxes are composed of an intricate mixture of lipids, including wax esters, triglycerides, hydrocarbons and other derivatives of aliphatic hydrocarbons (Lockey, 1959; Soliday, Blomquist and Jackson, 1974; Lockey, 1980; Lockey, 1985; Buckner, Mardaus and Nelson, 1996; Nelson, Fatland, Buckner et al., 1999; Lapointe, Hunter and Alessandro, 2004; Buschhaus and Jetter, 2011; Gołebiowski, Bogus, Paszkiewicz et al., 2011; Yocum, Buckner and Fatland, 2011). Advances in analytical chemistry techniques (GC-MS, FTIR, etc.) have enabled studies on the chemistry of surfaces to become more accessible. Many previous studies have reported that more than 40% of the total waxes extracted from whiteflies are wax esters, except for the exuviae of Aleyrodes singularis, Bemisia tabaci, Aleurodicus dugesii and Trialeurodes vaporariorum, which have long chain aldehydes and long chain alcohols as the main components (Buckner, Nelson and Mardaus, 1994; Nelson, Buckner and Fatland, 1994; Nelson, Walker, Buckner et al., 1997; Nelson, Guershon and Gerling, 1998; Buckner, Hagen and Nelson, 1999; Nelson, Fatland, Buckner et al., 1999; Nelson, Freeman and Buckner, 2000; Buckner and Hagen, 2003). In the case of beetles and grasshoppers, long chain hydrocarbons are the key cuticle constituents (Soliday, Blomquist and Jackson, 1974; Gołebiowski, Malinski, Nawrot et al., 2008; Geiselhardt, Geiselhardt and Peschke, 2009). The cuticles of some dustywings and booklouse species are mainly composed of fatty acids (Howard and Lord, 2003; Nelson, Freeman, Buckner et al., 2003). Recently, terpenes were detected on the surface of Acanthoscelides obtectus (Gołebiowski, Malinski, Nawrot et al., 2008). The length of the carbon chain in these terpenes varies from less than 10 to up to 30-35 carbon atoms. In the case of esters, some species can have up to 54 carbon atoms in their wax components. Insects have special mechanisms to adapt to tough environmental conditions (Van Dooremalen and Ellers, 2010). For example, the goldenrod gall fly Eurosta soliginis modifies the hydrocarbon backbone of their wax components by methylation to lower the melting and transition temperature during winter, and the potato beetle Leptinotarsa decemlineata stores a large amount of triacylglycerol for energy (Nelson and Lee Jr, 2004; Yocum, Buckner and Fatland, 2011).

Numerous investigations have been conducted on the chemistry of insect surfaces for various purposes. Much of this research has reported the functionality of cuticle chemistries with a view to providing an insight into the social communication of colonised insects, including colony membership, hierarchical dominance, fertility status 21 and task group membership (Moneti, Dani, Pieraccini et al., 1997; Liebig, Peeters, Oldham et al., 2000; Endler, Liebig, Schmitt et al., 2004; Ferveur, 2005; Howard and Blomquist, 2005; Dani, 2006; Martin and Drijfhout, 2009; Choe, Ramírez and Tsutsui, 2012). Recently, fatty acids, terpenoids, and chitins were found to perform significant roles as anti-microbial agents (Kramer and Muthukrishnan, 1997; Tapia, Vallejo, Gouiric et al., 1997; Powell and Raffa, 1999; Keeling and Bohlmann, 2006; Jarrold, Moore, Potter et al., 2007; Savluchinske-Feio, Nunes, Pereira et al., 2007; Carballeira, 2008; Gołebiowski, Maliński, Boguś et al., 2008; Bogus, Czygier, Golbiowski et al., 2010; Gołebiowski, Bogus, Paszkiewicz et al., 2010; Gołebiowski, Bogus, Paszkiewicz et al., 2011; Van Dyck, Caulier, Todesco et al., 2011). Some reports have distinguished insect species within the same order based on the differences in their hydrocarbon compositions (Buckner, Nelson, Hakk et al., 1984; Bernier, Carlson and Geden, 1998; Howard and Baker, 2003; Lapointe, Hunter and Alessandro, 2004; Saïd, Costagliola, Leoncini et al., 2005). Despite the long history of studying insect surfaces, there are very few works that have expressed the chemistry of the surface in terms of the context of surface wettability.

2.5. Hierarchical structure induced self-cleaning

2.5.1. Superhydrophobicity and self-cleaning surfaces in nature

In 1936, Fogg published the observation that very high water contact angles (WCA) would form on the upper surface of the leaves of Triticum (wheat) plants, representing the first reported natural surface that possessed WCAs above 150°. This property is now commonly known as ‘superhydrophobicity’ (Fogg, 1944). Since then, nature has been the source of many valuable templates used in the design of synthetic hydrophobic materials. Shark skin, bird feathers, gecko feet, plant leaves and insects are some examples which have been shown to exhibit highly hydrophobic properties. A synthetic replica of shark skin, which formed a WCA of 146°, has been used as a model for the reduction of drag in fluid flow environments (Bhushan, 2011; Bhushan, 2012). Gecko feet (WCA of 128° at the setae) are known to exhibit reversible adhesion with surfaces (Bhushan and Sayer, 2008; Liu, Du, Wu et al., 2012; Stark, Badge, Wucinich et al., 2013). Bird and duck feathers (WCA from 114° to 126°) possess corrugated surfaces that create air pockets that prevent water from touching the surface. These have been used as a model for water repellency treatments (Cassie and Baxter, 1944;

Bormashenko, Bormashenko, Stein et al., 2007; Liu, Chen and Xin, 2008). Each of these surfaces exhibited a WCA lower than 150° and as such are not considered superhydrophobic. From the published literature, it appears that plant leaves and the surfaces of some insects are the only natural surfaces that exhibit both superhydrophobic and self-cleaning properties.

2.5.1.1. Plant surface structures

As in the case with insect surfaces, plant surfaces are also covered by a cuticular surface layer. Depending on the plant species, this layer is secreted by the epidermal cells either at certain stages of their development throughout the lifetime of the plant, or on a single occasion only. The cuticular layer consists of two compartments: a layer of epicuticlar wax crystals and a ‘cutin’ layer, which can be considered to be analogous to the epicuticle and intracuticle present on the surfaces of insects. The composition of the plant epicuticular waxes is very similar to that found with insects, whilst ‘cutin’ is comprised of a polymer of predominantly ω- and mid-chain hydroxyl and epoxy C16 and C18 fatty acids together with glycerol (Samuels, Kunst and Jetter, 2008). Some plants that demonstrate superhydrophobic surfaces are presented in Figure 2.2. The upper surface of the lotus leaf is covered by a layer of microscale papillae which is, in turn, covered by an array of nanoscale asperities (Figure 2.2A) that are composed of epicuticular hydrophobic wax tubules (Nosonovsky and Bhushan, 2007b). The WCA and CAH are 164° and 3°, respectively (Koch, Bhushan, Jung et al., 2009).

Other plant leaf surfaces that exhibit superhydrophobic and self-cleaning properties include the Indian Canna (Figure 2.2B), Taro (Figure 2.2C) and perfoliate knotweed (Figure 2.2D). Indian Canna leaves are covered by many wax platelets, distributed randomly on a series of rod-like structures. This increases the proportion of air that can be trapped within the surface, producing water contact angles in excess of the 150° contact angle condition for superhydrophobicity (i.e. 165°). The surface of taro leaves (Figure 2.2C) is composed of a layer of elliptic protrusions, which are uniformly arranged on micro-scale cave-like structures, resulting in a WCA of 159° and CAH of 24

3°. Similar surface hydrophobicity characteristics were observed for perfoliate knotweed leaves (Guo and Liu, 2007).

The examples shown in Figure 2.2 are just a few of the plants that have been identified as possessing superhydrophobic surfaces. Neinhuis and Barthlott reported the static WCAs of 200 water repellent plant species (Neinhuis and Barthlott, 1997), most of which were classified as having superhydrophobic properties. The common feature shared by these surfaces is their very dense layer of three-dimensional cuticular wax crystals arranged either randomly or uniformly on their corresponding micro-scale surface features (papillae).

The wax compositions found on plant surfaces, together with their orientation, are species specific with their characteristic composition and orientation being used in taxonomic classification (Barthlott, Neinhuis, Cutler et al., 1998; Barthlott, Theisen, Borsch et al., 2003; Koch, Bhushan and Barthlott, 2008). The morphology of the waxes can range from being irregularly placed to being arranged in highly organised structures on the surface, prompting the question: what factors are responsible for determining the three dimensional structure of the surface wax crystals? Despite the lack of information available on these factors, it has been postulated that the cutin network may play a role in controlling the orientation of the wax crystals (Jeffree, 2006).

2.5.1.2. Insect surface structures

The surfaces of insect wings have also attracted a great deal of attention from researchers due to the sophisticated structures that exist on these surfaces. Such surfaces serve to afford the wings with unique properties, such as anti-fogging and anti-wetting behaviour. Insect surfaces have also been the source of a great deal of valuable information regarding the mechanisms responsible for superhydrophobicity and self- cleaning behaviour. In the past few years, insects such as the butterfly, cicada, water strider, dragonfly and damselfly have been recognised as insects that possess these two unique properties. Their outermost layer, the cuticle, is the barrier that directly interacts with environment. This layer has been optimised through millions of years of evolution to protect the insects from ever-changing environments. The cuticle is secreted by a single layer of epidermal cells, forming a lipophilic structure (Lockey, 1980; Lockey, 1985; Nelson and Blomquist, 1995; Buschhaus, Herz and Jetter, 2007; Jetter and Kunst, 2008; Samuels, Kunst and Jetter, 2008; Buckner, 2010; Moussian, 2010). This structure

25 consists of two major components, which can be distinguished by their solubility in organic solvents (Buschhaus and Jetter, 2011). The outermost layer of the cuticle, the epicuticle, is composed of a mixture of aliphatic hydrocarbons and their derivatives; these compounds contain one or more oxygen functional groups including esters, ketones, alcohols, aldehydes and fatty acids (Koch and Ensikat, 2008; Samuels, Kunst and Jetter, 2008). The intracuticular layer, located beneath the epicuticle is a mixture of chitin (poly-N-acetylglucosamine) and protein (Lockey, 1980; Lockey, 1985; Lockey, 1988). The mixture of organic components is self-organised in the epicuticular layer of the cuticle, producing either a smooth two-dimensional (2D) wax film or a three- dimensional (3D) wax crystal structure. These structures have afforded the organisms the ability to adapt to their environmental living conditions.

The surface of dragonfly wings is covered by an array of nanopillars that form a fractal structure (Figure 2.3). This structure enables the surface to retain pockets of air within its surface when it comes into contact with water, producing a high WCA. Moreover, low CAH and water droplet sliding angles less than 10°, indicating the ability to self-clean, can be observed in many different Odonata species (Ivanova, Hasan, Webb et al., 2012; Ivanova, Hasan, Webb et al., 2013; Pogodin, Hasan, Baulin et al., 2013a). These self-cleaning properties can be further accelerated by the wings being moved in turbulent conditions during flight (Nishimoto and Bhushan, 2013).

Figure 2.3. The surface morphologies of the Hemianax papuensis dragonfly wing. Credit photograph to © Arthur Chapman.

To date, there is not a universally adopted system for the description of the surface morphologies of insect wings. Byun et al. used the terms ‘layered cuticle’, ‘setae’, ‘denticles’ and ‘fractal’ to describe the morphological features present on the surfaces of the insect wings in their study. A majority of the insect surfaces characterised to date could be readily classified according to this description. The term ‘layered cuticle’ generally refers to a surface that contains scale-type structures that overlap, such as those typically found on butterfly wings. Surfaces with ‘setae’ contain high aspect ratio needles or hairs. ‘Denticle’ structures refer to tooth-like projections and these can vary in their morphology ranging from small hemispheres to taller pillars. ‘Fractal’ structures, as described by Byun et al., are not true fractals but composed of an irregular array of fine nanoscale protrusions (Byun, Hong, Saputra et al., 2009). Amongst these structural types, the presence of layered cuticles, denticles and fractal structures appear to result in the production of the most superhydrophobic surfaces, whilst the presence of setae alone on a surface does not induce superhydrophobic properties (Table 2.2). Fractal and layered cuticle typically display surface hierarchy and therefore exhibit WCA >150°, whilst setae and denticles do not.

Table 2.2. Micro- and nano-scale wax crystal morphologies on the epidermal cells of insect wing surfaces and their WCAa)

Structural Order Species WCA (°) morphology Coleoptera Allomyrina dichotoma Setae 54 Coleoptera Chrysolina virgata Setae 71 Isoptera Schedorhinotermes sp. Setae 71 Coleoptera Amphizoa sinica Setae 109 Hymenoptera Vespa simillima xanthoptera Setae 121 Hymenoptera Vespa dybowskii Setae 126 Hemiptera Meimuna microdon Denticle 140 Orthoptera Atractomorpha lata Denticle 148 Orthoptera Acrida cinerea cinerea Denticle 151 Odonata Hemicordulia tau Fractal 157 Odonata Hemianax papuensis Fractal 161 Lepidoptera Artogeia canidia Layered cuticle 162 Lepidoptera Papilio xuthus Layered cuticle 168 a)This table was modified and updated from (Byun, Hong, Saputra et al., 2009).

Terminology regarding structural morphology was also adopted from Byun et al.

The surface topologies of the wings of several insects are presented in Figure 2.4. The wings of the Nasutitermes sp., which belongs to the order Isoptera, exhibits superhydrophobic properties (WCA = 180°, Figure 2.4A) (Watson, Cribb and Watson, 2010). The surface of their wings is covered by micro-scale setae together with a star- shaped nanoscale structure (micraster), which produces a combined hierarchical structure that exhibits superhydrophobic properties. Similarly, the surface of the grasshopper wing, Acrida cinerea cinerea (WCA = 151°), is covered by micro-denticles that are enclosed by nanoscale wax-crystals (Figure 2.4B). High resolution SEM images indicate the presence of a binary structure (containing micro-and nano-features), which results in the wings possessing superhydrophobic properties. Cicada Meimuna opalifera (Walker) wings also exhibit superhydrophobicity (WCA 165°) despite not having a

28 hierarchical surface structure (Figure 2.4C) (Byun, Hong, Saputra et al., 2009). The surface is covered with a layer of dense nanopillars that result in a high degree of air entrapment on the surface. It is noteworthy that not all cicada wings surfaces exhibit the same properties; Hemiptera Meimuna microdon cicada wings are also covered by layer of denticle structures, however exhibits a WCA of only 140° (Table 2.2). In this case, the density, size, shape and composition of the denticles clearly contribute significantly to the extent of surface hydrophobicity. It is clear that the presence of surface topographical hierarchy is an important factor in imparting superhydrophobicity to a surface but other factors also play an important role. According to the Cassie-Baxter theory, the two important components to determine the hydrophobicity of a surface are the surface chemistry of the materials comprising the surface and the ability of the surface to entrap air, this ability being a function of the roughness and topology of the surface. The presence of a hierarchical surface clearly assists in the entrapment of air; however this is also a function of the density, size and shape of the hierarchical roughness features, as seen for the cicada wing surface shown in Figure 2.4C.

The superhydrophobicity of an insect wing surface, together with its ability to self- clean, are very important factors that contribute to an insect’s ability to survive. The nanoarray structures present on the surfaces of some insect wings such as those of the cicada and dragonfly afford the insect antireflective properties, which can assist in protecting them from attack from predators (Watson, Myhra, Cribb et al., 2008). The superhydrophobic and self-cleaning properties can assist to keep their surfaces clean and free from contaminants that may adversely impact their antireflective properties. Moreover, dragonfly wings were recently reported to possess bactericidal properties

(Ivanova, Hasan, Webb et al., 2012; Ivanova, Hasan, Webb et al., 2013). The mechanism by which this occurs is unclear, however it was postulated that this bactericidal behaviour was related to the wettability of the surface (Pogodin, Hasan, Baulin et al., 2013a).

2.5.2. Superhydrophobicity with additional properties

2.5.2.1. Superhydrophobic surfaces with high adhesive force

The hierarchical surface structure of the petals of some flowers imparts superhydrophobic properties to their surface, resulting in high CAH (greater than 10°) values. These surfaces possess sufficient micro- and nano-structure to induce superhydrophobic properties; however these surfaces also display a high adhesive force towards water. Water droplets in contact with these petal surfaces remain spherical in shape, however these water droplets do not roll off the surface, even when the petal is turned upside down (Figure 2.5A). This is referred to as the ‘petal effect’ (Feng, Zhang, Xi et al., 2008; Schulte, Droste, Koch et al., 2011). The micro- and nanostructure of the petals of the rose, Chinese Kafir lily and sunflower, have a smaller pitch value compared to that of the lotus leaf (i.e. pitch value is the distance between the micropapillae, which is approximately 20 µm compared to 40-50 µm of the lotus leaf). As a result, water droplets coming in contact with the surface can penetrate between the microstructure and partially penetrate between the nanostructures, inducing the ‘Cassie impregnating wetting state’. The wettability of this type of surface is less than that of a surface exhibiting the Wenzel regime but greater than that of a surface exhibiting the Cassie-Baxter regime of wetting (Bhushan and Her, 2010). This also implies that the CAH of such surfaces will increase with increasing wetting area, which is governed by the hierarchical structure of the surface.

Figure 2.5. Superhydrophobic surfaces with high water adhesion. (A) Surfaces that exhibit the ‘petal effect’ exhibit high water contact angles and high adhesion to water. (B) Rose petals, for which the effect is named, possess a similar microstructure to that of lotus leaves; however the nanostructure is comprised of cuticular folds rather than nanocrystals. (C) Gecko feet also exhibit the petal effect. The soles of their feet possess highly hierarchical surface structures that exhibit high water contact angles but the water adheres strongly to the surface. These structures also enable the geckoes to adhere to solid surfaces, enabling them to climb walls. ST: setae, SP: spatula, BR: branched. (Reproduced with permission from ACS publication and Elsevier (Gao, Wang, Yao et al., 2005; Feng, Zhang, Xi et al., 2008; Bhushan, 2012).

Bhushan and Her identified the surface structural properties that distinguish the ‘petal effect’ from the ‘lotus effect’ (Bhushan and Her, 2010). It was demonstrated that if a microstructure possesses a large pitch value, i.e. small peak to base height and low density nanostructure, water droplets coming in contact with the surface can impregnate the microstructure and partially wet the nanostructure. This will result in high CAH and the development of high adhesive forces that have minimal effect on the magnitude of

32 the high static CA (Figure 2.5B-B2). On the other hand, a microstructure possessing a small pitch and high density of nanostructure will create a surface that exhibits an increased propensity to form air pockets, leading to low adhesion of water droplets in contact with the surface. Like lotus leaves, rose petals, shown in Figure 2.5B-B2, are covered with micro papillae; however instead of having nanocrystals on the surface, the papillae on rose petals have cuticular wrinkles or folds that increase the degree of surface roughness. This structure again enables the surface to trap air-pockets, inducing superhydrophobicity; however this structure is not self-cleaning since they have large CAH (Figure 2.5 A, A1).

The foot attachment pads of several animals are capable of repeatedly attaching and detaching from a variety of surfaces. This is useful for locomotion on vertical walls or on inverted surfaces (Hu, Chan and Bush, 2003; Liu, Du, Wu et al., 2012). This dynamic attachment mechanism is referred to as reversible or ‘smart’ adhesion (Bhushan, 2012). There are two kinds of foot pads, either relatively smooth or with a hair-like coating. The former can be found in some amphibians, for example tree frogs, cockroaches and grasshoppers (Federle, Barnes, Baumgartner et al., 2006). The latter can be found in many insects, for example, beetles, spiders and lizards (Bhushan, 2012). Gecko feet have a hair-like surface and are considered an animal equivalent to the plant surfaces that exhibit the petal effect (Figure 2.5C-C3). These surfaces exhibit a static water contact angle larger than 150° whilst their adhesive force to water is reported to be approximately 66 μN (Liu, Du, Wu et al., 2012). The explanation for these properties originates from the morphology of the surface of the skin on their toes (Gao, Wang, Yao et al., 2005) (Figure 2.5C), which are covered by hundreds of thousands of setae (ST). Each ST contains hundreds of spatulae and their hierarchical structure ranges over 5 orders of magnitude, from macroscale (mm) to nanoscale (nm). This hierarchical structure allows geckos to adjust the shear adhesion of their feet when in contact with a range of surfaces under different environmental conditions (Stark, Badge, Wucinich et al., 2013).

2.5.2.2. Anisotropic superhydrophobic surfaces

Surfaces that possess directionally dependent or anisotropic superhydrophobicity have been the subject of a number of studies due to their potential application in microfluidic devices, their ability to form evaporation-driven surface patterns and their

33 potential to form coatings that are readily cleaned (Higgins and Jones, 2000; Liu, Zhai and Jiang, 2006; Liu, Yao and Jiang, 2010). Anisotropic superhydrophobicity has been observed on the surfaces of a number of plants and insects, including rice leaves and butterfly wings (Figure 2.6). These surfaces were found to exhibit properties associated with anisotropic hydrophobicity, that is low water drag, superhydrophobicity and the ability to self-clean (Bixler and Bhushan, 2012; Bixler and Bhushan, 2013), properties that arise from the unique hierarchical structures present on the surface. (Bixler and Bhushan, 2013).

Rice leaves possess a hierarchical structure similar to that of the lotus leaf, however the arrangement of the micro- and nanostructure on the rice leaves is anisotropic (Figure 2.6A-A2), whereas the arrangement is homogeneous on the upper surface of lotus leaves. The surfaces of rice leaves are patterned with sinusoidal grooves covered by 34 micropapillae that are arranged in a quasi-one-dimensional order parallel to the edge of the leaf (Figure 2.6 A, A1). These micropapillae also contain wax nanobumps on their surface (Figure 2.6 A2) (Feng, Li, Li et al., 2002). As a result of this hierarchical structure, water droplets making contact with the surface roll off the leaf along the parallel grooves rather than in a direction perpendicular to these grooves, enhancing the ability of the leaf to self-clean. The tilting angles associated with these two directions of movement have been reported to be 4° and 12°, respectively (Nishimoto and Bhushan, 2013).

Butterfly wing surfaces have been shown to demonstrate anisotropic superhydrophobicity and are known for their multifunctional properties such as structural colour, chemical sensing capacity and ability to fluoresce (Gu, Uetsuka, Takahashi et al., 2003; Vukusic and Sambles, 2003; Vukusic and Hooper, 2005). The directional adhesion of water droplets on the surface of the wings, both along and against the radial outward (RO) direction away from the body centre can be observed (Figure 2.6B-B2). When the wing is tilted along the RO direction, the water droplets initiated the process of ‘roll off’ at an angle of 9°, whereas when tilted perpendicular to the RO direction, the water droplets were tightly pinned to the wing surface. This unique property of butterfly wings arises from the alignment of the anisotropic shingle- like microscales, together with the radially aligned nanogrooves.

2.5.3. Hierarchical structure of superoleophobic surfaces

Not all surfaces that possess hierarchical structure are superhydrophobic. Several aquatic species exhibit superoleophobicity rather than superhydrophobicity, exhibiting an oil CA greater than 150° when submerged in water. In principle, superoleophobicity is very similar to superhydrophobicity, in that the oil CA on a surface is determined by the interfacial energies between the oil and the surface, the oil and the surrounding medium (i.e. water), and the surface and the medium. These organisms possess hierarchical surface structures that are self-cleaning, are antifouling and promote low- drag when moving through water (Bixler and Bhushan, 2013). The skin of a shark is one such example; it is covered by minute scales known as dermal denticles, which are shaped like small ribs (or ‘riblets’). They are positioned such that they align with the direction of fluid flow as the shark swims through the water. It is believed that sharks have the ability to remain clean due to their flexible, low drag riblet microstructure

35 together with a mucous layer on the skin (Bushnell and Moore, 1991; Bechert, Bruse, Hage et al., 1997; Dean and Bhushan, 2010). The design of the riblets, together with the mucous layer, enables the shark to move through water with reduced drag. This surface structure also assists in the protection of the sharks from abrasion, which in turn minimises the opportunities for microscopic organisms to adhere (Bhushan, 2012).

The scales of other fish species are further examples of self-cleaning surfaces in aquatic environments. Their living conditions can be subjected to pollution arising from oil leaks or garbage originating from branched rivers etc. (Hay, 1996). The body of the fish body is, however, well protected from the attachment of microorganisms and hence remains clean. The scales of fish display a similarity to the shark skin in that they are covered by a hierarchical structure, consisting of sector-like scales (diameter of 4-5 mm) covered by papillae (100-300 µm in length and 30-40 µm in width), and exhibit a particularly high oil contact angle in water (163°).

The surfaces of snail shells also possess the ability to remain clean, despite their dwelling environments and their appearance on rainy days. The surface of snail shells is comprised of a rough structure consisting of line grooves (pitch of 0.5 mm), smaller grooves crossing the line groove (pitch of 0.1 mm) and micro-grooves between the line grooves (pitch of 10 µm). As a result snail shells have a rough but regular hierarchical structure. There is one major difference, however, between these structures and those found on superhydrophobic surfaces that entrap air within their hierarchical structure; the snail shells trap water droplets within their rough surface. This macroscopic surface roughness, in fact, facilitates the entrapment of water, ensuring that the shells remain in a wet condition. This is a key factor that contributes to their ability to self-clean, in that their usually wetted surface is rarely able to be contaminated (Nishimoto and Bhushan, 2013). With this approaches, not only contaminants, but also biological organisms, i.e. bacterial cells, can also be cleaned by very similar mechanisms.

2.6. Bacteria

2.6.1. Bacterial physiology

Bacterial cells, which are classified as prokaryotes, were first observed in 1676 under a microscope by Anton van Leeuwenhoek. Bacterial cells assume four main morphologies, including coccus (spherical), bacillus (rod-shape), spirillum (spiral) and filamentous (elongated). Their general structure includes a cell wall surrounding genetic 36 materials, i.e. plasmids and chromosomes. Some bacteria have flagella which help them move in aquatic environments (Figure 2.7).

Figure 2.7. General structure of bacterial cells (Reproduced with permission for single use from ‘Molecular expression’ http://micro.magnet.fsu.edu/cells/bacteriacell.html).

Based on cell wall structure, bacteria are generally divided into Gram-positive and Gram-negative (Figure 2.8). The cell wall of Gram-positive bacteria consists of many layers of peptidoglycan, which have the ability to retain the violet coloured stain ‘crystal violet’ (Madigan, Martinko and Parker, 2009). Peptidoglycan is embedded in teichoic acid and accounts for about half of the dried mass of the cell wall (Shai, 2002; Epand, Rotem, Mor et al., 2008; Madigan, Martinko and Parker, 2009). The peptidoglycan is located outside the cytoplasmic membrane and provides a strong barrier from attachment of antibodies or antibacterial agents. The rigidity of the bacterial cell is due to multiple layers of peptidoglycan, which is a polymer of β (1-4)- linked N-acetyl glucosamine (GlcNAc) and N-acetyl muramic acid (MurNAc). It consists of four alternating L- and D-amino acid with all the lactyl groups substituted by stem peptides (Royet and Dziarski, 2007). Most Gram-positive bacteria have L- lysine as the third amino acid in their stem peptides composition (Royet and Dziarski, 2007). Teichoic acids are polymers which are linked by phosphodiester groups of

adjacent glycerol residues at position one and three or adjacent ribitol residues at position one and four. These make up a poly-glycerolphosphate or poly- ribitolphosphate backbone (Knox and Wicken, 1973). Most Gram-positive bacteria have their teichoic acid covalently bound to membrane glycolipids; these contain a lipoteichoic acid complex providing immunogenic ability which is derived from the teichoic acid component. However, the main role of teichoic acids is to attract cations, such as magnesium (Mg2+) and sodium (Na+), and to strengthen the rigidity of the cell membrane (Knox and Wicken, 1973).

Different from Gram-positive bacteria, the Gram-negative bacterial cell wall has a much thinner peptidoglycan layer. They contain additional outer membrane layers composed of phospholipids and lipopolysaccharides, which interact with the external environment. The negative charges that cover the cell wall are derived from lipopolysaccharides, which are often species specific for different bacterial strains. The toxicity of many Gram-negative bacterial cells is usually associated with the lipopolysaccharide layer (Beveridge, 1999). Even though their cell wall is thin, they are sufficiently strong to withstand approximately 3 atm of turgor pressure (Gumbart, Wiener and Tajkhorshid, 2007), and can be elastic enough to expand several times their

38 normal surface area (Koch and Woeste, 1992; Doyle and Marquis, 1994; Yao, Jericho, Pink et al., 1999). The differences in membrane structures of Gram-positive and Gram- negative bacteria cells cause them to behave differently when in contact with a surface. Once bacteria successfully invade a host surfaces they form aggregations called biofilms, which can be detrimental from both an economic and public health perspective. Therefore, it is important to understand the underlying interactions in order to find suitable approaches that control their formation and proliferation.

2.6.2. Biofilm formation

Biofilms are defined as the attachment and development of a microorganism community embedded in extracellular matrix on a substrate surface (O'Toole, Kaplan and Kolter, 2000). The organisms undergo a transition state between being free swimming in their native environment (planktonic cells) to being cells that form part of a surface-attached community. The essential factors for formation of biofilms are microbes and a substratum (Garrett, Bhakoo and Zhang, 2008). There are numerous advantages to bacteria from taking part in the formation of a biofilm. These advantages include: resistance against antibiotics (Schmidt, Winter and Gallert, 2012), resistance to disinfectants (Ryu and Beuchat, 2005; Simões, Simões and Vieira, 2009) and sheltering from dynamic environments (Liu and Tay, 2002; Di Iaconi, Ramadori, Lopez et al., 2005). The intercellular communication within the biofilm community enhances the regulation of gene expression, which enables the bacterial cells to temporally adapt their phenotypic variation to suit the conditions of the surrounding environment, such as the deficiency of nutrients (Dalton and March, 1998; Kjelleberg and Molin, 2002; Daniels, Vanderleyden and Michiels, 2004).

Biofilm formation can involve a single microbial species or multiple microbial species adhering onto a range of surfaces. On most environmental surfaces, mixtures of various species dominate biofilms. It is usually single species that are responsible for the infection of medical devices and implants (Behlau and Gilmore, 2008; Holmes, Doré, Saraswatula et al., 2008; Seo, Lee, Rayamahji et al., 2008; Wu, Sendamangalam, Xue et al., 2012; Bulgarelli, Schlaeppi, Spaepen et al., 2013). According to a public health report published in 2002 (Klevens, Edwards, Richards Jr et al., 2007), approximately 64% of hospital attending cases were due to the presence of a viable bacterial infection of medical devices and implants, which was associated with 100,000

39 mortalities annually in the US alone. Effort toward the studies of the nature of biofilms started over 3 decades ago with the discovery that in natural living conditions, microorganisms were predominantly found attached to surfaces (Geesey, Richardson, Yeomans et al., 1977).

The first recorded observation of a biofilm was published in 1933 by Henrici (Henrici, 1933). The impact of biofilm formation had, however, been recognised prior to this date as a result of ships being fouled in marine environments (Angst, 1923). It has been estimated that the costs associated with the fouling of ship hulls in the US Navy alone is approximately $180M-$260M per year, which represents only 0.5% of the total number of ships world-wide (Schultz, Bendick, Holm et al., 2011).

The basic development of biofilms is described by a two stage kinetic binding model (Figure 2.9). The first stage involves the initial reversible interaction between bacterial cells and the material surface, followed by the second stage when specific and non-specific interactions at the molecular level take place (Bos, Van Der Mei and Busscher, 1999; Lichter, Van Vlietpa and Rubner, 2009). The interactions that occur in stage two involve proteins that are expressed on bacterial surface and molecules on the material surfaces. The second stage occurs slowly and irreversibly once a mature biofilm is formed. Additional to these two main steps of biofilm maturation, O’Toole et al. proposed that the starvation response pathway can also be considered as part of the biofilm development (O'Toole, Kaplan and Kolter, 2000). This happens when the source of nutrients becomes depleted, and single microbial cells detach from the surface and return to their planktonic mode which may to infect new areas.

Figure 2.9. Schematic of the two stages biofilm formation on surfaces. Reproduced with permission from American Chemical Society (Lichter, Van Vlietpa and Rubner, 2009).

For these reasons, controlling bacterial attachment on material surfaces has been a long-standing challenge for science. Several approaches have been developed to limit colonisation of microbes on surfaces; however most of these approaches have focussed on the use of chemical-based methods, which has led to the new and rising problem of bacterial resistance. Preventing bacterial adhesion by modifying the topography of a surface, e.g. imparting a superhydrophobic and self-cleaning surface structure, has been identified as being a more attractive solution for controlling biofilm formation.

2.7. New approaches for controlling biofilm formation

2.7.1. Antibiofouling

As noted, biofouling has remained a complex and problematic issue for some time due to its significant impact on the economy and public health. As a result, the design of antibacterial materials has focussed on advancing strategies designed to limit the extent of bacterial colonisation onto a surface (Zhang, Wang and Levänen, 2013). Traditionally, antibacterial surfaces were designed so that they leached biocides

(commonly known as cytotoxins) from host cells, which caused the death of bacterial cells on site and in areas surrounding the surface. For example, peptides and chitosan have been used as chemical-based agents for controlling the bacterial colonisation of surfaces (Qi, Xu, Jiang et al., 2004; Gazit, 2007). Antibacterial metal nanoparticles such as Ag (Rai, Yadav and Gade, 2009), Cu (Hsiao, Chen, Shieh et al., 2006), and Mo (Yasuyuki, Kunihiro, Kurissery et al., 2010) have also been used for controlling bacterial attachment, as they are toxic to the bacteria in small concentrations. Their effects on human health and the environment, however, are of concern. Stronger chemical-based methods for disinfection of surfaces are being required in order to control bacterial colonisation, which has resulted in bacterial species developing resistance to such reagents.

More recently, new approaches that use photocatalytic metal oxides such as TiO2 (Gelover, Gómez, Reyes et al., 2006) and ZnO (Franklin, Rogers, Apte et al., 2007; Jones, Ray, Ranjit et al., 2008) have been employed. These materials produce highly reactive chemical species such as hydroxyl radicals, hydrogen peroxide and superoxide, which are lethal to Escherichia coli and some other types of bacterial cells (Maness, Smolinski, Blake et al., 1999; Ibáñez, Litter and Pizarro, 2003). These metal oxides are mainly activated by UVA light, which have limited their potential application (Fu, Vary and Lin, 2005; In, Orlov, Berg et al., 2007).

Superhydrophobic surfaces based on natural surfaces such as plant leaves and insect cuticles are current attractive approaches under study for controlling the extent of bacterial colonisation. Traditionally, materials that could induce bacterial cell death were considered antibacterial materials (Zhang, Wang and Levänen, 2013). Recently, however, many superhydrophobic surfaces have been categorised as being antibacterial materials due to their potential application in controlling bacterial attachment. These natural surfaces have become responsive to external living conditions and are therefore constantly in contact with pollutants as well as harsh living environments. Observing these unique surfaces allows us to understand their antimicrobial action and therefore design surfaces that exhibit resistance to bacterial colonisation.

2.7.2. Bactericidal activity of insect wings

Ivanova et al. recently found that the robust hexagonal arrays of 'nanopillars' present on the surface of the Psaltoda claripennis cicada wings exhibit bactericidal behaviour (Ivanova, Hasan, Webb et al., 2012). This array of nanopillars penetrated Pseudomonas aeruginosa cells in contact with the surface, killing them with great efficiency (Figure 2.10A, A1). The surface of the cicada wings retained its lethality against these Gram-negative pathogenic bacteria even after the surface was coated with a 10 nm-thick layer of gold, which indicated that the bactericidal properties of the surface were due to physical rather than chemical action. It was also reported that the wings consistently killed other Gram-negative bacteria, i.e., Branhamella cartarrhlis, E. coli, and Pseudomonas fluorescens, however Gram-positive cells (Bacillus subtilis, Planococcus maritimus, and Staphylococcus aureus) were resistant to the action of these nanopillars (Hasan, Webb, Truong et al., 2012a). These cicada wings were the first example of a surface that possessed bactericidal properties that relied on only a physical interaction with the bacteria.

To explain this phenomenon, biophysical models were constructed to describe the interaction between bacterial cells and the nanopatterned pillars present on the cicada wing surfaces (Pogodin, Hasan, Baulin et al., 2013a). Modelling calculations revealed that the nanopillars did not actually pierce the cell wall but rather the cell wall was stretched to the point of breakage in the regions between the nanopillars as the cells adsorbed onto the wing surface. It was also found that the more rigid the cell membrane, the more difficult it was for the cell wall to rupture, which was consistent with the persistence obtained with Gram-positive bacteria that have a more rigid peptidoglycan layer. This theory was supported experimentally by decreasing the rigidity of Gram- positive cells through the application of microwave radiation. B. subtilis, S. aureus and Planococcus maritimus bacteria were tested as representive bacterial species. After exposure to microwave radiation, all three bacterial species showed a higher level of

44 susceptibility to the action of the cicada wing surfaces (Pogodin, Hasan, Baulin et al., 2013a).

In contrast to cicada wings, dragonfly wings rupture a large range of bacterial species, including Gram-negative (Pseudomonas aeruginosa), Gram-positive (Staphylococcus aureus and Bacillus subtilis) and even endospores. The dragonfly wing surfaces, like the cicada wing surfaces, were covered by a layer of nanopillar-like structures that showed activity against all bacterial cells coming into contact with the surface, as demonstrated in Figure 2.10, B1-3. A synthetic material (black silicon), which mimicked the surface structure of the dragonfly wings also showed high effectiveness at killing the same bacterial cells (Ivanova, Hasan, Webb et al., 2013). This project will focus on understanding the mechanisms by which the superhydrophobic and self-cleaning insect surfaces can exhibit mechano-bactericidal properties with a view to designing artificial surfaces that have the ability to control bacterial attachment for potential use in biomedical device and other industrial applications.

Chapter 3. Materials and methods

______

3.1. Overview

This thesis focusses experimentally on the chemical and structural characterisations as well as the biological activity and physical surface properties of dragonfly (mainly Hemianax papuensis and Hemicordulia tau) wings. Additionally, experiments that attempt to model how these surface structures are generated are used to explore the mechanisms involved in their biological activities.

3.2. Materials

Chloroform (>99.8%, ethanol stabilised), n-eicosane (CH3(CH2)18CH3, anhydrous,

≥99%) and docosane (CH3(CH2)20CH3, 99%) were purchased from Sigma Aldrich (Castle Hill, NSW, Australia). Highly ordered pyrolytic graphite (HOPG - 0.8 grade, Atomgraph-Crystal, RF) was obtained from AtomGraphCrystal (ATC) (Moscow, Russian Federation). A p-type boron-doped 100 mm diameter commercial Si wafer with specific resistivity of 10-20 Ω cm-1, (100) oriented surface and 525 ± 25 µm thickness was obtained from Atecom Ltd (Taipei, Taiwan). Single-sided Kapton® polyimide tape, purchased from Proscitech (Thuringowa Central, QLD Australia), was used to cleave the HOPG surfaces, as it has been found to be suitable for experiments conducted over a wide range of temperatures.

3.3. Protocols for handing of insect wings

3.3.1. Fresh and aged insect wings

The Hemicordulia tau and Hemianax papuensis dragonfly wing samples, which commonly dwell in suburban regions of Melbourne, Australia, were provided by the Melbourne Museum. The Diplacodes bipunctata dragonfly and Ischnura heterosticta damselfly wing samples were obtained from the suburban regions of South Australia. These fresh samples were collected in early to mid-2011 and preserved in sterile Petri dishes at room temperature (ca. 22 °C). Aged Diplacodes melanopsis dragonfly and aged Xanthagrion erythroneurum damselfly wing samples that had been preserved at the Melbourne Museum in the 1970s were also used in this study. When required, the wings were aseptically removed from the body and stored at room temperature in sterile plastic containers.

3.3.2. Extraction of the lipids from dragonfly wings

An approximately 0.5 cm2 section of the dragonfly wings was used for stepwise extraction with chloroform. The epicuticular layer of the H. tau wing was extracted with chloroform for 10 s, 20 s, and 30 s periods, respectively, to investigate the impact of surface topography and surface chemistry on the wettability of the wing surface. The surface of H. papuensis wings was extracted for 10 s and 1 hr in chloroform to allow the chemical components that comprise the nanoscale structures of the wing to be identified. Ten second extractions were used in order to limit the extraction depth to just below the outer epicuticular surface, whereas the extractions performed over one hour ensured that the entire lipid fraction of the epicuticle was removed from the wing surface. The lipid-chloroform mixture was then filtered through glass wool to remove any contaminating particles that may have been present. The chloroform extracted wing samples were retained for further surface characterisation. A 20 µg volume of n- tetracosane was added to each of the chloroform fractions as an internal standard before performing GC-MS analysis.

3.4. Surface fabrication

Freshly cleaved highly ordered pyrolytic graphite (HOPG - 0.8 grade) surfaces were obtained by stripping their outermost layers with an adhesive tape. The lipids, eicosane (C20H42), docosane (C22H46), hexadecanoic acid (palmitic acid - C16H32O2), and octadecanoic acid (stearic acid - C18H36O2) were then individually dissolved in chloroform at a concentration of 3 mg mL-1. Approximately 10 µL of the alkane solution was then applied on the surfaces of the silicon wafer and freshly-cleaved HOPG, the lipids were allowed to self-assemble at ambient temperature (ca. 22°C).

3.5. Surface characterisation

3.5.1. Atomic force microscopy

Atomic force microscope (Innova, Veeco, USA) in tapping mode was employed to examine the surface topographic profiles. Phosphorus-doped silicon probes (MPP- 31120-10, Veeco) with a spring constant of 0.9 N m-1, radius of curvature (tip) of 8 nm and a resonance frequency of approximately 20 kHz were utilised for surface imaging. Scanning was carried out at a right angle to the axis of the cantilever at 1 Hz. The topographical data of the investigated surfaces were processed with first order horizontal and vertical levelling. In order to analyse the surface topography, different 48 surface roughness parameters were employed: the average roughness (Ra), root-mean- square (rms) roughness (Rq) and maximum peak height (Rmax) were performed using Gwyddion data processing software (Nečas and Klapetek, 2012). Two additional parameters, skewness (Rsk) and kurtosis (Rku), which are useful for describing surface morphology, have also been utilised. Skewness measures the symmetry of the height probability density function and kurtosis measures the ‘peakedness’ of the profile (Gadelmawla, Koura, Maksoud et al., 2002; Lamolle, Monjo, Lyngstadaas et al., 2009; Truong, Rundell, Lapovok et al., 2009; Ivanova, Truong, Wang et al., 2010).

Bearing statistics, including peak height (Rpk), valley depth (Rvk), core roughness

(Rk) and Rm were also analysed, as described elsewhere (Crawford, Webb, Truong et al., 2012). These were able to distinguish the differences between surfaces with similar average roughness. All images were produced using an Innova atomic force microscope (Veeco) and the raw data was then extracted to the Gwyddion software for processing (Nečas and Klapetek, 2012).

3.5.2. Scanning electron microscopy

3.5.2.1. Biological samples

The surface morphologies of the both lipid-extracted and untreated wings were observed using a field emission scanning electron microscope (FeSEM - SUPRA 40VP, Carl Zeiss GmbH, Jena, Germany) at fixed voltage of 3 kV. The wing sample was attached to a metallic substratum using conductive double-sided adhesive tape. Samples were sputter coated with gold using a JEOL NeoCoater (model MP-19020NCTR) prior to imaging, using a method described elsewhere (Ivanova, Hasan, Webb et al., 2012). To observe the structure of cross-sections of the wing membranes, the wing was snap- frozen in liquid nitrogen and then broken perpendicular to the main veins of the wing using tweezers. Broken wings were attached to metallic discs at one end, to enable the exposed end to face away from the substratum, enabling optimum imaging. Samples were then sputter coated with gold in an identical manner as above.

3.5.2.2. Synthetic surfaces

Samples for analysis were first coated with a thin layer of gold film (approximately 3 -5 nm thickness) using a Dynavac CS300. The high resolution scanning electron microscopy images were obtained using field-emission SEM (FESEM, SUPRA 40VP) at 3kV under 15,000× and 35,000× magnification. A two-dimensional fast Fourier- 49 transform was applied to the micrographs to identify the dominant patterns of alkane self-assembly.

3.5.2.3. Quantification of alignment using fast Fourier transforms (FFT) analysis

The quantitative topographical analysis of the SEM images was carried out using Gwyddion software (Open source, www.gwyddion.net).(Nečas and Klapetek, 2012) A two-dimensional fast Fourier-transform was applied to the micrographs to identify the dominant patterns of alkane self-assembly. Rectangular regions on top view SEM images with resolution of 744 × 502 pixels were processed for 2D FFT. The flat-top window function was chosen to obtain an insight into the spatial arrangements and direction of the self-assembled alkanes onto the surface of HOPG.

3.5.3. Raman spectroscopy

The self-assembled alkanes present on the surface of the graphite were immersed in 5 mL of MilliQ water for 1 hour. The air retained on the surfaces after this time was observed using an Alpha 300R Raman micro-spectrometer (WiTEC) with a 532.1 nm wavelength laser (hν = 2.33 eV). A water-immersion 63× objective lens (numerical aperture = 0.9, Zeiss) was used. A grid of 50 spectra × 50 spectra was acquired over a scanning area of 25 µm × 25 µm. The integration time for each spectrum was 0.5 seconds. Independent scanning was repeated twice on each of the two different samples (containing self-assembled eicosane and docosane). The signal was collected at an angle of 90° relative to the sample plane. The intensity of the spectral peak at 3100 – 3600 cm-1, which corresponds to the O-H bonds of water molecules, was used for qualitatively assessing the concentration of air bubbles (i.e. lack of water) present on the surface (Truong, Webb, Fadeeva et al., 2012; Webb, Boshkovikj, Fluke et al., 2013).

3.5.4. Wettability

The contact angles of Milli-Q water on both biological and synthetic surfaces was measured using the sessile drop method (Crawford, Koopal and Ralston, 1987; Van Oss, Good and Chaudhury, 1988; Guy, Crawford and Mainwaring, 1996; Öner and McCarthy, 2000). The contact angle measurements were carried out in air using an FTA1000 (First Ten Ångstroms, Inc., Portsmouth, VA, USA) instrument. An average of at least ten measurements was obtained for each sample, in duplicate. The evaluation of contact angles was performed by recording 50 images in 2 seconds with a Pelcomodel 50

PCHM 575–4 camera using FTA Windows Mode 4 software. As for the insect wings, to eliminate effect of the vein structure on the surface, the water droplet was placed on regions sufficiently large to accommodate the droplet footprint. The measurement was taken at ambient conditions of 21ºC and relative humidity of 60-70%.

3.6. Chemistry characterisation techniques

3.6.1. X-ray photoelectron spectroscopy

A VG ESCALAB 220i-XL X-ray Photoelectron Spectrometer equipped with a hemispherical analyser was used for XPS data acquisition. Samples underwent alternating data acquisition and ion beam etching cycles. Monochromatic Al Kα X-rays (1486.6 eV) at 220 W (22 mA and 10 kV) were used as incident radiation for data acquisition. Survey scans were carried out at pass energies of 100 eV over a binding energy range of 1200 eV. Base pressure in the analysis chamber was below 7.0 × 10-9 mbar and during sample depth profile analysis rose to 1.5 × 10-7 mbar. Ion beam etching was performed using a 4 keV Argon beam over a 2 mm × 2 mm area for intervals of 120 seconds. Data acquisition and etching was performed for 5 cycles, with a total etching time of 3000 seconds. XPS data were fitted using mixed Gaussian Lorentzian peak shape and linear background subtraction with CasaXPS.

3.6.2. Synchrotron radiation Fourier Transform Infrared spectroscopy (SR- FTIR)

3.6.2.1. FTIR in tranmission mode

The distribution of organic functional groups present across the wing membrane before and after chloroform extraction was analysed using SR-FTIR, on the Infrared microscopy beamline at the Australian Synchrotron. The samples were scanned in transmission mode over several areas of approximately 50 µm × 50 µm using a Bruker Hyperion 2000 FTIR microscope (Bruker Optic GmbH, Ettlingen, Germany), equipped with a narrow-band, high-sensitivity mercury cadmium telluride detector. A long-pass filter selected the detection range from 4950 cm-1 to 750 cm-1 and the microscope was fitted with a sample chamber purged with dry air to maintain low constant humidity (approximately 20%). OPUS version 6.5 software was employed for operating the microscope and spectrometer as well as for analysing the data.

3.6.2.2. FTIR in attenuated total reflection (ATR) mode

The infrared beam passes through the ATR crystal and reflects back, producing an evanescent wave which penetrates no more than a few micrometres into the upper surface of the sample (in this case ~1 µm). This method provides information on the chemical functionality of the surface in contact with the crystal (Figure 3.1).

Figure 3.1. Schematic diagram of Synchrotron the FTIR in ATR mode.

The ATR measurements were made with a hemispherical ATR element (germanium and zinc selenide (ZnSe)) with the flat side of the hemisphere pressed into contact with the sample (Beattie, Beaussart, Mierczynska-Vasilev et al., 2012). Similar with FTIR in transmission mode, the data was analysed using OPUS version 6.5 software.

3.6.3. Gas chromatography – mass spectrometry

Prior to GC-MS analysis, any hydroxyl-containing components of the wing extracts (e.g. alcohols and carboxylic acids) were converted to trimethylsilyl (TMSi) derivatives by means of reaction with N,N-bis-trimethylsilyl-triflouro-acetamide (BSTFA,

C8H21NOSi2) in the presence of pyridine for 45 minutes at 70°C according to previously 52 optimised laboratory protocols (Gołebiowski, Bogus, Paszkiewicz et al., 2010; Van Dooremalen and Ellers, 2010; Gołebiowski, Bogus, Paszkiewicz et al., 2011; Murphy and Axelsen, 2011; Yocum, Buckner and Fatland, 2011; Küçükbay, Kuyumcu, Bilenler et al., 2012). The composition of the sample extract was determined by injecting 1 µL of the solution onto a capillary GC column (Rxi-5SIL MS column fused silica; column 30.0 m, 0.25 mm i.d., df = 0.25 µm, Restek Corporation, Bellefonte PA, U.S.A.), using He as the carrier gas at constant flow of 1.4 mL min-1 and a mass spectrometric detector (GCMS-QP2010, Shimadzu Scientific Instruments, Columbia, U.S.A). The chromatograph was programmed as follows: on-column injection at 50°C, oven at 50°C for 5 minutes, then increasing to 280°C at the rate 3°C min-1, and then held for 30 minutes at 280°C. Mass spectra were recorded from m/z 14 to 800. The percentage composition reported for each component is based on the peak area from the Total Ion Current (TIC) chromatogram without standardisation. A standard mixture of n-alkanes

(C8-C40) in chloroform was used as an external standard to verify the retention times (Van Den Dool and Kratz, 1963). The retention indices (RI) for each component were calculated based on:

RTxn RT 0 RI100  n0 Equation 3.1. RTnn10 RT where n0 is the number of carbons for an n-alkane standard and n1 corresponds to the next n-alkane in the series. RTx, RTn0 and RTn1 are the retention times for the component being measured relative to the two closest n-alkane standards. The RI was compared with retention index databases (Sadtler Research Laboratories, 1984; Pacakova and Feltl, 1992) to identity the components not contained in the Wiley 7th Edition mass spectral database.

3.7. Bacterial strains, growth, and sample preparation

Pseudomonas aeruginosa ATCC 23246, Staphylococcus aureus CIP65.8T, Bacillus subtilis NCIMB 3610T and B. subtilis spores were used in this study. The bacterial strains were obtained from NCIMB (Aberdeen, UK), The Collection of Institut Pasteur (Paris, France), and American Type Culture Collection (Manassas, VA, USA). Prior to individual experiments, bacterial cultures were refreshed from stock onto nutrient agar (BD, Franklin Lakes, NJ, USA). To prepare spore, B. subtilis culture was left in the incubator at 37 °C for a week prior to experimentation to depress the oxygen levels,

53 hence stimulating the formation of spores. For individual experiments, bacterial cell suspensions of each strain were prepared in 5 mL nutrient broth (BD, Franklin Lakes, NJ, USA) from fresh culture grown overnight in an incubator at 37 °C. Bacterial cells were collected at the logarithmic stage of growth and the suspensions were adjusted to

OD600 = 0.3 for vegetative cells or OD600 = 1.0 for spores. The mounted insect wings were then immersed in approximately 5 mL bacterial suspension and incubated for 18 hours at room temperature (ca. 22 °C). The control samples consisted of the glass cover slip immersed in the same bacterial suspension.

3.8. Biological activity assays

3.8.1. Plate counting assay

Viability assays were performed by standard plate counts (Postgate, 1969) . P. aeruginosa, S. aureus and B. subtilis cells and spores were suspended in 5 mL of phosphate buffered saline (PBS) and adjusted to OD600 = 0.1. Re-suspended cells were diluted 1:10 and then incubated in 3.5 cm diameter wells, in triplicate, with each well containing a 1 cm2 area substrate sample of a D. bipunctata and H. papuensis wing, black silicon, smooth silicon wafer or a glass cover slip. The cell suspensions were then sampled (100 µL) at discrete time intervals (3 hours and 18 hours), serially diluted 1:10 and each dilution spread on three nutrient agar plates. Resulting colonies were then counted and the number of colony forming units per mL was calculated. The number of colony forming units was assumed to be equivalent to the number of live cells in suspension (Postgate, 1969). The maximum bactericidal efficiency was measured as the number of inactivated cells per square centimetre of sample per minute of incubation time, relative to the control surfaces.

3.8.2. Quantitation of bactericidal activity using confocal images for fabricated surfaces

Pseudomonas aeruginosa ATCC 9027 and Staphylococcus aureus CIP 65.8T were the bacterial strains used in this study. Three independent experiments were performed for each bacterium on each surface. In each, bacterial cell suspensions were prepared in 5 mL nutrient broth (BD, Franklin Lakes, NJ, USA) from fresh culture, grown overnight at 37°C. Bacterial cells were collected at the logarithmic stage of growth, and the suspensions were adjusted to OD600 = 0.1. The mounted self-assembled surfaces were then immersed in 5 mL of bacterial suspension and incubated for 2 hours at room 54 temperature (ca. 22°C). Control samples consisting of untreated, freshly cleaved HOPG were also immersed in the same bacterial suspensions.

3.9. Confocal laser scanning microscopy (CLSM)

Confocal laser scanning microscopy was used to visualise the proportions of live cells and dead cells using a LIVE/DEAD® BacLight™ Bacterial Viability Kit, L7012. SYTO® 9 permeated both intact and damaged membranes of the cells, binding to nucleic acids and fluorescing green when excited by a 485 nm wavelength laser. Propidium iodide only entered the cells that had sustained significant membrane damage (and therefore considered to be non-viable) and bound with higher affinity to the intracellular nucleic acids than the SYTO® 9. Bacterial suspensions were stained according to the manufacturer’s protocol and imaged using a Fluoview FV10i inverted microscope (Olympus, Tokyo, Japan) (Ivanova, Hasan, Webb et al., 2013).

3.10. Statistical analysis

The results obtained were expressed in terms of their mean values and the corresponding standard deviations following commonly used protocols. Statistical data processing was performed using paired Student’s two-tailed t-tests to evaluate the consistency of results (asterisked if p < 0.05) in figures and tables. To obtain the biological activities of bacterial cells onto fabricated surfaces, imaging processing software (the Gwyddion), was used to estimate the number of cells attached at each microscopic view. At least 17 individual scan areas were used for statistical analysis.

Chapter 4. Surface characterisation and functionality of the epicuticle of insect wings

______56

4.1. Overview

A number of natural surfaces possess superhydrophobic and self-cleaning properties that would be extremely beneficial in industrial applications. The surface of dragonfly wings is one of such examples. Here, the surface characteristics of the wings of various species of dragonflies, including more recently collected samples in 2011 and aged samples collected in the 1970s, were investigated in order to gain an insight into their surface structures, with a view to their potential applications of these surfaces. Surface characterisation methods, including water contact angle (WCA), scanning electron microscopy (SEM) and atomic force microscopy (AFM) which detail the wettability and morphology of these dragonfly wings. The chemical functionality of the surface of the wings was determined using the Fourier Transform Infrared beam line at the Australian Synchrotron and gas chromatography mass spectrometry (GCMS). It was determined that the surfaces of the wings are largely comprised of aliphatic hydrocarbons, which contribute to not only surface hydrophobicity of the surface, but also to the formation of the wing surface nano-structure.

4.2. Surface topology and wettability of dragonfly wings

4.2.1. Scanning electron microscopy

High-resolution scanning electron micrographs of a cross-section of the H. papuensis and H. tau dragonfly wings shown the presence of three distinct layers; these are the epicuticular layer on both the dorsal and ventral surface, and the intracuticular layer (procuticle) (Figure 4.1A). The epicuticular layers were estimated to be 180 - 400 nm in thickness, whilst the intracuticle was found to be approximately 1.5 µm thick, giving a total thickness of 2.0 – 2.5 µm for the wing membrane. A top view of the wing shown that the surface of consisted of multiple nanoscale pillars that have round tops with a diameter of approximately 80 ± 20 nm, and spaced approximately 180 ± 30 nm apart. These observations were consistent with those reported previously by Watson et al. (Watson, Watson, Hu et al., 2010) for the Rhyothemis phyyllischloe dragonfly species.

4.2.2. Atomic force microscopy

AFM was employed to measure the surface topography as shown in Figure 4.2. This method provided a comparison of the surface nano-structures of the different species of insect wings.

Figure 4.2. Surface topography of (A) Hemianax papuensis and (B) Hemicordulia tau dragonfly wings with their corresponding line profiles. Inset scale bar = 500 nm.

Typical two-dimensional images and cross-sectional profiles of 2.5 µm x 2.5 µm AFM scan areas of each of the wing surfaces are presented in Figure 4.2. Each cross- sectional profile was generated by selecting one line from the corresponding two- dimensional AFM scan and plotting peak height as the function of translation along the line. The mean planes of corresponding scans were defined as 0 in cross-sectional profiles. The nanopillars appeared to be more regular on the surface of H. papuensis compared to the of H. tau dragonfly wings. The line profile of H. papuensis showed that there are more splits within the same peaks compared to the line profile of the H. tau wing. The maximum height of the H. papuensis wing was approximately 130 - 250 nm whilst that of the H. tau wing was 150 - 300 nm in height. AFM is able to provide accurate dimensions of the nanopillars on the surfaces of the wings, where the data also highlights the consistency between dragonfly species, in which the general structures of the wing surface appear to be very similar.

4.2.3. Surface wettability

Dragonfly wings have been gaining significant research attention recently due to their superhydrophobic and self-cleaning properties. Whereas Ivanova et al. reported for the first time a mechanical-bactericidal surface that arises from the sophisticated structure of insect wings (Ivanova, Hasan, Webb et al., 2013). This new finding prompted further research here to understand the underlying mechanism(s) responsible for this mechanical-bactericidal phenomenon. As a starting point, the wettability of these surfaces was investigated. To eliminate the effect of the vein structure on the surface, a water droplet was placed onto the regions of the surface that were sufficiently large that the droplet footprint could be accommodated. Contact angles (CA) were measured using the sessile drop method. The image of the droplet was captured immediately after it came to rest on the surface and the contact angle (CA) determined. When a drop is allowed to fall onto the H. papuensis dragonfly wing surface, it bounces as a result of being repelled by the self-cleaning nature of the wing surface (Figure 4.3).

Figure 4.3. A water droplet coming into contact with the H. papuensis dragonfly wing surface. The surface is sufficiently hydrophobic for the droplet to bounce after it makes initial impact.

Hemianax papuensis and Hemicordulia tau dragonflies are common inhabitants of Australia. The wings of both dragonflies exhibited WCA greater than 150° (162°, and 151° respectively), which highlight their superhydrophobic properties. Water droplets are able to roll easily across the surface when it is tilted on an angle of less than 10° (6.3° in the case of H. papuensis and 2.3° for H. tau), which also supports their self- cleaning properties. It well established that superhydrophobic surfaces tend to adhere to the Cassie-Baxter regime, where a droplet sits on top of the textured surface with air

61 being trapped between the texture features. In order to maintain a stable superhydrophobic state, hierarchical structure is important (Bhushan, 2007; Berne, Weeks and Zhou, 2009). Therefore, further investigation into the surface characteristics of these dragonflies’ wings is suggested.

4.3. Correlation between surface chemistry and the wettability of a surface

4.3.1. Variations in the surface wettability of dragonfly wings

Three different dragonfly wings, two of which were collected in 2011 (Hemianax papuensis and Hemicordulia tau), and one collected in the 1970s (Diplacodes melanopsis) were analysed to compare the surface wettability between relatively fresh and aged samples. The WCA was found to be 162°, 151°, and 138° for H. papuensis, H. tau and D. melanopsis dragonfly wings, respectively. Whilst the two wings collected in 2011 both exhibited superhydrophobic properties (i.e. having a contact angle above 150°), the wing collected in the 1970s fell just short of this threshold as illustrated in Figure 4.4.

The fact that the aged dragonfly wing appeared to be less hydrophobic than the more recently collected wings in term of their wettability prompted two possible questions. The first was whether ageing of the wing might have affected the chemistry of the wing surface; the second was whether the differences that were detected were simply a function of the natural diversity between the species. In order to obtain an insight into these questions, various analytical methods, including Synchrotron FTIR, SEM and AFM were employed to compare and contrast their surface properties. FTIR analysis provides information regarding the chemistry across the wing membrane, which may reflect on any changes that may have taken place on the wing surface over time. In addition, SEM and AFM would provide an insight into the differences in surface architecture that could be correlated with data obtained using FTIR spectroscopy.

The SEM images of the three dragonfly wings show that all of the surfaces are covered by a layer of nanopillars (Figure 4.5). There are, however, some minor differences in the individual structures. The pillars on the surface of the D. melanopsis wings appeared to form larger clusters, whilst the more recently collected wings appeared to possess more distinct pillars with round tips (Figure 4.5). Based on these micrographs, there is no visible evidence to suggest that long-term storage had any impact on the surface structure of the wings.

Figure 4.5. High resolution scanning micrographs of (A) H. papuensis, (B) H. tau, (C) D. melanopsis dragonfly wings and their corresponding 3D image.

4.3.2. Relationship between the quantity of surface epicuticular lipids and wettability

Synchrotron FTIR in the transmission mode was used to determine the chemical composition of the wing membrane. Spectra representing three different species of insect wings including recently collected (H. papuensis, and H. tau) and aged (D. melanopsis) wings were found to contained three major bands at 3480-3230 cm-1, 3000- 2800 cm-1 and 1750-1480 cm-1 These bands corresponded to hydroxyl, alkyl hydrocarbons and amide groups (Figure 4.6, Table 4.1).

Table 4.1. Assignment of chemical components to the major absorption bands present in the IR spectra of the wing membranes

Band position (cm-1) Functional groups Characteristic Out-of-phase combination of the N-H in plane bend and the C-N stretching 1480-1580 Amide II vibration Carbonyl stretching band due to ketone, carboxylic acid (Barth, 2007) C=O stretch vibration causes CCN Amide I 1570-1700 deformation and the N-H in-plane bend Carbonyl groups (Jackson and Mantsch, 1995) Carbonyl stretching band (Lasch, Boese, 1700-1750 Ester Pacifico et al., 2002) CH , CH symmetric and asymmetric 2800-3000 Alkanes or alkyl groups 2 3 stretch Unsaturated and aromatic C-H stretch (C is sp2) (Coury and Dillner, 3000-3100 C-H 2008)

O–H and N-H stretching, which can be due O-H stretch from free to carbohydrates, carboxylic acids, amines 3200-3500 alcohol or phenol or alcohols/phenols (Caruso, Furlong, Ariga et al., 1998)

The spectra of all three wings were dominated by Amide I and Amide II absorption bands due to C=O bond stretching coupled to N-H bending (1610 - 1695 cm-1) and C-N stretching coupled to N-H bending (1480 - 1575 cm-1), respectively (Caruso, Furlong, Ariga et al., 1998; Brugnerotto, Lizardi, Goycoolea et al., 2001; Ganim, Hoi, Smith et al., 2008; Sajomsang and Gonil, 2010). The presence of amide groups can be attributed to chitin and proteins, as they represent the major structural components that make up the bulk of insect cuticle (Neville, Parry and Woodhead-Galloway, 1976; Hackman and Goldberg, 1987; Gorb, 1997; Kramer and Muthukrishnan, 1997; Kreuz, Arnold and Kesel, 2001). The broad, overlapping O-H and N-H stretching bands within 3500 - 3200 cm-1 are likely also to result from absorption by the chitin and protein components of the wings (Coury and Dillner, 2008). A small ester carbonyl stretching band was present at approximately 1735 cm-1 in all of the recorded spectra. This may be due to the

66 presences of carbonyl compounds, such as carboxylic acids and esters, on the wing membrane.

The C-H stretching region (2840 - 3000 cm-1) contains three bands representing the symmetric (s) and anti-symmetric (as) stretching vibrations of the CH2 and CH3 functional groups (Figure 4.6). The presence of C-H stretching bands with a prevalence of methylene bands is indicative of long-chain aliphatic hydrocarbons (Merk, Blume and Riederer, 1998; Zeier and Schreiber, 1999). Standardisation of the spectra using the Amide I peak was used to compare the proportion of aliphatic hydrocarbons relative to the thickness of the wing. The results revealed that their intensity reasonably aligned with their WCA. Both H. papuensis and H. tau exhibited high WCA, which also appeared to contain the higher proportion of aliphatic hydrocarbons. The opposite was the case for D. melanopsis, which exhibited the lowest WCA and hence, appeared with lower aliphatic hydrocarbon content. The data presented here suggested, not unexpectedly, that there was a correlation between the WCA and amount of aliphatic hydrocarbons; however this information does not provide an insight into whether the differences in both wettability and surface chemistry have arisen as a result of long periods of storage.

To investigate whether preservation had an impact on the amount of aliphatic hydrocarbon present on the wing surfaces, the wings of two species of damselfly, belonging to the Odonata order, were analysed. The species included samples collected in 2011 (Ischnura heterosticta) and aged samples collected in the 1970s (Xanthagrion erythroneurum). The infrared spectra and corresponding WCA analyses are presented in Figure 4.7.

Figure 4.7. Infrared absorbance spectra of two species of damselfly wings, Xanthagrion erythroneurum and Ischnura heterosticta, and their corresponding WCA.

The intensity of the CH2 stretching bands provides an indication of the amount of aliphatic hydrocarbons present on the wing surface.

The two damselfly wings including the aged sample X. erythroneurum, both exhibited a WCA greater than 150°, which indicated that long time storage did not affect the wettability of the surface. Also, the intensity of the C-H stretching regions in the IR spectra supports the suggestion that the quantity of aliphatic hydrocarbons is correlated with the extent of their wettability. These results also suggest that differences in the wettability between Odonata wings arose as a result of variations within species, which was also related to the amount hydrocarbons on the surface.

4.4. Dual roles of epicuticular lipids in determining the wettability of dragonfly wings

4.4.1. Wettability of dragonfly wings as a function of aliphatic hydrocarbon content

In previous experiments it was shown that there was a correlation between the surface wettability and the quantity of aliphatic hydrocarbons presented on the outermost layer of dragonfly wing. It was well-known that surface composition plays an important role in controlling surface wettability according to the Cassie-Baxter theory of wetting. Therefore, it is important to identify the precise composition present on the surface of the dragonfly wings. In order to do so, the epicuticular layer of dragonfly wings was extracted using chloroform, and then the mixtures were analysed using GCMS to identify their chemical components. The wettability of the dragonfly before and after the extraction process was measured to determine the extent to which the presence of aliphatic carbons affected the surface wettability.

Removal of the hydrophobic materials in the form of lipids (waxes) from the H. papuensis and H. tau dragonfly wing surfaces was predicted to lead to an increase in wettability. An area of approximately 0.5 mm × 0.5 mm of each wing was extracted with chloroform for 10, 20, and 30 s in order to investigate how the extraction process would affect the properties of the wings. The WCAs measured on the chloroform extracted wings decreased as a function of increasing chloroform extraction time. This decrease was accompanied by an increase in the contact angle hysteresis (CAH) (Figure 4.8). Within 10 s of extraction, the WCA of the wing fell below the threshold for superhydrophobic materials (shaded area, Figure 4.8), with further decreases in WCA and CAH occurring for the samples that were subjected to longer extraction times.

The chloroform extracted wings were found to remain relatively hydrophobic, which suggests that the newly exposed outer layer was composed of a material with low surface energy.

4.4.2. The chemical composition of untreated and chloroform extracted wings

Synchrotron FTIR spectroscopy was used in transmission mode to determine the chemical composition of the wing membranes, before and after extraction of the wing components using chloroform. The spectra in Figure 4.9 represent the C-H stretching region of the untreated and chloroform-extracted wings of H. tau dragonfly wings (Figure 4.9). The C-H stretching IR absorption bands appeared with relatively high intensity in the spectra of the untreated wings; however their intensity considerably decreased as the extraction time was increased. This observation suggests that the chloroform extraction removed the hydrophobic wax-like components present on the epicuticular layer of the wing membrane.

CH2 stretching bands is a good approximation of the concentration of aliphatic hydrocarbons, i.e. lipids (waxes) present on the wing surface. The spectra were acquired at the Australian Synchrotron.

The identities of the lipid components that were extracted from the epicuticular layer of Hemicordulia tau dragonfly wings were determined using GC-MS (Table 4.2). Data analysis confirmed the presence of aliphatic hydrocarbons and fatty acids; the major components of all fractions were n-alkanes (Figure 4.10). They represented approximately 71 % of the total compounds present in the chloroform-extracted fraction (10 s extraction time) and the proportion of n-alkanes in the total extraction remained constant for both the 20 and 30 s extraction times (71 % and 69 %, respectively). The extracted alkanes ranged in length from C10 to C34, and were dominated by alkanes with an even number of carbon atoms, with smaller quantities of C19, C23, C25, and C29 (Table

4.2). Among the n-alkanes, n-hexacosane (C26) alone accounted for 26.8 % of the total detected compounds in the 10 s extract, and this increased slightly to approximately 29 % in the 20 s extract, and dropped to 18 % in the 30 s extract. Methylalkanes were also prevalent amongst the aliphatic hydrocarbons extracted from the epicuticular layer of the wing membrane. They comprised up to 19 % of the total compounds present in the

10 s extract, however, their proportion decreased as the time of extraction increased. The only oxygen-containing compound detected in the 10 s extract was palmitic acid (hexadecanoic acid), which made up 7 % of the extracted epicuticular lipids. The proportion of palmitic acid in the wing extracts increased to 18 % and 14 % after 20 se and 30 s of extraction, respectively.

Table 4.2. Wax components of the Hemicordulia tau wing epicuticle isolated and identified by GCMS. Components that could not be unambiguously identified are grouped together.

Proportion (%) Formula Nomenclature 10 s 20 s 30 s

C10H22 n-decane 1.2

C11H24 4-methyldecane 0.1

C14H30 n-tetradecane 4.7 5.1 2.9

C16H34 3-methylpentadecane 0.3 1.0 0.6

C16H34 n-hexadecane 7.0 6.4 5.2

C18H38 3-methylheptadecane 0.9 2.5 1.9

C18H38 n-octadecane 7.7 7.8 7.7 3,15- C H 0.1 19 40 dimethylheptadecane

C19H40 n-nonadecane 2.3

C20H42 7-methylnonadecane 0.1

C20H42 8-methylnonadecane 0.1

C20H42 3-methylnonadecane 1.2 1.6 2.2

C20H42 n-eicosane 6.5 6.0 9.5

C16H32O2 Hexadecanoic acid 7.1 17.9 14.4

C21H44 n-heneicosane 2.8

C22H46 4-methylheneicosane 0.4

C22H46 3-methylheneicosane 0.7 1.6 1.9

C23H48 7,11-dimethylheneicosane 0.7 1.6 1.9

C22H46 n-docosane 12.2 11.2 13.5

C23H48 n-tricosane 0.1 1.8

C24H50 11-methyltricosane 0.1

C24H51 7-methyltricosane 0.1

C24H50 5-methyltricosane 0.1 0.1

C24H50 3-methyltricosane 12.0 2.2 5.9

C25H52 n-pentacosane 1.6 2.0

C27H56 7,11-dimethylpentacosane 2.2 0.3 1.9

C26H54 n-hexacosane 26.8 28.9 18.4

C28H58 2,6-dimethylhexacosane 0.1 0.3 0.1

C26H54 3-methylpentacosane 4.3 0.3 0.2

C28H58 n-octacosane 3.4 3.3 1.7

C29H60 n-nonacosane 0.1

C30H62 n-triacontane 2.1 0.3 0.9

C34H70 n-tetratriacontane 0.3 0.7 0.1 %Composition based on measured TIC peak areas without standardisation.

Aliphatic hydrocarbons represented approximately 93 % of the 10 s, 82 % of the 20 s, and 86 % of the 30 s extract. The change in proportions of aliphatic hydrocarbons allowed an insight into determining the location of certain compounds from different layers of the dragonfly wing epicuticle. For example, in the 10 s extraction, most compounds in the outermost layer of the wing surface were extracted and their proportions did not increase as the extraction time increased. This highlighted the importance that aliphatic hydrocarbons have in determining the wettability of dragonfly wings. Removal of a large portion of the n-alkanes from the surface in the 10 s extracted surface appeared to be the principal cause of the loss of the wing’s self-cleaning ability. Longer extraction times led to an increasing loss of lipid components, and this resulted in further decreases in the measured WCA. Even-numbered n-alkanes were present in high proportion in 10 s extract and this remained stable or slightly decreased in the longer extraction time samples. The dominance of even-numbered aliphatic hydrocarbons can be explained by the addition of acetyl groups to a fatty acid precursor, whereby chain extension occurs in multiples of C2, with the final length of the molecule being dependent on the fatty acid precursor (Howard and Blomquist, 2005; Samuels, Kunst and Jetter, 2008).

4.4.3. Changes in surface morphology on removal of surface aliphatic hydrocarbons

The surface morphology of the wing samples that were subjected to the 10 s chloroform extraction changed as a result of the extraction process, after which time slight changes to the pillar structures was observed. In these samples, the three distinct layers were still clearly visible, however the round tops on the pillar structures present on the untreated wings were no longer present (Figure 4.11C, D). The extraction process caused the pillars to sharpen, and become more sparse, suggesting that the outermost portions of the pillars were dissolved by chloroform extraction. As the extraction time was increased, the progressive loss of surface structure became more obvious. After 20 s (Figure 4.11E, F), the epicuticular layers of the wing appeared thinner in their cross-sectional view, and circular patterns became visible in the surface images. After 30 s of chloroform extraction, minimal structure remained, and this was scattered across the surface, whilst the internal structure remained unaffected (Figure 4.11G, H).

The changes in the surface morphologies observed in the electron micrographs were also reflected in the AFM images. Typical two-dimensional images and cross- sectional profiles of 5 µm × 5 µm scan areas of each of the wing surfaces and insets corresponding to 1 µm × 1 µm scan areas are presented in Figure 4.12. Each cross- sectional profile was generated by selecting one line from the corresponding two- dimensional AFM scan and plotting height as a function of translation along the line. The surface features in the profiles appeared to decrease in height and sharpness as the time of extraction increased. Untreated wing surface profiles contained features ranging from 50 - 150 nm in width. After 10 s of chloroform extraction, the features appeared rounder and shorter (50 - 100 nm). Notably, while scanning the surface of the chloroform-extracted wings, the AFM tip experienced decreased lateral deflection, i.e. the tip adhered to the wing surface to a lesser extent after lipid extraction, suggesting that less adhesive lipid material was present on the surface. As the time of extraction was increased to 20 and 30 s, the surface features decreased further to about 80 nm in height, appeared broader and were spaced further apart (Figure 4.12).

A statistical roughness analysis was performed on the data measured on both the untreated and chloroform extracted wings. Six statistical parameters, Rmax, Ra, Rq, Rsk,

Rku and Rdr, were calculated for each surface (Table 4.3) (Crawford, Webb, Truong et al., 2012). Rmax is the maximum roughness, is the vertical distance between the highest and lowest points on a topographical profile or surface. Ra and Rq are the average roughness; Ra, is the average deviation of the height values from the mean line/plane, whilst Rq is the root-mean-square deviation from the mean line/plane, i.e. the standard deviation from the mean. Skewness (Rsk) and kurtosis (Rku) describe the shape of the height distribution. Positive skewness indicates the surface having high peaks with shallower and broader valleys, whilst the surfaces with negative skewness tend to have small peaks with deep, narrow valleys. Kurtosis is a measure of the ‘sharpness’ of the height distribution, or the sharpness of the peaks in a profile or surface. Rdr is a measure of the surface area relative to the projected area of the scan(Crawford, Webb, Truong et al., 2012).

Table 4.3. Roughness analysis of the untreated and chloroform extracted H. tau wing surfacesa.

Roughness Untreated Chloroform extraction time (s) b statistics sample 10 s 20 s 30 s

Rmax (nm) 814 ± 72 409 ± 40 285 ± 14 332 ± 14

Ra (nm) 47 ± 3 38 ± 11 25 ± 3 40 ± 8

Rq (nm) 61 ± 3 50 ± 11 32 ± 3 49 ± 10

Rdr (%) 278 ± 51 157 ± 54 121 ± 3 143 ± 10 aResults presented are averages of three measurements taken from 25 µm2 scanning areas and errors are the standard deviations. bA technical description of the roughness parameters used can be found in Crawford et al. (Crawford, Webb, Truong et al., 2012).

It can be seen that Rdr decreased as the extraction time increased, and compared to the untreated sample. A similar trend was observed for Rmax, Ra, and Rq. The average roughness (Ra) of the untreated wing surface was found to be 47.0 nm, which gradually decreased to as low as 25.2 nm after 20 s of chloroform extraction. The roughness of the wings subjected to a 10 s chloroform extraction also appeared to decrease, however the changes were not significant. In contrast with the 10 s and the 20 s extracted samples, the roughness parameters of the 30 s extracted wing increased. Extended extraction time may have led to the removal of a greater amount of material from between the pillars than on their upper regions. This would enable the AFM tip to reach deeper into the sample, hence causing the height of the pillars to appear to increase in size. The loss of surface nanostructure with extended extraction time highlights the fact that the components of the wing that are rapidly soluble in chloroform, i.e. the epicuticular lipids, play a major role in forming the wing nanoarchitecture.

4.5. Summary

The dragonfly wing membranes consist of three distinct layers, including two epicuticles on the dorsal and ventral surfaces and an intracuticle located in between the epicuticles. Through Synchrotron FTIR spectroscopy, it was found that the epicuticular layers consisted mainly of aliphatic hydrocarbons, with protein and chitin making up the bulk of intracuticle. Combining the data obtained using two techniques, FTIR and WCA measurement, revealed a correlation between the quantity of lipids and wettability of the wing surface. Step-wise extractions of the wings using chloroform resulted in decreased surface hydrophobicity, with the WCA of the treated samples falling below the threshold that defines the condition of superhydrophobicity within 10 s of extraction. The loss of surface structure on the wing surface due to extraction with chloroform leads to two important conclusions. Firstly, it is not only the loss of hydrophobic material from the epicuticle that caused the wing to lose its superhydrophobic and self- cleaning properties; loss of structure was almost certainly also a contributing factor. Secondly, the removal of the wing nanostructures via chloroform extraction enables the conclusion to be drawn that the nanostructures present on the wing surfaces are composed of the materials that were identified in the chloroform extracts. It can be concluded that the nanoscale pillars present on the surface of Hemicordulia tau dragonfly wing membranes are composed primarily of aliphatic hydrocarbons, especially n-alkanes with even-numbered chain lengths ranging between C18 and C26, 79 and a relatively small proportion of palmitic acid. These components are vital in determining the superhydrophobic and self-cleaning properties of Hemicordulia tau wing membranes, as they not only present a hydrophobic lipid layer to the external environment, but are the primary building materials that form the nanostructures on the wing surface.

Chapter 5. Molecular organisation of dragonfly wings ______

5.1. Overview

As noted the chemical composition of dragonfly wings is a critical factor in determining their surface properties and the molecular organisation of the wing surfaces is an important parameter when it comes to constructing dragonfly wing-inspired materials. A combination of chromatographic and spectroscopic techniques was employed to analyse the wings. Lipids were extracted from the wing surfaces and analysed using gas chromatography-mass spectrometry (GCMS), whilst Synchrotron- sourced Fourier-transform infrared microspectroscopy (FTIR) and x-ray photoelectron spectroscopy (XPS) depth profiling were performed directly on the intact wings. The identities of over 60 chemical components that constitute the nanoscale structures on the surface of the wings of the dragonfly, Hemianax papuensis, were defined. The wing samples were immersed in chloroform for either 10 s or 1 h. Immersion of the wing in chloroform for just 10 s enabled the outermost components of the wing to be preferentially extracted, whilst the 1 h extractions enabled the analysis of the chemical composition of the entire epicuticular layer. To confirm that the intended regions of the epicuticle were successfully extracted, the pre- and post-extracted surfaces were analysed for changes in their surface topologies using SEM and AFM.

5.2. Physical effects on dragonfly wings resulting from extraction with chloroform

5.2.1. Changes in surface morphologies

Scanning electron micrographs confirmed that a chloroform extraction period of 10 s was sufficient to extract the outer epicuticle of the wing. It can be seen from cross- sections of the untreated and the 10 s extracted sample (Figure 5.1) that the wing surface structures remained largely unchanged after the 10 s extraction period. The cross- sectional views of the wing were obtained in order to visualise the wing surface as well as the internal structure of the dragonfly wing membrane. The extraction process did not appear to cause visible damage to the intracuticle, even after a 1 h extraction period. The nanostructure of the epicuticular layer, however, was completely eradicated after 1 h of extraction.

Surface view electron micrographs of the untreated and chloroform-extracted wings were also obtained, and are presented with their corresponding 3D images (Figure 5.2). The 3D images were generated from the micrographs using Gwyddion data processing software (Nečas and Klapetek, 2012). After a 10 s chloroform extraction period, the nanostructures on the wing surfaces were still visible, however the spatial distribution between individual pillars appeared to have increased. This observation was due to removal of material from the surface of these nanostructures during the extraction process, causing them to decrease in size. The wing surface appeared to be completely devoid of nanostructures after a 1 h chloroform extraction period. The results obtained here demonstrate that the 10 s extraction period was successful at removing just the outermost layer of lipids, whereas a 1 h extraction period enabled the entire epicuticle layer to be removed.

5.2.2. Changes in surface topographies

The changes in the surface topographies of the Hemianax papuensis wing surface that were observed in the electron micrographs were also reflected in the AFM analysis. Typical two-dimensional images and cross-sectional profiles of 1.0 µm × 1.0 µm AFM scan areas of each of the wing surfaces are presented in Figure 5.3. Each cross-sectional profile was generated by selecting one line from the corresponding two-dimensional

AFM scan and plotting height as a function of translation. The surface features in the profiles appeared to decrease in sharpness as the period of chloroform extraction increased. Untreated wing surface profiles contained features ranging from 50 - 100 nm in width. After 10 s of extraction with chloroform, the features appeared to have increased in their spatial distribution and the peaks appeared broader and higher compared to that of the untreated wing. These AFM data were in agreement with the SEM images. Extraction with chloroform may preferentially remove more material from the regions between the pillars rather than from the areas nearer to the apices of the pillars. This would enable the AFM tip to reach deeper into the surface of the samples that had been subjected to chloroform extraction for a 10 s period compared to that of the untreated samples, and therefore the height of the pillars appeared to increase in size. As the time of extraction was increased to 1 h, the surface nanopillars completely disappeared. The maximum height of the features on the 1 h extracted surface ranged from 5 - 10 nm. This most likely reflected the morphology of the intracuticular surface on which the epicuticular lipids had been assembled. It has been reported in the literature that the lipids found on the surface of wings are transported to the surface of insect cuticles through a series of pore channels, where they form the epicuticular nanopillar structures (Lockey, 1980; Lockey, 1985).

The roughness of the wings was quantitatively assessed by calculating a number of roughness parameters for each of the samples (Table 5.1). These parameters were Rmax,

Ra, Rq, Rsk, Rku, and Rdr, as detailed in Section 4.4.3. Relative to the untreated control

87 wing surface, the Rdr increased slightly after a 10 s extraction period and then sharply decreased when the extraction time was increased to 1 h. A similar trend was observed for Rmax, Ra and Rq. The average roughness (Ra) of the untreated wing surface was found to be 49.2 nm, which then increased to 67.5 nm after 10 s of extraction. The wing surface after a 1 h extraction period was found to be very smooth, with both Ra and Rq decreasing significantly (2.4 nm and 3.1 nm respectively) compared to both the untreated and samples that had been chloroform extracted for 10 s. These observations appear to correlate well with the selected line profiles presented in Figure 5.3.

Table 5.1. Roughness analysis of the untreated and chloroform extracted Hemianax papuensis wing surfacesa

Roughness Chloroform Chloroform Untreated statistics extraction for 10 s extraction for 1 h

Rmax (nm) 396 ± 30 491 ± 40 21 ± 5

Ra (nm) 49.2 ± 7.4 67.5 ± 5.8 2.4 ± 0.6

Rq (nm) 61.6 ± 7.6 83.2 ± 6.2 3.1 ± 0.8

Rsk 0.3 ± 0.1 0.4 ± 0.1 0.2 ± 0.1

Rku 2.6 ± 0.1 2.7 ± 0.2 2.7 ± 0.2

Rdr (%) 277 ± 24 306 ± 45 103 ± 2 aResults presented are averages of five measurements taken from 1 µm2 scanning areas and errors are the standard deviations.

The relative percentages of surface area compare to projected area (Rdr) were approximately 277 – 306 % in the case of the untreated and 10 s extracted wing (Table 5.1). This directly indicates the rough nanoscale nature of the surfaces of the wings. After the wings were extracted with chloroform for 1 h, removal of the lipids led to elimination of the nanostructure, and therefore the measured surface areas were found to be much closer to the projected area (i.e. Rdr = 103 %). This demonstrates that the underlying intracuticle of H. papuensis dragonfly wings is relatively smooth before the epicuticular lipids are transported to the surface to form the epicuticular structures.

5.3. The chemical composition of the Hemianax papuensis dragonfly wing epicuticle

Synchrotron-radiation FTIR microspectroscopy was initially used to confirm that subjecting the dragonfly wings to chloroform extraction led to a decrease in the lipid content on the surface of the wings but had minimal effect on the non-lipid components. Synchrotron radiation allows for the generation of small infrared beam spots that enable the analysis of small areas of the wing. Multiple scans can be conducted in a short period of time within a certain area to validate the consistency of the chemical composition as well as the reliability of quantitative changes of the chemistry over different chloroform extraction times. The characteristic FTIR spectra of the dragonfly wings before and after extraction with chloroform contained three major groups of bands (Figure 5.4) that were very similar to those obtained for the Hemicordulia tau dragonfly wings presented in Section 4.4.2. Of particular interest are the two peaks at -1 -1 2830 cm and 2930 cm , which correspond to symmetrical and antisymmetrical CH2 stretching vibrations. These two bands are of relatively high intensity in the spectra recorded for the untreated wings and are indicative of long chain aliphatic hydrocarbons with repeating CH2 units, i.e. lipid components. After 10 s of extraction with chloroform, the intensity of both peaks decreased visibly, indicating that the lipids were successfully extracted by the chloroform. After 1 h of extraction, the peaks decreased in intensity further but were still present. This is to be expected, as proteins, chitin and other organic molecules that constitute part of the wing membrane contribute to the extent of absorption in this region of the spectrum (Andersen and Weis-Fogh, 1964; Hepburn and Chandler, 1976; Neville, Parry and Woodhead-Galloway, 1976; Hackman and Goldberg, 1987; Gorb, 1997; Nelson and Charlet, 2003; Bogus, Czygier, Golbiowski et al., 2010; Sajomsang and Gonil, 2010). The small ester carbonyl peak, which appeared at approximately 1700 cm-1, also disappeared from the sample that was subjected to 1 h of chloroform extraction, whilst it was still visible in the spectrum for the wing that had been subjected to the shorter 10 s of chloroform extraction (Figure 5.4, indicated by arrows).

While surface analytical techniques such as FTIR microspectroscopy are very useful for generating maps and locating specific chemical functionalities, they usually return a signal or spectrum that represents a sum of all of the chemical components at that interface. This makes detailed molecular identification difficult or impossible, particularly with complex samples. To overcome this limitation, a chromatographic analysis of the lipid extracts was conducted using GCMS. Analysis of the GCMS data revealed that the major lipid components of the wings were comprised mostly of aliphatic hydrocarbons, with a significant contribution from fatty acids (Table 5.2).

Proportion (%) Formula RI* Chemical name 10 s 1 h

CHCl3 extraction CHCl3 extraction

C7H12O 965 2-Heptenal 0.1 0.2

C12H26 1200 n-dodecane 0.2

C14H30 1371 3-methyltridecane 0.1

C14H30 1400 n-tetradecane 5.4 2.3

C15H32 1540 3-methyltetradecane 0.1

C16H34 1571 3-methylpentadecane 1.5 0.7

C16H32 1593 1-hexadecene 0.1

C16H34 1600 n-hexadecane 10.6 5.7

C18H38 1728 8-ethylhexadecane 0.1

C18H38 1728 7-ethylhexadecane 0.1

C18H38 1772 3-methylheptadecane 4.2 1.5

C18H36 1793 1-octadecene 0.1

C18H38 1800 n-octadecane 10.3 6.7

C18H38 1799 6,12-dimethylheptadecane 10.3 6.7

C18H38 1888 7,12-dimethylhexadecane 0.1

C20H42 1888 7,12-dimethyloctadecane 0.1

C20H42 1915 6,7-dimethyloctadecane 0.1

C20H42 1921 5,6-dimethyloctadecane 0.1

C20H42 1928 6,10,14-trimethyloctadecane 0.2

C20H42 1944 7-methylnonadecane 0.1

C20H42 1944 8-methylnonadecane 0.1

C20H42 1966 2-methylnonadecane 0.2

C20H42 1973 3-methylnonadecane 2.1 1.7

C20H42 2000 n-eicosane 8.5 6.3

C16H32O1 2046 Hexadecanoic acid 24.9 32.7

C22H46 2116 7-ethyleicosane 0.2

C22H46 2133 11-methylheneicosane 0.1

C22H46 2142 5-methylheneicosane 0.1

C22H46 2142 7-methylheneicosane 0.1

C22H46 2142 9-methylheneicosane 0.1 (E)-3,7-dimethyl-2,6- C10H16O2 2162 0.1 octadienoic acid

C22H46 2170 3-methylheneicosane 1.7 1.9

C23H48 2170 7,11-dimethylheneicosane 1.7 1.9

C22H46 2200 n-docosane 6.9 6.7

C18H36O2 2244 Octadecanoic acid 13.0

C24H50 2282 9-propylheneicosane 0.2

C24H50 2315 9-ethyldocosane 0.3

C24H50 2373 3-methyltricosane 1.6 2.0

C25H52 2415 12-ethyltricosane 0.1

C25H52 2431 11-methyltetracosane 0.2 0.2

C26H54 2480 12-propyltricosane 0.1

C26H54 2507 11-ethyltetracosane 0.1

C27H56 2525 11,15-dimethylpentacosane 0.1

C26H54 2532 9-methylpentacosane 0.1

C26H54 2532 10-methylpentacosane 0.1

C26H54 2532 11-methylpentacosane 0.1

C26H54 2532 13-methylpentacosane 0.1

C26H54 2544 7-methylpentacosane 0.1

C26H54 2544 5-methylpentacosane 0.1

C27H56 2544 9,13-dimethylpentacosane 0.1

C26H54 2575 3-methylpentacosane 0.7 1.6

C26H54 2600 n-hexacosane 9.9 7.1

C28H58 2704 2,6-dimethylhexacosane 0.1

C28H58 2800 n-octacosane 2.9 4.9

C29H60 2900 n-nonacosane 0.1

C29H60 2973 3-methyloctacosane 0.4 1.0

C30H62 3000 n-triacontane 2.4 3.1

C32H66 3097 2,10-dimethyltriacontane 0.5 0.1

C32H66 3097 2,12-dimethyltriacontane 0.5 0.1

C31H64 3100 n-hentriacontane 0.1 0.1

C33H68 3172 7,23-dimethylhentriacontane 0.1 0.1

C32H66 3172 3-methylhentriacontane 0.1 0.1

C33H68 3172 7,25-dimethylhentriacontane 0.1 0.1

C32H66 3200 n-dotriacontane 0.3 2.1

C35H72 3369 7,19-dimethyltritriacontane 0.1

C35H72 3369 7,21-dimethyltritriacontane 0.1

C34H70 3400 n-tetratriacontane 0.1 0.1

C32H52O2 3578 Retinyl dodecanoate 0.1

C36H74 3600 n-hexatriacontane 0.1 % Composition based on measured TIC peak areas without standardisation. *RI = Retention index

Palmitic acid (Hexadecanoic acid) was found to be the major component of both the 10 s and 1 h extracts (approximately 24 % and 33%, respectively). A higher proportion of palmitic acid was detected in the 1 h extracts, suggesting that 10 s of extraction was not sufficient to remove this compound completely from the surface. This indicated that palmitic acid was localised not only to the outer epicuticular layer but was also present deeper within the wing. Stearic acid (Octadecanoic acid) was only present in the 1 h extract and had relatively large contribution to the total components of the epicuticular lipids in that sample (13%). Apart from these carboxylic acids components, straight chain alkanes as a group were the most abundant components, among which hexadecane, octadecane, eicosane and docosane were present in the greatest quantities. These six components (palmitic acid, stearic acid, hexadecane, octadecane, eicosane and docosane) were selected for study of their molecular organisation and assembly on synthetic substrata, which is the focus of Chapters 7 and 8.

Straight-chain alkanes made up approximately 50 % of the compounds detected in the 10 s chloroform extracts, with an additional 25 % being monomethylated (11 %) and dimethylated alkanes (14%). The remaining 23 % of the extract could be attributed to oxygenated compounds; more specifically this group was almost entirely comprised of palmitic acid. Palmitic acid, is a fatty acid widely distributed among plants, animals and insects (Smith, 1994; Milligan, Parenti and Magee, 1995; Dowhan, 1997; Buckner and Hagen, 2003; Carballeira, 2008; Samuels, Kunst and Jetter, 2008; Küçükbay, Kuyumcu, Bilenler et al., 2012; Mariod and Abdelwahab, 2012). After 1 h of chloroform extraction, the proportion of oxygenated compounds increased to 38 % of the identified compounds, largely due to the quantity of stearic acid. Since stearic acid was totally

93 undetected in the 10 s extracts, therefore it suggests that stearic acid is primarily situated deeper within the wing membrane, beneath the epicuticular surface.

Figure 5.5. Relative proportions of the major compound classes (top) and chain length (bottom) of dragonfly wing epicuticle components, extracted in 10 s and 1 h.

The aliphatic hydrocarbon constituents of both the 10 s and the 1 h extracts contained several main compounds in relatively high abundance. In addition, both extracts contained significant proportions of n-alkanes with even numbers of carbon atoms ranging from C14 to C30. Biosynthesis of n-alkanes occurs by the addition of acetyl groups to a fatty acyl precursor conjugated to co-enzyme A (Howard and Blomquist, 2005; Samuels, Kunst and Jetter, 2008). The chain-lengths of the resulting hydrocarbons are therefore in multiples of 2 and are dependent on the length of the fatty-acyl chain of the precursor. The composition of the minor constituents varied

94 significantly between the two extraction times, with considerably more compounds being detected in small quantities after 1 h of extraction. Most of these minor constituents were identified as various methyl-substituted alkanes, however three alkenes were detected in the 1 h extract; 1-hexadecene, 1-octadecene, and 1-eicosene. In addition, two esters, one aldehyde and one terpenoid were detected in very small proportions.

5.4. Molecular organisation of H. papuensis dragonfly wing epicuticle

X-ray photoelectron spectroscopy (XPS) is capable of nanometre resolution in the z- dimension (Wagner, 1978; Wagner, Gale and Raymond, 1979; 1983; Escobar Galindo, Gago, Duday et al., 2010). Depth-profiling can be performed using XPS by alternating data acquisition and ion-beam etching. The resulting spectra were found to be dominated by carbon (Figure 5.5). This was to be expected, as it was already well evidenced by FTIR and GCMS analyses that organic compounds were the major components of the wing membrane. Further analysis of the nitrogen and oxygen contents were of greater interest. After the first etching cycle, the oxygen content on the wing surface decreased to almost zero. After 6 etching cycles, the detected oxygen levels began to increase and quickly stabilised at around 2 %. The contribution of nitrogen components was insignificant until after the tenth etching cycle, after which it increased to around 6 %. This is an indication that after 10 etching cycles the protein and chitin structural components in the bulk of the wing were close enough in proximity to the surface to be detected in the spectra. Based on this observation, a tentative estimate of the etching rate can be made; dragonfly wings are typically ~3 μm thick (Kreuz, Arnold and Kesel, 2001; Song, Xiao, Bai et al., 2007; Wan, Cong, Wang et al., 2008; Chen, Wang, Ren et al., 2011) and from the images presented in Figure 4.2 the epicuticular layers of the wings of the H. papuensis dragonfly can be estimated to account for one tenth of the cross-section of the wing. Thus, in 10 etching cycles, i.e. 20 minutes, the wings were etched to a depth of approximately 300 nm, which corresponds to an etching rate of 15 nm min-1.

Figure 5.6. Atomic proportions of carbon, oxygen and nitrogen in the wings of Hemianax papuensis.

When one considers the XPS depth profiling data in conjunction with the GC-MS data, a tentative model of the molecular organisation of the wing membrane can be proposed. According to the XPS analysis, the oxygen content dropped to zero after just one etching cycle (Figure 5.5). From the GC-MS analysis, the only significant oxygen

96 containing compound in the 10 s lipid extracts was palmitic acid, which suggested the presence of a thin layer at the outer surface of the wing containing palmitic acid (Figure 5.6, Table 5.2). Lipid extracts after 1 h of chloroform extraction contained a significant proportion of stearic acid, along with an increased proportion of palmitic acid. XPS data indicated that the oxygen content only began to increase again after 6 etching cycles, suggesting the presence of a deeper layer that contained both palmitic acid and stearic acid. Aliphatic hydrocarbons were detected in large quantity in all extracts and in the regions between these two fatty acid-containing layers, carbon was the only element detected in XPS spectra.

Exposure of biological samples to ultra-high vacuum (UHV) can cause dehydration, which may conceivably lead to the collapse of the intracuticle. This work focuses on the epicuticular layer of the wing, which is much less hydrated as it does not contain polar components such as sugar or protein (Lapointe, Hunter and Alessandro, 2004). Therefore, UHV should not interfere with the reliability of data obtained here. Based on this information and the etching rate given above, a model of the epicuticle of the H. papuensis dragonfly wing is proposed that includes three distinct layers, one of which was never previously reported for insect epicuticles (Figure 5.7). It is proposed that the new layer be referred to as the mesoepicuticle, that it is positioned between the outer and inner epicuticles, and contains almost exclusively saturated aliphatic hydrocarbons.

5.1.Summary

While the outer epicuticle is generally accepted in literature to be composed of lipids, the precise identities of the compounds that form the surface nanostructures of the H. papuensis dragonfly wings have not previously been identified. Here, using a combination of surface analytical techniques, including SEM and AFM, and the complementary chemical analytical techniques of Synchrotron-sourced Fourier- transform infrared microspectroscopy (FTIR), gas chromatography-mass spectrometry (GCMS) and x-ray photoelectron spectroscopy (XPS), the epicuticular lipids of dragonfly wings were identified. Removal of the epicuticular lipids led to the disappearance of the surface nanostructures, revealing that the nanostructures themselves were composed of lipids and that the native surface structure of intracuticle was relatively smooth. XPS depth profiling, in conjunction with GCMS analysis of the

98 chloroform extracts, enabled the detection of three distinct layers within the epicuticle: the outer epicuticle, containing mainly palmitic acid; the mesoepicuticle, composed of aliphatic hydrocarbons; and the inner epicuticle, which contained palmitic acid and stearic acid. The identification of a third sub-layer of the epicuticle of insect wings, i.e. the mesoepicuticle, is now reported from this work.

Chapter 6. Nanostructural effects on antibacterial activity and wettability of dragonfly wings ______

100

6.1. Overview

In Chapters 4 and 5, the surface chemistry and the surface topography of different species of dragonfly wings were investigated and characterised in the context of their wettability. This chapter will address how this wettability relates to the antibacterial properties of the wings. The bactericidal activities of Hemianax papuensis and Diplacodes bipunctata dragonfly wings were evaluated to determine the consistency of bacterial activity between different dragonfly species. It was found that H. papuensis dragonfly wings were only moderately effective at inactivating some types of bacterial cells, whilst D. bipunctata dragonfly wings were highly lethal to all tested cell types. This chapter presents the results of work that was undertaken to determine whether the surface chemistry and/or surface topography of the dragonfly wings contribute to this variation in bactericidal activity. Complementary techniques were employed to investigate the mechanisms responsible for these differences; specifically, the surface morphologies of the wings were characterised using SEM and AFM, whilst the surface chemistry was assessed using Synchrotron-based FTIR and XPS.

The two dragonfly wing samples were found to be comprised of very similar chemical components, primarily aliphatic hydrocarbons. These compounds are hydrophobic in nature, partially contributing to the superhydrophobic properties of the wings. The surface roughness of the substrata was found to be the factor that played the greatest role in determining the differences in functionality between the wings of the two species. Despite the visual similarities of the wings, subtle differences were found in (i) their surface morphologies and (ii) their surface topographies. H. papuensis dragonfly wings possessed a higher surface area, were generally rougher and possessed a more hierarchical surface structure than the wings of the D. bipunctata wings, hence enabling the surface features to trap more air pockets within their structure, contributing to the surface hydrophobicity. For this reason, the surface of H. papuensis dragonfly wings were less accessible to bacterial attachment compare to D. bipunctata. The amount of trapped air determined how deep water droplets placed in contact with the surface could penetrate the surface in between the pillars, as evidenced in their differences in water contact angles. This may also be the case for bacterial cells; the depth to which they could impinge upon the surface would likely determine their susceptibility to the bactericidal action of the wings.

101

6.2. Bactericidal activities of dragonfly wings

The wings of the H. papuensis dragonfly were found to have moderate bactericidal activity. They were somewhat effective at killing P. aeruginosa, S. aureus cells and spores of B. subtilis, however the surface was found to be ineffective against vegetative B. subtilis cells. This can be observed in the confocal laser scanning microscopy (CLSM) images presented in Figure 6.1. On the other hand, D. bipunctata wings were found to be very effective against all tested cell types (P. aeruginosa, S. aureus, B. subtilis cells and spores). In general, the green fluorescence signal, which indicated the presence of live bacterial cells, was greater on the H. papuensis wings, while the cells on D. bipunctata wings almost exclusively exhibited red fluorescence, indicating the presence of dead cells. As for the largely resilient B. subtilis spores, a lower killing efficiency was observed on the wings of both dragonfly species, however, a greater number of dead cells were found to be present on the surface of D. bipunctata wings compared to that of H. papuensis wings. Yellow-fluorescing cells indicated the binding of both dyes, which is an indication that those cells are in fact dead, as the red propidium iodide is only able to enter cells with significant membrane damage.

The adhesion and morphology of bacterial cells after they had adsorbed onto the surfaces of the two dragonfly wings were observed using scanning electron microscopy (Figure 6.2). In both cases, P. aeruginosa cells appeared to sink deeply between the nano-features of the two dragonfly wings. The structure of these cells appeared severely deformed on both surfaces, which would explain the red fluorescence in CLSM images; as such strong deformation would very likely be accompanied by cell rupture. S. aureus, B. subtilis and spores also appeared deformed on the wings of D. bipunctata, however they were relatively unaffected by the action of the surface of the H. papuensis wings.

Bacterial viability assays in the form of serial dilutions and colony counts on agar plates which was employed to identify the bactericidal efficiency of the two dragonfly wings, based on the 3 hours assay. The data revealed that the bactericidal efficiency of the H. papuensis wings against S. aureus and B. subtilis spores after 3 hours was 5.3 × 104 and 4.5 × 104 cells killed per min-1 per cm-2 of wing surface, respectively (Figure 6.3). The next highest activity of this wing was against P. aeruginosa cells, however the efficacy against all types of cells was much less than that of D. bipunctata cells, which

102 in all cases was capable of killing at least twice as many cells in the same period of time.

The two dragonflies being studied are native to Australia. The surface morphologies of H. papuensis and D. bipunctata wings appeared to be highly similar when viewed using SEM, however the difference in their ability to kill bacterial cells is both striking and clear. Based on the confocal and SEM images, the surfaces of D. bipunctata wings were much more effective at killing bacterial cells than those of H. papuensis wings. In order to determine how such seemingly similar wing morphologies result in vastly different functionalities, various techniques were employed to further discriminate between the nanostructures present on each of these wing types.

103

6.3. Comparative surface chemistry of the dragonfly wings

Despite their differences in efficiency of killing bacterial cells, the elemental chemical compositions across the wing membranes of H. papuensis and D. bipunctata wings were highly similar. Typical infrared spectra containing the characteristic peaks of the two dragonflies across the wing membrane were dominated by the three major bands, which represent aliphatic hydrocarbons, carbonyl compounds and amides (Figure 6.4).

105

XPS is a surface sensitive method for quantifying elemental compositions, therefore it is an important technique for this study. The XPS data showed that the major elements present on the surface of the wing were carbon and oxygen. This is consistent with what was observed from FTIR data. The carbon peaks were dominant over the oxygen peaks, which indicated that there is much larger proportion of carbon at the outer surface of the wing. The high resolution scans of the C 1s peak showed the presence of, saturated carbon (i.e. C–C and C–O), and O-C꞊O chemical configurations, whereas two oxygen configurations were detected (O–C and O꞊C), consistent with the data presented in Chapter 5. This indicated that the major components present on the outer layer of the wings were aliphatic alkanes and fatty acids (Section 5.3 and 5.4).

106

The chemical analyses based on data obtained from Synchrotron-radiation sourced FTIR and XPS demonstrated the existence of high similarities between the two dragonfly species. No substantial differences in the chemical composition of the wings of H. papuensis and D. bipunctata could be detected, which further supports the concept that the bacterial killing action is mechanobactericidal in nature. It follows that the difference in functionality of the wings is based primarily on morphology and structure, therefore further investigation of their surface topographies were conducted in order to determine the cause of the variation in bactericidal activity. 107

6.4. Variation in surface structure

6.4.1. Surface topologies

The SEM images of the two dragonfly wings provide an indication of their surface morphology. The surfaces of the wings of both species appeared similar, however differences are evident visually (Figure 6.6). Here it can be seen in the high resolution SEM images that the nanopillars on the surface of H. papuensis wings appeared thicker and more densely packed than those of D. bipunctata wings. The tips of the pillars on the H. papuensis wings also appeared comparatively round and thick whereas the pillars appeared sharper and thinner in case of the D. bipunctata wings.

A 2D Fourier Transform (FT) analysis was applied to electron micrographs of both wings at various magnifications to provide a quantifiable measurement of their respective topographies. Fourier transformation is a tool in image analysis used to quantify the size and spacing between surface features (Burger and Burge, 2007). Figure 6.7 shows the FT power spectrum of SEM images of the wings of H. papuensis and D. bipunctata dragonflies.

108

109

The data obtained from FFT of the two wings provided quantitative estimate for the dimension of the pillars and their spatial distribution based on the vertical view of their scanning electron micrographs. This showed a maximum pillar diameter of 70 nm for the H. papuensis wing compared to about 100 nm for D. bipunctata wing; the latter exhibiting a greater degree of tailing to the larger diameters of the distribution, which was evidence of a higher level of aggregation of these features at the tips. Since the analysis is based on a vertical image from SEM, it only reflects patterning in the x- and y-plane. By imaging cross sections of the wing membranes of the two species, the dimensions of the pillars in the z-dimension can be directly measured (Figure 6.8).

110

The nanopillars on the H. papuensis wings appeared to have multiple layers with respect to their height, whilst those on the D. bipunctata wings appeared more regular. The height of pillars present on the H. papuensis wings ranged from 155 nm to 290 nm, whereas the pillars on the D. bipunctata wings fluctuated in height around 265 nm. This data correlated relatively well with the results obtained from the AFM analysis. Typical two-dimensional AFM images and cross-sectional profiles of 2.5 µm × 2.5 µm scan areas of the two dragonfly wings are presented in Figure 6.9. As can be seen in the two- dimensional AFM images, the pillars appeared irregular, with more outliers (bright and dark areas based on colour scales at the right of each AFM images) for H. papuensis compared to D. bipunctata wings. The zero in the line profiles was defined as the mean planes of corresponding scans. The surface features in the profiles appeared taller and

111 more irregular for the H. papuensis wings. The maximum features were 300 nm for H. papuensis and 250 nm for D. bipunctata wings.

Again, the six standard roughness parameters, i.e., Rmax, Ra, Rq, Rsk, Rku and Rdr, were calculated for each wing surface and these are summarised in Table 6.1. The maximum roughness (Rmax) of H. papuensis wings appeared larger in comparison to those of the D. bipunctata wings (481 nm compared to 371 nm). There were minor variations in Ra and Rq between the two species; e.g. the average roughness (Ra) of the H. papuensis wing was 50 nm, whereas this was 43.38 nm for the D. bipunctata wing.

Skewness (Rsk) and kurtosis (Rku), which have been used as an indication of the spatial variation in height, show that high similarity exists between the two species. Minor differences were observed with D. bipunctata wings possessing a slightly higher degree

112 of skewness. The Rdr measured 245% for H. papuensis and 204% for D. bipunctata wings, which demonstrated that the real surface area of the former species was significantly higher. These differences in surface roughness can be considered to be statistically significant in comparison to the surface roughness variations that can occur on wing surfaces within the same species, as indicated by the error values. Previously reported data also suggests that variations in the surface morphologies of insect wings within a single species is negligible (Watson, Cribb and Watson, 2011).

Table 6.1. Roughness analysis data for the two dragonfly wing surfacesa, H. papuensis and D. bipunctata dragonflies.

Roughness parameterb H. papuensis D. bipunctata

Rmax (nm) 482 ± 7 371 ± 8

Ra (nm) 50.4 ± 0.7 43.4 ± 0.8

Rq (nm) 64 ± 1 55 ± 1

Rsk 0.4 ± 0.1 0.4 ± 0.1

Rku 3.2 ± 0.1 3.2 ± 0.1

Rdr (%) 245 ± 6 204 ± 2 aResults presented are averages of five measurements taken from 2.5 µm2 scanning areas and errors are the standard deviations. bA technical description of the roughness parameters used can be found in Section 4.4.3 and in Crawford et al. (Crawford, Webb, Truong et al., 2012).

6.4.2. Surface morphologies

Other parameters often used to evaluate surfaces in materials engineering are the bearing statistics (Gadelmawla, Koura, Maksoud et al., 2002; Butt, Cappella and Kappl, 2005). Bearing area is determined as the area occupied by the surface in a single plane at a given depth measured from the highest point of the surface, and is usually expressed as the ratio of bearing area to the projected area of the plane, i.e. the bearing ratio, tp. The bearing ratios of the two dragonfly species were plotted as a function of depth (Figure 6.10). From these two graphs, a series of bearing statistics was extracted, i.e. peak height Rpk, core roughness Rk and valley depth Rvk, each of which is given in Table 6.2.

113

Table 6.2. Bearing statistics of the H. papuensis and D. bipunctata dragonfly wing surfaces.

Bearing parameter H. papuensis (nm) D. bipunctata (nm)

Rpk 28.6 28.3

Rk 135.6 98.2

Rvk 50.8 65.2

The parameters Rk, Rpk and Rvk are defined as core roughness, peak height, and valley depth, respectively (Crawford, Webb, Truong et al., 2012). In order to identify these parameters, the height was plotted as a function of the bearing area. The tangent with the maximum slope within the middle 40% of the bearing area was extended in both directions until it intersected tp = 0 and tp = 100%. The difference in height of those

114 two intercepts was identified as the core roughness. The value on the bearing curve at the depth of both tangential intercepts was located and tangents to these points were added to intersect tp = 0 at the base of the curve and tp = 100 at the upper end. The difference between these two additional intercepts and the intercepts of the original tangent gave the peak height Rpk (at the base of the curve) and valley depth Rvk (at the top of the curve). Detailed descriptions of these parameters can be found in the literature (Thomas, 1998; Gadelmawla, Koura, Maksoud et al., 2002; Crawford, Webb, Truong et al., 2012). All of the bearing parameters were larger for the H. papuensis compared to D. bipunctata wings, particularly with respect to the core roughness and valley depth. The bearing statistics suggested that the surface of the H. papuensis wings was more irregular and rougher than that of the D. bipunctata wings. Also, the core roughness Rk is lowest on the surface of D. bipunctata wing (98.2 compared to 135.6 of H. papuensis wing) suggesting that the sides of the D. bipunctata wing pillars are much steeper than those of the surface of H. papuensis wings, which was in agreement with the roughness analysis and electron micrographs previously discussed.

6.5. Structure-dependent wettability

An additional 2D Fourier Transform analysis was applied to SEM images of both wing samples at various magnifications to provide a further quantifiable measurement of their respective topographies. The FT power spectrum of SEM images at various magnifications of H. papuensis and D. bipunctata wings is given in Figure 6.11 and Figure 6.12, respectively. At the two lower magnifications (15000× and 35000×), the inflection points in the data indicated that the surface of H. papuensis wings had larger peak features compared to those of the D. bipunctata wings. Specifically at 15000× magnification, inflection points were observed at approximately 110 nm and 200 nm for D. bipunctata, and H. papuensis wings, respectively, which represents the average distance between the asperities on the surface. At higher magnification (35000×) a second inflection point was observed for each species, at approximately 42 nm and 70 nm for D. bipunctata and H. papuensis wings, respectively, which corresponds to the typical diameter of the asperities on each of the respective wings. At the highest magnification of 70000×, the most common length scale remains 42 nm for the D. bipunctata wings but a third feature was observed on the wings of the H. papuensis samples at 34 nm, which is indicative of the presence of a hierarchical structure (Figure 6.12). 115

116

The D. bipunctata wing possessed only a single scale of roughness, which consists of pillars that are approximately 42 nm in diameter and spaced 110 nm apart. Conversely, dual scales of roughness were detected on the H. papuensis wings, which consist of pillars approximately 70 nm in diameter and spaced approximately 200 nm apart, they are interwoven with bridges between the major asperities that are approximately 34 nm in width. This hierarchical structure of the H. papuensis wing indicated by this Fourier transform analysis matches well with the description of Cassie- Baxter structure (Cassie and Baxter, 1944; Nishimoto and Bhushan, 2013; Papadopoulos, Mammen, Deng et al., 2013; Yang, Jin, Liu et al., 2013; Zhang, Wang and Levänen, 2013) and hence gives a stable superhydrophobic and self-cleaning surface. Thus the H. papuensis wing exhibited a higher static water contact angle (161°) compared to that of the surface of the D. bipunctata wings (151°).

117

Insect wings, including cicada, damselflies, and dragonflies are well-known for their superhydrophobic nature. It was also well established in the literature that for superhydrophobic surfaces, water droplets are most likely to be in the Cassie-Baxter wetting state. Briefly, the Cassie-Baxter theory explains the wetting behaviours of water droplets on a composite or highly rough solid surface. According to this theory, changes in hydrophobicity of a surface can be induced by the combined effects of surface topography and surface chemistry (Cassie and Baxter, 1944; Shirtcliffe, McHale, Atherton et al., 2010; Su, Ji, Zhang et al., 2010; Bhushan, 2012). The wettability of a surface containing a large proportion of entrapped air is described by the Cassie-Baxter equation:

1 1 Equation 6.1. where θ is the observed water contact angle, θsmooth is the water contact angle on an ideally smooth surface with the same chemical composition, and f1 is the area fraction occupied by the surface. Essentially, f1 and θsmooth from equation (6.1) are used to describe the role of surface topography and surface chemistry, respectively. As described earlier, the chemical composition of the two dragonfly wings were highly similar, therefore the θsmooth values of the two can be treated as the same. Additionally, f1 is interchangeable with the bearing area at the depth of impingement by the water droplet. Assuming the same θsmooth for both wings, the theoretical wettability of the two dragonfly wing was plotted using equation 6.1 and plotted on the same axes as the bearing functions in order to demonstrate their correlation (Figure 6.13). The blue line represents the H. papuensis wings, whilst the red line represents the D. bipunctata wings.

118

At a given depth, the roughness features affect the wettability of the surfaces to different degrees. This occurs because the geometry of the surface features will determine the quantity of air that will remain trapped in between the pillars (assuming similar chemical compositions), which in turn limit how deep the water droplets can penetrate beneath the tips of the features. For example, from Figure 6.13, at 100 nm depths from the top of the pillar, the bearing percentage for D. bipunctata wings is approximately 3.8 % whereas it is 2.0 % for H. papuensis wings. From theoretical wettability graph (dashed lines, Figure 6.13), the WCA appeared to be higher on the surface of H. papuensis wings compare to that of the D. bipunctata wings. This trend is in agreement with the measured WCA of H. papuensis and D. bipunctata wings (161.8° ± 1.9° vs. 151.7° ± 2.8°).

The higher WCA that is observed on the surface of the H. papuensis wings arises from the decreased ability of the water phase to penetrate the surface, which is a result of the subtle differences in nanotopography identified through bearing analysis and Fourier transform analysis. The H. papuensis wings possess a greater surface roughness 119 hierarchy, which enables them to better support water droplets and maintain a superhydrophobic and self-cleaning state but does not afford the surface significant bactericidal activity. According to previously published work pertaining to bacterial attachment to cicada wings, cells that sit on the top of nanopillars and do not penetrate the surface appears to undergo significantly less rupture (Pogodin, Hasan, Baulin et al., 2013b). Thus, the more resistant cell types can be considered analogous to water droplets in the Cassie-Baxter state; they do not penetrate between the surface features sufficiently to deform to the point of rupture.

Psaltoda claripennis (cicada) wings have been shown to possess a periodic topography, consisting of a regular hexagonal array of spherically capped, conical, nanoscale pillars (Ivanova, Hasan, Webb et al., 2012). Cicadae belong to the order Hemiptera, while dragonflies belong to the order Odonata. Despite their taxonomical distance, they display some similarities in both their cuticular composition and structure. Cicada surfaces contain three distinct layers, i.e., epicuticular layers on both the dorsal and ventral surfaces, and a procuticle in between these surfaces. Their epicuticular layers consist of an array of nanopillars covered by wax components (Tobin, Puskar, Hasan et al., 2013). Cicada wings exhibit a WCA of approximately of 158°, which is similar to the WCA of 161° on the surface of the H. papuensis wing. According to the proposed mechanism of bactericidal activity of wing surfaces, it would be expected that cicada wings behave similarly to the wings of H. papuensis dragonfly, and display a decreased extent of antibacterial activity relative to that of the D. bipunctata wings (Ivanova, Hasan, Webb et al., 2012). The bactericidal activity of cicada wings against P. aeruginosa cells is approximately 2.0 × 105 cells cm-2 min-1 (Ivanova, Hasan, Webb et al., 2012), which is greater than that measured for the H. papuensis wing surface (3.5 × 104 cells cm-2 min-1) and about 60 % of that of the D. bipunctata wing surface (3.0 × 105 cells cm-2 min-1). The three wings exhibited a trend of decreasing bactericidal activity with increasing surface hydrophobicity, which is consistent with the bactericidal mechanisms proposed in the current work. Importantly, the range of bacterial species against which cicada wings surfaces are effective matched those that were susceptible to H. papuensis wings; only the more flexible/deformable Gram-negative cells were killed in each case.

Insect wings, including those of cicada, damselflies, and dragonflies, are known for their superhydrophobic properties (Hasan, Webb, Truong et al., 2012b; Ivanova, Hasan, 120

Webb et al., 2012; Pogodin, Hasan, Baulin et al., 2013b). The habitats in which they exist are ever-changing and this has dictated their need to develop sustainable coping strategies in order to withstand a number of environmental conditions, including the presence of pollutants and rainfall. As a result, individual species have over time, developed strategies that have resulted in the emergence of a diverse range of natural surface structures (Lafuma and Quéré, 2003; Burton and Bhushan, 2005; Callies and Quéré, 2005; Sun, Feng, Gao et al., 2005; Papadopoulos, Mammen, Deng et al., 2013). The D. bipunctata dragonfly (common name ‘Wandering Percher’) can be found in Australia, New Zealand and numerous Pacific Islands, whilst H. papuensis (common name ‘Australian Emperor’) dragonfly is found only in Australia. The H. papuensis dragonfly is very large, with a wing length being up to 10 cm. D. bipunctata dragonflies, on the other hand, are much smaller in wingspan (their wings are approximately 4 cm in length) and usually fly only short distances between perches. Their intricate surface structures are well matched to their living conditions. The surface area of the H. papuensis wings was found to be higher than that of the D. bipunctata wings, however their bearing ratio was much lower. The higher surface area is a result of the presence of intricate hierarchical structures that enables the entrapment of greater quantities of air between the pillars. According to Nosonovsky and Bhushan, the presence of hierarchical surface roughness is largely responsible for the superhydrophobicity and also serves to ensure that the wings are aerodynamic (Nosonovsky and Bhushan, 2007a). The H. papuensis wings exhibited higher WCA with smaller deviations across the wing surface, affording them their superhydrophobic state. Being a more active insect, having a higher WCA may enable H. papuensis dragonflies to obtain maximum benefit from the self-cleaning/antibiofouling properties associated with superhydrophobicity. The surface of the D. bipunctata wings, on the other hand, exhibited both superhydrophobic and bactericidal properties, which suggests that these surfaces would be able to withstand a greater variety of environmental conditions. This would appear to be the case; D. bipunctata dragonflies are found over a greater geographical region than H. papuensis dragonflies, suggesting that the combination of surface properties has indeed enabled D. bipunctata dragonflies to adapt to a wider range of habitats.

121

6.6. Summary

The surfaces of H. papuensis and D. bipunctata dragonfly wings exhibited different activities when put in contact with bacterial cells. The H. papuensis wings showed moderate bactericidal activities against S. aureus and B. subtilis spores but were not effective against P. aeruginosa and vegetative B. subtilis cells. D. bipunctata wings, on the other hand, showed strong bactericidal effects against all four cell types. This observation led to the requirement for evaluation of, and discrimination between, the wing surface characteristics. The results obtained demonstrated that the precise surface morphologies and surface topographies play a significant role in mechanobactericidal surfaces. The mechanisms that cause these bactericidal effects can be considered analogous to the wetting processes of a water droplet, i.e. the more a bacterium is able to ‘wet’ the surface, the more likely it is to be ruptured. The H. papuensis wings have a higher surface area but lower bearing ratio at the same depth, which indicated a higher degree of structural hierarchy and the ability to entrap a greater quantity of air. Consequently, cells are better able to safely rest on top of the surface, resulting in decreased bactericidal action.

122

Chapter 7. Systematic bacterial patterning using self-assembled microtextured alkanes ______

123

7.1. Overview

The role that surface chemistry plays in determining the wettability and structure on the surface of dragonfly wings was discussed in Chapter 4. In Chapter 5, the major components of the epicuticular lipids present on the surfaces of dragonfly wings were identified. It was shown that when the epicuticular lipids are removed through solvent extraction, the underlying surface of dragonfly wings is relatively smooth. It was also shown in Chapter 6 that minor alterations in the surface topology can change the surface functionality of a surface, which can impart both the ability to self-clean and bactericidal properties to the surface. The concepts allow the development of artificial surfaces with analogues properties to dragonfly wings. In this chapter, model surfaces were produced by recrystallisation of certain dragonfly wing components on smooth artificial substrata. Two of the major components of dragonfly epicuticles identified in Chapter 5, eicosane and docosane, were selected for use in self-assembly on highly ordered pyrolytic graphite (HOPG) (Figure 7.1).

Figure 7.1. The process of making ordered microcrystalline interfaces

Each alkane was dissolved in chloroform at a concentration of 3 mg/mL and 10 µL volumes of this solution was deposited onto the surface of two substrata, silicon wafer controls and HOPG. The solvent was allowed to evaporate, which resulted in these compounds recrystallising into patterned microcrystalline interfaces on the HOPG surface and random clusters on the silicon wafers. Consequently, the ordered microcrystals on the HOPG were shown to direct the attachment of two species of bacteria; spherical-shaped Gram positive Staphylococcus aureus and rod-shaped Gram negative Pseudomonas aeruginosa bacteria. These microtextured surfaces exhibited the ability to limit bacterial adhesion to a precise pattern that followed the air/water/lipid three phase line (TPL).

124

7.2. Thermodynamics of alkane recrystallisation on HOPG and silicon

The HOPG and silicon substrata both appeared to be very smooth before any lipid material was deposited onto the surface, as was observed in 2D AFM scans (Figure 7.1). A roughness analysis confirmed this to be the case (Table 7.1). Detailed explanations of the roughness parameters found in Table 7.1 are given in Section 4.4.3. The average roughness (Ra) of both surfaces was below 1 nm. The skewness (Rsk) and kurtosis (Rku) parameters can appear relatively high for smooth surfaces due to background noise, however in the case of HOPG this was partially mitigated due to the detection of a sheet edge, which increased the maximum height range.

125

Table 7.1. AFM roughness analysis of the graphite and silicon wafer surfacesa

Roughness parameters Graphite Silicon wafer

Rmax (nm) 4.4 ± 1.2 8.0 ± 0.8

Ra (nm) 0.74 ± 0.30 0.13 ± 0.03

Rq (nm) 0.88 ± 0.33 0.22 ± 0.05

Rsk -0.2 6.6

Rku 2.5 112.8

Rdr (%) 100 ± 0 100 ± 0 a Results presented are averages of five measurements taken from 25 µm2 scanning areas and the errors are the standard deviations.

Silicon wafers and graphite are known for their homogeneously smooth surfaces, which were confirmed from the data presented in Figure 7.1 and Table 7.1, and have been used as the substrata for many molecular assembly studies, including lipid self- assembly (Koch, Neinhuis, Ensikat et al., 2004; Koch, Dommisse and Barthlott, 2006; Koch and Ensikat, 2008). The adsorption patterns of compounds on their surfaces depend on their molecular organisation and surface properties. Koch and colleagues proposed polarity as one factor that determined how organic molecules adsorb onto surfaces. The wetting behaviours of both surfaces have been well established in the literature; whilst silicon wafers are very hydrophilic with a WCA of approximately 43° (Hermansson, Lindberg, Hok et al., 1991), the graphite WCA can range from 90° to 95°, which is considered slightly hydrophobic (Taherian, Marcon, Van Der Vegt et al., 2013). These properties may be important for defining the patterns of self-assembled molecules.

Chloroform was the solvent selected for these experiments. Being polar and lipophilic, it dissolved eicosane and docosane well and caused the solution to spread more readily and evenly across the HOPG surface than the silicon wafer surface. Hence, the evaporation rate was higher on HOPG due to the larger liquid surface area, which in turn led to a more uniform lipid distribution on the HOPG than on the silicon (Figure 7.2)

126

7.3. Patterns of the self-assembled crystals

7.3.1. Crystal morphology

The crystallisation patterns formed by pure eicosane and docosane on HOPG are shown in Figure 7.3. The aliphatic hydrocarbons recrystallised into similar rodlet patterns. Both the eicosane and docosane rodlets were preferentially oriented at approximately 30° or 60° relative to each other (as indicated by the arrows). The crystal orientation of the two alkanes on the surface of HOPG suggested an epitaxial relationship existing between the substrata and the adsorbed alkanes. Koch and colleagues have extensively studied the self-assembly of plant waxes on various substrata (Koch, Neinhuis, Ensikat et al., 2004; Koch, Barthlott, Koch et al., 2006; 127

Koch, Dommisse and Barthlott, 2006). They also demonstrated an epitaxial relationship between octacosan-1-ol, the major component present in wheat epicuticle waxes, and HOPG (Koch, Barthlott, Koch et al., 2006). The appearance of three rotational domains parallel to the main symmetry axes of HOPG was presented as evidence for their conclusions; this appearance is also reflected in this work.

Molecular dynamics (MD) studies of alkanes that have crystallised on the surface of HOPG substrata at the molecular level were reported by Yin et al (Yin, Wang, Qiu et al., 2001), where it was shown that a change in chain length of a C2H4 unit leads to a change in the respective interaction energies. It was found that as the alkane chain length increases, the adsorption energy per mole and the corresponding crystallisation energy (which is 40 – 60% less than the adsorption energy) increases. In addition, increasing the misalignment between the sp3 C–C bond length of the adsorbing alkane and the sp2 graphitic bond length resulted in a greater energy requirement for microcrystallite formation, which is consistent with the observed differences between eicosane and docosane (Figures 7.2 A, B). The difference in the microcrystallite dimensions between C20 and C22 alkanes appears to reflect the accommodation of the increased mismatch and concomitant strain with chain length. 128

There are two possible orientations of alkane molecules adsorbed on graphitic layers, which are referred to as either ‘zigzag’ or ‘armchair’ orientations (Figure 7.4). The zigzag orientation has been reported to be more energetically favourable than the armchair orientation (Yang, Berber, Liu et al., 2008). Both of these orientations lead to the possibility of three different molecular alignment directions, at angles of 60° relative to each other, hence the equilateral triangles formed by the alkane crystals (Figure 7.3). The zigzag and armchair orientations are rotationally opposed by an angle of 30°, giving rise to a total of six possible directions being adopted by the aligned alkane molecules (Figure 7.4).

Figure 7.5. Schematic representation of alkane alignment when they are adsorbed on graphite in both zigzag and armchair orientations. Two alkane molecules in the same orientation will align an angle of 60°, while molecules in the opposite orientations will be aligned at 30° or 90°.

A 2D Fast Fourier Transform (FFT) analysis (Burger and Burge, 2007) was applied to the electron micrographs obtained for the self-assembled eicosane and docosane in order to determine their dominant alignment directions. Over the areas of analysis, not all of the predicted directions were detected; mostly the crystallites were directed 60° from each other (Figure 7.5). In particular, the FFT of the docosane crystal structure revealed consistent alignment in three major directions, all of which were separated by 60°, indicating a preference for either the ‘zigzag’ or ‘armchair’ orientation (Figure 7.4).

129

Given that the ‘zigzag’ orientation is more energetically favourable at the molecular level, this is likely the dominant alignment pattern.

7.3.2. Eicosane and docosane crystal topographies

Analysis of the AFM cross-sectional profiles of the two recrystallised alkanes enabled the size and thickness of the rodlets, together with the overall roughness of the interfaces formed, to be determined (Figure 7.6, Table 7.2). Roughness statistics were calculated based on the alkane crystal surface topography over 50 µm × 50 µm AFM scanning areas (Table 7.2). The maximum height of eicosane crystals was 1.87 µm, larger than that of the docosane crystals, which was 1.50 µm. Docosane crystals were found to be generally longer and narrower than the eicosane crystals (Table 7.2).

130

Table 7.2. AFM roughness analysis of the eicosane and docosane crystals on HOPG surfacesa

Roughness parameters Eicosane Docosane

Rmax (µm) 1.9 ± 0.2 1.5 ± 0.2

Ra (µm) 0.3 ± 0.1 0.2 ± 0.0

Rq (µm) 0.3 ± 0.1 0.2 ± 0.1

Rdr (%) 110 ± 1 103 ± 5 Crystal dimensionsb Average length (µm) 3.7 4.3 Average crystal aspect ratio 2.1 2.7 a Results presented are averages of five measurements taken from 50 µm × 50 µm scanning areas and errors are the standard deviations. bCrystal less than 200 nm in width were excluded to minimise the effect of noise. 131

The high similarity between the two alkane molecules produces their similar patterns of crystallisation. When considering the crystallisation of alkane molecules on

HOPG at the molecular level, a difference of a single C2H4 unit can lead to an energy gain of approximate 0.09 to 0.12 eV (Yang, Berber, Liu et al., 2008) due to their misalignment with the graphite molecular structure. Longer hydrocarbon chain lengths would result in a larger longitudinal mismatch of the alkane molecule relative to the underlying graphite and hence a greater free energy of adsorption would result. This provides a possible explanation for the minor differences in the adsorption patterns for eicosane and docosane. Relative to eicosane, it was more energetically favourable for docosane molecules to associate with each other rather than the HOPG substratum. Any given docosane crystal would therefore be anchored by fewer template molecules than would eicosane crystals and thus the crystals formed would be narrower than those of the docosane.

7.4. Bacterial attachment to HOPG template-mediated self-organised alkane microcrystallites

Confocal laser scanning microscopy was employed to assess the influence of these interfacial topologies on cell attachment, using: the spherical, Gram-positive bacterium Staphylococcus aureus, and the rod-shaped, Gram-negative Pseudomonas aeruginosa as the model. The S. aureus cells attached and aligned well along the microcrystallite three phase contact line (TPL), whereas the rod-shaped P. aeruginosa cells, whilst being found to attach, were only partially aligned along the microcrystallite TPL (Figure 7.7). The aligned bacterial cells, particularly S. aureus cells, are clearly visible to follow the dominant 60° angle of the microcrystallites. Importantly, the bacterial cells resisted attachment to the underlying substratum in the interstices between the alkane microcrystallites.

132

Both species were found to attach in high numbers to the unmodified HOPG control surface (Figure 7.8) but in a random manner where no patterning was observed. It appeared that the introduction of the alkane microcrystallite topology to the surface limited the extent of adhesion of the bacteria to the substratum surface overall and those

133 bacteria that did attach did so in a highly structured pattern, reflecting the orientations of the microcrystallites. In addition, the topology prevented the attachment of bacteria to the underlying substratum between the lipid microcrystallites.

Figure 7.9. Attachment of S. aureus and P. aeruginosa cells on the HOPG surface. Bacterial cells attached randomly on the surface without any apparent pattern.

Further analysis of the bacterial cell attachment patterns (Figure 7.7) indicated that there were much fewer S. aureus cells attaching to the docosane surface than the eicosane surface and the directionally of the attached cells along the rodlets underneath was more clearly observed. As the width of the eicosane rodlets is greater, the S. aureus cells could attach to a greater extent. P. aeruginosa cells appeared to align to a greater extent on eicosane and only partially aligned on the docosane crystals. This pattern is likely to be due to the differences in the dimensions of the alkane rodlets and bacterial cells. Whist S. aureus can align well with docosane, P. aeruginosa can only partially align because of the restrictive geometry of the rodlets.

134

7.4.1. Local hydrophobicity for bacterial attachment and alignment

The specificity of bacterial attachment that results from the surface patterning on the substratum clearly arises from the structure of the interface in contact with the bacterial populations. As noted, considerable work has been carried out regarding the influence that various surface structures play on the extent of microbial attachment and subsequent biofouling, much of it based on the impact of surface superhydrophobicity. Eicosane and docosane, as lipid components, display water contact angles of approximately 100° on planar surfaces (Fox and Zisman, 1952), whereas the microcrystallite eicosane and docosane structures on these surfaces exhibited relatively high water contact angles (134° and 124° respectively). Whilst these contact angles are lower than the measured WCA of the superhydrophobic dragonfly wings (i.e. >150°) on which these alkane are present (Chapter 4 and 5), these angles are substantially higher than the corresponding WCA of planar lipid films and the underlying HOPG (ca. 90°) substratum surface. The large WCAs measured on eicosane and docosane microcrystalline surfaces are directly attributable to the ability of these topologies to entrap air, hence forming composite interfaces following the Cassie-Baxter wettability regime (Bhushan, Jung and Koch, 2009a; Papadopoulos, Mammen, Deng et al., 2013; Zhang, Wang and Levänen, 2013). Raman microspectroscopy confirmed the presence of air pockets within the interstices of the surfaces in contact with water (Figure 7.9A). The images were produced by colouring each pixel according to the intensity of the band corresponding to the O-H bonds of water molecules; hence areas of low intensity (black areas) indicated the absence of water in areas that were up to approximately 4 µm in diameter.

135

The behaviour of micro- and nano-particles at fluid-fluid interfaces are increasingly recognised as important from both a theoretical and technological viewpoint (Stamou, Duschl and Johannsmann, 2000; Fournier and Galatola, 2002). In this work, the understanding of this behaviour has been extended to include bacterial attachment at the three phase contact line of lipid microstructures. Figure 7.9B provides a schematic representation of spherical S. aureus cells attaching to the microcrystallite-water-air interface. To a first approximation (neglecting the extracellular matrix), the force of attraction of the bacteria to the alkane surface at equilibrium is dominated by van der Waals interactions over electrostatic and hydrodynamic forces, which results in an effective radial adhesion force Fadh. The water contact angle θ and the resultant capillary 136 force Fcap establishes the depth of penetration of the cell within the water-air interface.

The net force between Fadh and Fcap provides the air penetration angle ψ and illustrates the interplay between bacterial hydrophilicity/hydrophobicity and lipid adhesion. Loudet et al. (Loudet, Alsayed, Zhang et al., 2005) extended these studies to anisotropic (prolate ellipsoid) micron particles at fluid interfaces, where it was shown that interaction energies were considerably larger than their spherical counterparts. The data presented in Figure 7.7 indicated this was the case for the rod-shaped P. aeruginosa at the TPL, where alignment of the major axis of the bacteria with that of the microcrystallite was the dominant force determining the attachment pattern. The hydrophilicity of the bacteria, as shown by the water contact angles 72o and 45o for S. aureus and P. aeruginosa respectively, indicates their ability to penetrate into the air pocket, which is also consistent with the respective bacterial surface energies of 42.6 mJ/m2 and 48.7 mJ/m2 (Truong, Lapovok, Estrin et al., 2010). The relative bacterial surface contact angles also influence the degree to which cell attachment to eicosane and docosane microcrystallites will take place. The data presented in Figure 7.10A provides both the relative attachment numbers, as indicated by CSLM cell counting, together with the water contact angles of these microcrystallite surfaces (Figure 7.10B). Here it can be seen that the more hydrophobic S. aureus cells appeared to preferably attach on the more hydrophobic alkane (134o) whereas the less hydrophobic P. aeruginosa tended to attach more on the less hydrophobic surface (124o).

137

Figure 7.11. (A) Density of attached bacterial cells as inferred from confocal analysis and (B) Surface wettability of eicosane, docosane, S. aureus and P. aeruginosa. The S. aureus cells appear to more readily attach onto the eicosane, whereas the P. aeruginosa cells more readily attach onto the docosane.

138

Recently, it was reported that geometric (e.g. high-aspect ratio, nano-post arrays) and mechanical parameters were responsible for the spontaneous and distinct ordering of bacteria, which led to periodic orientations of the bacterial cells on the modified surfaces. Hochbaum and Aizenberg first proposed the directing of bacterial cells, through the mechanoselective adhesion of cells to nanoscale posts (Hochbaum and Aizenberg, 2010). Nanometre-scale periodic surface features were proposed as a possibility for controlling bacterial attachment at the single cell level (Epstein, Hochbaum, Kim et al., 2011). The alkane microcrystallite interfaces produced in this work were found to direct bacterial attachment but by an apparently different mechanism. When these model lipid interfaces were immersed in a bacterial suspension, the entrapment of air in the microcrystallite interstices caused the surface to undergo a transition to a Cassie-Baxter regime. This caused the bacterial cells to spontaneously respond, prompting them to preferentially align (Figure 7.9B), and importantly, attachment to the substratum between lipid interstices was prevented. A similar phenomenon has also been observed on the surface of titanium that was modified to mimic the surface structure of the lotus leaf (Truong, Webb, Fadeeva et al., 2012). Brennan and co-workers engineered particular surface microtopologies and developed a correlation between a roughness index based on Wenzel and Cassie-Baxter wetting parameters and the energetics of cell attachment using a statistical mechanical approach, which was shown to be consistent for a wide range of organisms modelled on a planar surface (Decker, Kirschner, Long et al., 2013).

7.5. Summary

A simple but effective fabrication technique, the self-assembly of aliphatic alkanes on the surface of HOPG, was employed to produce surfaces capable of directing bacterial attachment. The relationship between surface microstructure and the extent of bacterial adhesion that takes place represents a growing focus of interfacial research due to its broad impact, which ranges from biofilm formation to the binding of invasive pathogens. Both the composition and topology of substratum surfaces have been studied with a view to inhibiting the extent to which bacterial adhesion takes place. In the most part, this has been achieved through efforts to understand and replicate antifouling surfaces. In particular, natural surfaces, such as some insect wings with high aspect ratios, have been shown to provide self-cleaning, superhydrophobic and antibacterial functionalities. The microenvironment resulting from the spatial confinement of bacteria 139 significantly influences the subsequent attachment behaviour. In this chapter it was shown that air entrapment is the underlying mechanism that dictates the interaction and structuring of bacterial cells on microtextured C20 and C22 surfaces, prepared by exploiting the templating ability of highly ordered graphite. The introduction of alkane microcrystallite topologies to the substratum surface resulted in: (a) the limitation of the extent of adhesion of the bacteria overall, (b) predictable patterning of the bacteria that did adhere, and (c) the prevention of bacterial attachment to the underlying substratum. Both bacteria investigated in this study elegantly adopted an attachment pattern as they adhered at the TPL produced by the alkane microcrystallites, entrapped air and the surrounding aqueous medium. The relative hydrophobicity of the bacteria influenced their immersion into the air pockets on the surface, increasing the density of cell attachment to the respective model lipids. These microtextured surfaces were able to limit bacterial adhesion to areas of precise patterning at the lipid TPL. Surfaces such as these may have potential application as surfaces that can control or direct bacterial attachment to surfaces.

140

Chapter 8. Synthetic mechanobactericidal surfaces composed of dragonfly fatty acids ______

141

8.1. Overview

Aliphatic alkanes, specifically eicosane and docosane, were shown in the previous chapter to be capable of forming self-organised nanostructures on substrate surfaces that were able to control bacterial attachment via entrapment of air within the interstices on the surface. These surface structures also influenced the attachment direction of these cells according to the position of the three-phase line of contact with the surface. The alkane structures were, however, largely inert to the cells, with almost all of the attached cells being metabolically active. In addition to aliphatic alkanes, fatty acids were found to be major components of outer layer of the epicuticle. In this chapter, the two fatty acids, palmitic acid (hexadecanoic acid, C16H32O2) and stearic acid (octadecanoic acid,

C18H36O2) were self-assembled on the surface of HOPG to form microcrystallite structures that were subsequently evaluated for their biological activity.

8.2. Crystal patterns of carboxylic acids on graphite

As in the case with aliphatic alkanes, there was an epitaxial relationship developed between the carboxylic acids and the HOPG substratum upon attachment to the surface, however the resulting crystal patterns appeared very different from those of the alkanes. The images given in Figure 8.1 show high resolution scanning electron micrographs of the crystal structures of palmitic acid and stearic acid on HOPG substrates and their corresponding 3D simulated images, generated using Avizo® software. Palmitic acid recrystallised to form blade-type structures, whilst stearic acid formed broader, shorter structures that were less sharp in appearance.

142

Cross-sectional profiles generated from AFM scans of the two recrystallised fatty acids enabled an estimation of the size and thickness of the microblades (Figure 8.2). The palmitic acids microblades were typically 5 µm in length, whilst the stearic acid crystallites were approximately 1 - 2 µm in length. The line profiles of the fatty acids also confirmed that the palmitic acid microblades were sharper, and by comparison, the stearic acid structures tended to be flat across the peaks. In general, the palmitic acid surfaces consisted of features having high and sharp peaks with small valleys whilst stearic acid surfaces contain flattened high peaks with broader valleys in between.

143

A roughness analysis of 10 µm × 10 µm AFM scanning areas revealed that the stearic acid crystallites were generally taller than those of palmitic acid (Table 8.1). The maximum height (Rmax) of the palmitic acid crystallites was 430 nm, larger than that of the stearic acid crystals (340 nm), however the average and RMS roughness values (Ra and Rq, respectively) both appeared to be higher on the surface of stearic acid compared to that of palmitic acid. Statistically speaking, Ra and Rq are more representative of the surface as they describe the spread of values about the mean, whereas Rmax is derived from extreme surface roughness values. Kurtosis, Rku, provides an indication of the sharpness of the topographical features. As expected, the stearic acid surfaces possessed a lower Rku, which corresponds to a squarer feature shape, which is reflected by the flat- 144 topped features evident in the 2D line profiles. The skewness (Rsk) and surface areas

(Rdr) were very similar between the two surfaces.

Table 8.1. AFM roughness analysis of the eicosane and docosane crystals on HOPG surfacesa

Roughness statistics Palmitic acid Stearic acid

Rmax (nm) 431.4 ± 31.4 341.6 ± 23.7

Ra (nm) 37.8 ± 3.2 47.1 ± 3.1

Rq (nm) 48.94 ± 4.3 57.5 ± 4.1

Rsk 0.59 0.80

Rku 1.43 0.44

Rdr (%) 108.3 ± 1.8 107.6 ± 0.4 a Results presented are averages of five measurements taken from 10 µm × 10 µm scanning areas and errors are the standard deviations.

Additional topographical analyses demonstrated that the crystallite structures of palmitic acid were generally shorter, whilst stearic acid crystallites varied to a greater extent in length and width (Figure 8.3). The size distributions of the palmitic acid and stearic acid microcrystal populations spanned approximately 200 – 500 nm and 200 – 1700 nm in width, respectively. In terms of area coverage, the palmitic acid crystallites were typically below ~1000 nm or above ~1500 nm in width, with few crystallites being found between this range. The area coverage size distribution of the stearic acid microcrystals ranged from ~1000 nm to ~1700 nm. The results obtained in this study demonstrate that despite being very similar in their molecular formula, the packing arrangement of palmitic and stearic acids are very different, and hence the resulting surface topographies and morphologies are also quite different.

145

8.3. Thermodynamics of crystallisation

8.3.1. Directional crystals formed by the two fatty acids

A Fast Fourier Transform (FFT) analysis was performed on the data, and the results showed that stearic acid appeared to crystallise into random directions, however palmitic acid crystallised into more defined orientations (Figure 8.4, white ovals). The two main orientation directions were shifted on an angle of approximately 30° (or 150°) to each other, different to that observed for alkanes presented in Chapter 7, where there were six possible directional orientations. This indicated that the carboxylic acid functionalities introduced new localised adsorption states, which resulted in the formation of different crystallite patterns.

146

There are seven possibilities for fatty acid molecules to align on graphite based on the orientation of the functional group, either head to head or head to tail within a domain (Figure 8.5) (Yang, Berber, Liu et al., 2008). This modulation is based on monolayer calculations and the orientation of each molecule in a packed system, such as these, needs to be further investigated. In transforming from a 2D surface layer system into a 3D crystal system, the fatty acids undergo phase transitions. It is well established that these phases include: triclinic (Form A), monoclinic (Form C), and orthorhombic polytypes (forms B and E) (Kaneko, Kobayashi, Kitagawa et al., 1992; Moreno, Cordobilla, Calvet et al., 2007). Therefore, it may be expected that there were multimorphic crystal states within the same self-assembled fatty acids.

147

In self-assembled monolayers, depending on the temperature and humidity, fatty acids can go through various states of crystallinity (Kaneko, Kobayashi, Kitagawa et al., 1992; Bond, 2004; Gbabode, Negrier, Mondieig et al., 2006). As for a packed system, it is not clear which states are present; however it is expected that the two fatty acids investigated here differ in their crystal phases, which could account for their differences in morphology. Synchrotron radiation Fourier Transform Infrared Spectroscopy (FTIR) in soft-contact attenuated total reflection (ATR) mode was employed to gain an insight into their crystallisation structures. This method enables study of the chemical configuration of the crystallites at a very shallow depth, which is dependent on the refractive index of the ATR crystal used (in this case ≤ 1 µm).

Two-dimensional spectral maps of the fatty acid interfaces revealed spatial variations in the intensities of the symmetric and antisymmetric CH2 peaks and the carbonyl (C=O) peaks, indicating the heterogeneity of their distribution across the surface of the crystals (Figure 8.6). Additionally, while the expected carbonyl peak was observed in stearic acid, it was not present in palmitic acid spectra, suggesting a

148 different organisation at the molecular level. Given that FTIR in ATR mode allows the analysis of only the very top layers of the sample, the differences may be due to the carboxyl groups of the palmitic acid molecules being oriented inward toward the substratum.

A subtle peak shift was observed in the C-H stretching band of some of the spectra recorded for the stearic acid surface (Figure 8.7). This peak shift may be due to minor variation in the vibration energy of C-H bonds of stearic acid molecules in one crystal phase compared to another. A 2D map of the relative peak shift of the C-H stretching band demonstrated heterogeneous spatial distribution of the point of maximum intensity (Figure 8.7D). The highly sensitive scale on the right of the map indicated the relative magnitude of the shift. High colour scale intensity corresponded to spectra in which the

149 peak is located at slightly higher wavenumbers, whereas lower wavenumber peak positions are identified by low intensity of the colour scale. The high intensity spots somewhat reflect the crystallised structure visible on the SEM. This peak shift mapping is evidence to the presence of multimorphic crystal states within the same self- assembled fatty acids.

8.4. Biological activities

The biological activities of the two fatty acids crystallite structures were evaluated using Pseudomonas aeruginosa and Staphylococcus aureus bacteria. Both surfaces appeared to have bactericidal activities against both types of two bacterial cells; however palmitic acid produced stronger killing in both cases compared to stearic acid. At least 17 individual microscopic scanning views for each sample were used to estimate their bactericidal activities. As a result, the surface of palmitic acid showed strong deactivating effects on P. aeruginosa and S. aureus, with 71% death of the 150 former and 95% of the latter killed on contact. Stearic acid also exhibited moderately strong bactericidal behaviour against P. aeruginosa (63%), while being much lower against S. aureus, with 49% cells killed by the surface. Importantly both of the cells types appeared alive on HOPG control surfaces (Section 7.4). This confirmed that both of the surfaces were successful in inactivating the two bacterial cells types being investigated. The efficiency of the two surfaces at killing bacteria is illustrated graphically in Figure 8.8 and 8.9.

151

The palmitic acid surface unexpectedly showed greater activity against S. aureus cells than against P. aeruginosa (Figure 8.8), since the Gram-positive cell, S. aureus is usually considered to be the more mechanically robust of the two bacteria. Whereas previous works have indicated that the decreased rigidity of the cell wall was the Achilles heel of bacterial cells when coming into contact with mechanobactericidal surfaces, in this case it appeared that the bacteria with the softer cell wall was better able to cope with the bactericidal action of the fatty acid surfaces. It has been shown that the adsorption of P. aeruginosa cells on the mechanobactericidal nanopillar arrays of cicada wings did not pierce the cell walls, but rather stretched them to the point of rupture in the regions located between the nanopillars (Pogodin, Hasan, Baulin et al.,

152

2013a). In this case of these fatty acid microstructures, the softer P. aeruginosa cells may be able to sustain a greater extent of deformation before reaching the point of rupture. P. aeruginosa cells are also larger, which means they can share physical stress over a greater membrane area.

Both palmitic and stearic fatty acids are naturally-occurring metabolites of most organisms (including bacteria), suggesting that the observed bactericidal activity is probably not due to chemical functionality and therefore almost certainly mechanical in nature. In order to confirm this, the surfaces of these fatty acids were chemically modified with a thin layer of gold, which has a minimal effect on the surface structure.

8.5. Gold-coating of the fatty acids crystals

The surfaces of the two fatty acids were coated with a layer of gold with a thickness of approximately 3 nm. Similar approaches were employed to change surface chemistry of cicada wings (Ivanova, Hasan, Webb et al., 2012). The modified surfaces were then tested against the same two types of bacterial cells (Figure 8.10). The large number of red signals in the CLSM images indicated that the gold-coated surfaces also exhibited strong bactericidal activities, in fact both the numbers of cells attached to the surface and the proportion of dead cells were increased (Figures 8.11, 8.12). Despite the appreciable change in the surface chemistry of the two surfaces, the bactericidal effect of the gold-coated surfaces was preserved (Figure 8.10). This demonstrated that the physical surface structure of fatty acids, rather than their biochemistry, is primarily responsible for the lethality of crystallised structure for P. aeruginosa and S. aureus cells.

153

154

These results are similar to those of previous work that was published on the bactericidal activity of insect wings, i.e. cicada and dragonfly wings, where the mechanisms responsible for the bactericidal activity were found to be purely physical in nature. The surface structures of the wings of those insects were also found to retain their bactericidal activity when coated with thin layers of gold (Hasan, Webb, Truong et al., 2012a; Ivanova, Hasan, Webb et al., 2012; Pogodin, Hasan, Baulin et al., 2013a). To explain for this phenomenon, it was previously reported in literature that the metal layer indeed can enhance the fluorescence’s signals (Lee, Hahn, Yuen et al., 2011; Li, Busscher, van der Mei et al., 2013). On a metal surface, the fluorescence signal bound to bacterial cells is enhanced significantly, hence assisting in better quantification of bacterial adhesion. Even though the crystal structures of the two fatty acids did not resemble that of either cicada or dragonfly wings, their mechanisms appeared to be very similar.

8.6. Spatial wettability

It may be intuitively expected that the sharp-edged microblades would ‘cut’ the bacterial cells open. From the images presented in Figure 8.2 it can be seen that there are many spaces between the microblades, sufficient in size to safely accommodate a S. aureus cell (~1 µm in diameter). In this case it would seem that 95% killing efficiency exceeds reasonable expectations (Figure 8.10). This suggests that some phenomenon takes place that causes the cells to interact with the microblade edges to a greater degree than the substrate surface itself. Entrapment of air bubbles on a micro-/nanostructured surface has previously been shown to direct cell attachment, both in previously published literature and in Chapters 6 and 7. The accessibility of such surfaces to bacterial cells can be limited by the amount of air trapped in between the surface features (Carman, Estes, Feinberg et al., 2006; Scardino, Zhang, Cookson et al., 2009; Ivanova, Truong, Wang et al., 2010; Truong, Webb, Fadeeva et al., 2012; Hsu, Fang, Borca-Tasciuc et al., 2013; Webb, Boshkovikj, Fluke et al., 2013). From the information presented in Chapter 6, shown that the surface of two different species of dragonfly wings performed differently in terms of bacterial attachment through just a minor alteration of their surface morphologies and subsequently the amount of air that could be retained by the surface. These differences ultimately influenced the wettability and the biocidal activities of these surfaces. The crystallised structures of the two fatty acids in this study appeared very different in their surface topologies, but how their 156 differences affected the surface wettability and their bactericidal activities has not yet been addressed.

The two fatty acids were both moderately hydrophobic prior to gold coating, partially due to their long aliphatic hydrocarbon chain. Coating with the gold led to a different chemical composition of the surface, however the underlying structures were not affected. The WCA of the gold-coated palmitic acid surfaces increased by ~7° compared to the uncoated surface, however the WCA of stearic acid surfaces decreased by approximately 10° upon gold coating (Figure 8.13). All four surfaces, however, remained moderated hydrophobic, i.e. WCA > 90°. This indicated that the surface microstructure structure did significantly affect the WCA. This is also in agreement with their surface topographical analyses (Table 8.1) in that the surface of stearic acid was rougher than that containing palmitic acid and therefore contained a greater amount of space between features in which air may become trapped.

Figure 8.13. Water contact angles on native fatty acids and after coating with a thin layer of gold.

157

In Chapter 7 it was shown that air pockets trapped between the microscale surface features prevented bacterial cells from accessing the underlying substratum in these positions, and instead forced them to interact with the alkane microcrystallites and align at the three-phase contact line. A similar effect may explain the observations of bacterial behaviour on these fatty acid microcrystal interfaces as well, and particularly with the similarity of the systems. A wettability analysis in the form of water contact angle (WCA) measurements (Figure 8.13) and Raman microspectroscopy (Figures 8.14, 8.15) support this hypothesis. Both microcrystal interfaces displayed moderate to high hydrophobicity, with WCAs above 100°. These high contact angles are due to Cassie- Baxter wetting states, in that entrapped air forms a composite interface that is highly hydrophobic. Raman microspectroscopy was employed to image the crystals and air pockets submerged under water. The entrapped air was visualised by recording a grid of Raman spectra over 10 µm × 10 µm scanning areas, and assigning each position in the grid values corresponding to the integrated areas of the C–H str and O–H str regions, corresponding to alkyl hydrocarbon and water respectively (Figure 8.15). Positions that were relatively low in both cases (i.e. spaces not filled by either fatty acids or water) were considered to be indicative of entrapped air. By comparison, the palmitic acid microblades appeared to contain a lower proportion of localised air, but with a more consistent coverage.

158

Figure 8.14. Raman spectra of palmitic acid and stearic acid crystallites immersed under water for air imaging.

Figure 8.15. Image scans of fatty acid (FA) crystallites and water using Raman microspectroscopy. The regions, which are darker in both images, indicated the presence of air pockets.

The wetting phenomena at the three-phase line (TPL) shown in Chapter 7 directed the positioning of adhering cells, thus it can be inferred that an imbalance of the adhesion forces on a cell nearing the TPL results in a net force toward the TPL. If a cell were nearing two different TPLs, such as on either side of the narrow ridged palmitic acid microblades, the resulting net force would cause the cell membrane to experience a high stress distribution at the sharp edges of the palmitic acids peaks due to the very high radii of curvature upon wetting and adhesion contact (Figure 8.16). The opposing components of the applied forces may also aid in cell rupture.

159

8.7. Summary

Microcrystals of palmitic and stearic acids assembled on the surface of HOPG substrata display bactericidal activity against both Gram-negative, rod-shaped P. aeruginosa, and Gram-positive, spherical S. aureus. The HOPG substratum acts as a template, guiding the alignment of the initially adsorbed fatty acid molecules, ultimately resulting in an interface with defined structures. The microblades formed by palmitic acid were particularly effective at killing bacterial cells, up to 95% in the case of S. aureus. Much like the mechanobactericidal dragonfly wings on which they are based, the mechanisms of action appears to be closely related to their special wetting properties. Trapped air pockets force the cells to interact with the sharp edges of the

160 microblades, while the wetting action of the three phase line applies sufficient force to the cell, resulting in their rupture.

161

Chapter 9. General Discussion ______

162

9.1. Overview

Biofilm formation is an essential step for bacteria to successfully infect a host surface (Garrett, Bhakoo and Zhang, 2008). For bacteria, there are numerous advantages to biofilm development, including antibiotic and disinfectant resistance, and free sharing of nutrients. Once bacteria form a biofilm, it is very hard to remove (Liu and Tay, 2002; Di Iaconi, Ramadori, Lopez et al., 2005; Ryu and Beuchat, 2005; Simões, Simões and Vieira, 2009; Schmidt, Winter and Gallert, 2012), which is problematic in many cases. There have been some great advances made in the battle against biofilm formation, however there are still great challenges going forward, such as antibiotic resistance and the emergence of so-called ‘superbugs’ arising from the heavy usage of chemical-based methods for controlling bacteria. For this reason, new approaches need to be explored in order to efficiently control the bacterial colonisation of surfaces.

Over a decade ago, when Barthlott and Neinhuis introduced the concept of “self- cleaning”, as observed on the surface of lotus leaves, the “lotus effect” became a widespread ideal symbol for a natural surface that can exhibit superhydrophobic and self-cleaning properties (Barthlott and Neinhuis, 1997). Great efforts have been made since then to better understand the mechanisms responsible for such properties and to further explore the variety of aspects of these surfaces in nature. To date, insects and plant leaves are the two surfaces reported that truly exhibit superhydrophobicity and self-cleaning properties. The discovery of these surfaces have opened the way for the development of new alternatives for controlling bacterial colonisation, some of which have already achieved some success, examples of which include anti-wetting glass and self-cleaning paints.

Very recently it was reported that the surfaces of cicada and dragonfly wings efficiently kill various types of bacterial cells, due to purely physical means (Hasan, Webb, Truong et al., 2012a; Ivanova, Hasan, Webb et al., 2012; Ivanova, Hasan, Webb et al., 2013; Pogodin, Hasan, Baulin et al., 2013a). This finding was the first step towards developing novel chemical-free and environmentally friendly alternatives to controlling bacterial contamination, however there is a need for more thorough understanding of the mechanisms responsible for these phenomena.

Here, the molecular organisation of the surfaces of insect wings were characterised and it was determined that the spatial arrangement of the surface hydrophobic

163 compounds contributed to their superhydrophobic nature. These hydrophobic components are also responsible for forming the sophisticated surface structure on the surface of the wings, which in turn further boosted their surface hydrophobicity. Investigation of the mechanisms responsible for the bactericidal properties of the wings, it was found that (as is the case with superhydrophobic and self-cleaning properties), their antibacterial properties are also closely dependent on the surface wettability. The amount of air entrapped between the nanostructures of the surface determines the way in which bacterial cells sit on the surface. With precisely defined topographical dimensions, bacterial cells are ruptured, just as observed in the early discoveries of the mechanobactericidal activity of cicada wings. Based on the understanding gained regarding the significance of entrapped air to the performance of the surfaces, artificial surfaces that effectively controlled bacterial attachment were produced. As an outcome of this, it was found that surface wettability was important in determining the adhesion ability of bacteria. Bacterial cells aligned along the surface structure of self-assembled alkanes, due to the existence of limitation to the accessible surface area arising from the presence of entrapped air pockets. It was found that palmitic and stearic fatty acids self- assembled by similar means, but formed different structures, i.e. micro-blades, which were lethal to bacteria. The mechanisms of this phenomenon were found to be mechanobactericidal in nature, and similar to those of cicada and dragonfly wings.

9.2. Molecular organisation of the surface of insect wings

The natural diversity of the wings of Odonata species was determined to be relatively narrow. Synchrotron FTIR and XPS analyses revealed that the major chemical components of the dragonfly wings were qualitatively similar; all samples contained protein/chitin, as well as waxy compounds. These waxy components are important in determining the wettability of these natural surfaces (Chapter 4). It was also shown that the nanoscale structures present on the surface of the wings further enhanced the surface hydrophobicity and are composed entirely of these hydrophobic compounds (Chapters 4 and 5). It has been well documented that surface morphologies contribute to the extent of surface wettability (Shirtcliffe, McHale, Atherton et al., 2010; Papadopoulos, Mammen, Deng et al., 2013; Stark, Badge, Wucinich et al., 2013). Nosonovsky and Bhushan determined that surface hierarchy is a key requirement for optimally stable superhydrophobic states (Nosonovsky and Bhushan, 2007a; Nosonovsky and Bhushan, 2007b). 164

Characterisation of several different species of Odonata revealed that the surface chemistry not only contributes directly to the surface wettability but also to the surface topography (Chapter 5). For this reason, it is important to identify the contributing components that make up these sophisticated structures of these insect wings. Over 60 different compounds were identified to constitute the nanoscale structures on the surface of the wings of the Hemianax papuensis and Hemicordulia tau dragonflies. The major components on the wing surface were found to be fatty acids, predominantly palmitic acid and stearic acid, and n-alkanes with even numbered carbon chains ranging from C14 to C30, in which eicosane and docosane were found to be the major components. Aliphatic compounds were detected throughout the epicuticle, and these appeared to form a third discrete layer, the mesoepicuticle, that was separate from both the inner and outer epicuticles (Figure 9.1).

165

Figure 9.1. Proposed model of the epicuticle of dragonfly wing membranes. Three layers are contained within the epicuticle: the outer-epicuticle, the meso-epicuticle, and the inner-epicuticle.

9.3. Assessment of the antibacterial potential of superhydrophobic insect wing surfaces

Comparison of the antibacterial potential of dragonfly wings suggests some striking observations regarding the ability of nature to drive multiple successful evolutions for 166 coping with similar problems. In terms of biological activity, the wings of H. papuensis dragonfly were not particularly effective at killing bacterial cells, in fact the efficiency was rather poor compared to the bactericidal activity originally described for the, D. bipunctata dragonfly. This should not be considered to be a failure of the H. papuensis dragonfly wings to control bacterial colonisation; rather it is a variation on the same method that dragonflies have adapted to fit better with their living habitats. The H. papuensis wings were generally colonised by fewer bacterial cells, suggesting that their means of coping with bacterial contamination was more reliant on the process of self- cleaning the wing surface.

D. bipunctata dragonflies can be found in Australia, New Zealand and numerous Pacific Islands whilst H. papuensis dragonflies are primarily inhabitants of Australia (Figure 9.2). The common name of H. papuensis dragonfly is the Australian Emperor, and the D. bipunctata dragonfly is also known as the Wandering Percher, and these monikers accurately describe their behavioural characteristics as individuals in their territories. H. papuensis is a very large dragonfly, with a wingspan that can exceed 10 cm. During sunny days, the majority of their activities comprise flying over ponds or hawking along the forest for food. D. bipunctata dragonflies, on the other hand, are much smaller in size (wingspan of approximately 4 cm) and usually fly short distances between perches. The complicated surface structures of both insects are precisely designed to suit their respective living behaviours. H. papuensis was found to have more hierarchical structures that enable them to trap more air in between the pillars. Hierarchical roughness makes superhydrophobic states more stable, which can be very useful for an organism when air flow along their wings is crucial to its living behaviour (Nosonovsky and Bhushan, 2007a). The wings of H. papuensis dragonflies exhibited higher WCA, which indicates a larger amount of trapped air in the surface and hence assists in the self-cleaning properties that protect them from contaminants. The surfaces of the wings of D. bipunctata dragonflies, on the other hand, are both superhydrophobic and bactericidal, and the flexibility afforded by the combination of these properties may make them more adaptable to a wider range of living conditions, hence their larger geographical distribution.

167

9.4. Fabrication and evaluation of synthetic surfaces with the ability to control bacterial attachment

The alkanes and fatty acids that were identified as the main components present on the surfaces of dragonfly wings were used to make artificial wing surfaces. Highly ordered pyrolytic graphite (HOPG) was used as the substrata in order to produce an ordered surface structure, as it is well established in the literature to have being highly smooth with periodic molecular organisation. The interaction between alkanes and HOPG produced microarchitectures with preferential orientations of 30° and 60° in a 168 plane parallel to the substratum surface. The crystallites structure consequently controlled bacterial alignment on the surfaces (Chapter 7). Confocal laser scanning microscopy (CLSM) showed highly directed bacterial alignment for two bacterial species (spherical and rod-shaped cells), reflecting the preferential orientation of the microcrystallite–air–water interfaces to give linear and triangular bacterial patterning. The relative hydrophobicity of the two species of bacterial cells was shown to influence both the degree of immersion into the hydrophobic air pocket and the density of their attachment to the respective lipid microstructures. These microtextured surfaces that exhibit the ability to limit bacterial adhesion to a precise patterning at the lipid three phase lines could be used as a means to control the way in which bacterial colonisation of a surface takes place.

On the other hand, the interaction between fatty acids and HOPG produces microblades which resulted in the ability of the surface to kill bacterial cells. The mechanisms were found to be similar to that of dragonfly wings, i.e. mechanobactericidal. Several works have shown that entrapped air pockets can be instrumental in guiding the attachment of bacterial cells (Genzer and Efimenko, 2006; Scardino, Zhang, Cookson et al., 2009; Hsu, Fang, Borca-Tasciuc et al., 2013; Wu, Nakanishi, Cho et al., 2013; Vasudevan, Kennedy, Merritt et al., 2014), due to the cells’ inability to cross the air- water interfacial boundary. In the case of the palmitic acid microblades, entrapped air pockets prevented the attachment of cells to the regions in which they would otherwise find refuge from the bactericidal actions of the surface. Wetting phenomena at the three- phase line (TPL) of the system were shown in Chapter 7 to direct the positioning of the adhering cells, thus it can be inferred that an imbalance of adhesion forces on a cell nearing the TPL results in a net force toward the TPL. If a cell was nearing two different TPLs, such as on either side of the narrow ridged palmitic microblades, the resulting net force would cause the cell to be ruptured on the high curvature of the sharp edge. The opposing components of the applied forces may also aid in rupture the cell.

In nature, palmitic acid and stearic acid can be found in many organisms as part of their normal metabolic activity. This suggests that the prospects for biocompatibility of bactericidal surfaces made from these materials may be high. The fact that the ordered recrystallised structure can only be produced on ordered substrata, in this case HOPG, may limit the potential applications for these surfaces, however, the information

169 generated here will contribute to the future development of new materials that can be applied on a commercial scale.

170

Chapter 10. Conclusions and future directions ______

171

10.1. Summary and conclusions

Some naturally existing surfaces (e.g. lotus leaves and dragonfly wings) are able to maintain infection-free surfaces, and have themselves become templates for the development of synthetic superhydrophobic, and self-cleaning surfaces. For this reason, a large amount of work has been conducted on plant leaves, but much less time has been invested in studying the surfaces of insects, despite the fact they are required to cope with very similar living conditions. The objectives of this work were to reconstruct and evaluate model bactericidal surfaces based on insect wing-derived materials, with a particular view to differentiating between anti-biofouling and bactericidal activity.

The surfaces of insect wings have been shown to be covered by a layer of nanopillars, which directly influence their multi-functional properties, i.e. superhydrophobicity, self-cleaning and bactericidal activities. The molecular organisation of individual nanopillars on the surface of dragonfly wings were characterised and over 60 different substances were identified. The pillars consisted of up to 62 % of long chain aliphatic hydrocarbons, with the remaining 38 % being attributed to oxygenated compounds, largely due to the detection of fatty acids, i.e. palmitic acid and stearic acid. By combining the data from GCMS with depth-profiling XPS, the epicuticular layer of dragonfly wings could be divided in three distinct sub- layers, the inner epicuticle, the meso-epicuticle (now proposed in this work), and the outer epicuticle. Two factors were shown to simultaneously control the surface wettability; the precise chemical composition not only contributed directly to the hydrophobicity of the wing, but also determined the topology of the surface.

The precise molecular organisation of the surfaces of dragonfly wings determined their wettability, and subsequently their self-cleaning and bactericidal activities. The surface topographies of Hemianax papuensis and Diplacodes bipunctata dragonflies were characterised and compared to identify the correlation between surface topography and bactericidal activity. It was found that minor differences in the surface topologies could switch the performance of the wings from anti-biofouling to bactericidal, and it is postulated here that surface wettability plays significant role in controlling these mechanisms. The ability of the wings to entrap larger quantities of air was an indication of lower bactericidal ability; in such cases the wing surface tended to favour self-

172 cleaning activity. Slightly increased surface wettability tended to result in more effective bactericidal activity and cell rupture.

Having gained an understanding of the contributions of surface chemistry and surface topography on the superhydrophobicity and bactericidal activity of dragonfly wings, several of the major components present in the epicuticular layer of the dragonfly wings were selected for the fabrication of model surfaces. These components were self-assembled onto an ordered substratum, which was highly-ordered pyrolytic graphite (HOPG). Two alkanes, eicosane and docosane, were recrystallised onto the HOPG surfaces resulting in an assembly of patterned microrodlets, whilst the fatty acids, i.e. palmitic acid and stearic acid, recrystallised on the substrate to form microblade structures. The ordered alkane microcrystals controlled the attachment of two species of bacteria, Staphylococcus aureus and Pseudomonas aeruginosa, but did not appear to harm the bacteria in any way. The microcrystals created a three-phase interface with entrapped air pockets, which dictated the attachment and ordering of bacterial cells to certain regions of the surface. Similarly, the fatty acid microcrystals also caused the cells to align at the three-phase line, however the shape of the microcrystals in these regions generated a mechanobactericidal effect that led to cell death for each of these bacterial species.

In conclusion, this work has demonstrated the underlying mechanisms operating on a natural mechanobactericidal surface and allowed the modelling of the key features of these mechanisms, which can then be used to create synthetic analogues.

10.2. Future directions

While the current work has generated useful knowledge regarding the study of insect wings and the control of bacterial colonisation, further work is still needed to successfully apply this knowledge to real-world technology. For example, combinations of the various different lipid components would likely undergo different forms of self- assembly. With the correct selection of components, it may be possible to produce a precise structure with greater resemblance to that found on the surface of dragonfly wings. Different substrata, for examples plastics, metallic surfaces, etc. may also lead to different structures being formed with different properties.

One major drawback of HOPG as a substratum is its limitation in size and shape due to manufacturing reasons. A possible alternative may be a graphene coating, which 173 may provide a substrate that has greater flexibility in potential applications. The molecular structure of graphene sheets and an individual layer of HOPG are the same; therefore by successfully coating a single graphene sheet it may be possible to produce self-assembled lipid structures similar to those investigated here.

10.3. Final remarks

Preventing bacterial colonisation is an important aspect of everyday life. Bacterial infections can lead to severe damage to not only the economy but also to public health. Billions of dollars are spent annually cleaning the hulls of ships, and over 100,000 mortalities due to bacterial infections are reported annually, not to mention the significant costs associated with supporting healthcare systems for the treatment of infections. Recent discoveries in the areas of superhydrophobicity, self-cleaning, and mechanobactericidal surfaces are promising high-potential alternatives to current strategies for controlling the spread of bacteria and bacterial infections.

The data presented here contributes to the scientific knowledge regarding the superhydrophobicity, self-cleaning and mechanobactericidal properties of dragonfly wings. The information established in this research has shown that all of these three properties are highly interrelated to each other. Based on the data presented here, it may be possible to produce multi-tasking surfaces that possess both self-cleaning and bactericidal properties, if the multiple mechanisms employed by these insects can be simultaneously adapted.

174

Bibliography ______

175

(2014) Encyclopedia of Life. Accessed: 22 Dec 2014. Available from: http://www.eol.org

Andersen S.O., Weis-Fogh T. (1964) Resilin. A rubberlike protein in Arthropod cuticle. Advances in Insect Physiology 2,1.

Angst E.C. The fouling of ships bottoms by bacteria. Rep, Bur Constr Repair. Washington, DC: US Navy Department; 1923.

Arsene C., Schulz S., Van Loon J.J.A. (2002) Chemical polymorphism of the cuticular lipids of the cabbage white Pieris rapae. Journal of Chemical Ecology 28,2627.

Barth A. (2007) Infrared spectroscopy of proteins. Biochimica et Biophysica Acta - Bioenergetics 1767,1073.

Barthlott W., Neinhuis C. (1997) Purity of the sacred lotus, or escape from contamination in biological surfaces. Planta 202,1.

Barthlott W., Neinhuis C., Cutler D., Ditsch F., Meusel I., Theisen I., Wilhelmi H. (1998) Classification and terminology of plant epicuticular waxes. Botanical Journal of the Linnean Society 126,237.

Barthlott W., Theisen I., Borsch T., Neinhuis C. (2003) Epicuticular waxes and vascular plant systematics: integrating micromorphological and chemical data In: Stuessy T.F., Mayer V., Horandl E. Deep Morphology Toward a Renaissance of Morphology in Plant Systematics. (Gantner Verlag, Ruggell) pp:189.

Beattie D.A., Beaussart A., Mierczynska-Vasilev A., Harmer S.L., Thierry B., Puskar L., Tobin M. (2012) Synchrotron FTIR microscopy of Langmuir-Blodgett monolayers and polyelectrolyte multilayers at the solid-solid interface. Langmuir 28,1683.

Bechert D.W., Bruse M., Hage W., Van Der Hoeven J.G.T., Hoppe G. (1997) Experiments on drag-reducing surfaces and their optimization with an adjustable geometry. Journal of Fluid Mechanics 338,59.

Behlau I., Gilmore M.S. (2008) Microbial biofilms in ophthalmology and infectious disease. Archives of Ophthalmology 126,1572.

Berne B.J., Weeks J.D., Zhou R. (2009) Dewetting and hydrophobic interaction in physical and biological systems. Annual Review of Physical Chemistry 60,85.

176

Bernier U.R., Carlson D.A., Geden C.J. (1998) Gas chromatography/mass spectrometry analysis of the cuticular hydrocarbons from parasitic wasps of the genus Muscidifurax. Journal of the American Society for Mass Spectrometry 9,320.

Beveridge T.J. (1999) Structures of Gram-negative cell walls and their derived membrane vesicles. Journal of Bacteriology 181,4725.

Bhushan B. (2007) Adhesion of multi-level hierarchical attachment systems in gecko feet. Journal of Adhesion Science and Technology 21,1213.

Bhushan B. (2011) Biomimetics inspired surfaces for drag reduction and oleophobicity/philicity. Beilstein Journal of Nanotechnology 2,66.

Bhushan B. (2012) Bioinspired structured surfaces. Langmuir 28,1698.

Bhushan B., Her E.K. (2010) Fabrication of superhydrophobic surfaces with high and low adhesion inspired from rose petal. Langmuir 26,8207.

Bhushan B., Jung Y.C. (2008) Wetting, adhesion and friction of superhydrophobic and hydrophilic leaves and fabricated micro/nanopatterned surfaces. Journal of Physics Condensed Matter 20.

Bhushan B., Jung Y.C., Koch K. (2009a) Micro-, nano- and hierarchical structures for superhydrophobicity, self-cleaning and low adhesion. Philosophical Transactions Series A, Mathematical, Physical, and Engineering Sciences 367,1631.

Bhushan B., Jung Y.C., Koch K. (2009b) Self-cleaning efficiency of artificial superhydrophobic surfaces. Langmuir 25,3240.

Bhushan B., Sayer R.A. (2008) Gecko feet: Natural attachment systems for smart adhesion-mechanism, modeling, and development of bio-inspired materials. In: Bhushan B., Fuchs H., Tomitori M. Applied Scanning Probe Methods X: Biomimetics and Industrial Applications. (Springer, Berlin) pp.

Bico J., Marzolin C., Quéré D. (1999) Pearl drops. Europhysics Letters 47,220.

Bixler G.D., Bhushan B. (2012) Bioinspired rice leaf and butterfly wing surface structures combining shark skin and lotus effects. Soft Matter 8,11271.

Bixler G.D., Bhushan B. (2013) Fluid drag reduction and efficient self-cleaning with rice leaf and butterfly wing bioinspired surfaces. Nanoscale 5,7685.

177

Boeve J.L., Voigt D., Gorb S.N. (2011) Crystalline wax coverage of the cuticle in easy bleeding sawfly larvae. Arthropod Structure and Development 40,186.

Bogus M.I., Czygier M., Golbiowski M., Kdra E., Kucinska J., Mazgajska J., Samborski J., Wieloch W., Wloka E. (2010) Effects of insect cuticular fatty acids on in vitro growth and pathogenicity of the entomopathogenic fungus Conidiobolus coronatus. Experimental Parasitology 125,400.

Bond A.D. (2004) On the crystal structures and melting point alternation of the n-alkyl carboxylic acids. New Journal of Chemistry 28,104.

Bormashenko E., Bormashenko Y., Stein T., Whyman G., Bormashenko E. (2007) Why do pigeon feathers repel water? Hydrophobicity of pennae, Cassie-Baxter wetting hypothesis and Cassie-Wenzel capillarity-induced wetting transition. Journal of Colloid and Interface Science 311,212.

Bos R., Van Der Mei H.C., Busscher H.J. (1999) Physico-chemistry of initial microbial adhesive interactions - Its mechanisms and methods for study. FEMS Microbiology Reviews 23,179.

Briggs D., Seah M.P. (1983) Practical surface analysis: Auger and x-ray photoelectron spectroscopy. (John Wiley & Sons,

Brugnerotto J., Lizardi J., Goycoolea F.M., Argüelles-Monal W., Desbrières J., Rinaudo M. (2001) An infrared investigation in relation with chitin and chitosan characterization. Polymer 42,3569.

Buckner J.S. (2010) Oxygenated derivatives of hydrocarbons. In: Blomquist G.J., Bagnères A.-G. Insect hydrocarbons: Biology, biochemistry, and chemical ecology. (Cambridge University Press, Cambridge) pp:187.

Buckner J.S., Hagen M.M. (2003) Triacylglycerol and phospholipid fatty acids of the silverleaf whitefly: Composition and biosynthesis. Archives of Insect Biochemistry and Physiology 53,66.

Buckner J.S., Hagen M.M., Nelson D.R. (1999) The composition of the cuticular lipids from nymphs and exuviae of the silverleaf whitefly, Bemisia argentifolii. Comparative Biochemistry and Physiology Part B, Biochemistry & Molecular Biology 124,201.

178

Buckner J.S., Mardaus M.C., Nelson D.R. (1996) Curticular lipid composition of Heliothis virescens and Helicoverpa zea pupae. Comparative Biochemistry and Physiology Part B, Biochemistry & Molecular Biology 114,207.

Buckner J.S., Nelson D.R., Hakk H., Pomonis J.G. (1984) Long chain oxoaldehydes and oxoalcohols from esters as major constituents of the surface lipids of Manduca sexta pupae in diapause. The Journal of Biological Chemistry 259,8452.

Buckner J.S., Nelson D.R., Mardaus M.C. (1994) The lipid composition of the wax particles from adult whiteflies, Bemisia tabaci and Trialeurodes vaporariorum. Insect Biochemistry and Molecular Biology 24,977.

Bulgarelli D., Schlaeppi K., Spaepen S., Van Themaat E.V.L., Schulze-Lefert P. (2013) Structure and functions of the bacterial microbiota of plants. Annual Review of Plant Biology 64,807.

Burger W., Burge M.J. (2007) Digital image processing. (Springer, New York).

Burton Z., Bhushan B. (2005) Hydrophobicity, adhesion, and friction properties of nanopatterned polymers and scale dependence for micro- and nanoelectromechanical systems. Nano Letters 5,1607.

Buschhaus C., Herz H., Jetter R. (2007) Chemical composition of the epicuticular and intracuticular wax layers on adaxial sides of Rosa canina leaves. Annals of Botany 100,1557.

Buschhaus C., Jetter R. (2011) Composition differences between epicuticular and intracuticular wax substructures: How do plants seal their epidermal surfaces? Journal of Experimental Botany 62,841.

Bushnell D.M., Moore K.J. (1991) Drag reduction in nature. Annual Review of Fluid Mechanics 23,65.

Butt H.J., Cappella B., Kappl M. (2005) Force measurements with the atomic force microscope: Technique, interpretation and applications. Surface Science Reports 59,1.

Byun D., Hong J., Saputra, Ko J.H., Lee Y.J., Park H.C., Byun B.K., Lukes J.R. (2009) Wetting characteristics of insect wing surfaces. Journal of Bionic Engineering 6,63.

Callies M., Quéré D. (2005) On water repellency. Soft Matter 1,55.

179

Carballeira N.M. (2008) New advances in fatty acids as antimalarial, antimycobacterial and antifungal agents. Progress in Lipid Research 47,50.

Carman M.L., Estes T.G., Feinberg A.W., Schumacher J.F., Wilkerson W., Wilson L.H., Callow M.E., Callow J.A., Brennan A.B. (2006) Engineered antifouling microtopographies - Correlating wettability with cell attachment. Biofouling 22,11.

Caruso F., Furlong D.N., Ariga K., Ichinose I., Kunitake T. (1998) Characterization of polyelectrolyte-protein multilayer films by atomic force microscopy, scanning electron microscopy, and fourier transform infrared reflection-absorption spectroscopy. Langmuir 14,4559.

Cassie A.B.D., Baxter S. (1944) Wettability of porous surfaces. Transactions of the Faraday Society 40,546.

Chattopadhyay S., Huang Y.F., Jen Y.J., Ganguly A., Chen K.H., Chen L.C. (2010) Anti-reflecting and photonic nanostructures. Materials Science & Engineering R, Reports 69,1.

Chen Y., Wang X., Ren H., Li X. (2011) An organic junction between the vein and membrane of the dragonfly wing. Chinese Science Bulletin 56,1658.

Cheng L. (1973) Marine and freshwater skaters: Differences in surface fine structures. Nature 242,132.

Cheng Y.T., Rodak D.E., Wong C.A., Hayden C.A. (2006) Effects of micro- and nanostructures on the self-cleaning behaviour of lotus leaves. Nanotechnology 17,1359.

Choe D.H., Ramírez S.R., Tsutsui N.D. (2012) A silica gel based method for extracting insect surface hydrocarbons. Journal of Chemical Ecology 38,176.

Coury C., Dillner A.M. (2008) A method to quantify organic functional groups and inorganic compounds in ambient aerosols using attenuated total reflectance FTIR spectroscopy and multivariate chemometric techniques. Atmospheric Environment 42,5923.

Crawford R., Koopal L.K., Ralston J. (1987) Contact angles on particles and plates. Colloids and Surfaces 27,57.

180

Crawford R.J., Webb H.K., Truong V.K., Hasan J., Ivanova E.P. (2012) Surface topographical factors influencing bacterial attachment. Advances in Colloid and Interface Science 179-182,142.

Dalton H.M., March P.E. (1998) Molecular genetics of bacterial attachment and biofouling. Current Opinion in Biotechnology 9,252.

Dani F.R. (2006) Cuticular lipids as semiochemicals in paper wasps and other social insects. Annales Zoologici Fennici 43,500.

Daniels R., Vanderleyden J., Michiels J. (2004) Quorum sensing and swarming migration in bacteria. FEMS Microbiology Reviews 28,261.

Dean B., Bhushan B. (2010) Shark-skin surfaces for fluid-drag reduction in turbulent flow: A review. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 368,4775.

Decker J.T., Kirschner C.M., Long C.J., Finlay J.A., Callow M.E., Callow J.A., Brennan A.B. (2013) Engineered antifouling microtopographies: An energetic model that predicts cell attachment. Langmuir 29,13023.

Di Iaconi C., Ramadori R., Lopez A., Passino R. (2005) Hydraulic shear stress calculation in a sequencing batch biofilm reactor with granular biomass. Environmental Science and Technology 39,889.

Diaz C., Cortizo M.C., Schilardi P.L., de Saravia S.G.G., de Mele M.A.F.L. (2007) Materials Research 10,11.

Donlan R.M. (2001) Clinical Infectious Diseases 33,1387.

Dowhan W. (1997) Molecular basis for membrane phospholipid diversity: Why are there so many lipids? Annual Review of Biochemistry 66,199.

Doyle R.J., Marquis R.E. (1994) Elastic, flexible peptidoglycan and bacterial cell wall properties. Trends in Microbiology 2,57.

Ellington C.P. (1999) The novel aerodynamics of insect flight: Applications to micro-air vehicles. Journal of Experimental Biology 202,3439.

Endler A., Liebig J., Schmitt T., Parker J.E., Jones G.R., Schreier P., Hölldobler B. (2004) Surface hydrocarbons of queen eggs regulate worker reproduction in a social

181

insect. Proceedings of the National Academy of Sciences of the United States of America 101,2945.

Ensikat H.J., Ditsche-Kuru P., Neinhuis C., Barthlott W. (2011) Superhydrophobicity in perfection: The outstanding properties of the lotus leaf. Beilstein Journal of Nanotechnology 2,152.

Epand R.M., Rotem S., Mor A., Berno B., Epand R.F. (2008) Bacterial membranes as predictors of antimicrobial potency. Journal of the American Chemical Society 130,14346.

Epstein A.K., Hochbaum A.I., Kim P., Aizenberg J. (2011) Control of bacterial biofilm growth on surfaces by nanostructural mechanics and geometry. Nanotechnology 22.

Escobar Galindo R., Gago R., Duday D., Palacio C. (2010) Towards nanometric resolution in multilayer depth profiling: A comparative study of RBS, SIMS, XPS and GDOES. Analytical and Bioanalytical Chemistry 396,2725.

Federle W., Barnes W.J.P., Baumgartner W., Drechsler P., Smith J.M. (2006) Wet but not slippery: Boundary friction in tree frog adhesive toe pads. Journal of the Royal Society Interface 3,689.

Feng L., Li S., Li Y., Li H., Zhang L., Zhai J., Song Y., Liu B., Jiang L., Zhu D. (2002) Super-hydrophobic surfaces: From natural to artificial. Advanced Materials 14,1857.

Feng L., Zhang Y., Xi J., Zhu Y., Wang N., Xia F., Jiang L. (2008) Petal effect: A superhydrophobic state with high adhesive force. Langmuir 24,4114.

Ferrari M., Ravera F. (2010) Surfactants and wetting at superhydrophobic surfaces: Water solutions and non aqueous liquids. Advances in Colloid and Interface Science 161,22.

Ferveur J.F. (2005) Cuticular hydrocarbons: Their evolution and roles in Drosophila pheromonal communication. Behavior Genetics 35,279.

Fogg G.E. (1944) Diurnal fluctuation in a physical property of leaf cuticle. Nature 154,515.

182

Fournier J.B., Galatola P. (2002) Anisotropic capillary interactions and jamming of colloidal particles trapped at a liquid-fluid interface. Physical Review E - Statistical, Nonlinear, and Soft Matter Physics 65,031601/1.

Fox H.W., Zisman W.A. (1952) The spreading of liquids on low-energy surfaces. III. Hydrocarbon surfaces. Journal of Colloid Science 7,428.

Franklin N.M., Rogers N.J., Apte S.C., Batley G.E., Gadd G.E., Casey P.S. (2007)

Comparative toxicity of nanoparticulate ZnO, bulk ZnO, and ZnCl2 to a freshwater microalga (Pseudokirchneriella subcapitata): The importance of particle solubility. Environmental Science and Technology 41,8484.

Fu G., Vary P.S., Lin C.T. (2005) Anatase TiO2 nanocomposites for antimicrobial coatings. Journal of Physical Chemistry B 109,8889.

Fürstner R., Barthlott W., Neinhuis C., Walzel P. (2005) Wetting and self-cleaning properties of artificial superhydrophobic surfaces. Langmuir 21,956.

Gadelmawla E.S., Koura M.M., Maksoud T.M.A., Elewa I.M., Soliman H.H. (2002) Roughness parameters. Journal of Materials Processing Technology 123,133.

Ganim Z., Hoi S.C., Smith A.W., Deflores L.P., Jones K.C., Tokmakoff A. (2008) Amide I two-dimensional infrared spectroscopy of proteins. Accounts of Chemical Research 41,432.

Gao H., Wang X., Yao H., Gorb S., Arzt E. (2005) Mechanics of hierarchical adhesion structures of geckos. Mechanics of Materials 37,275.

Garrett T.R., Bhakoo M., Zhang Z. (2008) Bacterial adhesion and biofilms on surfaces. Progress in Natural Science 18,1049.

Gazit E. (2007) Self-assembled peptide nanostructures: The design of molecular building blocks and their technological utilization. Chemical Society Reviews 36,1263.

Gbabode G., Negrier P., Mondieig D., Leger J.M., Moreno E., Calvet T., Cuevas-Diarte M.A. (2006) Crystal structure of the B′ form of nonadecanoic acid. Analytical Sciences: X-ray Structure Analysis Online 22,x269.

183

Geesey G.G., Richardson W.T., Yeomans H.G., Irvin R.T., Costerton J.W. (1977) Microscopic examination of natural sessile bacterial populations from an alpine stream. Canadian Journal of Microbiology 23,1733.

Geiselhardt S.F., Geiselhardt S., Peschke K. (2009) Comparison of tarsal and cuticular chemistry in the leaf beetle Gastrophysa viridula (Coleoptera: Chrysomelidae) and an evaluation of solid-phase microextraction and solvent extraction techniques. Chemoecology 19,185.

Gelover S., Gómez L.A., Reyes K., Teresa Leal M. (2006) A practical demonstration of

water disinfection using TiO2 films and sunlight. Water Research 40,3274.

Genzer J., Efimenko K. (2006) Recent developments in superhydrophobic surfaces and their relevance to marine fouling: A review. Biofouling 22,339.

Gibbs A.G. (1998) Water-proofing properties of cuticular lipids. American Zoologist 38,471.

Gołebiowski M., Bogus M.I., Paszkiewicz M., Stepnowski P. (2010) The composition of the free fatty acids from Dendrolimus pini exuviae. Journal of Insect Physiology 56,391.

Gołebiowski M., Bogus M.I., Paszkiewicz M., Stepnowski P. (2011) Cuticular lipids of insects as potential biofungicides: Methods of lipid composition analysis. Analytical and Bioanalytical Chemistry 399,3177.

Gołebiowski M., Maliński E., Boguś M.I., Kumirska J., Stepnowski P. (2008) The cuticular fatty acids of Calliphora vicina, Dendrolimus pini and Galleria mellonella larvae and their role in resistance to fungal infection. Insect Biochemistry and Molecular Biology 38,619.

Gołebiowski M., Malinski E., Nawrot J., Stepnowski P. (2008) Identification and characterization of surface lipid components of the dried-bean beetle Acanthoscelides obtectus (Say) (Coleoptera: Bruchidae). Journal of Stored Products Research 44,386.

Gorb S.N. (1997) Porous channels in the cuticle of the head-arrester system in dragon/damselflies (Insecta: Odonata). Microscopy Research and Technique 37,583.

184

Gorb S.N. (1999) Serial elastic elements in the damselfly wing: Mobile vein joints contain resilin. Naturwissenschaften 86,552.

Gorb S.N., Kesel A., Berger J. (2000) Microsculpture of the wing surface in Odonata: Evidence for cuticular wax covering. Arthropod Structure and Development 29,129.

Gorb S.N., Tynkkynen K., Kotiaho J.S. (2009) Crystalline wax coverage of the imaginal cuticle in Calopteryx splcndcns (Odonata: Calopterygidae). International Journal of Odonatology 12,205.

Grimaldi D., Engel M.S. (2005) Insects take to the skies. Evolution of the insects. Cambridge University Press) pp:155.

Gu Z.Z., Uetsuka H., Takahashi K., Nakajima R., Onishi H., Fujishima A., Sato O. (2003) Structural color and the lotus effect. Angewandte Chemie - International Edition 42,894.

Gumbart J., Wiener M.C., Tajkhorshid E. (2007) Mechanics of force propagation in TonB-dependent outer membrane transport. Biophysical Journal 93,496.

Guo Z., Liu W. (2007) Biomimic from the superhydrophobic plant leaves in nature: Binary structure and unitary structure. Plant Science 172,1103.

Guy D.W., Crawford R.J., Mainwaring D.E. (1996) The wetting behaviour of several organic liquids in water on coal surfaces. Fuel 75,238.

Hackman R.H., Goldberg M. (1987) Comparative study of some expanding arthropod cuticles: The relation between composition, structure and function. Journal of Insect Physiology 33,39.

Hasan J., Webb H.K., Truong V.K., Pogodin S., Baulin V.A., Watson G.S., Watson J.A., Crawford R.J., Ivanova E.P. (2012a) Selective bactericidal activity of nanopatterned superhydrophobic cicada Psaltoda claripennis wing surfaces. Applied Microbiology and Biotechnology 97,9257.

Hasan J., Webb H.K., Truong V.K., Watson G.S., Watson J.A., Tobin M.J., Gervinskas G., Juodkazis S., Wang J.Y., Crawford R.J., Ivanova E.P. (2012b) Spatial variations and temporal metastability of the self-cleaning and superhydrophobic properties of damselfly wings. Langmuir 28,17404.

185

Hay M.E. (1996) Marine chemical ecology: What's known and what's next? Journal of Experimental Marine Biology and Ecology 200,103.

Henrici A.T. (1933) Studies of freshwater bacteria. I.A direct microscopic technique. Journal of Bacteriology 25,277.

Hepburn H.R., Chandler H.D. (1976) Material properties of arthropod cuticles: The arthrodial membranes. Journal of Comparative Physiology B 109,177.

Hermansson K., Lindberg U., Hok B., Palmskog G. Wetting properties of silicon surfaces. Transducer' 911991. p. 193

Herminghaus S. (2000) Roughness-induced non-wetting. Europhysics Letters 52,165.

Higgins A.M., Jones R.A.L. (2000) Anisotropic spinodal dewetting as a mute to self- assembly of patterned surfaces. Nature 404,476.

Hochbaum A.I., Aizenberg J. (2010) Bacteria pattern spontaneously on periodic nanostructure arrays. Nano Letters 10,3717.

Holmes A., Doré C.J., Saraswatula A., Bamford K.B., Richards M.S., Coello R., Modi N. (2008) Risk factors and recommendations for rate stratification for surveillance of neonatal healthcare-associated bloodstream infection. Journal of Hospital Infection 68,66.

Hooper I.R., Vukusic P., Wootton R.J. (2006) Detailed optical study of the transparent wing membranes of the dragonfly Aeshna cyanea. Optics Express 14,4891.

Howard R.W., Baker J.E. (2003) Cuticular hydrocarbons and wax esters of the ectoparasitoid Habrobracon hebetor: Ontogenetic, reproductive, and nutritional effects. Archives of Insect Biochemistry and Physiology 53,1.

Howard R.W., Blomquist G.J. (2005) Ecological, behavioral, and biochemical aspects of insect hydrocarbons. Annual Review of Entomology 50,371.

Howard R.W., Lord J.C. (2003) Cuticular lipids of the booklouse, Liposcelis bostrychophila: Hydrocarbons, aldehydes, fatty acids, and fatty acid amides. Journal of Chemical Ecology 29,615.

Hsiao M.T., Chen S.F., Shieh D.B., Yeh C.S. (2006) One-pot synthesis of hollow

Au3Cu1 spherical-like and biomineral botallackite Cu2(OH)3Cl flowerlike

186

architectures exhibiting antimicrobial activity. Journal of Physical Chemistry B 110,205.

Hsu L.C., Fang J., Borca-Tasciuc D.A., Worobo R.W., Moraru C.I. (2013) Effect of micro- and nanoscale topography on the adhesion of bacterial cells to solid surfaces. Applied and Environmental Microbiology 79,2703.

Hu D.L., Chan B., Bush J.W.M. (2003) The hydrodynamics of water strider locomotion. Nature 424,663.

Hu H.M.S., Watson G.S., Cribb B.W., Watson J.A. (2011) Non-wetting wings and legs of the cranefly aided by fine structures of the cuticle. Journal of Experimental Biology 214,915.

Huang J., Wang X., Wang Z.L. (2006) Controlled replication of butterfly wings for achieving tunable photonic properties. Nano Letters 6,2325.

Ibáñez J.A., Litter M.I., Pizarro R.A. (2003) Photocatalytic bactericidal effect of TiO2 on Enterobacter cloacae. Comparative study with other Gram (-) bacteria. Journal of Photochemistry and Photobiology A: Chemistry 157,81.

In S., Orlov A., Berg R., García F., Pedrosa-Jimenez S., Tikhov M.S., Wright D.S., Lambert R.M. (2007) Effective visible light-activated B-doped and B,N-codoped

TiO2 photocatalysts. Journal of the American Chemical Society 129,13790.

Ivanova E.P., Hasan J., Webb H.K., Truong V.K., Watson G.S., Watson J.A., Baulin V.A., Pogodin S., Wang J.Y., Tobin M.J., Löbbe C., Crawford R.J. (2012) Natural bactericidal surfaces: Mechanical rupture of Pseudomonas aeruginosa cells by cicada wings. Small 8,2489.

Ivanova E.P., Hasan J., Webb K.W., Gervinskas G., Juodkazis S., Truong V.K., Wu A.H.F., Lamb R.N., Baulin V.A., Watson G.S., Watson J.A., Mainwaring D.E., Crawford R.J. (2013) Bactericidal activity of black silicon. Nature Communications,2838

Ivanova E.P., Truong V.K., Wang J.Y., Bemdt C.C., Jones R.T., Yusuf I.I., Peake I., Schmidt H.W., Fluke C., Barnes D., Crawford R.J. (2010) Impact of nanoscale roughness of titanium thin film surfaces on bacterial retention. Langmuir 26,1973.

187

Jackson M., Mantsch H.H. (1995) The use and misuse of FTIR spectroscopy in the determination of protein structure. Critical Reviews in Biochemistry and Molecular Biology 30,95.

Jarrold S.L., Moore D., Potter U., Charnley A.K. (2007) The contribution of surface waxes to pre-penetration growth of an entomopathogenic fungus on host cuticle. Mycological Research 111,240.

Jeffree C.E. (2006) In: Riederer M., Müller C. Biology of the plant cuticle. (Blackwell Pub., Oxford, UK) pp:11.

Jetter R., Kunst L. (2008) Plant surface lipid biosynthetic pathways and their utility for metabolic engineering of waxes and hydrocarbon biofuels. Plant Journal 54,670.

Jones N., Ray B., Ranjit K.T., Manna A.C. (2008) Antibacterial activity of ZnO nanoparticle suspensions on a broad spectrum of microorganisms. FEMS Microbiology Letters 279,71.

Jung Y.C., Bhushan B. (2006) Contact angle, adhesion and friction properties of micro- and nanopatterned polymers for superhydrophobicity. Nanotechnology 17,4970.

Kaneko F., Kobayashi M., Kitagawa Y., Matsuura Y. (1992) Collective displacement of acyl chains on E → C phase transition of stearic acid. Journal of Physical Chemistry 96,7104.

Keeling C.I., Bohlmann J. (2006) Diterpene resin acids in conifers. Phytochemistry 67,2415.

Kesel A.B., Philippi U., Nachtigall W. (1998) Biomechanical aspects of the insect wing: An analysis using the finite element method. Computers in Biology and Medicine 28,423.

Kjelleberg S., Molin S. (2002) Is there a role for quorum sensing signals in bacterial biofilms? Current Opinion in Microbiology 5,254.

Klevens R.M., Edwards J.R., Richards Jr C.L., Horan T.C., Gaynes R.P., Pollock D.A., Cardo D.M. (2007) Estimating health care-associated infections and deaths in U.S. Hospitals, 2002. Public Health Reports 122,160.

Knox K.W., Wicken A.J. (1973) Immunological properties of teichoic acids. Bacteriological Reviews 37,215.

188

Koch A.L., Woeste S. (1992) Elasticity of the sacculus of Escherichia coli. Journal of Bacteriology 174,4811.

Koch K., Barthlott W., Koch S., Hommes A., Wandelt K., Mamdouh W., De-Feyter S., Broekmann P. (2006) Structural analysis of wheat wax (Triticum aestivum, c.v. 'Naturastar' L.): from the molecular level to three dimensional crystals. Planta 223,258.

Koch K., Bhushan B., Barthlott W. (2008) Diversity of structure, morphology and wetting of plant surfaces. Soft Matter 4,1943.

Koch K., Bhushan B., Ensikat H.J., Barthlott W. (2009) Self-healing of voids in the wax coating on plant surfaces. Philosophical Transactions Series A, Mathematical, Physical, and Engineering Sciences 367,1673.

Koch K., Bhushan B., Jung Y.C., Barthlott W. (2009) Fabrication of artificial Lotus leaves and significance of hierarchical structure for superhydrophobicity and low adhesion. Soft Matter 5,1386.

Koch K., Dommisse A., Barthlott W. (2006) Chemistry and crystal growth of plant wax tubules of lotus (Nelumbo nucifera) and nasturtium (Tropaeolum majus) leaves on technical substrates. Crystal Growth and Design 6,2571.

Koch K., Ensikat H.J. (2008) The hydrophobic coatings of plant surfaces: Epicuticular wax crystals and their morphologies, crystallinity and molecular self-assembly. Micron 39,759.

Koch K., Neinhuis C., Ensikat H.J., Barthlott W. (2004) Self assembly of epicuticular waxes on living plant surfaces imaged by atomic force microscopy (AFM). Journal of Experimental Botany 55,711.

Kramer K.J., Muthukrishnan S. (1997) Insect chitinases: Molecular biology and potential use as biopesticides. Insect Biochemistry and Molecular Biology 27,887.

Kreuz P., Arnold W., Kesel A.B. (2001) Acoustic microscopic analysis of the biological structure of insect wing membranes with emphasis on their waxy surface. Annals of Biomedical Engineering 29,1054.

189

Küçükbay F.Z., Kuyumcu E., Bilenler T., Yildiz B. (2012) Chemical composition and antimicrobial activity of essential oil of Achillea cretica L. (Asteraceae) from Turkey. Natural Product Research 26,1668.

Lafuma A., Quéré D. (2003) Superhydrophobic states. Nature Materials 2,457.

Lamolle S.F., Monjo M., Lyngstadaas S.P., Ellingsen J.E., Haugen H.J. (2009) Titanium implant surface modification by cathodic reduction in hydrofluoric acid: Surface characterization and in vivo performance. Journal of Biomedical Materials Research Part A 88,581.

Lapointe S.L., Hunter W.B., Alessandro R.T. (2004) Cuticular hydrocarbons on elytra of the Diaprepes root weevil Diaprepes abbreviatus (L.) (Coleoptera: Curculionidae). Agricultural and Forest Entomology 6,251.

Lasch P., Boese M., Pacifico A., Diem M. (2002) FTIR spectroscopic investigations of single cells on the subcellular level. Vibrational Spectroscopy 28,147.

Lichter J.A., Van Vlietpa K.J., Rubner M.F. (2009) Design of antibacterial surfaces and interfaces: Polyelectrolyte multilayers as a multifunctional platform. Macromolecules 42,8573.

Liebig J., Peeters C., Oldham N.J., Markstädter C., Hölldobler B. (2000) Are variations in cuticular hydrocarbons of queens and workers a reliable signal of fertility in the ant Harpegnathos saltator? Proceedings of the National Academy of Sciences of the United States of America 97,4124.

Liu H., Zhai J., Jiang L. (2006) Wetting and anti-wetting on aligned carbon nanotube films. Soft Matter 2,811.

Liu K., Du J., Wu J., Jiang L. (2012) Superhydrophobic gecko feet with high adhesive forces towards water and their bio-inspired materials. Nanoscale 4,768.

Liu K., Yao X., Jiang L. (2010) Recent developments in bio-inspired special wettability. Chemical Society Reviews 39,3240.

Liu Y., Chen X., Xin J.H. (2008) Hydrophobic duck feathers and their simulation on textile substrates for water repellent treatment. Bioinspiration and Biomimetics 3.

Liu Y., Tay J.H. (2002) The essential role of hydrodynamic shear force in the formation of biofilm and granular sludge. Water Research 36,1653.

190

Lockey H.K. (1959) The thickness of some insect epicuticular wax layers. The Journal of Experimental Biology 37,316.

Lockey K.H. (1980) Insect cuticular hydrocarbons. Comparative Biochemistry and Physiology Part B, Biochemistry & Molecular Biology 65,457.

Lockey K.H. (1985) Insect cuticular lipids. Comparative Biochemistry and Physiology Part B, Biochemistry & Molecular Biology 81,263.

Lockey K.H. (1988) Lipids of the insect cuticle: origin, composition and function. Comparative Biochemistry and Physiology Part B, Biochemistry & Molecular Biology 89,595.

Loudet J.C., Alsayed A.M., Zhang J., Yodh A.G. (2005) Capillary interactions between anisotropic colloidal particles. Physical Review Letters 94.

Ma M., Hill R.M. (2006) Superhydrophobic surfaces. Current Opinion in Colloid and Interface Science 11,193.

Madigan M.T., Martinko J.M., Parker J. (2009) Brock biology of microorganisms. 12th ed. New Jersey, Prentice Hall, Peason Education).

Maness P.C., Smolinski S., Blake D.M., Huang Z., Wolfrum E.J., Jacoby W.A. (1999)

Bactericidal activity of photocatalytic TiO2 reaction: Toward an understanding of its killing mechanism. Applied and Environmental Microbiology 65,4094.

Mariod A.A., Abdelwahab S.I. (2012) Sclerocarya birrea (Marula), an African tree of nutritional and medicinal uses: A review. Food Reviews International 28,375.

Martin S., Drijfhout F. (2009) A review of ant cuticular hydrocarbons. Journal of Chemical Ecology 35,1151.

Mei H., Luo D., Guo P., Song C., Liu C., Zheng Y., Jiang L. (2011) Multi-level micro- /nanostructures of butterfly wings adapt at low temperature to water repellency. Soft Matter 7,10569.

Merk S., Blume A., Riederer M. (1998) Phase behaviour and crystallinity of plant cuticular waxes studied by Fourier transform infrared spectroscopy. Planta 204,44.

Milligan G., Parenti M., Magee A.I. (1995) The dynamic role of palmitoylation in signal transduction. Trends in Biochemical Sciences 20,181.

191

Moneti G., Dani F.R., Pieraccini G., Turillazzi S. (1997) Solid-phase microextraction of insect epicuticular hydrocarbons for gas chromatographic/mass spectrometric analysis. Rapid Communications in Mass Spectrometry 11,857.

Moreno E., Cordobilla R., Calvet T., Cuevas-Diarte M.A., Gbabode G., Negrier P., Mondieig D., Oonk H.A.J. (2007) Polymorphism of even saturated carboxylic acids from n-decanoic to n-eicosanoic acid. New Journal of Chemistry 31,947.

Moussian B. (2010) Recent advances in understanding mechanisms of insect cuticle differentiation. Insect Biochemistry and Molecular Biology 40,363.

Murphy R.C., Axelsen P.H. (2011) Mass spectrometric analysis of long-chain lipids. Mass Spectrometry Reviews 30,579.

Nakajima A., Hashimoto K., Watanabe T. (2001) Recent studies on super-hydrophobic films. Monatshefte fur Chemie 132,31.

National Research Infrastructure for Australia (NCRIS). (2014) Atlas of Living Australia. Accessed: 22 Dec 2014. Available from: http://www.ala.org.au/

Nečas D., Klapetek P. (2012) Gwyddion: An open-source software for SPM data analysis. Central European Journal of Physics 10,181.

Neinhuis C., Barthlott W. (1997) Characterization and distribution of water-repellent, self-cleaning plant surfaces. Annals of Botany 79,667.

Nelson D.R., Blomquist L.G. (1995) Insect waxes. In: Hamilton R.J., Christie W.W. Waxes: Chemistry, Molecular Biology and Functions. (The Oily Press Ltd Hamilton RJ, Dundee) pp:1.

Nelson D.R., Buckner J.S., Fatland C.L. (1994) The composition of external lipids from adult whiteflies, Bemisia tabaci and Trialeurodes vaporariorum. Comparative Biochemistry and Physiology Part B, Biochemistry & Molecular Biology 109,293.

Nelson D.R., Charlet L.D. (2003) Cuticular hydrocarbons of the sunflower beetle, Zygogramma exclamationis. Comparative Biochemistry and Physiology Part B, Biochemistry & Molecular Biology 135,273.

Nelson D.R., Fatland C.L., Buckner J.S., Freeman T.P. (1999) External lipids of adults of the giant whitefly, Aleurodicus dugesii. Comparative Biochemistry and Physiology Part B, Biochemistry & Molecular Biology 123,137.

192

Nelson D.R., Freeman T.P., Buckner J.S. (2000) Waxes and lipids associated with the external waxy structures of nymphs and pupae of the giant whitefly, Aleurodicus dugesii. Comparative Biochemistry and Physiology Part B, Biochemistry & Molecular Biology 125,265.

Nelson D.R., Freeman T.P., Buckner J.S., Hoelmer K.A., Jackson C.G., Hagler J.R. (2003) Characterization of the cuticular surface wax pores and the waxy particles of the dustywing, Semidalis flinti (Neuroptera: Coniopterygidae). Comparative Biochemistry and Physiology Part B, Biochemistry & Molecular Biology 136,343.

Nelson D.R., Guershon M., Gerling D. (1998) The surface wax composition of the exuviae and adults of Aleyrodes singularis. Comparative Biochemistry and Physiology Part B, Biochemistry & Molecular Biology 119,655.

Nelson D.R., Lee Jr R.E. (2004) Cuticular lipids and desiccation resistance in overwintering larvae of the goldenrod gall fly, Eurosta solidaginis (Diptera: Tephritidae). Comparative Biochemistry and Physiology Part B, Biochemistry & Molecular Biology 138,313.

Nelson D.R., Walker G.P., Buckner J.S., Fatland C.L. (1997) Composition of the wax particles and surface wax of adult whiteflies: Aleuroplatus coronata, Aeurothrixus floccosus, Aleurotithius timberlakei, Dialeurodes citri, Dialeurodes citrifolii, and Parabemisia myricae. Comparative Biochemistry and Physiology Part B, Biochemistry & Molecular Biology 117,241.

Neville A.C., Parry D.A., Woodhead-Galloway J. (1976) The chitin crystallite in arthropod cuticle. Journal of Cell Science 21,73.

Nishimoto S., Bhushan B. (2013) Bioinspired self-cleaning surfaces with superhydrophobicity, superoleophobicity, and superhydrophilicity. RSC Advances 3,671.

Nosonovsky M., Bhushan B. (2005) Roughness optimization for biomimetic superhydrophobic surfaces. Microsystem Technologies 11,535.

Nosonovsky M., Bhushan B. (2007a) Hierarchical roughness makes superhydrophobic states stable. Microelectronic Engineering 84,382.

193

Nosonovsky M., Bhushan B. (2007b) Multiscale friction mechanisms and hierarchical surfaces in nano- and bio-tribology. Materials Science & Engineering R, Reports 58,162.

Nosonovsky M., Bhushan B. (2008) Biologically inspired surfaces: Broadening the scope of roughness. Advanced Functional Materials 18,843.

O'Toole G., Kaplan H.B., Kolter R. (2000) Biofilm formation as microbial development. Annual Review of Microbiology 54,49.

Öner D., McCarthy T.J. (2000) Ultrahydrophobic surfaces. Effects of topography length scales on wettability. Langmuir 16,7777.

Ong Y.L., Razatos A., Georgiou G., Sharma M.M. (1999) Langmuir 15,2719.

Pacakova V., Feltl K. (1992) Chromatographic retention indices: An aid to the identification of organic compounds. (Ellis Horwood, London).

Papadopoulos P., Mammen L., Deng X., Vollmer D., Butt H.J. (2013) How superhydrophobicity breaks down. Proceedings of the National Academy of Sciences of the United States of America 110,3254.

Patankar N.A. (2004) Mimicking the lotus effect: Influence of double roughness structures and slender pillars. Langmuir 20,8209.

Peng W., Hu X., Zhang D. (2011) Bioinspired fabrication of magneto-optic hierarchical architecture by hydrothermal process from butterfly wing. Journal of Magnetism and Magnetic Materials 323,2064.

Perez Goodwyn P., Maezono Y., Hosoda N., Fujisaki K. (2009) Waterproof and translucent wings at the same time: Problems and solutions in butterflies. Naturwissenschaften 96,781.

Pogodin S., Hasan J., Baulin V.A., Webb H.K., Truong V.K., Nguyen T.H.P., Boshkovikj V., Fluke C.J., Watson G.S., Watson J.A., Crawford R.J., Ivanova E.P. (2013a) Biophysical model of bacterial cell interactions with nanopatterned cicada wing surfaces. Biophysical Journal 104,835.

Pogodin S., Hasan J., Baulin V.A., Webb H.K., Truong V.K., Phong Nguyen T.H., Boshkovikj V., Fluke C.J., Watson G.S., Watson J.A., Crawford R.J., Ivanova E.P.

194

(2013b) Biophysical model of bacterial cell interactions with nanopatterned cicada wing surfaces. Biophysical Journal 104,835.

Postgate J.R. (1969) Methods in microbiology. (Academic Press, New York) pp:611

Powell J.S., Raffa K.F. (1999) Effects of selected Larix laricina terpenoids on Lymantria dispar (Lepidoptera: Lymantriidae) development and behavior. Environmental Entomology 28,148.

Qi L., Xu Z., Jiang X., Hu C., Zou X. (2004) Preparation and antibacterial activity of chitosan nanoparticles. Carbohydrate Research 339,2693.

Rai M., Yadav A., Gade A. (2009) Silver nanoparticles as a new generation of antimicrobials. Biotechnology Advances 27,76.

Rajabi H., Moghadami M., Darvizeh A. (2011) Investigation of microstructure, natural frequencies and vibration modes of dragonfly wing. Journal of Bionic Engineering 8,165.

Ramsay J.A. (1935) The Evaporation of Water from the Cockroach. The Journal of Experimental Biology 12,373.

Riedel M., Eichner A., Jetter R. (2003) Slippery surfaces of carnivorous plants: Composition of epicuticular wax crystals in Nepenthes alata Blanco pitchers. Planta 218,87.

Rothstein J.P. (2010) Slip on superhydrophobic surfaces. Annual Review of Fluid Mechanics 42,89.

Royet J., Dziarski R. (2007) Peptidoglycan recognition proteins: Pleiotropic sensors and effectors of antimicrobial defences. Nature Reviews Microbiology 5,264.

Ryu J.H., Beuchat L.R. (2005) Biofilm formation and sporulation by Bacillus cereus on a stainless steel surface and subsequent resistance of vegetative cells and spores to chlorine, chlorine dioxide, and a peroxyacetic acid-based sanitizer. Journal of Food Protection 68,2614.

Sadtler Research Laboratories. (1984) The sadtler standard gas chromatography retention index library. (Sadtler Research Laboratories, Philadelphia, PA).

195

Saïd I., Costagliola G., Leoncini I., Rivault C. (2005) Cuticular hydrocarbon profiles and aggregation in four Periplaneta species (Insecta: Dictyoptera). Journal of Insect Physiology 51,995.

Sajomsang W., Gonil P. (2010) Preparation and characterization of α-chitin from cicada sloughs. Materials Science & Engineering C 30,357.

Samuels L., Kunst L., Jetter R. (2008) Sealing plant surfaces: Cuticular wax formation by epidermal cells. Annual Review of Plant Biology 59,683.

Sarrazin M., Vigneron J.P., Welch V., Rassart M. (2008) Nanomorphology of the blue iridescent wings of a giant tropical wasp Megascolia procer javanensis (Hymenoptera). Physical Review E, Statistical, Nonlinear, and Soft Matter Physics 78.

Savluchinske-Feio S., Nunes L., Pereira P.T., Silva A.M., Roseiro J.C., Gigante B., Marcelo Curto M.J. (2007) Activity of dehydroabietic acid derivatives against wood contaminant fungi. Journal of Microbiological Methods 70,465.

Scardino A.J., Zhang H., Cookson D.J., Lamb R.N., de Nys R. (2009) The role of nano- roughness in antifouling. Biofouling 25,757.

Schmidt S., Winter J., Gallert C. (2012) Long-term effects of antibiotics on the elimination of chemical oxygen demand, nitrification, and viable bacteria in laboratory-scale wastewater treatment plants. Archives of Environmental Contamination and Toxicology 63,354.

Schulte A.J., Droste D.M., Koch K., Barthlott W. (2011) Hierarchically structured superhydrophobic flowers with low hysteresis of the wild pansy (Viola tricolor) - New design principles for biomimetic materials. Beilstein Journal of Nanotechnology 2,228.

Schultz M.P., Bendick J.A., Holm E.R., Hertel W.M. (2011) Economic impact of biofouling on a naval surface ship. Biofouling 27,87.

Seo Y.S., Lee D.Y., Rayamahji N., Kang M.L., Yoo H.S. (2008) Biofilm-forming associated genotypic and phenotypic characteristics of Staphylococcus spp. isolated from animals and air. Research in Veterinary Science 85,433.

196

Shai Y. (2002) Mode of action of membrane active antimicrobial peptides. Biopolymers - Peptide Science Section 66,236.

Shevtsova E., Hansson C., Janzen D.H., Kjærandsen J. (2011) Stable structural color patterns displayed on transparent insect wings. Proceedings of the National Academy of Sciences of the United States of America 108,668.

Shirtcliffe N.J., McHale G., Atherton S., Newton M.I. (2010) An introduction to superhydrophobicity. Advances in Colloid and Interface Science 161,124.

Simões M., Simões L.C., Vieira M.J. (2009) Species association increases biofilm resistance to chemical and mechanical treatments. Water Research 43,229.

Simonsen T.J. (2001) The wing vestiture of the non-ditrysian Lepidoptera (Insecta). Comparative morphology and phylogenetic implications. Acta Zoologica 82,275.

Smith S. (1994) The animal fatty acid synthase: One gene, one polypeptide, seven enzymes. FASEB Journal 8,1248.

Soliday C.L., Blomquist G.J., Jackson L.L. (1974) Cuticular lipids of insects. VI. Cuticular lipids of the grasshoppers Melanopulus sanguinipes and Melanoplus packardii. Journal of Lipid Research 15,399.

Song F., Xiao K.W., Bai K., Bai Y.L. (2007) Microstructure and nanomechanical properties of the wing membrane of dragonfly. Materials Science & Engineering A 457,254.

Stamou D., Duschl C., Johannsmann D. (2000) Long-range attraction between colloidal spheres at the air-water interface: The consequence of an irregular meniscus. Physical Review E - Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics 62,5263.

Stark A.Y., Badge I., Wucinich N.A., Sullivan T.W., Niewiarowski P.H., Dhinojwala A. (2013) Surface wettability plays a significant role in gecko adhesion underwater. Proceedings of the National Academy of Sciences of the United States of America 110,6340.

Su Y., Ji B., Zhang K., Gao H., Huang Y., Hwang K. (2010) Nano to micro structural hierarchy is crucial for stable superhydrophobic and water-repellent surfaces. Langmuir 26,4984.

197

Sun M., Watson G.S., Zheng Y., Watson J.A., Liang A. (2009) Wetting properties on nanostructured surfaces of cicada wings. Journal of Experimental Biology 212,3148.

Sun T., Feng L., Gao X., Jiang L. (2005) Bioinspired surfaces with special wettability. Accounts of Chemical Research 38,644.

Suter R.B., Rosenberg O., Loeb S., Wildman H., Long Jr J.H. (1997) Locomotion on the water surface: Propulsive mechanisms of the fisher spider Dolomedes triton. Journal of Experimental Biology 200,2523.

Taherian F., Marcon V., Van Der Vegt N.F.A., Leroy F. (2013) What is the contact angle of water on graphene? Langmuir 29,1457.

Tapia A.A., Vallejo M.D., Gouiric S.C., Feresin G.E., Rossomando P.C., Bustos D.A. (1997) Hydroxylation of dehydroabietic acid by Fusarium species. Phytochemistry 46,131.

Thomas T.R. (1998) Rough surfaces. 2nd ed. ( Imperial College Press, London).

Tobin M.J., Puskar L., Hasan J., Webb H.K., Hirschmugl C.J., Nasse M.J., Gervinskas G., Juodkazis S., Watson G.S., Watson J.A., Crawford R.J., Ivanova E.P. (2013) High spatial resolution mapping of superhydrophobic cicada wing surface chemistry using infrared microspectroscopy and infrared imaging at two synchrotron beamlines. Journal of Synchrotron Radiation 20,482

Truong V.K., Lapovok R., Estrin Y.S., Rundell S., Wang J.Y., Fluke C.J., Crawford R.J., Ivanova E.P. (2010) The influence of nano-scale surface roughness on bacterial adhesion to ultrafine-grained titanium. Biomaterials 31,3674.

Truong V.K., Rundell S., Lapovok R., Estrin Y., Wang J.Y., Berndt C.C., Barnes D.G., Fluke C.J., Crawford R.J., Ivanova E.P. (2009) Effect of ultrafine-grained titanium surfaces on adhesion of bacteria. Applied Microbiology and Biotechnology 83,925.

Truong V.K., Webb H.K., Fadeeva E., Chichkov B.N., Wu A.H.F., Lamb R., Wang J.Y., Crawford R.J., Ivanova E.P. (2012) Air-directed attachment of coccoid bacteria to the surface of superhydrophobic lotus-like titanium. Biofouling 28,539.

198

Van Den Dool H., Kratz J.D. (1963) A generalization of the retention index system including linear temperature programmed gas-liquid partition chromatography. Journal of Chromatography A 11,463.

Van Dooremalen C., Ellers J. (2010) A moderate change in temperature induces changes in fatty acid composition of storage and membrane lipids in a soil arthropod. Journal of Insect Physiology 56,178.

Van Dyck S., Caulier G., Todesco M., Gerbaux P., Fournier I., Wisztorski M., Flammang P. (2011) The triterpene glycosides of Holothuria forskali: Usefulness and efficiency as a chemical defense mechanism against predatory fish. Journal of Experimental Biology 214,1347.

Van Oss C.J., Good R.J., Chaudhury M.K. (1988) Additive and nonadditive surface tension components and the interpretation of contact angles. Langmuir 4,884.

Vasudevan R., Kennedy A.J., Merritt M., Crocker F.H., Baney R.H. (2014) Microscale patterned surfaces reduce bacterial fouling-microscopic and theoretical analysis. Colloids and Surfaces B: Biointerfaces 117,225.

Vukusic P., Hooper I. (2005) Biomaterials: Directionally controlled flourescence emission in butterflies. Science 310,1151.

Vukusic P., Sambles J.R. (2003) Photonic structures in biology. Nature 424,852.

Vukusic P., Wootton R.J., Sambles J.R. (2004) Remarkable iridescence in the hindwings of the damselfly Neurobasis chinensis chinensis (Linnaeus) (Zygoptera: Calopterygidae). Proceedings of the Royal Society of London Series B, Biological Sciences 271,595.

Wagner C.D. (1978) X-ray photoelectron spectroscopy with x-ray photons of higher energy. Journal of Vacuum Science and Technology 15,518.

Wagner C.D., Gale L.H., Raymond R.H. (1979) Two-dimensional chemical state plots: A standardized data set for use in identifying chemical states by x-ray photoelectron spectroscopy. Analytical Chemistry 51,466.

Wagner T., Neinhuis C., Barthlott W. (1996) Wettability and contaminability of insect wings as a function of their surface sculptures. Acta Zoologica 77,213.

199

Wakeling J.M., Ellington C.P. (1997) Dragonfly flight. II. Velocities, accelerations and kinematics of flapping flight. Journal of Experimental Biology 200,557.

Wan Y.-l., Cong Q., Jin J.F., Wang X.-j. (2009) Microstructure and wettability of dragonfly wings. Journal of Jilin University (Engineering and Technology Edition) 39,732.

Wan Y.-l., Cong Q., Wang X.-j., Yan Z. (2008) The wettability and mechanism of geometric non-smooth structure of dragonfly wing surface. Journal of Bionic Engineering 5,40.

Wang Z.J. (2005) Dissecting insect flight. Annual Review of Fluid Mechanics 37,183.

Watson G.S., Cribb B.W., Watson J.A. (2010) How micro/nanoarchitecture facilitates anti-wetting: An elegant hierarchical design on the termite wing. Acs Nano 4,129.

Watson G.S., Cribb B.W., Watson J.A. (2011) Contrasting micro/nano architecture on termite wings: Two divergent strategies for optimising success of colonisation flights. PLoS ONE 6,Art. No. e24368.

Watson G.S., Myhra S., Cribb B.W., Watson J.A. (2008) Putative functions and functional efficiency of ordered cuticular nanoarrays on insect wings. Biophysical Journal 94,3352.

Watson G.S., Watson J.A., Hu S., Brown C.L., Cribb B.W., Myhra S. (2010) Micro and nanostructures found on insect wings - Designs for minimising adhesion and friction. International Journal of Nanomanufacturing 5,112.

Webb H.K., Boshkovikj V., Fluke C.J., Truong V.K., Hasan J., Baulin V.A., Lapovok R., Estrin Y., Crawford R.J., Ivanova E.P. (2013) Bacterial attachment on sub- nanometrically smooth titanium substrata. Biofouling 29,163.

Webb H.K., Truong V.K., Hasan J., Fluke C., Crawford R.J., Ivanova E.P. (2012) Roughness parameters for standard description of surface nanoarchitecture. Scanning.

Wenzel R.N. (1936) Resistance of solid surfaces to wetting by water. Industrial and Engineering Chemistry 28,988.

Wootton R.J. (1992) Functional morphology of insect wings. Annual Review of Entomology 37,113.

200

Wu A.H.F., Nakanishi K., Cho K.L., Lamb R. (2013) Diatom attachment inhibition: Limiting surface accessibility through air entrapment. Biointerphases 8,1.

Wu M.Y., Sendamangalam V., Xue Z., Seo Y. (2012) The influence of biofilm structure and total interaction energy on Escherichia coli retention by Pseudomonas aeruginosa biofilm. Biofouling 28,1119.

Xiao K., Bai K., Wang W., Song F. (2007) Experimental study on the microstructure and nanomechanical properties of the wing membrane of dragonfly. Acta Mechanica Sinica/Lixue Xuebao 23,281.

Yan Y.Y., Gao N., Barthlott W. (2011) Mimicking natural superhydrophobic surfaces and grasping the wetting process: A review on recent progress in preparing superhydrophobic surfaces. Advances in Colloid and Interface Science 169,80.

Yang S., Jin X., Liu K., Jiang L. (2013) Nanoparticles assembly-induced special wettability for bio-inspired materials. Particuology 11,361.

Yang T., Berber S., Liu J.F., Miller G.P., Tománek D. (2008) Self-assembly of long chain alkanes and their derivatives on graphite. Journal of Chemical Physics 128.

Yao X., Jericho M., Pink D., Beveridge T. (1999) Thickness and elasticity of gram- negative murein sacculi measured by atomic force microscopy. Journal of Bacteriology 181,6865.

Yasuyuki M., Kunihiro K., Kurissery S., Kanavillil N., Sato Y., Kikuchi Y. (2010) Antibacterial properties of nine pure metals: a laboratory study using Staphylococcus aureus and Escherichia coli. Biofouling 26,851.

Yin S., Wang C., Qiu X., Bo X., Bai C. (2001) Theoretical study of the effects of intermolecular interactions in self-assembled long-chain alkanes adsorbed on graphite surface. Surface and Interface Analysis 32,248.

Yocum G.D., Buckner J.S., Fatland C.L. (2011) A comparison of internal and external lipids of nondiapausing and diapause initiation phase adult Colorado potato beetles, Leptinotarsa decemlineata. Comparative Biochemistry and Physiology Part B, Biochemistry & Molecular Biology 159,163.

201

Yoshimitsu Z., Nakajima A., Watanabe T., Hashimoto K. (2002) Effects of surface structure on the hydrophobicity and sliding behavior of water droplets. Langmuir 18,5818.

Zeier J., Schreiber L. (1999) Fourier transform infrared-spectroscopic characterisation of isolated endodermal cell walls from plant roots: Chemical nature in relation to anatomical development. Planta 209,537.

Zhang X., Wang L., Levänen E. (2013) Superhydrophobic surfaces for the reduction of bacterial adhesion. RSC Advances 3,12003.

Zhang Z., Yang M.J., Pawliszyn J. (1994) Solid-phase microextraction. Analytical Chemistry 66,844A.

Zhao H.X., Yin Y.J., Zhong Z. (2011) Assembly modes of dragonfly wings. Microscopy Research and Technique 74,1134.

202