Structure, Properties and Applications Of

STRUCTURE, PROPERTIES AND APPLICATIONS OF

LAYERED MATERIALS:

MULTILAYERED FILMS AND AEROGEL COMPOSITES

MINGZE SUN

Submitted in partial fulfillment of the requirements

For the degree of Doctor of Philosophy

Dissertation Advisor: Prof. David Schiraldi

Department of Macromolecular Science and Engineering

CASE WESTERN RESERVE UNIVERSITY

January, 2018 CASE WESTERN RESERVE UNIVERSITY SCHOOL

OF GRADUATE STUDIES

We hereby approve the thesis/dissertation of

Mingze Sun

candidate for the degree of Doctor of Philosophy *.

Committee Chair

David Schiraldi

Committee Member

Eric Baer

Committee Member

Gary Wnek

Committee Member

Emily Pentzer

Date of Defense

September 19, 2017

*We also certify that written approval has been obtained

for any proprietary material contained therein.

Table of Contents

Title Page ...... i Committee Approval Sheet ...... ⅱ Copyright ...... iii Table of Contents ...... iⅴ List of Tables ...... iⅹ List of Figures ...... ⅺ Acknowledgement ...... xx Abstract ...... xⅻ

Chapter 1 ...... 1 Introduction ...... 1 1.1 Layered materials ...... 1 1.2 Multilayer films ...... 2 1.3 Clay aerogels ...... 4 1.4 Objectives ...... 6 1.5 References ...... 11 Chapter 2 ...... 16 Fabrication and modification of multilayer films: HDPE/EVOH ...... 16 2.1 Introduction ...... 16 2.1.1 Background ...... 16 2.1.2 Barrier improvement methods ...... 18 2.1.2.1 Layer-by-Layer ...... 19 2.1.2.2 Parylene ...... 21 2.1.2.3 Inorganic barrier materials ...... 22 2.1.2.4 Plasma Treatment ...... 26 2.1.3 Permeation principles and mechanism ...... 27 2.1.4 Permeability measurement ...... 32 2.1.4.1 Mocon ...... 32 2.1.4.2 Calcium corrosion test ...... 33

2.2 Experimental Section ...... 36 2.2.1 Multilayer film fabrication ...... 36 2.2.2 Modification of multilayer films ...... 37 2.2.2.1 Layer-by-layer deposition ...... 37 2.2.2.2 Parylene C deposition ...... 38 2.2.2.3 Inorganic barrier materials deposition ...... 40 2.2.2.4 Plasma Pre-treatment ...... 42 2.2.3 Barrier Performance Investigation ...... 42 2.2.4 Mechanical and Transparency Performance Investigation ...... 45 2.3 Multilayer Films’ performance ...... 47 2.3.1 Raw film...... 47 2.3.2 Barrier improvement performance ...... 52 2.3.3 Rolling Test ...... 63 2.3.4 Transparency Performance ...... 65 2.4 Summary ...... 67 2.5 Acknowledgement ...... 67 2.6 References ...... 68 Chapter 3 ...... 78 Integration of barrier films with organic photovoltaic devices ...... 78 3.1 Introduction ...... 79 3.1.1 Overview of organic solar cell packages ...... 79 3.1.2 Degradation mechanism of organic photovoltaic devices ...... 80 3.1.3 Overview of existing solutions and products ...... 82 3.1.4 Current lab work and project objective ...... 84 3.2 Experimental section ...... 90 3.2.1 Materials ...... 90 3.2.2 Device Fabrication ...... 91 3.2.3 Device Characterization ...... 94 3.2.4 Barrier Film Encapsulation ...... 95 3.3 Results and discussions ...... 97 3.3.1 Influence from encapsulation process ...... 97 3.3.2 Different recipes and substrates influence on performance of solar cell packages...... 100

3.3.3 The importance of barrier films (with or without films)...... 105 3.3.4 Different sealing skills ...... 107 3.3.5 Long-term test for the same packages under different conditions ...... 110 3.4 Summary ...... 110 3.5 Acknowledgement ...... 111 3.6 References ...... 113 Chapter 4 ...... 120 Effects of feather-fiber reinforcement on clay aerogel structure, property and applications ...... 120 4.1 Introduction ...... 120 4.2 Experimental Section ...... 122 4.2.1 Materials ...... 122 4.2.2 Hydrogel preparation ...... 123 4.2.3 Compression samples...... 123 4.2.4 Thermal conductive samples...... 123 4.2.5 Characterization ...... 124 4.2.6 Mechanical testing ...... 124 4.2.7 Surface area, porosity and pore size ...... 124 4.2.8 SEM and μCT ...... 126 4.2.9 Thermal conductivity test ...... 127 4.3 Results and discussion ...... 130 4.3.1 Different types of fibers and fiber amounts ...... 130 4.3.2 Different freezing conditions and influence of morphology...... 135 4.3.3 Thermal insulation performance ...... 139 4.4 Conclusion ...... 141 4.5 Acknowledgement ...... 142 4.6 References ...... 143 Chapter 5 ...... 147 The relation between rheological properties of gels and mechanical properties of aerogels...... 147 5.1 Introduction ...... 148 5.1.1 Materials optimization Loop ...... 148 5.1.2 Rheological properties ...... 150

5.1.3 Objectives ...... 152 5.2 Experimental Section ...... 152 5.2.1 Materials ...... 152 5.2.2 Hydrogel preparation ...... 153 5.2.3 Viscosity testing ...... 153 5.2.4 Aerogel formation ...... 154 5.2.5 Mechanical testing ...... 154 5.2.6 SEM and Con-focal...... 155 5.3 Results and discussion ...... 156 5.3.1 Different processing rate (stirring speed) effect ...... 156 5.3.2 Different standing time effect ...... 158 5.3.3 Different molecular weights effect ...... 161 5.4 Summary ...... 164 5.5 Acknowledgement ...... 165 5.6 References ...... 166 Chapter 6 ...... 168 Reactive extrusion of bio-based paint material: synthesis part ...... 168 6.1 Introduction ...... 168 6.2 Experimental Section ...... 170 6.2.1 Materials ...... 170 6.2.2 Reaction Procedure ...... 172 6.2.3 Material Characterization...... 173 6.3 Results and discussion ...... 175 6.3.1 FT-IR characterization ...... 175 6.3.2 Quantitative solubility test ...... 176 6.3.3 Different methods to pre-treat soy flour ...... 177 6.3.4 Additional trial for a coating based on hot-press procedure ...... 179 6.4 Summary ...... 181 6.5 Acknowledgement ...... 182 6.6 References ...... 183 Chapter 7 ...... 185 Conclusions and Future work ...... 185

vii

7.1 Conclusions ...... 185 7.2 Future work ...... 187 7.2.1 Multilayer barrier films ...... 187 7.2.2 Aerogel composites ...... 188 7.2.3 Bio-based paint filler...... 188 7.3 References ...... 190 Bibliography ...... 191

viii

List of Tables

Table 2.1 Permeability of polymers commonly used in packaging [6].……….………....18

Table 2.2 Summary of ALD Al2O3 layer deposition conditions………………...………41

Table 2.3 Introduction about two systems of multilayer films used for barrier projects…48

Table 2.4 Control film’s barrier properties……………………………………………...49

Table 2.5 Barrier data of a series of multilayer films with different ratios of

HDPE/EVOH…………………………………………………………………………….50

Table 2.6 Barrier data about various multilayer films modified by Layer-by-

Layer…………………………………………………………………………..………....54

Table 2.7 Water Vapor Barrier improvement of 4 mil multilayer films by Parylene C with or without pre-treatment………………...……………………………………………….58

Table 2.8 Water Vapor Battier improvement on a series of multilayer films by a method combined Al2O3 and Parylene C…………………………………………………………61

Table 2.9 Barrier Data of modified films before and after rolling tests at certain times

(default=100 times)………………………………………………………………………64

Table 3.1 Performance parameters of solar cells before and after UV curing or Heat @ 90 ℃ in air for different periods………………………………..……………..………………..99

Table 4.1 Comparison on mechanical strength (compressive modulus) of different fibers reinforced aerogels……………………………………………………………………...131

Table 4.2 Comparison on mechanical strength of different fibers reinforced aerogels

(incorporated with 2.5% PVOH)……………………………………………………….132

Table 4.3 Mechanical strength of different fibers reinforced aerogels frozen at different temperatures. (incorporated with 5% Clay and 2.5% PVOH)…………………………..136

Table 4.4 Surface Area and porosity of different fibers reinforced aerogels frozen at different temperatures. (incorporated with 5% Clay and 2.5% PVOH)……….……….138

Table 4.5 Thermal conductivity value of different fibers reinforced aerogels frozen at different conditions. (incorporated with 5% Clay and 2.5% PVOH)…………….…….140

Table 6.1 Quantitative solubility tests based on yield products mixed with various amounts of clay…………………………………..…………..…………………………..……....176

Table 6.2 A quantitative solubility tests for synthesized products made from regular soy protein, acid-leached soy protein and alkaline-leached soy protein……………..……..178

List of Figures

Figure 1.1 Illustration of multilayer co-extrusion process showing how to fabricate a two- component film system with 65 core layers and surface layers…………………...……....3

Figure 1.2 Processing principle of co-extrusion for a two-component multilayer film system. Adapted from reference [12] ...……..………………………………………...…...4

Figure 1.3 Illustration of the structure of montmorillonite clay [24]……………………….5

Figure 1.4 SEM images of aerogel frozen at different temperatures to show an organized lamellar structure [25] …………………………………………………...………………….6

Figure 1.5 Illustration to exhibit the study strategy to build a “bridge” correlating the rheology of gel with the structure and properties of its product, aerogel…………………………………………………………………………….….……9

Figure 2.1 Oxygen permeability VS ratio of fractional free volume (FFV)/cohesive energy density (CED) for some widely used commercial polymers [4] ………...………….…….17

Figure 2.2 Water Vapor Transmission Rate (WVTR) requirements for different applications [7]………………………………………………………………………….....19

Figure 2.3 Layer-by-layer process illustration (a) and nano brick wall structure of 5 QLs on film substrate: MMT (red), PEI (blue) and PAA(green) (b) [13]………………...…….21

Figure 2.4 Illustration of E-beam Vapor Deposition process [26]……………………...…24

Figure 2.5 Illustration of ALD process based on two precursors: TMA and DI-water

[31]………………………………………………………………………………………...25

Figure 2.6 Illustration to show the path of gas transmission in a film system………….………………………………………………………………………...27

Figure 2.7 Picture shows MOCON PERMATRAN-W® Model 3/33 unit used for WVTR test (a) and illustration to explain the Mocon set-up work flow briefly, using water vapor as the permeant gas (b) [40] …………………….…………………………………………33

Figure 2.8 Illustration of Top view (a) and Front view (b) about Ca sensor device fabrication and Ca devices with barrier films before (c) and after (d) oxidation by the permeated water vapor…………………………………………………………………...34

Figure 2.9 (a) Setup for the co-extrusion of the multilayer A-B-A-B system. (b) Schematic of a cross-section of 4 mil multilayer films consisting of 65 core layers of EVOH/HDPE and two HDPE skin layers………………………………………..………………….…...37

Figure 2.10 Flow of Parylene C vapor deposition process………………………..……..39

Figure 2.11 Experimental results of parylene C film thickness with parylene dimer weight

[16]………………………………………………………………...………………………39

Figure 2.12 Relationship and derived equation of purge time with deposition temperature………………………………………………………………………………42

xii

Figure 2.13 Illustration: Top view of Ca test devices (a) and a real Ca test device (b); A home-made chamber creating a sealed environment with RH~99% (c) An electrical conductance measurement set-up (d)………………………………………………...…..45

Figure 2.14 A home-made rolling test setup composed by two rolling motors…………46

Figure 2.15 Barrier data (OTR & WVTR) for films with different EVOH/HDPE ratios…………………………………………………………………………………...…51

Figure 2.16 DSC data about 1 mil 65 Layers multilayer film with different ratios. Degrees of crystallinity are 67%, 62%, 59%, 58%, 60% and 62% (from top to down)…….……..51

Figure 2.17 AFM image to illustrate multilayer films’ A-B-A-B-A structure…….…….52

Figure 2.18 Illustration of the clay nano-brick wall structure on target film…………….54

Figure 2.19 Contact angle measurement for unmodified multilayer films (a) and films after

Ar plasma treatment for 10s (b), 60s (c), and 600s (d)…………………………...... 56

Figure 2.20 Schemes to display HDPE crosslink mechanism by Ar plasma……….…....57

Figure 2.21 AFM image to display a conformal and pin-hole free 2 μm Parylene C deposition on HDPE surface layer of a 4 mil multilayer film…………………………...57

Figure 2.22 Mocon curves about two films with (blue) and without (red) 30nm Al2O3 deposition...... 59

Figure 2.23 A stimulated structure to illustrate to combine organic and inorganic materials to fabricate a tortuous gas permeation path [37]…………………………………….……...59

xiii

Figure 2.24 AFM images comparison on deposition of Al2O3 by E-beam (a) and ALD (b).

Same recipe applies on multilayer films: Film/Al2O3/Parylene/Al2O3/Parylene…………62

Figure 2.25 A representative time dependent conductance data of Ca sensor. Effective

WVTR and Lag time can be determined……………………………………...…………62

Figure 2.26 Visible light transparency as a function of wavelength for a series of films with different barrier improvement methods…………………………………………….66

Figure 3.1 An increasing forecast production capacity (a) and an increasing efficiency of

OPV can make a boost for OPV industry and market development [7]…………………..80

Figure 3.2 Illustration of a ITO-free solar cell device’ s structure (a) and a picture to display the real device………………………………………………………………………….…82

Figure 3.3 A scheme to show lamination and lacquering machine: (1) unwinder (2) anilox roller, (3) roller coating system, (4) thermal drying unit, (5) UV drying unit, (6) laminating unit, (7) corona treatment, (8) unwinder, (9) rewinder [18]..………………………………83

Figure 3.4 Some commercial flexible solar cell panels or tape: (a) Seeed® 1 W thin-film flexible solar panel (b)PowerFilm® 7 W rollable solar charger (c) Uni-solar ® 136 W flexible solar panel. (d) InfinityPV® 4 Watt Rollable solar foils and tapes [19-22]..……...84

Figure 3.5 A standard BHJ solar cell structure to show a sandwich architecture where active layer is inserted into the middle of the whole device. And an illustration image and

TEM image on the left exhibit the work principle and morphology of BHJ layer. Adapted from Reference [26] ……………………………………………………………………...86

xiv

Figure 3.6 The mechanism of organic solar cells. (1) Photoexcitation to generate excitons;

(2) Diffusion of excitons; (3) Dissociation of excitons; (4) Free charge transferred and collected by electrode [32]……………..………………………………………………….88

Figure 3.7 A characteristic IV curve used to illustrated the key parameters of organic solar cell[32]. This curve is measured from a real solar cell and obtained from IV station…………………………………………………………………………………….90

Figure 3.8 Pre-treatment process on ITO-PET substrates has been illustrated in these figures……………………………………………………………………………………94

Figure 3.9 Illustration to show two methods used to seal the solar cell devices: cured epoxy sealing (a) and EVA film lamination sealing (b) and (c). An EVA film has been pre-cut to fit the device and film size shown as (b)………………………………………………....97

Figure 3.10 Data about VOC (a), FF(b), JSC(c), PCE(d) and efficiency life (%) (e) of two type of organic solar cells vs storage time at the same condition: storing in dark, in the air…………………………………………………………………………………….....103

Figure 3.11 Power conversion efficiency of PET flexible solar cells from different batches.

(a). A 2.6-3 times difference on sheet resistance of two type substrates, from data sheet and

4-point measurements (b)……………………………………………...……..………...104

Figure 3.12 Efficiency life as a function of storing time for organic solar cells with different structures encapsulated with or without unmodified 4 mil film….……….…..106

Figure 3.13 Efficiency life as a function of storing time for inverted organic solar cells encapsulated with three different 4 mil films…………………………………………..107

Figure 3.14 Similar organic solar cell packages sealed by EVA (red) and UV-cured epoxy

(blue) had a different power conversion efficiency performance after a 10-day critical test

(a). Solar cell packages sealed by epoxy had an obvious color change.………………...109

Figure 3.15 Efficiency life as a function of storing time for inverted organic solar cells encapsulated by a 4 mil film, then stored at different conditions………………….……112

Figure 4.1 Schematic of the heat flow meter setup for thermal conductivity test….…..127

Figure 4.2 Picture of the heat flow meter setup for thermal conductivity test………….128

Figure 4.3 Compressive stress-strain curves for 5% clay aerogel incorporating varying amounts of keratin fibers Comparison on mechanical strength (compressive modulus) of different fibers reinforced aerogels…………………………………………...………....130

Figure 4.4 Examples to display inflection points or plateau forms in 5% Clay/2.5%PVOH aerogels with varied amount of soy silk fibers (top) and in 5% Clay aerogel with 1% hemp fibers (bottom) [25]…...………………………………………………………………….133

Figure 4.5 Stress-strain curves for 5% clay/2.5% PVOH aerogel incorporated with varied amount of keratin fibers……………………………………………………………...…134

Figure 4.6 Morphology comparison of different fibers: (a) keratin fiber; (b) Hemp fiber;

xvi

Figure 4.7 Comparison of a 5% Clay/2.5% PVOH aerogel (a) and a 5%Clay/2.5%

PVOH/1% keratin fibers dispersed randomly (b) and in an oriented manner ……..…..135

Figure 4.8 a) Lamellar and granular inner structure of 5%Clay 2.5% PVOH aerogels without fibers, frozen at -78 ℃ (a) and -178 ℃ (c). 10 times zoom in to exhibit details of each sample (b) and (d)……………………………………………………………..…..137

Figure 4.9 Inner structure of 5%Clay 2.5% PVOH aerogels with 1% keratin-fibers adding, frozen at a) -78 ℃ and b) -178 ℃; with 3% keratin-fibers adding, frozen at c) -78 ℃ and d) -178 ℃……………………………………………………………..…………….…..139

Figure 4.10 Keratin fibers dispersion in 5%Clay/2.5%PVOH incorporated with 3% fibers, conducted by μCT scanning…………………………………………………………….141

Figure 5.1 The strategy to study material science: Materials optimization loop………148

Figure 5.2 Water three phase diagram to display some common routes: freeze-drying, conventional drying, freezing, etc[4]………………………………….…………………149

Figure 5.3 Oscillatory frequency sweep measurement on a representative hydrogel system: complex viscosity, elastic modulus and loss modulus vs frequency……...……………..152

Figure 5.4 Illustration of aerogel fabrication and rheology test………………………..155

Figure 5.5 Viscosity of hydrogels (a), compressive moduli of respective aerogels (b), density vs. stirring speeds (c), specific modulus vs. stirring speeds (d)……………..…157

xvii

Figure 5.6 SEM images to show the difference inner structure morphology of aerogels made from same composition: 5%Alginate/5%Clay/5%PVOH with different stirring rate,

350 rpm (a) and 1000 rpm (b)……………………………………………………..……158

Figure 5.7 Graphs to display the related properties of hydrogels storing for a various standing time. Viscosity (a), mechanical performance (b), different density (c) increasing trend of specific modulus (d)………………………………………………….………..159

Figure 5.8 Confocal microscopy images to show different morphology of aerogels made from hydrogel with various standing time: 0 day, freeze-immediately (a) and 3-day-later freezing (b)……………………………………………………………….…………….160

Figure 5.9 A confocal 3D image to show a 5%Clay 2.5%PVOH aerogel’s internal structure………………………………………………………………………………...161

Figure 5.10 Graphs to exhibit the viscosity properties of same composition hydrogels prepared with various Mw PVOH solution as well as mechanical performance of respective aerogel products………………………………………………..…………………….…162

Figure 5.11 A series of SEM images as well as layer thickness and spacing information exhibited the influence of hydrogel viscosity based on different Mw PVOH………….164

Figure 6.1 Scheme to describe a reaction that performic acid migrates into the soybean oil and reacts to generate epoxidized soybean oil. Adapted from reference [5]….………...169

Figure 6.2 Scheme of an open-ring reaction between gliadin protein and epoxidized soybean oil………………..…………………………………………………………….170

xviii

Figure 6.3 Acid-leached powder (left) and alkaline-leached powder (right) after freeze-dry process…………………………………...…………………………………………...…172

Figure 6.4 Images to show the reaction setup (a) and product which is ready for characterization or further extrusion as filler…………………………………..……….173

Figure 6.5 Illustration to display the procedure of a quantitative solubility test……….174

Figure 6.6 the FT-IR spectra of the reaction products of epoxidized soybean oil with soy protein (red) and the control group, pure epoxidized soybean oil (black)……………....175

Figure 6.7 Scheme to illustrate a proposed mechanism about a reaction between maleic anhydride (MA), polyethylenimine (PEI) and soy proteins with amino groups (SP-NH2).

Adapted from reference [12] …………………………………………………………….180

Figure 6.8 Figures to display the procedure to synthesize the product contained “MA-PEI-

SP”, fabrication of product (a) and (b); samples with or without hot-press curing (c); Solve the mixture into acetone (left) and xylene (right) (d)……………………………….…181

Figure 7.1 Trials composed soybean flour, epoxidized soy bean oil and Pebax® elastomer have been extruded together(a); the product stripe showed an impressive tensile strength

(b)…………………………………………………………………………………….....189

xix

Acknowledgement

When I start to write the following words, a lot of names come to my mind, as well as lovely smiling faces and warm hearts…

First of all, this thesis and related work, cannot be done well without the endless support and encouragement from my academic dad, Professor David Schiraldi. I would like to express my sincere gratitude to him for guidance in the research and care in my oversea life. I am so luck and honored to be your students, really proud of that.

I would also like to acknowledge my gratitude to my thesis committee members:

Professor Eric Baer for his guidance and important discussions on barrier films, as well as tips and advice on professional communication style; Professor Gary Wnek for his expanded view on polymer development and inspired thoughts; and Professor Emily

Pentzer for her valuable time and kindly cooperation, as well as solid background instruction.

I would like to thank Dr. Ina Martin and Nichole Hoven for their assistance and training of using facilities of the MORE Center and MFL center at CWRU.

I appreciate help from the past and present members in our research group for helping me in lab and in life. Thanks to Hongbing Chen for being a guide of lab and life when I first came to CWRU; to Hua Sun for her inspiration and support and to Yuxin Wang for her motivation and encouragement. I also want to thank Rocky Viggiano, Matt Herbert,

Taneisha Deans, Kimberly Degracia, Rakim Tyler, Mark Holland, who help me a lot on my project and my life. It’s so warm to live in such a big family.

I would also like to thank all my friends at CWRU to help me go through difficult periods and share the happiness and sorrows. Special notes of thanks go to Xueliu Fan and Cong Zhang, for a great friendship never fading; to Jun Zhang, for the brotherhood I can count on; to Bin Liu and Sandra Pejic, for helping me widen the view in Physics and

Chemistry; to Sunsheng Zhu and Ci Zhang, for sharing happiness and hardship together; to the rest of friends for their support and friendship. You all make me stronger.

My deepest appreciation goes to my families for their love. Thanks to my parents and grandma, loving me, helping me and supporting me unconditionally, which makes me become the unique individual that I am. Thanks to my girlfriend, Danni. Thank you for building a new me you will be proud of.

At last, I am grateful for having such a great time here. It’s an irreplaceable treasure in my life, having so many friendly people, amazing tours, enriched experiences and fantasy stories. It goes away with leaving campus, but stays always in my heart.

xxi

Structure, Properties and Applications of Layered Materials:

Multilayered Films and Aerogel Composites

Abstract

MINGZE SUN

Materials with layered structures are of increased importance in current the disciplines of solid state chemistry, physics and engineering, due to their integration of single layers and flexibility in design and process. Many current technologies utilize layered materials to realize functions, such as barrier and layered multifunctional foam-like films. Two such classes of layered materials are studied in this present work: multilayered polymer barrier films and layered clay aerogel composites.

Multilayered barrier films

A polymer-based multilayer film system and related modified versions have prepared in this part of the thesis. These materials were designed with superior barrier properties, and can be used for encapsulation of organic photovoltaic (PV) devices. Such encapsulation is important because the active elements of the PV devices are easy to degrade and lose function when exposed to oxygen and water vapor. In addition to barrier properties,

xxii integrated with flexible organic PV devices have the requirements of flexibility and transparency. Multilayered films with “inorganic-organic” external coatings exhibit superior barrier properties as well as stable mechanical performance and 90+% transparency to the UV-vis spectrum. Applying those barrier films over the organic solar cells can elongate the shelf time of devices from several days to 3 months or more.

Aerogel composite systems

Clay/PVOH aerogels fabricated via freeze-drying, are ultra-low density and foam-like materials. Reinforcement by keratin-fibers can further improve their mechanical performance as well as insulation properties, not only from adding materials, but also from inner morphology change which can be observed through SEM and µCT. Those structural changes which make an impact on properties come from processing condition changes. A study has been conducted to determine the relationship between rheological properties of gels under different processing conditions and the performance of the final aerogel products made from those gels. A link, correlating rheology with final properties, may make it possible to predict aerogels’ performance before freeze drying.

xxiii

Chapter 1

Introduction

1.1 Layered Materials

Materials possessing layered structures have been extensively investigated in recent years, impacting the disciplines of solid state chemistry, physics and engineering.

Layered materials were initially used as the test media of intercalation chemistry where chemists could conduct sequential steps to reach kinetically stable products; researchers in physics utilized layered compounds to study novel phenomenon made more apparent by reduced dimensionality[1]. Many technological areas now utilize layered materials to realize properties, such as gas barrier or optical data storage[2][3].

Layered materials, with their special structures, consisting of stacked sheets, provide a significant opportunity to develop new type of materials with a tunable accessibility to the sites and properties, high surface areas and controllable pore sizes[4]. Because of those unique structural advantages, layered materials have a higher chance than single component polymer materials with bulk structure to satisfy the increasing complex requirement for smart and multifunction application such as multilayer capacitors, adhesives, coating, dielectric materials, structural composites and food packaging applications[5-9]; they have a wide level of dimensionality starting from the molecular- to the macro-level fabrication creating wide options to combine different materials together physically or chemically[10]. Considering this wide definition, the two main parts in this work: multilayered polymer films, and layered clay aerogel composites can be both

1 counted as belonging to this family of materials, due to their unique layered structures and morphologies. Bio-based paint in Chapter 6, from technical view, can be regard as a layered coating once it applies on the substrate. The properties of layered materials introduced above, utilizing chemical bonding and/or physical interactions to fabricate as an entire product with multifunction, will be studied in detail.

1.2 Multilayer films

Multilayer co-extrusion technology is an important path to produce polymer films with multilayer structure in a single continuous processing step[11]. Two or more polymers are melted separately and joined, extruded together through a die to form a single structure with multiple layers (see Figure 1.1)[12]. Utilizing multipliers to increase the inner layers exponentially, this technology can be used to fabricate polymer films with as many as thousands of layers which individual layer thickness is as small as a few nanometers[13].

Figure 1.2 shows a detailed processing principle of co-extrusion for a two-component multilayered film system: Two different polymer melts, A and B, are combined together in the feedblock as an entire melt flow with 2 layers (A-B), this flow goes through several layer-multiplying die elements[14]; each element splits the flow vertically, spreads the two separated flow horizontally and recombines together to form a flow with 4 layers (A-B-

A-B); then this recombined flow goes through the second layer-multiplying die element to repeat the same process to form a new flow with 8 layers. By that analogy, the product film can possess 2n+1 individual layers if applied n die elements. If the total thickness of film has been set as a constant, the reduction in layer thickness is achieved with more layers formed.

As introduced previously, the individual layer thickness in a multilayer film system can be varied from several nanometers to micrometers. By varying processing conditions, the individual layer thickness is manipulated, which result in the morphology change: polymer spherulites become more confined and orientated as the inner layer thickness decrease. Processing conditions make an impact on films’ inner structure, and reflects from property performance: multilayer films with more confined and orientated spherulites in each individual layer displayed the best gas barrier property[15]. In addition to applications in gas barrier, related studies of multilayer films have also been conducted and applied on lenses and in the data storage field due to their unique optical and dielectric properties[8][16]. This microlayering technology has been constantly improved over decades and benefit from the inspiring work and pioneering efforts of Professors

Eric Baer and Anne Hiltner[5,8.13-16].

Figure 1.1 Illustration of multilayer co-extrusion process showing how to fabricate a

two-component film system with 65 core layers and surface layers.

Figure 1.2 Processing principle of co-extrusion for a two-component multilayered film

system. Adapted from reference[12].

1.3 Clay aerogels

Aerogels are porous materials made from a gel, where the liquid within replaced by air without destroying its coherent network structures[17]. These materials are ultralight solids possessing low densities and low thermal conductivities, due to most of volume in aerogels is occupied by air. First described by Kissler, the removal of volatile organic solvents from gels formed the first-generation of these low-density porous, foam-like materials[20-22]. The use of a freeze drying process method was introduced by Mackenzie and Call, in the 1950s to fabricate montmorillonite clay aerogels, in turn leading to the wide-spread development of polymer/Clay composite aerogels in the 2000s. Clay aerogels, based on montmorillonite (MMT) clay, combined with polymers, belong to a family of layered materials with properties similar to those of conventional foams. Clay minerals consist of layers which have partial negative charges on the faces and partial positive charges on the edge, which provide then chance to form a linked edge to face structure (house of cards) due to opposite charges (see Figure 1.3)[23][24].

Figure 1.3 Illustration of the structure of montmorillonite clay[24]

Utilizing this inherent property, clay aerogels exhibit well-organized lamellar structured with pores, as a response to the change of freezing rate and concentration (see

Figure 1.4)[25][26]. Similar lamellar structure and granular layer structure of clay aerogel are proposed and will be observed through SEM in Chapter 4. Because of their layered and porous structures, clay aerogel can be used as thermal insulation and as absorbent compositions[27-30]. Before converting the lab results into commercial products, clay aerogels need to improve their mechanical strengths, due to their inherent brittleness in most cases. Incorporating with polymers is one of the methods to realize the mechanical improvement. Another method is adding a second or third filler to improve the final product strength. A wide range of inspiring results have been published over the past decade, showing structural cooperation of clay with poly(vinyl alcohol), cellulose, and even carbon nanotubes to substantially improve mechanical properties[31-35]. In Chapter 4, a new method which uses bio-based keratin fibers to improve mechanical properties as well as thermal insulation properties will be discussed in detail.

Figure 1.4 SEM images of aerogel frozen at different temperatures to show an

organized lamellar structure[25].

1.4 Objectives

In this thesis, we focus on the study of structure, properties and applications of layered materials. Two main parts will be separately presented in detail: (1) Part Ⅰ: multilayered barrier films used for flexible organic electronic devices, including Chapter 2 and

Chapter 3; and (2) Part Ⅱ: clay aerogel composites reinforced by fibers and the relation between aerogel properties vs hydrogel rheological properties, including Chapter 4 and

Chapter 5. The last Chapter 6 will independently introduce a synthesis work about soy- based paint, which can be used as a filler for extrusion with polymer carrier.

Part Ⅰ: multilayered barrier films used for flexible organic electronic devices

In Chapter 2, a detailed introduction about a polymer based multilayer film and related modified version will be given. The raw and modified films were designed to encapsulate organic photovoltaic devices in order to increase their lifetimes and maintain their performance, through protecting core devices away from oxygen and water vapor

6 degradation. In addition to superior barrier properties, the films under study in this work also need to be durable during frequent use and transparent to absorb visible light efficiently. The barrier performance of the subject materials was testing using a commercial

Mocon permatran, and using a home-made Ca-corrosion-test setup. The durability of the films was tested using a multiple-pass rolling test then verified the property through barrier changes. A UV-vis spectrometer has been utilized to measure the film’s transparency. Once we confirmed that we have the ability to fabricate superior barrier films with durable mechanical and transparency properties, the films were integrated with organic photovoltaic devices to evaluate their performance.

In Chapter 3, the multilayer barrier films studied in Chapter 2 were used to encapsulate organic solar cells, in order to evaluate film’s performance in a real application. A variety of 4 mil films were used due to their inherent low transmission rate to oxygen and water vapor. A relation between barrier films and organic solar cells’ lifetime performance was the expected objective of this part of the project. The core objective in this chapter therefore is to evaluate organic solar cells’ performance and their lifetimes under different conditions. The encapsulated solar cells, we called as packages, have been set in different conditions, with varied humidity level and UV-irradiation level. The device performance would be tested periodically and calculated from efficiency life (%), a relative change in solar cells’ power conversion efficiency (PCE) with time. This could be a classic application case about the path “design-processing-property-application” of multilayered polymer films.

Part Ⅱ: clay aerogel composites

In Chapter 4, a detailed clay aerogel system, reinforced by keratin-fibers, will be introduced and analyzed in detail. A series of low density, highly porous clay/poly(vinyl alcohol) composite aerogels reinforced by keratin fibers, were fabricated via a convenient and eco-friendly freeze drying method. The introduction of keratin fibers improved the mechanical performance, as well as decreased thermal conductivities, allowing these materials to have potential utility in thermal insulation. Different types of bio-fibers have been compared to confirm the unique and important role of keratin fibers; various processing conditions were applied in order to study processing/structure/property relationships in these materials. Compression tests and thermal conductivity tests were used to confirm the mechanical properties and insulation properties, respectively. And

SEM observations and µCT scanning suggest the change of inner structure of aerogel, when incorporated with keratin fibers, which can be used to explain the reasons of property change in mechanical and insulation.

In Chapter 5, the relationship between rheology property of gels, and the mechanical properties of aerogel fabricated from the precursor gel above is explored. Clay, poly(vinyl alcohol) and ammonium alginate have been used to prepare a series of the colloidal suspension under different conditions then freeze-dry to obtain target products.

We have verified that different compositions and conditions applied on those colloidal suspension systems may bring different layer structures morphology of aerogel products then different mechanical performance. These phenomena could be observed using scanning electron microscopy and confocal microscopy, then re-confirmed from mechanical tests. Emphases are placed on exploring the different factors in rheology and collecting sufficient sample evidence to support our idea. It is expected to establish a

“bridge” correlating the rheology with the structure and properties and realize making a preliminary forecast before converting suspension into product (see Figure 1.5).

Figure 1.5 Illustration to exhibit the study strategy to build a “bridge” correlating the

rheology of gel with the structure and properties of its product, aerogel.

Part Ⅲ: Reactive extrusion of bio-based paint material (synthesis work)

Chapter 6 introduces a corporate-sponsored project related to a new bio-based paint material. This project contains two major parts which were carried out in two research groups within the Department: (1) filler materials synthesis (mainly talked in the present chapter) and (2) extrusion with polymer carrier materials. The present work focuses on the synthesis work on reaction between epoxy soybean oil and soy protein, to synthesize a filler product. FT-IR was used to verify the reaction and a quantitative solubility test was applied to see if the product can dissolve in common organic solvents, acetone and xylene, which are frequently used in paints. Different pre-treatment conditions for the raw materials have been examined to evaluate their influences on final product quality. A

9 extra recipe trial which used as an adhesive coating at hot-press condition has shown as an extensive study.

1.5 References

[1]. Johnson D, Clarke S, Koumoto K. Layered materials. Semiconductor Science and

Technology. 2014 Jun;29(6):060301.

[2]. Liu RY, Schiraldi DA, Hiltner A, Baer E. Oxygen‐barrier properties of cold‐

drawn polyesters. Journal of Polymer Science Part B: Polymer Physics. 2002

May 1;40(9):862-877.

[3]. Ryan C, Christenson CW, Valle B, Saini A, Lott J, Johnson J, Schiraldi D, Weder

C, Baer E, Singer KD, Shan J. Roll‐to‐Roll Fabrication of Multilayer Films for

High Capacity Optical Data Storage. Advanced Materials. 2012 Oct

2;24(38):5222-5226.

[4]. Centi G, Perathoner S. Catalysis by layered materials: A review. Microporous

and mesoporous materials. 2008 Jan 1;107(1):3-15.

[5]. Zhou Z, Carr J, Mackey M, Yin K, Schuele D, Zhu L, Baer E. Interphase/interface

modification on the dielectric properties of polycarbonate/poly (vinylidene

fluoride‐ co ‐ hexafluoropropylene) multilayer films for high ‐energy density

capacitors. Journal of Polymer Science Part B: Polymer Physics. 2013 Jun

15;51(12):978-991.

[6]. Carpenter, Diane R., et al. "Adhesive-backed breathable layered materials." U.S.

Patent No. 5,277,954. 11 Jan. 1994.

[7]. Yamada, Satoru, et al. "Solar cell module having a surface coating material of

three-layered structure." U.S. Patent No. 5,650,019. 22 Jul. 1997.

[8]. Carr JM, Mackey M, Flandin L, Schuele D, Zhu L, Baer E. Effect of biaxial

orientation on dielectric and breakdown properties of poly (ethylene

terephthalate)/poly (vinylidene fluoride ‐ co ‐ tetrafluoroethylene) multilayer

films. Journal of Polymer Science Part B: Polymer Physics. 2013 Jun

1;51(11):882-896.

[9]. Riganakos KA, Koller WD, Ehlermann DA, Bauer B, Kontominas MG. Effects

of ionizing radiation on properties of monolayer and multilayer flexible food

packaging materials. Radiation Physics and Chemistry. 1999 May 31;54(5):527-

40.

[10]. Wang J, Cheng Q, Tang Z. Layered nanocomposites inspired by the structure and

mechanical properties of nacre. Chemical Society Reviews. 2012;41(3):1111-

1129.

[11]. Langhe D, Ponting M. Manufacturing and Novel Applications of Multilayer

Polymer Films. William Andrew; 2016 Jan 4.

[12]. Dooley J. Determining the processability of multilayer coextruded structures.

InANTEC-CONFERENCE PROCEEDINGS- 2007 (Vol. 4, p. 2369).

[13]. Mackey M, Flandin L, Hiltner A, Baer E. Confined crystallization of PVDF and

a PVDF‐TFE copolymer in nanolayered films. Journal of Polymer Science Part

B: Polymer Physics. 2011 Dec 15;49(24):1750-1761.

[14]. Liu RY, Jin Y, Hiltner A, Baer E. Probing Nanoscale Polymer Interactions by

Forced ‐ Assembly. Macromolecular rapid communications. 2003 Nov

1;24(16):943-948.

[15]. Zhang G, Lee PC, Jenkins S, Dooley J, Baer E. The effect of confined spherulite

morphology of high-density polyethylene and polypropylene on their gas barrier

properties in multilayered film systems. Polymer. 2014 Aug 18;55(17):4521-

4530.

[16]. Beadie G, Shirk JS, Rosenberg A, Lane PA, Fleet E, Kamdar AR, Jin Y, Ponting

M, Kazmierczak T, Yang Y, Hiltner A. Optical properties of a bio-inspired

gradient refractive index polymer lens. Optics express. 2008 Jul

21;16(15):11540-11547.

[17]. Alemán JV, Chadwick AV, He J, Hess M, Horie K, Jones RG, Kratochvíl P,

Meisel I, Mita I, Moad G, Penczek S. Definitions of terms relating to the structure

and processing of sols, gels, networks, and inorganic-organic hybrid materials

(IUPAC Recommendations 2007). Pure and Applied Chemistry. 2007 Jan

1;79(10):1801-1829.

[18]. Hua DW, Anderson J, Di Gregorio J, Smith DM, Beaucage G. Structural analysis

of silica aerogels. Journal of non-crystalline solids. 1995 Jun 2;186:142-148.

[19]. Aegerter M, Leventis N, Koebel MM, editors. Aerogels handbook. Springer

Science & Business Media; 2011 Jun 10

[20]. Kistler SS. Coherent expanded-aerogels[J]. The Journal of Physical Chemistry,

1932, 36(1): 52-64.

[21]. Mackenzie RC. Clay-Water Relationships. Nature 1953, 171, 681-683

[22]. Call F. Preparation of Dry Clay-Gels by Freeze drying. Nature 1953, 172, 126.

[23]. Hebbar RS, Isloor AM, Ismail AF. Preparation and evaluation of heavy metal

rejection properties of polyetherimide/porous activated bentonite clay

nanocomposite membrane. RSC Advances. 2014;4(88):47240-47248.

[24]. Ismadji S, Soetaredjo FE, Ayucitra A. Natural clay minerals as environmental

cleaning agents. InClay Materials for Environmental Remediation 2015 (pp. 5-

37). Springer International Publishing.

[25]. Wang Y, Gawryla MD, Schiraldi DA. Effects of freezing conditions on the

morphology and mechanical properties of clay and polymer/clay aerogels.

Journal of Applied Polymer Science. 2013 Aug 5;129(3):1637-1641.

[26]. Bandi S, Schiraldi DA. Glass transition behavior of clay aerogel/poly (vinyl

alcohol) composites. Macromolecules. 2006 Sep 19;39(19):6537-6545.

[27]. Hostler SR, Abramson AR, Gawryla MD, Bandi SA, Schiraldi DA. Thermal

conductivity of a clay-based aerogel. International Journal of Heat and Mass

Transfer. 2009 Jan 31;52(3):665-669.

[28]. Reim M, Körner W, Manara J, Korder S, Arduini-Schuster M, Ebert HP, Fricke

J. Silica aerogel granulate material for thermal insulation and daylighting. Solar

Energy. 2005 Aug 31;79(2):131-139.

[29]. Schiraldi DA, Gawyrla MD, Bandi SA, inventors; Schiraldi David A, Gawyrla

Matthew D, Bandi Suheel A, assignee. Absorbent compositions with clay

aerogels and methods for forming absorbent compositions. United States patent

application US 12/012,248. 2008 Feb 1.

[30]. Gawryla MD, Schiraldi DA. Novel absorbent materials created via ice

templating. Macromolecular Materials and Engineering. 2009 Sep

14;294(9):570-574.

[31]. Bandi S, Bell M, Schiraldi DA. Temperature-responsive clay aerogel− polymer

composites. Macromolecules. 2005 Nov 1;38(22):9216-9220.

[32]. Chen HB, Hollinger E, Wang YZ, Schiraldi DA. Facile fabrication of poly (vinyl

alcohol) gels and derivative aerogels. Polymer. 2014 Jan 14;55(1):380-384.

[33]. Gawryla MD, Liu L, Grunlan JC, Schiraldi DA. pH tailoring electrical and

mechanical behavior of polymer–clay–nanotube aerogels. Macromolecular rapid

communications. 2009 Oct 1;30(19):1669-1673.

[34]. Gawryla MD, van den Berg O, Weder C, Schiraldi DA. Clay aerogel/cellulose

whisker nanocomposites: a nanoscale wattle and daub. Journal of Materials

Chemistry. 2009;19(15):2118-2124.

[35]. Viggiano III RP, Schiraldi DA. Fabrication and mechanical characterization of

lignin-based aerogels. Green Materials. 2014 Sep;2(3):153-158.

Chapter 2

Fabrication and modification of multilayer films: HDPE/EVOH

2.1 Introduction

The main objective of this Chapter is to provide a detailed introduction about a polymer- based multilayer film system and related modified versions, which can encapsulate organic photovoltaic devices, protecting the photoactive materials from oxygen and water vapor in order to maintain a long-lifetime performance under different climate conditions. The focus is on film systems, introducing methods to improve barrier, flexural, and transparency properties. Once performance requirements are met, organic photovoltaic devices will be encapsulated with the target films, and the entire system life-time performance evaluated.

More details for this integration will be discussed in Chapter 3.

2.1.1 Background

Organic electronic devices, especially flexible ones, are promising technologies for industrial and consumer applications. Advantages such as low manufacturing cost, reduced environmental impact, and ease of integration into various medias have enabled these organic advanced devices for practical applications. Stability and degradation issues of organic materials currently limit wide use of organic photovoltaics; the permeation of water vapor and oxygen into core layers of electronic devices, especially some organic photovoltaic devices, can lead to their damage or malfunction, due to chemical reactions

16 between active layers and these permeants [1]. Devices of interest include displays, such as

LCD, OLED and photovoltaic modules. In order to protect core devices, barrier films are widely applied in the fields above to elongate the life-time of the devices.

In our current work, we use ethylene-vinyl alcohol copolymer (EVOH) and high density polyethylene (HDPE) as two key barrier materials combined and thermoformed through co-extrusion technology as a multilayer structure. EVOH exhibits excellent low oxygen transmission rate (OTR) values under dry conditions, due to a permeable structure with a high cohesive energy density (CED) and very low fractional free volume[2][4] (see Figure

2.1): the former is used to evaluate the solubility of permeant in the matrix. When CED is higher, the affinity between a potential permeant and the matrix becomes worse, a low solubility from this chemical disparity causes a low permeability; in the other physical view, a low fractional free volume in polymer matrix offers limit microcavities for permeants movement[3]. EVOH was selected to be the candidate of oxygen barrier materials for this reason.

Figure 2.1 Oxygen permeability VS ratio of fractional free volume (FFV)/cohesive

energy density (CED) for some widely used commercial polymers[4]

EVOH has, however, been found to lose partial oxygen barrier under very humid conditions (Relative humidity >75%) due to swelling of the polymer[5]. A hydrophobic polymer layer, HDPE, was utilized in the present work to sandwich the EVOH layers, in a multilayer film structure, to maintain a high oxygen barrier in humid conditions. HDPE offers a significantly better barrier against water vapor than most commonly used polymers

(see Table 2.1), as a result of olefinic hydrophobic character. And its nature of high density and well-packed inner chains also provides the chance to be a good candidate for water vapor barrier materials[4].

Table 2.1 Permeability of polymers commonly used in packaging[6].

Water vapor Oxygen permeability permeability Polymer @23℃ 50% / 0% RH @23℃ 85% RH [cm3mm/(m2 day atm)] [g mm/(m2 day)] Poly(ethylene terephthalate) (PET) 1-5 0.5-2 Polypropylene (PP) 50-100 0.2-0.4 Polyethylene (PE) 50-200 0.5-2 Polystyrene (PS) 100-150 1-4 Poly(vinyl chloride) (PVC) 2-8 1-2 Poly(ethylene naphthalate0 (PEN) 0.5 0.7 Polyamide (PA) 0.1-1 (dry) 0.5-10 Poly(vinyl alcohol) (PVA) 0.02 (dry) 30 Ethylene vinyl alcohol (EVOH) 0.001-0.01 (dry) 1-3 Poly(vinylidene chloride) (PVDC) 0.01-0.3 0.1

2.1.2 Barrier improvement methods

The permeability of a polymer film to oxygen and moisture can be varied in a range of

10-3~102 [cm3 mm/(m2 day atm)] and 10-1~102 [g mm/(m2 day atm)], respectively. It can be noted clearly that common gas barrier films based on polymer can be tuned to meet the basic barrier requirement for food packaging area, from Figure 2.2[7]. For advanced electronic applications, the encapsulation film needs to be further to improve barrier performance.

Figure 2.2 Water Vapor Transmission Rate (WVTR) requirements for different

applications[7]

Various fabrication techniques have been applied on thin films, aiming to improve their barrier performance. In this section, we will provide details about various methods and pre- treatment used in current work.

2.1.2.1 Layer-by-Layer

Layer-by-Layer (LbL) assembly is a simple and highly tailorable method to coating ultrathin layers on target substrates. Through immersion or spraying with cationic and anionic aqueous mixtures alternatively, these multifunctional layers have been built up on the substrate surface to realize its functions[8]. Previous work has demonstrated that this

19 method has a potential to modify the surface in order to enhance related properties such as antimicrobial, electrical conductivity, and flame retardancy[9-12]. In our current work, negative-charged clay nanoplatelets have been used to form LbL structure system with corporation of positive-charged polymer solution. Priolo, et al have demonstrated the spreading of clay layers provides gas molecules a chance to be trapped in the polymer- filled matrix, which is full of channels perpendicular to the molecules diffusion paths[13].

Inspired by his work, we cooperated with Grunlan Group at Texas A&M University, and applied LbL coating on our multilayer films.

Multifunctional layers were produced with positively-charged polyethylenimine (PEI) and negatively-charged montmorillonite clay (MMT) and poly (acrylic acid) PAA[14]. The multilayer film substrate was dipped, rinsed and dried in four different charged solutions as illustrated in Figure 2.3a. One cycle of four-layer deposition, PEI/PAA/PEI/MMT, is referred to as a quadlayer (QL), seen from Figure 2.3b. We named this recipe as MMT-

QL deposition. Similar recipe based on vermiculite clay (VMT) with large-aspect-ratio, named as VMT-QL, has also applied to compare the barrier improvement performance.

The experimental process and results will be talked in Section 2.3 and Section 2.4 with details.

Figure 2.3 Layer-by-layer process illustration (a) and nano brick wall structure of 5

QLs on film substrate: MMT (red), PEI (blue) and PAA(green) (b)[13].

2.1.2.2 Parylene

Vacuum deposition method is another important and widely used techniques to create barrier layers on substrates. This deposition method can be broadly classified into physical vapor deposition (PVD) and chemical vapor deposition (CVD). In this section, we will focus to talk about one of chemical vapor deposition to mainly improve water vapor barrier:

Parylene C deposition. The rest of techniques we also applied, such as e-beam deposition and atomic layer deposition (ALD), will be explained briefly in Section 2.1.2.3.

Parylene is a common name for Poly-para-xylylene polymers used as moisture and dielectric barriers[15]. The two most common types of parylene are: Parylene N and

Parylene C. Parylene C has been widely used as a water barrier coating on micro-electro- mechanical systems and biomedical implantable devices due to its high chemical stability, cheap cost and relatively low water vapor transmission rate[16][17]. Its relatively high flexibility and inherent transparency, and its ease in deposition under vacuum at room temperature are reasons for parylene use in photovoltaic applications[18]. Due to the uniqueness of the vapor phase deposition, the parylene polymers forms from a gaseous monomer without an intermediate liquid stage. Consequently, a conformal thin films will deposit on target substrate with free pin-holes or voids at ambient temperature[19]. And because of its high molecular weight, linear, crystalline polymer nature with all carbon backbone, it makes Parylene C quite hydrophobic, stable and highly resistant to permeation from most solvents and most chemical attack[20].

Although Parylene C coating layers have many advantages to make them candidates for water vapor barrier material, Madakasira et al. demonstrated that they may not be highly effective barriers for organic electronic devices requiring minimum WVTR requirement as low as 10-3 g/m2·day[19]; hence it is necessary to continue looking for a material or material combination which can meet this requirement.

2.1.2.3 Inorganic barrier materials

A material used as a barrier coating needs to be considered whether it has appropriate bulk properties such as density, to be qualified as an effective barrier. Inorganic barrier

22 materials, such as metal or metal oxide coatings, can be better options than organic materials[21]. To improve barrier for photovoltaic applications, opaque and thick coatings need to be eliminated to ensure modified films maintain transparency property in visible light range.

As transparent inorganic barrier materials, metal oxides (Al2O3), silicon oxides (SiOx)

[22-25] and silicon nitrides (SiNx) have widely used in decades as barrier coating layers . In this section, we will focus on introduction about several methods to deposit inorganic barrier materials.

Physical Vapor Deposition—Electron beam deposition (For Al2O3, SiO2, etc)

Physical vapor deposition, known as PVD, is a process which is carried out at low temperature and without chemical reactions, depositing target anode atoms directly on substrate to form thin layers[25]. Two methods can be used to obtain the gaseous phase, thermal evaporation for low-melting point materials such as Al, Ca, etc. and electron beam bombardment for high-melting point materials such as Al2O3, SiO2, etc. Most of PVD processes applied in our current work were E-beam; for example, Al2O3 pellets were bombarded with an electron beam under vacuum and released gaseous phase atoms. These target atoms then precipitate on modified films as a solid form (Figure 2.4) producing an e-beam coating layer[26]. The advantage of this technique is deposition rates as fast as several micrometers per minute. A fast, random vapor deposition offers a chance to generate pin-holes in coating layers. From AFM images in the Results section and associated barrier data, the potential defects from E-beam deposition layers are clearly evident.

Figure 2.4 Illustration of E-beam Vapor Deposition process[26].

Chemical Vapor Deposition—Atomic Layer Deposition (For Al2O3)

Chemical Vapor Deposition, known as CVD, is a process applied to form a high quality, thin layers with superior performance, making use of varying operating pressures, precursor types, and substrate status to apply the products of a chemical reaction. Volatile precursors react or decompose on target substrate to generate the desired layer materials, and unreacted precursors and by-product are removed by gas flow.

PECVD, short for Plasma-Enhanced Chemical Vapor Deposition, has been widely applied for fabrication of SiOx and SiNx as inorganic barrier layers. Although it has been widely recommended by many authors to fabricate a dense and pin-hole free inorganic layers, a relatively higher deposition temperature (~250 ℃) has limited its usage on our multilayer films[27-29]. For the present work, we used atomic layer deposition (ALD) to deposit a uniform and conformal barrier layers under temperature below 100 ℃. ALD is

24 known for its pinhole-free inorganic layers, which makes this technology a promising method to fabricate a barrier layer. The reasons for us to choose ALD to deposit Al2O3 are:

1. It can be deposited with different precursors based on Al with high reliability and short cycle times[30]; 2. Its deposition temperature can be tuned in a range of 30~300 ℃, which is compatible with target film’s usage; 3. Equipment, Savanah 100 (Cambridge Nanotech), was available in the MFL center at Case Western Reserve University. Trimethylaluminum

(TMA) and Deionized water have been chosen as molecular precursors to provide Al and

O sources; those two precursors are inserted into the chamber as gaseous phase alternately.

One precursor is present and reacts with the functional group kept on the substrate surface, then the purging gas flow removes the by-product/residual precursor and introduce the other precursor, to react with the new functional group left on surface (Figure 2.5)[31]. The whole process can be controlled that each time only one precursor’s reaction on the surface to form a monolayer, in order to produce a high quality, conformal and pin-hole free barrier layers after repeating the steps at target cycles.

Figure 2.5 Illustration of ALD process based on two precursors: TMA and DI-water[31]

2.1.2.4 Plasma Treatment

Plasma treatment was used in the current work to clean surfaces prior to further deposition steps. In addition to removing oil and fingerprint from the films, we surprisingly observed oxygen barriers for multilayer films, especially for those with HDPE surface layers, improved slightly. Coincidentally, Rossi et al. have demonstrated that argon cold plasma surface treatments on HDPE films can result in not only surface modification but also a decrease in its permeability values[32]. It is also recommended to carry out a pre- treatment by plasma before further modification in our following work. For the details parameter and discussion on the reasons to improve barrier, it will be explained with details in Section 2.2.2.4 and Section 2.3.2.

It has been a trend to combine organic and inorganic materials together to form a hybrid or multilayer barrier layer, for the achievement of high barrier performance[33-35]. This strategy is straightforward: inorganic barrier materials have an inherent low transmission rate to permeant as low as 10-5, but are brittle and easy to generate pinholes and defects during the deposition. Organic barrier materials possess relatively inferior barrier performance but they’re uniform and easy to generate a pin-hole free layers. Sandwich the two different barrier materials can overcome each other’s disadvantages and perform their strength. Ghosh et al. have published that parylene keeps water from condensing on the

[36] inorganic barrier layers such as Al2O3 or SiOx ; parylene films are also relatively flexible and durable protection layers, with a modulus close to 4 GPa. Parylene can remove the

26 stress from brittle inorganic barrier layers during frequently usage and rolling, and therefore warrants investigation for barrier improvements for our multilayer films[37].

2.1.3 Permeation principles and mechanism

A pressure or temperature gradient or a concentration gradient, are the main factors which lead the transport of gases or liquids[38]. The transportation process through polymer films can be illustrated by following steps: 1) small gas molecules condense on the surface of the film and dissolve into the film at the side of higher potential (pressure, concentration), the concentration of dissolved gas is proportional to its pressure; 2) gas molecules diffuse in and through the film driven by a concentration or pressure; 3) the gas molecules arrive to the other side of film and desorbed from the other surface of the film, completing the transportation through the films. The gas transmission process can be divided as dissolution, diffusion and desorption as Figure 2.6 shows[39].

Figure 2.6 Illustration to show the path of gas transmission in a film system.

The permeation rate of one gas can be impacted by “dissolution” and “diffusion” steps.

The permeability P, can be calculated as the product of solubility coefficient S and diffusion coefficient D: The permeation rate can be calculated by the permeability P, which is equal to the product of solubility coefficient S and diffusion coefficient D:

푃 = 푆 ∙ 퐷 EQ (2.1)

To a polymer based film, S is the amount of gas per unit volume of polymer film in equilibrium with a unit pressure of gas; D is affected by factors such as penetrant size, chain packing and polymer chain segment mobility; the molecular diffusion through the film is usually the rate-determining step in the process since it is a slow movement[38]. It is therefore significant to study “diffusion” as a dominant step to control the whole transmission rate. In order to make the theory session clear and straightforward, we separate the case as single layer film and multilayer film.

Case I: Single layer film

For Case I, the diffusion process can be described using Fick’s first law[40]:

퐽 = −퐷 ∙ ∇푐 EQ (2.2)

Where J is the molar flux (mol m-2s-1) and c is the gas concentration (mol m-3). In the case of gas vapor, the gas concentration at equilibrium status can be related to the partial pressure, ∆푝,

푐 = 푆 ∙ ∆푝 EQ (2.3)

Equation (2.2) can be combined with (2.3) and rewritten as

퐽= −퐷 ∙ ∇푐

= −퐷 ∙ d푐⁄d푥

= −퐷 ∙ 푆 ∙ ∆푃⁄∆퐿 EQ (2.4)

From Equation (2.4), the total flux of permeant is proportional to the pressure difference,

∆푃, between the two sides of the film, and inversely proportional to film’s thickness, ∆퐿.

However, the Equation (2.4) is an ideal derivation with assumption D and S independent and the process is at an equilibrium status, the case of steady state. The other state, known as transient one, is also important for transmission study because it reflects certain lag time for a barrier film and protected equipment are expected to benefit from a long lag time.

In order to model permeation process through a single layer film at transient status, Fick’s second law has been applied due to the assumption that concentration c is a function that depends on location x and time t. Graff, et al has developed a solution with certain assumptions such as: the vapor is saturated in the carrier gas at fixed concentration C1, concentration at downstream side is 0 and the gas is noncondensable[41]; his derivation combined with Fick’s second law becomes Equation (2.5),

푥 2퐶 1 푛휋푥 2 2 2 퐶(푥, 푡) = 퐶 (1 − ) − 1 ∑∞ sin( )푒−퐷푛 휋 푡/푙 EQ (2.5) 1 푙 휋 푛=1 푛 푙

where x means the location and 푙 is the film’s thickness. Then the total diffusing mass through this single layer film at this case can be integrated as

푡 푡 푄(푡) = ∫ 퐽 (푥 = 푙, 푡)푑푡 = ∫ 퐷 ∙ 푐 (푥 = 푙, 푡)푑푡 푡=0 푡=0

퐷푡퐶 푙퐶 2푙퐶 (−1)푛 2 2 2 = 1 − 1 − 1 ∑∞ 푒−퐷푛 휋 푡/푙 EQ (2.6) 푙 6 휋2 푛=1 푛2

When t is increasing to limit, EQ (2.6) will becomes a simple derivation referred to as the lag time,

퐷퐶 푙2 푄(푡 → ∞) = 1 (푡 − ) EQ (2.7) 퐿 6퐷

Then the lag time, L for a single-layer barrier film is the time it switches from transient to steady state, written as Equation (2.8),

푙2 L = EQ (2.8) 6퐷

Permeability P can be rewritten as Equation (2.9) based on EQ (2.1) and EQ (2.4),

∆퐿 푃 = 퐷푆 = 퐽 ∙ EQ (2.9) ∆푝

Case Ⅱ: Multilayer film

Compared to single layer film system, multilayer one has a more complicated architecture and different compositions. A series resistance model, known as the ideal laminate theory, can be applied to obtain the permeability of total multilayer film 푃푡표푡푎푙 under steady state. Combined with each layer’s permeability 푃푛, and thickness 퐿푛 , 푃푡표푡푎푙 can be described in the form of Equation (2.10) ,

퐿 푃 = 푡표푡푎푙 EQ (2.10) 푡표푡푎푙 퐿 퐿 퐿 1 + 2 + ⋯ + 푛 푃1 푃2 푃푛

where n denotes the layer numbers[42]. This total permeability is derived based on an assumption that a noncondensable gas flux penetrate through multilayer film under steady

30 state and from equation it helps to predict a multilayer film system’s barrier performance in a certain range. In the real test, the permeability can also be impacted by layer morphology, compatible issue and layer materials degradation, etc. Even to the similar structure, different number of layers even make a difference on transparency performance; this will be further discussed in the Section 2.3.

Similar as single layer film system, in case Ⅱ the lag time for multilayer film system has been revised as Equation (2.11)[41] ,

2 3 푛 푙푖 푛 푙푚 푚−1 푙푖 푖−1 ∑푖=1 { ∑푚=1 [ ∏푗=1 퐾푗] − 2 ∏푗=1 퐾푗} + 2퐷푖 퐷푚 3퐷푖 2 푙 푙훽 푙 푙 ∑푛 { 푖 ∏푖−1 퐾 ∑푛 [ ∑푛 ( 푚 ∏푚−1 퐾 ) − 훽 ]} 푖=1 퐷 푗=1 푗 훽=푖+1 훽−1 푚=훽 퐷 푗=1 푗 2퐷 푖 ∏ 퐾푗 푚 훽 퐿 = 푗=1 EQ (2.11) 푛 푙푖 푖−1 ∑푖=1 [ ∏푗=1 퐾푗] 퐷푖

Different from the single layer case, the lag time for multilayer film system depends on not only thickness and diffusion coefficient, but also solubility related parameter, 퐾푗 =

푆 푗 , which makes the lag time much more complicated. It can be noted through the Mocon 푆푗+1 and Ca-test data in the Results section that lag-time data can be observed.

Through design and modification of the films with extra barrier layers, utilizing related parameters such as layer thickness and different barrier materials’ permeability or solubility, we can control the permeation rate through the films and then control the amount of water vapor and oxygen exposed to the organic photovoltaic devices.

2.1.4 Permeability measurement

It is essential to have an accurate and reliable set-up to measure the OTR and WVTR in order to evaluate high quality barrier layers’ performance. Films’ barrier performance will determine its protection performance on elongate shelf-lifetime of organic electronic devices. Thus, a typical commercial system, Mocon (Figure 2.7(a)), has been chosen and widely used to analyze single layer or multilayer films’ permeation performance, oxygen permeability and water vapor permeability.

2.1.4.1 Mocon

This commercial tool has adapted ADTM F1249 for measuring OTR and WVTR respectively. The principle of this tool is shown in Figure 2.7 (b). The target film sample is clamped in a test plate between two chambers. On one side of the film, Chamber B, the system sweeps with nitrogen to eliminate the test gas, and on the other side, Chamber A, the test gas, oxygen or water vapor is allowed to flow into Chamber B through the film to be test until it reaches equilibrium. The permeated test gas then flows with carrier gas N2 into a pressure-modulated infrared sensor and absorb infrared energy, which can be measured by sensor[37,43-44]. Then this sensor generates an electrical signal to reflects the permeated gas concentration, for example, water vapor concentration. Then the system will calculate the rate of water vapor permeation by comparing the ratio of produced electrical signal to the signal generated from standard calibration film with known WVTR.

Figure 2.7 Picture shows MOCON PERMATRAN-W® Model 3/33 unit used for WVTR test (a) and illustration to explain the Mocon set-up work flow briefly, using water vapor as the permeant gas (b)[40].

Organic electrical devices place higher demands on barrier performance than typical applications, which becomes a challenge to commercial tools for acquiring a barrier data.

Measuring an ultra-low permeation rate as low as 10-4 even 10-6 g/m2·day with accuracy and reliability is quite difficult to current commercial system, necessitating solutions to measure permeation associated with high barrier performance; we introduce here the calcium corrosion test[44].

2.1.4.2 Calcium corrosion test

The calcium corrosion test is a widely used lab method to investigate ultra-high permeation barriers, especially for WVTR values below 10-3 g/cm2·day. The basic strategy to use calcium is quite straightforward: Calcium (Ca) is an active element which is conductive and opaque when in metallic form. As water vapor or oxygen permeates

33 through the barrier layers and meet with Ca metal, an oxidation reaction happens and generates a non-resistive and transparent metal hydroxide, Ca(OH)2. Through recording or monitoring the consumption of Ca through electrical conductance change (electrical method) or transparency change (optical method), it is possible to use this indirect method to calculate the corrosion rates of Ca, then the permeation rate of water vapor or oxygen.

It needs to be noted that distinguishing the attribution of Ca consumption from water or oxygen is difficult. Former publications from different authors have demonstrated calcium has a main reaction with water, although different sources claim that it is possible for Ca react with oxygen only[44-47]. All authors agree that water may play the dominant role to consume Ca and it can be an effective method to measure WVTR. For the discussion and measurement below, we will assume all the consumption of Ca is from water vapor and in this term to calculate WVTR.

Figure 2.8 Illustration of Top view (a) and Front view (b) about Ca sensor device

fabrication and Ca devices with barrier films before (c) and after (d) oxidation by the

permeated water vapor.

The reaction between Ca and water vapor is shown as below:

Ca + 2H2O (g) → Ca(OH)2 + H2 EQ (2.12)

A transparent and insulated calcium hydroxide will form after oxidation of Ca. Figure

2.8 (a) and (b) show the same Ca sensor device before and after oxidation by the permeated water vapor. Ca has been deposited between two gold conductive stripes used to connect the power, on the other side voltage on the Ca sensor will increase with the resistance increase at constant current (see Figure 2.8 (c) and (d)). Through measuring the resistance of the Ca as a function of time, the rate of Ca consumed can be converted into a rate of water vapor transmission rate according to the Equation (2.13):

1 푑 ( ) 푙 퐴퐶푎 푀퐶푎 푅 WVTR = −nδ𝜌퐶푎( )( )( )( ) EQ(2.13) 푤 퐴푓푖푙푚 푀퐻2푂 푑푡

where n is the molar equivalent of the degradation reaction, here, n is assumed to be 2 since the main reaction is oxidation by water. δ is Ca resistivity, equal to 3.4×10-8 Ωm and

3 𝜌퐶푎 is the density of Ca, 1.55 g/cm . 푙 and w are the length and width of deposited Ca, respectively. 퐴퐶푎 and 퐴푓푖푙푚 are the area of Ca and area of covered film which is exposed

to air and not sealed by epoxy glue. 푀퐶푎 and 푀퐻2푂 are the molar masses of water vapor

1 and calcium, which are 18 and 40 g/mol respectively. 푑( )/푑푡 can be obtained from the 푅

35 slope of a linear fit to the inverted resistance, conductance, versus time; examples will be shown in the Results section.

2.2 Experimental Section

2.2.1 Multilayer film fabrication

Multilayer films utilized the two polymer materials in this study: high density polyethylene (HDPE) from Dow, used as the moisture barrier material and ethylene vinyl alcohol (EVOH 27, Ethylene content 27%) from Kuraray, used as the oxygen barrier material. The densities of HDPE and EVOH were 0.965 and 1.2 g/cm3, respectively. Melt

Mass-Flow Rate (MFR) of polymers studied, according to ASTM D1238, were 8.3 g/10 min for HDPE and 4.0 g/10min for EVOH.

Multilayers of HDPE and EVOH, with 65 core layers with/without extra HDPE skin layers, were extruded from the melt by forced assembly co-extrusion technology which incorporates layer-multiplying technology. The one with HDPE skin layers is denoted as

Raw Film A (see Figure 2.9), and the other one without HDPE skin layers is denoted as

Raw Film B.

During the extrusion process, one extruder contains EVOH and the other one contains

HDPE. It needs to be noted that prior to co-extrusion, EVOH resins need to be dried under vacuum for 48 hours or more to get rid of moisture. In this case, 5 multipliers were used to prepare the target structure to obtain 65 layers, and the volumetric composition of EVOH and HDPE has been set as 50/50 by equalizing the pump rate. The same type of HDPE

36 surface layer has been used after 5th multiplier to provide extra water barrier and protect

EVOH from rapid degradation. A processing temperature of 220-230℃ was used and a 14 inch exit die was used to obtain the conformal polymer films. The total film thickness has been set as 4mil, controlled by speed and viscosity of melt at exit.

Figure 2.9 (a) Setup for the coextrusion of the multilayer A-B-A-B system. (b)

Schematic of a cross-section of 4 mil multilayer films consisting of 65 core layer of

EVOH/HDPE and two HDPE skin layers

2.2.2 Modification of multilayer films

2.2.2.1 Layer-by-layer deposition

The multilayer film was cut in 5x7’’ size and loaded on a programmable home-built robotic dipping system. The pH of cationic PEI solution is set as 10 and anionic PAA

37 solution is set as 4, using 1 M NaOH and HCl respectively before LbL coating. For MMT or VMT suspensions, they were left unaltered and kept as negatively-charged materials.

Starting from dipping into cationic PEI solution for 5min, the film then rinses, dries and dips into anionic PAA solution for another 5min, followed by rinsing and drying[48].

Continually dipping into PEI and Clay solution for 5min each to complete the initial quad layer (QL). That’s how the first deposition cycle forms. For the rest of cycles, they are completed with 1min dips in 4 solutions as the sequence, PEI/PAA/PEI/MMT(VMT), with rinsing and drying between each dipping. The assembled films are referred to as (MMT-

QL) or (VMT-QL). Schematic of the LbL coating process can be seen from previous

Figure 2.3 (a).

2.2.2.2 Parylene C deposition

Parylene deposition is carried out in Specialty Coating Systems (SCS) PDS 2010 deposition system. Parylene C, from Kisco diX-C parylene dimer, was used as the source material for the deposited film and loaded in the vacuum chamber, followed by heating, sublimating around 175 ℃ under vacuum to form into a dimer gas. Then the gas is pyrolized to cleave into monomer molecules in the pyrolizer furnace around 690 ℃, and deposits on the multi-layer film’s surface in the room-temperature chamber (Figure 2.10).

Multilayer film has gone through Argon plasma and A-174 adhesion promoter pretreatment prior to placing into the room-temperature chamber. The thickness of deposited parylene is measured by profilometer. It is important to mention that the thickness of parylene is not always to be measured by profilometer as the thickness is mostly linearly proportioned to

38 the usage of dimer (see Figure 2.11). The target thickness can be achieved by applying different usages of dimer, for example, 2 μm Parylene layer will be deposited on films by adding 3.5 g Parylene dimer into vaporizer chamber.

Figure 2.10 Flow of Parylene C vapor deposition process

Figure 2.11 Experimental results of parylene C film thickness with parylene dimer

weight[16]

2.2.2.3 Inorganic barrier materials deposition

As Section 2.1.2.3 introduces, inorganic barrier materials have been applied on films by two main paths: E-beam vapor deposition and Atomic Layer Deposition (ALD). For the former, it is an easy and relatively rough process compared to the latter. Target film samples were cut into 11 cm × 11 cm specimens, were fixed on the substrate and loaded in a vacuum Angstrom chamber, then an acceleration voltage as high as 8 kV was applied on tungsten filament to give off an electron beam to attack on target oxides, Al2O3 or SiO2 directly. Through tuning the voltage, the deposition rate can be controlled at a slow value, here is 0.5 Å/s, to get a relatively uniform deposition although it consumes more time.

Once it reaches the target thickness, here we set it as 30nm, the whole program will stop automatically and chamber can be vented to obtain the modified films.

For ALD, a Sananah 100 (Cambridge Nanotech) was used to deposit Al2O3 on target films. Limited by the area, the film needs to be cut to the dimensions 10 cm × 10 cm and fixed on the substrate with Kapton® tape. The deposition was performed at 80 ℃ and

5×10-2 Torr in order to be compatible by target films and the whole procedure, even organic devices for future use. Trimethylaluminum (TMA) was used as the precursor, and water was used as the oxidant. One cycle of TMA exposure, N2 purge, water exposure, N2 purge was referred as one cycle deposition and the timing sequence for this cycle was 0.015 s, 58 s, 0.015 s and 58 s (see Table 2.2). The thickness can be estimated through product of total cycles and deposition rate per cycle. For example, a 30 nm ALD-Al2O3 layer can be achieved by set 300 cycles since the deposition rate is 0.1 Å/s due to lower temperature setting. The exact thickness is determined by profilometer once deposition is done. Process

40 time depends on the temperature we set because the reaction takes longer time at lower temperature (See Figure 2.12).

Table 2.2 Summary of ALD Al2O3 layer deposition conditions

Parameter Value Unit

Chamber Temperature 80 ℃

Chamber Pressure 0.05 Torr

TMA pulse 0.015 sec

N2 purge and wait 58 sec

DI-water pulse 0.015 sec

N2 purge and wait 58 sec

Target cycles 300 cycles

Run time 9.8 hrs

Figure 2.12 Relationship and derived equation of purge time with deposition

temperature.

2.2.2.4 Plasma Pre-treatment

For plasma pre-treatment, Plasma Etch Reactive Ion Etching system was used to carry out an Argon plasma treatment on target films. The film was cut into 11 cm ×11 cm specimens and placed on a stage in the vacuum chamber with pressure as low as 0.1 Torr; treatment was then performed using argon gas at a flow rate of 25 sccm for 30 min at power

= 25 W. The post-treatment film was then kept in a sealed bag for further modification once they were removed from the chamber.

2.2.3 Barrier Performance Investigation

For the effective O2 measurement, oxygen transmission rate (OTR) for multilayer films with/without barrier improvement (raw film) were measured with a MOCON Ox-Tran 2/20

42 unit (ST or SH module, Minneapolis, MN,USA) in accordance with ASTM Standard

F1927[49]. The measurements were run under 1 atm pressure at 23 ℃ with a 100% relative humidity. The instrument was calibrated at 23 ℃ with a certified Mylar® film which has known oxygen transport parameters. Water vapor transmission rate (WVTR) for multilayer films with/without barrier improvement were measured with a MOCON PERMATRAN-

W® Model 3/33 unit. The measurements were run at 1atm pressure at 37 ℃ with a 100% relative humidity. The instrument was calibrated with a certified Mylar® film which has known water vapor transport parameter. Target films were cut by mold in order to fit

Mocon equipment. For each test, two same flat sheet film samples were used in parallel to average the barrier data. Samples were clamped into the diffusion cell, sealing using vaccum grease. The OTR or WVTR data were calculated by normalizing the flow rate at a steady state at least for 24 hrs. Once the barrier property was too low to measure by

MOCON (the minimum reliable OTR and WVTR data are 0.01 cc/m2 ·day and 0.01 g/ m2

·day, respectively), Ca corrosion tests were introduced to evaluate high barrier quality of modified barrier films. As we explained previously, the Ca corrosion test cannot distinguish between the oxidation reactions with O2 and H2O and because water vapor is highly reactive with Ca, the Ca corrosion test is mainly used to measure an effective H2O transmission rate[18].

Basic concept of Ca corrosion test has been introduced in Section 2.1.4.2. There are two methods used to conduct an effective WVTR measurement. In our current work, we adopted the electrical method than optical one since the former measuring the conductance change provides an accurate and continuous quantitative data for further analysis. In this experiment, a Ca sensor with an area of 18 × 5 mm2 and two gold electrical stripes were

43 coated on a glass substrate for conductance measurement as shown in Figure 2.13 (a). The glass substrate sonicated in acetone for 15 min and dried by N2 flow, prior to sputter- coating 30 nm gold stripes. After coated with gold, the substrate need to be cleaned by UV- ozone at 80 ℃ for 10 min. It is important to avoid contaminations existing on the substrate to lead a potential local oxidation to Ca sensor and finally influence measuring effective

WVTR. The cleaned substrate with gold stripes was covered by a shadow mask and transferred to a vacuum deposition Angstrom system. 300 nm Ca was deposited through the mask between two gold stripes. The deposition rate was monitored by a RC quartz monitor crystal and set at the rate of 0.5 Å/s initially to make sure a uniform base layer, then 2-3 Å/s when the thickness was over 50 nm. When the thermal evaporation of Ca completed, the substrate, modified with gold and calcium was removed from the chamber and transferred in the nitrogen glovebox, since Ca is really sensitive to oxygen and water vapor. Removing the shadow mask, the substrate was covered by target multilayer encapsulation film and sealed with a UV-cured epoxy (purchased from Epoxy Technology,

OG-142), which has been demonstrated as a barrier sealing material having a low WVTR

(see Section 2.3.2). After 15 min UV-light curing, a Ca-sensor test device was fabricated as shown as Figure 2.13 (b). This device was connected with metal wires and kept in a home-made chamber, which had a relative humidity ~99% and a room temperature 23 ℃ monitored by a RH reading sensor (see Figure 2.13 (c) and (d)). Provided a constant current 0.02 A by a Keithley DC power unit, the voltage applied on the Ca sensor changed with resistance. A Labview program was used to record the voltage change with the time and finally could be plot as a conductance change with the time. Then as explained in

44 previous Section 2.1.4.2, the effective WVTR could be calculated through the Equation

(2.13).

Figure 2.13 Illustration: Top view of Ca test devices (a) and a real Ca test device (b); A

home-made chamber creating a sealed environment with RH~99% (c) An electrical

conductance measurement set-up (d).

2.2.4 Mechanical and Transparency Performance Investigation

Once it was confirmed that the modified film systems had stable low barrier property by

Mocon or Ca test, it was essential to verify the mechanical quality and transmission quality

45 because these encapsulation films were designed for flexible, organic photovoltaic devices application. For mechanical testing, a rolling test was introduced to verify films and modification layers’ quality after at least 100 cycles at a small rolling radius. For optical properties, a UV-vis spectrometer was used to measure the transparency percentage in visible light range.

Mechanical Testing

A homemade auto-rolling machine was set up in the lab. Two rolling motors were fixed on a table and had a 10 cm gap between two shafts. It needed to be noted here that the space between shafts were adjustable according to the length of film sample. Each of the motor was connected with a metal rolling shaft with a radius 1/8 inch, which was also the rolling radius for film samples. This set-up can be seen in Figure 2.14: the target films needed to be cut in widths less than 10 cm, then fixed on Shaft A. Both motors were starting at the same time until the film almost completed rolling on the shaft A, then the direction was reverse. The number of rolling and time for each cycle could be set in a program related to motors.

Figure 2.14 A home-made rolling test setup composed by two rolling motors

Optical property testing

The percent transparency can be measured by a UV-vis spectrometer under ambient conditions. Two clean glass plates were pre-set in the spectrometer and system was calibrated as the base line (the system recognized two glass plate’s transparency as 100%).

The sample films were then clamped between glass plates and transmission between 400 nm to 800 nm, the visible light range, was measured. A good barrier film used for photovoltaic application needs to have a higher transparency percentage in this range because the light absorption of organic PV materials focuses in the visible spectrum[50].

Once the test was completed, a plot of transparency percentage vs wavelength was obtained; data during the visible spectrum, especially in the region of 400-633nm wavelength was used to evaluate film’s transparency performance because this region is considered to be efficiently absorbable for organic photovoltaic devices[51].

2.3 Multilayer Films’ performance

2.3.1 Raw film

As introduced in Section 2.2.1, a series of multilayer structure films composed of alternating EVOH and HDPE were co-extruded in different batches. Two systems were applied mainly to study the properties related to barrier, mechanical and transparency.

Table 2.3 exhibits the basic information about these two systems. All the data of raw films and modified ones are based on these two systems.

Table 2.3 Basic introduction about two systems multilayer film used for barrier project.

Sample System 1 System 2 EVOH, from Kurary, Ethylene content 27% Materials HDPE, from Dow, ρ=0.965 g/cm3

Film Structure

Film Image

Thickness 1mil (25.4 μm) 4mil (101.6 μm) OTR (cc/m2 day, 0.19±0.03 Below limit 100% O2, 23℃) WVTR (g/m2 day, 2.97±0.05 1.1±0.1 100%RH, 37℃) Transparency 90 65 (%)

Compared to the 1mil control bulk film (see Table 2.4), multilayer films have low OTR values, close to that of pure EVOH control films and with relatively lower WVTR values

48 close to that of pure HDPE films. It can be demonstrated that multilayer film can increase oxygen and water vapor barrier at the same time when co-extruding EVOH and HDPE.

EVOH plays a dominant role to oxygen barrier and HDPE contribute primarily to water vapor barrier.

Table 2.4 Control film’s barrier properties.

OTR WVTR Film Type Material (cc/m2 day) (g/m2 day) HDPE 49.2±3.2 2.9±0.3 1mil, Control film EVOH 0.16±0.03 91.5±6.2

We have also designed different core layers and different ratio of usage for HDPE and

EVOH in order to obtain an optimum structure. Table 2.5 shows barrier data and a series of films from System 1. With HDPE usage increasing, the ratio of HDPE increasing, and the permeability to water vapor is decreased, whereas the permeability to oxygen increases due to the decreased levels of EVOH (see Figure 2.15). Crystallinity information can be obtained from DSC curves on samples (see Figure 2.16). Multilayer films generally possess lower crystallinity than control films due to quick heat release rate from thinner core layers. The Tm of HDPE shifted to lower temperatures, which was a further evidence of HDPE confinement during multilayer co-extrusion[52]. A related study can be conducted through X-ray to study the crystal morphologies based on pattern angle. Zhang, et al demonstrated that HDPE could be confined by PC and HP030 to form a confined spherulite morphology, increasing the tortuosity for gas to diffusion. EVOH has also been confirmed to confine LLDPE during the extrusion[53][54]. HDPE has a higher chance to be confined by

EVOH to achieve a better morphology, then a better barrier property. From the morphology

49 and barrier data view, the optimum ratio option for multilayer has been chosen as HDPE:

EVOH=50:50 to keep a balance on barrier performance. The AFM image of this kind of film displays a clear separation between two core layers and similar thickness for each inner layer (see Figure 2.17). We have found there is no significant difference on barrier performance between 65 and 129 layers however it has a higher chance for mismatch between inner layers if increasing the number of layers during co-extrusion. From the real processing it is difficult to obtain a transparent and high quality barrier films with core layers over 257.

Table 2.5 Barrier data of a series of multilayer film with different ratio of HDPE/EVOH

Film Type Thickness of layers (nm) (Based on OTR WVTR 2 2 1mil, 65 core EVOH HDPE (cc/m day) (g/m day) layers) E/H=30/70 234 546 0.25±0.03 2.63±0.17 E/H=40/60 312 468 0.24±0.01 2.81±0.09 E/H=50/50 390 390 0.19±0.01 2.97±0.05 E/H=60/40 468 312 0.19±0.03 4.27±0.13 E/H=70/30 546 234 0.17±0.01 7.83±0.31 1mil, 129 layers, 198 198 0.20±0.02 3.42±0.08 E/H=50/50

Figure 2.15 Barrier data (OTR & WVTR) for films with different EVOH/HDPE ratios

Figure 2.16 DSC data about 1mil 65Layers multilayer film with different ratio.

Degrees of crystallinity are 67%, 62%, 59%, 58%, 60% and 62% (from top to down)

Figure 2.17 AFM image to illustrate multilayer films’ A-B-A-B-A structure

2.3.2 Barrier improvement performance

From Section 2.3.1 we have designed and fabricated two multilayer film systems and found the highest barrier performance is from thicker system 2, with OTR below Mocon limit (< 10-2 cc/m2day) and WVTR = 1.1 g/m2day. For advanced electronic applications, the encapsulated film needs to be further modified and significantly pushed down to improve barrier performance for organic electronic devices usage. Minimum barrier requirement for this application has been set as OTR<10-3 cc/m2 day and WVTR<10-3 g/m2 day[17]. Pure multilayer raw films are not qualified enough to meet such a low requirement.

Then further modification methods will be applied and evaluated the effect in this section.

Layer-by-layer

This deposition method has been conducted in Professor Jaime Grunlan’s research lab at

Texas A&M University. Various thicknesses (0.25mil, 0.5mil and 1mil) of films in System

1 and 4mil films in System 2 were used to verify the layer-by-layer barrier performance.

Table 2.6 exhibits barrier data of four different films before and after layer-by-layer deposition. As mentioned previously, an ordered “nano brick wall” structure is expected to form after multi cycles quad-layer deposition, although it is difficult to identify each MMT- clay layer (Figure 2.18). This image demonstrates that clay platelets deposit in a highly oriented fashion, which results in the excellent oxygen barrier behavior. From Table 2.6, a significant oxygen barrier improvement can be seen when comparing the OTR data before and after LbL deposition. For water vapor barrier, limited improvement was observed with LbL coating - this approach will clearly not improve water vapor barrier properties. An updated LbL process based on vermiculite clay (VMT), which has a large aspect ratio compared to MMT clay, was expected to contribute further improvement; while the WVTR was decreased, it was still far away from the target value of 10-3 g/cm2 day and transparency was also compromised.

Table 2.6 Barrier data about various multilayer films modified by Layer-by-Layer

OTR after Film Type OTR WVTR WVTR after LbL (cc/m2 day) (g/m2 day) (Based on LbL (g/m2 day) 100%O , 2 100%RH, E/H=50/50) 2 (cc/m day) 100%RH, 37℃ 23℃ 100%O2, 23℃ 37℃ 0.25mil 1.1±0.1 0.025±0.05 ------

0.5mil 0.6±0.1 Below limit ------2.82 2.71 (MMT-QL) (VMT-QL) 1mil 0.19±0.03 Below limit 2.97±0.05

1.06 0.92 (MMT-QL) (VMT-QL)

4mil Below limit Below limit 1.1±0.1

Figure 2.18 Illustration of Clay nano-brick wall structure on target film

Parylene C

As introduced in Section 2.1.2.1, Parylene C has been widely used as a water barrier coating on micro-electro-mechanical systems due to its high chemical stability, low cost and relatively low water vapor transmission rate. It provides an opportunity to be applied onto a polymer film substrate to achieve low water barrier target[16][17]

Despite the known barrier improvements obtained with Parylene C, its stable chemical structure makes it difficult to obtain reliable interface adhesion. In the absence of chemical bond sites, only mechanical adhesion is available during its deposition. An appropriate surface modification, such as pretreatment or introducing adhesion promoting agents, therefore needs to be utilized to enhanced adhesion.

In our current work, we applied an argon plasma treatment and A-174 silane adhesion promoting agent before Parylene C deposition. The advantages of adopting Ar-plasma treatment are: 1) as a method of surface-cleaning, plasma pre-treatment helps to eliminate accumulated substrate contaminants which may diminish coating quality; 2) it enables to increase HDPE surface energy, acquiring a high wettability to improve adhesion between

HDPE and A-174. Figure 2.19 exhibits water contact angle changes before and after Ar- plasma treatment. Unmodified HDPE, with low surface energy, has clearly changed from a hydrophobic to hydrophilic surface; 3) the plasma treatment provides a chance to crosslink surface layer. A surface crosslinking of HDPE may increase crystallinity locally in order to improve gas barrier[56]. The crosslinking mechanism in argon can be seen from

Figure 2.20. The use of Silane A-174 has been recommended because the molecule of silane can develop a robust chemical bond with the substrate, especially after plasma-

55 pretreatment, in order to help improve the surface adhesion capacity of Parylene C. Table

2.7 shows a barrier improvement with combination of Parylene C and pre-treatment.

Parylene C is an entirely conformal and pinhole-free layers which can be confirmed by

AFM image (Figure 2.21). The optimal combination can improve water vapor barrier from 1.1 g/m2 day to 0.5 g/m2 day. With increasing amount of Parylene usage, water vapor barrier has improved in a limit scale. And another fact is the thicker the extra layer is, the higher chance of delamination during flexible tests. Single Parylene C layers are insufficient to be used only as a barrier layer to satisfy the critical requirement due to the organic materials’ nature. An optimal Parylene C usage is also need to be considered to avoid layer leakage (too thin) or layer delamination (too thick).

Figure 2.19 Contact angle measurement for unmodified multilayer film (a) and films

after Ar plasma treatment for 10s (b), 60s (c), and 600s (d)

Figure 2.20. Schemes to HDPE crosslink mechanism by Ar plasma

Figure 2.21 AFM image to display a conformal and pin-hole free 2 μm Parylene C

deposition on HDPE surface layer of 4mil multilayer film.

Table 2.7 Water Vapor Barrier improvement of 4mil multilayer films by Parylene C with

or without pre-treatment.

Parylene C Usage WVTR Film Type Pre-treatment (g/m2 day) (μm) 100%RH, 37℃ 4mil Raw Film No No 1.1±0.1 4mil Raw Film No 3 0.84±0.4 4mil Raw Film 10min Ar+A174 3 0.63±0.02 4mil Raw Film No 5 0.73±0.05 4mil Raw Film 10min Ar+A174 5 0.51±0.05

Combination between inorganic materials and Parylene C

From section above we can tell the application of Parylene C, an organic barrier material, is not enough to improve water vapor barrier solely. And inorganic barrier materials have been mentioned previously as an effective barrier which can low down the transmission rate to 10-5 g/m2 day even lower but disadvantages are obvious as its advantages: they’re brittle and easy to form pinholes and leakage during frequently using. Figure 2.22 shows

Mocon monitoring curves, which reflect clearly that inorganic Al2O3 used solely can improve barrier starting from very low WVTR then degrade to a constant value due to

Al2O3 has been reported to slowly dissolved and corroded by water ingress to form defects[57] [58]. Therefore, a study of barrier improvement by combination of Parylene C and inorganic barrier layers has been conducted. Parylene C plays a role as smoothing, protecting, defect-decoupling and strengthening layer[59]. The simulative structure, shown as Figure 2.23, is expected to achieve a high barrier performance by utilizing each parts’

58 advantages to create a tortuous permeation path[37]. Results and discussion on this study will be demonstrated as below.

Figure 2.22 Mocon curves about two films with (blue) and without (red) 30nm Al2O3

deposition.

Figure 2.23. A stimulated structure to illustrate to combine organic and inorganic

materials to fabricate a tortuous gas permeation path[37].

Table 2.8 shows a barrier improvement of a series of films modified by a combination method based on Al2O3 and Parylene C. This combination method has significantly improved the WVTR from 10-1 to 10-4 g/m2 day under certain recipe. For the similar recipe, the techniques used to deposit Al2O3 are also essential to ensure to get a high quality of inorganic barrier layers. E-beam, as one of physical vapor deposition, has a rough deposition although it can be controlled at a very slow deposition rate. For the similar recipe, the one with ALD deposition displays a much lower WVTR than E-beam one.

Through AFM image (see Figure 2.24) it can be noted that ALD-Al2O3 has an incline to form a denser and conformal deposition due to its mechanism introduced in Section

2.1.2.3. A less pin-hole and more conformal inorganic layer helps improve the entire water barrier performance, even under the similar recipe.

For the WVTR over 0.01 g/m2day, the sample can be test through Mocon to get a reliable data. For the modified samples which has a better data below Mocon limit, we adopt Ca- corrosion test mentioned in Section 2.1.4.2. After fabricating the Ca sensor device in glovebox (full of nitrogen), the target modified film has been cut, covered on Ca sensor and sealed by Epoxy. It has been demonstrated previously that this Epoxy has a really low water vapor permeation: A Ca sensor covered by a glass and sealed by this Epoxy can survive 30 days more and Ca is still opaque and conductive. Assuming glass is non- permeable and as the only path for permeation, this epoxy can keep a long-time as a low barrier sealant. Therefore, the only path for water vapor permeates in our device is the target film. Figure 2.25 shows a representative time dependent conductance data of Ca

1 sensor. The slope of the conductance vs time corresponds to the value of 푑( )/푑푡 used in 푅

Equation 2.13. And two slopes reflecting in different areas help to acquire Lag time, which

60 mentioned in Section 2.1.3, to distinguish two states where water vapor goes through:

Transient and steady-state.

Table 2.8 Water Vapor Battier improvement on a series of multilayer films by a method

combined Al2O3 and Parylene C

# of Deposition WVTR Barrier Test Film Type combination 2 Recipe (g/m day) Method layers 23℃, >95% RH

N/A N/A 0.78±0.03 Mocon E-beam 30nm 1 0.12±0.04 Al2O3 + 2 µm 1mil, 5.5×10-2 ± Parylene C 2 65Layers, <0.0002 E/H=50/50 NO surface 1 2.3×10-2 layer Ca-Test ALD 30nm -3 Al2O3 + 2 µm 2 8±1×10 Parylene C 3 4.6×10-4

N/A N/A 0.28±0.01 Mocon 4mil, 65 Layers 100nm SiO +2 2 1 0.044 E/H=50/50 µm Parylene C 1/6 hdpe Ca-Test surface layer ALD 30nm Al2O3 + 2 µm 2 0.002 Parylene C

Figure 2.24 AFM images comparison on deposition of Al2O3 by E-beam (a) and ALD

(b). Same recipe applies on multilayer films: Film/Al2O3/Parylene/Al2O3/Parylene.

Figure 2.25 A representative time dependent conductance data of Ca sensor. Effective

WVTR and Lag time can be determined.

In summary, we have successfully to evaluate different barrier improvement methods through barrier tests and morphology characterization. An optimized recipe based on combination between inorganic barrier materials, Al2O3 and organic barrier materials,

-4 2 Parylene C has proposed and acquires a low water vapor barrier as low as 10 g/m day.

Once it’s possible to fabricate and modified a barrier film, then the flexibility test, used to verify the mechanical durability and UV-spectrum test, used to confirm not lose effective spectrum, are recommended applied on the modified films.

2.3.3 Rolling Test

A home-made setup has been installed as Section 2.2.4 introduction. After certain times rolling on the shafts, the tested films have been removed and conducted a barrier test,

Mocon or Ca-test, to confirm the quality and durability after frequently rolling and using.

Table 2.9 displays a barrier data before and after 100 times or 300 times rolling. It can be seen that films modified by LbL might be not durable after rolling: The fact that an undetected barrier data after LbL modification can be retained after rolling demonstrated the nano-brick wall clay structure has suffered damages and cracked. The more times rolling, the more areas cracked and leaked. For the samples deposited with Parylene C, data before and after rolling can keep in a same range, which confirmed that Parylene has a protective role as well as helps with stress relaxation during rolling to avoid severe cracks.

It’s also can be confirmed from the films modified by inorganic-organic combination method. Parylene C passes through frequent rolling tests and plays an important role for films barrier quality and durability.

Table 2.9 Barrier Data of modified films before and after rolling tests at certain times (default=100 times)

OTR WVTR Modificati OTR WVTR Film Type (cc/[m2·da (g/[m2 on (after rolling) (after rolling) y]) day]) 0.5 mil 65L 0.13±0. 0.35±0. 5X MMT- Below No 02 04 ------QL Limit surface 100X 100X film 2.97±0. N/A 0.19±0.03 0.21±0.04 2.90±0.12 1mil 65L 05 No 5X MMT- surface QL Below 2.87±0. Below limit 2.91 ± 0.04 film + 2 μm limit 05 Parylene C 1.2±0. 1.1±0. Below 1.1±0.0 N/A Below limit 1 1 limit 5 100X 300X 4 mil 65L 5X MMT- Below Below limit 1.06 1.1 1/6 QL limit surface 5 μm Below 0.73±0. Below limit 0.78±0.08 film Parylene C limit 04 (top&dow ALD n) 30nm Below Al O + 2 Below limit 0.002 0.0067 2 3 limit µm Parylene C

2.3.4 Transparency Performance

As an encapsulation film used for organic photovoltaic device, a superior transparency property is essential and significant to the whole device since many organic materials absorb the energy from the visible light to convert into the electricity. A transparent film helps to avoid losing the energy absorption from visible spectrum. UV-vis spectroscopy

(Figure 2.26) reveals that, 1mil films, unmodified or modified, exhibit excellent transparency throughout the visible light spectrum. For the modification methods, it can be observed that LbL method lost less compared to parylene. Even with a sophisticated

“inorganic-organic” combination’s application, it still has ~83% visible light transparency with losing~5% compared to raw film. An order of magnitude improvement on barrier, however, can offset the small lose on visible-spectrum. It also can be noted obviously that for the same number of core layers, film in system 1, 1mil film, has a much better transparency performance than film in system 2, 4 mil film. A higher thickness and introduction of HDPE surface layer can both be the reasons to loss transparency.

Continuously modifying thicker film based on system 2 may provide a product with low barrier and high durability, it helps less to obtain a transparent encapsulation film. To sum up, a film with high barrier and superior transparency should be the goal as the future encapsulation product.

Figure 2.26 Visible light transparency as a function of wavelength for a series of films

with different barrier improvement methods.

2.4 Summary

This chapter has provided details on introduction to explain why we need a barrier film for organic photovoltaic devices and what kind of requirement we need to meet. With this goal, OTR< 10-3 and WVTR< 10-4, we designed and fabricated multilayer films based on water barrier material, HDPE and oxygen barrier material, EVOH. Based on this film substrate, we keep pushing the barrier down by various methods and treatment, Layer-by-

Layer, Parylene, Plasma treatment and combination method integrated Parylene with

Al2O3. Finally, we have successfully to modify a barrier film which can meet the requirement to barrier, transparency and mechanical durability. In the next chapter, we will apply multilayer film on the organic solar cells to evaluate its protection performance. For modified film, considering the modification is not easy to process batch, it is recommended strongly to apply the high barrier films as a future work.

2.5 Acknowledgement

The current work in this chapter cannot be done without support from three supervisors,

Dr.Schirlaidi, Dr.Baer and Dr.Olah. A special thanks to Clips for funding support. The support from former graduate student, Dr. Matt Herbert, Dr. Guojun Zhang and teamwork with Sunsheng Zhu and Ci Zhang has been acknowledged. A sincere appreciation to Ina

Matin from More Center and Nichole Hoven from MFL. The help and cooperation with

Grunlan’s group, TAMU has also been recognized and thanks!

2.6 References

[1]. Charton C, Schiller N, Fahland M, Holländer A, Wedel A, Noller K.

Development of high barrier films on flexible polymer substrates. Thin Solid

Films. 2006 Apr 28;502(1):99-103.

[2]. Lagaron JM, Catalá R, Gavara R. Structural characteristics defining high barrier

properties in polymeric materials. Materials Science and Technology. 2004 Jan

1;20(1):1-7.

[3]. Barton AF. Solubility parameters. Chemical Reviews. 1975 Dec;75(6):731-753.

[4]. Duncan TV. Applications of nanotechnology in food packaging and food safety:

barrier materials, antimicrobials and sensors. Journal of colloid and interface

science. 2011 Nov 1;363(1):1-24.

[5]. K.L. Yam, Gas permeation of packaging materials, in: K.L. Yam (Ed.), The

Wiley Encyclopedia of Packaging Technology, third ed., John Wiley and Sons,

Inc., New York, 2009, pp. 551–555.

[6]. Andrew W. Permeability and other film properties of plastics and elastomers.

Plastics Design Library. New York: Norwich. 1995:58-60.

[7]. Barrier materials & encapsulation for OPV, Institut National De L’Energie

Solaire

http://www.sophiari.eu/fileadmin/SOPHIA_docs/documents/Workshops/Works

hop_on_OPV/Cros_INES_SOPHIA_EERA_OPV_WS_111108.pdf (accessed

September 2, 2017).

[8]. Decher G, Eckle M, Schmitt J, Struth B. Layer-by-layer assembled

multicomposite films. Current opinion in colloid & interface science. 1998 Feb

1;3(1):32-39.

[9]. Dvoracek CM, Sukhonosova G, Benedik MJ, Grunlan JC. Antimicrobial

behavior of polyelectrolyte− surfactant thin film assemblies. Langmuir. 2009 Jun

18;25(17):10322-10328.

[10]. Lutkenhaus JL, Hammond PT. Electrochemically enabled polyelectrolyte

multilayer devices: from fuel cells to sensors. Soft Matter. 2007;3(7):804-816.

[11]. Park YT, Ham AY, Grunlan JC. High electrical conductivity and transparency in

deoxycholate-stabilized carbon nanotube thin films. The Journal of Physical

Chemistry C. 2010 Mar 15;114(14):6325-6333.

[12]. Carosio F, Laufer G, Alongi J, Camino G, Grunlan JC. Layer-by-layer assembly

of silica-based flame retardant thin film on PET fabric. Polymer Degradation and

Stability. 2011 May 31;96(5):745-750.

[13]. Priolo MA, Gamboa D, Holder KM, Grunlan JC. Super gas barrier of transparent

polymer−clay multilayer ultrathin films. Nano letters. 2010 Nov 3;10(12):4970-

4974.

[14]. Priolo MA, Holder KM, Greenlee SM, Grunlan JC. Transparency, gas barrier,

and moisture resistance of large-aspect-ratio vermiculite nanobrick wall thin

films. ACS applied materials & interfaces. 2012 Oct 10;4(10):5529-5533.

[15]. Wikipedia contributors, "Parylene," Wikipedia, The Free Encyclopedia,

https://en.wikipedia.org/w/index.php?title=Parylene&oldid=788330598

(accessed September 2, 2017).

[16]. Kuo HI, Zhang R, Ko WH. Development of micropackage technology for

biomedical implantable microdevices using parylene C as water vapor barrier

coatings. InSensors, 2010 IEEE 2010 Nov 1 (pp. 438-441). IEEE.

[17]. Xie X, Rieth L, Williams L, Negi S, Bhandari R, Caldwell R, Sharma R,

Tathireddy P, Solzbacher F. Long-term reliability of Al2O3 and Parylene C

bilayer encapsulated Utah electrode array based neural interfaces for chronic

implantation. Journal of neural engineering. 2014 Mar 24;11(2):026016.

[18]. Kim N, Potscavage WJ, Sundaramoothi A, Henderson C, Kippelen B, Graham S.

A correlation study between barrier film performance and shelf lifetime of

encapsulated organic solar cells. Solar Energy Materials and Solar Cells. 2012

Jun 30;101:140-146.

[19]. Madakasira P, Inoue K, Ulbricht R, Lee SB, Zhou M, Ferraris JP, Zakhidov AA.

Multilayer encapsulation of plastic photovoltaic devices. Synthetic Metals. 2005

Nov 15;155(2):332-335.

[20]. Why Use Parylene?, Parylene Engineering,

http://www.paryleneengineering.com/why_use_parylene.htm (accessed

September 2, 2017).

[21]. Schaepkens M, Kim TW, Gün Erlat A, Yan M, Flanagan KW, Heller CM,

McConnelee PA. Ultrahigh barrier coating deposition on polycarbonate

substrates. Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and

Films. 2004 Jul;22(4):1716-1722.

[22]. Dameron AA, Davidson SD, Burton BB, Carcia PF, McLean RS, George SM.

Gas diffusion barriers on polymers using multilayers fabricated by Al2O3 and

rapid SiO2 atomic layer deposition. The Journal of Physical Chemistry C. 2008

Mar 27;112(12):4573-4580.

[23]. Fang L, Shen M, Yang J, Li Z. The effect of SiO2 barrier layer on the dielectric

properties of CaCu3Ti4O12 films. Journal of Physics D: Applied Physics. 2005

Nov 18;38(23):4236.

[24]. Chiang CC, Wuu DS, Lin HB, Chen YP, Chen TN, Lin YC, Wu CC, Chen WC,

Jaw TH, Horng RH. Deposition and permeation properties of SiNX/parylene

multilayers on polymeric substrates. Surface and Coatings Technology. 2006

May 22;200(20):5843-5848.

[25]. Wikipedia contributors, "Electron-beam physical vapor deposition," Wikipedia,

The Free Encyclopedia, https://en.wikipedia.org/w/index.php?title=Electron-

beam_physical_vapor_deposition&oldid=797652091 (accessed September 2,

2017).

[26]. Wikimedia Commons contributors, "File:Electron Beam Deposition 001.jpg,"

Wikimedia Commons, the free media repository,

https://commons.wikimedia.org/w/index.php?title=File:Electron_Beam_Deposit

ion_001.jpg&oldid=199384249 (accessed September 2, 2017).

[27]. Chen TN, Wuu DS, Wu CC, Chiang CC, Chen YP, Horng RH. High-performance

transparent barrier films of SiOx/SiNx stacks on flexible polymer substrates.

Journal of The Electrochemical Society. 2006 Oct 1;153(10):F244-248.

[28]. Garcia-Cespedes J, Thomasson S, Teo KB, Kinloch IA, Milne WI, Pascual E,

Bertran E. Efficient diffusion barrier layers for the catalytic growth of carbon

nanotubes on copper substrates. Carbon. 2009 Mar 31;47(3):613-621.

[29]. Aubry E, Ghazzal MN, Demange V, Chaoui N, Robert D, Billard A. Poisoning

prevention of TiO2 photocatalyst coatings sputtered on soda-lime glass by

intercalation of SiNx diffusion barriers. Surface and Coatings Technology. 2007

Jun 25;201(18):7706-7712.

[30]. Groner MD, Fabreguette FH, Elam JW, George SM. Low-temperature Al2O3

atomic layer deposition. Chemistry of Materials. 2004 Feb 24;16(4):639-645.

[31]. Klumbies H. Encapsulations for Organic Devices and their Evaluation using

Calcium Corrosion Tests.

[32]. Rossi A, Incarnato L, Tagliaferri V, Acierno D. Modification of barrier properties

of polymeric films of LDPE and HDPE by cold plasma treatment. Journal of

polymer engineering. 1995;14(2-3):191-197.

[33]. Jung K, Bae JY, Park SJ, Yoo S, Bae BS. High performance organic-inorganic

hybrid barrier coating for encapsulation of OLEDs. Journal of Materials

Chemistry. 2011;21(6):1977-1983.

[34]. Amberg-Schwab S, Hoffmann M, Bader H, Gessler M. Inorganic-organic

polymers with barrier properties for water vapor, oxygen and flavors. Journal of

sol-gel science and technology. 1998 Jan 1;13(1):141-146.

[35]. Kim TW, Yan M, Erlat AG, McConnelee PA, Pellow M, Deluca J, Feist TP,

Duggal AR, Schaepkens M. Transparent hybrid inorganic/organic barrier

coatings for plastic organic light-emitting diode substrates. Journal of Vacuum

Science & Technology A: Vacuum, Surfaces, and Films. 2005 Jul;23(4):971-977.

[36]. Ghosh AP, Gerenser LJ, Jarman CM, Fornalik JE. Thin-film encapsulation of

organic light-emitting devices. Applied physics letters. 2005 May

30;86(22):223503.

[37]. Kim N. Fabrication and characterization of thin-film encapsulation for organic

electronics. Georgia Institute of Technology; 2009.

[38]. Barrier Propertities of Polymers, Polymer Properties Database,

http://polymerdatabase.com/polymer%20physics/Permeability.html (accessed

September 2, 2017).

[39]. Gas Barrier Seminar, High Gas-Barrier Resin Soarnol,

http://www.soarnol.com/eng/solution/solution050303.html (accessed September

2, 2017).

[40]. Wikipedia contributors, "Fick's laws of diffusion," Wikipedia, The Free

Encyclopedia,

https://en.wikipedia.org/w/index.php?title=Fick%27s_laws_of_diffusion&oldid

=797952086 (accessed September 2, 2017).

[41]. Graff GL, Williford RE, Burrows PE. Mechanisms of vapor permeation through

multilayer barrier films: Lag time versus equilibrium permeation. Journal of

Applied Physics. 2004 Aug 15;96(4):1840-9.

[42]. Zeman S, Kubík L. Permeability of polymeric packaging materials. Technical

Sciences/University of Warmia and Mazury in Olsztyn. 2007(10):26-34.

[43]. Permeation Basices, Article: PPS 18-5102, Permeation,

http://www.mocon.com/assets/documents/pps-18-5102-permeation-basics.pdf

(accessed September 2, 2017).

[44]. Nisato G, Bouten PC, Slikkerveer PJ, Bennett WD, Graff GL, Rutherford N,

Wiese L. Evaluating high performance diffusion barriers: The calcium test.

InProc. Int. Display Workshop/Asia Display 2001 Oct 16 (Vol. 1435).

[45]. Gregg SJ, Jepson WB. 186. The oxidation of calcium in moist oxygen. Journal of

the Chemical Society (Resumed). 1961:884-888.

[46]. Cros S, Firon M, Lenfant S, Trouslard P, Beck L. Study of thin calcium electrode

degradation by ion beam analysis. Nuclear Instruments and Methods in Physics

Research Section B: Beam Interactions with Materials and Atoms. 2006 Sep

30;251(1):257-260.

[47]. Paetzold R, Winnacker A, Henseler D, Cesari V, Heuser K. Permeation rate

measurements by electrical analysis of calcium corrosion. Review of Scientific

Instruments. 2003 Dec;74(12):5147-5150.

[48]. Song Y, Tzeng P, Grunlan JC. Super Oxygen and Improved Water Vapor Barrier

of Polypropylene Film with Polyelectrolyte Multilayer Nanocoatings.

Macromolecular rapid communications. 2016 Jun 1;37(12):963-8.

[49]. ASTM F. 07. Standard Test Method for Determination of Oxygen Gas

Transmission Rate, Permeability and Permeance at Controlled Relative Humidity

Through Barrier Materials Using a Coulometric Detector. American National

Standards Institute. Aug. 2007.

[50]. Lunt RR, Bulovic V. Transparent, near-infrared organic photovoltaic solar cells

for window and energy-scavenging applications. Applied Physics Letters. 2011

Mar 14;98(11):61.

[51]. Yoshida K, Oku T, Suzuki A, Akiyama T, Yamasaki Y. Fabrication and

Characterization of PCBM: P3HT Bulk Heterojunction Solar Cells Doped With

Germanium Phthalocyanine or Germanium Naphthalocyanine. Materials

Sciences and Applications. 2013 Apr 24;4(04):1.

[52]. Minick J, Moet A, Baer E. Morphology of HDPE/LDPE blends with different

thermal histories. Polymer. 1995 May 1;36(10):1923-32.

[53]. Zhang G, Lee PC, Jenkins S, Dooley J, Baer E. The effect of confined

crystallization on high-density poly (ethylene) lamellar morphology. Polymer.

2014 Jan 30;55(2):663-672.

[54]. Zhang G, Xu H, MacInnis K, Baer E. The structure-property relationships of

LLDPE–EVOH blend films fabricated by multiplication extrusion. Polymer.

2015 Jan 28;57:117-124.

[55]. Rand BP, Richter H, editors. Organic Solar Cells: fundamentals, devices, and

upscaling. CRC Press; 2014 Aug 26.

[56]. Kim JW, Choi HS. Surface crosslinking of high‐density polyethylene beads in a

modified plasma reactor. Journal of applied polymer science. 2002 Mar

28;83(13):2921-2929.

[57]. Abdulagatov AI, Yan Y, Cooper JR, Zhang Y, Gibbs ZM, Cavanagh AS, Yang

RG, Lee YC, George SM. Al2O3 and TiO2 atomic layer deposition on copper for

water corrosion resistance. ACS applied materials & interfaces. 2011 Nov

16;3(12):4593-601.

[58]. Nehm F, Klumbies H, Richter C, Singh A, Schroeder U, Mikolajick T, Mö nch T,

Hoßbach C, Albert M, Bartha JW, Leo K. Breakdown and protection of ALD

moisture barrier thin films. ACS applied materials & interfaces. 2015 Sep

30;7(40):22121-22127.

[59]. Chen TN, Wuu DS, Wu CC, Chiang CC, Chen YP, Horng RH. Improvements of

permeation barrier coatings using encapsulated parylene interlayers for flexible

76 electronic applications. Plasma Processes and Polymers. 2007 Feb 22;4(2):180-

185.

Chapter 3

Integration of barrier films with organic photovoltaic devices

In this chapter, the barrier films which we designed, fabricated and modified in Chapter

2 were integrated with organic photovoltaic devices to evaluate films’ protection performance. A variety of 4 mil films from system 2 were used to encapsulate organic solar cells in this work due to their inherent low oxygen/water vapor transmission rate (The best

-3 2 option of 4-mil films from system 2 we have chosen can have an OTR<10 cc/m day and

WVTR= 0.14±0.02 g/m2 day at 23 ℃, 100%RH). The focus of this work is on barrier’s effect on devices more than transparency, because high transparency can be achieved with barrier improvement on 1mil multilayered films. A relation between barrier films and organic solar cells’ lifetime performance was the expected objective of this part of the project. The core objective in this chapter therefore is to evaluate organic solar cells’ performance and their lifetimes under different conditions. The organic solar cells’ background and selection will be introduced Section 3.1. The procedure for this device fabrication and encapsulation with barrier films will be explained in Section 3.2. The device performances will be evaluated from efficiency life (%), a relative change in solar cells’ power conversion efficiency (PCE) with time, periodically and related results and discussion will be stated in Section 3.3.

3.1 Introduction

3.1.1 Overview of organic solar cell packages

Starting with the announcement of Bell lab of the first practical silicon solar cell in 1954, this application of chemistry and physics has grown steadily in importance for half a century[1]. Inorganic crystalline-based solar cells have become a relatively mature area[2]; existing commercial PV devices are mostly based on crystalline silicon wafers or ribbons[3].

Organic or polymeric photovoltaic devices, based on composites of conjugated polymers and fullerene derivatives, have attracted enormous interest as the need for renewable, flexible devices has grown. The key benefits of such organic systems include low-cost manufacturing, flexible fabrication, reduced weight and low environmental impact; most of these advantage provide potential opportunities to become a disruptive technology in photovoltaic market[4-6]. Figure 3.1 shows a gradually increasing need from the market (a) and an increasing higher power conversion efficiency from lab data (b). Both of the factors become a big boost for organic PV industry and market[7].

Figure 3.1 An increasing forecast production capacity (a) and a increasing efficiency of

OPV can make a boost for OPV industry and market development[7]. [Data source and adapted from “Recent OPV Technology and Market Forecast (2009-2020)”, SNR research,

2013]

3.1.2 Degradation mechanism of organic photovoltaic devices

Compared to a mature silicon solar cell technology and market, organic photovoltaic devices are promising, but face an enormous challenge to enter and conquer the market.

Increasing the power conversion efficiency of OPV, improving the stability and lifetimes of these devices are big challenges that need to be overcome[8].

The organic materials used for photovoltaic active layer have been found to be unstable under ambient conditions[9]. Jorgensen, et al proposed typical mechanisms of OPV degradation, separating them into two categories: chemical degradation and physical & mechanical degradation[10]. The latter one focuses on morphology control and morphological stability which can be controlled by tuning related parameters, such as applying a longer annealing treatment. Chemical degradation may involve a series of chemical reactions which result in decrease performance of the system; polymer photo- oxidation, transport layer corrosion, electrode degradation, migration of atom from the electrodes, reaction between transfer layer and active layer are all possible sources of degradation. A major reason for chemical degradation, as concluded by a number of authors, is the diffusion of oxygen and water vapor[11-14]. The reaction of water vapor and oxygen permeated into the active layers of electronic devices can result in a damage and malfunction. Oxygen can be activated by UV illumination to form the superoxide or hydrogen peroxide, which may attack organic materials in active polymers aggressively[15].

When oxygen and water diffuse into the device, they may bind oxygen atoms to vinyl bonds, which break the conjugation and lead to formation of carbonyl group; and they can also oxidize metal electrode to generate a thin insulating oxide barriers to hinder electric conduction and lead a delamination between electrode and semiconductor layers. It has been demonstrated that oxygen and water vapor lead to most significant degradation for organic photovoltaic devices[16]. For that purpose, we integrated films from Chapter 2 with

81 those core devices: we hope to postpone the permeation of molecular O2 and H2O into the solar cell devices by encapsulating with a transparent barrier film.

3.1.3 Overview of existing solutions and products

A great deal of work has been carried out with the goal of converting organic solar cells into real products. Galagan et al. has published work on the use of an ITO-free recipe, deposited on a barrier PEN film (seen as Figure 3.2), for a flexible organic solar cell[17].

The concept was brilliant, but lacked the necessary lifetime testing.

Figure 3.2 Illustration of a ITO-free solar cell device’ s structure (a) and a picture to

display the real device.[17]

Charton et al introduced the concept of developing inorganic organic hybrid polymer lacquers, which were then used to cover Fraunhofer ISC devices[18]; this work is illustrated

82 in Figure 3.3; machine and method allowed massive production of flexible barrier layers utilizing roll to roll process and a final thickness was between 2 to 4 μm. These barrier films have been used to cover OLEDs housed in the glovebox, with claimed lifetimes of

2-3 months.

Figure 3.3 A scheme to show lamination and lacquering machine: (1) unwinder (2)

anilox roller, (3) roller coating system, (4) thermal drying unit, (5) UV drying unit, (6)

laminating unit, (7) corona treatment, (8) unwinder, (9) rewinder[18].

For commercial flexible product, most of the solar cell devices which have an acceptable cost are made of amorphous silicon (See Figure 3.4 a-c), such as the Seeed® 1 W thin- film flexible solar panel which has a 10% efficiency and costs $16.98[19]; The PowerFilm®

7 W rollable solar charger which cost $131.99 can be used for camp lighting[20]; The Uni- solar ® 136 W flexible solar panel can be installed on the top of roof[21]. Relatively flexible performance is from amorphous silicon type so it cannot be regarded as commercial organic solar panels. InfinityPV® claims to have invented a product based on printed organic solar cells which can work as a 4 Watt Rollable solar charger (See Figure 3.4 d)[22];

That product, which is close to the original intention of organic solar cell., utilizes roll-to- roll process and flexible PET as substrate. Organic solar modules are mounted on the film, and sealed with an acrylic-based adhesive to yield a system operating at 2-4% efficiently, but at the price of €175/meter, which for the lifetime it is not mentioned. In summary, it is still hard to fully converted the work in the lab into a real product. And we expect our barrier films can be the potential candidate to use as the flexible substrate, which can be the base as well as provide essential barrier, flexible and transparent property.

Figure 3.4. Some commercial flexible solar cell panels or tape: (a) Seeed® 1 W thin-film

flexible solar panel (b)PowerFilm® 7 W rollable solar charger (c) Uni-solar ® 136 W

flexible solar panel. (d) InfinityPV® 4 Watt Rollable solar foils and tapes[19-22]

3.1.4 Current lab work and project objective

In this section, an introduction to the fabrication work and characterization methods used in this project is presented. With this base of solar cell knowledge, it will be helpful to fabricate organic photovoltaic devices with high quality and integrate with barrier films.

Conventional Organic Solar Cell Fabrication

In a typical organic solar cell device, the polymer semiconductor layer, which is widely known as the “active layer”, is inserted between a transparent conducting electrode (TCE) and a metal electrode[23]. For the active layer, poly(3-hexyl thiophene) (P3HT), a hole- transporting conjugated polymer assembled into nanowires to form nanoscale domains of charge carrying materials, as well as phenyl-C61-butyric acid methyl ester (PCBM), an electron-transporting fullerene derivatives which could be uses as a soluble electron conducting semiconductor when blended with the former, are commonly used as bulk heterojunction (BHJ) blends in the active layer of polymer solar cells[24][25]. BHJ blends require only a single layer to create electron donor-acceptor heterojunction, making this type of organic solar cell (BHJ-solar cell) greatly simplified through a solution processing process. Figure 3.5 shows the architecture of a BHJ solar cell; it can be seen clearly that the electrons and holes generated from photoexitation have separated in the active layer and flowed into TCE and metal electrodes[26]. For the TCE layer, indium-tin-oxide (ITO) is the most common example, and aluminum (Al) or silver (Ag) are examples of typical metal electrodes. In addition to these three parts, there is a hole transfer layer between the active layer and the ITO electrode, as well as an electron transfer layer between the active layer and the metal electrode. For the classic structure, a poly(ethylene dioxythiophene): polystyrene sulfonic acid (PEDOT:PSS) facilitated the hole transfer layer and a lithium fluoride (LiF) work as the electron transfer layer. The latter was demonstrated to improve electron injection in some organic devices, prevent a direct contact or some chemical reactions between donor and metal and reduces the serial resistance of the interface between metal and donor[23].

Figure 3.5 A standard BHJ solar cell structure to show a sandwich architecture where active layer is inserted into the middle of the whole device. And an illustration image and

TEM image on the left exhibit the work principle and morphology of BHJ layer. Adapted

from Reference [26]

Inverted Organic Solar Cell Fabrication

An inverted OSC has been demonstrated by many researchers to exhibit improved air stability over conventional structures[27-29]. Different from conventional photovoltaic systems, electrons in an OSC are collected by the ITO electrode, and holes are collected from top metal electrode in inverted model; that is the reason that such solar cells are referred to as having inverted structures. In our inverted OSC model, we chose P3HT:

PCBM as active layer, and zinc oxide (ZnO) as a transparent electron selective layer between ITO and active layer. The layer between the active layer and metal electrode, in this case, was the hole selective layer, molybdenum oxide (MoO3). We used silver as the

86 top electrode since once oxidation happens, the formation of silver oxide exhibits a higher work function to enhance hole collection, and compared to other metals that might have been selected, silver is much more stable in the presence of air[30]. In Section 3.3 we have compared two different type of solar cells’ performance and given out the explanation.

Mechanism of organic solar cells

Almost all kind of solar cells, no matter silicon-based or organic one, have similar basic working principles: A energy source, light has been provided and absorbed as photons by donor material to form excitons, electron-hole pairs, through photoexcitation. Then these pairs move, separate or recombine at donor-acceptor interface. For the fully separated charge carriers, they will be conducted to respective electrodes with the help of internal electric field. The electrical potential difference between electrons and holes will be displayed as an electrical power source: photo-voltage and photo-current when connecting to an external circuit. Figure 3.6 illustrates the working mechanism of organic solar cells.

Two transfer layers have been introduced previously and they can play a role to stop excitons diffusing and recombining at the interface, as well as help moving the targets, holes or electrons to respective electrode[31].

Figure 3.6 The mechanism of organic solar cells. (1) Photoexcitation to generate

excitons; (2) Diffusion of excitons; (3) Dissociation of excitons; (4) Free charge

transferred and collected by electrode. [32]

Characterization of organic solar cells

From the introduction above we have shown the working principle of solar cell devices; a standard, reproducible tool to for quantifying the energy the devices transfer from light to electricity is also needed. The I-V curve is the classic method to test and show the solar cells’ basic parameters in order to quantify the devices’ quality. A typical curve is shown as Figure 3.7; the short circuit current ISC is calculated when the voltage equals 0 and the open circuit voltage VOC is the voltage when there is no current going through the devices.

According to the basic equation, P=IV, the power generated from this cell can be calculated and shown as the blue curve in Figure 3.7. From the PV curve one can tell easily that there

88 is a point at which the cell can provide the maximum power, and voltage and current at that point are regarded as VMP and IMP. Fill Factor (FF) is used to evaluate the quality of the solar cell and can be written as Equation 3.1[33].

P FF = max × 100% EQ (3.1) ISCVOC

Through comparing the ratio of maximum power (Pmax) to the theoretical power, the product of ISC and VOC, a fill factor of the device can be obtained to evaluate if the device provides as much as power it could to electricity. Power conversion efficiency (PCE) is a key parameter for solar cell’s performance, which can be used to evaluate the efficiency how much solar power transferred into the electricity. PCE can be written as the ratio of the maximum output electrical power to the input solar power and can be calculated through Equation 3.2.

P FF ∙ I V PCE(η) = max × 100%= SC OC × 100% EQ (3.2) Pin Pin

2 Where Pin is the power from incident light, the unit is in W/m , with the surface area of the solar cell. In the result section, we will use this definition as indication of the solar cell performance. In order to compare this parameter, it is recommended strongly on test

condition to guarantee Pin as a constant value.

In summary, we have known the basic knowledge of organic solar cells, the working mechanism and potential reasons to lead a degradation. Through related tests we can keep monitoring solar cell packages’ performance through some key parameters. With these tools we can start fabrication and encapsulation, to ultimately evaluate the packages’ quality.

Figure 3.7 A characteristic IV curve used to illustrated the key parameters of organic solar cell[32]. This curve is measured from a real solar cell and obtained from IV station.

3.2 Experimental section

Experimental procedures including solar cell devices fabrication, measurements, encapsulation and long-term performance testing, are discussed in this chapter.

3.2.1 Materials

The ITO coated glass and PET substrates, were purchased from Xinyan Technology and

Sigma Aldrich, with ~20 Ω/square and ~60 Ω/square resistances, respectively. The glass-

ITO substrate had a L-shaped pattern which was compatible with the mask used as cathode deposition. PET-ITO had no specific pattern and had to be etched by solvent, which will

90 be introduced in next part of this thesis. Poly(3,4-ethylenedioxythiophene): poly(styrenesulfonate) (PEDOT: PSS) aqueous dispersion was purchased from Heraeus.

The electron donor material, Poly(3-hexylthiophene-2,5-diyl) (P3HT), and electron acceptor, phenyl-C61-butyric acid methyl ester (PCBM), were purchased from Rieke materials, Inc. and SES research, respectively. The solvent, 1,2-dichloribenzene, used to dissolve P3HT: PCBM, was purchased from Sigma-Aldrich. The remaining solvents, acetone, ethanol and isopropyl alcohol were all from Fisher. For conventional device fabrication, the electron transfer layer, LiF, was purchased from Materion and the metal electrode Al pellets from Kurt J. Lesker. For inverted device fabrication, zinc acetate dihydrate, which is used to deposit the ZnO transparent electron selective layer. and MoO3 for hole selective layer, were both purchased from Sigma-Aldrich. Metal electrode, Ag pellets were purchased from Kurt J. Lesker.

3.2.2 Device Fabrication

Both conventional solar cell and inverted solar devices were investigated and compared, in order to obtain a stable device as test media. The fabrication methods for these two types are described as follows:

Conventional Solar Cells

Conventional solar cells were the first type we used for test media. The solar cell with device architecture: “Glass – ITO - PEDOT: PSS - P3HT: PCBM – LiF - Al” was prepared according to the procedure below. The ITO coated glass substrates were first cleaned thoroughly with following solvents under 10 min sonication each: DI-water with

91 detergent/DI-water rinsing/acetone/ethanol/ isopropyl alcohol, dried by blowing nitrogen, and finally cleaned using an UV-Ozone cleaner @ room temperature for 15min. After cleaning by various methods, the cleaned substrates were covered with a thin PEDOT: PSS layer by spin coating, followed by annealed in ambient at 150 ℃ for 10 min. Then the rest of deposition, fabrication and characterization would be processed in a glove box system full of nitrogen. The P3HT: PCBM solution was prepared 1 night in advance in 1,2- dichloribenzene at a ratio of 1:0.8 and a final concentration of 20 mg/mL. The active layer was spin-coated from this blend aqueous solution of P3HT: PCBM at 600 rpm for 1 min and 2000 rpm for extra 2 sec, then annealed for 5 min at 140 ℃ to result in a 120 to 150 nm layer. Then LiF (1 nm) and Al (100 nm) were thermal evaporated at rate of 0.5 A/s and

3 A/s, as sequence through Angstrom Engineering Evovac Deposition System under vacuum pressure below 4x10-6 Torr.

Inverted Solar Cells

The inverted solar cells are the other type we used for an advanced test media with stable performance. The solar cell with device architecture: “Glass – ITO – ZnO - P3HT: PCBM

- MoO3 - Ag” was prepared using a similar procedure to that used for conventional cell cleaning steps. After cleaning using the various solvents and UV-Ozone, a ZnO layer, used as electron selective layer, was deposited through spin-coating a ZnO precursor solution at

4000rpm for 40s, followed by annealing at 150 ℃ for 7 minutes at room environment in order to convert zinc acetate to zinc oxide. This precursor solution was prepared 1 day in advance by dissolving 0.25 M zinc acetate dehydrate in 0.25 M monoethanolamine and 2- methoxyethanol[34]. Then the ZnO-deposited substrates had been transferred into the

92 glovebox full of nitrogen to complete the rest of photoactive layer’s deposition and rest fabrication. The active layer had a similar preparation and deposition steps as conventional solar cell, and had been spin-coated with a thickness of 120~150 nm. After annealing and cleaning, MoO3 (10nm) and Ag (80nm) were thermal evaporated at rate of 0.25 A/s and

2.5 A/s, as sequence through Angstrom Engineering Evovac Deposition System under vacuum pressure below 4x10-6 Torr.

Organic Solar Cells based on PET substrates

The fabrication steps discussed above were based on ITO-Glass substrates which has a

L-shape ITO pattern to be compatible with the metal electrode mask. The ITO-PET substrate sheet had to be pre-cut as the same size as a ITO-glass substrate at a dimension as 14.8 mm × 24.8 mm and pre-etched to achieve the same ITO-L-pattern as a glass substrate. First, the pre-cut ITO-PET piece need to be covered over the desired target L- pattern using an electric tape and exposed the rest of ITO ready to be etched (See Figure

3.8 a). The acid etch solution, a mixture of 1:1 ratio of hydrochloric acid and DI water, in a fume hood, then placed this etch solution on the hotplate at 60 ℃ for 10 min. A tweezer was used to dip the exposed ITO part completely into the acid-water bath for 10 sec, then the samples were out and rinsed it with DI water. After removing the protective electric tape, a very sharp line outlining the ITO against the PET substrate will be shown as Figure

3.8 b. A further test can be conducted using an Ohm meter to verify the resistance across covered area and etched area should be infinite theoretically. This etched ITO-PET piece was then taped with Kapton® double-sided tape on the same size glass substrate, to form

93 a rigid substrate for rest of deposition (See Figure 3.8 c). An organic solar cell based on a flexible PET substrate, was then fabricated as shown in Figure 3.8 d.

Figure 3.8 ITO-PET substrates pre-treatment process has been illustrated in these

figures.

3.2.3 Device Characterization

Once the organic solar cells were coated with the top metal electrode layer, the devices were moved in the glovebox (at this time organic solar cell kept function without exposing to the air) and set for a J-V Characterization. This characterization was carried out in the

94 glovebox also and under 1.5 AM solar illumination at intensity of 100 mW/cm2 using Oriel

Sol2A solar simulator. A Keithley 2400 IV station had been used to enable the measurement of current density-voltage (J-V) curves. Through varying the applied voltage in a certain range to observe how the corresponding current change. Key data such as open- circuit voltage (VOC), short-circuit current density (JSC), fill factor (FF) and power conversion efficiency (PCE) were all calculated through the J-V curves under illumination.

3.2.4 Barrier Film Encapsulation

Once the organic solar cells were characterized as working, these devices needed to be encapsulated with our barrier films, fabricated in Chapter 2, taken out of the glovebox, stored under different conditions to observe the solar cells, then the films’ performance tested periodically. There are two methods we have used to seal the encapsulation films:

UV-cured epoxy and ethylene vinyl acetate film.

UV-cured epoxy

UV-cured epoxy was purchased from Epoxy Technology and has been qualified as possessing a low water vapor transmission rate, as was confirmed in Chapter 2. Once a piece of organic solar cell had finished its initial J-V curve measurement, it would be put on the center of a barrier film with the dimension 6 cm× 8cm. A 1ml syringe was used to extract epoxy and drop them around the device. Then the UV-light was used to cure the epoxy frame for 30 sec for an initial hardening to avoid flowing over the device once we drop more around the frame. This step was followed by application of more epoxy around initial epoxy frame; the other piece of barrier film with same size had been covered over

95 the flowing epoxy and device in the center. UV-light was applied for another 10 min; at this point, the epoxy-sealed package with structure “Film-Epoxy/solar cell device/Epoxy-

Film” was fabricated as shown in Figure 3.9 (a). It needs to be noted that the entire process was carried out in a nitrogen glovebox.

EVA thermal lamination

EVA films were purchased from Interlayer Solutions and were pre-cut to the same size as barrier films, 6 cm× 8 cm. In order to avoid having the molten EVA flow over the device, the center of EVA film would be cut an empty area with a dimension 2 cm × 3 cm, a little bit larger than the device size (See Figure 3.9 b). Different from epoxy sealing in the glovebox, EVA sealing had to be performed outside of the glovebox due to the size of the lamination machine. The solar cell’s quality therefore was impacted inevitably and this influence would be talked about in next section. The vacuum lamination machine has been pre-heat to the target temperature 100 ℃, an optimized temperature which can melt the

EVA and avoid films and devices further degradation. The solar cells would be removed from the glovebox and loaded in the center of barrier film and EVA film immediately, covered with the other barrier films and subjected to vacuum lamination. After 10 min sealing, this EVA-sealed package with structure “Film-EVA/solar cell device/EVA-Film” had been fabricated as the Figure 3.9 (c) shown.

3.3 Results and discussions

3.3.1 Influence from encapsulation process

Organic solar cells are easy to degrade and sensitive not only to oxygen and water vapor, but also to heat, plasma, organic solvent and physical scratching. UV-light curing is an

97 essential step for epoxy sealing, the same as heat for EVA sealing. The concern when we sealed organic solar cell devices was whether UV-light shining or heat from EVA lamination made a negative impact to the devices’ performance.

For epoxy sealing, solar cells need to be sealed and cured under the UV-light for at least

10 min. We then compared the key parameters (VOC, JVC, FF and PCE) extracted from the same batch solar cells before and after UV-light irradiation (see Table 3.1). Those averaged data, based on 8 devices, showed little variability. Devices maintained 98.7% power conversion efficiency after a 10-min UV irradiation, which means this curing step could be compatible with solar cell packages’ encapsulation.

For EVA film lamination, however, a necessary heating step made an impact on solar cell performance. Most workers who have heated solar cell devices to improve their performance were annealing on purpose and at a mild temperature in a shorter time, such as 4~10 min[35-37]. In our case, heating from lamination cannot avoid and what we could do was shortening the heat time if we could pre-heat EVA film to target temperature then integrated with devices. So when heated at 90 ℃, our solar cell’s performance showed a slightly decrease on measured value after an optimized 10 min lamination and a drastic decrease happened after a normal 90 min lamination process. Li, et al had published his work that annealing organic solar cells with similar recipe as ours at 110 ℃ for 10 min displayed the best performance as a postproduction treatment, however, a longer time post- treatment had not evaluated in his work. The optimized temperature was set depends on active layer surface through AFM image and he claimed 110 ℃ annealing produced the roughest surface which benefits a rougher metal-polymer interface for more efficient charge collection[38]. A higher temperature would degrade the performance in this case.

Annealing 10 min at 90 ℃ in air might not impact the performance significantly due to the explanation from Li’s work and ISC inherent stability. A 90 min 90 ℃ annealing in air, however, would accelerate the degradation process especially reacted with O2 and water vapor in air at a high reaction rate for a longer period. Even though the same condition annealing in glovebox, we measured that polymer layer would still degrade in a 2 hours annealing @ 90 ℃ to result in a low PCE ~ 1.84% for the device.

Table 3.1 Performance parameters of solar cells before and after UV curing or Heat @

90 ℃ in air for different periods.

UV-light shining for 10 Performance parameters Initial value min

VOC (V) 0.6 ± ≤0.01 0.6 ± ≤0.01

2 JSC (mA/cm ) 10.25 ± 1.0 10.18 ± 1.0 FF 0.52 ± ≤0.02 0.52 ± ≤0.02 Power Conversion 3.24 ± 0.4 3.20 ± 0.4 Efficiency (%)

Performance Heat @ 90 ℃ for Heat @ 90 ℃ for Initial value parameters 10 min 90 min

VOC (V) 0.6 ± ≤0.01 0.58 ± ≤0.01 0.43 ± 0.03

2 JSC (mA/cm ) 13.36 ± 0.7 12.98 ± 0.6 11.72 ± 0.8 FF 0.50 ± 0.05 0.51 ± 0.04 0.44 ± 0.01 Power Conversion 3.92 ± 0.2 3.84 ± 0.4 2.12± 0.24 Efficiency (%)

In conclusion, both 10 min UV-light irradiation and 10 min 90 ℃ heating had less impact on solar cell performance according to our measurement on the key performance parameters before and after the two methods sealing. Heating at 110 ℃, following to a

99 published optimized temperature, and a fast lamination process was used to shorten the time in air. A longer heating time outside definitely would degrade organic solar cell’s performance due to degradation and reaction. UV-light irradiation and a revised heating process were therefore acceptable to organic solar cells and then could be compatible with the devices.

3.3.2 Different recipes and substrates influence on performance of solar cell

packages

Once we confirmed the sealing process had no significant impact to core organic devices, those encapsulation skills could be applied to investigate a long-term lifetime performance of the sealed packages integrated with organic solar cell devices. Considering frequently lamination outside cause more instability due to heat, oxygen and water vapor, measured or calculated results from most cases below were from the packages sealed by UV-cured epoxy.

As we introduced previously, two recipes for organic solar devices’ fabrication were used: a conventional and an inverted process. Figure 3.10 (a-e) shows the evolution of the open circuit voltage (VOC), Fill Factor (FF), short circuit current density (JSC) and power conversion efficiency (PCE) of two types of organic solar cells with storage time. In the case of conventional and inverted solar cell, both of VOC and FF could remain at about 80% their initial value for most time during test period, as seen from Figure 3.10 (a) and (b).

JSC and PCE, however, behaved quite different among two type devices. Due to reaction with oxygen and water vapor, both devices tended to degrade and reflected from their PCE

100 change. Inverted solar cells exhibited better and stable performances, which could still possess about 60% of their initial values we named as efficiency life (%) after 300 hrs in air, as shown in Figure 3.10 (e). In contrast, conventional solar cells exhibited a faster degradation rate and died after 96 hours even earlier in the same condition as inverted one.

Many scholars found the similar phenomenon and gave the explanations: Han et al. claimed that the top Al electrode used in conventional solar cell underwent severe damage and produced bubble-like features to hinder the charge collection[39]; Jia et al. stated the phenomenon acidic PEDOT:PSS layer was detrimental to ITO anode[40]: 1) it may corrode the ITO; 2) PSS was likely to react with the conjugated polymer in active layer; 3)

PEDOT:PSS may form discontinuous surface morphology and unstable electronic structure; as well as using the air-sensitive low work function cathode Al, they were all reasons for fast degradation of conventional solar cells[10][29][42]. For an inverted cell, it avoided using those two interior materials and used Ag to replace Al, as well as inserted

ZnO to inhibited electrode oxidation and kept high electron mobility[43][44]. Those reasons could result in a better performance of inverted solar cells. From Equation 3.2 we have been told that PCE is proportional to the product of JSC, VOC and FF if the incident light

2 power is constant, in our case it is 100 mW/cm . From Figure 3.3 (c) and (d) we could observe during a long-term test, oxygen and water vapor would degrade the active layer and it would be reflected from JSC. Thus, JSC, PCE and efficiency life had a synergistic similar trend in each type of solar cells. Thus, for the following long-term test, it is recommended to use efficiency life (%), the concept reflecting normalized efficiency in real-time, as the standard to evaluate solar cell packages’ performance.

101

Both cases above were devices based on ITO-glass substrates, we have tried 3 batches to use the same recipe as conventional solar cell to fabricate devices on ITO-PET glass substrates. From Figure 3.11 (a), it should be noted that for the best PET batch 3 which adopted chemical etch method introduced in Section 3.2.2, these flexible solar cell devices had an average power conversion efficiency as 1.4%, highest as 1.8%. Compared to the same recipe, rigid solar cell devices based on ITO-glass had an average PCE as 4%, highest as 5%. Two reasons could be concluded to explain interior performance for flexible solar cell: Subjectively, we have no equipment fixed for flexible substrate deposition. Without a rigid base taped by double-side tape, the flexible PET would curve and distort during spin- coating since vacuum applied led to a deformation; with a rigid base, when we peeled the

PET devices from the rigid base, an inevitable physical distortion would result in an uncertain structural failure. Objectively, ITO-PET substrate we purchased has a sheet resistance as 60 Ω/square, and ITO-glass substrate we purchased was 20 Ω/square. And from real test on the substrates through 4-point resistance measurement (see Figure

3.11(b)), PET and Glass substrates were 163 and 62 Ω/square, respectively. Although the data were different from they claimed, the approximately 3 times difference on resistance could be reflected from the difference of PCE performance of solar cells based on PET or glass. Therefore, appropriate equipment for deposition and low resistance substrates applied could be summarized as keys to fabricate an effective, flexible solar cells.

102

Figure 3.10. Data about VOC (a), FF(b), JSC(c), PCE(d) and efficiency life (%) (e) of two

type of organic solar cells vs storage time at the same condition: storing in dark, in the

air.

103

Figure 3.11 Power conversion efficiency of PET flexible solar cells from different batches.

(a). A 2.6-3 times difference on sheet resistance of two type substrates, from data sheet and

4-point measurement (b)

104

3.3.3 The importance of barrier films (with or without films)

From Figure 3.10 we found an organic solar cell device without barrier could only survive less than half month at a relatively mild conditions. That’s why the barrier films we modified in Chapter 2 are important and essential to maintain a longer life-time of those devices. Figure 3.12 shows solar cells performance with and without barrier films. For the same type of solar cells, the shelf-lifetime of the device encapsulated with barrier films was much longer (at least 4 times) compared to the devices without films’ protection. In the case of a 4 mil barrier film from system 2 encapsulation structure, efficiency life can still obtain ~60% after 3 months. We also conducted a same batch of devices encapsulated by different films (Film A, B and C) from system 2, with various WVTR data as 0.6 ± 0.1, 1.1

± 0.1 and 1.3 ± 0.1 g/m2 day at 37 ℃ and RH=100%, respectively. All the films were all the unmodified films. It can be observed that with water barrier of films improved, the shelf-lifetime of protected devices was improved. As seen in Figure 3.13, for the devices encapsulated with Film A with WVTR=0.6 g/m2 day could result in a less than 40% degradation with ~2500 hours. Considering the films we modified in Chapter 2 can achieve a better WVTR lower than 10-4 g/m2 day, we believe that modified films may further improve solar cells’ performance through blocking more oxygen and water vapor effectively, even over ~10000 hours. Currently we had three films with different WVTR

(we assumed WVTR played a dominant role in current degradation test), and once we got encapsulation test data with more films with various WVTR, we could establish a basic quantitative relationship between water vapor barrier and elongated solar cell lifetime.

105

Figure 3.12 Efficiency life as a function of storing time for organic solar cells with

different structures encapsulated with or without unmodified 4 mil film.

106

Figure 3.13 Efficiency life as a function of storing time for inverted organic solar cells

encapsulated with three different 4 mil films.

3.3.4 Different sealing skills

In Section 3.3.1 we have learned that 10 min UV-light irradiation and 10 min revised lamination process could be compatible with the devices. For a transient encapsulation, both methods were acceptable however, for a long-term lifetime performance, we have found different solar cell packages’ performance under a critical condition. Figure 3.14 (a) exhibits two types of solar cell packages’ performance before and after a 10-day critical test under 48℃, RH=30% and 1 sun output power condition. Both types are the same batch inverted solar cells encapsulated by 4 mil Film A from system 2 and only one difference was the sealing method, red bar representing EVA-sealing while blue bar representing

107 epoxy-sealing. Under the same critical condition and same package, the data should be close while it was observed epoxy sealing packages behaved interior than EVA sealing.

From original packages we expected epoxy one should be a little bit better than EVA one because 10 min UV made less impact than a 10-min heat outside. Through a 10-day critical test it had been observed as a contrary result: EVA sealing performed much more stable than epoxy one and average efficiency life was 9% higher. The only likely cause was potential leaking happens from the sealing area. Figure 3.14 (b) shows the packages before and after tests and it has been displayed that epoxy had turned to yellow color which indicates epoxy aging while EVA had no obvious change. Epoxy might have a higher chance to leak from yellow aging layers, which resulted in an interior performance than

EVA sealing one. Thus for a long-term observation without frequently tests, it was recommended to adopt EVA sealing packages especially under conditions with strong UV light exposure.

108

Figure 3.14 Similar organic solar cell packages sealed by EVA (red) and UV-cured

epoxy (blue) had a different power conversion efficiency performance after a 10-day

critical test (a). Solar cell packages sealed by epoxy had an obvious color change on sealing layers after critical test, which from a long-term UV irritation leading aging issue

while EVA-sealing one had no obvious change.

109

3.3.5 Long-term test for the same packages under different conditions

Currently we have completed to fabricate inverted solar cell device packages encapsulated by multilayer barrier films. Then those packages would go through a series tests under different climate conditions. Figure 3.15 displays four different climate conditions we used to test solar cell packages. Different efficiency life change had shown with storing time under various conditions: Under an ideal condition which had a pretty low humidity, the device with protection films may keep a high performance as 75% of initial value after 90 days (Figure 3.15, black). In a relatively ideal condition with 40-50% humidity the packages had a 40% degradation due to more water vapor penetration and reaction with devices (Figure 3.15, red). In a real-world test we left the packages on the top of the building during high humidity summer time and tested them after 18 days for a one-time test, the average performance of device packages can maintain a 60% efficiency life after close to 3 weeks’ shining and showering (Figure 3.15, blue). In the critical condition, we could set a fixed environment with the help of Q-sun solar simulator chamber and the device packages sealed by EVA could survive as 40% life left (Figure 3.15, pink).

A critical condition may help to do the accelerated test for solar cell packages’ performance.

Once we applied the films with modified barrier technique and improved the sealing skill, the revised solar cell packages are expected to maintain a high shelf-lifetime even at critical condition.

3.4 Summary

110

In this chapter we have provided a basic introduction about organic solar cells: its structure, development and market. All the current fabrication need a great barrier as a premise.

That’s the value to apply the barrier films we fabricated in Chapter 2. Integrated with barrier polymer films, solar cell devices had been fabricated and encapsulated as packages, and tested under different conditions. We have evaluated different recipes, different substrates, different sealing skills to expect obtain the constant method for packaging.

Currently, inverted solar cell based on glass substrates could be the best option as a stable test media. Encapsulated by a 4 mil raw film with a WVTR=0.6 g/m2 day, the device packages could maintain a 60% power efficiency of initial value after 3 months in air. In future work, we may improve flexible solar cell’s fabrication skills and apply our modified films with much lower WVTR as low as 10-4 g/m2 day to protect core devices. It can be expected to achieve a flexible solar cell packages which could survive a regular using for over 3 months.

3.5 Acknowledgement

A special thanks to my three supervisors, Dr. Schiraldi, Dr. Baer and Dr. Olah who guided and instructed me on this project. The solar cell fabrication work would not have gone well without the help of my solar knowledge teachers, Sandra Pejic, Bin Liu and Kyle Peters.

A special thanks to CLiPS for funding support. A sincere appreciation to Dr. Ina Matin and

Dr. Timothy Peshek from the MORE Center as well as Professor Roger French from SDLE research center!

111

Figure 3.15 Efficiency life as a function of storing time for inverted organic solar cells

encapsulated by a 4 mil film, then stored at different conditions.

112

3.6 References

[1]. Chapin DM, Fuller CS, Pearson GL. A new silicon p‐n junction photocell for

converting solar radiation into electrical power. Journal of Applied Physics. 1954

May 1;25(5):676-677.

[2]. Sun S. Recent Progress of Organic Photovoltaics. Advances in Solar Energy.

2007;17:74.

[3]. Surek T. Crystal growth and materials research in photovoltaics: progress and

challenges. Journal of Crystal Growth. 2005 Feb 15;275(1):292-304.

[4]. Zhang Y, Yip HL, Acton O, Hau SK, Huang F, Jen AK. A simple and effective

way of achieving highly efficient and thermally stable bulk-heterojunction

polymer solar cells using amorphous fullerene derivatives as electron acceptor.

Chemistry of Materials. 2009 Jun 8;21(13):2598-2600.

[5]. Chambon S, Derue L, Lahaye M, Pavageau B, Hirsch L, Wantz G. MoO3

thickness, thermal annealing and solvent annealing effects on inverted and direct

polymer photovoltaic solar cells. Materials. 2012 Nov 27;5(12):2521-2536.

[6]. Scharber MC, Sariciftci NS. Efficiency of bulk-heterojunction organic solar cells.

Progress in polymer science. 2013 Dec 31;38(12):1929-1940.

113

[7]. Lessons Learnt Report: Market & Technology Analysis for Organic

Photovoltaics; https://arena.gov.au/assets/2014/06/Market-Technology-

Analysis_-Lessons-Learnt.pdf (accessed September 2, 2017)

[8]. Jørgensen M, Norrman K, Krebs FC. Stability/degradation of polymer solar cells.

Solar Energy Materials and Solar Cells. 2008 Jul 31;92(7):686-714

[9]. Yu D, Yang YQ, Chen Z, Tao Y, Liu YF. Recent progress on thin-film

encapsulation technologies for organic electronic devices. Optics

Communications. 2016 Mar 1;362:43-49.

[10]. Jørgensen M, Norrman K, Gevorgyan SA, Tromholt T, Andreasen B, Krebs FC.

Stability of polymer solar cells. Advanced materials. 2012 Feb 2;24(5):580-612.

[11]. Chen HY, Hou J, Zhang S, Liang Y, Yang G, Yang Y, Yu L, Wu Y, Li G.

Polymer solar cells with enhanced open-circuit voltage and efficiency. Nature

photonics. 2009 Nov 1;3(11):649-653.

[12]. Girtan M, Rusu M. Role of ITO and PEDOT: PSS in stability/degradation of

polymer: fullerene bulk heterojunctions solar cells. Solar Energy Materials and

Solar Cells. 2010 Mar 31;94(3):446-450.

[13]. Gevorgyan SA, Jørgensen M, Krebs FC. A setup for studying stability and

degradation of polymer solar cells. Solar Energy Materials and Solar Cells. 2008

Jul 31;92(7):736-745.

114

[14]. Cai W, Gong X, Cao Y. Polymer solar cells: recent development and possible

routes for improvement in the performance. Solar Energy Materials and Solar

Cells. 2010 Feb 28;94(2):114-127.

[15]. Krebs FC, Jørgensen M, Norrman K, Hagemann O, Alstrup J, Nielsen TD,

Fyenbo J, Larsen K, Kristensen J. A complete process for production of flexible

large area polymer solar cells entirely using screen printing—first public

demonstration. Solar Energy Materials and Solar Cells. 2009 Apr 30;93(4):422-

441.

[16]. Norrman K, Madsen MV, Gevorgyan SA, Krebs FC. Degradation patterns in

water and oxygen of an inverted polymer solar cell. Journal of the American

Chemical Society. 2010 Nov 5;132(47):16883-16892.

[17]. Galagan Y, Rubingh JE, Andriessen R, Fan CC, Blom PW, Veenstra SC, Kroon

JM. ITO-free flexible organic solar cells with printed current collecting grids.

Solar Energy Materials and Solar Cells. 2011 May 31;95(5):1339-1343.

[18]. Charton C, Schiller N, Fahland M, Holländer A, Wedel A, Noller K.

Development of high barrier films on flexible polymer substrates. Thin Solid

Films. 2006 Apr 28;502(1):99-103.

[19]. 2V 1W Thin-film Flexible Solar Panel, Seeed,

http://www.seeedstudio.com/depot/2V-1W-Thinfilm-Flexible-Solar-Panel-p-

1856.html (accessed September 2, 2017)

115

[20]. Military Foldable and Rollable Solar Panels, PowerFilm Solar,

http://www.powerfilmsolar.com/military/products__systems/foldable-and-

rollable-solar/ (accessed September 2, 2017)

[21]. Uni-solar PVL-136 (136W) Solar Panel,

http://www.solardesigntool.com/components/module-panel-solar/Uni-

solar/154/PVL-136/specification-data-sheet.html (accessed September 2, 2017)

[22]. Infinity PV foil, Infinity PV, https://www.infinitypv.com/products/opv/foil

(accessed September 2, 2017)

[23]. Goswami DY. Advances in Solar Energy: an Annual Review of Research and

Development. Earthscan; 2007.

[24]. Pentzer EB, Bokel FA, Hayward RC, Emrick T. Nanocomposite “Superhighways

” by Solution Assembly of Semiconductor Nanostructures with Ligand ‐

Functionalized Conjugated Polymers. Advanced Materials. 2012 May

2;24(17):2254-2258.

[25]. Backer SA, Sivula K, Kavulak DF, Fréchet JM. High efficiency organic

photovoltaics incorporating a new family of soluble fullerene derivatives.

Chemistry of materials. 2007 Jun 12;19(12):2927-2929.

[26]. Cheng YJ, Yang SH, Hsu CS. Synthesis of conjugated polymers for organic solar

cell applications. Chemical reviews. 2009 Sep 28;109(11):5868-5923.

116

[27]. Chambon S, Derue L, Lahaye M, Pavageau B, Hirsch L, Wantz G. MoO3

thickness, thermal annealing and solvent annealing effects on inverted and direct

polymer photovoltaic solar cells. Materials. 2012 Nov 27;5(12):2521-2536.

[28]. He Z, Zhong C, Su S, Xu M, Wu H, Cao Y. Enhanced power-conversion

efficiency in polymer solar cells using an inverted device structure. Nature

Photonics. 2012 Sep 1;6(9):591-595.

[29]. Hau SK, Yip HL, Baek NS, Zou J, O’Malley K, Jen AK. Air-stable inverted

flexible polymer solar cells using zinc oxide nanoparticles as an electron selective

layer. Applied Physics Letters. 2008 Jun 23;92(25):225.

[30]. Kim, J.B.; Kim, C.S.; Kim, Y.S.; Loo, Y.-L. Oxidation of silver electrodes

induces transition from conventional to inverted photovoltaic characteristics in

polymer solar cells. Appl. Phys. Lett. 2009, 95, 183301

[31]. Klumbies H. Encapsulations for Organic Devices and their Evaluation using

Calcium Corrosion Tests. (2013)

[32]. Working Principle of Organic Solar Cells, https://www.iapp.de/organische-

elektronik.de/en/?OSCs___Working_Principle (accessed September 2, 2017)

[33]. Part II – Photovoltaic Cell I-V Characterization Theory and LabVIEW Analysis

Code, http://www.ni.com/white-paper/7230/en/ (accessed September 2, 2017)

[34]. Mao Z. Molecular Structures and Device Properties of Organic Solar Cells

(Doctoral dissertation, Case Western Reserve University, 2014).

117

[35]. Padinger F, Rittberger RS, Sariciftci NS. Effects of postproduction treatment on

plastic solar cells. Advanced Functional Materials. 2003 Jan 16;13(1):85-88

[36]. Kim YH, Sachse C, Machala ML, May C, Müller‐Meskamp L, Leo K. Highly

conductive PEDOT: PSS electrode with optimized solvent and thermal post‐

treatment for ITO‐free organic solar cells. Advanced Functional Materials. 2011

Mar 22;21(6):1076-1081.

[37]. Kim H, So WW, Moon SJ. The importance of post-annealing process in the

device performance of poly (3-hexylthiophene): Methanofullerene polymer solar

cell. Solar Energy Materials and Solar Cells. 2007 Apr 16;91(7):581-587.

[38]. Li G, Shrotriya V, Yao Y, Yang Y. Investigation of annealing effects and film

thickness dependence of polymer solar cells based on poly (3-hexylthiophene).

Journal of Applied Physics. 2005 Aug 15;98(4):043704.

[39]. Han D, Yoo S. The stability of normal vs. inverted organic solar cells under

highly damp conditions: comparison with the same interfacial layers. Solar

Energy Materials and Solar Cells. 2014 Sep 30;128:41-47.

[40]. Jia X, Jiang Z, Chen X, Zhou J, Pan L, Zhu F, Sun Z, Huang S. Highly efficient

and air stable inverted polymer solar cells using LiF-modified ITO cathode and

MoO3/AgAl alloy anode. ACS applied materials & interfaces. 2016 Feb

2;8(6):3792-3799

[41]. Wei W, Zhang C, Chen D, Wang Z, Zhu C, Zhang J, Lu X, Hao Y. Efficient

“light-soaking”-free inverted organic solar cells with aqueous solution processed

118

low-temperature ZnO electron extraction layers. ACS applied materials &

interfaces. 2013 Dec 12;5(24):13318-13324.

[42]. Zhang C, Pang S, Heng T, You H, Han G, Lu G, He F, Jiang Q, Zhang J. Stable

inverted low-bandgap polymer solar cells with aqueous solution processed low-

temperature ZnO buffer layers. International Journal of Photoenergy. 2016 Oct

9;2016.

[43]. Sun, Y.; Takacs, C. J.; Cowan, S. R.; Seo, J. H.; Gong, X.; Roy, A.; Heeger, A.

J.Efficient, Air-Stable Bulk Heterojunction Polymer Solar Cells Using MoO(x)

as The Anode Interfacial Layer Adv. Mater. 2011, 23 (19) 2226–2230

[44]. Kyaw AK, Sun XW, Jiang CY, Lo GQ, Zhao DW, Kwong DL. An inverted

organic solar cell employing a sol-gel derived ZnO electron selective layer and

thermal evaporated MoO3 hole selective layer. Applied Physics Letters. 2008 Dec

1;93(22):221107.

119

Chapter 4

Effects of feather-fiber reinforcement on clay aerogel structure, property and applications

A series of low density, highly porous clay/poly(vinyl alcohol) composite aerogels, incorporating keratin fibers derived from chicken feathers, were fabricated via a convenient and eco-friendly freeze drying method. The basic introduction about aerogels has been given in Section 4.1. The introduction of keratin fibers improved the mechanical performance, as well as decreased thermal conductivities, allowing these materials to have potential utility in thermal insulation. Different types of bio-fibers have been compared to confirm the unique and important role of keratin fibers; various processing conditions were applied in order to study processing/structure/property relationships in these materials.

Those detailed results, analysis and discussion are described in Section 4.3. The procedure for this type of aerogel fabrication and characterization is explained in Section 4.2. The mechanical and thermal insulation properties have been demonstrated to increase with amount of keratin fibers.

4.1 Introduction

Aerogels, ultralow-density (0.1 g/cm3 or less) porous materials, are of interest in packaging, absorption and insulation fields. [1-4] First described by Kissler, the removal of volatile organic solvents from gels formed the first-generation of these low-density porous, foam-like materials.[5] The use of a freeze drying process method was introduced by

120

Mackenzie and Call, in the 1950s to fabricate montmorillonite clay aerogels, in turn leading to the wide-spread development of polymer/Clay composite aerogels in the 2000s. [6][7]

Such composite aerogels have been shown to exhibit useful mechanical applications and flammability.[8] In order to improve aerogel mechanical properties, fillers such as crystalline cellulose nanofibers and glass fabrics have been successfully added.[12][13]

Finlay et al. reported the incorporation of biologically based fibers to reinforce mechanical property of aerogels. [14] One of the difficulties shown in this work is detachment between fibers and matrix during compression, which showed a compatible problem. The chemical and physical nature of introduced fibers need to be further study to match the matrix and exhibit a stable performance.

Keratin fibers, derived from chicken feathers, were utilized in this present work, in order to take advantage of their low cost, low density, good thermal properties and wide range of sources. [15] Keratin is insoluble and durable, and is different from other natural fibers such as hemp, silk and soy silk. Keratin fibers from chicken feathers have a complex hierarchical structure including helical microfibrils and macrofibrils, [16][17] creating cites to connect/entangle with added matrix and playing a significant role in improving reinforced systems’ mechanical properties. Utilizing those special physical macrostructures, keratin fibers have been applied into different polymer matrices to improve relative mechanical strengths: Barone et al. have utilized keratin feather fiber to mix with low-density polyethylene matrix to reinforce polymer’s mechanical performance.[18] Martinez-Hernandez et al have demonstrated that a poly(methyl methacrylate) matrix can also be reinforced with keratin biofibers.[19] Besides mechanical properties enhancement, keratin fibers have been studied as a green natural filling material

121 for superior thermal insulating properties.[20][21] The opportunity to combine keratin fibers with high porous aerogel composites could lead to a trend to utilize bio-waste to improve novel materials on mechanical and thermal properties.[22]

In the current work, the clay aerogels reinforced by keratin fibers has been described and fabricated in a robust and environmentally-benign method. Compared to other biologically based fibers, keratin fibers have reinforced clay/poly(vinyl alcohol)composite system, exhibiting higher compressive modulus and useful mechanical strength. A durable mechanical performance has been displayed through stress-strain curves and corresponding reasons have been stated from physical and chemical views. Different conditions, leading to differences in aerogel morphologies, were investigated to evaluate their impact on mechanical and thermal properties. Related thermal conductivity tests based on reinforced aerogels were studied as a potential application as an insulation material, reported for the first time in fiber-reinforcement aerogel system, to the best of our knowledge.

4.2 Experimental Section

4.2.1 Materials

All materials used here were added without further purification. Sodium montmorillonite (PGW grade; Nanocor Inc.), poly(vinyl alcohol) (PVOH, Mw ≈ 108,000,

99.7% hydrolyzed, and a polydispersity index of PDI ≈ 1.7; Polysciences, Inc.), keratin fiber (Featherfiber Corporation) were all used as received. Deionized water was prepared using a Barnstead RO pure low-pressure, reverse osmosis system.

122

4.2.2 Hydrogel preparation

Loadings of clay, PVOH and keratin fiber are given as percentages, based on mass of

DI water employed. For example, in order to formulate a hydrogel containing 5 wt% clay and 2.5 wt% PVOH, a quantity of 200.0 mL of deionized water and 20.0 g sodium montmorillonite were placed into a 1.0 L blender, and then mixed at a high speed (22,000

RPM) for approximately 1 min to wet the clay completely and create a 10 wt% clay hydrogel. Then 200 mL of 5 wt% PVOH solution was added slowly with low speed mixing to create the clay/polymer hydrogel. For the one contained different amount of keratin fibers, appropriate amount of fibers, for example: 4 g, was added into a 400 mL hydrogel made above then mixed with low speed, to obtain fibers/clay/polymer gels comprising 1 wt% fiber, 5 wt% clay and 2.5 wt% PVOH.

4.2.3 Compression samples

Approximately 12 mL of the mixture was transferred into 5 dram (18.5 mL) polystyrene vials and subsequently frozen in a solid carbon dioxide/ethanol bath (-80 ℃), or in a liquid nitrogen bath (-178 ℃). The frozen samples were then dried in the VirTis Advantage EL-

85 freeze dryer, which has a shelf temperature of 25 ℃ and an ultimate chamber pressure of 5 μbar to sublime the ice. The entrie freeze drying process may take 72 - 96 hours, to insure completion of the process.

4.2.4 Thermal conductive samples

Approximately 350 mL of the mixture was poured into a 15 × 15 × 3.8 cm mold constructed of polytetrafluoroethylene frame, covering with cling film and immersing in

123 either a solid carbon dioxide/ethanol bath (-78 ℃), or a liquid nitrogen bath (-178 ℃) until the sample is completely frozen. The frozen mold and sample was then placed in the freeze dryer which has an original temperature of ca. -70 ℃, then the vacuum was applied and temperature of the shelf was raised to 25 ℃ for the duration of drying and the platen-type aerogels were acquired after 3-4 days.

4.2.5 Characterization

After the freeze drying process, dried samples were removed from the vials or mold, and cut with a vertical bandsaw to make sure they had a standard cylinder or flat-platen shape for compression and thermal conductivity tests.

4.2.6 Mechanical testing

Compression testing was conducted on the prepared cylindrical aerogel samples

(measuring about 20 mm in diameter and 20 mm in height), using an Instron model 5565 universal testing machine, fitted with a 1 kN load cell. These tests were conducted at a constant strain rate of 10 mm/min. Five to eight samples for each batches with various compositions were tested for reproducibility. The initial compressive modulus was calculated from the slope of the linear area, obtained from the initial part of the stress-strain curve.

4.2.7 Surface area, porosity and pore size

124

Brunauer–Emmett–Teller analysis (BET test) was applied to study adoption of gas molecules on a solid surface. Aerogel’s surface area was measured using nitrogen gas by a

Micromeritics Tristar Ⅱ unit (Norcross, GA).

Porosity was determined using a density method as Equation 4.1 described:

푉푝표푟푒푠 푃표푟표푠푖푡푦 = × 100% 푉푎푒푟표푔푒푙푠

푚ℎ푦푑푟표푔푒푙 − 푚푎푒푟표푔푒푙 = ( ) × 100% EQ (4.1) 𝜌푖푐푒 × 푉푎푒푟표푔푒푙

The volume of aerogel, 푉푎푒푟표푔푒푙, was calculated from the cylinder aerogel’s height and diameter. The volume of pores, 푉푝표푟푒푠 , which came from the sublimation of ice, was

[23] 3 acquired from the density of ice, 𝜌푖푐푒 = 0.92 g/cm and the mass of freezing water, which was obtained from difference between hydrogel mass, 푚ℎ푦푑푟표푔푒푙 and aerogel mass,

푚푎푒푟표푔푒푙.

For pore size, Cai et al. developed a model for spontaneous inhibition in homogenous porous media and obtained an analytical solution used to describe the relationship between accumulated weight (M ) of liquid imbibed into the porous system and liquid uptake time

(t ) as Equation 4.2 described[31]:

2 2 2 2 3 𝜌푙 퐴 휙 (푆푤푓 − 푆푤푖) 훼 푟푎푒𝜎 cos 휃 푀2 = 푡 EQ (4.2) 2휇휏2

Where 휙 is aerogel’s porosity and 퐴 is the average cross-sectional area of porous keratin-enhanced aerogel, which can be measured directly from dimension of samples. 𝜌푙 is the density of up-taken liquid, 𝜎 is the liquid surface tension and 휇 is the fluid viscosity, those three parameters can be obtained from the datasheet of liquid used in this experiment,

125

DOW corning 200 fluid-100cst. and 휏 is the tortuosity which can be calculated from the

Equation 4.3:

휙 휏 = EQ (4.3) 1 − (1 − 휙)1/3

훼 is the demensionless geometry correction factor, which has a value of 1 for circular cross section. 휃 is the contact angle, which can be measured through contact angle measurement. Swf is the wetting-phase saturation behind the imbibition front, and Swi is the initial wetting-phase saturation which can be regarded as 0 at the initial point in this case. Swf can be calculated from the Equation 4.4, shown below, when the up-taken oil has totally penetrated aerogel samples and reached the saturation point (liquid mass has reached equilibrium):

푚푙 × 𝜌푠 × 𝜌푎푒푟표푔푒푙 푆푤푓 = EQ (4.4) 푚푎푒푟표푔푒푙 × 𝜌푙(𝜌푠 − 𝜌푎푒푟표푔푒푙)

Where 푚푙 is the mass of up-taken oil reaching the saturation point, which can be recorded by balance and 푚푎푒푟표푔푒푙 is the mass of aerogel sample used to absorb oil. 𝜌푠, the density of solid part of aerogel system, can be derived from porosity and density of aerogel and air. In this way, we got the slope of plot 푀2 (푡) and calculated the average/effective pore radius of porous aerogel system, 푟푎푒. The value of 푟푎푒 can be also demonstrated from

SEM images to describe the inner structure morphology.

4.2.8 SEM and μCT

Morphological microstructure of aerogel and keratin fibers was observed using a JEOL

SEM. The observed samples were pre-treated by fracturing in liquid nitrogen then coated

126 with 5nm gold before testing. In order to observe keratin fibers distribution in the whole aerogel systems, micro computed tomography (μCT) was applied to display the inner morphology without damaging the aerogel sample. SkyScan 1172 microCT was used to acquire a series of projection images and reconstructed the obtained images into virtual 3D models.

4.2.9 Thermal conductivity test

The thermal conductivity of the aerogel is measured with a conventional heat flow meter setup based on ASTM C 518 as shown in Figure 4.1. This method and model has been demonstrated working well for relatively low thermal conductivity materials, especially compatible with aerogel systems which has an averaged thermal conductivity smaller than

0.1 W/mK.[24] Target aerogel sample designed as a thin board type (15 × 15 × 1 cm) was inserted between two boards of same material with known thermal conductivity, here are nylon plates with 푘0 = 0.25 W/mK. Target samples were compressed completely to avoid the influence from thermal contact resistance, shown as Figure 4.2.

Figure 4.1 Schematic of the heat flow meter setup for thermal conductivity test

127

Figure 4.2 Picture of the heat flow meter setup for thermal conductivity test

Heat was sent through these layer samples with in an approximated one-dimensional manner, by applying on known nylon reference plates and unknown aerogel sample with high aspect ratio (length/width to thickness is maintained at least 10:1, even higher) and evenly distributed to three plates (two nylon plates and one aerogel plate) via a metal heat expander/spreader. In order to lead one-dimensional conduction through the stacks more than losing heat through sides, proper thermal insulation by PS foams and cotton had been set up around the sides and over the heater. Then the whole setup has been kept in an enclosed space. The temperature across each of the three plates could be recorded using four thermocouples, located at the center of two sides of nylon plates. Thin thermocouple wires and grooves formed on the reference plates are encouraged to eliminate the gap

128 between two plates in order to ignore the heat loss. The rate of heat transfer Q across the each plate was written according to Fourier’s Law, shown as Equation 4.5

∆푇 Q = kA EQ (4.5) 퐿

The heat transfer Q across all three plates ideally was equivalent, assuming a perfect one-dimensional model with no heat loss from the sides. And in actual tests, if the temperature drops across the two refnce nylon plates (∆푇0,1, ∆푇0,2) were close enougostly

∆푇0,2 was slightly smaller than ∆푇0,1, this setup can be regarded as an ideal 1-D model due to similar heat transfer Q across the reference plates. Based on this theory we could derive the Equation 4.6:

푄0 ≅ 푄푠 EQ (4.6)

Where 푄0 was average heat transfer across two nylon plates and 푄푠 was actual heat transfer across aerogel stack, and in realty, the value of 푄푠 would be between the ones across two nylon plates. Therefore, the averaged of 푄0 gave a good and actuate measurement of heat flux through the target samples. Combining with Equation 4.5,

Equation 4.6 can be rewritten as Equation 4.7:

∆푇0 ∆푇푠 푘0퐴0 ≅ 푘푠퐴푠 EQ (4.7) 퐿0 퐿푠 where ∆푇0 is the average of two temperature drops across nylon plates (∆푇0,1, ∆푇0,2), ∆푇푠 is the temperature drops across aerogel samples; 퐿0 and 퐿푠 are the thickness of nylon plates and aerogel plate, respectively; 퐴0 and 퐴푠 are the area of nylon plates and aerogel plate.

The thermal conductivity of the target aerogel sample, 푘푠, could be calculated based on known material, 푘0, and combination of thickness, area, as well as temperature drops. In order to obtain the reliable data at a dynamic equilibrium status, a constant heat flux was

129 applied overnight (~12 hrs) to ensure temperatures detected by thermocouples reached steady-state values. Data was gathered for one hour to obtain an average temperature drop for each section.

4.3 Results and discussion

4.3.1 Different types of fibers and fiber amounts

The compressive stress-strain curves of the aerogels incorporating various amounts of keratin fibers is shown in Figure 4.3, and exhibits the same classical fiber-reinforced aerogel behavior as reported previously by Finlay et al. [14] With increasing amounts of fibers, the compressive moduli increased monotonically, with values similar to those observed with hemp or silk fiber reinforcement (Table 4.1).

Figure 4.3 Compressive stress-strain curves for 5% clay aerogel incorporating varying amounts of keratin fibers Comparison on mechanical strength (compressive modulus) of

different fibers reinforced aerogels

130

Table 4.1 Comparison on mechanical strength (compressive modulus) of different fibers

reinforced aerogels

Fiber Types Fiber Content Hemp[14] Silk[14] Keratin (based on 5% Clay) Compressive Compressive Compressive Modulus (kPa) Modulus (kPa) Modulus (kPa)

0 14±8 14±8 14±8 1 179±26 166±20 186±20 3 ------607±70 5 1292±277 1771±327 ----

Based on the performance of keratin fiber-reinforced clay aerogels, experiments incorporating the water-soluble polymer, PVOH, were carried out. Clay/polymer/fiber composites were reported to significantly improve aerogel mechanical strengths and similar results were obtained in the present system as well (Table 4.2). It can be noted obviously that keratin fiber demonstrated superior reinforcement to those fibers (hemp, soy silk, silk) previously tested.

131

Table 4.2 Comparison on mechanical strength of different fibers reinforced aerogels

(incorporated with 2.5% PVOH)

Fiber Fiber Types Content Hemp[14] Silk[14] Keratin (based on 5% Clay, Compressive Compressive Compressive Density Density Density 2.5% Modulus Modulus Modulus PVOH)

0 0.07 1600±208 0.07 1600±208 0.067 3500±580

1 0.082 2688±207 0.084 3056±229 0.077 11700±780

3 ------0.083 13700±1400

5 0.101 8560±487 0.121 8857±1521 0.105 15700±1700

Modulus in kPa, Density in g/cm3

The small inflection points or plateaus (Figure 4.4), which had been observed in the stress-strain curves of other-fiber reinforced aerogels, were not present in the aerogels which incorporated keratin fibers (Figure 4.5). Those points or plateau were attributed to delamination of the fiber from the matrix during the compression. It is clearly to see that keratin fiber had the stronger connection and adhesion to Clay/PVOH matrix than the rest of fibers. From its chemical nature, protein-based keratin, was expected to interact more tightly with polar polymer/clay matrix through hydrogen bond, whereas hemp, a cellulosic- nature fiber, may not expected to adhere tightly to the polar matrix;[25] Keratin also possessed many “knots” or “hooks”, dispersed evenly on the fibers, whereas other fibers are quite smooth and straight on their surface (shown in Figure 4.6). Those special

132 structures provide an opportunity for the keratin fibers to entangle with the matrix and renders them more difficult to detach from aerogels during compression. Both chemical and physical factors therefore contribute to keratin fibers’ superior enhancement performance.

Figure 4.4 Examples to display inflection points or plateau forms in 5%

Clay/2.5%PVOH aerogels with varied amount of soy silk fibers (top)[14] and in 5% Clay

aerogel with 1% hemp fibers (bottom)[25]

133

Figure 4.5 Stress-strain curves for 5% clay/2.5% PVOH aerogel incorporated with varied amount of keratin fibers

Figure 4.6 Morphology comparison of different fibers: (a) keratin fiber; (b) Hemp fiber;

Even dispersion of fibers in clay/polymer matrixes is difficult, because of the fiber dimensions and high matrix suspension viscosities. Although keratin fibers can be separated among the matrix through mixing the suspension with relatively lower speed, random and unorientated fiber dispersion existed within the intermediate hydrogel and the subsequent aerogel, as is confirmed from SEM images (Figure 4.7b) and μCT video

(Supplementary information). A comparison SEM images of aerogel inner structure as is shown in Figure 4.7, demonstrates the difference morphology with and without keratin

134 fibers: fibers penetrating the inner layers brought the instabilities and minimal disruption, whereas, spanning through the layers would enhance the strength. (Figure4.7c).

Figure 4.7 Comparison of a 5% Clay/2.5% PVOH aerogel (a) and a 5%Clay/2.5%

PVOH/1% keratin fibers dispersed randomly (b) and in an oriented manner (c).

4.3.2 Different freezing conditions and influence of morphology

In previous work reported by Wang et al. it has been demonstrated the freezing process used for aerogel fabrication significantly influences material morphology.[28] External temperature and internal viscosity have been two main aspects to determine the freezing rate, which includes ice nucleation and propagation. Different freezing conditions have been applied to hydrogels with varying amounts of keratin fibers in the present work.

Material density and mechanical property values are presented in Table 4.3. Regardless of freezing conditions, the densities of aerogels prepared with same amount of keratin fibers are similar. Samples frozen at the relatively higher temperature (-78 ℃) exhibited higher moduli than those with same composition frozen at a lower temperature (-178 ℃). In order to verify this phenomenon, same recipe of aerogel samples with 1% keratin fibers were frozen at -20 ℃ (freezing chamber of lab-use refrigerator), a much higher freezing temperature. The same conclusion can be also obtained that higher freezing conditions provide higher chance to fabricate stronger aerogel samples.

135

Table 4.3 Mechanical strength of different fibers reinforced aerogels frozen at different

temperatures. (incorporated with 5% Clay and 2.5% PVOH)

Keratin Fiber Content 0% 1% 3% 5% (based on 5% Clay, 2.5%PVOH)

Freezing -78 -178 -20 -78 -178 -78 -178 -78 -178 Temperature

0.06 0.07 0.067 5 0.077 0,077 6 0.083 0.083 0.105 Bulk ± crac ± ± ± ± ± ± ± Density <0.00 k 0.00 0.001 0.001 0.00 0.001 0.001 0.002 1 1 1

1260 1370 1020 1170 15700 Compressiv 3500 2000 0 7400 0 0 0 ±170 ---- e Modulus ±580 ±700 ± ±950 ± ± ±780 0 1000 1400 1300

Temperature in ℃, Modulus in kPa, Density in g/cm3

Morphological studies of aerogels frozen at different temperature were conducted with

SEM. Corresponding graphs, shown as Figures 4.8 and 4.9, have exhibited details which can be used as an explanation for mechanical performance, from inner structure view.

Samples frozen at -78 ℃ displayed a typical lamellar structure, as illustrated in Figure

4.8a, 4.9a and 4.9c, presumably due to the dominant effect of ice growth winning over ice nucleation, although ice nucleation still happened locally to form pores in each layer, shown as Figure 4.8b. Whereas relatively lower temperatures, such as -178 ℃, provide higher freezing rates, reducing the ice crystal growth time and hindering ice propagation.

Thus, these lower temperatures produce smaller ice crystals, then form smaller pores sizes

136 after the freeze drying process, as illustrated in Figure 4.8d, 4.9b and 4.9d. It is also noted that samples frozen at -178℃ have a higher tendency to crack or form defects, as displayed in Figure 4.8c, resulting from the interactions among rapid freezing rate, random freezing directions and high viscosity from PVOH and keratin fibers.

Figure 4.8 a) Lamellar and granular inner structure of 5%Clay 2.5% PVOH aerogels without fibers, frozen at -78 ℃ (a) and -178 ℃ (c). 10 times zoom in to exhibit details of

each sample (b) and (d)

The porosity of an aerogel is the percentage of the volume of pores over total volume of sample. Aerogels generally exhibit high porosities and can be simply calculated through

Equation 4.1, based on density. All the samples in Table 4.4 display porosities over 91%.

BET data provides information about surface area: with increasing amount of fibers, surface area of aerogels freezing in same condition increased, which attributed to keratin

137 fiber’s introduction to create more cites; whereas another obvious factor to impact the surface area is the freezing temperature in our research: it has been clearly shown aerogels frozen at lower temperature are inclined to produce smaller pores, (Figure 4.8 and 4.9).

With similar bulk volumes and porosities, the aerogels having smaller voids are expected to have higher surface areas. The BET data reported herein show that aerogel samples frozen from liquid nitrogen possess almost double the surface areas as those of the same composition frozen at solid carbon dioxide/ethanol bath temperatures.

Table 4.4 Surface Area and porosity of different fibers reinforced aerogels frozen at

different temperatures. (incorporated with 5% Clay and 2.5% PVOH)

Keratin Fiber Content (based on 0% 1% 3% 5% Clay, 2.5%PVOH)

Freezing -78 -178 -78 -178 -78 -178 Temperature 9.90± 18.56± 10.52± 19.48± 12.67± 21.03± Surface Area 0.04 0.09 0.09 0.22 0.09 0.13 Porosity 92.68 93.27 92.28 92.40 91.56 91.87

Temperature in ℃, Modulus in kPa, Density in g/cm3

138

Figure 4.9 Inner structure of 5%Clay 2.5% PVOH aerogels with 1% keratin-fibers adding, frozen at a) -78 ℃ and b) -178 ℃; with 3% keratin-fibers adding, frozen at c) -78

℃ and d) -178 ℃.

4.3.3 Thermal insulation performance

Thermal conductivities of aerogel samples produced with or without keratin fibers were measured and calculated through the heat flow meter method described in Section 4.2.9

(Table 5). Aerogel samples (5% Clay/2.5% PVOH) frozen at -78 ℃ exhibited thermal conductivity values as low as 0.075 W/mK, same composition samples frozen in liquid nitrogen has a relatively lower value, 0.070 W/mK, which can be attributed to more small- size pores formed in the aerogel. As a well-known nature thermal insulation material, different amount of keratin’s introduction improved thermal insulation effect and thermal conductivity values have been decreased to 0.066 W/mK for 1% fiber reinforced aerogels,

139 and 0.058 W/mK for 3% fiber reinforced one. Different freezing methods made an impact on thermal conductivity of keratin-reinforced aerogels as well. Figure 4.8 and 4.9 have shown a trend that pores generated from frozen at -178 ℃ were smaller than the one frozen at -78 ℃, which means the former were expected to postpone more air or more heat flow through aerogel samples than the latter, in order to create a better insulation effect.

Table 4.5 Thermal conductivity value of different fibers reinforced aerogels frozen at

different conditions. (incorporated with 5% Clay and 2.5% PVOH)

Keratin Fiber Content (based on 0% 1% 3% 5% Clay, 2.5%PVOH)

Freezing -78 -178 -78 -178 -78 -178 Temperature Thermal 0.075± 0.070± 0.070± 0.066± 0.067± 0.058± Conductivity, k 0.001 <0.001 0.03 0.002 0.003 0.002

Temperature in ℃, Thermal Conductivity, k in W/mK

Increasing levels of keratin introduced into the aerogel systems helps to decrease thermal insulation, at the same time the hydrogel processing became much more difficult once fiber contents above 3%. Increasing viscosity led to uneven fiber distributions and increased levels of defects in the associated hydrogels and aerogels. Figure 4.10 and the video show the dispersion of keratin fibers in aerogels, obtained from micro-CT. Fibers disperse randomly in the aerogel systems and due to freezing process, keratin fibers are inclined to

140 accumulate on the edge of sample. Micro-CT has also been demonstrated as an intuitive method to discover fiber dispersion and inner structure of aerogels without cutting or damaging samples.[29][30]

Figure 4.10 Keratin fibers dispersion in 5%Clay/2.5%PVOH incorporated with 3% fibers, conducted by μCT scanning

4.4 Conclusion

A series of highly porous, low density aerogels based on aqueous mixtures of sodium montmorillonite clay and poly(vinyl alcohol) were produced incorporating differing levels of keratin feather fibers, using an eco-friendly freeze drying process. These reinforced aerogels exhibited an increasing compressive modulus values and lower thermal conductivity values with levels of fiber contents. Keratin fibers used in the present work played a more significant role to enhance aerogels compared to other fiber fillers, because

141 of: 1) physical “knots” or “hooks” in their structure, leading to entanglement with the matrix; 2) a high level of interaction with the polar polymer/clay matrix through hydrogen bond due to the nature of protein. With 5% keratin fiber adding, 5%Clay 2.5%PVOH aerogels can be reinforced from 3.9 MPa to 15.7 MPa, close to 4-fold improvement, and without delamination of fibers from the matrix during compression. Meanwhile, different processing conditions, played a role to control the morphology of aerogel inner structure, then impacted the mechanical property even in same composition. Lamellar structure, which was inclined to generate at relatively higher freezing temperature (-78 ℃), provided a stable layer structure than random granular structure formed at relatively lower temperature (-178 ℃). Due to porous inner structure and well-known insulation nature, aerogels incorporated with keratin fibers had an application to use as a low thermal conductivity material. With 3% fiber adding, thermal conductivity can be decreased from

0.075 W/mK to 0.058 W/mK, suggesting use of these reinforced materials in insulation applications.

4.5 Acknowledgement

Many thanks are due to Steven Hostler and Boran Zhao for their collaboration in thermal conductive test and pore size characterization, Also, I would like to gratefully acknowledge

Dr. Hossein Ghassemi’s encouragement and support. The supervision and suggestions from Dr. Schiraldi and Dr. Abramson are gratefully appreciated.

142

4.6 References

[1]. Randall D, Lee S. The Polyurethanes Book, vol. 14. Wiley: New York 2002.

[2]. Bandi S, Bell M, Schiraldi DA. Temperature-responsive clay aerogel− polymer

composites. Macromolecules. 2005 Nov 1;38(22):9216-9220.

[3]. Gawryla MD, Liu L, Grunlan JC, Schiraldi DA. pH tailoring electrical and

mechanical behavior of polymer–clay–nanotube aerogels. Macromolecular rapid

communications. 2009 Oct 1;30(19):1669-1673.

[4]. Akimov YK. "Fields of application of aerogels." Instruments and Experimental

Techniques 46, no. 3 (2003): 287-299.

[5]. Kistler SS. Coherent expanded-aerogels[J]. The Journal of Physical Chemistry,

1932, 36(1): 52-64.

[6]. Mackenzie RC. Clay-Water Relationships. Nature 1953, 171, 681-683.

[7]. Call F. Preparation of Dry Clay-Gels by Freeze drying. Nature 1953, 172, 126.

[8]. Wang TW, Sun H, Long J, Wang YZ and Schiraldi DA. "Biobased Poly (furfuryl

alcohol)/Clay Aerogel Composite Prepared by a Freeze drying Process." ACS

Sustainable Chemistry & Engineering 4.5 (2016): 2601-2605.

[9]. Chen HB, Wang YZ, Sánchez-Soto M, Schiraldi DA. Low flammability, foam-

like materials based on ammonium alginate and sodium montmorillonite clay.

Polymer. 2012;53(25):5825-5831.

143

[10]. Arndt EM, Gawryla MD, Schiraldi DA. Elastic, low density epoxy/clay aerogel

composites. Journal of Materials Chemistry. 2007;17(33):3525-3529.

[11]. Sun H, Schiraldi DA, Chen D, Wang D, Sánchez-Soto M. Tough polymer

aerogels incorporating a conformal inorganic coating for low flammability and

durable hydrophobicity. ACS applied materials & interfaces. 2016 May

11;8(20):13051-13057.

[12]. Gawryla MD, van den Berg O, Weder C, Schiraldi DA. Clay aerogel/cellulose

whisker nanocomposites: a nanoscale wattle and daub. Journal of Materials

Chemistry. 2009;19(15):2118-2124.

[13]. Wang Y, Al‐Biloushi M, Schiraldi DA. Polymer/clay aerogel‐based glass fabric

laminates. Journal of Applied Polymer Science. 2012 May 15;124(4):2945-2953.

[14]. Finlay K, Gawryla MD, Schiraldi DA. Biologically based fiber-reinforced/clay

aerogel composites. Industrial & Engineering Chemistry Research. 2008;47(3):

615-619

[15]. Martínez-Hernández AL, Velasco-Santos C, De-Icaza M, Castano VM.

Dynamical–mechanical and thermal analysis of polymeric composites reinforced

with keratin biofibers from chicken feathers. Composites Part B: Engineering.

2007;38(3):405-410.

[16]. Martinez-Hernandez AL, Velasco-Santos C, De Icaza M, Castano VM.

Microstructural characterisation of keratin fibres from chicken feathers.

International journal of environment and pollution. 2005; 23(2):162-178.

144

[17]. Martínez-Hernández AL, Velasco-Santos C, de Icaza M, Castaño VM.

Hierarchical microstructure in keratin biofibers. Microscopy and Microanalysis.

2003 Aug;9(S02):1282-1283.

[18]. Barone JR, Schmidt WF. Polyethylene reinforced with keratin fibers obtained

from chicken feathers. Composites Science and Technology. 2005;65(2):173-

181.

[19]. Martínez-Hernández AL, Velasco-Santos C, de Icaza M, Castaño VM.

Mechanical properties evaluation of new composites with protein biofibers

reinforcing poly (methyl methacrylate). Polymer. 2005;46(19):8233-8238.

[20]. Gao J, Yu W, Pan N. Structures and properties of the goose down as a material

for thermal insulation. Textile Research Journal. 2007;77(8):617-626.

[21]. Mao N, Russell SJ. The thermal insulation properties of spacer fabrics with a

mechanically integrated wool fiber surface. Textile Research Journal. 2007

Dec;77(12):914-922.

[22]. Wang Y, Schiraldi DA. Foam-like materials produced from milk and sodium

montmorillonite clay using a freeze-drying process. Green Materials. 2013

Mar;1(1):11-15.

[23]. Lide DR. Handbook of Chemistry and Physics, 86th edn. CRC. 2005.

[24]. Hostler SR, Abramson AR, Gawryla MD, Bandi SA, Schiraldi DA. Thermal

conductivity of a clay-based aerogel. International Journal of Heat and Mass

Transfer. 2009;52(3):665-669.

145

[25]. Finlay KA, Gawryla MD, Schiraldi DA. Effects of fiber reinforcement on clay

aerogel composites. Materials. 2015;8(8):5440-5451.

[26]. Surapolchai W, Schiraldi DA. The effects of physical and chemical interactions

in the formation of cellulose aerogels. Polymer bulletin. 2010;65(9):951-960.

[27]. Viggiano III RP, Schiraldi DA. Fabrication and mechanical characterization of

lignin-based aerogels. Green Materials. 2014;2(3):153-158

[28]. Wang Y, Gawryla MD, Schiraldi DA. Effects of freezing conditions on the

morphology and mechanical properties of clay and polymer/clay aerogels.

Journal of Applied Polymer Science. 2013;129(3):1637-1641.

[29]. Neugebauer A, Chen K, Tang A, Allgeier A, Glicksman LR, Gibson LJ. Thermal

conductivity and characterization of compacted, granular silica aerogel. Energy

and buildings. 2014;79:47-57.

[30]. Martins M, Barros AA, Quraishi S, Gurikov P, Raman SP, Smirnova I, Duarte

AR, Reis RL. Preparation of macroporous alginate-based aerogels for biomedical

applications. The Journal of Supercritical Fluids. 2015;106:152-159.

[31]. Cai J, Perfect E, Cheng CL, Hu X. Generalized modeling of spontaneous

imbibition based on Hagen–Poiseuille flow in tortuous capillaries with variably

shaped apertures. Langmuir. 2014 May 1;30(18):5142-5151.

146

Chapter 5

The relation between rheological properties of gels and mechanical properties of aerogels

The relationship between rheological property of gels, and the mechanical properties of aerogels fabricated from the precursor gels is explored in this chapter. It is very important to understand rheological properties because they have a determined effect on the form of aerogels and their processing behaviors. In this project, we used clay, poly(vinyl alcohol) and ammonium alginate to prepare a series of the colloidal suspension under different conditions then freeze dried the gels to obtain target aerogel products. We have verified that different compositions and conditions applied on those colloidal suspension systems may bring different layer structures morphology of aerogel products then different mechanical performance. These phenomena could be observed from scanning electron microscope and confocal microscopy, then re-confirmed from mechanical tests. Emphases are placed on exploring the different factors in rheology, for instance, how the molecular weight of the polymer changes the rheological behavior, and provide sufficient evidence to support our idea. It is expected that we will be able to establish a “bridge” correlating the rheology with the structure and properties enabling one to make a preliminary forecast of final properties before converting suspension into product. The background knowledge of rheology and related aerogels will be introduced briefly in Section 5.1. The procedure for colloidal suspension preparation, aerogel fabrication and characterization will be

147 explained in Section 5.2. Those detailed results, analysis and discussion will be exhibited in the Section 5.3.

5.1 Introduction

5.1.1 Materials optimization Loop

Material science is a multidisciplinary field which studies in part the relationship among processing, structure, properties and performance of materials[1][2]. The materials optimization loop or materials optimization triangle, seen in Figure 5.1, describes the experimental strategy of every material, which will be made into a part by a process, spontaneously creating a multilevel structure, which leads a series of properties, characteristics or behaviors in certain environment; it can finally be used in different end uses, according to its performance[3]. By learning this loop, suitable process conditions could be selected to fabricate materials with certain structure then properties, and finally meet desired performance.

Figure 5.1 The strategy to study material science: Materials optimization loop

148

We have studied the optimization loop, utilizing aerogels as the target material. Two main process procedures exist during the aerogels’ fabrication. One is freeze drying; water could be removed from a frozen clay-polymer hydrogel through a sublimation process following a three phase-diagram shown in Figure 5.2[4].

Figure 5.2 Water three phase diagram to display some common routes: freeze-drying,

conventional drying, freezing, etc[4].

149

Water is extracted from solid ice to water vapor directly without going through liquid status, avoiding inner structure collapse from water aggregation due to capillary forces[5].

The other step is hydrogel processing; that is also the procedure we can alter manually to achieve certain properties of hydrogel, reflecting from ultimate aerogel product. Rheology, which is used as a primary experimental method to study hydrogel’s viscoelastic property, becomes the target we will study in this chapter[6].

5.1.2 Rheological properties

To assess hydrogel viscoelastic properties, small deformation rheology experiments were conducted. The measurements need to be carried out in linear viscoelastic region which ensures the sample properties are independent of imposed strain or stress. Through measuring linear viscoelastic properties, it helps us bridge the gap between molecular structure and product performance[8]. One approach of the typical rheological tests to discover gels’ viscoelastic properties is oscillatory frequency sweep measurements, running under a constant oscillation amplitude (constant strain or stress) and constant temperature. The samples’ behavior of viscoelastic parameters, such as storage modulus

(G’), loss modulus (G”), the tangent of the phase angle (tanδ) and complex viscosity (η*) were monitored and recorded against frequency.

A detailed introduction about viscoelastic parameters can be found from a number of sources. To re-emphasize, G’ means storage (or elastic) modulus which is a measure of energy stored during a deformation, here is oscillatory shear; it can be simply seen as reversibly stored energy due to elastic deformation; while G” means loss (or viscous) modulus which is a measure of energy dissipated during a deformation. It also can be

150 simply seen as irreversibly dissipated energy attributed from viscous flow. η*, is a parameter based on complex modulus (G*) (see Equation 5.1), an overall resistance to deformation, integrated G’ and G” and written as Equation 5.2,

G* η* = EQ (5.1) ω

G* = G'+iG" EQ (5.2)

G* or η*, for this integration reason, are good visible indicators to attributes such the flexibility or stiffness of a material during formation. One more advantage of complex viscosity is, when plotted as a function of angular frequency, it can be equated to a shear viscosity vs shear rate profile for some materials utilizing Cox-Merz rule[9]. This process also provides a way to calculate the effective viscosity/shear rate profiles when it is hard to test effective viscosity from traditional methods. For our system, we have tested various hydrogel matrix and found the typical profiles as shown in Figure 5.3, which is a representative profile of our hydrogel matrix and G’> G”, consistent with a matrix which behaves like an elastic solid[10]. No cross-over point between G’ and G” observed; no region shift, the matrix behaved like a stable visco-elastic solid during the frequency range and it was a typical behavior of soft gel or dispersion with weak structure. For our work in this chapter, therefore, we focus on evaluating complex viscosity (η*) performance against frequency, as a target parameter to study hydrogel’s rheological property.

151

Figure 5.3 Oscillatory frequency sweep measurement on a representative hydrogel

system: complex viscosity, elastic modulus and loss modulus vs frequency.

5.1.3 Objectives

In the following contents, we aimed to develop the strategy of materials optimization loop into our study: through evaluating the hydrogel processing property performances to build a bridge to structure and mechanical property of aerogels. Different hydrogel systems and influence factors were applied to lead the complex viscosity changes; by monitoring

η* performance we expect to predict final product’s performance without further fabrication.

5.2 Experimental Section

5.2.1 Materials

152

All materials used here were added without further purification. Sodium montmorillonite (PGW grade; Nanocor Inc.), various molecular weight poly(vinyl alcohol)

(PVOH, Mw ≈ 31,000~50,000 (low Mw) /130,000 (high Mw) /146,000~180,000 (ultra high Mw), 99+% hydrolyzed, from Sigma Aldrich, as well as Mw ≈ 108,000 (medium

Mw), 99.7% hydrolyzed and a polydispersity index of PDI ≈ 1.7, from Polysciences, Inc.), and ammonium alginate (Pfaltz & Bauer)were all used as received. Deionized water was prepared using a Barnstead RO pure low-pressure, reverse osmosis system.

5.2.2 Hydrogel preparation

Loadings of clay, PVOH and alginate were given as percentages, based on mass of DI water employed. For example, in order to formulate a hydrogel containing 5 wt% clay and

5 wt% PVOH, a quantity of 200.0 mL of deionized water and 20.0 g sodium montmorillonite were placed into a 1.0 L blender, and then mixed at a high speed (22,000

RPM) for approximately 2 min to wet the clay completely and created a 10 wt% clay hydrogel. 200 mL of 10 wt% PVOH solution was then added slowly with low speed mixing to create the clay/polymer hydrogel with 5 wt% clay and 2.5 wt% PVOH. For aerogels which contained 5% ammonium alginate, appropriate amounts of raw materials, for example, 20 g, were added into a 400 mL hydrogel made above then mixed by egg-beater with various speed tuning through voltage transformer, to obtain alginate/clay/polymer gels comprising 5 wt% alginate, 5 wt% clay and 5 wt% PVOH. The whole procedure for this part can be described in Figure 5.4 (a).

5.2.3 Viscosity testing

153

The complex viscosity of the hydrogel specimen was measured using an advanced rheometric expansion system (ARES) strain controlled rheomoeter (TA instruments). 1~2 ml hydrogel specimens were carefully loaded between two parallel plate fixtures, which the bottom plate is fixed and the top one rotates with an imposed torque. The test was set to run an oscillation frequency sweep from (0.1 rad/s to 100 rad/s), with presetting a constant amplitude. Before starting a test, the excess gel need to be cleaned to ensure a clear and smooth interface. And results typically plotted as G’, G’’ or complex viscosity vs frequency. The procedure can be illustrated in Figure 5.4 (b).

5.2.4 Aerogel formation

Approximately 12 mL of the mixture was transferred into 5 dram (18.5 mL) polystyrene vials and subsequently frozen by immersing the vials into a solid carbon dioxide/ethanol cooling bath (-80 ℃). The frozen samples were then dried in the VirTis Advantage EL-85 freeze dryer, which has a shelf temperature of 25 ℃ and an ultimate chamber pressure of

5 μbar to sublime the ice. The entire freeze drying process may take 72 - 96 hours, to insure completion of the process. This procedure can be seen from Figure 5.4 (c).

5.2.5 Mechanical testing

Compression testing was conducted on the cylindrical aerogel samples (measuring about

20 mm in diameter and 20 mm in height), using an Instron model 5565 universal testing machine, fitted with a 1 kN load cell. These tests were conducted at a constant strain rate of 10 mm/min. Five to eight samples for each batch with various compositions were tested

154 for reproducibility. The initial compressive modulus was calculated from the slope of the linear area, obtained from the initial part of the stress-strain curve.

Figure 5.4 Illustration of aerogel fabrication and rheology test.

5.2.6 SEM and Con-focal

155

Morphological microstructures of the aerogels were observed using a JEOL scanning electron microscope. The samples were pre-treated by fracturing in liquid nitrogen then sputter coated with 5nm gold before testing. In order to observe the inner structure as a 3D view, a Leica confocal laser microscope was applied to display the inner morphology without breaking samples inside. The specimens were pre-cut by razor and double-taped on the glass slide in order to load for scanning. Through multi-scanning in a pre-set range, one 3D image could be obtained by integration all acquired 2D X-Y pictures in Z-direction.

5.3 Results and discussion

5.3.1 Different processing rate (stirring speed) effect

The first factor which may impact rheological property is stirring speed, which is used to control the processing rate to disperse the hydrogel matrix. The hydrogel system we used was 5% ammonium alginate/ 5% clay/ 5% PVOH (Mw=31k). Ammonium alginate was selected to amplifying the effect of viscosity due to the relatively inconspicuous viscosity performance of the original clay-PVOH matrix system. Figure 5.5 (a) shows that the hydrogel exhibited a general trend of increasing viscosity with stirring speed increase.

Figure 5.5 (b) shows the similar phenomenon that the modulus of aerogels made from various stirring rate of hydrogel, had a similar trend of increasing with stirring speed increase. The viscosity of hydrogel stirring at 650 and 800 rpm had a similar performance which was also reflected simultaneously from a similar mechanical strength of the respective aerogel samples. All the aerogels were produced from the same concentration and composition of “alginate-clay-PVOH” matrix which resulted in similar aerogel

156 densities (see Figure 5.5 (c)) and the specific modulus had the same performance as overall modulus (see Figure 5.5 (d)).

Figure 5.5 Viscosity of hydrogels (a), compressive moduli of respective aerogels (b),

density vs. stirring speeds (c), specific modulus vs. stirring speeds (d).

The processing conditions that were examined above brought about a change in the viscosity then finally, the aerogel mechanical properties; the morphology of aerogel inner structure undoubtedly become the bridge to link the former and the latter. Figure 5.6 presents SEM images of an aerogel sample produced from hydrogels stirring at 350 rpm and 1000 rpm. From the inner structures, it is evident that the higher stirring speed mixed the matrix more thoroughly and coated the matrix more uniformly. As stirring speeds increase, molecular chains have a higher chance to entangle, increasing viscosity. More

157 networks and layer structures were linked and generated during the mixing. Attributed to more organized structures formation, the respective mechanical performance has been improved and reflected from increasing modulus.

Figure 5.6 SEM images to show the difference inner structure morphology of aerogels made from same composition: 5%Alginate/5%Clay/5%PVOH with different stirring rate,

350 rpm (a) and 1000 rpm (b).

5.3.2 Different standing time effect

The second factor we will evaluate is the standing time between mixing and freezing of the hydrogels. We generally freeze the hydrogel samples immediately or shortly after stirring. The hydrogel system we used in this experiment was 5% Clay/ 2.5% PVOH

(Mw=31k) one. In this part of work, we left the hydrogel for 0 day (immediately freeze), 1 day (then freeze) and 3 days (then freeze). The parts we tested for aerogel modulus were from the lower halves of the monoliths, and correspondingly we took a sample from bottom of hydrogel for rheology testing.

158

Figure 5.7 (a) and (b) show that viscosities of hydrogels and moduli of respective aerogels increased in the same trend with standing time increased. Different from stirring speed effect, the density of hydrogels changed with longer standing time (see Figure 5.7

(c)) implied the composition changed at the bottom of hydrogel. The specific modulus had also an increased trend to demonstrate the modulus increased excluding density influence from Figure 5.7 (d).

Figure 5.7 Graphs to display the related properties of hydrogels storing for a various standing time. Viscosity (a), mechanical performance (b), different density (c) increasing

trend of specific modulus (d).

159

Density changes implied the composition changes which resulted from clay sedimentation. Since our hydrogels are a type of high viscous suspension, and after standing for a sufficient period of time, clay sediments would be inclined to deposit to bottom because of gravity. It can be explained why the bottom aerogel from hydrogel left for 3 days had a higher density. The structure of lower part had enhanced by more clay, bringing higher modulus and specific modulus. From confocal images (seen from Figure

5.8), we have observed the aerogels from immediately-freezing samples kept the classic lamellar structure while samples fabricated after 3 days wait prior to freezing exhibited a stack of clay deposition[11-14]. The aerogel modulus increased as a result of the matrix stack effect and after longer time such as a week, no more organized structure existed and collapse happened inside, which resulted in a poor mechanical performance.

Figure 5.8 Confocal microscopy images to show different morphology of aerogels made from hydrogel with various standing time: 0 day, freeze-immediately (a) and 3-day-

later freezing (b).

160

We have also used confocal microscopy to scan multi-images in order to integrate 3D images which show inner structures without destroying them. Figure 5.9 illustrates a 50x- multi-scanning image integration as a 50-layer images stacking. A clear lamellar structure of aerogels with big layer spacing has exhibited through this motion graph.

Figure 5.9 A confocal 3D image to show a 5%Clay 2.5%PVOH aerogel’s internal

structure

5.3.3 Different molecular weights effect

The third factor we will discuss is the molecular weights of PVOH used in this study.

The hydrogel system we used for this comparison was 3% Clay/ 5% PVOH (Mw varied) one. Viscosity of the hydrogel system can be tuned through changing molecular weight of

PVOH we used. Figure 5.10 shows that the viscosity increased with PVOH molecular weight, resulting then in an aerogel modulus increase, similar to the conclusions of

161

Lamison et al. [15]. From the view of viscosity, higher Mw of PVOH means more chances for local polymer chain entanglement. Even without the cooperation of clay, the preparation of different Mw PVOH solutions could also bring about this phenomenon: the higher the molecular weight of PVOH one used, the harder to prepare respective PVOH solutions.

Figure 5.10 Graphs to exhibit the viscosity properties of same composition hydrogels

prepared with various Mw PVOH solution as well as mechanical performance of

respective aerogel products.

When mixing with higher Mw PVOH solutions, the systems viscosities inevitably increased, which led a granular layer structure formation when freezing hydrogels. Figure

5.11 displays a series of aerogel internal morphologies and related layer thicknesses and layer spacings. All the aerogels prepared in this comparison study were based on 5%

Clay/2.5%PVOH. The only difference in this study was the molecular weight of PVOH used: (a) 31-50k, low Mw, (b) 108k, medium Mw and (c) 146-180k, high Mw. All aerogels in this comparison exhibited a layered structure and small pores in each layer (Figure 5.11

(a),(b) and (c) SEM images). These pores are attributed to the polymer-clay-water

162 interaction during ice crystallization, which hinders ice growth during freezing into large platelets (lamellar structure); the higher the viscosity in the matrix, the more ice crystallization which forms, due to the difficulty to push the matrix aside during ice platelets formation. SEM images show increasing pores generated with increased viscosity.

The layer thicknesses and layer spacing counts measured through SEM images could also demonstrate thickness and spacing decreased due to more pores, which finally were attributed to high viscosity system. The greater number of pores and granular structure formation from high viscosity matrixes could also give an insight to fabricate samples used for thermal insulation, utilizing the special granular structures to capture more air and hinder the heat flow transfer.

163

Figure 5.11 A series of SEM images as well as layer thickness and spacing information exhibited the influence of hydrogel viscosity based on different Mw PVOH.

5.4 Summary

In this chapter we have demonstrated that the viscosity of hydrogels, which were later freeze dried to form aerogels, impacts aerogel structure and mechanical properties.

Through tuning different factors such as stirring speed, standing time and the Mw of polymers used, we have observed the similar performance and successfully established the qualitative relationship between viscosity of hydrogels and modulus of aerogels so that one

164 could predict aerogel properties based on its hydrogel viscosity alone, thereby saving time and energy.

5.5 Acknowledgement

Many thanks are due to Hongbing Chen for useful discussions. As the first project I have done, I sincerely express appreciation to my supervisor’s trust and my trainers’ help on confocal microscopy, SEM and rheometry, Cory Christenson, Cong Zhang and Jesse

Gadley, respectively. Participation of Emma Klinkhamer and Lauren Walters are also acknowledged.

165

5.6 References

[1]. Olson GB. Computational design of hierarchically structured materials. Science.

1997 Aug 29;277(5330):1237-1242.

[2]. Callister Jr WD, Rethwisch DG. Fundamentals of materials science and

engineering: an integrated approach. John Wiley & Sons; 2012.

[3]. Importance of Materials Science in Understanding Failures,

http://www.industrialheating.com/articles/93256-importance-of-materials-

science-in-understanding-failures (accessed September 3, 2017)

[4]. Freeze drying process, http://image.fatare.com/freeze-drying-process/ (accessed

September 3, 2017)

[5]. Gawryla MD. Low density materials through freeze-drying: clay aerogels and

beyond… (Doctoral dissertation, Case Western Reserve University, 2009).

[6]. Yan C, Pochan DJ. Rheological properties of peptide-based hydrogels for

biomedical and other applications. Chemical Society Reviews. 2010;39(9):3528-

3540.

[7]. Macosko CW. Rheology: principles, measurements, and applications. Wiley-vch;

1994.

[8]. Rheology Theory and Applications, TA, http://www.tainstruments.com/wp-

content/uploads/CA-2016-Rheology.pdf (accessed September 3, 2017)

166

[9]. Viscosity & Rheology Glossary,

http://www.rheologyschool.com/advice/rheology-glossary (accessed September

3, 2017)

[10]. Mezger TG. The rheology handbook: for users of rotational and oscillatory

rheometers. Vincentz Network GmbH & Co KG; 2006.

[11]. Gawryla MD, van den Berg O, Weder C, Schiraldi DA. Clay aerogel/cellulose

whisker nanocomposites: a nanoscale wattle and daub. Journal of Materials

Chemistry. 2009;19(15):2118-2124.

[12]. Pojanavaraphan T, Schiraldi DA, Magaraphan R. Mechanical, rheological, and

swelling behavior of natural rubber/montmorillonite aerogels prepared by freeze-

drying. Applied Clay Science. 2010

[13]. Gawryla MD, Schiraldi DA. Novel absorbent materials created via ice

templating. Macromolecular Materials and Engineering. 2009 Sep

14;294(9):570-574.

[14]. Wang Y, Gawryla MD, Schiraldi DA. Effects of freezing conditions on the

morphology and mechanical properties of clay and polymer/clay aerogels.

Journal of Applied Polymer Science. 2013 Aug 5;129(3):1637-1641.

[15]. Lamison K, Gawryla MD, Schiraldi DA. The effect of molecular weight on poly

(vinyl alcohol)/clay aerogel composite properties. Polymer Preprints. 2007

Mar;48(1).

167

Chapter 6

Reactive extrusion of bio-based paint material: synthesis part

One part of work related to a new bio-based paint material project is discussed in this chapter. This project contains two major parts which were carried out in two research groups within the Department: (1) filler materials synthesis (the present chapter) and (2) extrusion with polymer carrier materials (carried out by the Maia group). The following contents will focus on the first part, lab work on reaction between epoxy soybean oil and soy protein, which was used to synthesize the bio-based filler to extrude with polymer carriers in second extrusion part. The concept of this new bio-based paint filler and related background will be introduced briefly in Section 6.1. The materials information, reaction procedure and characterization details will be explained in Section 6.2. Results, analysis and discussion will be presented in Section 6.3.

6.1 Introduction

The use of new type, bio-based materials in daily life, to replace the traditional materials from fossil resources, has been the interests of many labs for decades[1]. These activities are driven by the idea for a sustainable resource supply and green economy.

Among all the bio-based raw materials, soybean has been one of the most significant green sources in recent years[2]. Soybean contains 20+% soybean oil, which composed of

168 saturated and unsaturated triglycerides as well as 50+% protein, which include more than

[3] 18 amino acids . Utilizing its high contents of protein, many applications such as adhesives, paints, insect sprays and food substitutes have adopted soy flour as an important material to conduct a series of reactions. Epoxidized soybean oil is referred to a non-naturally occurring epoxy oil which is prepared through treating soybean oil by epoxidizing the double bonds in molecular chains. In situ performic/peracetic acid processes appear to be used widely to epoxidize vegetable oil, including soybean oil (see

Figure 6.1)[4][5].

Figure 6.1 Scheme to describe a reaction that performic acid migrates into the soybean

oil and reacts to generate epoxidized soybean oil. Adapted from reference[5].

This new type of epoxidized oil is expected to be more reactive than the original unmodified soybean oil with double bonds, since it contains more energetically favorable sites for reaction. The reactive parts of soy flour are expected to be amino groups or amine groups in soy proteins, which have a minimum 53+% content in the 7B ADM soy flour we purchased[6]. Figure 6.2 shows a reaction scheme of one of protein representative, gliadin, with the oxirane groups in epoxidized soybean oil. Since the gliadin has the primary amide groups which are not nucleophiles to activate ring-open, a catalyzed, ZnCl2 has been used

169 to activate the oxirane carbons toward amidolysis by coordination to the epoxy oxygen, linking the primary amide functional groups of proteins to generates amidohydroxy triglycerides, which could be further utilized as an assistant agents or bio-based filler for next extrusion step[7].

Figure 6.2 Scheme of an open-ring reaction between gliadin protein and epoxidized

soybean oil

6.2 Experimental Section

6.2.1 Materials

170

Epoxidized soybean oil (ChemSeed), Bakers soy flour (or 7B soy flour; ADM) polyethylenimine (PEI) and maleic anhydride (MA) (Sigma Aldrich), sodium montmorillonite (PGW grade, Nanocor Inc) and Zinc chloride (TCI) were all used as received. The soy flour we used was a commercial mixture including protein, fat and carbohydrates, and was the soy protein source in our work.

Pre-treatment of soy protein

We aimed to react with protein in ESBO, pre-treating raw materials appeared to be significant to collect more proteins effectively and unfold the chains, exposing more amino or amine groups for reaction. In the Section 6.3.3 soy flour’s performance after acid or alkaline leaching is discussed. Three methods of commercial preparation used currently to obtain soybean protein concentrates, yielding products with protein content as high as 90+% are: 1) aqueous alcohol leach; 2) dilute acid leach (pH=4.5); 3) moist heat then water leach[8]. We adopted method 2 and prepared a pre-treatment solution with 2 M acetic acid and 1 M sodium acetate to tune the pH to 4.5, where the soy protein has its lowest possible solubility characteristics[9]. The acidic solution was then neutralized with 1 M sodium hydroxide to reach pH= 7, frozen, then freeze dried produce an acid-leaching soy protein.

If additional sodium hydroxide was added until the solution reached pH= 9 (to unfold protein chains), the freeze dried samples are referred to as alkaline-leached soy protein.

Due to the freeze drying process, the soy protein would form chunks and need to be ground into powder before using to react with epoxidized soybean oil (see Figure 6.3).

171

Figure 6.3 Acid-leached powder (left) and alkaline-leached powder (right) after freeze-

dry process.

6.2.2 Reaction Procedure

69g epoxidized soybean oil was added to a dry 500mL three-necked round bottom flask equipped with an overhead stirrer. The flask then was heated in a water bath to reach an equilibrium temperature, 70℃. The oil was stirred vigorously and purged with N2 for 15 min during this step. To half of the mixture, 20g Soy Flour, 0.75g ZnCl2 and 0.4g Clay were mixed well first outside then added into the flask through an addition funnel. The rest of mixture was added continuously to the reaction after 3 hours until the epoxy bands were consumed. The basic set up can be seen in Figure 6.4 (a). The epoxidized soybean oil with the additives described above were allowed to react nitrogen at 70℃ for another 3 hours, then a yellow, viscous mixture from the flask was removed (seen as Figure 6.4 (b)). This product could be used for further characterization or sent to the extrusion group as filler material.

172

Figure 6.4 Images to show the reaction setup (a) and product which is ready for

characterization or further extrusion as filler.

6.2.3 Material Characterization

The solubility of the products in acetone and xylene, typical paint/coating solvent, at room temperature were examined. 15 g of the viscous products, whose preparation is described above, were dispersed in 100 ml acetone (upper phase)/ xylene (lower phase) with the help of a magnetic stir bar (Figure 6.5 top-left), and a vacuum filter has been utilized to separate the solving solution and residue unsolved solids. (Figure 6.5 ①②③).

The remaining solution in the flask was centrifuged; a 2 ml solution that had been extracted and centrifuged for 20 min, with the sediments settled at the bottom of the centrifuge tube, allowed for removal of the top residue solvents, acetone or xylene, then the rest of sediments were dried under vacuum in a 100 ℃ oven. The final weights of residues should correspond to the solutes that had been in solution.

173

Characterization of reaction products was carried out by Fourier transform infrared spectroscopy (FT-IR) using a Bomem Michaelson MB 110 spectrophotometer equipped with a deuterated triglycine sulfate detector[10]. 32 scans were added per spectrum at a resolution of 4 cm-1, after using dry air to purge the spectrometer. Then reacted products were ground with KBr powder and compressed into a 13 mm pellet for analysis. The detailed results could be seen in Section 6.3.

Figure 6.5 Illustration to display the procedure of a quantitative solubility test.

174

6.3 Results and discussion

6.3.1 FT-IR characterization

Analysis by FT-IR was used to determine if chemical reactions occured after 7 hours

-1 stirring under N2 environment (see Figure 6.6). C=O stretching at 1630 cm (amide I), and

N-H bending at 1530 cm-1 (amide II), demonstrated that proteins were in the product.

ESBO showed a characteristic peak at 823 cm-1 from the epoxy. Decreases in the oxirane peak intensity was interpreted to mean that ring opening reaction had occurred. The other peaks observed in the FTIR spectra, 955, 1010, 1105 (ether, antisymmetric stretch), and a decrease in C-O-C intensity can be explained as corresponding to increased concentrations of OH groups generated as epoxy ring open.

Figure 6.6 the FT-IR spectra of the reaction products of epoxidized soybean oil with soy

protein (red) and the control group, pure epoxidized soybean oil (black)

175

6.3.2 Quantitative solubility test

The reaction products were subjected to quantitative solubility testing to see if they could dissolve in organic solvents typical for further paint applications. Table 6.1 shows the solubility of different trials of synthesized products. The detailed amount information has shown in the table. ESBO and SP were short for epoxidized soybean oil and soy protein, respectively.

Table 6.1 Quantitative solubility tests based on yield products mixed with various

amount of clay

Acetone- Xylene- Sample Acetone Xylene Centrifuge Centrifuge

69g ESBO+41g 9.3g/100ml 3.5g/L 9.7g/100ml 1.4g/L SP

5.3g/50ml 10.3g/L 4.6g/50ml 5.1g/L 69g ESBO + 41

g SP+0.4gClay 5.1g/50ml 9.4g/L 5.0g/50ml 6.4g/L

(1% on SP) 5.8g/50ml 10.4g/L 5.0g/50ml 6.7g/L

5.1g/50ml 19.5g/L 5g/50ml 10.3g/L 69g ESBO + 41

g SP+0.8gClay 5.3g/50ml 14.1g/L 5.2g/50ml 8.6g/L

(2% on SP) 5.2g/50ml 14.2g/L 4.9g/50ml 8.1g/L

5.4g/50ml 13.8g/L 5.1g/50ml 9.1g/L

176

69g ESBO + 41 6.2g/60ml 15.7g/L 5.5g/50ml 8.8g/L

g SP+1.2gClay 5.5g/50ml 14.9g/L 5.2g/50ml 8.1g/L (3% on SP)

When dissolving the target samples with acetone, it was observed that the products changed color from yellow to white while when dissolved in xylene detailed their same color (see Figure 6.5 top-left one), suggesting that acetone could decolorize soybean flour as it has been demonstrated to decolorize corn zein[11]. The decolorization could also provide an opportunity for the reaction products to be more easily colored in paint products.

From Table 6.1 it can be seen that adding clay in certain amount helped the products dissolve in acetone or xylene. All of the products exhibited higher solubility in acetone than in xylene due to the better match of polarities with somewhat polar acetone. These data lead to the recommendation to add certain amounts of clay into the reaction and help to improve the solubility in organic solvents.

6.3.3 Different methods to pre-treat soy flour

As can be seen from Figure 6.3, soy flour after pre-treatment by acid or alkaline then freeze drying, produces chunks of product, rather than powder. These chunks needed to be ground as fine as possible in order to react with ESBO sufficiently. Table 6.2 shows a quantitative solubility tests for products made from regular soy protein, acid-leached soy protein and alkaline-leached soy protein. Products exhibited higher solubility in acetone than in xylene, regardless of pre-treatment of the soy protein or not. Product synthesized from alkaline-leached soy protein showed higher solubility than with acid-leached material,

177 suggesting that the alkaline solutions helped unfold protein chains in order to expose more reaction cites. Compared to the reactions with raw materials without pre-treatment, alkaline-leached starting material showed no obvious improvement in solubility (perhaps due to the chunky, low surface area nature of this materials). It would be a good comparison in future if we could acquire a finely powdered, alkaline-leached soy protein.

Table 6.2 A quantitative solubility tests for synthesized products made from regular

soy protein, acid-leached soy protein and alkaline-leached soy protein.

Acetone Acetone- Xylene Xylene- Sample (status) Centrifuge (status) Centrifuge

9.3g/100ml 9.7g/100ml 69g ESBO + Disperse 3.5g/L Disperse 1.4g/L 41 g SP White Yellow

69g ESBO + 5.1g/50ml 5g/100ml 41 g Disperse 14.1g/L Disperse 8.6g/L SP+0.8gClay White Yellow (2% on SF)

23g ESBO + 5.3g/50ml 5.5g/60ml 13 g Acid- Disperse 8.82g/L Disperse 4.86g/L leached White Yellow SP+0.3gClay

178

23g ESBO + 5.2g/50ml 5.7g/60ml 13 g Alkaline- Disperse 13.82g/L Disperse 6.66g/L leached White Yellow SP+0.3gClay

6.3.4 Additional trial for a coating based on hot-press procedure

Except for traditional paint, soy protein has been also reported to be coated on substrates through hot-press curing. A proposed reaction mechanism is shown in Figure 6.7[12]. The reaction between maleyl groups of MA and amino groups of PEI can form amide-linked groups in the solution (Figure 6.8 (a)). With added soy protein, MA could further react with amino groups to from a highly cross-linked water resistant adhesive network during hot-pressing then curing on the substrates (Figure 6.8(c))[13]. It has been clearly observed that wooden sticks dipped in the reaction mixtures described in this section, then hot- pressed at 150 ℃, 150 psi, (Figure 6.8(b)) formed a resin which was water-resistant.

Without hot-pressing, no color change and no water-resistant adhesive production was observed. The product shown in Figure 6.8 b was designed to be coated onto wood through hot-press curing. As for solubility, even though acetone was still a better solvent than

Xylene for this product (see Figure 6.8 (d)), it was even worse than the synthesized product in Section 6.3.2, which demonstrated it didn’t fit for paint application.

179

Figure 6.7. Scheme to illustrated a proposed mechanism about a reaction between maleic anhydride (MA), polyethylenimine (PEI) and soy proteins with amino groups (SP-

NH2). Adapted from reference[12].

180

Figure 6.8 Figures to display the procedure to synthesize the product contained “MA-

PEI-SP”, fabrication of product (a) and (b); samples with or without hot-press curing (c);

Solve the mixture into acetone (left) and xylene (right) (d)

6.4 Summary

Soy protein in soy flour has been successfully reacted with epoxidized soybean oil to generate a product which can be used as a bio-based paint filler, in order to extrude with

181 polymer carriers in second phase. This kind of green bio-based material could be further studied as an adhesive coating through hot-press process. Pre-treatment through acid- alkaline leaching helped to expose more reaction cites while the chucks formed during treatment would hinder the reaction. It was recommended to grind the acid-base-leaching soybean chunks before reacted with epoxidized soybean oil.

6.5 Acknowledgement

This project was supported by Sherwin-Williams. Suggestions and guidance from

Gamini S. Samaranayake, Philip. J. Ruhoff, and Duke Rao is gratefully acknowledged.

Many thanks are due to Hossein Ghassemi for discussions as well as to Sidney Carson and

Joseph Kestner for help on extrusion work.

182

6.6 References

[1]. Hablot E, Graiver D, Narayan R. Biobased industrial products from soybean

biorefinery. InGreen Polymer Chemistry: Biocatalysis and Materials II 2013 (pp.

305-329). American Chemical Society.

[2]. Wool R, Sun XS. Bio-based polymers and composites. Academic Press; 2011

Aug 30.

[3]. Liu Y, Li K. Development and characterization of adhesives from soy protein for

bonding wood. International Journal of Adhesion and Adhesives. 2007 Jan

31;27(1):59-67.

[4]. Webster DC, Sengupta PP, Chen Z, Pan X, Paramarta A, inventors; Ndsu

Research Foundation, assignee. Highly functional epoxidized resins and coatings.

United States patent US 9,096,773. 2015 Aug 4.

[5]. Epoxidation of soybean oil to obtain ESBO ， Eurochem Engineering,

http://www.eurochemengineering.com/Epoxidation-of-soybean-oil-to-obtain-

ESBO.aspx (accessed September 3 2017)

[6]. GMO Status of ADM Soy Protein Products, ADM, http://manghisbread.com/wp-

content/uploads/2014/01/non-gmo-soy-flour.pdf (accessed September 3 2017)

[7]. Harry-O'kuru RE, Mohamed AA, Gordon SH, Xu J, inventors; The United States

Of America, As Represented By The Secretary Of Agriculture, assignee.

Elastomer product from epoxidized vegetable oil and gliadin. United States

patent US 8,575,289. 2013 Nov 5.

183

[8]. Jeon J, Lee JH, Kim S. Soybean-based green adhesive for environment-friendly

furniture material. 한국가구학회지. 2011 Jul;22(3):174-82.

[9]. Schmitz Jr JF. Enzyme modified soy flour adhesives. Iowa State University; 2009.

[10]. Ohashi S, Cassidy F, Huang S, Chiou K, Ishida H. Synthesis and ring-opening

polymerization of 2-substituted 1, 3-benzoxazine: the first observation of the

polymerization of oxazine ring-substituted benzoxazines. Polymer Chemistry.

2016;7(46):7177-7184.

[11]. Sessa DJ, Eller FJ, Palmquist DE, Lawton JW. Improved methods for

decolorizing corn zein. Industrial crops and products. 2003 Jul 31;18(1):55-65.

[12]. Huang J. Development and characterization of new formaldehyde-free soy flour-

based adhesives for making interior plywood (Doctoral dissertation).

[13]. Huang J, Li K. A new soy flour-based adhesive for making interior type II

plywood. Journal of the American Oil Chemists' Society. 2008 Jan 1;85(1):63-

70.

184

Chapter 7

Conclusions and Future work

7.1 Conclusions

The emphasis of this thesis has been an effort to study two classes of layered materials according to the strategy of material optimization loop: “Processing- structure- properties”. From previous work we have found the properties of layered materials, either multilayered films or clay aerogel composites, are directly affected from their structure or morphology, which is determined by corresponding processing methods and related conditions. Therefore, such systems can be tuned and designed according to an appropriate process path or conditions which fulfill the product end-use requirements:

 For the multilayer barrier film systems, in order to achieve the goal of barrier

requirement, OTR< 10-3 and WVTR< 10-4, we designed and fabricated

multilayered films based on water barrier material, HDPE and an oxygen barrier

material, EVOH. When the raw multilayer films could not further reduce the

WVTR down, we designed a series of extra coating layers which could modify

the original multilayer films; these coatings made use of layer by layer application

of clay and polymer, parylene, plasma treatment, and a combination method

integrating parylene with alumina. When we realized that thicker films brought

about better barrier properties at only a small loss of transparency, we processed

185

thinner multilayer films then made up barrier loss by “inorganic-organic” multi-

stack coating. We ultimately were successful in achieving a multilayer film with

superior properties for barrier, durability and transparency by a co-extrusion and

coating processing path. Integrating the modified films above with organic solar

cell devices to fabricate solar cell packages, the devices’ shelf-lifetime were

increased through barrier film’s protection, hindering the permeation of oxygen

and water vapor. That is an achievement; Chapter 2 and 3, in a word, have built a

complete flow of a multilayered product, from design, processing, to properties

improvement then application in the real world.

 For aerogel systems, different fabrication conditions, such as freezing temperature

and the molecular weight of raw materials, have directly made the impact on

aerogel products’ structure. Due to the lamellar and granular structure formation,

specific properties such as mechanical one and thermal insulation have been

improved, then brought a potential application as an insulation material.

Incorporating with keratin fibers, a kind of natural bio-based insulation fibers,

have enhanced the current structure from physical entanglement and chemical

interaction view, which finally is shown as improvement on the properties.

 Based on that classic optimization loop, we are eager to implant this strategy on

study of aerogel and build a link between processing gels and properties of

aerogels because the cost of time and energy during freeze-drying process limits

its tremendous application potential in industry. Through tuning different

parameters related to gel’s viscosity, we have successfully established the

186

qualitative relationship between viscosity of hydrogels and modulus of aerogels

so that one could make a prediction about aerogel product’s quality based on its

hydrogel’s viscosity performance.

 For soybean-based paint, it has been confirmed via FT-IR and quantitative

solubility test that soybean protein could be reacted with epoxidized soybean oil

to generate a product used as a paint filler. This type of green bio-based material

could be further studied as an adhesive coating through hot-press. It has a

potential to be integrated with polymer carrier such as polystyrene in second

extrusion phase, for further modification.

7.2 Future work

7.2.1 Multilayer barrier films

All the present work is based on 1 or 4 mil thick HDPE-EVOH multilayer barrier films. As a semicrystalline material, HDPE shows a superior water vapor barrier and an acceptable transparency, both due to its crystal domain size[1]. The adding of nucleating agents could decrease the crystal sizes as well as promote the crystallization, the former helps to increase the transparency due to generating small spherulites which do not scatter visible light and the latter aids to improve the barrier property due to formation of organized crystal packing[2][3]. An optimized barrier film recipe improves the raw film’s quality and performance as a protection layer to organic PV devices. Most of encapsulated packages in Chapter 3 were based on 4-mil raw film, which had the best barrier property in raw films so far. Modified films, with extra “inorganic-organic” stacks

187 coating, should be applied on the devices directly in future. And a quantitative relationship between barrier property and solar cells’ shelf-lifetime should be the important simulation work, to predict solar cell’s performance based on films with ultralow barrier property.

7.2.2 Aerogel composites

Keratin has been confirmed to be a good filler to add into aerogel composites to improve mechanical property as well as thermal insulation properties. During this research we found that the amount and the size of fibers plays a role in impacting the formation of hydrogels[4]. Pore size, within the aerogel, is influenced by processing conditions has been described in Chapter 4; other methods, such as supercritical drying can be tried to achieve a nano-scale pores, utilizing for thermal insulation or catalyst sites[5].

For rheology part of this thesis, we have established a qualitative relationship, which in the future, could be the bases on a predictive model which matches hydrogel system rheological behavior and composition, to make accurate predictions of aerogel properties.

7.2.3 Bio-based paint filler

This synthesis work was conducted under nitrogen protection, which is probably not reasonable for a real industrial process. A modified reaction needs to be tried under ambient atmosphere, or even directly conducted as a reactive extrusion with polymer carriers. Some extrusion trials, composed of soybean flour, epoxidized soybean oil and

Pebax®, have been carried out, producing elastic strips which showed strong mechanical

188 strengths (see Figure 7.1). potentially providing a path forward for future reactive extrusion research which could be translated into a in industrial process[6].

Figure 7.1 Trials composed soybean flour, epoxidized soy bean oil and Pebax®

elastomer have been extruded together(a); the product stripe showed an impressive

tensile strength (b)

189

7.3 References

[1]. Kotani T, Taka T, Saito Y, inventors; Showa Denko Kabushiki Kaisha, assignee.

High density polyethylene type transparent film and process for production

thereof. United States patent US 4,954,391. 1990 Sep 4.

[2]. Osman MA, Atallah A. High‐Density Polyethylene Micro‐and Nanocomposites:

Effect of Particle Shape, Size and Surface Treatment on Polymer Crystallinity

and Gas Permeability. Macromolecular rapid communications. 2004 Sep

9;25(17):1540-1544.

[3]. Bartczak Z, Galeski A, Pracella M. Spherulite nucleation in blends of isotactic

polypropylene with high-density polyethylene. Polymer. 1986 Apr 1;27(4):537-

543.

[4]. Finlay KA, Gawryla MD, Schiraldi DA. Effects of fiber reinforcement on clay

aerogel composites. Materials. 2015 Aug 21;8(8):5440-5451.

[5]. Kocon L, Despetis F, Phalippou J. Ultralow density silica aerogels by alcohol

supercritical drying. Journal of Non-Crystalline Solids. 1998 Apr 30;225:96-100.

[6]. Wang N, Yu J, Ma X. Preparation and characterization of thermoplastic

starch/PLA blends by one‐step reactive extrusion. Polymer International. 2007

Nov 1;56(11):1440-1447.

190

Bibliography

1. 2V 1W Thin-film Flexible Solar Panel, Seeed,

http://www.seeedstudio.com/depot/2V-1W-Thinfilm-Flexible-Solar-Panel-p-

1856.html (accessed September 2, 2017)

2. Abdulagatov AI, Yan Y, Cooper JR, Zhang Y, Gibbs ZM, Cavanagh AS, Yang RG,

Lee YC, George SM. Al2O3 and TiO2 atomic layer deposition on copper for water

corrosion resistance. ACS applied materials & interfaces. 2011 Nov 16;3(12):4593-

601.

3. Aegerter M, Leventis N, Koebel MM, editors. Aerogels handbook. Springer

Science & Business Media; 2011 Jun 10

4. Akimov YK. "Fields of application of aerogels." Instruments and Experimental

Techniques 46, no. 3 (2003): 287-299.

5. Alemán JV, Chadwick AV, He J, Hess M, Horie K, Jones RG, Kratochvíl P, Meisel

I, Mita I, Moad G, Penczek S. Definitions of terms relating to the structure and

processing of sols, gels, networks, and inorganic-organic hybrid materials (IUPAC

Recommendations 2007). Pure and Applied Chemistry. 2007 Jan 1;79(10):1801-

1829.

191

6. Amberg-Schwab S, Hoffmann M, Bader H, Gessler M. Inorganic-organic polymers

with barrier properties for water vapor, oxygen and flavors. Journal of sol-gel

science and technology. 1998 Jan 1;13(1):141-146.

7. Andrew W. Permeability and other film properties of plastics and elastomers.

Plastics Design Library. New York: Norwich. 1995:58-60.

8. Arndt EM, Gawryla MD, Schiraldi DA. Elastic, low density epoxy/clay aerogel

composites. Journal of Materials Chemistry. 2007;17(33):3525-3529.

9. ASTM F. 07. Standard Test Method for Determination of Oxygen Gas

Transmission Rate, Permeability and Permeance at Controlled Relative Humidity

Through Barrier Materials Using a Coulometric Detector. American National

Standards Institute. Aug. 2007.

10. Aubry E, Ghazzal MN, Demange V, Chaoui N, Robert D, Billard A. Poisoning

prevention of TiO2 photocatalyst coatings sputtered on soda-lime glass by

intercalation of SiNx diffusion barriers. Surface and Coatings Technology. 2007

Jun 25;201(18):7706-7712.

11. Backer SA, Sivula K, Kavulak DF, Fréchet JM. High efficiency organic

photovoltaics incorporating a new family of soluble fullerene derivatives.

Chemistry of materials. 2007 Jun 12;19(12):2927-2929.

12. Bandi S, Bell M, Schiraldi DA. Temperature-responsive clay aerogel− polymer

composites. Macromolecules. 2005 Nov 1;38(22):9216-9220.

192

13. Bandi S, Schiraldi DA. Glass transition behavior of clay aerogel/poly (vinyl

alcohol) composites. Macromolecules. 2006 Sep 19;39(19):6537-6545.

14. Barone JR, Schmidt WF. Polyethylene reinforced with keratin fibers obtained from

chicken feathers. Composites Science and Technology. 2005;65(2):173-181.

15. Barrier materials & encapsulation for OPV, Institut National De L’Energie Solaire

http://www.sophiari.eu/fileadmin/SOPHIA_docs/documents/Workshops/Worksho

p_on_OPV/Cros_INES_SOPHIA_EERA_OPV_WS_111108.pdf (accessed

September 2, 2017).

16. Barrier Propertities of Polymers, Polymer Properties Database,

http://polymerdatabase.com/polymer%20physics/Permeability.html (accessed

September 2, 2017).

17. Bartczak Z, Galeski A, Pracella M. Spherulite nucleation in blends of isotactic

polypropylene with high-density polyethylene. Polymer. 1986 Apr 1;27(4):537-

543.

18. Barton AF. Solubility parameters. Chemical Reviews. 1975 Dec;75(6):731-753.

19. Beadie G, Shirk JS, Rosenberg A, Lane PA, Fleet E, Kamdar AR, Jin Y, Ponting

M, Kazmierczak T, Yang Y, Hiltner A. Optical properties of a bio-inspired gradient

refractive index polymer lens. Optics express. 2008 Jul 21;16(15):11540-11547.

20. Cai J, Perfect E, Cheng CL, Hu X. Generalized modeling of spontaneous imbibition

based on Hagen–Poiseuille flow in tortuous capillaries with variably shaped

apertures. Langmuir. 2014 May 1;30(18):5142-5151.

193

21. Cai W, Gong X, Cao Y. Polymer solar cells: recent development and possible

routes for improvement in the performance. Solar Energy Materials and Solar Cells.

2010 Feb 28;94(2):114-127.

22. Call F. Preparation of Dry Clay-Gels by Freeze drying. Nature 1953, 172, 126.

23. Callister Jr WD, Rethwisch DG. Fundamentals of materials science and

engineering: an integrated approach. John Wiley & Sons; 2012.

24. Carosio F, Laufer G, Alongi J, Camino G, Grunlan JC. Layer-by-layer assembly of

silica-based flame retardant thin film on PET fabric. Polymer Degradation and

Stability. 2011 May 31;96(5):745-750.

25. Carpenter, Diane R., et al. "Adhesive-backed breathable layered materials." U.S.

Patent No. 5,277,954. 11 Jan. 1994.

26. Carr JM, Mackey M, Flandin L, Schuele D, Zhu L, Baer E. Effect of biaxial

orientation on dielectric and breakdown properties of poly (ethylene

terephthalate)/poly (vinylidene fluoride‐co‐tetrafluoroethylene) multilayer films.

Journal of Polymer Science Part B: Polymer Physics. 2013 Jun 1;51(11):882-896.

27. Centi G, Perathoner S. Catalysis by layered materials: A review. Microporous and

mesoporous materials. 2008 Jan 1;107(1):3-15.

28. Chambon S, Derue L, Lahaye M, Pavageau B, Hirsch L, Wantz G. MoO3 thickness,

thermal annealing and solvent annealing effects on inverted and direct polymer

photovoltaic solar cells. Materials. 2012 Nov 27;5(12):2521-2536.

194

29. Chapin DM, Fuller CS, Pearson GL. A new silicon p‐n junction photocell for

converting solar radiation into electrical power. Journal of Applied Physics. 1954

May 1;25(5):676-677.

30. Charton C, Schiller N, Fahland M, Holländer A, Wedel A, Noller K. Development

of high barrier films on flexible polymer substrates. Thin Solid Films. 2006 Apr

28;502(1):99-103.

31. Chen HB, Hollinger E, Wang YZ, Schiraldi DA. Facile fabrication of poly (vinyl

alcohol) gels and derivative aerogels. Polymer. 2014 Jan 14;55(1):380-384.

32. Chen HB, Wang YZ, Sánchez-Soto M, Schiraldi DA. Low flammability, foam-like

materials based on ammonium alginate and sodium montmorillonite clay. Polymer.

2012;53(25):5825-5831.

33. Chen HY, Hou J, Zhang S, Liang Y, Yang G, Yang Y, Yu L, Wu Y, Li G. Polymer

solar cells with enhanced open-circuit voltage and efficiency. Nature photonics.

2009 Nov 1;3(11):649-653.

34. Chen TN, Wuu DS, Wu CC, Chiang CC, Chen YP, Horng RH. High-performance

transparent barrier films of SiOx/SiNx stacks on flexible polymer substrates.

Journal of The Electrochemical Society. 2006 Oct 1;153(10):F244-248.

35. Chen TN, Wuu DS, Wu CC, Chiang CC, Chen YP, Horng RH. Improvements of

permeation barrier coatings using encapsulated parylene interlayers for flexible

electronic applications. Plasma Processes and Polymers. 2007 Feb 22;4(2):180-

185.

195

36. Cheng YJ, Yang SH, Hsu CS. Synthesis of conjugated polymers for organic solar

cell applications. Chemical reviews. 2009 Sep 28;109(11):5868-5923.

37. Chiang CC, Wuu DS, Lin HB, Chen YP, Chen TN, Lin YC, Wu CC, Chen WC,

Jaw TH, Horng RH. Deposition and permeation properties of SiNX/parylene

multilayers on polymeric substrates. Surface and Coatings Technology. 2006 May

22;200(20):5843-5848.

38. Cros S, Firon M, Lenfant S, Trouslard P, Beck L. Study of thin calcium electrode

degradation by ion beam analysis. Nuclear Instruments and Methods in Physics

Research Section B: Beam Interactions with Materials and Atoms. 2006 Sep

30;251(1):257-260.

39. Dameron AA, Davidson SD, Burton BB, Carcia PF, McLean RS, George SM. Gas

diffusion barriers on polymers using multilayers fabricated by Al2O3 and rapid SiO2

atomic layer deposition. The Journal of Physical Chemistry C. 2008 Mar

27;112(12):4573-4580.

40. Decher G, Eckle M, Schmitt J, Struth B. Layer-by-layer assembled multicomposite

films. Current opinion in colloid & interface science. 1998 Feb 1;3(1):32-39.

41. Dooley J. Determining the processability of multilayer coextruded structures.

InANTEC-CONFERENCE PROCEEDINGS- 2007 (Vol. 4, p. 2369).

42. Duncan TV. Applications of nanotechnology in food packaging and food safety:

barrier materials, antimicrobials and sensors. Journal of colloid and interface

science. 2011 Nov 1;363(1):1-24.

196

43. Dvoracek CM, Sukhonosova G, Benedik MJ, Grunlan JC. Antimicrobial behavior

of polyelectrolyte− surfactant thin film assemblies. Langmuir. 2009 Jun

18;25(17):10322-10328.

44. Epoxidation of soybean oil to obtain ESBO ， Eurochem Engineering,

http://www.eurochemengineering.com/Epoxidation-of-soybean-oil-to-obtain-

ESBO.aspx (accessed September 3 2017)

45. Fang L, Shen M, Yang J, Li Z. The effect of SiO2 barrier layer on the dielectric

properties of CaCu3Ti4O12 films. Journal of Physics D: Applied Physics. 2005 Nov

18;38(23):4236.

46. Finlay K, Gawryla MD, Schiraldi DA. Biologically based fiber-reinforced/clay

aerogel composites. Industrial & Engineering Chemistry Research. 2008;47(3):

615-619

47. Finlay KA, Gawryla MD, Schiraldi DA. Effects of fiber reinforcement on clay

aerogel composites. Materials. 2015;8(8):5440-5451.

48. Freeze drying process, http://image.fatare.com/freeze-drying-process/ (accessed

September 3, 2017)

49. Galagan Y, Rubingh JE, Andriessen R, Fan CC, Blom PW, Veenstra SC, Kroon

JM. ITO-free flexible organic solar cells with printed current collecting grids. Solar

Energy Materials and Solar Cells. 2011 May 31;95(5):1339-1343.

50. Gao J, Yu W, Pan N. Structures and properties of the goose down as a material for

thermal insulation. Textile Research Journal. 2007;77(8):617-626.

197

51. Garcia-Cespedes J, Thomasson S, Teo KB, Kinloch IA, Milne WI, Pascual E,

Bertran E. Efficient diffusion barrier layers for the catalytic growth of carbon

nanotubes on copper substrates. Carbon. 2009 Mar 31;47(3):613-621.

52. Gas Barrier Seminar, High Gas-Barrier Resin Soarnol,

http://www.soarnol.com/eng/solution/solution050303.html (accessed September 2,

2017).

53. Gawryla MD, Liu L, Grunlan JC, Schiraldi DA. pH tailoring electrical and

mechanical behavior of polymer–clay–nanotube aerogels. Macromolecular rapid

communications. 2009 Oct 1;30(19):1669-1673.

54. Gawryla MD, Schiraldi DA. Novel absorbent materials created via ice templating.

Macromolecular Materials and Engineering. 2009 Sep 14;294(9):570-574.

55. Gawryla MD, van den Berg O, Weder C, Schiraldi DA. Clay aerogel/cellulose

whisker nanocomposites: a nanoscale wattle and daub. Journal of Materials

Chemistry. 2009;19(15):2118-2124.

56. Gawryla MD. Low density materials through freeze-drying: clay aerogels and

beyond… (Doctoral dissertation, Case Western Reserve University, 2009).

57. Gevorgyan SA, Jørgensen M, Krebs FC. A setup for studying stability and

degradation of polymer solar cells. Solar Energy Materials and Solar Cells. 2008

Jul 31;92(7):736-745.

198

58. Ghosh AP, Gerenser LJ, Jarman CM, Fornalik JE. Thin-film encapsulation of

organic light-emitting devices. Applied physics letters. 2005 May

30;86(22):223503.

59. Girtan M, Rusu M. Role of ITO and PEDOT: PSS in stability/degradation of

polymer: fullerene bulk heterojunctions solar cells. Solar Energy Materials and

Solar Cells. 2010 Mar 31;94(3):446-450.

60. GMO Status of ADM Soy Protein Products, ADM, http://manghisbread.com/wp-

content/uploads/2014/01/non-gmo-soy-flour.pdf (accessed September 3 2017)

61. Goswami DY. Advances in Solar Energy: an Annual Review of Research and

Development. Earthscan; 2007.

62. Graff GL, Williford RE, Burrows PE. Mechanisms of vapor permeation through

multilayer barrier films: Lag time versus equilibrium permeation. Journal of

Applied Physics. 2004 Aug 15;96(4):1840-9.

63. Gregg SJ, Jepson WB. 186. The oxidation of calcium in moist oxygen. Journal of

the Chemical Society (Resumed). 1961:884-888.

64. Groner MD, Fabreguette FH, Elam JW, George SM. Low-temperature Al2O3

atomic layer deposition. Chemistry of Materials. 2004 Feb 24;16(4):639-645.

65. Hablot E, Graiver D, Narayan R. Biobased industrial products from soybean

biorefinery. InGreen Polymer Chemistry: Biocatalysis and Materials II 2013 (pp.

305-329). American Chemical Society.

199

66. Han D, Yoo S. The stability of normal vs. inverted organic solar cells under highly

damp conditions: comparison with the same interfacial layers. Solar Energy

Materials and Solar Cells. 2014 Sep 30;128:41-47.

67. Harry-O'kuru RE, Mohamed AA, Gordon SH, Xu J, inventors; The United States

Of America, As Represented By The Secretary Of Agriculture, assignee. Elastomer

product from epoxidized vegetable oil and gliadin. United States patent US

8,575,289. 2013 Nov 5.

68. Hau SK, Yip HL, Baek NS, Zou J, O’Malley K, Jen AK. Air-stable inverted flexible

polymer solar cells using zinc oxide nanoparticles as an electron selective layer.

Applied Physics Letters. 2008 Jun 23;92(25):225.

69. He Z, Zhong C, Su S, Xu M, Wu H, Cao Y. Enhanced power-conversion efficiency

in polymer solar cells using an inverted device structure. Nature Photonics. 2012

Sep 1;6(9):591-595.

70. Hebbar RS, Isloor AM, Ismail AF. Preparation and evaluation of heavy metal

rejection properties of polyetherimide/porous activated bentonite clay

nanocomposite membrane. RSC Advances. 2014;4(88):47240-47248.

71. Hostler SR, Abramson AR, Gawryla MD, Bandi SA, Schiraldi DA. Thermal

conductivity of a clay-based aerogel. International Journal of Heat and Mass

Transfer. 2009 Jan 31;52(3):665-669.

72. Hua DW, Anderson J, Di Gregorio J, Smith DM, Beaucage G. Structural analysis

of silica aerogels. Journal of non-crystalline solids. 1995 Jun 2;186:142-148.

200

73. Huang J, Li K. A new soy flour-based adhesive for making interior type II plywood.

Journal of the American Oil Chemists' Society. 2008 Jan 1;85(1):63-70.

74. Huang J. Development and characterization of new formaldehyde-free soy flour-

based adhesives for making interior plywood (Doctoral dissertation).

75. Importance of Materials Science in Understanding Failures,

http://www.industrialheating.com/articles/93256-importance-of-materials-

science-in-understanding-failures (accessed September 3, 2017)

76. Infinity PV foil, Infinity PV, https://www.infinitypv.com/products/opv/foil

(accessed September 2, 2017)

77. Ismadji S, Soetaredjo FE, Ayucitra A. Natural clay minerals as environmental

cleaning agents. InClay Materials for Environmental Remediation 2015 (pp. 5-37).

Springer International Publishing.

78. Jeon J, Lee JH, Kim S. Soybean-based green adhesive for environment-friendly

furniture material. 한국가구학회지. 2011 Jul;22(3):174-82.

79. Jia X, Jiang Z, Chen X, Zhou J, Pan L, Zhu F, Sun Z, Huang S. Highly efficient

and air stable inverted polymer solar cells using LiF-modified ITO cathode and

MoO3/AgAl alloy anode. ACS applied materials & interfaces. 2016 Feb

2;8(6):3792-3799

80. Johnson D, Clarke S, Koumoto K. Layered materials. Semiconductor Science and

Technology. 2014 Jun;29(6):060301.

201

81. Jørgensen M, Norrman K, Gevorgyan SA, Tromholt T, Andreasen B, Krebs FC.

Stability of polymer solar cells. Advanced materials. 2012 Feb 2;24(5):580-612.

82. Jørgensen M, Norrman K, Krebs FC. Stability/degradation of polymer solar cells.

Solar Energy Materials and Solar Cells. 2008 Jul 31;92(7):686-714

83. Jung K, Bae JY, Park SJ, Yoo S, Bae BS. High performance organic-inorganic

hybrid barrier coating for encapsulation of OLEDs. Journal of Materials Chemistry.

2011;21(6):1977-1983.

84. K.L. Yam, Gas permeation of packaging materials, in: K.L. Yam (Ed.), The Wiley

Encyclopedia of Packaging Technology, third ed., John Wiley and Sons, Inc., New

York, 2009, pp. 551–555.

85. Kim H, So WW, Moon SJ. The importance of post-annealing process in the device

performance of poly (3-hexylthiophene): Methanofullerene polymer solar cell.

Solar Energy Materials and Solar Cells. 2007 Apr 16;91(7):581-587.

86. Kim JW, Choi HS. Surface crosslinking of high‐density polyethylene beads in a

modified plasma reactor. Journal of applied polymer science. 2002 Mar

28;83(13):2921-2929.

87. Kim N, Potscavage WJ, Sundaramoothi A, Henderson C, Kippelen B, Graham S.

A correlation study between barrier film performance and shelf lifetime of

encapsulated organic solar cells. Solar Energy Materials and Solar Cells. 2012 Jun

30;101:140-146.

202

88. Kim N. Fabrication and characterization of thin-film encapsulation for organic

electronics. Georgia Institute of Technology; 2009.

89. Kim TW, Yan M, Erlat AG, McConnelee PA, Pellow M, Deluca J, Feist TP, Duggal

AR, Schaepkens M. Transparent hybrid inorganic/organic barrier coatings for

plastic organic light-emitting diode substrates. Journal of Vacuum Science &

Technology A: Vacuum, Surfaces, and Films. 2005 Jul;23(4):971-977.

90. Kim YH, Sachse C, Machala ML, May C, Müller‐Meskamp L, Leo K. Highly

conductive PEDOT: PSS electrode with optimized solvent and thermal post ‐

treatment for ITO‐free organic solar cells. Advanced Functional Materials. 2011

Mar 22;21(6):1076-1081.

91. Kim, J.B.; Kim, C.S.; Kim, Y.S.; Loo, Y.-L. Oxidation of silver electrodes induces

transition from conventional to inverted photovoltaic characteristics in polymer

solar cells. Appl. Phys. Lett. 2009, 95, 183301

92. Kistler SS. Coherent expanded-aerogels[J]. The Journal of Physical Chemistry,

1932, 36(1): 52-64.

93. Klumbies H. Encapsulations for Organic Devices and their Evaluation using

Calcium Corrosion Tests. (2013)

94. Kocon L, Despetis F, Phalippou J. Ultralow density silica aerogels by alcohol

supercritical drying. Journal of Non-Crystalline Solids. 1998 Apr 30;225:96-100.

203

95. Kotani T, Taka T, Saito Y, inventors; Showa Denko Kabushiki Kaisha, assignee.

High density polyethylene type transparent film and process for production thereof.

United States patent US 4,954,391. 1990 Sep 4.

96. Krebs FC, Jørgensen M, Norrman K, Hagemann O, Alstrup J, Nielsen TD, Fyenbo

J, Larsen K, Kristensen J. A complete process for production of flexible large area

polymer solar cells entirely using screen printing—first public demonstration. Solar

Energy Materials and Solar Cells. 2009 Apr 30;93(4):422-441.

97. Kuo HI, Zhang R, Ko WH. Development of micropackage technology for

biomedical implantable microdevices using parylene C as water vapor barrier

coatings. InSensors, 2010 IEEE 2010 Nov 1 (pp. 438-441). IEEE.

98. Kyaw AK, Sun XW, Jiang CY, Lo GQ, Zhao DW, Kwong DL. An inverted organic

solar cell employing a sol-gel derived ZnO electron selective layer and thermal

evaporated MoO3 hole selective layer. Applied Physics Letters. 2008 Dec

1;93(22):221107.

99. Lagaron JM, Catalá R, Gavara R. Structural characteristics defining high barrier

properties in polymeric materials. Materials Science and Technology. 2004 Jan

1;20(1):1-7.

100. Lamison K, Gawryla MD, Schiraldi DA. The effect of molecular weight on poly

(vinyl alcohol)/clay aerogel composite properties. Polymer Preprints. 2007

Mar;48(1).

204

101. Langhe D, Ponting M. Manufacturing and Novel Applications of Multilayer

Polymer Films. William Andrew; 2016 Jan 4.

102. Lessons Learnt Report: Market & Technology Analysis for Organic Photovoltaics;

https://arena.gov.au/assets/2014/06/Market-Technology-Analysis_-Lessons-

Learnt.pdf (accessed September 2, 2017)

103. Li G, Shrotriya V, Yao Y, Yang Y. Investigation of annealing effects and film

thickness dependence of polymer solar cells based on poly (3-hexylthiophene).

Journal of Applied Physics. 2005 Aug 15;98(4):043704.

104. Lide DR. Handbook of Chemistry and Physics, 86th edn. CRC. 2005.

105. Liu RY, Jin Y, Hiltner A, Baer E. Probing Nanoscale Polymer Interactions by

Forced ‐ Assembly. Macromolecular rapid communications. 2003 Nov

1;24(16):943-948.

106. Liu RY, Schiraldi DA, Hiltner A, Baer E. Oxygen‐barrier properties of cold‐drawn

polyesters. Journal of Polymer Science Part B: Polymer Physics. 2002 May

1;40(9):862-877.

107. Liu Y, Li K. Development and characterization of adhesives from soy protein for

bonding wood. International Journal of Adhesion and Adhesives. 2007 Jan

31;27(1):59-67.

108. Lunt RR, Bulovic V. Transparent, near-infrared organic photovoltaic solar cells for

window and energy-scavenging applications. Applied Physics Letters. 2011 Mar

14;98(11):61.

205

109. Lutkenhaus JL, Hammond PT. Electrochemically enabled polyelectrolyte

multilayer devices: from fuel cells to sensors. Soft Matter. 2007;3(7):804-816.

110. Mackenzie RC. Clay-Water Relationships. Nature 1953, 171, 681-683

111. Mackey M, Flandin L, Hiltner A, Baer E. Confined crystallization of PVDF and a

PVDF‐TFE copolymer in nanolayered films. Journal of Polymer Science Part B:

Polymer Physics. 2011 Dec 15;49(24):1750-1761.

112. Macosko CW. Rheology: principles, measurements, and applications. Wiley-vch;

1994.

113. Madakasira P, Inoue K, Ulbricht R, Lee SB, Zhou M, Ferraris JP, Zakhidov AA.

Multilayer encapsulation of plastic photovoltaic devices. Synthetic Metals. 2005

Nov 15;155(2):332-335.

114. Mao N, Russell SJ. The thermal insulation properties of spacer fabrics with a

mechanically integrated wool fiber surface. Textile Research Journal. 2007

Dec;77(12):914-922.

115. Mao Z. Molecular Structures and Device Properties of Organic Solar Cells

(Doctoral dissertation, Case Western Reserve University, 2014).

116. Martinez-Hernandez AL, Velasco-Santos C, De Icaza M, Castano VM.

Microstructural characterisation of keratin fibres from chicken feathers.

International journal of environment and pollution. 2005; 23(2):162-178.

206

117. Martínez-Hernández AL, Velasco-Santos C, de Icaza M, Castaño VM. Hierarchical

microstructure in keratin biofibers. Microscopy and Microanalysis. 2003

Aug;9(S02):1282-1283.

118. Martínez-Hernández AL, Velasco-Santos C, de Icaza M, Castaño VM. Mechanical

properties evaluation of new composites with protein biofibers reinforcing poly

(methyl methacrylate). Polymer. 2005;46(19):8233-8238.

119. Martínez-Hernández AL, Velasco-Santos C, De-Icaza M, Castano VM.

Dynamical–mechanical and thermal analysis of polymeric composites reinforced

with keratin biofibers from chicken feathers. Composites Part B: Engineering.

2007;38(3):405-410.

120. Martins M, Barros AA, Quraishi S, Gurikov P, Raman SP, Smirnova I, Duarte AR,

Reis RL. Preparation of macroporous alginate-based aerogels for biomedical

applications. The Journal of Supercritical Fluids. 2015;106:152-159.

121. Mezger TG. The rheology handbook: for users of rotational and oscillatory

rheometers. Vincentz Network GmbH & Co KG; 2006.

122. Military Foldable and Rollable Solar Panels, PowerFilm Solar,

http://www.powerfilmsolar.com/military/products__systems/foldable-and-rollable-

solar/ (accessed September 2, 2017)

123. Minick J, Moet A, Baer E. Morphology of HDPE/LDPE blends with different

thermal histories. Polymer. 1995 May 1;36(10):1923-32.

207

124. Nehm F, Klumbies H, Richter C, Singh A, Schroeder U, Mikolajick T, Mö nch T,

Hoßbach C, Albert M, Bartha JW, Leo K. Breakdown and protection of ALD

moisture barrier thin films. ACS applied materials & interfaces. 2015 Sep

30;7(40):22121-22127.

125. Neugebauer A, Chen K, Tang A, Allgeier A, Glicksman LR, Gibson LJ. Thermal

conductivity and characterization of compacted, granular silica aerogel. Energy and

buildings. 2014;79:47-57.

126. Nisato G, Bouten PC, Slikkerveer PJ, Bennett WD, Graff GL, Rutherford N, Wiese

L. Evaluating high performance diffusion barriers: The calcium test. InProc. Int.

Display Workshop/Asia Display 2001 Oct 16 (Vol. 1435).

127. Norrman K, Madsen MV, Gevorgyan SA, Krebs FC. Degradation patterns in water

and oxygen of an inverted polymer solar cell. Journal of the American Chemical

Society. 2010 Nov 5;132(47):16883-16892.

128. Ohashi S, Cassidy F, Huang S, Chiou K, Ishida H. Synthesis and ring-opening

polymerization of 2-substituted 1, 3-benzoxazine: the first observation of the

polymerization of oxazine ring-substituted benzoxazines. Polymer Chemistry.

2016;7(46):7177-7184.

129. Olson GB. Computational design of hierarchically structured materials. Science.

1997 Aug 29;277(5330):1237-1242.

130. Osman MA, Atallah A. High‐Density Polyethylene Micro‐and Nanocomposites:

Effect of Particle Shape, Size and Surface Treatment on Polymer Crystallinity and

208

Gas Permeability. Macromolecular rapid communications. 2004 Sep 9;25(17):1540-

1544.

131. Padinger F, Rittberger RS, Sariciftci NS. Effects of postproduction treatment on

plastic solar cells. Advanced Functional Materials. 2003 Jan 16;13(1):85-88

132. Paetzold R, Winnacker A, Henseler D, Cesari V, Heuser K. Permeation rate

measurements by electrical analysis of calcium corrosion. Review of Scientific

Instruments. 2003 Dec;74(12):5147-5150.

133. Park YT, Ham AY, Grunlan JC. High electrical conductivity and transparency in

deoxycholate-stabilized carbon nanotube thin films. The Journal of Physical

Chemistry C. 2010 Mar 15;114(14):6325-6333.

134. Part II – Photovoltaic Cell I-V Characterization Theory and LabVIEW Analysis

Code, http://www.ni.com/white-paper/7230/en/ (accessed September 2, 2017)

135. Pentzer EB, Bokel FA, Hayward RC, Emrick T. Nanocomposite “Superhighways”

by Solution Assembly of Semiconductor Nanostructures with Ligand ‐

Functionalized Conjugated Polymers. Advanced Materials. 2012 May

2;24(17):2254-2258.

136. Permeation Basices, Article: PPS 18-5102, Permeation,

http://www.mocon.com/assets/documents/pps-18-5102-permeation-basics.pdf

(accessed September 2, 2017).

209

137. Pojanavaraphan T, Schiraldi DA, Magaraphan R. Mechanical, rheological, and

swelling behavior of natural rubber/montmorillonite aerogels prepared by freeze-

drying. Applied Clay Science. 2010

138. Priolo MA, Gamboa D, Holder KM, Grunlan JC. Super gas barrier of transparent

polymer−clay multilayer ultrathin films. Nano letters. 2010 Nov 3;10(12):4970-

4974.

139. Priolo MA, Holder KM, Greenlee SM, Grunlan JC. Transparency, gas barrier, and

moisture resistance of large-aspect-ratio vermiculite nanobrick wall thin films. ACS

applied materials & interfaces. 2012 Oct 10;4(10):5529-5533.

140. Rand BP, Richter H, editors. Organic Solar Cells: fundamentals, devices, and

upscaling. CRC Press; 2014 Aug 26.

141. Randall D, Lee S. The Polyurethanes Book, vol. 14. Wiley: New York 2002.

142. Reim M, Körner W, Manara J, Korder S, Arduini-Schuster M, Ebert HP, Fricke J.

Silica aerogel granulate material for thermal insulation and daylighting. Solar

Energy. 2005 Aug 31;79(2):131-139.

143. Rheology Theory and Applications, TA, http://www.tainstruments.com/wp-

content/uploads/CA-2016-Rheology.pdf (accessed September 3, 2017)

144. Riganakos KA, Koller WD, Ehlermann DA, Bauer B, Kontominas MG. Effects of

ionizing radiation on properties of monolayer and multilayer flexible food packaging

materials. Radiation Physics and Chemistry. 1999 May 31;54(5):527-40.

210

145. Rossi A, Incarnato L, Tagliaferri V, Acierno D. Modification of barrier properties

of polymeric films of LDPE and HDPE by cold plasma treatment. Journal of

polymer engineering. 1995;14(2-3):191-197.

146. Ryan C, Christenson CW, Valle B, Saini A, Lott J, Johnson J, Schiraldi D, Weder

C, Baer E, Singer KD, Shan J. Roll‐to‐Roll Fabrication of Multilayer Films for High

Capacity Optical Data Storage. Advanced Materials. 2012 Oct 2;24(38):5222-5226.

147. Schaepkens M, Kim TW, Gün Erlat A, Yan M, Flanagan KW, Heller CM,

McConnelee PA. Ultrahigh barrier coating deposition on polycarbonate substrates.

Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films. 2004

Jul;22(4):1716-1722.

148. Scharber MC, Sariciftci NS. Efficiency of bulk-heterojunction organic solar cells.

Progress in polymer science. 2013 Dec 31;38(12):1929-1940.

149. Schiraldi DA, Gawyrla MD, Bandi SA, inventors; Schiraldi David A, Gawyrla

Matthew D, Bandi Suheel A, assignee. Absorbent compositions with clay aerogels

and methods for forming absorbent compositions. United States patent application

US 12/012,248. 2008 Feb 1.

150. Schmitz Jr JF. Enzyme modified soy flour adhesives. Iowa State University; 2009.

151. Sessa DJ, Eller FJ, Palmquist DE, Lawton JW. Improved methods for decolorizing

corn zein. Industrial crops and products. 2003 Jul 31;18(1):55-65.

211

152. Song Y, Tzeng P, Grunlan JC. Super Oxygen and Improved Water Vapor Barrier

of Polypropylene Film with Polyelectrolyte Multilayer Nanocoatings.

Macromolecular rapid communications. 2016 Jun 1;37(12):963-8.

153. Sun H, Schiraldi DA, Chen D, Wang D, Sánchez-Soto M. Tough polymer aerogels

incorporating a conformal inorganic coating for low flammability and durable

hydrophobicity. ACS applied materials & interfaces. 2016 May 11;8(20):13051-

13057.

154. Sun S. Recent Progress of Organic Photovoltaics. Advances in Solar Energy.

2007;17:74.

155. Sun, Y.; Takacs, C. J.; Cowan, S. R.; Seo, J. H.; Gong, X.; Roy, A.; Heeger, A.

J.Efficient, Air-Stable Bulk Heterojunction Polymer Solar Cells Using MoO(x) as

The Anode Interfacial Layer Adv. Mater. 2011, 23 (19) 2226–2230

156. Surapolchai W, Schiraldi DA. The effects of physical and chemical interactions in

the formation of cellulose aerogels. Polymer bulletin. 2010;65(9):951-960.

157. Surek T. Crystal growth and materials research in photovoltaics: progress and

challenges. Journal of Crystal Growth. 2005 Feb 15;275(1):292-304.

158. Uni-solar PVL-136 (136W) Solar Panel,

http://www.solardesigntool.com/components/module-panel-solar/Uni-

solar/154/PVL-136/specification-data-sheet.html (accessed September 2, 2017)

159. Viggiano III RP, Schiraldi DA. Fabrication and mechanical characterization of

lignin-based aerogels. Green Materials. 2014 Sep;2(3):153-158.

212

160. Viscosity & Rheology Glossary, http://www.rheologyschool.com/advice/rheology-

glossary (accessed September 3, 2017)

161. Wang J, Cheng Q, Tang Z. Layered nanocomposites inspired by the structure and

mechanical properties of nacre. Chemical Society Reviews. 2012;41(3):1111-1129.

162. Wang N, Yu J, Ma X. Preparation and characterization of thermoplastic starch/PLA

blends by one ‐ step reactive extrusion. Polymer International. 2007 Nov

1;56(11):1440-1447.

163. Wang TW, Sun H, Long J, Wang YZ and Schiraldi DA. "Biobased Poly (furfuryl

alcohol)/Clay Aerogel Composite Prepared by a Freeze drying Process." ACS

Sustainable Chemistry & Engineering 4.5 (2016): 2601-2605.

164. Wang Y, Al‐Biloushi M, Schiraldi DA. Polymer/clay aerogel‐based glass fabric

laminates. Journal of Applied Polymer Science. 2012 May 15;124(4):2945-2953.

165. Wang Y, Gawryla MD, Schiraldi DA. Effects of freezing conditions on the

morphology and mechanical properties of clay and polymer/clay aerogels. Journal

of Applied Polymer Science. 2013 Aug 5;129(3):1637-1641.

166. Wang Y, Gawryla MD, Schiraldi DA. Effects of freezing conditions on the

morphology and mechanical properties of clay and polymer/clay aerogels. Journal

of Applied Polymer Science. 2013 Aug 5;129(3):1637-1641.

167. Wang Y, Schiraldi DA. Foam-like materials produced from milk and sodium

montmorillonite clay using a freeze-drying process. Green Materials. 2013

Mar;1(1):11-15.

213

168. Webster DC, Sengupta PP, Chen Z, Pan X, Paramarta A, inventors; Ndsu Research

Foundation, assignee. Highly functional epoxidized resins and coatings. United

States patent US 9,096,773. 2015 Aug 4.

169. Wei W, Zhang C, Chen D, Wang Z, Zhu C, Zhang J, Lu X, Hao Y. Efficient “light-

soaking”-free inverted organic solar cells with aqueous solution processed low-

temperature ZnO electron extraction layers. ACS applied materials & interfaces.

2013 Dec 12;5(24):13318-13324.

170. Why Use Parylene?, Parylene Engineering,

http://www.paryleneengineering.com/why_use_parylene.htm (accessed September

2, 2017).

171. Wikimedia Commons contributors, "File:Electron Beam Deposition 001.jpg,"

Wikimedia Commons, the free media repository,

https://commons.wikimedia.org/w/index.php?title=File:Electron_Beam_Depositio

n_001.jpg&oldid=199384249 (accessed September 2, 2017).

172. Wikipedia contributors, "Electron-beam physical vapor deposition," Wikipedia,

The Free Encyclopedia, https://en.wikipedia.org/w/index.php?title=Electron-

beam_physical_vapor_deposition&oldid=797652091 (accessed September 2,

2017).

173. Wikipedia contributors, "Fick's laws of diffusion," Wikipedia, The Free

Encyclopedia,

https://en.wikipedia.org/w/index.php?title=Fick%27s_laws_of_diffusion&oldid=7

97952086 (accessed September 2, 2017).

214

174. Wikipedia contributors, "Parylene," Wikipedia, The Free Encyclopedia,

https://en.wikipedia.org/w/index.php?title=Parylene&oldid=788330598 (accessed

September 2, 2017).

175. Wool R, Sun XS. Bio-based polymers and composites. Academic Press; 2011 Aug

30.

176. Working Principle of Organic Solar Cells, https://www.iapp.de/organische-

elektronik.de/en/?OSCs___Working_Principle (accessed September 2, 2017)

177. Xie X, Rieth L, Williams L, Negi S, Bhandari R, Caldwell R, Sharma R, Tathireddy

P, Solzbacher F. Long-term reliability of Al2O3 and Parylene C bilayer encapsulated

Utah electrode array based neural interfaces for chronic implantation. Journal of

neural engineering. 2014 Mar 24;11(2):026016.

178. Yamada, Satoru, et al. "Solar cell module having a surface coating material of three-

layered structure." U.S. Patent No. 5,650,019. 22 Jul. 1997.

179. Yan C, Pochan DJ. Rheological properties of peptide-based hydrogels for

biomedical and other applications. Chemical Society Reviews. 2010;39(9):3528-

3540.

180. Yoshida K, Oku T, Suzuki A, Akiyama T, Yamasaki Y. Fabrication and

Characterization of PCBM: P3HT Bulk Heterojunction Solar Cells Doped With

Germanium Phthalocyanine or Germanium Naphthalocyanine. Materials Sciences

and Applications. 2013 Apr 24;4(04):1.

215

181. Yu D, Yang YQ, Chen Z, Tao Y, Liu YF. Recent progress on thin-film

encapsulation technologies for organic electronic devices. Optics Communications.

2016 Mar 1;362:43-49.

182. Zeman S, Kubík L. Permeability of polymeric packaging materials. Technical

Sciences/University of Warmia and Mazury in Olsztyn. 2007(10):26-34.

183. Zhang C, Pang S, Heng T, You H, Han G, Lu G, He F, Jiang Q, Zhang J. Stable

inverted low-bandgap polymer solar cells with aqueous solution processed low-

temperature ZnO buffer layers. International Journal of Photoenergy. 2016 Oct

9;2016.

184. Zhang G, Lee PC, Jenkins S, Dooley J, Baer E. The effect of confined spherulite

morphology of high-density polyethylene and polypropylene on their gas barrier

properties in multilayered film systems. Polymer. 2014 Aug 18;55(17):4521-4530.

185. Zhang G, Lee PC, Jenkins S, Dooley J, Baer E. The effect of confined

crystallization on high-density poly (ethylene) lamellar morphology. Polymer. 2014

Jan 30;55(2):663-672.

186. Zhang G, Xu H, MacInnis K, Baer E. The structure-property relationships of

LLDPE–EVOH blend films fabricated by multiplication extrusion. Polymer. 2015

Jan 28;57:117-124.

187. Zhang Y, Yip HL, Acton O, Hau SK, Huang F, Jen AK. A simple and effective

way of achieving highly efficient and thermally stable bulk-heterojunction polymer

216

solar cells using amorphous fullerene derivatives as electron acceptor. Chemistry of

Materials. 2009 Jun 8;21(13):2598-2600.

188. Zhou Z, Carr J, Mackey M, Yin K, Schuele D, Zhu L, Baer E. Interphase/interface

modification on the dielectric properties of polycarbonate/poly (vinylidene fluoride‐

co ‐ hexafluoropropylene) multilayer films for high ‐ energy density capacitors.

Journal of Polymer Science Part B: Polymer Physics. 2013 Jun 15;51(12):978-991.

217