MODIFICATION OF ALKYD RESINS AND SEED OIL BASED REACTIVE

DILUENTS FOR HIGH PERFORMANCE COATINGS

A Dissertation

Presented to

The Graduate Faculty of The University of Akron

In Partial Fulfillment

of the Requirements for the Degree

Doctor of Philosophy

Brittany A. Pellegrene

August, 2019

MODIFICATION OF ALKYD RESINS AND SEED OIL BASED REACTIVE

DILUENTS FOR HIGH PERFORMANCE COATINGS

Brittany A. Pellegrene

Dissertation

Approved: Accepted:

Advisor Department Chair Dr. Mark Soucek Dr. Sadhan Jana

Committee Member Interim Dean of the College Dr. Thein Kyu Dr. Ali Dhinojwala

Committee Member Dean of Graduate the School Dr. Younjin Min Dr. Chand Midha

Committee Member Date Dr. Tianbo Liu

Committee Member Dr. Chelsea Monty-Bromer

ii

ABSTRACT

Alkyds, one of the most commonly used binders for coating systems, are modified polyesters derived from seed oils. They find utility in several coating applications, including architectural, industrial and wood coatings. Formulation involves the use of reactive diluents to decrease the viscosity and trigger the autoxidative curing mechanism of the alkyds to avoid the use of volatile organic compounds (VOCs). This work studies the modification of alkyds and reactive diluents and the differences between the coating performance of these additives.

Two differently functionalized alkyds and reactive diluents were synthesized and formulated into high solids alkyds coatings. Alkoxysilane and fluorine functionalities were chosen to improve adhesion, hardness, and chemical and corrosion resistance of the coating system. The resulting coatings were analyzed for performance, tensile properties, corrosion resistance and weatherability. ESEM-EDX was used to observe the distribution of the fluorine and alkoxysilane in the cross-section of the coating. Stratification was observed for the modified reactive diluents at high concentrations, and these coatings showed improved adhesion and corrosion resistance. The modified alkyds performed better in terms of mechanical properties, but stratification was not observed.

iii

Next, the moisture sensitivity of alkoxysilanes was studied by looking into the effect of various relative humidity conditions on the curing and performance of alkoxysilane-functionalized alkyd coatings. These coatings were evaluated for drying time, adhesion, hardness and mechanical properties. At high humidity, the alkoxysilane- functional reactive diluents dried more quickly and formed harder coatings than the unmodified control. The functionalized alkyds showed enhanced adhesion and tensile strength at high humidity.

Thirdly, fluorinated alkyds and reactive diluents were compared to understand the effects of molecular weight and viscosity on the stratification and performance of these coatings. The additives were found to create more hydrophobic and chemically resistant coatings without detrimental effects on other properties. The reactive diluents showed more stratification, with decreased mechanical properties. The fluorinated alkyd system improved corrosion resistance without sacrificing other coating properties. In this case, the added mobility of the reactive diluent did not improve the final properties of the coatings.

In the final part of this work, novel soybean oil-based reactive diluents were developed, based on an ene reaction between soybean oil and maleic anhydride, which was characterized using FT-IR, 1H-, 13C- and 2D NMR spectroscopy and MALDI-ToF mass spectrometry. The resulting maleated soybean oil was further modified with various nucleophiles, containing allyl ether or methacrylate functionalities. The efficacy of these reactive diluents in decreasing viscosity was evaluated, as well as the crosslink density, tensile properties and coating performance. These novel reactive diluents efficiently reduced viscosity without a significant decrease in performance of the resulting coatings.

iv

ACKNOWLEDGMENTS

I would like to thank the wonderful people without whom this would not have been possible.

Firstly, I would like to acknowledge my advisor, Dr. Mark Soucek, and thank him for his counsel, encouragement and support throughout my time here. I’d also like to thank my committee members, Dr. Thein Kyu, Dr. Tianbo Liu, Dr. Younjin Min and Dr. Chelsea

Monty-Bromer for their comments and advice. I would like to especially thank Dr. Monty-

Bromer for her help with modeling corrosion data. I would like to extend gratitude to Dr.

Kevin Cavicchi for the financial support in these last few semesters and for always offering guidance and ideas. Additionally, to the staff and technicians in the department, your help was always greatly appreciated, especially Dr. Paula Watt, Christopher Paige, and Thomas

Quick. I also want to thank Diana Woolf for her willingness to go above and beyond for students, and for her support and encouragement to me throughout this process.

I would like to express sincere gratitude to my colleagues in the Department of

Polymer Engineering. Thank you to Dr. Ryan Salata for your mentorship and for teaching me the ways of alkyds. To Dr. Nick Teo and Marisa Tukpah, thank you for your friendship and for many lunches and study sessions. To my lab mates in Dr. Soucek’s group, especially Anisa Cobaj, thank you so much for your support, advice and understanding. I

v

would also like to especially thank Ted Hammer for his experimental support and for always being willing to help.

I want to express sincere gratitude to my parents and family for their constant love and encouragement, as well as for instilling in me the work ethic and follow-through necessary to achieve my goals. I definitely could not have made it this far without you.

And last, but definitely not least, to my very best friends, Emily Brahler, Audrey Fletcher,

Anna Holdren, J.D., and Becky VanVoorhis, thank you for everything, for letting me complain and being there for me throughout these last 15 plus years of friendship. I owe a lot to each one of you.

Finally, and above all, I want to express my sincere gratitude and appreciation to

Dillon Lloyd for his unwavering support, encouragement and the unending sacrifices he has made throughout this journey and over the last 9 years. Your love, advice and encouragement have been instrumental in helping me reach this achievement.

vi

TABLE OF CONTENTS

Page

LIST OF FIGURES ...... x

LIST OF TABLES ...... xvii

CHAPTER

...... 1

...... 4

2.1 Environmentally Friendly Coatings ...... 4

2.2 Seed Oils ...... 5

2.3 Linseed oil ...... 8

2.4 Autoxidation and driers ...... 8

2.5 Alkyd Resins ...... 11

2.6 High Solids Formulations ...... 17

2.7 Reactive Diluents ...... 18

2.8 Ene Reactions ...... 20

2.9 Inorganic/Organic Hybrid Coatings ...... 21

2.10 Fluorinated Polymers ...... 25

2.11 Self-Stratifying Coatings ...... 26

2.12 Corrosion...... 28

2.13 Exterior Durability ...... 31

vii

...... 36

3.1 Abstract ...... 36

3.2 Introduction ...... 37

3.3 Experimental ...... 39

3.4 Results ...... 47

3.5 Discussion ...... 72

3.6 Conclusions ...... 77

...... 80

4.1 Abstract ...... 80

4.2 Introduction ...... 80

4.3 Experimental ...... 83

4.4 Results ...... 87

4.5 Discussion ...... 98

4.6 Conclusions ...... 102

...... 104

5.1 Abstract ...... 104

5.2 Introduction ...... 104

5.3 Experimental ...... 106

5.4 Results ...... 111

viii

5.5 Discussion ...... 125

5.6 Conclusions ...... 127

...... 129

6.1 Abstract ...... 129

6.2 Introduction ...... 129

6.3 Experimental ...... 132

6.4 Results ...... 141

6.5 Discussion ...... 163

6.6 Conclusions ...... 168

6.7 Acknowledgments ...... 168

...... 170

...... 173

...... 203

APPENDIX A. NMR SPECTRA OF SOYBEAN OIL AND MALEATED

SOYBEAN OIL ...... 204

ix

LIST OF FIGURES

Figure Page

Figure 2.1. Structure of triglyceride, the main component in seed oil. The R groups are

fatty acid chains...... 5

Figure 2.2. Autoxidation process...... 10

Figure 2.3. Epoxide intermediates that form during the autoxidation process...... 11

Figure 2.4. Monoglyceride process for alkyd synthesis...... 12

Figure 2.5. Fatty acid process for alkyd synthesis...... 13

Figure 2.6. Reaction set up for alkyd synthesis, including nitrogen flow, mechanical

stirring, a Dean-Stark trap and condenser...... 14

Figure 2.7. The ene reaction...... 20

Figure 2.8. Sol-gel process for alkoxysilanes...... 23

Figure 3.1. Synthesis scheme for alkoxysilane functional alkyd (ASLOA)...... 48

Figure 3.2. 1H-NMR spectrum of alkoxysilane functional alkyd (ASLOA)...... 49

Figure 3.3. FT-IR spectrum of alkoxysilane functional alkyd (ASLOA)...... 49

Figure 3.4. MALDI MS of alkoxysilane modified alkyd (bottom), compared with

unmodified alkyd (top)...... 50

Figure 3.5. Reaction scheme for fluorinated alkyd (FLOA)...... 51

Figure 3.6. 1H-NMR of fluorinated alkyd (FLOA)...... 51

x

Figure 3.7. MALDI-MS of fluorinated alkyd (FLOA)...... 51

Figure 3.8. Reaction scheme for alkoxysilane functionalized tung oil (ASTO)...... 52

Figure 3.9. FT-IR spectrum of alkoxysilane functional tung oil (ASTO)...... 53

Figure 3.10. 1H-NMR of alkoxysilane functional tung oil (ASTO)...... 53

Figure 3.11. MALDI MS of alkoxysilane functional tung oil (ASTO)...... 54

Figure 3.12. Reaction scheme for fluorinated tung oil (FTO)...... 54

Figure 3.13. 1H-NMR spectrum of fluorinated tung oil (FTO)...... 55

Figure 3.14. MALDI MS of fluorinated tung oil (FTO)...... 55

Figure 3.15. Comparison of tensile properties between the reactive diluent series (left) and

the modified alkyd series (right)...... 59

Figure 3.16. Impedance modulus of the modified alkyd series initially (left) and after 10

weeks (left)...... 60

Figure 3.17. Impedance modulus of the RD series, initially (left) and after 10 weeks (right).

...... 61

Figure 3.18. Comparisons of impedance modulus for the modified alkyds and reactive

diluent series...... 61

Figure 3.19. Equivalent electric circuit for EIS modeling of these coatings...... 63

Figure 3.20. Nyquist plots and fittings for control over 6 weeks...... 64

Figure 3.21. Nyquist plots and fittings for ST-10 over 6 weeks...... 64

Figure 3.22. Nyquist plots and fittings for STRD-10 over 6 weeks...... 65

Figure 3.23. Percent change in pendulum hardness after 500 h of weathering, with reactive

diluents (left) and modified alkyds (right)...... 66

xi

Figure 3.24. EDX dot maps of the cross section of the modified alkyd coating series. ... 69

Figure 3.25. EDX dot maps of fluorine and silicon for the modified reactive diluent series,

STRD 2.5-10...... 71

Figure 3.26. EDX dot maps of fluorine and silicon for the modified reactive diluent series,

STRD 15-20...... 72

Figure 3.27. Schematic of stratifying behavior in reactive diluent series (top) and modified

alkyd series (bottom)...... 76

Figure 4.1. Percent change in pendulum hardness when compared to the control alkyd.

ASTO series (left) and ASLOA series (right)...... 92

Figure 4.2. Pull-off adhesion for ASTO (left) and ASLOA series (right)...... 92

Figure 4.3. Tensile properties of the ASTO (left) and ASLOA (right) series, cured at 25

%RH...... 94

Figure 4.4. Tensile properties of the ASTO (left) and ASLOA (right) series, cured at 75

%RH...... 95

Figure 4.5. Percent change in modulus (left) and tensile strength (right), compared to the

control sample...... 95

Figure 4.6. Gel content results of the ASTO (left) and ASLOA (right) series, cured at 25

%RH...... 97

Figure 4.7. Gel content of the ASTO (left) and ASLOA series (right), cured at 75 %RH.

...... 97

Figure 4.8. Gel content percent change...... 98

Figure 5.1. Reaction scheme for FLOA (a) and FTO (b)...... 113

xii

Figure 5.2. 1H-NMR spectrum of FLOA...... 114

Figure 5.3. MALDI-TOF MS of FLOA...... 114

Figure 5.4. SEM-EDX mapping of a cross-section of FLOA-30 (left) and FTO-2.5 (right)

...... 116

Figure 5.5. Tensile strength (top), strain at break (middle) and modulus (bottom) for FLOA

(left) and FTO (right) coatings...... 118

Figure 5.6. Contact angle for FLOA coatings...... 120

Figure 5.7. Bode plots for FLOA coatings after 2 h of exposure to 3.5 wt.% NaCl solutions.

...... 121

Figure 5.8. Bode plots for FLOA coatings after 37 days of exposure to 3.5 wt.% NaCl

solutions...... 121

Figure 5.9. Bode plots comparing the impedance modulus of FLOA0 over 37 days of

exposure...... 122

Figure 5.10. Bode plots comparing the impedance modulus of FLOA20 over 37 days of

exposure...... 122

Figure 5.11. Bode plots of the impedance of FTO coatings, initially...... 123

Figure 5.12. Bode plots of impedance for the FTO coatings, after 42 days of exposure.

...... 124

Figure 5.13. Bode plots of impedance over time for FTO-0...... 124

Figure 5.14. Bode plots of impedance over time for FTO-7.5...... 125

Figure 6.1. Acrylated epoxidized soybean oil...... 131

Figure 6.2. Reaction scheme for MA-SBO...... 141

xiii

Figure 6.3. MALDI-ToF MS of SBO (top) and MA-SBO (bottom), showing no high

molecular weight species in MA-SBO...... 142

Figure 6.4. Reaction scheme of ring-opening model reaction...... 143

Figure 6.5. Viscosity of (a.) MA-SBO with varying degrees of maleation, and (b.) after

functionalization of MS-SBO with ethylene glycol monopropyl ether...... 143

Figure 6.6. FT-IR spectra overlay of SBO (top) and the MA-SBO (bottom)...... 145

Figure 6.7. FT-IR spectra of real-time reaction monitoring for the maleation reaction

between 1600-1950 cm-1 (top) and 780-1000 cm-1 (bottom)...... 145

Figure 6.8. 1H NMR spectra of MA-SBO...... 147

Figure 6.9. 13C NMR spectra of MA-SBO...... 147

Figure 6.10. 2D NMR spectra of MA-SBO with HSQC (left) and HMBC (right)...... 148

Figure 6.11. 2D HMBC NMR spectra of MA-SBO (left) which shows coupling between

the anhydride carbonyl carbons and the protons in the 5-membered ring that are not

present in SBO (right)...... 148

Figure 6.12. 2D HMBC NMR spectra of MA-SBO (left) which shows coupling between

the conjugated double bonds and the anhydride ring; the unmodified SBO (right)

does not exhibit this coupling behavior...... 149

Figure 6.13. MALDI-MS of SBO (top) and MA-SBO (bottom)...... 150

Figure 6.14. Reaction products for MA-SBO-based reactive diluents...... 151

Figure 6.15. Real-time FT-IR monitoring of the reaction of AGE and DEGMBE with MA-

SBO from 1600 to 1900 cm-1...... 152

xiv

Figure 6.16. FT-IR spectra overlay of MA-SBO (top) and the TMPDAE- functionalized

MA-SBO (bottom)...... 152

Figure 6.17. FT-IR spectra overlay of MA-SBO (top) and the HEMA-functionalized MA-

SBO (bottom)...... 153

Figure 6.18. FT-IR spectra overlay of MA-SBO (top) and AGE with DEGMBE (bottom).

...... 153

Figure 6.19. FT-IR spectra overlay of MA-SBO (top) and GMA with DEGMBE (bottom).

...... 153

Figure 6.20. 1H-NMR spectrum of TMPDAE-functionalized MA-SBO...... 154

Figure 6.21. 1H-NMR spectrum of HEMA-functionalized MA-SBO...... 155

Figure 6.22. 1H-NMR spectrum of AGE-functionalized MA-SBO...... 155

Figure 6.23. 1H-NMR spectrum of GMA-functionalized MA-SBO...... 156

Figure 6.24. Viscosity of the formulations with increasing amounts of RDs...... 157

Figure 6.25. Gel content of the coatings analyzed by Soxhlet extraction...... 159

Figure 6.26. Tensile properties, including strain at break (top, left), modulus (top, right)

and tensile strength (bottom)...... 160

Figure 6.27. Pendulum hardness (top, left), pencil hardness (top, right) and pull-off

adhesion (bottom)...... 163

Figure 6.28. Various configurations following the maleation reaction...... 165

Figure A.1. 1H NMR spectrum of SBO...... 204

Figure A.2. 13C NMR spectrum of soybean oil...... 204

Figure A.3. 1H NMR and 2D HMBC NMR spectra of MA-SBO...... 205

xv

Figure A.4. 2D HSQC NMR spectra of MA-SBO (left) and SBO (right)...... 205

Figure A.5. 2D HMBC NMR spectra of MA-SBO (right) and SBO (left)...... 206

xvi LIST OF TABLES

Table Page

Table 2.1. Structures of the fatty acids in common seed oils.3,19 ...... 6

Table 2.2. Fatty acid content of various seed oils.3 ...... 7

Table 2.3. Typical viscosity for different coating application techniques...... 19

Table 3.1. Modified alkyd formulations. All amounts are in g...... 45

Table 3.2. Modified tung oil reactive diluent alkyd formulations. All amounts are in g. 45

Table 3.3. Viscosity measurements of alkyd, modified alkyds and modified reactive

diluents...... 56

Table 3.4. Coatings properties of the modified alkyd series...... 57

Table 3.5. Coatings properties of the reactive diluent series...... 58

Table 3.6. Coating capacitance calculated from Equation (12), using the equivalent circuit

shown in Figure 3.19...... 63

Table 3.7. Water uptake, calculated from Equation (13), using the equivalent circuit shown

in Figure 3.19...... 63

Table 3.8. Gloss changes following 500 h of weathering...... 67

Table 3.9. Stratification toward the surface (↑) for modified alkyd series...... 68

Table 3.10. Stratification of fluorine and silicon toward the surface (↑) or the substrate (↓)

interface of the coating for the reactive diluent series...... 72

xvii Table 4.1. Formulations with alkoxysilane modified alkyd (ASLOA). All amounts are in

g...... 86

Table 4.2. Formulations with alkoxysilane modified tung oil (ASTO). All amounts are in

g...... 86

Table 4.3. Drying times of samples cured in different relative humidity...... 89

Table 4.4. Coating properties of samples cured at 25 %RH...... 90

Table 4.5. Coating properties of samples cured at 75 %RH...... 91

Table 4.6. Glass transition temperature of the samples cured at varying relative humidity.

...... 96

Table 5.1. Formulations for FLOA- and FTO-containing alkyd coatings...... 110

Table 5.2. Viscosity of the coating components...... 116

Table 5.3. Coating test results of FLOA and FTO for hardness, flexibility and contact angle.

...... 119

Table 5.4. Coatings properties for FLOA and FTO for chemical resistance and adhesion.

...... 119

Table 6.1. Comparison of experimental and theoretical AN to determine the degree of

maleation. aTheoretical AN was determine from Equation (16)...... 144

Table 6.2. The crosslink density, ν calculated from Equation (18) using the Tg,u and Tg,

measured by DSC...... 159

Table 6.3. Coating properties...... 162

xviii CHAPTER I

INTRODUCTION

Alkyd resins were once the most widely used coatings, and they were first discovered by Keinle in the 1920s.1,2 Alkyds are modified polyesters, which are synthesized from polyhydric alcohol, polybasic acid, and monobasic fatty acids from triglycerides (seed oils). Alkyds mainly find use as the binder material in coatings, and they are unique because of their adjustable properties and versatility for many different modifications, which allow for fine-tuning to tailor properties for various end-use applications. Alkyds also have an autoxidative curing process, which involves crosslinking between the pendant fatty acid chains from the seed oils. The main advantages of alkyds are high gloss, few defects during film formation, low cost and good solvent resistant.

However, alkyds are lacking in hydrolytic stability and outdoor durability, and organic solvents are usually necessary to reach the proper application viscosity.1–5

To reduce the need for organic solvents or volatile organic components (VOCs),

the coating industry has developed high solids coatings, which utilize little to no VOCs.

Alkyds are an excellent candidate for high solids coatings because the molecular weight

and viscosity of the resin can be easily adjusted, and they are compatible with several

common reactive diluents, which are additives that have the ability to crosslink into the

coating network during curing. These compounds effectively lower viscosity without

volatilizing, replacing the need for VOCs in a coating. Initially, seed oils themselves were

1 used as reactive diluents, and then research began to modify seed oils in a variety of ways

to create more reactive and efficient diluents. Commercially, Cargill® introduced Dilulin® as a reactive diluent based on a Diels-Alder modification of linseed oil to create norbornene groups on the fatty acid chains.6 Several other seed oil modifications have also been

reported.7–14

The first objective (Chapter 3) of this work was to evaluate the performance and stratification of two functionalized alkyds and two functionalized reactive diluents used as additives in high solids alkyd coating formulations. The combination of alkoxysilane and fluorine functionalities was chosen to improve adhesion, hardness, and chemical and corrosion resistance of the coating system. Looking at both modified alkyds and reactive diluents allows for a comparison of the coating properties and stratification based on the molecular weight and viscosity of the additives. The coatings were evaluated for performance, mechanical properties, corrosion resistance and weatherability. The cross- section of the coating was also observed under ESEM-EDX spectroscopy to analyze the distribution of fluorine and alkoxysilane in the cross-section of the coating. It was expected that alkoxysilane-functionalized components would migrate toward the substrate to interact with free hydroxyl groups on the aluminum surface, and the fluorinated components would move toward the air interface due to their low surface energy.

Due to the ability of the alkoxysilane functionality to react with water, Chapter 4 focused on the performance of alkoxysilane additives when cured in varying humidity conditions. These components were used in alkyd coating formulations with increasing concentrations, and then were cured in 25 and 75 % relative humidity (RH). The performance of these coatings was evaluated for drying time, hardness, adhesion and

2 tensile properties. The difference in performance of modified alkyds and reactive diluents was analyzed.

Comparison of properties between the fluorinated alkyd and fluorinated reactive diluent when used as additives was carried out in Chapter 5. The fluorinated alkyd was synthesized to have fluorine functionality in the backbone of the polymer, whereas the reactive diluent had the functionality in a flexible, fatty acid chain. The effect of the architecture of these coatings, as well as molecular weight and viscosity was explored via stratification studies, contact angle, tensile properties and corrosion resistance of these coatings.

Lastly, Chapter 6 describes the development of novel reactive diluents based on maleated soybean oil. These reactive diluents were developed to be solvent-free in both synthesis and formulation of the coatings. Through a facile ene reaction between maleic anhydride and soybean oil, a succinic anhydride group is grafted to the soybean oil. This was confirmed by FT-IR, 1H-, 13C-, and 2D NMR spectroscopy. The anhydride was further functionalized using nucleophilic groups, including various allyl ether and methacrylate functionalities, which were chosen to improve crosslinking in the final coating. The coatings were analyzed for performance, mechanical properties and crosslink density.

3 CHAPTER II

BACKGROUND

2.1 Environmentally Friendly Coatings

Coatings are an essential part of everyday life that often go overlooked. Vehicles, buildings, infrastructure and electronics require coatings. Coatings can be categorized in several ways, including by their appearance, function, and type of binder (inorganic or organic). Organic coatings are complex mixtures, usually made up of binders, volatile components, pigments, and additives. The binder is the organic matrix that generally makes up most of the coating mixture. It is the polymer component of the coating that adheres to

the substrate and form the hardened coating layer. The volatile component is necessary in

most coatings to aid in the application process and evaporate during and after the application.3 This volatile component is often an organic solvent, but concerns about

health, safety and the environment are leading many coatings companies to develop

alternatives to using volatile organic compounds (VOCs) in their coating formulations.

Reducing these organic volatiles has been a driving force in the coating industry, leading

to a wider use of waterborne coatings and high solids formulations, where the binder is

concentrated and a volatile compound is used in a very small amount.15 Pigments in

coatings refer to the use of solid particles dispersed in the coating matrix. The main use of

pigments is to provide color or opacity to the coating, but they can also be used for abrasion

4 resistance or rheology modification. Additives are other materials used in coatings, while

can include among others, catalysts, surface or rheology modifiers and stabilizers.3

Creating environmentally friendly coatings is a two-fold process. First, creating more water-based coatings without volatile organic components leads to better health and safety.15 Additionally, creating coatings from renewable resources, rather than petrochemical feedstocks could lead to more sustainable materials in the long term. These approaches are both necessary in a more environmentally-conscious society. Many companies are working toward water-based versions of previously solvent-based coating systems. Another approach is to use seed oils as a feedstock for the organic binders in coatings. This creates a more sustainable process, which can utilize seeds that may not be used as a food source, such as linseeds.15

2.2 Seed Oils

Seed oils are naturally occurring triglycerides, which are the triesters of glycerol

and fatty acids as seen in Figure 2.1. The fatty acid composition of these seed oils varies

based on the type of seed and the region and conditions during growth of the seed. Resins

based on seed oils are one of the oldest known coating materials and were still used as the

binders in most paints through the early twentieth century. Their use has decreased, but

they are still used in a variety of applications, including alkyds and epoxy esters.1,3,15

O O

R1 O O R3 O O

R2 Figure 2.1. Structure of triglyceride, the main component in seed oil. The R groups are fatty acid chains.

5 Seed oil-based resins have several advantages over other resins, including their low

cost, abundance in the market, low toxicity and ability to crosslink in the presence of

oxygen. The most commonly used oils are linseed, safflower, soybean, tung, tall, and

sunflower. The difference in these oils is based on the amounts and types of fatty acids

present. The fatty acids vary by number of carbons, number and location of double bonds,

cis (c-) or trans (t-) configuration of the double bonds and the location of these configurations.3,16–23

Table 2.1 gives the structures, names and double bond information on commonly

occurring fatty acids. The variance between these groups and small differences in

molecular weight make characterization difficult, but it is usually done using high pressure

liquid chromatography (HPLC) or gas chromatography (GC). The fatty acid content of

seed oils tends to vary from batch-to-batch, and growing conditions and location can

significantly affect the fatty acid content in seed oils, even of the same species. It is

important for producers of seed oil-based resins to find a reliable supplier of these oils to

decrease batch-to-batch differences.16–23

Table 2.1. Structures of the fatty acids in common seed oils.3,19 R = Fatty Double Structure Acid Bond Stearic CH (CH ) COOH Saturated Acid 3 2 16 Palmitic CH (CH ) COOH Saturated Acid 3 2 14 Oleic CH (CH ) CH=CH(CH ) COOH 9c Acid 3 2 7 2 7 Linoleic CH (CH ) CH=CHCH CH=CH(CH ) COOH 9c, 12c Acid 3 2 4 2 2 7 Linolenic CH CH CH=CHCH CH=CHCH CH=CH(CH ) COOH 9c, 12c, 15c Acid 3 2 2 2 2 7 α-Eleostearic CH (CH ) CH=CHCH=CHCH=CH(CH ) COOH 9t, 11c, 13t Acid 3 2 3 2 7

6 Seed oils can be characterized as drying or non-drying oils. Drying oils have a certain concentration of 1,4-dienes, which comprise diallylic methylene groups and react with oxygen to form solid films. The diallylic methylene groups are the crosslinking sites in the fatty acid chains, and the amount present per molecule determines the drying ability of a seed oil. If the average number of diallylic sites per molecule is over 2.2, that seed oil is considered to oil to be drying. More commonly used is the drying index, which is calculated from the percentage of two specific fatty acids within the seed oil, as seen in

Equation (1).3,16

������ ����� = (%�������� ����) + 2(%��������� ����) (1)

For a nonconjugated oil to be considered drying, the drying index must be over 70.3

Table 2.2 lists the fatty acid content of common seed oils, which is used to calculate the drying index. As is seen in Table 2.1, linoleic acid contains one diallylic methylene group, and linolenic acid contains two of these groups. Therefore, this calculation is a measure of the amount of diallylic methylene groups within a molecule. Because these are crosslinking sites, the average number of these groups is directly related to the number average

̅ 3,16,24,25 functionality � of the triglyceride.

Table 2.2. Fatty acid content of various seed oils.3 Seed Oils Saturateda Oleic Linoleic Linolenic Other Linseed21,22 10 22 16 52 - Safflower26 11 13 75 1 - Soybean23 15 25 51 9 - Tung27 5 8 43 80b - Tall, American3,28 8 46 41c 3 2d Tall, European3,28 2.5 30 45 1 14e aSaturated fatty acids mixtures of stearic and palmitic acids. b�-Eleostearic acid. cLinoleic plus geometric and conjugated isomers dRosin ePinoleic acid

7 2.3 Linseed oil

Linseed oil, also known as flaxseed oil is used for several different applications.

The oil is obtained by pressing the seeds of the flax plant, Linum usitatissimum, which is

grown mainly in China, Canada, and Kazakhstan for food and fiber. Textiles made from

flax are commonly used as bedsheets, undergarments and table linens. Flaxseeds and

linseed oil are both used as nutritional supplements as a source of linolenic acid, which is

an omega-3 fatty acid.21,22,29

In addition to the plant’s uses as fibers and nutritional supplements, the seeds can be pressed to extract the oil. This oil can be used in several different applications, but is

most commonly used in coatings applications. Linseed oil contains triglycerides of fatty

acids, including linolenic, linoleic, oleic, stearic, palmitic and myristic acids. The amount

of these acids can vary slightly, but the typical amounts are shown in Table 2.2. The high

content of linolenic and linoleic acids make linseed oil a good drying oil, which lends to

its use as a coating and as a common raw material for alkyds and epoxy esters.3,21,22,6

2.4 Autoxidation and driers

Drying oils polymerize and form solid films in oxygen by a process known as autoxidation. This process is a free radical chain process that takes place in three stages: initiation, propagation and termination. The process takes place relatively slowly because naturally present antioxidants must first be consumed in an induction period. Following this, oxygen is taken up and there is about a 10% weight gain in the coating. Finally, the autocatalytic crosslinking reactions take place. In this stage, hydroperoxides and the carbon-carbon double bonds are consumed as crosslinks are formed. These reactions occur very slowly, and are often are catalyzed by metal driers. These driers are metal salts, which 8

help to decompose the hydroperoxides by cycling the oxidation states. In non-conjugated fatty acids, such as linolenic and linoleic acids, free radicals abstract the hydrogen from the diallylic methylene group. The radical formed is stabilized by the delocalization of the electron by the double bonds, and the oxygen attacks this radical to form hydroperoxides.

The decomposition of these hydroperoxides leads to more radicals, which abstract hydrogens from other diallylic methylene groups, and this leads to crosslinks. In conjugated drying oils, such as tung oil, drying occurs much more rapidly because the oxygen undergoes a 1,4-addition, leading to the formation of an unstable cyclic peroxide, which decomposes quickly into peroxy free radicals.3,30–35

The alkoxy and peroxy radicals are the main sites where termination takes place, creating new peroxy and ether linkages, as shown in lines 5 and 6 of Figure 2.2. Many studies have been done to determine the main crosslinks that form during autoxidation.31,33,34,36–40 Studies using 1H- and 13C-NMR have shown that the predominant crosslinks are peroxy linkages,38,40 and mass spectroscopic studies showed that about 5% of the linkages were new C-C bonds.40 However, looking at FT-IR spectroscopy and FT

Raman analysis, only ether and C-C bonds were shown to be formed. Additionally, epoxide groups are present in the reaction mixture during curing. The formation of these is shown in Figure 2.3. These are shown to reach a maximum in about 5 days, but disappear after about 100 days. This suggests that these epoxides continue to react, most likely with carboxyl groups that form from oxidation of aldehydes. These reactions form ester crosslinks.33,34,36

Metal salts, called driers, are used to catalyze the autoxidation process to make it occur within hours, rather than days or weeks.3,31,39,41–45 Drier packages are typically made

9 of three different metal salts, known as primary (surface) driers, secondary (through curing) driers, and auxiliary driers. Primary driers are typically cobalt, manganese or iron salts.

Secondary driers are zirconium, bismuth, barium or aluminum. Finally, the auxiliary driers are commonly calcium, potassium, lithium or zinc. The ratio of driers is extremely important to proper curing of the coating. If the ratio of primary to secondary is too high, the coating surface will dry too quickly, and this change in free volume at the surface will not match that at the substrate and will result in wrinkling. The proper balance of this ratio is necessary to achieve full curing from the substrate to the surface.3,31

Initiation Hydroperoxide OOH O O2 decomposition OO or (1) R R' R R' R R' R R' 1 2 3 4 Hydrogen abstraction O2 + 2 3 (2) 1 + 2 or 3 R R' R R' 5 6 Termination R' OH O + + O O O R2 2 (3) 3 + 3 R O O R R' R1 R2 7 8 R1 R' R 6 + 6 2 (4) R R1 9

R' R1 (5) 6 + 4 R O R2 10 R' (6) O R2 4 + 4 R O R1 11 Figure 2.2. Autoxidation process.

10 OH OOH O O (7) R1 R2 R R 1 2 R1 R2 2 4 12

OO R O + 2 O O R + (8) R1 R2 R3 R4 3 R1 O R R 1 2 R3 R4 14 4 15 3 13 R4

O OH O O2 13 O O O (9) 4 + + + R5 R R R H 1 2 5 R5 OO R 5 O R3 R4 16 17 18 19 20 15

Figure 2.3. Epoxide intermediates that form during the autoxidation process.

The most commonly used primary drier is cobalt because it is highly active and

efficient in both solvent- and waterborne coatings. However, health and toxicity concerns

are leading to a phasing out of cobalt driers. Replacements are generally manganese or

iron-based salts. The manganese salts are generally dark in color and can cause

discoloration in clear or white coats, but they are still commonly used, especially in

pigmented systems.43,45 Lead salts used to be used as secondary driers, but after health

concerns became apparent, zirconium became the most commonly used.39 Calcium is the

most commonly used auxiliary drier, and it promotes drying by increasing the efficiency

of the other two driers, by preferentially binding to pigment surfaces thereby decreasing

adsorption of active driers.3,4,31

2.5 Alkyd Resins

Alkyd resins were once the most widely used coatings, and they were first

discovered by Keinle in the 1920s.1,2 Alkyds are modified polyesters, which are synthesized from polyhydric alcohol, polybasic acid, and monobasic fatty acids from

11 triglycerides (seed oils). Alkyds mainly find use as the binder material in coatings, and they

are unique because of their adjustable properties and versatility for many different

modifications. These modifications allow for fine-tuning of the properties to tailor them

for various end-use applications. The main advantages of alkyds are high gloss, few defects during film formation, low cost and good solvent resistant. However, alkyds are lacking in hydrolytic stability and outdoor durability, and they have a slow autoxidative curing mechanism.1–5 Synthesis of alkyds can be done by the two-step monoglyceride process or

the one-step fatty acid process. The fatty acid process offers better control, but the

monoglyceride process is more cost effective.3,4,24,25 This research used the fatty acid

process. The monoglyceride process is shown in Figure 2.4, and the fatty acid process is

shown in Figure 2.5.

O O LiOH HO OH R1 O O R3 + HO OH O O OH heat O O R R2 Seed Oil Glycerol Monoglyceride

O O O O O O O HO OH xylene O + HO O O O OH + H2O O heat O O R O O R n R Monoglyceride Phthalic Anhydride Alkyd Figure 2.4. Monoglyceride process for alkyd synthesis.

12 O O O O O O O O + HO OH + xylene HO O O O OH + H2O HO R OH O OH heat O Fatty Acids Glycerol Phthalic R n Anhydride Alkyd Figure 2.5. Fatty acid process for alkyd synthesis.

The fatty acid process allows for the use of polyols other than glycerol, but it

requires the use of fatty acids, rather than the seed oil. The seed oil is saponified to produce

fatty acids, but the cost to separate the acids from the reaction mixture leads to increased

costs of synthesizing alkyds by this method. In this process, there is still a mixture of fatty

acids, but the transesterification step is skipped, which reduces some of the uncertainty in

the final product. For the fatty acid process, the polyol, fatty acids and dibasic acid are

added at the start of the reaction, and the esterification of the aliphatic and aromatic acids

happens simultaneously at temperatures from 220-260 °C. Evolved water is removed via a

Dean-Stark trap as the reaction takes place to drive the polymerization forward.4,24,25

Alkyds are characterized by three main distinctions: oil length, modification and

oxidation. The oil length refers to the amount of seed oil or fatty acid used in the synthesis.

It is essentially a measure of polyester backbone content to fatty acid content. Modification

refers to the addition of other functional groups. Modified alkyds will have undergone

further reactions to form products such as, styrenated alkyds, uralkyds or alkyd-acrylic

hybrids. Oxidation is whether the seed oil or fatty acid combination used is considered

drying or non-drying. An oxidizing alkyd would be made with a drying oil, such as linseed

oil, and would have the ability to cure autoxidatively.1–5,24,25

13 2.5.1 Alkyd Resin Synthesis Theory

For laboratory scale synthesis, alkyds are reacted in a three- or four-neck round bottom flask under inert atmosphere with mechanical stirring, heating, a Dean-Stark trap and a condenser as shown in Figure 2.6.

Figure 2.6. Reaction set up for alkyd synthesis, including nitrogen flow, mechanical stirring, a Dean-Stark trap and condenser.

Due to the versatility of the synthesis techniques for alkyds, a variety of starting materials can be used. The polyol and dibasic acid are most commonly varied, and in many industrial setting, multiple starting materials are used in a single alkyd. For this study, linseed oil fatty acids were used for their high drying index and relatively low cost.

Glycerol was used as the polyol, as it is the most commonly used in industrial settings.

Phthalic anhydride was chosen as the dibasic acid, and it is also most commonly used. The esterification of the anhydride is rapid due to ring strain. In the fatty acid process, all reactants are added at the start of the reaction, and are heated slowly to 220 – 260 °C for the esterification to begin. A Dean-Stark trap is added on the reaction set-up to collect the

14 water produced during the esterification reaction. Trapping the water prevents the reaction

from going to equilibrium, and drives polymerization forward.2–5,24,25

The amounts of starting materials needed can be calculated using a series of equations to predict the molecular weight, oil length and gel point of the alkyd. Firstly, alkyds are generally formulated to have an excess of hydroxyl groups, and the ratio of hydroxyl equivalents, eb, to acid equivalents, ea, is first calculated as R to be greater than

1, as shown in Equation (2).24 � � = � (2)

The alkyd constant, K, is another important parameter for calculating starting

materials in an alkyd formulation. It is a predictor of gelation during the alkyd cook, and

should be close to 1 to avoid gelling. In practical formulating, K is usually slightly above

1 to ensure stability during storage. The alkyd constant is the ratio of total molar equivalents

24 of all raw materials, m0, to total acid equivalents, ea, as shown in Equation (3). � � = (3) �

The reaction progress can be monitored in several ways, including viscosity, water

evolution and acid value. In industrial settings, reactions are generally monitored by

viscosity because it is the quickest estimation of progress. Gardner-Holdt bubble viscosity

is commonly used, as it allows for comparison of the reaction product to be compared with

a mineral oil samples of standard viscosity.46–48 However, the acid value gives a more

accurate determination of what is going on, by indicating the molecular weight of the alkyd.

Acid value is determined by a standard test method (ASTM D1639-90) and is defined as

the mg KOH / g resin.49 The approximation of molecular weight was derived by Patton,24

and shows that the average molecular weight, Mav, is the starting weight of the materials, 15 w0, divided by the difference of the initial acid equivalents, ea, from the total equivalent

moles, m0, plus the total residual acid, as calculated by the product of the weight of the

starting materials and the acid value, AV, divided by the molecular weight of potassium

hydroxide, as shown in equation (4). � � = � ∗ �� (4) (� − � ) + ( ) 56100

The water evolved, WE, during the reaction can be calculated theoretically by using the molar equivalents of acid added to the reaction and multiplying them by the molecular weight of water, as shown in Equation (5),

� � �� = + ∗ 18.01 (5) � �

where ea1 is the acid equivalents of the fatty acids, and F1 is the functionality of the fatty

acids, and ea2 is the acid equivalents of phthalic anhydride, and F2 is the functionality of

the phthalic anhydride. Using this theoretical amount, the yield of the alkyd reaction, Y,

can also be calculated by Equation (6),

� = � − �� (6)

where w0 is the weight of all starting materials, and WE is the water evolved.

Lastly, the theoretical hydroxyl value can be calculated to determine approximately

how many free hydroxyl groups will be present on the final alkyd. This equation takes into

account acid value, AV, yield, Y, equivalents of acid and base, ea and eb, and the molecular

weight of potassium hydroxide in terms of mg/mol as shown in Equation (7).

(� + � ) ∗ 56100 �� = �� + (7) �

16 Based on these calculations, the starting materials for long linseed oil and medium

linseed oil alkyds were chosen to maximize molecular weight, while preventing gelling.24

2.6 High Solids Formulations

Solventborne coatings contain a large amount of volatile organic compounds

(VOCs), which have begun to be phased out due to environmental, health and safety

concerns. The industry has been largely driven to using waterborne systems.50 However, for some coatings, emulsions or dispersions are difficult to stabilize and do not possess the same properties as their solventborne counterparts. High solids coatings are formulated to possess as little VOC as possible, while still retaining the excellent properties afforded by solventborne coatings. The strategy for creating these coatings is to decrease the molecular weight of the binder, therefore lowering the viscosity and reducing the amount of solvent needed to reach application viscosity. As the small amount of solvent evaporates, the lower molecular weight polymer can crosslink to form a film. However, this presents several challenges. When the binder has a lower starting molecular weight, it is difficult for it to build up the appropriate molecular weight through crosslinking, which could result in poor film formation. Because there is little solvent, the physical drying process is almost non- existent.1,3,5,50 Hardness and glass transition are dependent on the build-up of molecular

weight after the film is tack-free, whereas in conventional solventborne coating, these

properties are present quickly upon drying due to the high molecular weight of the binder.

Additionally, the lower viscosity of the binder in high solids coatings allows for more

defects to form than in conventional higher molecular weight coating systems.51–54 High solids coatings are usually formulated by combining multiple techniques to achieve the

17 best properties. Lowering molecular weight of the resin, addition of reactive diluents and

use of HAP exempt solvents are the most common methods.1,3,5

Reactive diluents are commonly used with alkyd resins because they can be easily

synthesized from seed oils, which are very compatible with the alkyd resin. The reactive

diluents help to lower the viscosity, but instead of volatilizing upon drying, they crosslink

into the forming film.1,3,5,55 These will be discussed in more detail in a later section. HAP exempt solvents are commonly used to introduce volatiles without labeling them as VOCs.

There are several exempt solvents, including acetone and methyl amyl ketone (MAK).56

Aliphatic hydrocarbons, ketones, and alcohols are all HAP exempt and can lower the

viscosity without decreasing the solids content of the coating.3

Alkyd resins are well-suited for formulation as high solids coatings. They are well-

positioned as the market is moving toward more plant-based and renewable products with

lower VOCs. The molecular weight can be decreased while the molecular weight

distribution is increased, creating a coating with lower viscosity. The oil length can be

increased, while can come at the cost of increased drying time. However, this can be

combatted by using a slightly higher cost seed oil, such as tung or linseed oil, which have

higher crosslinking functionality. Additionally, alkyds can be baked for even faster curing

times.3,5

2.7 Reactive Diluents

To achieve high solids coatings, reactive diluents are an attractive choice due to their ability to react and form part of the film upon curing, while also decreasing the viscosity of the coating. Typical viscosities for various coatings applications is shown in

18

Table 2.3. Linseed oil is one of the first used reactive diluents because it can react easily into the system via autoxidation reactions.5

Table 2.3. Typical viscosity for different coating application techniques. Application Viscosity Range Type (Pa-s) General3 0.05 – 1 Spray3 0.05 – 0.15 Brush57 0.1 – 0.25 Roller58 5 – 50 Inks3 5 – 10

Choosing an effective reactive diluent for an alkyd system requires it to meet certain physical and chemical specifications. One of the most important criteria is the presence of reactive sites that can crosslink into the resin system, unlike a plasticizer. It also must react rapidly so it does not retard the curing of the film, and it should also not affect coatings properties, such as hardness, gloss, flexibility or adhesion. The reactive diluent should also have low volatility so it does not evaporate as a solvent would, and it must not release any toxic or harmful degradation products. Additional considerations include no color, economic in cost when compared to solvents and good storage stability.59

For alkyd coatings, several reactive diluents have been reported. These include sucrose octasoyate,60 octadienyl fumarate/succinate derivatives,59 hexadienol and hexanedienal derivatives,61 octadienyl ethers,62 and 2-butene-1,4-diol derivatives.55 These groups have many things in common, including easily abstracted hydrogens, vinyl groups, low molecular weight, liquid at room temperature and a lack of hydroxyls, amines, carbamates and thiols, functional groups that increase viscosity through their strong interactions.

19 In addition to these reactive diluents, there are several based on different seed oils

and modification of these oils. There is a variety of chemistries available to modify seed

oils, including epoxidation, Diels-Alder reactions, and addition of acrylates and methacrylates.7,63 Methacrylated fatty acids have been used as a replacement for styrene in

vinyl ester resins, as shown by La Scala et al.9 Biermann et al.12 showed that various esters

of calendula and tung oils show excellent properties as reactive diluents.

Wutticharoenwong et al.10,11 have shown extensive research into using Diels-Alder

chemistry to modify tung oil with various functionalities, including fluorine and

alkoxysilane. Additionally, Nalawade et al.13,14 used conjugated soybean oil to create various reactive diluents for high solids alkyd coatings. Cargill® introduced Dilulin® as a reactive diluent based on a Diels-Alder modification of linseed oil to create norbornene groups on the fatty acid chains.6

2.8 Ene Reactions

The ene reaction is often overlooked in organic chemistry, especially when compared to the similar Diels-Alder reaction. The ene reaction occurs between an alkene with an allylic hydrogen (ene) and any double or triple bond (enophile). The pericyclic reaction creates a new σ-bond, the ene double bond migrates and there is a 1,5-hydrogen shift, as shown in Figure 2.7.64–66

Enophile

H H Ene

Figure 2.7. The ene reaction.

20 The ene reaction can take place with the unsaturation in the fatty acids in seed oils.

A room temperature ene reactions with soybean oil and diethyl azodicarboxylate (DEAD) has been shown by Biswas et al.67 to create a self-curing coating system. The reaction was monitored by 13C-NMR and viscosity measurements to show that the system continues to thicken for up to 14 days. Eren and Küsefoğlu68 used an ene reaction with plant oil and

paraformaldehyede to create maleated plant oil products that can be polymerized for use

as composite resins. The mechanical properties were found to be comparable to high-

performance unsaturated polyester resins, making these a bio-based alternative to those

petroleum-based materials.69

Maleic anhydride (MA) has been grafted onto polymers using the ene reaction, where the maleic anhydride acts as the enophile. The grafting of this group is of considerable interest due to its versatile reactivity and compatibility.66,70–72 Sclavons et al.71

showed the ene-grafting of MA on polypropylene with enriched double bonds. The grafting

mechanism of MA to polybutene-1 was researched by Zhao et al.,72 using comonomers to

make the interactions more efficient. Benn et al.70 studied the kinetics of the ene reaction

between MA and various alkenes, looking at the effect of reaction temperature, solvents,

and steric hindrance. Grafting of MA creates a versatile anhydride group that can be used

to modify the polymer with various other functionalities.

2.9 Inorganic/Organic Hybrid Coatings

Sol-gel chemistry was first used for inorganic materials to make glass or ceramics.

A sol is a dispersion of colloidal particles within a liquid, and a gel is a 3D network. The

sol-gel process involves curing these suspended particles to form a continuous gel.73,74

Ebelman75 first demonstrated this process with tetraethyl orthosilicate (TEOS) in acidic 21 conditions to form a glassy material. The precursors in these processes are generally metal alkoxides, such as alkoxysilanes. These form an interconnected 3D network through simultaneous hydrolysis and condensation reactions, as shown in Figure 2.8. The final properties of these networks are heavily dependent on pH, catalyst, concentration and temperature of the reaction.73,74,76

The sol-gel process can take place in both acidic and basic conditions.77,78 The hydrolysis reaction dominates in acidic conditions and proceeds in two steps in an SN2

mechanism. The acid catalyst attacks and leads to a positive charge on the alkoxysilane,

causing electron density to be drawn away from silicon, making it more susceptible to

attack from water. The rate-determining step involves the attack of the oxygen in water on

the silicon atom. This leads to a pentacoordinate transition state, which has a partially

positive charge. The rate of this hydrolysis reaction is significantly impacted by steric

effects. The silanol group that forms can then undergo condensation reactions with an

alkoxide or another silanol, forming Si-O-Si linkages. The siloxane particles will then

condense into a gel. When the reaction conditions are basic, the hydroxyl anion will attack

the silicon atom directly by an SN2 mechanism. The alkoxide group repels the hydroxyl anion, which significantly slows the reactions. Under basic conditions, the hydrolysis reaction is much slower.74,77

22 Hydrolysis

OR OH

OR Si OR OR Si OR + H2O + ROH

OR OR [1] [2]

OH OH

OR Si OR + H2O OR Si OH + ROH

OR OR [2] [3] Condensation OH OH OH

2 OR Si OH OR Si O Si OH + H2O

OR OR OR [3] [4]

OH OR OH OR

OR Si OH + RO Si OR OR Si O Si OR + ROH

OR OR OR OR [3] [1] [5]

Gelation O O OR

[1] + [2] + [3] + [4] + [5] + ... OR Si O Si O Si O

O O O

O Si O Si O Si OR

O O O

OR Si O Si O Si OR

OR O RO

Figure 2.8. Sol-gel process for alkoxysilanes.

The use of these groups in coatings was shown by Ni et al.,82 who developed a moisture-curing system with alkoxysilane-functionalized isocyanurate coatings. The

23 incorporation of this chemistry led to advantages in many properties, including adhesion,

crosslink density, glass transition temperature, and flexibility. Soucek and coworkers32,82–

89 have reported several novel inorganic/organic hybrid coating systems based on drying

oils and alkyds. The metal alkoxide forms the inorganic network through hydrolysis and

condensation reactions, while the organic network is formed through the autoxidation

reaction of the fatty acids in the drying oils or alkyds. In epoxidized soybean oil-based

inorganic/organic hybrid coatings, the hardness, tensile strength and flexibility were

increased, while the adhesion and impact resistance decreased with increased sol-gel

precursor content.85 Various precursors were used in this study, based on titanium and zirconium, and the use of mixed precursors enhanced the mechanical properties of the coatings due to a synergistic effect.

When alkoxysilane-modified resins are exposed to humidity or atmospheric moisture, they begin to hydrolyze and form sol-gel networks. This can be used as a primary or secondary curing mechanism in coatings systems with applications in high humidity environments. Silane end-capped soybean oil polyurethanes were synthesized by

Baghdachi et al.,90 and their curing was studied under differing relative humidity and

temperature. It was found that increased relative humidity promoted faster crosslinking,

and thus tack-free time of the coating, while lower relative humidity severely retarded these

crosslinking reactions. Chang et al.91 patented an acrylic-silane coating with curing time accelerated by increased humidity. Emmerling et al.92 show an alkoxysilane-terminated polyurethane with moisture-curing capabilities. The effect of an acid-catalyst on the curing of polyurea/polysiloxane organic/inorganic hybrid coatings at high humidity was shown

24 by Ni et al.82,93 To our knowledge, no literature on the effect of curing at varying humidity

on the final coating properties has been shown with alkoxysilane-modified alkyds.

2.10 Fluorinated Polymers

Fluoropolymers have gained attention as a coating material due to the excellent

barrier properties afford by these coatings. They provide resistance to organic solvents,

chemicals, and water due to their low surface energy, and they are also resistant to high

temperatures. The high performance properties afforded by these coatings make them an

attractive choice for a variety of applications. Using fluorine-modified materials as coating

additives can provide the same performance without the drawbacks of difficult

processibility that is common with fluoropolymers.94,95

The unique properties of fluoropolymers are based on the nature of the carbon- fluorine bond, which has a bond energy of ~480 kJ mol-1, compared to the carbon-carbon

bond energy of ~290-330 kJ mol-1.96 Alkyl fluorides are 102-106 times less reactive than

alkyl chlorides in solvolysis and displacement reactions due to the poor leaving group

ability of fluorine.97 Because fluorine has a high ionization potential and low polarizability,

fluoropolymers have low intermolecular interactions, the fluorinated groups generally

migrate to the air interface to lower the surface energy.98

The use of fluorinated coatings is relegated to certain applications due to high cost and low processibility. Teflon™ developed by DuPont in the 1930s is an example of a linear fluoropolymer, poly(tetrafluoroethylene) (PTFE). As a coating, PTFE exhibits the desirable properties mentioned earlier, including chemical resistance, but it has very limited solubility.99 To apply PTFE as a coating, it must be dispersed as polymer particles

25 and sintered at temperatures greater than 350 °C to achieve a uniform film. This severely

limits the application potential.100

There are many other ways to incorporate fluorine functionality into coatings systems. Vinylidene fluoride (VDF) can be copolymerized with a hydroxy-functional monomer, such as 2-hydroxyethyl methacrylate (HEMA), and then crosslinked with a poly(isocyanate) to form a coating with excellent adhesion and corrosion resistance properties, when compared with poly(vinylidene fluoride) (PVDF) homopolymer.101

Wutticharoenwong et al.10,11 synthesized a fluorine-functional reactive diluent based on tung oil that showed properties improved by the fluorine functionality. Additionally, fluorine-functionalized alkyds synthesized by Thanamongkollit and Soucek102 showed

improved surface properties.

2.11 Self-Stratifying Coatings

Self-stratifying coatings are semi-compatible polymer blends with the ability to

form heterogeneous micro-structures upon curing. The idea of these coatings was first

introduced by Funke in 1973,103 and it has found applications in a variety of fields,

including automotive,104 antimicrobial,105,106 aerospace and anti-fouling.107 Previous

studies have shown stratifying coatings with epoxy-acrylic copolymers,104,108 alkyd-acrylic

copolymers,109 polyurethane,106,107,110,111 and latexes.109

The forces that drive self-stratification include solubility, surface tension, interfacial tension, evaporation rates and curing kinetics.112–114 Ultimately, thermodynamic incompatibility between the resins involved can lead to this stratification phenomena. To better understand this, the Flory-Huggins theory regarding the free energy of mixing of polymeric materials can be employed to better understand these phenomena. This can be 26 applied to polymer-polymer or polymer-solvent systems, as well as to three component systems, such as polymer-polymer-solvent.115 Enthalpy is the driving force of the segregation if the molecular weight of each polymer or block is roughly equal.116 There must be a gradient throughout the coating cross-section to drive the separation taking place.117,118 Another component that must be considered is the migration phenomena, meaning that as the oligomeric system turns polymeric, the movement is limited.118

Additionally, as solvent evaporates, the incompatibility in the polymer systems may become more pronounced and lead to more separation.119

In thermosetting polymers, the stratification is more complicated by the drying versus curing relationship. In curing, or crosslinking, the phase behavior and separation dynamics are the dominant factors. In drying, the solvent evaporates from the system by diffusing from the substrate-coating interface to the coating-air interface. As the solvent diffuses, hardening of the coating is also occurring, which restricts the movement of the components trying to segregate. The segregation during drying is time-dependent and relies on the speed of solvent evaporation for the systems to orient properly. The separation of the coatings is dependent on the competition between the diffusion of the stratifying groups and the curing of the coating.119

Exploration of several commercially available unpigmented resins for their ability to self-stratify was carried out by Benjamin et al.120 They found that the stratification was dependent on the solvent used and the phase separation of the resins. Of the many systems tested, the epoxide/acrylate and epoxide/fluoroethylene/vinyl ether copolymers showed the highest degree of stratification. The epoxide was cured with an amine curative that seemed to help drive the phase separation. A self-stratifying polyurethane was created by

27 Baghdachi et al.110 that separated into two distinct layers upon curing. This system utilized a silsesquioxane with epoxide functionality, a hydroxyl functional fluorinated fluoroethylene alkyl vinyl ether, and a polyurethane. The fluorinated polymer separated to the air interface, and silsesquioxane stayed at the substrate interface.

2.12 Corrosion

Corrosion is an electrochemical and chemical process caused by exposure to water and electrolytes, such as salt. Steel and aluminum are commonly used metals in applications ranging from industrial to commercial, and they are susceptible to corrosion.

This process degrades the metal into brittle pieces, leading to problems with the integrity of corroded structures. Once a metal part has been corroded, it must be replaced or more damage will be caused.3,121,122

The cost of corrosion is astronomical and is of worldwide concern. Corrosion costs

stem from the expense of replacement, the waste of resources used on the corroded product

and the inconvenience it causes. The economic cost of corrosion for each country is

estimated to be between 1 – 3.5 % of its gross national product. For the United States alone,

the cost is estimated to be close to $300 billion per year, if one includes the indirect costs

as well. There is great need for a process which can dramatically slow corrosion at a

reasonable cost to manufacturers. To this point, many processes have been adopted and

proposed, but is necessary to first understand the electrochemical and chemical reactions

that cause corrosion.122

Steel is an alloy of iron, carbon and many other metals with an exact composition

that varies depending on its end usage. Steel is one of the most commonly used metals due

to its low cost and high tensile strength. The base metal of steel is iron, which can form 28

two different crystal structures, depending on its temperature. At very high temperatures, iron forms the face-centered cubic structure, and at lower temperatures, it forms the body- centered cubic structure. Carbon and the other elements added to the steel act as hardeners by preventing dislocations of the crystalline iron structure.123

Electrochemical reactions take place when two pieces of different metals are connected with a conductive wire and submerged into an electrolyte, usually a water solution containing dissolved salts. The reaction can then take place spontaneously, causing oxidation of metal atoms at the cathode and dissolution of iron at the anode.3,121,122

Because of the nonuniform surface of steel, certain areas are anodic while others are cathodic. The composition and processing conditions of steel determine the stresses and morphology that contribute to its anode-cathode pairs. High internal stress leads to a higher susceptibility to corrosion. Cold-rolled steel, which is most commonly used for its high strength, has high internal stresses and thus corrosion susceptibility.121

On an untreated and uncoated piece of steel, water containing trace amounts of salt is adsorbed thinly to the surface. This creates an electrolyte for electrochemical reactions to take place. These reactions differ in the presence and absence of oxygen. When oxygen is absent, iron(II) ions are released as the anodic products of corrosion and hydrogen is formed at the cathode as seen in Equations (8) and (9). The hydrogen stops dissolution of iron by cathodic polarization, unless the electrolyte is acidic.3,121,124

�����: �� → �� + 2� (8)

���ℎ���: 2� + 2� → � (9)

When oxygen is present, hydroxyl anions are formed at the cathode and dissolution of iron continues, as shown in Equations (10) and (11).121,125

29 � + �� + 4� → 4�� (10)

2�� + � + 2�� → 2�� + 4�� (11)

The amount of oxygen at the steel surface also plays an important role in the rate

of corrosion. At low oxygen concentrations, the corrosion rate will increase with increase

in the concentration of dissolved oxygen. However, at a critical concentration, the

corrosion rate will begin to decline with increase in dissolved oxygen due to a passivation

effect.121

In addition to these factors, corrosion is also affected by the concentration of the

dissolved salts in the water. A higher concentration of salts leads to a greater conductivity

and a faster corrosion rate. Metals in acidic conditions will degrade much faster due to the

solubility of iron in strong acid, even without electrochemical reactions. Lastly, high

temperatures will also speed up corrosion by increasing the rate of chemical reactions.3,121

2.12.1 Electrochemical Impedance Spectroscopy (EIS)

Electrochemical impedance spectroscopy (EIS) is a tool used to look at the barrier

properties of a coating. It is designed to measure corrosion kinetics and mechanisms. This process has the capability to give a more quantitative analysis of corrosion than what more qualitative tests like salt spray can.126 The EIS test creates a capacitor, where the coating is

a non-conductive medium between two conducting plates. The conductors, in this case, are

the salt water solution held on the coating surface and the aluminum panel. The impedance

is measured as a modulus and is usually presented in a Bode plot. This plot displays the

logarithm of the impedance modulus versus the frequency. The impedance modulus is a

measure of the non-conducting properties of the coating. As the coating begins to fail, it

30 will no longer form a non-conductive barrier between the salt water and the aluminum

panel, making the impedance modulus decrease.126,127

2.13 Exterior Durability

Coatings used in outdoor applications must have exterior durability to avoid property changes when exposed to sunlight, rain and other weather events. Common changes that occur with weathering include discoloration, chalking, embrittlement and loss of adhesion. Outdoor exposure often leads to photoinitiated or hydrolytic degradation, depending on the functional groups that are present. The changing temperature and humidity can also cause cracking when the coatings expand and shrink. The rates and

occurrence of these processes are highly dependent on the site of exposure and the time of

year.3,128,129

2.13.1 Photodegradation

Photoinitiated oxidative degradation is the main cause of weathering with outdoor

coatings. Even without functional groups that directly absorb UV radiation, there are still

peroxides and ketones present in the coatings that will absorb it. These compounds are

present in virtually all resins used for coatings, as well as in organic substances. Peroxides

absorb UV light, forming more hydroperoxides and peroxides. These will further dissociate

with exposure to sunlight and yield more free radicals, making this an autocatalytic process.

To minimize the effects of photodegradation, coatings exposed to sunlight should avoid or

reduce the amount of abstractable hydrogens that are present in the resin. Some functional

groups are more susceptible to photodegradation. From most unstable to most stable, the

list is as follows: amines, allyl ethers > esters, alcohols, urethanes, allyls and benzyls >

31 esters, tertiary alkyds > secondary alkyls > primary alkyls >> methylsiloxanes. The order is determined by the energy of bond dissociation, activating effects of the heteroatoms, neighboring groups and steric hindrance. Silicon and fluorine-functionalized resin provide significant improvement to photo-stability, which is due to the resistance of silicone to photo-oxidation and the reduction of C-H functionality in fluorine-containing resins.130

Aromatic groups, on the other hand, absorb UV radiation, and lead to the continued formation of free radicals.3,128,129,1313,128,129,131

Several additives have been developed to mitigate the effects of photoinitiated oxidative degradation in coatings. These materials help to absorb UV light, scavenge radicals, reduce the amount of hydroperoxides present in the resin and complex transition metals to avoid catalysis of unwanted reactions. Usually, these additives are used in combination and work synergistically to provide the needed exterior durability to the coating formulation. The main additives used include antioxidants, metal-complexing agents and hindered-amine light stabilizers (HALS). Antioxidants are used to reduce the amount of hydroperoxides present by reducing them to alcohols, which are subsequently oxidized into harmless products. The metal-complexing agents are used to remove transition metal ions before they can catalyze the peroxide decomposition reactions. In resins with oxidative curing mechanisms, such as alkyds, these transition metals are added to catalyze curing. By that same mechanism, the photo-degradation reactions can take place, so it is important to use as little as possible in outdoor coatings. Lastly, HALS act both as chain-breaking antioxidants and metal complexing agents. In combination, these additives significantly improve the exterior durability of coatings.3,132

32

2.13.2 Hydrolysis

Outdoor coatings are also susceptible to hydrolytic degradation. This is especially dependent on the functionality of the coating system and the design of the polymers being used. The functional groups most vulnerable to hydrolysis are as follows from most vulnerable to least: esters > carbonates > ureas > urethanes >> ethers. UV radiation and acidic conditions can accelerate hydrolytic degradation. In outdoor coatings, exposure to sunlight and UV radiation is virtually unavoidable. Acid rain can also cause etching and hydrolysis of resin, leading to further degradation.3,129 Steric hindrance and hydrophobicity in the polymer structure can provide more hydrolytic stability.133–138 Silicon-containing coatings, which are resistant to photodegradation, are susceptible to hydrolysis at their crosslinking sites. The electronegativity of oxygen facilitates the nucleophilic attack by water on silicon. However, this reaction is reversible so the crosslinks can reform.

Extended exposure to high humidity or water can cause the coating to soften.139,140

2.13.3 Other Failure Modes

The main failure of outdoor coatings comes from degradation of the polymer by photoinitiated oxidation or hydrolysis reactions. However, there are several other problems that can arise from outdoor exposure. Microcracking occurs from rapid temperature changes, especially for automotive coatings.141 On wood, coatings must be able to undergo uneven extensions when wood contracts and expands. Using oil-based coatings on wood can also cause significant blistering. Wood has absorbed water, and when it is heated by the sunlight, the vapor pressure of the water is increased and cannot escape due to the low water permeability of the oil-based coating, causing the coating to blister.142 In latex house paints, they can become stickier in high humidity conditions and retain dirt that cannot be

33 easily removed. Lastly, growth of fungi causing mold and mildew in coatings can be a problem as well.3,128,129

2.13.4 Weather Testing

Weathering presents expansive problems for coatings that are not easily solved or tested. There are no weathering tests that provide reliable results, and often customer complaints and field testing are the only way to truly understand the weatherability of a coating. Accelerated outdoor tests are commonly done in certain regions. This involved fencing off an area where coated panels are exposed to sunlight at certain angles. For architectural and industrial coatings, they are most commonly placed in Florida, facing south at a 45º angle. To further accelerate the testing, black box exposure is sometimes used, where the coating is mounted on a black box to increase the temperature significantly during sunlight exposure. Photocells are used to gauge the amount of UV radiation the coating is exposed to over this timeframe, and the coatings are periodically tested. The coatings are evaluated for ease of cleaning, changes in gloss or color, chalking and failure.

This testing takes place over 2-5 years, depending on the final application of the coating.3,128,129,141,143

Because outdoor testing is so time consuming, there are accelerated weathering devices for use in labs. These instruments expose the coatings to temperature, humidity and UV light in alternating intervals as programmed by the user. In some devices, acid rain conditions can also be accounted for by using dilute acid solutions in the water spray.

However, there is no correlation between the accelerated testing device and actual outdoor exposure. It cannot take into account the variability of outdoor environments and account for loss of stabilizers over long periods of time. However, it is useful as a first

34 approximation of how a coating might react to outdoor exposure, and it is much quicker than a 2-5-year test.3,129,141,143–145

35 CHAPTER III

EFFECTS OF MODIFIED ALKYDS AND REACTIVE DILUENTS ON THE

PERFORMANCE AND STRATIFICATION OF ALKYD COATINGS

3.1 Abstract

Performance and stratification of two coating systems were evaluated with previously reported alkoxysilane and fluorine modified alkyds and reactive diluents. The combination of alkoxysilane and fluorine functionalities was chosen to improve adhesion, hardness, and chemical and corrosion resistance of the coating system. A comparison between modified alkyds and reactive diluents was carried out in terms of the effects of the molecular weight and viscosity of these components on coating properties and stratification. The coatings were formulated with an unmodified alkyd as the base resin, and the modified alkyds and reactive diluents were used as additives in increasing amounts.

The systems were evaluated for coating and tensile properties, corrosion resistance, weatherability and stratification. The modified alkyd coating system showed improved performance in tensile properties and corrosion resistance, while there was little stratification in this system. The modified reactive diluents showed improved properties and corrosion resistance at high loadings, where it was found that stratification was occurring.

36 3.2 Introduction

Alkyd resins are derived from renewable seed oils and are able to undergo facile

modifications. Alkyd resins were once the most widely used coatings,1,2 and they were first introduced by Kienle146 in the 1920s. Alkyds are modified polyesters, synthesized from

polyhydric alcohol, polybasic acid, and monobasic fatty acids from the triglycerides in seed

oils.1,3 Alkyds are often formulated with reactive diluents, which are low viscosity

additives with the ability to crosslink into the alkyd system during film formation. Many

iterations of reactive diluents based on seed oils have been shown previously and can be

modified to add a desired functionality into the coating system.10–14

The modification of alkyds with various functionalities has been shown previously.

The use of a fluorinated polyol, 2,2,3,3,4,4,5,5-octafluoro-1,6-hexanediol as a partial replacement for glycerol during alkyd synthesis was shown by Salata et al.147 to create

corrosion resistant, partially stratifying alkyd coatings. Fluorine functionalized reactive

diluents were also shown by Wutticharoenwong et al.10,11 and Nalawade et al.13,14 These

additives were shown to create alkyd coatings with improved corrosion resistance,

chemical resistance and surface properties. Fluoropolymers are often difficult to process

and have high cost associated with them, which severely limits their application

potential.94,95 However, using a small amount of fluorinated material as an additive allows

for the improvement of properties without the obstacles of processibility and prohibitive

cost. The low surface energy of fluorinated materials drives them to the air/coating interface.98,110,117,119,148

Salata et al.149 showed the use of 3-(triethoxysilyl)-propyl isocyanate to create an alkoxysilane functional alkyd coating with increased adhesion and mechanical properties.

37 Alkoxysilane functionalities in reactive diluents were also shown previously by

Wutticharoenwong et al.10,11 and Nalawade et al.13,14 to create coatings with increased mechanical properties and hardness. Ni et al.82 developed a moisture-curing system with alkoxysilane-functionalized isocyanurate coatings. The incorporation of this chemistry led to advantages in many properties, including adhesion, crosslink density, glass transition temperature, and flexibility. Soucek and coworkers32,82–89 have reported several novel

inorganic/organic hybrid coating systems based on drying oils and alkyds. The metal

alkoxides used form an inorganic network through hydrolysis and condensation reactions,

while the organic network is formed through the autoxidation reaction of the fatty acids in

the drying oils or alkyds. The alkoxysilane groups have often been found to increase

adhesion in direct-to-metal coatings, possibly due to the condensation and hydrolysis

reactions with free hydroxyl groups at the substrate surface.82,150–154

Self-stratifying coatings were first introduced by Funke in 1973,103 and they have found applications in a variety of fields, including automotive,104 antimicrobial,105,106

aerospace and anti-fouling.107 A multi-layer or gradient structure can be created from a single coat system, where it forms an undercoat and a top-coat without the cost associated with applying two coats to the same substrate. Coatings can be formulated to achieve stratification by using two incompatible resins in a common solvent, where the separation begins taking place during evaporation of the solvent. Previous studies have shown stratifying coatings with epoxy-acrylic copolymers,104,108 alkyd-acrylic copolymers,109

polyurethane,106,107,110,111 and latexes.109 However, this method can introduce many

obstacles for commercial use, including the use of solvents increasing volatile organic

38 compounds (VOCs) in the formulation, long-term stability of the two incompatible polymers, and adhesion between the layers.

In this study, performance and stratification of two coating systems were evaluated with previously reported alkoxysilane and fluorine modified alkyds and reactive diluents.

The alkoxysilane functionality was chosen to promote adhesion and mechanical properties of the coating, while the fluorine functionality should improve hardness, chemical and corrosion resistance. Studying the differences between modified alkyds and reactive diluents in terms of coating properties and stratification will help to elucidate the effects of molecular weight and viscosity. Formulations were made with equal amounts of alkoxysilane and fluorine functional components. They were used as additives in an unmodified alkyd in increasing amounts, up to 10 wt.% each. The coating performance, tensile properties, corrosion resistance and weatherability were evaluated as a function of concentration of the modified alkyds or reactive diluents. The stratification was observed using environmental scanning electron microscopy with energy dispersive X-ray spectroscopy (ESEM-EDX) to look at the concentration of fluorine and silicon over the cross-section of these coatings.

3.3 Experimental

3.3.1 Materials

Linseed oil fatty acids were provided by Alnor Oil Company. Glycerol, phthalic anhydride, xylene, lithium hydroxide, potassium hydroxide, methanol, acetone, acetic anhydride, pyridine, 3-(triethoxysilyl)-propyl isocyanate, dibutyltin dilaurate,

2,2,3,3,4,4,5,5-octafluoro-1,6-hexanediol, tung oil, phenothiazine, and 2,2,2-trifluoroethyl

39 methacrylate were all obtained from Millipore Sigma. Methacryloxypropyl

trimethoxysilane was obtained from Gelest. BYK 333 was obtained from BYK. Borschi®

OXY-coat, 12% zirconium Hex-Cem®, and 5% Calcium Hex-Cem® driers were obtained

from OM Group. AQ-36 aluminum panels were purchased from the Q-Lab Corp. All

materials were used as received.

3.3.2 Instrumentation

All reactions were performed under a nitrogen atmosphere unless otherwise noted.

1H-NMR (300 MHz) spectra (�, ppm) were obtained using a Varian Mercury 300 spectrometer in CDCl3 solvent and analyzed with ACD/NMR Processor. FT-IR spectra

were recorded using a Nicolet iS50 spectrometer using an attenuated total reflection (ATR)

diamond attachment and analyzed using OMNIC software. MALDI-ToF-MS spectra were

recorded on a Bruker Ultra-Flex III MALDI-ToF/ToF mass spectrometer (Bruker,

Billerica, MA) equipped with a Nd:YAG laser emitting at 355 nm. The instrument was

operated in positive ion mode. Samples were dissolved in THF to a final concentration of

10 mg mL-1. Trans-2-[3-(4-tert-butylphenyl)-2-methyl-2-propenylidene] malononitrile

(DCTB) (20 mg mL-1) served as matrix and sodium trifluoroacetate (NaTFA) (10 mg mL-

1) as cationizing agent. The latter two were prepared and mixed in the ratio 10:1 (v/v),

respectively. Matrix:salt and sample solutions were applied onto the MALDI-ToF-MS

target plate using sandwich method. Bruker's FlexAnalysis software was used for analysis.

Viscosity measurements were recorded with a Brookfield LV-II+ Pro Viscometer using

Rheocalc V2.6 software for data analysis. Scanning electron microscopy was done using a

FEI Quanta 200 environmental scanning electron microscopy equipped with EDAX

energy-dispersive x-ray spectroscopy. Glass transition temperature was evaluated using

40 TA Instruments DSC Q2000. Tensile testing was performed on an Instron 5567 tensiometer

with a 100 N load cell. Contact angle was measured on ramé-hart Model 500 Advanced

Goniometer. Electrochemical impedance spectroscopy was performed using a Gamry

Reference 600+. Weathering data was obtained using an Atlas Ci 4000 Weather-Ometer, equipped with a xenon arc lamp. Gloss was measured using a BYK Micro-Tri Gloss Meter.

Color was characterized by a Konica Minolta CM-5 spectrophotometer.

3.3.3 Alkyd Resin Synthesis (LLOA)

Long oil alkyd was synthesized according to the fatty acid method.24 Linseed oil

fatty acids (140.2 g, 0.5 mol), glycerol (50.0 g, 0.5 mol), phthalic anhydride (74.1 g, 0.5

mol), and xylene (10 mL) were added to a 500-mL 4-neck round-bottomed flask. The flask

was equipped with nitrogen flow, mechanical stirring, a condenser, and a Dean-Stark trap.

The reaction was heated over 1.5 h to 250 ºC, using a heating mantle with a temperature

controller. Water began to evolve into the Dean-Stark trap after 1 h. After the first 4 h of

reaction, the acid number (ASTM D 1639-90)49 was measured every hour until it was less

than 10 mg KOH/ g resin. The reaction was taken down following 5 h of reaction, and 14

mL of water had evolved. The acid number was measured to be 8 mg KOH/ g resin. The

1 1 product was analyzed by H-NMR and MALDI MS. H-NMR: 0.88-0.99 (-CH3), 1.25-

1.29 (-CH2-), 1.59 (-CH2-CH2-CO-O-), 2.04-2.09 (-CH2-CH=), 2.31 (-CH2-CO-O-), 2.80

(=CH-CH2-CH=), 3.68-3.77 (-CH2-OH), 4.14-4.47 (-CO-O-CH2-), 5.35 (-CO-O-CH- and

–CH=CH-), 7.53-7.71 (ArH).

3.3.4 Alkoxysilane Modified Alkyd Resin Synthesis (ASLOA)

This synthesis was carried out as previously reported by Salata et. al.149 Linseed oil fatty acids (145.3 g, 0.5 mol), glycerol (43.0 g, 0.5 mol), phthalic anhydride (63.0 g, 0.4 41

mol), and xylene (10 mL) were added to a 500-mL 4-neck round-bottomed flask. The flask was equipped with nitrogen flow, mechanical stirring, a condenser, and a Dean-Stark trap.

The reaction was heated over 2 h to 250 ºC, using a heating mantle with a temperature controller. Water began to evolve into the Dean-Stark trap after 1 h. After the first 4 h of reaction, the acid number (ASTM D 1639-90)49 was measured every hour until it was less than 10 mg KOH/ g resin. The reaction was taken down following 8 h of reaction, and 14 mL of water had evolved. The acid number was measured to be 9 mg KOH/ g resin.

Following the synthesis, the hydroxyl value test (ASTM D 1957-86)155 was performed, and it was found that the hydroxyl value is 51 mg KOH/ g resin. From this, the amount of 3-

(triethoxysilyl)-propyl isocyanate (TESPIC) needed for a 15 wt.% addition was calculated.

In a 250-mL three-neck round-bottomed flask, the alkyd (80.3 g), TESPIC (3.0 g, 12.1 mmol) and as a catalyst, dibutyltin dilaurate (0.5 mL) were added. The flask was equipped with nitrogen flow, mechanical stirring and a condenser. The flask was heated to 75 ºC, using an electric heating mantle with a temperature controller. The reaction was monitored by FT-IR for the disappearance of the N=C=O vibration at 2270 cm-1, and it was found that after the 1.5 h, the reaction was complete. The product was analyzed by FTIR, 1H-NMR,

1 and MALDI MS. H-NMR: 0.60-0.65 (-CH2-Si-), 0.85-0.99 (-CH3), 1.20-1.30 (-CH2- and

CH3-CH2-O-Si-), 1.60 (-CH2-CH2-CO-O-), 2.02-2.09 (-CH2-CH=), 2.29-2.37 (-CH2-CO-

O-), 2.78-2.80 (=CH-CH2-CH=), 3.16 (-CH2-NH-C(O)-O-), 3.78-3.85 (-CH2-OH and (-

CH2-O-Si), 4.14-4.54 (-CO-O-CH2-), 5.34-5.46 (-CO-O-CH- and –CH=CH-), 7.53-7.71

(ArH).

42 3.3.5 Fluorine Modified Alkyd Resin Synthesis (FLOA)

This synthesis was carried out as previously reported by Salata et. al.147 Linseed oil fatty acids (86.1 g, 0.3 mol), glycerol (26.3 g, 0.3 mol), phthalic anhydride (34.1 g, 0.2 mol) and xylene (8 mL) were added to a 500-mL four-neck round-bottomed flask. The flask was equipped with nitrogen flow, mechanical stirring, a Dean-Stark trap and a condenser. The reaction was heated to 250 ºC over 1.5 h, using an electric heating mantle with a temperature controller. Thirty minutes after heating was started, 2,2,3,3,4,4,5,5- octafluoro-1,6-hexanediol (4.02 g, 15.3 mmol) was added to the reaction mixture.

Following 4 h of reaction, the acid number was measured every hour until it was below 15 mg KOH/ g resin. The reaction was taken down after 6 h, when the acid number was measured to be 13 mg KOH/ g resin. There was 9 mL of water evolved. The product was

1 1 analyzed by H-NMR and MALDI MS. H-NMR: 0.85-0.99 (-CH3), 1.25-1.29 (-CH2-),

1.59 (-CH2-CH2-CO-O-), 2.01-2.03 (-CH2-CH=), 2.28-2.34 (-CH2-CO-O-), 2.78-2.80

(=CH-CH2-CH=), 3.71-3.73 (-CH2-OH), 4.15-4.57 (-CO-O-CH2- and -CF2-CH2-O-CO-),

5.34-5.35 (-CO-O-CH- and –CH=CH-), 7.53-7.71 (ArH).

3.3.6 Alkoxysilane Modified Tung Oil Synthesis (ASTO)

This synthesis was carried out as previously reported by Wutticharoenwong et al.10

Tung oil (87.7 g, 0.3 mol), methacryloxypropyl trimethoxysilane (29.9 g, 0.1 mol) and

phenothiazine (0.1 g, 0.6 mmol) were added to a 3-neck 250-mL round-bottomed flask.

The flask was equipped with nitrogen flow, mechanical stirring and a condenser. The flask

was heated to 180 ºC using an electric heating mantle with a temperature controller. The

reaction was heated for 3 h at 180 ºC, and the temperature was then turned to 200 ºC for

another 2 h. The reaction was monitored by 1H-NMR. The product was analyzed by FT-

43 1 1 IR, H-NMR, and MALDI MS. H-NMR: 0.62-0.72 (-CH2-Si-), 0.86-0.91 (-CH3), 1.06-

1.07 (CH3-C-), 1.25-1.38 (-CH2-), 1.60 (-CH2-CH2-CO-O-), 1.93 (-CH2-, cyclic), 2.05-

2.10 (-CH2-CH=CH-), 2.28-2.33 (-CH2-CO-O-), 2.76 (-CH=CH-CH2-CH=CH-), 3.56-

3.61 (CH3-O-Si-), 4.00-4.09 (-CH2-CO-O-), 4.10-4.31 (-CH-CH2-O-), 5.26-5.54 (-

CH=CH-, non-conjugated), 5.60-5.74 (-CH2-CH=CH-, conjugated), 5.94-6.19 (-CH-CH2-

O- and -CH=CH-, conjugated), 6.32-6.47 (-CH=CH-, conjugated).

3.3.7 Fluorine Modified Tung Oil Synthesis (FTO)

This synthesis was carried out as previously reported by Wutticharoenwong et al.10

Tung oil (60.2 g, 0.2 mol), 2,2,2-trifluoroethyl methacrylate (13.7 g, 8.1 mmol), and

phenothiazine (0.1 g, 0.5 mmol) were added to a 3-neck 250-mL round-bottomed flask.

The flask was equipped with nitrogen flow, mechanical stirring and a condenser. The flask was heated to 180 ºC using an electric heating mantle with a temperature controller. The reaction was heated for 4 h at 180 ºC and was monitored by 1H-NMR. The product was

1 1 analyzed by H-NMR and MALDI MS. H-NMR: 0.86-0.91 (-CH3), 1.11-1.12 (CH3-C-),

1.25-1.38 (-CH2-), 1.60 (-CH2-CH2-CO-O-), 1.97 (-CH2-, cyclic), 2.05-2.15 (-CH2-

CH=CH-), 2.28-2.33 (-CH2-CO-O-), 2.76 (-CH=CH-CH2-CH=CH-), 3.18 (-CH-CH=CH-

CH-C), 4.10-4.32 (-CH-CH2-O-), 4.48-4.57 (-CH2-CF2-), 5.24-5.42 (-CH=CH-, non- conjugated), 5.61-5.74 (-CH2-CH=CH-, conjugated), 5.94-6.22 (-CH-CH2-O- and -

CH=CH-, conjugated), 6.32-6.43 (-CH=CH-, conjugated).

3.3.8 Viscosity Measurements

Viscosity measurements were carried out using a Brookfield LV-II+ Pro

Viscometer with a speed of 10 rpm and 2.20 s-1 as the shear rate. The SC4-25 spindle was used for these measurements. 44 3.3.9 Formulations and Film Preparation

The coatings were formulated with equal amounts of each modified alkyd or

reactive diluent. A drier package was included in each formulation at 2 w.%, which

contained a 1:17:2 ratio of Borschi® OXY-coat, zirconium Hex-Cem, and calcium Hex-

Cem. BYK 333, a wetting agent was also included at 1.5 wt.%. The amounts of driers and

additives used were determined by ladder studies. The formulations are shown in Table 3.1

and Table 3.2. The coatings were cast using a drawdown bar at 2 mils on aluminum AQ-

36 panels. The formulations were also cast on glass panels at 4 mils for tensile testing.

They were cured at room temperature for 2 weeks prior to testing.

Table 3.1. Modified alkyd formulations. All amounts are in g. Formulations: Control ST-5 ST-10 ST-15 ST-20 LLOA 10 9.5 9 8.5 8 ASLOA 0 2.5 0.5 0.75 1 FLOA 0 2.5 0.5 0.75 1 Acetone 0.5 0.5 0.5 0.5 0.5 BYK 333 0.15 0.15 0.15 0.15 0.15 Driers 0.2 0.2 0.2 0.2 0.2

Table 3.2. Modified tung oil reactive diluent alkyd formulations. All amounts are in g. Formulations: STRD-2.5 STRD-5 STRD-7.5 STRD-10 STRD-15 STRD-20 LLOA 9.75 9.5 9.25 9 8.5 8 ASTO 0.125 0.25 0.375 0.5 0.75 1 FTO 0.125 0.25 0.375 0.5 0.75 1 Acetone 0.5 0.5 0.5 0.5 0.5 0.5 BYK 333 0.15 0.15 0.15 0.15 0.15 0.15 Driers 0.2 0.2 0.2 0.2 0.2 0.2 3.3.10 Coating Properties

Coatings tests were completed following ASTM standards. The pendulum (ASTM

D 4366-16)156 and pencil hardness (ASTM D 3363-05),157 MEK chemical resistance

(ASTM D 5402-15),158 pull-off (ASTM D 4541-17)159 and crosshatch adhesion (ASTM D

45 3359-17),160 and Gardner impact (ASTM D 2794-93)161 were all tested. A ramé-hart Model

500 Advanced Goniometer was used to measure contact angle with 2 µL drops of DI water

on the coating. The tests were repeated 5 times when no failure mode was specified by the

ASTM method (such as, pull-off adhesion and pendulum hardness). The results are

reported as an average with a 95 % confidence interval.

3.3.11 Tensile Testing

Tensile testing was carried out with an Instron Tensiometer 5567. The load cell

used was 100 N, and hydraulic clamps were used for the testing, with an extension rate of

10 mm min-1. A micrometer was used for measurement of sample dimensions, which were around 25 x 10 x 0.06 mm. Strain at break, modulus and tensile strength were evaluated for each sample. Young’s modulus was calculated from the instantaneous modulus at 0.1 mm mm-1 of strain. Five samples at minimum were testing, and the mean was found with

a 95 % confidence interval.

3.3.12 Electrochemical Impedance Spectroscopy (EIS)

Electrochemical impedance spectroscopy (EIS) was completed using a Gamry

Reference 600+. A 3.5 wt.% NaCl solution in water was held on the coating surface

throughout the testing. The tests were done over 10 weeks.

3.3.13 Weathering

An Atlas Ci 4000 Weather-Ometer with a xenon arc lamp was used for accelerated

weather testing of the coatings. Prior to testing, the coatings were evaluated for color,162

gloss163 and pendulum hardness.156 The panels were placed in the weathering chamber,

facing the lamp. The panels were exposed to a cycle of light, humidity and water spray,

46 according to ASTM D6695-16.164 The irradiance is kept at 0.35 W/(m2-nm) at 340 nm. The samples are exposed to 18 h of continuous like exposure, cycled between 102 min. at 63

ºC and 50 % relative humidity (RH) and 18 min. with water spray. This is followed by 6 h of darkness at 24 ºC and 95 % RH. This cycling was continued for 500 h of exposure, and the coatings were removed and tested for color, gloss and pendulum hardness.

3.3.14 Energy Dispersive X-Ray Spectroscopy

A FEI Quanta 200 environmental scanning electron microscope with an EDAX energy-dispersive x-ray was used for analysis of the stratification of the coatings. Samples were prepared by removal from the substrate and then were broken under liquid nitrogen to maintain the integrity of the cross-section for the analysis. Carbon tape was used to place the sample of the SEM mount in an upright position. Dot maps were obtained of the fluorine and silicon in the cross-section of the coating. The amounts of each were then analyzed in ImageJ by dividing the resulting micrographs into quadrants.

3.4 Results

The objective of this study was to create a highly functional coating system with improved adhesion, corrosion resistance and mechanical properties. The performance of modified alkyds versus modified tung oil reactive diluents was studied. The modifications shown here have been reported previously,10,147,149 but they have not been combined into

one coating system. The combination of alkoxysilane and fluorine functional alkyds will

create coatings with improved adhesion, chemical and corrosion resistance. In one series

of experiments, amounts of alkoxysilane modified alkyd and fluorine modified alkyd were

varied from 2.5 wt.% up to 10 wt.% of each component. In a second series, amounts of

47 alkoxysilane modified and fluorine modified tung oil reactive diluents were varied from

1.25 wt.% up to 10 wt.% of each component. The coatings were tested for tensile and

coating performance, as well as barrier properties and corrosion resistance using EIS and accelerated weathering. The films were also evaluated using environmental SEM-EDX on

the cross-section to look at the ability of each component to partially self-stratify.

3.4.1 Synthesis and Characterization

The alkoxysilane modified alkyd (ASLOA) was synthesized to have 15% of the

free hydroxyl groups functionalized by TESPIC. This amount of modification was shown

previously by Salata et al.149 to create an alkyd with acceptable application viscosity. The

reaction scheme is shown in Figure 3.1, and was carried out according to Salata et al.149

The progress of the reaction was monitored by FT-IR, looking at the disappearance of the

N=C=O vibration at 2270 cm-1. Figure 3.2 shows the product of the reaction following 100 min. at 75 °C. There is no vibration at 2270 cm-1, signaling that the reaction is complete

because all the isocyanate has reacted. The product was also characterized by 1H-NMR spectroscopy and MALDI MS. The 1H-NMR spectrum (Figure 3.3) shows a resonance around 0.6 ppm, corresponding to the hydrogens in the CH2-Si. There is also a slight

resonance at 3.2 ppm that corresponds to the hydrogen adjacent to the carbamate linkage.

MALDI MS (Figure 3.4) shows the weight difference of 247 g mol-1 corresponding to the addition of TESPIC. corresponding to the addition of TESPIC.

O O O O O C O O O O N HO O O O OH Dibutyltin Dilaurate O O HO O O O OH + O OH Heat O O O R n NH O Si O R n O

O Si O O Figure 3.1. Synthesis scheme for alkoxysilane functional alkyd (ASLOA). 48 Figure 3.2. FT-IR spectrum of alkoxysilane functional alkyd (ASLOA).

ThisreportwascreatedbyACD/NMRProcessorAcademicEdition.Formoreinformationgotowww.acdlabs.com/nmrproc/ SFLOA_062618 1.25 1.0 O O O O

HO O O O OH 0.9 O O O O 0.8 R HN 1.30

0.7

O Si O 0.6 O

0.5

0.4 1.22 NormalizedIntensity 7.26 5.35 0.3 5.34 2.31 0.88 2.03 2.02 2.80 1.20 2.78

0.2 0.97 1.60 0.97 2.29 7.71 2.35 7.53 0.85 0.99 3.81 3.83 2.09

0.1 4.32 4.17 4.31 3.78 4.37 4.15 4.28 4.27 4.14 4.42 2.37 5.46 4.48 0.63 0.60 3.16 4.54 0.65 0.07 0 4.00 5.38 2.41 0.71 0.08 3.08 2.98 4.88 3.35 20.76 3.53 0.15

7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 ChemicalShift(ppm)

No. (ppm) 1(Hz) Height No. (ppm) (Hz) Height No. (ppm) (Hz) Height No. (ppm) (Hz) Height Figure1 0.07 3.3. 20.5H-NMR0.0047 spectrum14 1.25 of 375.0alkoxysilane1.0000 27 functional3.78 1134.1 alkyd0.0405 (ASLOA).40 4.32 1294.8 0.0482 2 0.60 180.3 0.0104 15 1.30 388.7 0.7441 28 3.81 1140.8 0.0663 41 4.37 1309.4 0.0385 3 0.63 187.9 0.0110 16 1.60 480.1 0.1474 29 3.83 1147.8 0.0524 42 4.42 1323.8 0.0268 4 0.65 195.5 0.0058 17 2.02 604.2 0.1899 30 3.85 1154.9 0.0160 43 4.48 1342.2 0.0146 5 0.85 256.1 0.0880 18 2.03 609.8 0.1997 31 4.11 1232.2 0.0176 44 4.54 1359.8 0.0058 6 0.88 262.9 0.2019 19 2.09 627.3 0.0512 32 4.13 1238.3 0.0234 45 5.34 1601.3 0.2748 7 0.94 282.8 0.0835 20 2.29 686.5 0.1341 33 4.14 1240.6 0.0274 46 5.35 1603.7 0.2751 8 0.95 284.2 0.0859 21 2.31 692.0 0.2047 34 4.15 1244.249 0.0381 47 5.46 1637.3 0.0149 9 0.97 290.1 0.1520 22 2.35 703.2 0.0951 35 4.17 1250.0 0.0453 48 7.26 2176.3 0.3163 10 0.97 291.5 0.1404 23 2.37 711.1 0.0261 36 4.21 1262.3 0.0246 49 7.53 2256.2 0.0886 11 0.99 297.4 0.0816 24 2.78 834.3 0.1739 37 4.27 1278.7 0.0312 50 7.71 2310.7 0.0975 12 1.20 360.1 0.1775 25 2.80 839.6 0.1826 38 4.28 1282.8 0.0373 13 1.22 367.1 0.3652 26 3.16 947.6 0.0069 39 4.31 1290.7 0.0417 Figure 3.4. MALDI MS of alkoxysilane modified alkyd (bottom), compared with unmodified alkyd (top).

The fluorinated alkyd (FLOA) was synthesized to have 10 mol % fluorinated polyol

in the backbone. Using a small amount of the fluorinated additive helps to reduce cost, and

at this low level, there will be little effect on the processibility of the alkyd, but it will still

have a significant effect on the overall performance of the coating. The reaction scheme can be seen in Figure 3.5. The product was characterized by 1H-NMR spectroscopy and

MALDI MS. As is seen in Figure 3.6, the 1H-NMR spectrum does not show any significant difference with the fluorine-functionality added. The CH2-CF2 group has a resonance that overlaps with that of the protons (CH2-O) in the monoglyceride in the backbone at 4.3 ppm.

To better elucidate the structure, MALDI MS was employed. As seen in Figure 3.7, the molecular weight differences of 260 g mol-1 indicate that the fluorinated diol adds in the backbone as a repeat unit.

50 O O O F F O F F xylene + HO OH + OH HO HO R OH F F F F heat

O O O O O O F F F F O OH HO O O O F F O F F O R n Figure 3.5. Reaction scheme for fluorinated alkyd (FLOA).

O O O O O O F F F F O OH HO O O O F F O F F O R n

O O O O

HO O O O OH O OH O R

14 Figure 3.6. 1H-NMR of fluorinated alkyd (FLOA).

Figure 3.7. MALDI-MS of fluorinated alkyd (FLOA). 51 The alkoxysilane modified tung oil (ASTO) was functionalized by a Diels-Alder

reaction between the conjugated bonds in �-eleostearic acid and methacrylate (Figure 3.8), as previously reported by Wutticharoenwong et al.10 The reactive diluent was formulated to have about 25 mol% alkoxysilane functionality. Looking at the FT-IR spectrum in

Figure 3.9, an intense Si-O stretching can be observed at 1080 cm-1. An Si-C stretching can also be seen around 810 cm-1. The 1H-NMR spectrum (Figure 3.10) shows a resonance at

0.6 ppm, corresponding to CH2-Si, and a resonance at 3.5 ppm for the Si-O-CH3 protons.

There is also a resonance at 1.1 ppm caused by the C-CH3 group from the methacrylate

Diels-Alder reaction. The changes in splitting between 5.5 - 6.1 ppm are caused by the non-

conjugated double bonds in the cyclohexene ring that is formed by the Diels-Alder

reaction. The MALDI MS is shown in Figure 3.11, and the ionization and splitting

correspond to the molecular weight of ASTO.

O Si O O O O O O O O O Si O O O O O O O O Phenothiazine Heat O O

O O Figure 3.8. Reaction scheme for alkoxysilane functionalized tung oil (ASTO).

52 Figure 3.9. FT-IR spectrum of alkoxysilane functional tung oil (ASTO).

ThisreportwascreatedbyACD/NMRProcessorAcademicEdition.Formoreinformationgotowww.acdlabs.com/nmrproc/ SFTO_110618 1.29 3.57 1.0 O

Si O 3.56 O 0.9 O O

0.8 O O O 0.7 1.25 O

0.6 O

0.5 O 1.33 0.89 0.4 0.88 NormalizedIntensity 2.30

0.3 1.36 0.87 0.86 0.91 2.28 1.93 1.60 0.2 2.33 1.07 2.08 1.38 4.11 2.10 4.14 4.16 2.05 1.06 6.12 4.27 7.26 4.26 3.61 5.54 6.16 5.33 0.66 4.12 4.10 3.61 6.06 6.10 6.08 5.98 5.54 5.39 5.35 5.53 0.69 4.31 4.30 5.53 5.42 6.13 0.68 0.1 6.01 5.26 6.19 1.72 6.36 6.37 1.73 1.75 1.12 4.03 6.47 1.52 2.76 2.36 1.48 1.02

0 2.32 5.54 2.291.85 5.15 5.03 1.32 8.33 0.37 6.11 9.62 1.60 9.49 42.97 2.08 8.99 2.39

7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ChemicalShift(ppm)

No. (ppm) Value 1 AbsoluteValue Non-NegativeValue No. (ppm) Value AbsoluteValue Non-NegativeValue Figure1[0.5667..0.7379] 32.38757348.10. H-7.64083680e+7NMR of alkoxysilane2.38757348 functional10[3.4891..3.6547]8.33362579 tung oil2.66697008e+8 (ASTO). 8.33362579 2[0.7607..0.9719]8.99383640 2.87825408e+8 8.99383640 11[3.9172..4.0713]1.31794643 4.21776040e+7 1.31794643 3[0.9890..1.1032]2.08383632 6.66880080e+7 2.08383632 12[4.0713..4.3624]5.03378201 1.61093696e+8 5.03378201 4[1.1375..1.4400]42.97442627 1.37528986e+9 42.97442627 13[4.9618..5.4698]5.14911318 1.64784592e+8 5.14911318 5[1.4400..1.8681]9.49155521 3.03753664e+8 9.49155521 14[5.4698..5.5725]1.84748840 5.91242840e+7 1.84748840 6[1.8681..1.9480]1.59854019 5.11573120e+7 1.59854019 15[5.5725..5.7780]2.28656435 7.31758160e+7 2.28656435 7[1.9480 ..2.2277]9.62455177 3.08009888e+8 9.62455177 16[5.8808..6.2404]5.53594494 1.77164176e+8 5.53594494 8[2.2277..2.3932]6.11460352 1.95682720e+8 6.11460352 17[6.2689..6.5143]2.32003188 7.42468640e+7 2.32003188 9[2.6843..2.8042]0.37149695 1.18888380e+7 0.37149695

No. (ppm) (Hz) Height No. (ppm) (Hz) Height No. (ppm) (Hz) Height No. (ppm) (Hz) Height 1 0.61 183.2 0.0184 4 0.65 195.2 0.0470 7 0.69 206.1 0.0609 10 0.86 258.2 0.1979 2 0.62 186.7 0.0294 5 0.66 197.9 0.0759 8 0.70 208.7 0.0304 11 0.87 261.1 0.2475 3 0.63 190.0 0.0427 6 0.68 202.5 0.0510 9 0.72 214.6 0.0422 12 0.88 264.0 0.3442

53 ASTO 1145.837

+Na = 1145.8

1392.945

247.108 Da

1089.756 249.119 Da 375.447 Da 249.11 Da 1642.064 248.118 Da 2017.511 1768.399 2266.621 2514.739

1200 1400 1600 1800 2000 2200 2400 2600 m/z Figure 3.11. MALDI MS of alkoxysilane functional tung oil (ASTO).

The fluorine modified tung oil (FTO) was modified using a Diels-Alder reaction

between the conjugated bonds in �-eleostearic acid and methacrylate (Figure 3.12), as previously reported by Wutticharoenwong et al.10 The reactive diluent was formulated to have about 25 mol% fluorine functionality. The 1H-NMR spectrum in Figure 3.13 shows the resonance at 4.5 ppm corresponds to the CH2-CF3, and the new non-conjugated

double bond in the cyclohexene is shown to disrupt the double bond splitting from 5.5 –

6.1 ppm. There is also a resonance at 1.1 ppm caused by the C-CH3 group from the methacrylate Diels-Alder reaction. The MALDI MS in Figure 3.14 shows an ionization corresponding to the molecular weight of FTO.

F O F O F O O F O O O F O O F O O Phenothiazine O Heat O O

O O

Figure 3.12. Reaction scheme for fluorinated tung oil (FTO). 54 ThisreportwascreatedbyACD/NMRProcessorAcademicEdition.Formoreinformationgotowww.acdlabs.com/nmrproc/ FTO_2_061118.esp 1.29 1.0 F O F 0.9 O F O 0.8 O O 0.7 O

0.6

O 1.25

0.5 O 1.33 0.89 1.35 0.88 0.4 NormalizedIntensity 1.36 2.30

0.3 0.87 0.91 0.91 2.28 2.08 6.06 1.97 0.86 1.60 2.33 0.2 1.38 2.10 4.14 4.16 4.28 4.26 6.12 5.98 6.16 5.69 6.02 5.33 5.67 4.12 4.51 5.35 4.54 5.70 4.30 4.10 4.32 5.39 5.64 5.70 6.22 5.71 0.1 7.26 5.26 6.36 6.32 1.12 6.37 1.11 5.74 6.41 4.57 2.76 6.97 3.18 0 0.13 1.90 7.93 3.84 4.50 0.91 4.45 0.11 0.48 6.79 13.64 6.95 44.760.93 9.01

7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0 ChemicalShift(ppm)

No. (ppm) Value 1 AbsoluteValue Non-NegativeValue No. (ppm) Value AbsoluteValue Non-NegativeValue Figure1[0.7937..0.9663] 39.00908566.13. H-4.07163168e+8NMR spectrum9.00908566 of fluorinated9[4.0613..4.3604] tung4.45162010 oil 2.01189744e+8(FTO). Reprinted4.45162010 from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.0172[1.0641..1.1676]0.93110055 4.20808360e+7 0.93110055 with10[4.4410 permission..4.6078]0.91358465 from4.12892080e+7 Elsevier.0.91358465 3[1.1791..1.4553]44.76173019 2.02299418e+9 44.76173019 11[5.1601..5.4707]4.50109196 2.03425632e+8 4.50109196 4[1.5128..1.7084]6.95123434 3.14159136e+8 6.95123434 12[5.4937..5.7469]3.84420180 1.73737648e+8 3.84420180 5[1.9040..2.2319]13.64047241 6.16477440e+8 13.64047241 13[5.8964..6.2704]7.93283415 3.58522272e+8 7.93283415 6[2.2319..2.3815]6.78503704 3.06647904e+8 6.78503704 14[6.2761..6.5407]1.89562535 8.56722720e+7 1.89562535 7[2.6864..2.8187]0.47930241 2.16619420e+7 0.47930241 15[6.8997..7.0455]0.12681283 5.73127150e+6 0.12681283 8[3.0610..3.2302]0.11354328 5.13155750e+6 0.11354328 FTO No. (ppm) (Hz) Height No. (ppm) (Hz) Height No. (ppm) (Hz) Height No. (ppm) (Hz) Height 1 0.86 258.8 0.1818 5 0.91 271.3 0.2195 9 1.25 373.8 0.5373 13 1.36 406.9 0.3335 2 0.87 262.0 0.2256 6 0.91 +Na272.5 = 1064.50.2113 10 1.29 387.9 1.0000 14 1.38 414.8 0.1669 3 0.88 264.6 0.3948 7 1.11 333.1 0.0301 11 1.33 399.6 0.4568 15 1.60 479.2 0.1720 4 0.89 265.8 0.4133 8 1.12 336.3 0.0410 12 1.35 403.7 0.4010 16 1.97 591.9 0.1832

1145.847

1063.733 250.136 Da

1395.983 373.415 Da 1769.398

1200 1400 1600 1800 2000 2200 2400 m/z Figure 3.14. MALDI MS of fluorinated tung oil (FTO). Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier.

The viscosity of each component was measured and is shown in Table 3.3. The alkyds have much higher viscosity than the reactive diluents, as expected. The alkoxysilane

55 modified components also have higher viscosity than their counterparts, likely due to

hydrolysis and condensation reactions taking place with the alkoxysilane groups and any

amount of water present.

Table 3.3. Viscosity measurements of alkyd, modified alkyds and modified reactive diluents. Viscosity (cP s) LLOA 26,000 ASLOA 47,000 FLOA 11,000 ASTO 1,600 FTO 800

3.4.2 Coating Formulation

Each formulation contains equal amounts of both an alkoxysilane modified component and a fluorine modified component. These were used as additives with an unmodified alkyd serving as the base resin. To study the effect of molecular weight and viscosity on the properties of the coatings, one set of formulations was made with the modified alkyds, and one with the reactive diluents. The modified alkyd formulations started at 2.5 wt.% of each component and went up to 10 wt.% in increments of 2.5 wt.%.

For the reactive diluent series, the formulations started with 1.25 wt.% of each component and increased up to 10 wt.% of each component. Smaller amounts of reactive diluents were used to mitigate any negative effects caused by the addition of lower molecular weight

components into the coating formations. The exact formulations can be seen in Table 3.1

and Table 3.2.

3.4.3 Coating Performance

To better understand the effects of these components on the coating performance,

the properties were evaluated by standard test methods following curing. The properties 56 for the modified alkyd series can be seen in Table 3.4. Compared to the control, all the

properties increased or showed no change. The hardness, chemical resistance and contact

angle are all shown to increase with increasing modified alkyd content, which was expected due to the FLOA content. Pull-off adhesion shows no change from the control, which indicates that there were no significant reactions between ASLOA and free hydroxyl groups at the aluminum substrate surface. The other coating properties, including cross- hatch adhesion, and reverse and direct impact resistance, all measured at the upper limit of

the test.

Table 3.4. Coatings properties of the modified alkyd series. Pendulum Pencil Pull-Off Crosshatch Hardness Hardness Adhesion Adhesion (s) (lb/in2) Control 17 ± 2 3H 250 ± 30 5B ST-5 17 ± 2 4H 5B 220 ± 20 ST-10 18 ± 2 3H 5B 210 ± 20 ST-15 16 ± 1 2H 5B 220 ± 40 ST-20 15 ± 1 4H 5B 210 ± 20 MEK Direct Impact Reverse Impact Contact Double Rubs Resistance Resistance Angle (lb/in) (lb/in) (º) Control 100 ± 20 > 40 > 40 94 ± 1 ST-5 170 ± 60 > 40 > 40 100 ± 1 ST-10 > 200 > 40 > 40 104.5 ± 0.6 ST-15 > 200 > 40 > 40 105.2 ± 0.5 ST-20 > 200 > 40 > 40 105 ± 1

The coating properties for the modified tung oil reactive diluent series are shown

in Table 3.5. Hardness, adhesion, chemical resistance and contact angle are all shown to increase with increasing RD content. Pencil hardness and pull-off adhesion show a

significant increase for RD loadings over 10 wt.%, which could imply that stratification is

occurring. Fluorine moving toward the surface would cause the increase in hardness, and

57 adhesion would increase due to the reaction of the alkoxysilanes with free hydroxyl groups

at the substrate surface. Other coating properties, including cross-hatch adhesion and

impact resistance measured at the upper limit of the test.

Table 3.5. Coatings properties of the reactive diluent series. Pendulum Pencil Pull-Off Crosshatch Hardness Hardness Adhesion Adhesion (s) (lb/in2) Control 17 ± 2 3H 250 ± 30 5B STRD-2.5 41 ± 2 F 200 ± 60 5B STRD-5 40 ± 3 2H 180 ± 30 5B STRD-7.5 27 ± 2 2H 190 ± 40 5B STRD-10 27 ± 3 2H 270 ± 70 5B STRD-15 25 ± 2 6H 390 ± 20 5B STRD-20 27 ± 3 6H 320 ± 60 5B MEK Direct Impact Reverse Impact Contact Double Rubs Resistance Resistance Angle (lb/in) (lb/in) (º) Control 100 ± 20 > 40 > 40 94 ± 1 STRD-2.5 70 ± 20 > 40 > 40 106.7 ± 0.6 STRD-5 200 ± 10 > 40 > 40 105.3 ± 0.3 STRD-7.5 200 ± 10 > 40 > 40 105.0 ± 0.2 STRD-10 200 ± 10 > 40 > 40 104.7 ± 0.3 STRD-15 200 ± 10 > 40 > 40 107 ± 2 STRD-20 200 ± 10 > 40 > 40 108 ± 2

3.4.4 Tensile Properties

Evaluating the tensile properties adds insight into the crosslink density and

mechanical performance of these coatings. The tensile properties of both series are shown

in Figure 3.15. The modified alkyd series has consistently higher strain at break, modulus

and tensile strength compared to the RD series. This was expected because the addition of

lower molecular weight components tends to have a detrimental effect on the overall

mechanical properties of the coatings.11,165 With increasing content of the modified alkyds, there is no additional improvement in properties, indicating that only a small amount of

58

these additives are needed. For the modified tung oil series, the modified coatings performed slightly better than the control in terms of tensile strength and modulus, which reach a maximum at STRD-5, followed by a decrease in properties with higher loadings of the RDs.

Strain at Break, STRD Strain at Break, ST 1.6 1.6

1.4 1.4

1.2 1.2

1 1

0.8 0.8

0.6 0.6

Strain at Break, mm/mm Strain at Break, 0.4 0.4 Strain at Break, mm/mm Strain at Break, 0.2 0.2

0 0 Control STRD-2.5 STRD-5 STRD-7.5 STRD-10 Control ST-5 ST-10 ST-15 ST-20

Modulus, STRD Modulus, ST 4.5 4.5 4 4 3.5 3.5

3 3 2.5 2.5 2 2 Modulus, MPa 1.5 Modulus, MPa 1.5 1 1 0.5 0.5 0 0 Control STRD-2.5 STRD-5 STRD-7.5 STRD-10 Control ST-5 ST-10 ST-15 ST-20

Tensile Strength, STRD Tensile Strength, ST

4.5 4.5 4 4 3.5 3.5 3 3 2.5 2.5 2 2 1.5 1.5 Tensile MPa Strength, 1 Tensile MPa Strength, 1 0.5 0.5 0 0 Control STRD-2.5 STRD-5 STRD-7.5 STRD-10 Control ST-5 ST-10 ST-15 ST-20

Figure 3.15. Comparison of tensile properties between the reactive diluent series (left) and the modified alkyd series (right).

59 3.4.5 Corrosion Resistance

Corrosion resistance was analyzed by electrochemical impedance spectroscopy

(EIS) to understand the effects of these additives on the barrier properties of the coatings.

Measurements were taken over several weeks to see how the barrier properties changed

over time, as shown in Figure 3.16. For the modified alkyd series, initially, each sample had a higher impedance modulus than the control. A higher impedance modulus indicates that the coating is providing a more effective barrier between the salt solution and the

substrate. The modified alkyds all had impedance moduli an order of magnitude higher

than that of the control at low frequency. Following 10 weeks of exposure, ST-10 still

showed the highest impedance modulus at low frequency, however, all the samples showed

similar results.

ST Comparisons, initially Comparison of Self-Stratifying Impedance after 10 weeks

100000 100000 Control

ST-5 10000 10000 Control ST-10 ST-5 1000 ST-15 1000 ST-10 ST-20 ST-15

100 100 ST-20 Impedance Modulus, Ohm Modulus, Impedance Impedance Modulus, Ohm Modulus, Impedance

10 10 0.01 0.1 1 10 100 1000 10000 100000 0.01 0.1 1 10 100 1000 10000 100000 Frequency, Hz Frequency, Hz

Figure 3.16. Impedance modulus of the modified alkyd series initially (left) and after 10 weeks (left).

For the modified tung oil series, the results are shown in Figure 3.17. The initial testing showed that STRD-10 had the highest impedance modulus, 9.7 x 104 Ohms, while

the control and other samples were all about the same. After 10 weeks of exposure, STRD-

10 still had the highest impedance modulus, and the control and other RD-containing

60

samples have decreased significantly. This indicates that STRD-10 is providing excellent barrier properties, and this would likely translate to higher corrosion resistance.

STRD Comparisons, initially STRD Comparisons after 10 weeks

100000 100000 Control

STRD-2.5 Control 10000 10000 STRD-5 STRD-2.5

1000 STRD-7.5 STRD-5 1000 STRD-10 STRD-7.5 100 100 STRD-10 Impedance Modulus, Ohm Impedance Modulus, Ohm

10 0.01 0.1 1 10 100 1000 10000 100000 10 0.01 0.1 1 10 100 1000 10000 100000 Frequency, Hz Frequency, Hz

Figure 3.17. Impedance modulus of the RD series, initially (left) and after 10 weeks (right).

Comparing the two series, it can be seen that the reactive diluent series is consistently lower in impedance modulus than the modified alkyd series. However, STRD-

10 has a very similar impedance modulus to ST-10 throughout the testing and performs even better after 10 weeks of exposure (Figure 3.18). This shows that the same barrier properties can be achieved with modified reactive diluents, rather than modified alkyds.

Comparison of Self-Stratifying Impedance Spectroscopy, after 10 weeks 100000

Control

10000 ST-5 ST-10 STRD-5 1000 STRD-10 Impedance Modulus, Ohm 100

10 0.01 0.1 1 10 100 1000 10000 100000 Frequency, Hz Figure 3.18. Comparisons of impedance modulus for the modified alkyds and reactive diluent series.

These samples were also analyzed for their capacitance and water uptake, using modeling with an equivalent electric circuit, which is shown in Figure 3.19. This circuit

61 represents diffusion through the coating in terms of a constant phase element (CPE), and also shows a secondary resistor and capacitor in series that represent delamination at the coating/substrate interface. The fittings on the Nyquist plots are shown in Figure 3.20,

Figure 3.21 and Figure 3.22. Using the constant phase element parameters Y0 and �, the coating capacitance (CC) and water uptake of the coating can be calculated using equations (12) and (13),

� = �(� ) (12)

� log ( , ) �, (13) ����� ������ (%) = ∗ 100% log (80) where � is the angular frequency at which |Zimag| reaches a maximum on the

Nyquist plot, �, is the capacitance at time t and �, is the initial capacitance. The water uptake can be calculated from this equation because the capacitance of the coating is represented by equation (14),

� � � � = (14) � where � is the permittivity of free space, � is the permeability of the coating, and A is the area of the coating surface and d is the thickness. The dielectric constant of the coating is one order of magnitude less than water, generally, so as water permeates the coating, the capacitance increases, as the other parameters remain constant.

Because of this, water uptake can be evaluated from the change in coating capacitance over time, divided by log(80), where 80 is the dielectric constant of water. The values for capacitance and water uptake over 6 weeks are shown in Table 3.6 and Table 3.7.

62 Figure 3.19. Equivalent electric circuit for EIS modeling of these coatings.

Table 3.6. Coating capacitance calculated from Equation (12), using the equivalent circuit shown in Figure 3.19.

Sample, CC Initially 1 week 2 weeks 3 weeks 5 weeks Control 1.08 x 10-6 0.91 x 10-6 1.05 x 10-6 0.84 x 10-6 1.29 x 10-6 ST-10 1.45 x 10-7 6.22 x 10-7 10.2 x 10-7 15.4 x 10-7 8.45 x 10-7 STRD-10 1.16 x 10-6 10.5 x 10-6 29.2 x 10-6 22.4 x 10-6 4.59 x 10-6

Table 3.7. Water uptake, calculated from Equation (13), using the equivalent circuit shown in Figure 3.19. Sample, 1 week 2 weeks 3 weeks 5 weeks Water Uptake (%) Control -4.04 -0.64 -5.77 4.04 ST-10 33.2 44.5 54.0 40.2 STRD-10 50.3 73.6 67.6 31.4

The capacitance and water uptake of the control alkyd do not change significantly

over the period of testing, which was also found to be the case for the impedance modulus,

shown in the Bode plots. The systems containing modified components were found to have

increasing capacitance over the time of the test, indicating an increase in water uptake. ST-

10 shows an increase over the first three weeks of exposure before a slight decrease was

found. For STRD-10, the water uptake was found to increase over the first two weeks,

before leveling off and decreasing. From the Bode plots of impedance, it was found that

63

despite this increase in water uptake, the impedance remained high at the low frequency

limit, indicating the coatings were still providing a good barrier to corrosion.

Control

Figure 3.20. Nyquist plots and fittings for control over 6 weeks.

ST-10

Figure 3.21. Nyquist plots and fittings for ST-10 over 6 weeks.

64 STRD-10

Figure 3.22. Nyquist plots and fittings for STRD-10 over 6 weeks.

3.4.6 Weathering

Weathering testing provides more information on how these coatings will perform

when exposed to outdoor conditions. Exposure to sunlight, heat and moisture can cause

color and gloss changes, as well as changes in the properties of the coatings.3,128,129,131 For

this reason, the panels were analyzed for color, gloss and pendulum hardness prior to being

placed in the accelerated weathering environment for 500 h.

The color change was analyzed by the L, a, b values of the panels before and after

weathering. The color difference was calculated from ∆E, using the Equation (15).

∆� = [(∆�∗) + (∆�∗) + (∆�∗)]/ (15)

If ∆E ~ 2.3, this is considered a “just noticeable difference,” where the human eye can just

begin to notice a color change. Below this value, color differences are not recognizable and

65 could also be contributed to measurement error.162,166 For these coatings, ∆E < 1.5 in every case, indicating that no significant color change is occurring over the weathering period.

The gloss was also measured before and after weathering. Gloss was measured at a

20º angle, which is used specifically for coatings with high gloss.163,167 The gloss changes

in percentages are shown in Table 3.8. It was found that with the addition of the fluorinated

and alkoxysilane functional additives, the gloss changes were less significant than that of

the control. However, in all cases, the gloss was slightly decreased. This loss of gloss is

likely due to increased surface roughness as the coatings are exposed to the harsh

conditions of the accelerated weather testing.167,168

In addition to the appearance of the coating following weathering, changes in the hardness of the surface were also evaluated by pendulum hardness. It was found that significant hardening took place during the accelerated testing, as shown in Figure 3.23.

The effect was more significant on the modified alkyds. The increased hardness is a result of embrittlement of the alkyd coating during the weathering process. This occurs due to increased photo-oxidation reactions in the alkyd, resulting in more crosslinking and hardening of the coating.168–170

Percent Hardness Change, Reactive Diluent Series Percent Hardness Change, Modified Alkyd Series

500 500

400 400

300 300

200 200 Percent Change, % Change, Percent Percent Change, % Change, Percent 100 100

0 0 Control STRD-2.5 STRD-5 STRD-7.5 STRD-10 STRD-15 STRD-20 Control ST-5 ST-10 ST-15 ST-20 Figure 3.23. Percent change in pendulum hardness after 500 h of weathering, with reactive diluents (left) and modified alkyds (right).

66 Table 3.8. Gloss changes following 500 h of weathering. Percent Change in 20º Gloss % Control -31 % STRD-2.5 -20 % STRD-5 -17 % STRD-7.5 -8.7 % STRD-10 -27 % STRD-15 0.0 % STRD-20 -17 % ST-5 -33 % ST-10 -18 % ST-15 -18 % ST-20 -20 % 3.4.1 Stratification

Finally, any potential stratification in these films was analyzed by EDX to observe the distribution of fluorine and silicon within the cross-section of the coating. The series

with modified alkyds was not expected to show any significant stratification, due to the

higher molecular weight, and therefore, lower mobility of these additives within the bulk.

Some very slight stratification of both the fluorine modified and alkoxysilane modified

alkyds did occur, and both additives moved toward the air/coating interface. The proportion

of each additive is quantified in Table 3.9, and the micrographs are shown in Figure 3.24.

In all cases, FLOA showed a more significant concentration at the surface, which is likely

due to the lower surface energy and viscosity of this additive. There is at least a 20%

increase at the air interface for the fluorinated component in all cases. For the alkoxysilane

modified alkyds, there was less movement, and at most, a 20% increase was observed at

the surface for the alkoxysilane. For the alkoxysilane components, less mobility is expected

due to the higher viscosity of this additive, and the competing stratification forces. The

alkoxysilane is a low surface energy additive,171,172 and thus, will have some drive toward

67

the air interface. However, it also has the ability to react with any free hydroxyl groups present at the substrate interface,154 creating a driving force toward the substrate interface as well. These competing forces could account for this difference in stratification ability.

Table 3.9. Stratification toward the surface (↑) for modified alkyd series. F Si ST-5 ↑ 33 % ↑ 15 % ST-10 ↑ 46 % ↑ 11 % ST-15 ↑ 24 % ↑ 13 % ST-20 ↑ 21 % ↑ 20 %

68 F Si ST-5 Surface 33% increase 168 ± 5 15% increase 761 ± 5

205 ± 5 710 ± 5

169 ± 5 658 ± 5

126 ± 5 662 ± 5 5 µm Substrate

F Si ST-10 Surface 46% increase 86 ± 5 11% increase 418 ± 5

76 ± 5 414 ± 5

65 ± 5 433 ± 5

59 ± 5 378 ± 5 5 µm Substrate

F Si ST-15 Surface 24% increase 164 ± 5 13% increase 775 ± 5

114 ± 5 608 ± 5

108 ± 5 560 ± 5

132 ± 5 683 ± 5 Substrate 5 µm

F Si ST-20 Surface 21 % increase 52 ± 5 20 % increase 218 ± 5

59 ± 5 213 ± 5

58 ± 5 222 ± 5

5 µm 43 ± 5 182 ± 5 Substrate

Figure 3.24. EDX dot maps of the cross section of the modified alkyd coating series.

69 Based on the coating properties and the corrosion resistance data, it was expected

that at high concentrations of the RDs, stratification was occurring in both directions. This

series was expected to show more stratification, due to the lower molecular weight and viscosity of the tung oil components. At low loadings, the stratification behavior is similar to that seen in the modified alkyd series, where both components move toward the air interface. However, the RDs show a higher degree of stratification than the modified alkyds at these low loadings. Table 3.10 quantifies the percentages of fluorine and silicone movement in both directions, and the micrographs are shown in Figure 3.25 and Figure

3.26. For STRD-5, there is a 48% increase of FTO and a 28% increase of ASTO at the air

interface. At higher loadings of the reactive diluents, the bidirectional stratification

occurred with FTO moving toward the air interface and ASTO moving toward the substrate

interface. FTO still shows more movement than ASTO, especially at higher loadings. For

STRD-20, there is a 75% increase in fluorine at the air interface, and an 11% increase in

alkoxysilane at the substrate. The discrepancies in the stratification of each component can

be explained by the difference in their viscosity. FTO has a lower viscosity than ASTO and

is therefore, more mobile within the coating during the curing process. As aforementioned,

ASTO has competing forces driving it toward the surface and the substrate, and in the case

of the modified tung oil components, ASTO is also more miscible with FTO than with the

larger alkyd, and therefore will be partially driven toward the air interface with FTO.

70 F Si STRD-2.5 Surface 47 % increase 56 ± 5 1 % increase 386 ± 5

33 ± 5 196 ± 5

31 ± 5 250 ± 5

10 µm 38 ± 5 384 ± 5 Substrate

F Si STRD-5 Surface 48 % increase 83 ± 5 28 % increase 327 ± 5

83 ± 5 309 ± 5

61 ± 5 268 ± 5

56 ± 5 255 ± 5 2.5 µm Substrate

F Si STRD-7.5 Surface 25% increase 50 ± 5 23% increase 433 ± 5

44 ± 5 385 ± 5

35 ± 5 339 ± 5

40 ± 5 353 ± 5 Substrate 5 µm

F STRD-10 Surface 53% increase 119 ± 5 4051 ± 5

130 ± 5 4200 ± 5

134 ± 5 4576 ± 5

Si 4287 ± 5 78 ± 5 5 µm 7% increase Substrate

Figure 3.25. EDX dot maps of fluorine and silicon for the modified reactive diluent series, STRD 2.5-10. 71 F STRD-15 Surface 44% increase 26 ± 5 134 ± 5

26 ± 5 114 ± 5

22 ± 5 134 ± 5

18 ± 5 Si 137 ± 5 9% increase Substrate 5 µm

F STRD-20 Surface 75 % increase 35 ± 5 322 ± 5

35 ± 5 298 ± 5

39 ± 5 345 ± 5

20 ± 5 S 357 ± 5 10 µm 11 % increase Substrate Figure 3.26. EDX dot maps of fluorine and silicon for the modified reactive diluent series, STRD 15-20.

Table 3.10. Stratification of fluorine and silicon toward the surface (↑) or the substrate (↓) interface of the coating for the reactive diluent series. F Si STRD-2.5 ↑ 47 % ↑ 1 % STRD-5 ↑ 48 % ↑ 28 % STRD-7.5 ↑ 25 % ↑ 23 % STRD-10 ↑ 53 % ↓ 7 % STRD-15 ↑ 44 % ↓ 9 % STRD-20 ↑ 75 % ↓ 11 %

3.5 Discussion

The objective of this study was to create high performance alkyd coatings with

enhanced durability, adhesion and corrosion resistance. Evaluation of modified alkyds and

reactive diluents as additives for this purpose was carried out. The first series used both 72 fluorine and alkoxysilane modified alkyds and the second series used modified tung oil-

based reactive diluents with the same functionalities. The coating performance, tensile

properties, corrosion resistance and weatherability were analyzed. Due to significant

improvements in adhesion, hardness and corrosion resistance for some of the RD samples,

stratification of the fluorine and silicon was evaluated for both coating systems. Based on

previous studies of stratifying coatings, it was expected there would be an increase in

fluorine concentration at the air/coating interface and an increase in alkoxysilane

concentration at the coating/substrate interface.110 The low surface energy of fluorine has

been shown to drive migration to the coating surface during curing.110,117,119,148

Alkoxysilanes also have low surface energy,171,172 but it is also expected that the alkoxysilane will interact and react with free hydroxyl groups at the substrate interface through condensation and hydrolysis reactions,154 driving a slightly higher concentration

at the substrate than at the surface.111 However, this increased concentration would be very slight due to the competing forces on the alkoxysilane components.

Overall, it was expected that the modified alkyd series would perform better in regard to coating, tensile and corrosion resistance properties, and the RD series was

expected to show a decrease in those properties at higher loadings. The lower molecular

weight and viscosity of RDs has been shown previously to be detrimental to coating

properties, especially at high loadings. The polyester backbone in the alkyd provides the

only rigid feature of this polymer structure, and the curing mechanism involves building

up molecular weight through autoxidative curing of the pendant fatty acid chains in the

alkyd and the tung oil. The tung oil has more reactive sites per chain and a lower molecular

weight, meaning that as it begins to crosslink, it can build up molecular weight quickly.

73 However, at high loadings of the reactive diluent, this could create a situation where the

film begins to have more triglyceride character than polyester, leading to diminished tensile

properties and crosslink density.11,165

The coating properties, however, show results dissimilar to what was expected for

these systems. At high loadings of the RDs, it was found that the pull-off adhesion and

pencil hardness were significantly improved. The pull-off adhesions of STRD-10, -15, and

-20 are higher than any other sample tested, including the modified alkyd (ST) series.

Additionally, the pencil hardness was found to increase as a function of RD content. This could be due to the increase in concentration of fluorine at the surface as the RD content increases.110,117,119,148 Comparing this to the modified alkyd series, the pencil hardness

shows no significant change from the control. At high loadings of the RDs (STRD-15 and

-20), the pencil hardness is higher than for any of the samples in the modified alkyd series.

With the increased adhesion from higher loadings of RDs, the corrosion resistance

should also improve. From the EIS measurements, it was found that at high loading of RDs,

the corrosion resistance was at least as good as or better than that of the modified alkyd

series. STRD-10 performed better than ST-10 over 10 weeks of exposure. The addition of

either fluorine or alkoxysilane functionalized additives has previously been shown to

increase corrosion resistance in coatings.124,147,149 The other RD samples had significantly diminished performance when compared to the control and the modified alkyd series. The modified alkyd series showed better corrosion resistance overall, but at high loadings of the RDs, this performance could be matched or slightly exceeded.

Because of these significant improvements in adhesion, hardness and corrosion resistance, it was expected that at high loadings of the RDs stratification was occurring

74 bidirectionally. As aforementioned, the FTO would migrate to the air/coating interface due to the low surface energy, and the ASTO would become trapped at the substrate surface, following reaction with free hydroxyl groups present there. The increased hardness, adhesion and corrosion resistance found at higher loadings of these RDs led to these assumptions, which were confirmed by ESEM-EDX spectroscopy.

The RD series was expected to show stratification at loadings about 10 wt.% of total RDs, but not below that level, due to the coating properties and corrosion resistance that was found. For STRD-5, an increase in both fluorine (48 %) and alkoxysilane (28 %) concentrations was found at the air/coating interface. At loadings above 5 wt.% of each component, bidirectional stratification was observed, and the alkoxysilane concentration was found to increase at the coating/substrate interface, while the fluorine concentration remained elevated at the air/coating interface. This implies that with a higher concentration of ASTO, the reaction between the alkoxysilane and the free hydroxyl groups at substrate occurs,111,154 creating a slight increase in concentration of ASTO at the coating/substrate interface. However, this concentration at the substrate interface remains lower than that of fluorine at the surface due to the added crosslinking reactions that ASTO can undergo during curing. Reaction between alkoxysilane groups leads to a larger proportion of ASTO remaining in the bulk of the coatings, rather than at either interface. In addition, the low surface energy of the alkoxysilane component still drives it to the air/coating interface.171,172 At low loadings, where the ASTO does not react as significantly at the substrate, it is driven to the surface with FTO. The tung oil components are miscible in the base alkyd, however, they are more miscible with each other than with the alkyd. For these reasons, the concentration of the alkoxysilane at the substrate is less significant than that

75 of fluorine at the surface, and the miscibility also explains the higher concentration of FTO

and ASTO at the air/coating interface when the concentration is too low to initiate reactions

at the substrate. A schematic of the stratification is shown in Figure 3.27

After casting During curing Cured coating

STRD

ST

Alkoxysilane functional RD Alkoxysilane functional alkyd Unmodified alkyd

Fluorine functional RD Fluorine functional alkyd Solvent

Figure 3.27. Schematic of stratifying behavior in reactive diluent series (top) and modified alkyd series (bottom).

Observation of the cross-section of the modified alkyd coating series by ESEM-

EDX showed that it did not result in any significant stratification, as expected. A slightly

higher concentration of fluorine (~30 %) and alkoxysilane (~15 %) were determined to be

at the air/coating interface, as shown in Figure 3.27. These small amounts are driven to the

surface by their low surface energy,110,117,119,148,171,172 but the high molecular weight and

viscosity of these components limited their mobility. Furthermore, the modified alkyds

have the fluorine and alkoxysilane functionality in the backbone, where the polymer is

more rigid and various configurations may be more difficult to achieve during the curing

process. Additionally, as the polymer begins to vitrify, the backbone becomes locked into

a position, where it cannot reorient itself. However, the functionality in the reactive diluents

is on a flexible fatty acid chain, and even if another chain becomes crosslinked into the 76 coating network, it could still have enough mobility to orient it in the proper position. The

modified alkyd series will also reach a vitrified state prior to the reactive diluent series due

to the rigidity and high Tg of the polyester backbones in the alkyd. For these reasons, little stratification is observed in the modified alkyd system.

The increased adhesion and corrosion resistance where bidirectional stratification was observed is further evidence of the reaction between ASTO and the free hydroxyl groups at the substrate surface. The alkoxysilane groups can undergo hydrolysis and condensation reactions with each other and the free hydroxyl groups, forming inorganic domains within the organic network. These domains, especially when involving the substrate surface, would lead to higher pull-off adhesion, as shown here. The increased pencil hardness is also due to the increase in concentration of fluorine at the surface as the

RD content increases.110,117,119,148

While stratification of the RDs provided improvement to coating performance and

corrosion resistance, other properties were not significantly affected by the stratification.

Tensile properties and weatherability were much more dependent on the modified alkyd

versus modified RD characteristics than on the stratification of the coating. The tensile

properties showed that with modified alkyds, particularly higher modulus and tensile

strength could be achieved than with the RDs.

The weatherability of these coatings was also not affected by stratification. The

modified components all showed more gloss retention than the control alkyd, likely due to the fluorinated components.173 Embrittlement was observed from increases in hardness, especially for the modified alkyd series. This series showed a more pronounced increase in pendulum hardness as a function of the loadings of the modified alkyds. This could be

77 due to photo-oxidation reactions in the modified alkyd components, including the

alkoxysilane component, in the high humidity environment of the weatherometer.168–170

The RD series was found to have a slightly larger change in hardness compared to the

control, but it was not as dramatic as for the modified alkyd series. This was likely due to

the lower molecular weight of the RDs, leading to a less dramatic build-up of molecular

weight during the additional crosslinking that occurred in the accelerated weathering

environment.12,174

Overall, the adhesion, hardness and corrosion resistance were significantly affected by the partial bidirectional stratification of the fluorine and alkoxysilane functionalities.

Despite the low molecular weight and high loadings of the RDs, when stratification occurred, these properties were the same or better than all the other coating formulations tested. However, stratification did not have a measurable effect on the tensile properties or weathering of these coatings. The modified alkyd series had significantly better tensile properties. However, weathering caused hardening of the modified alkyd series more so than the RDs. Further work could be done to combine these additives in various ways to tune to the properties to specific performance applications. Here, their utility and performance has been shown systematically.

3.6 Conclusions

Performance and stratification of two coating systems were evaluated with previously reported alkoxysilane and fluorine modified alkyds and reactive diluents. With

loadings of the reactive diluents over 10 wt.% based on the total resin content, stratification

of the fluorine functionalized component toward the air/coating interface was observed, as

well as a slight increase in alkoxysilane functionality at the coating/substrate interface. 78 Increased adhesion and corrosion resistance for these partially bidirectionally stratified

systems evidences that the alkoxysilane components are undergoing reactions with free

hydroxyl groups at the substrate surface. The pencil hardness was also shown to increase

significantly in systems that showed bidirectional stratification. Stratification, however, was not found to have any significant impacts on tensile properties or weatherability. The modified alkyd series, which did not show signs of stratification, had improved tensile properties and corrosion resistance when compared to the control alkyd and the RD samples at lower loadings, where stratification did not occur bidirectionally. Overall, the coating performance was significantly improved with the addition of only 10 wt.% of the

RDs.

79 CHAPTER IV

EFFECT OF HUMIDITY ON ALKOXYSILANE FUNCTIONAL ALKYD COATINGS

4.1 Abstract

The effect of relative humidity on the curing of alkoxysilane functional alkyds and

reactive diluents was evaluated. Alkoxysilane functional alkyds and tung oil-based reactive

diluents were synthesized and used as additives in varying amounts in alkyd coatings.

Looking at modified alkyd versus modified reactive diluent allows for the comparison of

the humidity effect on high and low molecular weight alkoxysilane functional additives.

The coatings were cured at 25 and 75% relative humidity, and evaluated for drying time,

coating and tensile properties and gel content. It was found that at higher humidity, the

alkoxysilane containing samples improved properties significantly when compared to the

control alkyd.

4.2 Introduction

Concerns of depleting petroleum feedstocks, the environment and cost fluctuations

has created interest in the use of renewable resources, and plant-derived polymer materials.

In the coatings field, alkyds have seen a resurgence with the desire to remove the largely

petrochemical based polymer materials.1,15,50 Alkyds are made from seed oils, and also

boast a relatively low cost, high performance solution. They can be formulated as high solids coatings, using little to no volatile organic components (VOCs) to achieve

80 application viscosity.1,4,5 Alkyds have the largest market share of bio-based resins in the

coatings industry, but they are limited by their durability.3

Inorganic/organic hybrid alkyd coatings have been explored as a way to increase the durability of alkyds.32,83,175 These hybrid systems have been extensively studied for a

variety of other resins,73,176,177 including polyimide,178 epoxide,85,89 polyurethane,82 and

polyester.179 The materials are designed using sol-gel chemistry, an approach where the inorganic components undergo hydrolysis and condensation reactions to form a crosslinked gel product within the network of organic crosslinks. The water content has significant effects on the final products as well. At low water content, the inorganic component forms linear chains with alkoxyl groups present, and these are less susceptible to further hydrolysis due to steric hindrance from the alkyl groups. If the water content is sufficient, then the inorganic component will completely hydrolyze, and the linear species formed initially will entangle and form a gel.74,76–78,80,180,181

Ni et al.82 showed the use of inorganic/organic hybrids when developing a moisture- curing system with alkoxysilane-functionalized isocyanurate coatings. The incorporation of this chemistry led to improvements in many properties, including adhesion, crosslink density, glass transition temperature, and flexibility. Soucek and coworkers32,82–89 have

reported several novel inorganic/organic hybrid coating systems based on drying oils and

alkyds. The metal alkoxide forms the inorganic network through hydrolysis and

condensation reactions, while the organic network is formed through the autoxidation

reaction of the fatty acids in the drying oils or alkyds. In epoxidized soybean oil-based

inorganic/organic hybrid coatings, the hardness, tensile strength and flexibility were

81 increased, while the adhesion and impact resistance decreased with increased sol-gel

precursor content.85

When alkoxysilane-modified resins are exposed to humidity or atmospheric moisture, they begin to hydrolyze and form sol-gel networks. This can be used as a primary or secondary curing mechanism in coatings systems with applications in high humidity environments. Silane end-capped soybean oil polyurethanes were synthesized by

Baghdachi et al.,90 and their curing was studied under differing relative humidity and

temperature. It was found that increased relative humidity promoted faster crosslinking,

and thus tack-free time of the coating, while lower relative humidity severely retarded these

crosslinking reactions. Chang et al.91 patented an acrylic-silane coating with curing time accelerated by increased humidity. Emmerling et al.92 shows an alkoxysilane-terminated polyurethane with moisture-curing capabilities. The effect of an acid-catalyst on the curing of polyurea/polysiloxane organic/inorganic hybrid coatings at high humidity was shown by Ni et al.82,93

Herein, a study is completed to look at the effect of humidity on the curing of

inorganic/organic hybrid coatings with alkoxysilane functional alkyds or reactive diluents

(RDs). The use of a higher molecular weight functional alkyd was compared to that of the

lower molecular weight reactive diluent to see if the mobility of the alkoxysilanes would

affect the performance of these materials. Additionally, the alkoxysilane functional alkyd

could be used in much higher amounts than the RD with the weight percent being varied

from 20 to 100 wt.% based on total resin content. The weight percent of the alkoxysilane

functional RD was varied from 10 to 50 wt.%. The films were cured at 25 and 75% relative

82 humidity (RH). The materials were then analyzed for coating performance, tensile

properties and gel content.

4.3 Experimental

4.3.1 Materials

Supreme linseed oil was provided by Cargill. Linseed oil fatty acids were provided

by Alnor Oil Company. Glycerol, phthalic anhydride, xylene, lithium hydroxide,

potassium hydroxide, methanol, acetone, acetic anhydride, pyridine, 3-(triethoxysilyl)-

propyl isocyanate, dibutyltin dilaurate, tung oil, and phenothiazine were all obtained from

Millipore Sigma. Methacryloxypropyl trimethoxysilane was obtained from Gelest. BYK

333 was obtained from BYK. Borschi® OXY-coat, 12% zirconium Hex-Cem®, and 5%

Calcium Hex-Cem® driers were obtained from OM Group. AQ-36 aluminum panels were purchased from the Q-Lab Corp. All materials were used as received.

4.3.2 Instrumentation

All reactions were performed under a nitrogen atmosphere unless otherwise noted.

1H-NMR (300 MHz) spectra (�, ppm) were obtained using a Varian Mercury 300 spectrometer in CDCl3 solvent and analyzed with ACD/NMR Processor software. FT-IR spectra were recorded using a Nicolet iS50 spectrometer using an attenuated total reflection

(ATR) diamond attachment and OMNIC software was use for analysis. MALDI-ToF-MS spectra were recorded on a Bruker Ultra-Flex III MALDI-ToF/ToF mass spectrometer

(Bruker, Billerica, MA) equipped with a Nd:YAG laser emitting at 355 nm. The instrument was operated in positive ion mode. Samples were dissolved in THF to a final concentration of 10 mg mL-1. Trans-2-[3-(4-tert-butylphenyl)-2-methyl-2-propenylidene] malononitrile 83 (DCTB) (20 mg mL-1) served as matrix and sodium trifluoroacetate (NaTFA) (10 mg mL-

1) as cationizing agent. The latter two were prepared and mixed in the ratio 10:1 (v/v), respectively. Matrix:salt and sample solutions were applied onto the MALDI-ToF-MS target plate using sandwich method. Bruker's FlexAnalysis software was used for analysis.

Viscosity measurements were recorded with a Brookfield LV-II+ Pro Viscometer using

Rheocalc V2.6 software for data analysis. The humidity was kept constant using a

Memmert HPP 110 humidity chamber. A PTC Instruments 24-Hour Timer was used for measuring drying time of the coatings in the humidity chamber. Glass transition temperature was evaluated using TA Instruments DSC Q2000. Tensile testing was performed on an Instron 5567 tensiometer with a 100 N load cell.

4.3.3 Alkyd Resin Synthesis (LLOA)

The linseed long oil alkyd was prepared using a standard monoglyceride process.

Linseed oil (170.05 g, 0.19 mol) and glycerol (40.13 g, 0.44 mol) were added to a 500-mL round bottomed flask equipped with an overhead mechanical stirrer, N2 inlet purge through

the vapor space, and a water-cooled condenser. The reaction was heated to 120 ºC at which point, LiOH (0.29 g, 12.11 mmol) was introduced. The reaction temperature was increased to 240 ºC whereupon it was held for ~1 h. An aliquot was taken, cooled to room temperature, and diluted with methanol (3:1, MeOH:resin). This step was repeated until a clear solution was obtained, signifying the formation of monoglyceride.

The reaction was subsequently cooled to 100 ºC and a Dean-Stark trap, filled with xylenes, was introduced to the reactor. Phthalic anhydride (80.13 g, 0.541 mol) was then added to the reaction, slowly, and the mixture was slowly heated to 220 ºC. Reaction progress was monitored via acid number calculations and continued until an acid number

84 of < 20 mg KOH/ g resin was obtained. The alkyd resin was then poured into a glass jar

1 and stored under an atmosphere of N2. H NMR: 0.85-1.00 (-CH3), 1.25-1.30 (-CH2-), 1.60

(-CH2-CH2-CO-O-), 2.01-2.09 (-CH2-CH=), 2.31-2.37 (-CH2-CO-O-), 2.78-2.80 (=CH-

CH2-CH=), 4.14-4.47 (-CO-O-CH2-), 5.35 (-CO-O-CH- and –CH=CH-), 7.53-7.70 (ArH).

4.3.4 Alkoxysilane-Modified Alkyd Resin Synthesis (ASLOA)

The synthesis and characterization of this modified alkyd is detailed in Section

3.3.4.

4.3.5 Alkoxysilane-Modified Tung Oil Synthesis (ASTO)

The synthesis and characterization of this modified tung oil is detailed in Section

3.3.6.

4.3.6 Viscosity Measurements

A Brookfield LV-II+ Pro Viscometer with a SC4-25 spindle was used for viscosity

analysis. The spindle was rotated at a speed of 10 rpm with a shear rate of 2.20 s-1.

4.3.7 Formulations and Film Preparation

Coating formulations were prepared as shown in Table 4.1 and Table 4.2. All

coatings contain a drier package made with 1 part Borschi® OXY-coat, 17 parts zirconium

Hex-Cem, and 2 parts calcium Hex-Cem, which was added in 2 wt.% based on the total

resin content. BYK 333 surface modifier was added at 1.5 wt.% based on resin content.

The amounts of driers and surface modifier were chosen based on ladder studies. The

formulations were cast on aluminum AQ-36 panels, using a drawdown bar at a 2 mil wet

thickness. The coatings were also cast on glass panels at 4 mils for tensile testing. The

85 panels were allowed to cure for 10 days at room temperature in varying relative humidity:

25 and 75 %RH.

Table 4.1. Formulations with alkoxysilane modified alkyd (ASLOA). All amounts are in g. LLOA- ASLOA- ASLOA- ASLOA- ASLOA- ASLOA- Formulations: 100 20 40 60 80 100 LLOA 10 8 6 4 2 0 ASLOA 0 2 4 6 8 10 Acetone 0.5 0.5 0.5 0.5 0.5 0.5 BYK 333 0.15 0.15 0.15 0.15 0.15 0.15 Driers 0.2 0.2 0.2 0.2 0.2 0.2

Table 4.2. Formulations with alkoxysilane modified tung oil (ASTO). All amounts are in g. Formulations ASTO-10 ASTO-20 ASTO-30 ASTO-40 ASTO-50 LLOA 9 8 7 6 5 ASTO 1 2 3 4 5 Acetone 0.5 0.5 0.5 0.5 0.5 BYK 333 0.15 0.15 0.15 0.15 0.15 Driers 0.2 0.2 0.2 0.2 0.2

4.3.8 Coating Properties

Coatings tests were carried out according to standards; pendulum (ASTM D 4366-

16)156 and pencil hardness (ASTM D 3363-05),157 and pull-off adhesion (ASTM D 4541-

17)159 were all analyzed. At least five replicates were completed for each test, except in cases where the standard mentions a specific failure mode (such as, pencil hardness), and an average was calculated with a 95 % confidence interval.

4.3.9 Tensile Testing

An Instron Tensiometer 5567 with hydraulic clamps was used for tensile testing with a 100 N load cell. Samples were measured with a micrometer and were about 25 x 10 x 0.06 mm for each sample. For each test, the rate of extension was 10 mm min-1.

Evaluation of tensile strength, strain at break and Young’s modulus was carried out.

Modulus was determined by the instantaneous modulus at a strain of 0.1 mm mm-1. Each 86 sample was tested a minimum of five times, and the average was taken with a 95 %

confidence interval.

4.3.10 Thermal Properties

Glass transition temperatures were evaluated with a TA Instruments Q2000 DSC.

Aluminum hermetic pans were used for the testing. The temperature was ramped to 100 ºC

at a rate of 10 ºC min-1 and held for 5 min. Following that, it is cooled to -80 ºC, and then

is ramped to 100 ºC at the same rate. The glass transition temperature is analyzed from the

last cycle, using TA Universal Analysis software.

4.3.11 Soxhlet Extraction

Soxhlet extractions were completed to evaluate gel content. Acetone was used as

the solvent, and the experiment was carried out over 24 h. A vacuum oven at 50 ºC was

used to dry the samples overnight before the final weight was analyzed. The samples

initially weighed about 0.1 g. At least three replicates were taken for each sample, and the

average was taken with a 95 % confidence interval.

4.4 Results

The objective of this study was to look at the effect of curing in different relative

humidity on the overall properties of alkoxysilane functionalized coatings. The

functionalizations shown here were previously reported by Salata et al.149 and

Wutticharoenwong et al.10,11 Alkoxysilanes are known to react with water and free hydroxyl groups,81 so this study was designed to see if coatings containing various amounts of alkoxysilane functional alkyds or reactive diluents showed improved coating properties when cured in high humidity conditions. The alkoxysilane functional alkyds (ASLOA) 87 were formulated from 0 to 100 wt.% using an unmodified long oil alkyd (LLOA) as the

base resin. The reactive diluent series was formulated with alkoxysilane functional RDs

(ASTO) in varying amounts from 10 to 50 wt.% based on total resin content. These coatings were cured at 25 and 75 % RH, and were evaluated for coating performance, tensile properties and gel content. It was found that at higher humidity, the alkoxysilane containing samples improved properties significantly when compared to the control alkyd.

4.4.1 Synthesis and Characterization

The synthesis and characterization are described in detail in Section 3.4.1, where

the same ASLOA and ASTO components were used in the study on stratification. The

viscosity of each component was also tabulated in Table 3.3.

4.4.2 Film Formation

Coatings were formulated with LLOA as the base alkyd and ASLOA content

ranging from 0 to 100 wt.%, in 20 wt.% increments. A second set of RD coating

formulations were made with up to 50 wt.% of ASTO, in increments of 10 wt.%. These

coatings were cast and cured at 25 or 75 %RH. To observe the effect of humidity on the

curing of these coatings, drying times were measured, and the results can be seen in Table

4.3. As expected, the drying of the unmodified alkyd was found to be faster at 25 %RH, indicating that the autoxidation reactions are inhibited at high humidity. This has been previously found for various autoxidation reactions, where high humidity is thought to decreased the oxygen permeability, and therefore, the curing reactions are slightly retarded by the high moisture content.182,183 With the addition of ASTO, it was found that the drying time did not change significantly in 25 %RH conditions, likely due to the curing mechanism being dominated by autoxidation. At 75 %RH, the drying of ASTO-containing 88 samples is significantly faster than the LLOA-100 control. The addition of only 10 wt.% of ASTO decreased the drying time by 50 %.

This trend is also observed with low ASLOA content. At 20 wt.%, the drying time is again significantly decreased from the control, and from the same sample cured at 25

%RH. However, at higher ASLOA content, the drying time was unchanged by the varying humidity conditions. Comparing ASLOA-60 to the LLOA-100 control, the drying time at either humidity is slower with the addition of ASLOA. However, because there is more

ASLOA content than LLOA content, it should be compared to ASLOA-100 as the control.

ASLOA-60 and ASLOA-100 show no change in their curing time in the various humidity conditions. It could be possible that in these samples, there is a balance reached between the retardation of the autoxidative reactions and the secondary moisture curing, causing an equilibrium between the drying times at these humidity conditions.

Table 4.3. Drying times of samples cured in different relative humidity. Tack-Free Tack-Free Time Time 25 %RH 75 %RH (h) (h) LLOA-100 11 ± 1 14 ± 1 ASTO-10 10 ± 1 7 ± 1 ASTO-30 8 ± 1 10 ± 1 ASLOA-20 13 ± 1 8 ± 1 ASLOA-60 16 ± 1 16 ± 1 ASLOA-100 16 ± 1 17 ± 1

4.4.3 Coating Properties

Following curing under different humidity conditions, the coated panels were tested

using standard methods. Table 4.4 shows the results for the 25 %RH samples. At high

loadings of ASTO, the properties were diminished, likely due to the high content of lower

89 molecular weight RD.11,165 The ASLOA series does not show significant changes in

hardness as the alkoxysilane functional content is increased. At this low relative humidity,

it is unlikely that significant moisture curing reactions are occurring with the alkoxysilane.

Therefore, the crosslinking in these systems is mainly based on the autoxidative reactions

of the components. However, the pull-off adhesion was found to decrease, especially at

high loadings of ASLOA. If compared to the ASLOA-100 control for this series, it can be

seen that at low humidity, the control alkyd LLOA-100 is contributing more effectively to the pull-off adhesion. A similar result in adhesion for inorganic/organic systems was shown by Teng et al.,85,86 where a slight decrease in adhesion was found as the inorganic

component was increased.

Table 4.4. Coating properties of samples cured at 25 %RH. Pendulum Pencil Pull-Off Adhesion 25 %RH Hardness Hardness (lb in-2) (s) LLOA-100 20 ± 2 B 220 ± 50 ASTO-10 18 ± 3 F 250 ± 40 ASTO-20 18 ± 2 2H 250 ± 40 ASTO-30 18 ± 2 H 180 ± 30 ASTO-40 11 ± 4 F 100 ± 10 ASTO-50 11 ± 1 3B 100 ± 10 Pendulum Pencil Pull-Off Adhesion 25 %RH Hardness Hardness (lb in-2) (s) LLOA-100 20 ± 2 B 220 ± 50 ASLOA-20 17 ± 2 HB 220 ± 30 ASLOA-40 18 ± 2 H 210 ± 40 ASLOA-60 19 ± 1 2H 190 ± 20 ASLOA-80 22 ± 3 4H 160 ± 50 ASLOA-100 22 ± 2 3H 110 ± 10

The coatings properties for the samples cured at 75 %RH are shown in Table 4.5.

It was found that hardness increased with increasing ASTO content, except for at 50 wt.%,

90 where properties began to diminish. The pull-off adhesion of the ASTO series was found

to remain the same, up to 30 wt.%, after which the properties were found to decline. The

ASLOA series also showed hardness and pull-off adhesion increasing as a function of

alkoxysilane content.

Table 4.5. Coating properties of samples cured at 75 %RH. Pendulum Pencil Pull-Off Adhesion 75 %RH Hardness Hardness (lb in-2) (s) LLOA-100 8.2 ± 0.8 2H 240 ± 30 ASTO-10 11.8 ± 0.8 F 260 ± 30 ASTO-20 14.2 ± 0.7 H 210 ± 10 ASTO-30 16.9 ± 0.7 2H 210 ± 40 ASTO-40 28 ± 3 6B 100 ± 10 ASTO-50 12 ± 2 6B 80 ± 30 Pendulum Pencil Pull-Off Adhesion 75 %RH Hardness Hardness (lb in-2) (s) LLOA-100 8.2 ± 0.8 H 240 ± 30 ASLOA-20 9.7 ± 0.9 F 280 ± 30 ASLOA-40 11 ± 1 2H 280 ± 30 ASLOA-60 10.6 ± 0.7 2H 240 ± 60 ASLOA-80 12 ± 1 2H 290 ± 10 ASLOA-100 16 ± 1 2H 270 ± 30

It was expected that properties of the alkoxysilane-containing samples cured at 75

%RH would have significantly improved properties over those cured at 25 %RH. However,

the hardness was found to be slightly diminished for the 75 %RH samples. Despite that, if

the percent change compared to the control alkyds (LLOA-100) are evaluated for these

various humidity conditions, it is found that those cured at higher humidity offer a much

greater improvement over the control, as shown in Figure 4.1. This difference would imply that the moisture curing is having a significant effect on the overall hardness of the coatings at higher humidity, but the retardation of the autoxidative crosslinking at this humidity is

91 causing the hardness to diminish when compared directly to the 25 %RH samples. The

adhesion of the ASTO series did not change significantly between the various curing

conditions, but the ASLOA series showed significant improvement at higher humidity

curing. At high ASLOA content, the adhesion was much higher when cured at 75 %RH

versus 25 %RH, as shown in Figure 4.2.

Pendulum Hardness, ASTO Pendulum Hardness, ASLOA 200 200

150 150

100 100

50 50 Percent Change, % Change, Percent 0 % Change, Percent 0

-50 -50 ASTO-10 ASTO-20 ASTO-30 ASTO-40 ASTO-50 ASLOA-20 ASLOA-40 ASLOA-60 ASLOA-80

25 %RH 75 %RH 25 %RH 75 %RH

Figure 4.1. Percent change in pendulum hardness when compared to the control alkyd. ASTO series (left) and ASLOA series (right).

Pull-Off Adhesion, ASTO Pull-Off Adhesion, ASLOA 350 350 2 2 - - 300 300

250 250

200 200

150 150 Off Adhesion, lb in Off Adhesion, Off Adhesion, lb in Off Adhesion, 100 100 - - Pull Pull 50 50

0 0 Control ASTO-10 ASTO-20 ASTO-30 ASTO-40 ASTO-50 Control ASLOA-20 ASLOA-40 ASLOA-60 ASLOA-80 ASLOA-100

25 %RH 75 %RH 25 %RH 75 %RH

Figure 4.2. Pull-off adhesion for ASTO (left) and ASLOA series (right).

4.4.4 Tensile Properties

To better understand the mechanical properties provided by the alkoxysilane

functional components, tensile properties were evaluated. It was expected that curing at

higher humidity would result in more crosslinking of the alkoxysilane components, which 92 have moisture curing capabilities. It was also expected that higher alkoxysilane

functionality in either system would improve the overall tensile properties.84,86,149 The

results for samples cured at 25 %RH can be seen in Figure 4.3, and those at 75 %RH are in Figure 4.4. It was found that increasing additive content led to less flexible and more mechanically robust films, regardless of curing conditions. The strain at break was found to decrease with increased alkoxysilane content for both ASLOA and ASTO whereas the modulus and tensile strength were increased. The samples with high proportions of ASTO were too soft to create adequate tensile samples, which is likely due to the very high content of RD present in these samples.

Comparing the tensile properties of the various curing conditions, it was found that there was a slight decrease in performance at the 75 %RH curing condition. However, recalling the trend in coating hardness, the modulus and tensile strength were compared as a percent change, based on the control alkyd, LLOA-100. The results are shown in Figure

4.5. There was significant improvement for the samples cured at 75 %RH in the modulus and tensile strength when compared to the control, especially at high loadings of the alkoxysilane component. Additionally, tensile strength of ASLOA-80 and -100 cured at 75

%RH are found to be significantly higher than their counterparts cured at 25 %RH when compared directly. These results indicate that the secondary moisture curing contributes significantly to the mechanical properties of the films when cured in high humidity conditions.

93 Strain at Break, ASTO, 25 %RH Strain at Break, ASLOA, 25%RH 0.9 0.9 0.8 0.8 0.7 0.7

0.6 0.6 0.5 0.5 0.4 0.4

0.3 0.3 0.2 0.2 Strain at Break, mm/mm at Break, Strain mm/mm at Break, Strain 0.1 0.1 0 0 LLOA-100 ASTO-10 ASTO-20 ASTO-30 LLOA-100 ASLOA-20 ASLOA-40 ASLOA-60 ASLOA-80 ASLOA-100

Modulus, ASTO, 25 %RH Modulus, ASLOA, 25 %RH

5 5

4 4

3 3

2 2 Modulus, MPa Modulus, MPa Modulus,

1 1

0 0 LLOA-100 ASTO-10 ASTO-20 ASTO-30 LLOA-100 ASLOA-20 ASLOA-40 ASLOA-60 ASLOA-80 ASLOA-100

Tensile Strength, ASTO, 25 %RH Tensile Strength, ASLOA, 25 %RH 2 2 1.8 1.8 1.6 1.6 1.4 1.4 1.2 1.2 1 1 0.8 0.8 0.6 0.6 Tensile Strength,MPa Tensile Tensile Strength, MPa Strength, Tensile 0.4 0.4 0.2 0.2 0 0 LLOA-100 ASTO-10 ASTO-20 ASTO-30 LLOA-100 ASLOA-20 ASLOA-40 ASLOA-60 ASLOA-80 ASLOA-100

Figure 4.3. Tensile properties of the ASTO (left) and ASLOA (right) series, cured at 25 %RH.

94 Strain at Break, ASTO, 75 %RH Strain at Break, ASLOA, 75 %RH

1.2 1.2

1 1

0.8 0.8

0.6 0.6

0.4 0.4 Strain at Break, mm/mm at Break, Strain Strain at Break, mm/mm at Break, Strain 0.2 0.2

0 0 LLOA-100 ASTO-10 ASTO-20 ASTO-30 LLOA-100 ASLOA-20 ASLOA-40 ASLOA-60 ASLOA-80 ASLOA-100

Modulus, ASTO, 75 %RH Modulus, ASLOA, 75 %RH

4 4 3.5 3.5 3 3 2.5 2.5 2 2 Mosulus, MPa Mosulus,

1.5 MPa Mosulus, 1.5

1 1 0.5 0.5 0 0 LLOA-100 ASTO-10 ASTO-20 ASTO-30 LLOA-100 ASLOA-20 ASLOA-40 ASLOA-60 ASLOA-80 ASLOA-100

Tensile Strength, ASTO, 75 %RH Tensile Strength, ASLOA, 75 %RH

2.5 2.5

2 2

1.5 1.5

1 1 Tensile Strength, MPa Strength, Tensile Tensile Strength, MPa Strength, Tensile 0.5 0.5

0 0 LLOA-100 ASTO-10 ASTO-20 ASTO-30 LLOA-100 ASLOA-20 ASLOA-40 ASLOA-60 ASLOA-80 ASLOA-100

Figure 4.4. Tensile properties of the ASTO (left) and ASLOA (right) series, cured at 75 %RH.

Modulus, % Change Tensile Strength, % Change 600 600 100 100 - 500 - 500

400 400

300 300

200 200

100 100

0 0 Percent Change,based on LLOA Change,based Percent

Percent Change, based on LLOA based Change, Percent ASTO-10 ASTO-20 ASTO-30 ASLOA-20 ASLOA-40 ASLOA-60 ASLOA-80 ASTO-10 ASTO-20 ASTO-30 ASLOA-20 ASLOA-40 ASLOA-60 ASLOA-80

25 %RH 75 %RH 25 %RH 75 %RH

Figure 4.5. Percent change in modulus (left) and tensile strength (right), compared to the control sample. 95 4.4.5 Thermal Properties

The glass transition temperature is known to increase with higher crosslink

density.184–189 As evaluated by DSC, the glass transition temperatures are shown in Table

4.6. It was found that when cured at higher humidity, samples containing alkoxysilane

’ additives had increased Tg s, except at low alkoxysilane content, where at 10 wt.% of

ASTO and 20 wt.% of ASLOA, the Tg still decreased slightly at higher humidity. However, it was still increased over the control alkyd cured under at the same conditions.

Table 4.6. Glass transition temperature of the samples cured at varying relative humidity.

Sample Tg (25 %RH) Tg (75 %RH) LLOA-100 0.8 ± 0.1 -5.9 ± 0.6 ASTO-10 -2.8 ± 0.3 -3.8 ± 0.4 ASTO-30 -9.2 ± 0.9 -5.8 ± 0.6 ASLOA-20 -2.7 ± 0.3 -3.9 ± 0.4 ASLOA-60 -6.2 ± 0.6 11 ± 1 ASLOA-100 -6.0 ± 0.6 -2.0 ± 0.2

4.4.6 Gel Content

The gel content of the coatings is a measure of the crosslink density and the curing

reactions that are taking place. Admittedly, the base alkyd used for these systems does not

have particularly high gel content. However, this helps to more effectively observe the contributions of the additives being used. The results of gel content at 25 %RH curing are shown in Figure 4.6. The addition of ASTO results in a slight increase in gel content after

curing at 25 %RH. The ASLOA series shows a significant increase as a function of ASLOA

content in the samples.

96 Gel Content, ASTO Gel Content, ASLOA 80 80

70 70

60 60

50 50

40 40

30 30 Gel Content, % Content, Gel Gel Content, % Content, Gel 20 20

10 10

0 0 LLOA-100 ASTO-10 ASTO-20 ASTO-30 ASTO-40 ASTO-50 LLOA-100 ASLOA-20 ASLOA-40 ASLOA-60 ASLOA-80 ASLOA-100 Figure 4.6. Gel content results of the ASTO (left) and ASLOA (right) series, cured at 25 %RH.

At 75 %RH, the control alkyd, LLOA-100, has very poor gel content, indicating

poor crosslinking in these conditions. However, with the addition of ASTO, the gel content

increases, except at 40 wt.%, where it drops off to around that of the control. When ASLOA

is added to the system, the gel content steadily increases as a function of ASLOA content.

These results are shown in Figure 4.7.

Gel Content, ASTO Gel Content, ASLOA 80 80

70 70

60 60

50 50

40 40

Gel Content, % Content, Gel 30 % Content, Gel 30

20 20

10 10

0 0 LLOA-100 ASTO-10 ASTO-20 ASTO-30 ASTO-40 ASTO-50 ASTO-10 ASLOA-20 ASLOA-40 ASLOA-60 ASLOA-80 ASLOA-100

Figure 4.7. Gel content of the ASTO (left) and ASLOA series (right), cured at 75 %RH.

Comparing the results of the two systems, it can be seen that the gel contents are

very similar to the control for the 25 %RH samples, whereas there is a significant increase

for the samples cured at 75 %RH. This can be seen in Figure 4.8, where the percent change

from the control for each sample set is plotted. The addition of ASTO accounts for a

97 significant increase in gel content. However, the ASTO-40 does not show any significant change from the control, or the 25 %RH sample. ASLOA provides significant improvements, especially at high

Gel Content, % Change 120 100 - 100

80

60

40

20

0 Percent Change, based on LLOA based Change, Percent

ASTO-10 ASTO-20 ASTO-30 ASTO-40 ASTO-50 ASLOA-20 ASLOA-40 ASLOA-60 ASLOA-80

25 %RH 75 %RH

Figure 4.8. Gel content percent change.

4.5 Discussion

The objective of this work was to look at the effect of humidity on the curing of alkoxysilane functionalized alkyd coatings. It is well-known that silanes can react with water in hydrolysis and condensation reactions to form crosslinked networks.81 These hybrid coatings studied here have a dual-curing mechanism with the autoxidation of the alkyd and the sol-gel process of the alkoxysilane components. This creates an inorganic/organic hybrid coating that can form densely packed inorganic domains within the crosslinked polymer network.73,86,176,190 It was expected that at high humidity curing conditions, coatings with high loadings of alkoxysilane functional components would cure more quickly, have increased gel content and improved properties, especially over control samples cured in these same conditions.

To analyze this hypothesis, a series of alkoxysilane-functional formulations using either modified reactive diluent (ASTO) or modified alkyd (ASLOA) were created. These

98 formulations contained increasing amount of each component, up to 50 wt.% of ASTO and

up to 100 wt.% of ASLOA. These formulations were cured at 25 and 75 %RH, and

following curing, several coating properties were analyzed to better understand the effects

of these conditions on the autoxidative curing process of alkyds and the secondary moisture

curing of the alkoxysilane components. Additionally, the difference between a

functionalized alkyd and RD was explored through these performance properties.

The drying time of these systems was the first clue into the effects of humidity on these curing mechanisms. At 75 %RH, it was found that the control alkyd cured much slower than it did at 25 %RH. This implies that the high humidity has a significant effect on the autoxidative curing mechanism.182,183 However, with the addition of only 10 wt.% of ASTO, the tack-free time at 75 %RH was 50 % improved from the control alkyd and was also faster than the identical sample cured at 25 %RH. The addition of ASTO provided a more significant improvement in drying time when compared to the ASLOA samples.

The modified tung oil has significantly lower viscosity than the modified alkyd, making it more mobile within the bulk of the coating. This mobility gives the alkoxysilane groups a better chance of colliding with each other and any atmospheric moisture at the surface of the coating. Additionally, the location of the alkoxysilane in the rigid alkyd backbone, rather than on a pendant fatty acid chain, leads to even less mobility for the alkoxysilane functional alkyds.

Evaluation of coating properties further indicated the positive effects of the alkoxysilane functional components in high humidity curing. As was the case with the drying time, the alkyd cured at 75 %RH showed significantly lower hardness than that cured at 25 %RH. This softening could be due both to the retardation of the autoxidative

99 curing mechanism and the absorption of water into the film.182,183 Because of this, all of the coatings cured at 75 %RH were slightly softer than those cured at 25 %RH. However, considering the improvement over the control alkyd, those containing alkoxysilane components all showed positive increases in hardness, with up to a 200 % increase over the control. Whereas, in the 25 %RH samples, there was a slight decrease in hardness compared to the control in almost every case. Interestingly, ASTO provided harder coatings than ASLOA, especially under high humidity curing. This result could also be attributed to the mobility of the ASTO within the coating network, which allows it to provide more crosslinking at the surface due to the higher moisture content at the air interface of the coating.

Coatings containing ASLOA resulted in better adhesion to the substrate when cured at 75 %RH. The ASTO-containing samples did not show any changes between the two humidity conditions, but for the ASLOA samples, there was a significant increase in adhesion that was only observed in high humidity curing conditions. At 25 %RH, the adhesion was actually found to decrease with increasing ASLOA content. Teng et al.85,86

also observed a slight decrease in adhesion for inorganic/organic hybrid coatings, but there

is no indication of the humidity conditions during curing. However, when cured at 75

%RH, the adhesion increases significantly as a function of ASLOA content. This could be

due to increased moisture curing reactions as water is absorbed into the film from the high

humidity atmosphere.

The use of alkoxysilane functional additives has previously been shown to improve

tensile properties of the coatings. This is thought to be due to the formation of harder

inorganic segments within the organic framework of the polymer during curing.84,86,149

100 Because of this, it was anticpated that this effect would be observed under both curing

conditions, but that it would be more pronounced under higher humidity curing. The

flexibility was found to decreasing slightly and the modulus and tensile strength increasing

significantly with the addition of more alkoxysilane-functional additives. Additionally, it

was found that the percent changes versus the control sample for both the modulus and

tensile strength were much more significant for those samples cured at 75 %RH. At high

ASLOA content, the tensile strength was much greater than any samples cured at 25 %RH.

ASLOA samples also consistently showed higher modulus and tensile strength than their

ASTO counterparts. This is likely due to the detrimental effects of the low molecular

weight components in the system. It is usually expected that for high loadings of a reactive

diluent, the coating properties begin to diminish due to the lower molecular weight and

higher flexibility of the component, especially for tensile properties.11,165 While the addition of ASTO did not decrease these properties, they were found to be lower than those for the ASLOA series.

Lastly, the glass transition temperature and gel content were evaluated to relate the coating and tensile properties to the overall crosslink density of these systems. Glass transition temperature, which is indicative of crosslink density, was found to increase for alkoxysilane-containing samples cured at 75 %RH. As expected, based on the other performance properties, the control alkyd has a higher Tg and gel content when cured at 25

%RH. All alkoxysilane-containing samples showed increased Tg when cured at 75 %RH,

except for ASTO-10 and ASLOA-20, which had a slight decrease, likely due to the high

proportion of control alkyd present in these samples. However, the gel content of both these

samples at 75 %RH did increase over the control, indicating that more complete curing did

101 occur. This agrees with the other properties tested, which also showed improvement in

ASTO-10 and ASLOA-20 when cured at 75 %RH. The gel content also showed that the with alkoxysilane additives present, the gel content was significantly improved over the control alkyd when curing at 75 %RH.

Overall, it was found that the addition of alkoxysilane functional components provide improvement in properties when cured in high humidity conditions. As an additive,

ASTO provided faster curing and harder coatings following high humidity curing, while

ASLOA provided improved adhesion, tensile properties and crosslink density. The use of a small amount of ASTO or ASLOA in an alkyd system being cured at high humidity can result in faster drying time, and improved hardness, adhesion and tensile properties over the control. Additionally, the use of these additives, regardless of the curing condition, provides significant improvement to modulus and tensile strength, especially at high loadings of ASLOA. However, to see significant improvements over identical samples cured at lower humidity, a large amount of ASLOA is needed, and the ASTO samples never reached improved properties over their counterparts at lower humidity. It is also important to note that above 30 wt.% of ASTO, properties were significantly diminished, and these high RD contents resulted in poor coatings.

4.6 Conclusions

The effect of relative humidity on the curing and properties of alkoxysilane functional alkyd coatings was evaluated. It was found that while high humidity has a detrimental effect on the curing and performance of unmodified alkyds, the use of alkoxysilane functional components in the formulations can provide improved performance at high humidity when compared to the control alkyd. The use of ASTO 102 provided faster curing and harder coatings at high humidity, while ASLOA provided

improvements in adhesion, tensile properties and gel content. The addition of only 10 wt.%

of ASTO provided a 50 % faster curing time at high humidity, as well as an improvement

in hardness, modulus and tensile strength over the control sample. These properties continue to improve at up to 30 wt.% of ASTO in the formulation. Additionally, it was found that the addition of ASLOA provided better adhesion and tensile performance when cured at higher humidity. A high proportion of ASLOA, 80-100 wt.%, was necessary to achieve higher tensile strength and adhesion at 75 %RH curing when compared directly to samples cured at lower humidity.

103 CHAPTER V

COMPARISON OF FLUORINATED ALKYD AND FLUORINATED REACTIVE

DILUENT AS ADDITIVES IN ALKYD COATINGS

5.1 Abstract

A novel fluorine functionalized alkyd was developed using 2,2,3,3,4,4,5,5- octofluoro-1,6-hexanediol as a repeat unit in the alkyd backbone. The fluorinated alkyd was used as an additive for high solids formulations of alkyd coatings. Its performance was compared to that of a previously reported fluorinated tung oil reactive diluent. The coatings were compared to gain a better understanding of the effects of molecular weight and viscosity on the stratification and performance of the coatings. Using a scanning electron microscope with energy dispersive X-ray spectroscopy, the distribution of the fluorinated components in the cross-section of the coating was evaluated. Coatings tests were also performed to evaluate hardness, adhesion and chemical resistance of the coatings, as well as contact angle. Lastly, electrochemical impedance spectroscopy was carried out to evaluate the corrosion protection provided by these additives.

5.2 Introduction

Alkyd resins were once the most widely used coatings,1,2 and they were first discovered by Kienle146 in the 1920s. Alkyds are modified polyesters, which are synthesized from polyhydric alcohol, polybasic acid, and monobasic fatty acids from

104 triglycerides (seed oils).1,3 Alkyds are often formulated with reactive diluents, which are low viscosity additives with the ability to crosslink into the alkyd network during film formation. Many iterations of reactive diluents based on seed oils have been shown previously, as seed oils can be easily modified to add a desired functionality into the coating system.10–14

Modifying seed oils for reactive diluents (RDs) is a common approach to high

solids coatings. Due to their plurality of reactive functional groups, e.g. unsaturation,

hydroxyl groups, ester groups, etc., seed oils can be chemically modified in a variety of

ways.15,16,19,191–193 This creates a plethora of functionalities that can be introduced to high

solids coatings. Fluorine functionalized reactive diluents were shown by

Wutticharoenwong et al.10,11 and Nalawade et al.13,14 These additives improved corrosion

and chemical resistance, as well as surface properties of the alkyd coatings.

Fluorinated alkyd coatings have been shown previously in several studies. Most

often, a fluorinated ester is included in the formulation to provide improved surface

properties.194–196 D. Anton100 showed the use of a simple fluorinated ester with an

unsaturated alkyl chain that could easily crosslink into the alkyd network. This ester also

migrated to the surface of the coating, where it was found to reduce surface energy and

contact angle significantly. Fluorine can also be introduced to the polymer chain through

the use of fluorinated diols197–199 or by introducing perfluoroalkyl functionalized acrylate monomers.200–204

While the fluorine functionality can significantly improve coating performance,

fluoropolymers are often difficult to process and have high cost associate with them, which

severely limits their application potential.94,95 However, using only a small amount of

105 fluorinated material as an additive allows for the improvement of properties without the

obstacles of processibility and prohibitive cost. The low surface energy of fluorinated materials also drives them to migrate toward the air/coating interface.98,110,117,119,148

This work focused on the comparison of stratification and properties of a

fluorinated tung oil reactive diluent and a fluorinated alkyd. The fluorinated alkyd (FLOA)

synthesized for this work utilizes a fluorinated diol to add functionality to the polymer

backbone. The fluorinated tung oil was previously reported by Wutticharoenwong et al.10

and is used here as a RD. 1H-NMR spectroscopy and MALDI MS were used to elucidate

the structure of these components. Using energy dispersive X-ray spectroscopy (EDX), the

cross-section of the coatings was analyzed for the distribution of the fluorinated

components. The coatings were also evaluated for tensile properties, general coating

performance and corrosion resistance.

5.3 Experimental

5.3.1 Materials

Supreme linseed oil was provided by Cargill. Linseed oil fatty acids were provided

by Alnor Oil Company. Glycerol, phthalic anhydride, xylene, lithium hydroxide,

potassium hydroxide, methanol, acetone, 2,2,3,3,4,4,5,5-octafluoro-1,6-hexanediol, tung

oil, phenothiazine, and 2,2,2-trifluoroethyl methacrylate were all obtained from Millipore

Sigma. BYK 333 was obtained from BYK. Borschi® OXY-coat, 12% zirconium Hex-

Cem®, and 5% Calcium Hex-Cem® driers were obtained from OM Group. AQ-36 aluminum panels were purchased from the Q-Lab Corp. All materials were used as received.

106 5.3.2 Instrumentation

All reactions were performed under a nitrogen atmosphere unless otherwise noted.

1H-NMR (300 MHz) spectra (�, ppm) were obtained using a Varian Mercury 300 spectrometer in CDCl3 solvent and analyzed with ACD/NMR Processor. FT-IR spectra

were recorded using a Nicolet iS50 spectrometer using an attenuated total reflection (ATR)

diamond attachment and analyzed using OMNIC software. MALDI-ToF-MS spectra were

recorded on a Bruker Ultra-Flex III MALDI-ToF/ToF mass spectrometer (Bruker,

Billerica, MA) equipped with a Nd:YAG laser emitting at 355 nm. The instrument was operated in positive ion mode. Samples were dissolved in THF to a final concentration of

10 mg mL-1. Trans-2-[3-(4-tert-butylphenyl)-2-methyl-2-propenylidene] malononitrile

(DCTB) (20 mg mL-1) served as matrix and sodium trifluoroacetate (NaTFA) (10 mg mL-

1) as cationizing agent. The latter two were prepared and mixed in the ratio 10:1 (v/v),

respectively. Matrix:salt and sample solutions were applied onto the MALDI-ToF-MS

target plate using sandwich method. Bruker's FlexAnalysis software was used for analysis.

Viscosity measurements were recorded with a Brookfield LV-II+ Pro Viscometer using

Rheocalc V2.6 software for data analysis. Scanning electron microscopy was done using a

FEI Quanta 200 environmental scanning electron microscopy equipped with EDAX

energy-dispersive x-ray spectroscopy. Tensile testing was performed on an Instron 5567

tensiometer with a 100 N load cell. Contact angle was measured on ramé-hart Model 500

Advanced Goniometer. Electrochemical impedance spectroscopy was performed using a

Gamry Reference 600+.

107 5.3.3 Alkyd Resins Synthesis (LLOA)

The alkyd was prepared using the monoglyceride process.24 The ratio of linseed oil, glycerol, and phthalic anhydride content was calculated to achieve a medium oil length alkyd resin.3 Linseed oil (160 g) and glycerol (45 g, 0.49 mol, f=3) were charged in a 4-

neck round bottom flask equipped with a reflux condenser, mechanical stirrer, and gas inlet

and stirred under nitrogen gas for 15 min before the heat was increased to 120 °C. At this

point, where lithium hydroxide (0.3g, 0.013mol) was added and the heat was increased to

240 °C. After 1 h, a 1 mL aliquot was removed, cooled, and mixed with 3 mL of methanol.

This was repeated every 15 min, until the mixed solution became clear, indicating the

formation of monoglyceride.20

The reaction mixture was cooled to 120 °C and a Dean-Stark apparatus was added

in line with the reflux condenser. The apparatus was filled with p-xylene to displace the

condensing water. Phthalic anhydride (96.3 g, 0.65 mol, f=2) was added slowly and the

temperature was increased to 220 °C. The reaction was monitored via acid number by

ASTM D1639-90.21 When an acid number of ~12 was attained after 6 h, the reaction was

1 stopped and cooled. H-NMR: 0.92 (CH3-), 1.27 (-CH2-), 2.03 (-CH2-CH=CH-), 2.29 (-

CH2-CO-O), 2.79 (-CH=CH-CH2-CH=CH-), 4.12 (-CH-CH2-O-), 5.35 (-CH=CH-), 7.54

(Ar-H), 7.74 (Ar-H).

5.3.4 Synthesis of Fluorinated Linseed Oil Alkyd (FLOA)

Linseed oil (85 g, 0.39 mol), glycerol (22.5 g, 0.48 mol), and lithium hydroxide

(0.3 g, 0.1 wt.%) were charged into a 4-neck round bottom flask equipped with a reflux condenser, mechanical stirrer, and gas inlet and stirred under nitrogen gas for 15 min. before increasing the temperature to 240 °C. The formation of monoglyceride was

108 confirmed after 75 min. by dissolving the mixture in three parts methanol to one part

monoglyceride. This mixture was cooled and a Dean-Stark trap was added in line with the

reflux condenser and filled with p-xylene. Phthalic anhydride was charged (38 g, 0.26 mol) slowly, and then the reaction was heated to 220 °C. After 2 hours, 2,2,3,3,4,4,5,5-

octafluoro-1,6-hexanediol (4 g, 0.02 mol) was added. The reaction was continued until an

acid number of 22 was reached. The product was analyzed by 1H-NMR and MALDI MS.

1 H-NMR: 0.92 (CH3-), 1.29 (-CH2-), 2.05 (- CH2-CH=CH-), 2.33 (-CH2-CO-O-), 2.81 (-

CH=CH-CH2-CH=CH-), 3.5 (-CH2-CF2-), 4.12 (-CH-CH2-O-), 5.39 (-CH=CH-), 7.57

(Ar-H), 7.74 (Ar-H).

5.3.5 Synthesis of Fluorinated Tung Oil

The synthesis and characterization of this modified tung oil is detailed in Section

3.3.7.

5.3.6 Formulations and Film Preparation

Ten samples were prepared as shown in Table 5.1, using the unmodified alkyd

(LLOA) as a control. The fluorinated linseed oil alkyd (FLOA) was added in increments

of 10 % of the total resin, up to 40 %. The fluorinated tung oil (FTO) reactive diluent was

added with increments of 2.5 % of total resin, up to 10 %. A ratio of 1:17:2 of Borchi®

OXY to zirconium to calcium drier package was used to eliminate the need for cobalt

driers. Acetone was used as a solvent, between 5 and 15 wt. %. BYK 333 was used as a

surface tension modifier. The amounts of additives and driers were optimized using ladder

studies. The coatings were applied at 2-4 mils to aluminum panels for testing coating

properties, corrosion resistance, and film formation, and to glass panels for testing tensile

properties. 109 Table 5.1. Formulations for FLOA- and FTO-containing alkyd coatings. Component FLOA-0 FLOA-10 FLOA-20 FLOA-30 FLOA-40 FLOA 0 g 1 g 2 g 3 g 4 g MLOA 10 g 9 g 8 g 7 g 6 g Acetone 1.5 g 1.5 g 1.5 g 1.5 g 1.5g Driers 0.2 g 0.2 g 0.2 g 0.2 g 0.2 g BYK 333 0.15 g 0.15 g 0.15 g 0.15 g 0.15 g Component FTO-0 FTO-2.5 FTO-5 FTO-7.5 FTO-10 FTO 0 g 0.25 g 0.5 g 0.75 g 1 g MLOA 10 g 9.75 g 9.5 g 9.25 g 9 g Acetone 0.5 g 0.5 g 0.5 g 0.5 g 0.5g Driers 0.2 g 0.2 g 0.2 g 0.2 g 0.2 g BYK 333 0.15 g 0.15 g 0.15 g 0.15 g 0.15 g 5.3.7 Viscosity Measurements

Viscosity of the alkyds and reactive diluent was measured using a Brookfield LV-

II+ Pro Viscometer. A SC4-25 spindle was used at a rotating speed of 10 rpm, 2.20 s-1

shear rate.

5.3.8 Energy Dispersive X-ray Spectroscopy

Scanning electron microscopy was done using a FEI Quanta 200 environmental

scanning electron microscopy equipped with an EDAX energy-dispersive x-ray spectroscopy. Samples were prepared by peeling from the substrate and breaking under liquid nitrogen. This keeps the cross-section in-tact for analysis. The sample was placed on the SEM stub using carbon tape to hold the cross-section up-right for detection. Dot maps are created by analyzing the fluorine present in the cross-section. The dot maps were divided into quadrants and the amount of fluorine was counted using ImageJ software.

5.3.9 Coating Properties

Coatings tests were done using the following ASTM standards. The pendulum hardness (ASTM D 4366-16),156 pencil hardness (ASTM D 3363-05),157 MEK chemical

110 resistance (ASTM D 5402-15),158 pull-off adhesion (ASTM D 4541-17),159 crosshatch

adhesion (ASTM D 3359-17),160 and Gardner impact (ASTM D 2794-93)161 were measured. Contact angle was evaluated using 2 µL drops of DI water on the coating surface with a ramé-hart Model 500 Advanced Goniometer. Each test was carried out at least 5 times, except in cases where a specific failure mode is specificied (such as, pencil hardness and crosshatch adhesion), and an average was taken with a 95 % confidence interval.

5.3.10 Tensile Properties

Tensile testing was completed using an Instron Tensiometer 5567. A 100 N load cell was used, and the instrument was equipped with hydraulic clamps. Extension was done at a rate of 10 mm min-1. Sample dimensions were generally 25 x 10 x 0.06 mm. Tensile strength, Young’s modulus and strain at break were determined. Each sample was tested at least five times, and the average was calculated with a 95 % confidence interval.

5.3.11 Electrochemical Impedance Spectroscopy

A Gamry Reference 600+ was used for the electrochemical impedance spectroscopy (EIS) with 3.5 wt.% NaCl in DI water held on the coating.

5.4 Results

The modification of linseed oil alkyd with a fluorinated diol was carried out to study the effect of this modified alkyd on coating performance. A fluorinated tung oil was also synthesized based on previous work from Wutticharoenwong et al.10 These two materials were compared for their ability to stratify, coating performance, tensile properties and corrosion resistance. As expected, the lower viscosity fluorinated tung oil showed increased migration toward the surface of the coating, whereas the fluorinated alkyd did 111 not stratify significantly. The coating properties of hardness and chemical resistance were

improved in all cases. However, the fluorinated alkyd coatings outperformed the

fluorinated tung oil in terms of corrosion resistance. Both of these additives have

significant potential for improvement of coating performance, even with little material

added.

5.4.1 Synthesis and Characterization

Fluorinated resins provide corrosion protection and chemical resistance to coatings

by the high chemical and thermal stability and low surface energy provided by the fluorine.

The addition of fluorine into the backbone of the alkyd resin was chosen to avoid disrupting

the diallylic methylene groups in the pendant fatty acid chains that are involved in the

autoxidative crosslinking mechanism of the alkyd resins. The synthesis was carried out by

replacing a small portion of glycerol with fluorinated diol during the step-growth

polymerization of the alkyd. The reaction scheme can be seen in Figure 5.1, a. The resulting fluorinated alkyd (FLOA) was then characterized by 1H-NMR spectroscopy and MADLI-

TOF mass spectrometry. The 1H-NMR spectrum, shown in Figure 5.2, compares the unmodified alkyd with the fluorine-modified alkyd. A new resonance appears around 3.5 ppm that corresponds to the protons adjacent to the fluorinated groups. MALDI-TOF mass spectrometry also showed consistent weight differences, corresponding to the addition of the fluorinated diol as a repeat unit in the polymer (Figure 5.3).

The fluorinated tung oil (FTO) was synthesized by a Diels-Alder reaction with a fluorinated methacrylate and the conjugated double bonds in tung oil, according to the procedure from Wutticharoenwong et al.10 Tung oil was chosen as the seed oil due to the conjugation in the ⍺-eleostearic fatty acid chains and the ability for a quick autoxidative

112

curing process. Using a Diels-Alder reaction is a quick and facile method for addition of a fluorinated methacrylate group, and it can be done without the use of solvent. The reaction scheme is seen in Figure 5.1, b. The 1H-NMR spectrum and MALDI MS can be seen in

Figure 3.13 and Figure 3.14, respectively.

a. O O O O F F F F OH HO O O OH + HO O F F F F O R n 2,2,3,3,4,4,5,5-octofluoro- Alkyd 1,6-hexanediol

O O O O F F F F OH HO O O O O F F F F O R n Fluorinated Alkyd (FLOA)

O b. O F O O F F O Heat, Diels-Alder Reaction O O Phenothiazine O Tung Oil

O F O F O F O

O

O O

O F O O F F Fluorinated Tung Oil (FTO)

Figure 5.1. Reaction scheme for FLOA (a) and FTO (b). Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier.

113 F F F F O O O O OH + HO O O OH HO O F F F F C O n R

F F F F O O O O OH HO O O O O F F F F C O n Unmodified alkyd R Fluorinated alkyd

Figure 5.2. 1H-NMR spectrum of FLOA. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier.

1788.390 Intens. [a.u.]

1566.274 F F F F O O O O OH 6000 + HO O O OH HO O F F F F C O n 2012.503 R Mw= 262 Da

1305.992 O F F F F 4000 1528.099 O O O 2271.783 OH O O HO 2049.678 O 1344.165 O 260 F F2232.608F F 2493.896 C O n R 260

1750.208 1826.567 260 2000 260

2454.714 260 260 1658.138 2533.073 1970.308 1880.259 1436.026 1605.446 260 2140.549 2310.962 260

0

1400 1600 1800 2000 2200 2400 2600 m/z Figure 5.3. MALDI-TOF MS of FLOA. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier. 114 5.4.2 Film Formation and Stratification

To evaluate the properties of these materials, FLOA was used as an additive with

an unmodified alkyd. The FLOA was used in amounts up to 40 wt.%, based on the total

resin content, in increments of 10 wt.%. The FTO was added up to 10 wt.%, in 2.5 wt.%

increments. The difference in the amounts of the additives was chosen for two reasons. It

was expected that the FLOA, being of higher molecular weight and viscosity, would not

stratify as completely. Thus, adding more may help to improve the stratification. The FTO,

having low molecular weight and viscosity, would be able to migrate throughout the bulk

of the coating and lead to stratification. However, this same minimal molecular weight

could lead to detrimental effects on the crosslinking of the alkyd coating.11,165 The functionalized tung oil has less sites for autoxidative curing, and this could create more dangling chains and decrease the hardness or mechanical properties of the coatings, especially at high loadings of the FTO.11 For this reason, the FTO was used in smaller

proportions to avoid diminishing coating properties.

The viscosity of these materials was analyzed to observe the effect of the viscosity

changes on the stratification of the coatings. It is expected that lower viscosity will lead to

more stratification due to the easier mobility of that material within the bulk of the coating.

From the results in Table 5.2, the viscosity of the fluorinated tung oil is almost two orders

of magnitude lower than that of the fluorinated alkyd.

The films were analyzed using EDX over the cross-section of the coating to study

the distribution of the fluorinated components. The fluorinated components are driven to

the air interface of the coating by their low surface energy. As the solvent is evaporating,

the fluorinated components migrate toward the surface and continue until they are trapped

115

by the crosslinking reactions of the alkyd. The detected fluorine atoms are shown in a dot map in Figure 5.4, where a representative sample from each set is shown. The FLOA sample shows a 40 % increase in fluorine in the surface quadrant versus the substrate, and the FTO sample shows a 105 % increase. This indicates that the FTO is able to migrate toward the surface more efficiently than the FLOA, likely due to the differences in viscosity and molecular weight.

Table 5.2. Viscosity of the coating components. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier. Sample Viscosity, cP-s MLOA 37,400 ± 400 FLOA (100%) 10,300 ± 300 FTO (100%) 760 ± 20

5 µm 5 µm

Figure 5.4. SEM-EDX mapping of a cross-section of FLOA-30 (left) and FTO-2.5 (right). Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier.

5.4.3 Tensile Properties

The tensile properties were analyzed to gain a better understanding of the effect of the various components on the mechanical properties of the film. The FLOA is not

116 expected to have a significant effect on these mechanical properties. Because the

fluorinated component is in the alkyd backbone, it does not have a detrimental effect on

the crosslinking mechanism of the alkyd. As seen in Figure 5.5, the strain at break and

tensile strength do not change significantly with the inclusion of FLOA. The modulus,

however, increases as function of FLOA content. This increase in flexibility is due to the

slightly lower molecular weight of the fluorinated alkyd versus the unmodified control

alkyd. The flexibility in the polymer backbone is also from the fluorinated diol.

The FTO samples show a significant drop off in mechanical properties when loaded

over 5 wt.%, as shown in Figure 5.5. This is likely due to the low molecular weight of the

FTO. Similar results were observed from Wutticharoenwong et al.11 The tensile strength, strain at break and modulus all show these same significant drop offs. The tensile strength and modulus both show an increase with the addition of a small amount of FTO, which could be attributed to an increase in flexibility from the lower molecular weight components.

117 2 1.8 1.6 1.4 1.2 1 0.8 0.6 Tensile strength, MPA 0.4 0.2 0 F-0 F-10 F-20 F-30 F-40

1.6

1.4

1.2

1

0.8

0.6

0.4 Strain at break, mm/mm

0.2

0 F-0 F-10 F-20 F-30 F-40

3.5

3

2.5

2

1.5 Modulus, MPa 1

0.5

0 F-0 F-10 F-20 F-30 F-40 Figure 5.5. Tensile strength (top), strain at break (middle) and modulus (bottom) for FLOA (left) and FTO (right) coatings. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier.

5.4.4 Coating Performance

Coating performance was analyzed by a series of general coatings tests from standard methods. The formulations were cast onto aluminum panels and cured over 2 weeks prior to this testing. Table 5.3 and Table 5.4 show the results of this testing for

FLOA and FTO samples. The addition of FLOA was found to improve the pencil hardness,

118 and the contact angle was also found to increase significantly as fluorine content was increased, as shown in Figure 5.6. The chemical resistance was also improved, except for at very high loadings, where FLOA-40 showed a slight decrease in chemical resistance, but it was still higher than the control alkyd.

Table 5.3. Coating test results of FLOA and FTO for hardness, flexibility and contact angle. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier. Pendulum Pencil Direct Reverse Contact Sample Hardness Hardness Impact Impact Angle FLOA-0 8.4 ± 0.8 4B > 40 > 40 81 ± 1 FLOA-10 5.6 ± 0.6 4B > 40 > 40 101.1 ± 0.6 FLOA-20 7.0 ± 0.7 3B > 40 > 40 101.5 ± 0.8 FLOA-30 5.6 ± 0.6 3B > 40 > 40 101.7 ± 0.5 FLOA-40 8.4 ± 0.8 2B > 40 > 40 101.8 ± 0.6 Pendulum Pencil Direct Reverse Contact Sample Hardness Hardness Impact Impact Angle FTO-0 17 ± 2 3H > 40 > 40 94 ± 1 FTO-2.5 17 ± 1 HB > 40 > 40 97 ± 2 FTO-5 14.0 ± 0.9 F > 40 > 40 104.5 ± 0.3 FTO-7.5 14 ± 1 F > 40 > 40 104.6 ± 0.7 FTO-10 17 ± 2 2H > 40 > 40 104.8 ± 0.4

Table 5.4. Coatings properties for FLOA and FTO for chemical resistance and adhesion. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier. Sample MEK Double Rubs Pull-Off Adhesion Crosshatch Adhesion FLOA-0 90 ± 30 220 ± 50 5B FLOA-10 > 200 290 ± 20 5B FLOA-20 > 200 280 ± 30 5B FLOA-30 > 200 280 ± 20 5B FLOA-40 110 ± 10 230 ± 70 5B Sample MEK Double Rubs Pull-Off Adhesion Crosshatch Adhesion FTO-0 110 ± 60 270 ± 40 5B FTO-2.5 > 200 280 ± 30 5B FTO-5 > 200 250 ± 30 5B FTO-7.5 > 200 220 ± 20 5B FTO-10 > 200 240 ± 40 5B

119 Figure 5.6. Contact angle for FLOA coatings. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier.

The FTO samples showed less significant trends with regard to the coating

properties. The hardness was found to decrease slightly, due to the lower molecular weight

of the RD. The chemical resistance, however, is significantly improved with the

incorporation of FTO. The contact angle also increases as a function of FTO concentration.

The other coating properties, including impact resistance, crosshatch adhesion and pull-off

adhesion do not show any significant changes with FTO as an additive.

5.4.5 Electrochemical Impedance Spectroscopy (EIS)

Corrosion resistance was evaluated by studying the barrier properties of the coatings, as measured by EIS. The coatings were exposed to a 3.5 wt.% NaCl salt solution over several weeks, and the impedance modulus was measured several times over this period. The impedance modulus indicates how much resistance is provided by the coating when trying to pass an electrical current between the salt solution and the aluminum substrate. The Bode plots for FLOA coatings after 2 h of exposure and 37 days are shown in Figure 5.7 and Figure 5.8, respectively. Initially, FLOA-20 and FLOA-40 show the highest low frequency impedance, above 2 x 106 Ohms. The control sample was approximately 4 x 105 Ohms, an order of magnitude lower than the modified samples. After

37 days of exposure, the FLOA modified coatings showed effective barrier properties against the corrosive environment. The control, however, showed a significant loss in 120 impedance modulus after exposure. Figure 5.8 shows a trend of decreasing low frequency impedance with decreasing FLOA content at 37 days of exposure. Figure 5.9 and Figure

5.10 compare the Bode plots of impedance for the control to FLOA-20. The control begins

to significantly decrease in impedance modulus after only 3 days of exposure, whereas

FLOA-20 remains relatively high after 37 days.

a.)

1.E+06

FLOA0 FLOA10

1.E+05 FLOA20 FLOA30 FLOA40

1.E+04 Impedance (Ohm)

1.E+03 0.01 0.1 1 10 100 1000 10000 100000 Frequency, 1/s

Figure 5.7. Bode plots for FLOA coatings after 2 h of exposure to 3.5 wt.% NaCl solutions. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier.

1.E+07 a.

1.E+06 FLOA0 FLOA10 FLOA20 1.E+05 FLOA30 FLOA40 Impedance (Ohm)

1.E+04

1.E+03 0.01 0.1 1 10 100 1000 10000 100000 Frequency, 1/s

Figure 5.8. Bode plots for FLOA coatings after 37 days of exposure to 3.5 wt.% NaCl solutions. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier. 121

1.E+06 FLOA0

2 h

24 h 1.E+05 72 h 140 h 240 h 550 h 1.E+04 Impedance (Ohm) Impedance 888 h

1.E+03 0.01 0.1 1 10 100 1000 10000 100000 frequency, 1/s

Figure 5.9. Bode plots comparing the impedance modulus of FLOA0 over 37 days of exposure. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier.

1.E+07 FLOA20

2 h 1.E+06

24 h 72 h 140 h 1.E+05 240 h 550 h

Impedance (Ohm) Impedance 888 h 1.E+04

1.E+03 0.01 0.1 1 10 100 1000 10000 100000 frequency, 1/s

Figure 5.10. Bode plots comparing the impedance modulus of FLOA20 over 37 days of exposure. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier.

The Bode plots for FTO are shown in Figure 5.11 and Figure 5.12 for 2 h and 42 days of exposure, respectively. Initially, all samples have relatively similar low frequency 122 impedance, but the high frequency impedance is much higher for the FTO coatings.

However, following 42 days of exposure, the barrier protection of all coatings was decreased significantly. The FTO coatings did not show the same level of performance as the FLOA coatings. After 42 days of exposure, FTO-2.5 and FTO-5 had a significant decrease in impedance modulus, while the other samples, including the control, showed similar impedance moduli at low frequency. Figure 5.13 and Figure 5.14 show a comparison between the control and FTO-7.5 over the 42 days of the exposure.

1E+06 FTO-0 FTO-2.5 1E+05 FTO-5 FTO-7.5

1E+04 FTO-10 Impedance Impedance (Ohm) 1E+03

1E+02 0.01 0.1 1 10 100 1000 10000 100000 Frequency, 1/s

Figure 5.11. Bode plots of the impedance of FTO coatings, initially. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier.

123 1E+05 FTO-0 FTO-2.5 1E+04 FTO-5 FTO-7.5 FTO-10 1E+03 Impedance Impedance (Ohm) 1E+02

1E+01 0.01 0.1 1 10 100 1000 10000 100000 Frequency, 1/s

Figure 5.12. Bode plots of impedance for the FTO coatings, after 42 days of exposure. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier.

1E+06 FTO-0

2 h 1E+05 170 h 340 h 500 h 1E+04 670 h 1100 h Impedance Impedance (Ohm) 1E+03

1E+02 0.01 0.1 1 10 100 1000 10000 100000 Frequency, 1/s

Figure 5.13. Bode plots of impedance over time for FTO-0. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier. 124 1E+06 FTO-7.5 2 h 1E+05 170 h 340 h 500 h 1E+04 670 h 1100 h Impedance Impedance (Ohm) 1E+03

1E+02 0.01 0.1 1 10 100 1000 10000 100000 Frequency, 1/s

Figure 5.14. Bode plots of impedance over time for FTO-7.5. Reprinted from Salata et al. Prog. Org. Coat. DOI:10.1016/j.porgcoat.2019.04.017 with permission from Elsevier.

5.5 Discussion

The objective of this research was the synthesis and characterization of a newly

developed fluorinated alkyd for use as an additive in alkyd coatings. This fluorinated alkyd

was then compared with a fluorinated tung oil reactive diluent that was previously reported

by Wutticharoenwong et al.10 The two sets of coatings were evaluated for the stratification of the fluorinated components, coating performance, tensile properties and corrosion resistance. The molecular weight and viscosity differences between these two additives should have a profound effect on the properties of the coatings, especially in terms of stratification.

The synthesis of FLOA placed a fluorinated diol in the backbone of the alkyd polymer. This placement of the fluorinated group was chosen to avoid reducing the

125 crosslinking sites present in the alkyd. If a fluorinated group was to be added by grafting

across unsaturated bonds in the pendant fatty acid chains, this would lead to the

consumption of functional crosslinking sites. However, by avoiding this, the fluorine group is now in the rigid backbone of the polymer, rather than the more flexible pendant fatty acid chains. This could be causing even more problems for the stratification process because the fluorinated group will have more difficulty migrating through the bulk of the film due to the rigidity and lower number of configurations available with this polymer architecture.

Comparing the FLOA and FTO structures shows that the tung oil is much more flexible, especially with respect to the position of the fluorinated group. This, in combination with the lower molecular weight and viscosity of the FTO component, will lead to more mobility of the FTO throughout the cross-section of the coating. This was observed in the stratification results, where the FTO showed more than twice the increase in fluorinated content at the surface than was shown for the FLOA coatings. This was further evidenced by the contact angle measurements, which showed that the FTO coatings had higher contact angles than the FLOA coatings, especially at high loadings.

Despite the stratification results of the FTO coatings, some of the properties are shown to diminish with the inclusion of the reactive diluent. These are properties dependent on the crosslink density of the coatings. The tensile properties, in particular, are shown to decrease significantly at high loadings of the RDs, which is likely due to lowered crosslink density from these low molecular weight components. The FLOA coatings show a significant improvement in modulus with increased FLOA content, indicating more flexible coatings.

126 Another important factor in the performance of fluorinated coatings is the resistance to corrosion and improved barrier properties. It was found that the FLOA coatings showed excellent corrosion resistance over the course of the testing. When comparing these to the control alkyd without any fluorine content, it was shown that the FLOA coatings stayed consistently around the same level of impedance, while the control dropped significantly over this timeframe. The FTO coatings, on the other hand, did not all outperform the control. At higher loadings, the impedance modulus stayed about the same as the control, but significant reductions were seen for lower loadings of the RD. This improvement at higher loadings is likely due to the stratification and the hydrophobicity of the fluorinated components at higher loadings.

5.6 Conclusions

A novel fluorinated alkyd was synthesized and characterized for its performance as a coating additive in terms of stratification and coating properties. It was compared to a previously reported fluorinated tung oil reactive diluent to gain a better understanding of the role of molecular weight, viscosity, and placement of the fluorinated group in the overall performance of the coating. It was found that FTO exhibited better stratification, due to the lower viscosity and molecular weight of this component. However, it was found that the tensile properties and corrosion resistance were more significantly improved for

FLOA, indicating that the higher molecular weight may improve performance in these areas. For all the formulations, the increased concentration of fluorine at the surface produced improvement in hardness and chemical resistance without compromising other coating properties. This study indicates that these additives, even used in low

127 concentrations, can provide significant improvement to performance, proving that FLOA or FTO could be used as a low cost, bio-derived additive.

128 CHAPTER VI

MALEATED SOYBEAN OIL DERIVATIVES AS VERSATILE REACTIVE

DILUENTS

6.1 Abstract

Several different bio-based reactive diluents were prepared from soybean oil and their behavior as diluents in long oil alkyds was evaluated. The maleated soybean oil derivatives were prepared in a two-step fashion: (1) Maleation of soybean oil with maleic anhydride via an ene reaction and (2) Nucleophilic acyl substitution of the grafted succinic anhydride group. The structures were characterized using 1H, 13C and 2D HSQC and

HMBC NMR spectroscopy, FT-IR (ATR) spectroscopy, and MALDI-ToF-MS. Brookfield viscosity measurements were conducted to determine their efficiency as diluents. Crosslink density, gel content and tensile properties were evaluated to understand the effect of the diluents on mechanical properties. General coatings tests were also conducted to analyze the performance of the alkyd formulations.

6.2 Introduction

Petroleum-based chemicals are the predominate source of raw materials used in coatings applications. Unfortunately, these depend on depleting fossil fuels and constantly fluctuate in price due to the volatility of the oil and gas industry. Furthermore, the detrimental impact that non-renewable resources have on the environment has led to more

129 stringent regulations in recent years.205,206 This prompted a substantial amount of

investigation on potential replacements for conventionally prepared resins, crosslinkers,

and additives within the coatings industry.50 Currently, plant and animal feedstocks serve

as one of the most viable alternatives. These sources are cheap, abundant, non-toxic, bio-

renewable, and more bio-degradable as well.15,19

One of the most common green chemistry approaches is the utilization of seed oils.

Alkyd resins are seed oil-based polymers that are frequently used in coatings. Discovered in the 1920s, alkyds were once one of the most prominent resin systems, but the high VOC content of these coatings led to a significant decrease in their market share by the 1950s.1–

3 With the more recent push to move away from petrochemical-based materials, however,

alkyds have experienced a resurgence.1,15 Alkyds are also an excellent candidate for high

solids coatings, especially with the use of seed oil-based reactive diluents (RDs). These

RDs decrease the viscosity of the coatings and then crosslink into the film during the curing

process. The compatibility of alkyds with seed oils, and their common crosslinking

mechanism allows for an easy transition to using RDs in place of solvents.1,4,5

Seed oils are commonly used to develop high solids coatings. Due to their plurality of reactive functional groups, e.g. unsaturation, hydroxyl groups, ester groups, etc., seed oils can be chemically modified in a variety of ways.15,16,19,191–193 One of the most

successful examples is acrylated epoxidized soybean oil (AESO; Figure 6.1). AESO is

prepared from the reaction of acrylic acid with epoxidized soybean oil and has found utility

130 in coatings, adhesives, and composites.207–213 It is commercially available under the brand name Ebecryl 860 from UCB Chemicals Company.214

Figure 6.1. Acrylated epoxidized soybean oil.

Previous work on RDs by Soucek et al.10,11,13,14 has focused on functionalizing tung

oil and conjugated soybean oil through a Diels-Alder reaction. Various monomers were

employed as dienophiles with functionalities that included alkoxysilane, triallyl ether, and

fluorine. Alkoxysilane and triallyl ether diluents improved the crosslink density, glass

transition temperature (Tg), and tensile properties; whereas, fluorine functional diluents adjusted surface energy and enhanced the solvent resistance. However, there were some drawbacks to these approaches. Tung oil is much less readily available than soybean oil

(SBO) and tends to yellow quickly due the high degree of conjugated unsaturation. SBO,

contrarily, is the most abundant seed oil available. Unfortunately, it must first be

conjugated in the presence of an expensive, air-sensitive catalyst for the Diels-Alder

functionalization to occur.13

Herein, we describe a facile, solvent-free approach to four different maleated SBO derivatives for use as RDs in alkyd coatings. Maleic anhydride (MA) was first grafted onto

SBO via an ene reaction to afford maleated SBO (sometimes referred to as maleinized

131 SBO, MA-SBO). The resulting grafted succinic anhydride group was subsequently ring- opened to yield monoalkyl half ester-half acid adducts with functional sites capable of participating in the curing process. Optionally, to show the utility of MA-SBO, the half ester-half acid was further reacted with a glycidyl ether compound to introduce additional functionality. The SBO derivatives were characterized using 1H, 13C and 2D HSQC and

HMBC NMR spectroscopy, FT-IR (ATR) spectroscopy, and MALDI-ToF-MS. The diluent efficiency was evaluated for each formulation and compared against a control alkyd and a control series, which consisted of unmodified SBO as the RD. General coatings tests

(according to ASTM standards) were conducted to evaluate the influence these diluents had on the properties of the alkyd coating.

6.3 Experimental

6.3.1 Materials

Maleic anhydride (MA; 99.0%), phenothiazine (≥98.0%), 2-Hydroxyethyl methacrylate (HEMA; 97.0%), trimethylolpropane diallyl ether (TMPDAE; 90.0%), diethylene glycol monobutyl ether (DEGMBE; 99%), allyl glycidyl ether (AGE; ≥99.0%), glycidyl methacrylate (GMA; 97%), pyridine (≥99%) glycerol (≥99.0%), phthalic anhydride (99.0%), Lithium hydroxide monohydrate (LiOH; ≥98.0%), 1,8- diazabicyclo[5.4.0]undec-7-ene (DBU; 98%) and m-xylene (≥ 99.0%) were obtained from

Sigma Aldrich. Hexanes (95%), methanol (≥99.0%) and diethyl ether (≥99.0%) were obtained from Fisher Scientific. Soybean oil (SBO, Technical grade) and supreme linseed oil were kindly provided by Cargill. Borschi® OXY-coat, 12% Zirconium Hex-Cem®, and

5% Calcium Hex-Cem® driers were obtained from OM Group. All reagents and solvents

132 were used as received without further purification. AQ-36 aluminum panels were

purchased from the Q-Lab Corp.

6.3.2 Instrumentation

All reactions were conducted under an atmosphere of N2 using standard Schlenk

line techniques unless otherwise noted. 1H (300 MHz) NMR were recorded on a Varian

Mercury 300 spectrometer, 1H (750 MHz) NMR and 2D HSQC and HMBC experiments were recorded on a Varian INOVA 750 spectrometer, and 13C (500 MHz) NMR were

recorded on a Varian NMRS 500 spectrometer. All experiments were done using CDCl3 in

solvent. Spectra were analyzed with ACD/NMR processor software. Fourier-transform

Infrared (FT-IR) spectroscopy was recorded on a Nicolet iS50 spectrometer using an

attenuated total reflection (ATR) diamond attachment. A total of 16 scans at 4 cm-1

resolution were recorded for each sample. All spectra were analyzed using OMNIC

software. MALDI-ToF-MS spectra were recorded on a Bruker Ultra-Flex III MALDI-

ToF/ToF mass spectrometer (Bruker, Billerica, MA) equipped with a Nd:YAG laser emitting at 355 nm. The instrument was operated in positive ion mode. Samples were dissolved in THF to a final concentration of 10 mg mL-1. Trans-2-[3-(4-tert-butylphenyl)-

2-methyl-2-propenylidene] malononitrile (DCTB) (20 mg mL-1) served as matrix and

sodium trifluoroacetate (NaTFA) (10 mg mL-1) as cationizing agent. The latter two were

prepared and mixed in the ratio 10:1 (v/v), respectively. Matrix:salt and sample solutions

were applied onto the MALDI-ToF-MS target plate using sandwich method. Bruker's

FlexAnalysis software was used for analysis. A TA Instruments model Q2000 differential

scanning calorimeter (DSC) was used to quantify glass transition temperatures of the

polymers; TA Universal Analysis software was used to analyze the thermograms. Viscosity

133 measurements were recorded with a Brookfield LV DV-II+ Pro Viscometer using

Rheocalc V2.6 software for data analysis. A SC4-25 spindle was used at a rotating speed

of 10 rpm, 2.20 s-1 shear rate.

6.3.3 Synthesis of maleated soybean oil (MA-SBO) (1A)

In a typical maleation reaction, SBO (220.50 g, 0.25 mol) and MA (29.49, 0.30

mol) were added to a 500 mL round bottomed flask equipped with an overhead mechanical

stirrer, N2 inlet, and a condenser. The reaction mixture was heated to 180 ºC to melt MA

and sparged with N2 for 15 min under mild agitation. The N2 sparge was subsequently

removed and replaced with a slow N2 purge through the vapor space of the round bottomed

flask. The clear, light yellow reaction mixture was then slowly increased to 220 ºC where

it was held 5 h. Reaction progress was monitored via FT-IR spectroscopy, following the

disappearance of the characteristic MA absorbance band ~840 cm-1. When changes were

no longer observed, the residual, unreacted maleic anhydride was removed via vacuum

distillation at 150 °C to yield 1A as a reddish-amber oil (240.89 g, 96.36% yield). 1H NMR:

δH (CDCl3; 750 MHz)/[ppm]: 0.81-0.99 (-CH3), 1.26-1.38 (-CH2-), 1.61-1.62 (-CH2-CH2-

CO-O-), 2.01-2.09 (=CH-CH2-), 2.15-2.19 (-CH2-CH=CH-CH=), 2.30-2.32 (-CH2-CO-O-

), 2.75-2.81 (=CH-CH2-CH=), 2.89-2.93 (-CH2-CO-O-, cyclic), 3.00-3.10 (-CH-CH-CO),

3.14-3.23 (-CH-CH-CO-, cyclic), 4.14-4.31 (-CO-O-CH2-), 5.26-5.61 (-CO-O-CH- and –

13 CH=CH-), 5.81-5.90 (-CH=CH-CH=), 6.41 (-CH=CH-CH=). C NMR: δC (CDCl3; 500

MHz)/[ppm]: 13.79-14.23 (-CH3), 20.52 (=CH-CH2-CH3), 22.04-22.65 (-CH2-CH2-CH3),

24.82-24.85 (CO-CH2-CH2-), 25.51-25.61 (-CH=CH-CH2-CH=CH-), 27.18-27.20 (-CH2-

CH2-CH=CH-), 28.95-29.74 (-CH2-), 31.22-31.35 (-CH2-, cyclic) 31.50-31.90 (-CH2-CH2-

CH3), 34.01-34.17 (CO-CH2-; CO-CH-CH-, anhydride), 42.46-45.47 (-CH-, cyclic), 62.80 134 (-CH2-O-CO-), 68.89 (-CH-O-CO-), 127.09 (-CH=CH-CH2-CH=CH-), 127.73-128.27 (-

CH=CH-CH2-CH=CH-CH2-CH=CH-; -CH-CH=CH-CH=CH-), 129.65-130.19 (-CH2-

CH=CH-CH2-CH=CH-; -CH-CH=CH-CH=CH-), 131.93 (-CH=CH-CH2-CH3; -CH-

CH=CH-CH=CH-), 172.77-173.23 (-CO-O-; -CO-O-CO-).

6.3.4 Model ring opening reactions of maleated soybean oil with ethylene

glycol monopropyl ether (1B)

To evaluate viscosity and experimentally quantify the number of moles of maleic anhydride that were successfully grafted onto SBO (i.e. degree of maleation), model reactions were conducted with ethylene glycol monopropyl ether. In a typical model reaction, 1A (13.31 g, 0.01 mol) and ethylene glycol monopropyl ether (1.70 g, 0.02 mol)

were added to a 100 mL round bottomed flask and placed into a 120 °C oil bath. The

reaction was magnetically stirred for 12 h. Once no additional changes were observed in

the FT-IR spectra, the reaction was poured off and stored under N2 to yield 1B as a light

1 orange oil (14.81 g, 98.24% yield). H NMR: δH (CDCl3; 300 MHz)/[ppm]: 0.85-0.93 (-

CH3), 1.24-1.29 (-CH2-), 1.58-1.62 (-CH2-CH2-COO-), 1.99-2.03 (=CH-CH2-), 2.28-2.32

(-CH2-CO-O-), 2.75-2.79 (=CH-CH2-CH=), 3.39-3.63 (-CH2-O-CH2-), 4.11-4.26 (-CO-

O-CH2-), 5.28-5.39 (-CO-O-CH- and –CH=CH-).

6.3.5 Synthesis of trimethylolpropane diallyl ether functionalized,

maleated soybean oil (1C)

A mixture of 1A (16.80 g, 16.85 mmol), trimethylolpropane diallyl ether (3.59 g,

16.75 mmol), and phenothiazine (3.10 mg, 0.02 mmol) as inhibitor was added to a 100-mL

round bottomed flask and placed into a 120 ºC oil bath. The reaction was magnetically

stirred for 12 h. Aliquots were withdrawn for FT-IR throughout the course of the reaction 135 to observe changes in the carbonyl region. The reaction is considered complete following

the consumption of the grafted anhydride. Following this, the reaction was poured off and

1 stored under N2 to yield 1C as a light orange-amber oil (19.68 g, 96.50% yield). H NMR:

δH (CDCl3; 300 MHz)/[ppm]: 0.82-0.91 (-CH3), 1.25-1.37 (-CH2-), 1.60 (-CH2-CH2-CO-

O-), 2.01-2.05 (=CH-CH2-), 2.29-2.33 (-CH2-CO-O-), 2.74-2.80 (=CH-CH2-CH=), 3.30-

3.32 (-CH2-O-CH2-), 3.61 (-CH2-O-CH2-), 3.92-3.94 (-O-CH2-CH=), 4.15-4.28 (-CO-O-

CH2-), 5.15-5.39 (-CO-O-CH- and –CH=CH- and –CH=CH2), 5.82-5.91 (-O-CH2-CH=).

6.3.6 Synthesis of HEMA functionalized, maleated soybean oil (1D)

A mixture of 1A (17.88 g, 17.94 mmol), HEMA (2.29 g, 17.59 mmol), and phenothiazine (0.20 g, 1.02 mmol) was added to a 100-mL round bottomed flask and placed into a 120 ºC oil bath. The reaction was magnetically stirred for 8 h under a breathing air atmosphere to prevent gelation. Once no additional changes were observed in the FT-IR spectra, the reaction was poured off and stored under N2 to yield 1D as an orange-amber

1 oil (19.63 g, 96.32% yield). H NMR: δH (CDCl3; 300 MHz)/[ppm]: 0.85-0.97 (-CH3),

1.25-1.30 (-CH2-), 1.60 (-CH2-CH2-CO-O-), 1.94 (-CH3), 2.01-2.05 (=CH-CH2-), 2.28-

2.33 (-CH2-CO-O-), 2.75-2.79 (=CH-CH2-CH=), 4.15-4.32 (-CO-O-CH2-), 5.26-5.40 (-

CO-O-CH- and –CH=CH-), 5.59 (-CH=CH2), 6.13 (-CH=CH2).

6.3.7 Synthesis of diethylene glycol monobutyl ether and allyl glycidyl

ether functionalized, maleated soybean oil (1E)

A mixture of 1A (17.63 g, 17.61 mmol), diethylene glycol monobutyl ether (2.84

g, 17.51 mmol), and phenothiazine (8.20 mg, 0.04 mmol) was added to a 100-mL round

bottomed flask and placed into a 100 ºC oil bath. The reaction was magnetically stirred for

136 5 h while aliquots were withdrawn from the reaction periodically. Once no additional

changes were observed in the FT-IR spectra, allyl glycidyl ether (2.01 g, 17.63 mmol) and

DBU (3 drops) were added to the round bottomed flask. After 12 h, the reaction was poured

1 off and stored under N2 to yield 1E as a dark amber oil (20.93 g, 93.06% yield). H NMR:

δH (CDCl3; 300 MHz)/[ppm]: 0.85-0.92 (-CH3), 1.25-1.30 (-CH2-), 1.58-1.60 (-CH2-CH2-

CO-O-), 2.01-2.05 (=CH-CH2-), 2.28-2.33 (-CH2-CO-O-), 2.75-2.78 (=CH-CH2-CH=),

3.45-3.67 (-CH2-O-CH2-), 4.02 (-CH-OH), 4.13-4.30 (-CO-O-CH2-), 5.28-5.39 (-CO-O-

CH- and –CH=CH- and –CH=CH2), 5.87-5.92 (-O-CH2-CH=).

6.3.8 Synthesis of diethylene glycol monobutyl ether functionalized,

maleated soybean oil (1F)

A mixture of 1A (15.35 g, 15.33 mmol), diethylene glycol monobutyl ether (2.48

g, 15.29 mmol), and phenothiazine (0.10 g, 0.51 mmol) was added to a 50-mL round

bottomed flask and placed into a 100 ºC oil bath. The reaction was magnetically stirred

overnight (16 h). Once the reaction showed no additional changes in the FT-IR spectra, the

reaction was poured off and stored under N2 to yield 1F as a reddish-amber oil (16.41 g,

1 91.54% yield). H NMR: δH (CDCl3; 300 MHz)/[ppm]: 0.85-0.97 (-CH3), 1.25-1.30 (-

CH2-), 1.58-1.60 (-CH2-CH2-COO-), 2.01-2.05 (=CH-CH2-), 2.28-2.33 (-CH2-CO-O-),

2.75-2.79 (=CH-CH2-CH=), 3.47-3.72 (-CH2-O-CH2-), 4.11-4.30 (-CO-O-CH2-), 5.28-

5.39 (-CO-O-CH- and –CH=CH-).

137 6.3.9 Synthesis of diethylene glycol monobutyl ether and glycidyl

methacrylate functionalized, maleated soybean oil (1G)

A mixture of 1F (13.37 g, 11.50 mmol), glycidyl methacrylate (1.63 g, 11.46 mmol), and dimethylbenzylamine (5 drops) was added to a 50-mL round bottomed flask

and placed into a 100 ºC oil bath. The reaction was magnetically stirred for 24 h under a

breathing air atmosphere to prevent gelation. Once the reaction showed no additional

changes in the FT-IR spectra, the reaction was poured off and stored under N2 to yield 1G

1 as a dark red-amber oil (14.24 g, 94.89% yield). H NMR: δH (CDCl3; 300 MHz)/[ppm]:

0.85-0.92 (-CH3), 1.25-1.30 (-CH2-), 1.58-1.60 (-CH2-CH2-CO-O-), 1.95 (-CH3), 2.01-

2.05 (=CH-CH2-), 2.28-2.33 (-CH2-CO-O-), 2.74-2.79 (=CH-CH2-CH=), 3.45-3.72 (-

CH2-O-CH2), 4.11-4.32 (-CO-O-CH2-), 5.26-5.37 (-CO-O-CH- and –CH=CH-), 5.60 (-

CH=CH2), 6.14 (-CH=CH2).

6.3.10 Synthesis of linseed long oil alkyd resin

The synthesis and characterization of this long oil alkyd is detailed in Section 4.3.3.

6.3.11 Total acid number calculations

A theoretical acid number (AN) can be calculated that corresponds to the amount of maleic anhydride that was charged at the beginning of the reaction, and it can be calculated via Equation (16).

�ℎ��������� ��

� �� (16) 56.1 ��� ��� ∗ 1,000 � ∗ ��� �� ���� ������ �� ������� = � �� �������

138 AN was measured experimentally with titrations were performed according to

ASTM D 1639-90,49 using Equation (17). All results were obtained in duplicate.

�� �� 56,100 ��� ∗ �� ��� �� ���� �������� �� ∗ 0.1 � ��� �� ���� (17) = ��� � �� �����

6.3.12 Formulation and Film Formation

The coatings were formulated with alkyd, RD and a combination of metal driers.

The driers used were Borschi® OXY-coat, 12% Zirconium Hex-Cem®, and 5% Calcium

Hex-Cem® with a ratio of 1:17:2 by weight. The RDs were added in 10 wt.% increments, relative to alkyd content up to 30 wt.%. A control set was made using unmodified SBO as a RD in the same amounts. Each formulation also included 2 wt.% of the drier package.

Additionally, BYK 333, a surface-tension reducer was used as an additive at 1.5 wt.%. The formulations were mixed on a roller-mill, and then applied to AQ-36 aluminum panels, and

allowed to dry in ambient conditions for 2 weeks prior to testing.

6.3.13 Viscosity Testing

Viscosity was measured using a Brookfield DV-II+ Pro viscometer with a SC4-25

spindle, and Rheocalc V2.6 software was used for the analysis. All samples were measured

in triplicate following 1 min. of run time at 10 rpm, and 2.20 s-1.

6.3.14 Thermal Properties

TA Instruments Q2000 DSC was used to collect glass transition temperatures of

the samples before and after curing, using hermetically sealed aluminum pans with 5-10

mg of sample. The temperature was increased to 100 ºC and held for 5 min, and then cooled

139 to -70 ºC. The temperature was then ramped to 100 ºC at 10 ºC min-1 to measure the glass

transition temperature. Analysis was done using TA Universal Analysis software.

6.3.15 Soxhlet Extraction

Gel content was determined from Soxhlet extractions. Experiments were conducted

over the course of 24 h using acetone as the solvent. Samples were subsequently dried

overnight in an 80 °C vacuum oven prior to obtaining the final mass; initial sample weights

were approximately 0.2 g. All Soxhlet extractions were performed in triplicate. The

average was calculated with a 95 % confidence interval.

6.3.16 Tensile Testing

Tensile testing was done using an Instron Tensiometer 5567 with a 100 N load cell

and hydraulic clamps. Sample dimensions were measured using a micrometer, and were

approximately 25 x 10 x 0.06 mm. The rate of extension was kept constant at 10 mm min-

1. The tensile properties of tensile strength, Young’s modulus and strain at break were determined. Young’s modulus was calculated using the instantaneous modulus at a strain of 0.1 mm mm-1. A minimum of five replicates was completed for each sample. The

average was calculated with a 95 % confidence interval.

6.3.17 Coatings Properties

Coatings tests were performed according to ASTM standards on AQ-36 aluminum panels. The pendulum (ASTM D 4366-16)156 and pencil hardness (ASTM D 3363-05),157

MEK chemical resistance (ASTM D 2794-93),158 pull-off (ASTM D 4541-17)159 and

crosshatch adhesion (ASTM D 3359-17),160 and Gardner impact (ASTM D 2794-93)161

140 were all evaluated. A minimum a of five samples were tested for each system. The average

was calculated with a 95 % confidence interval.

6.4 Results

6.4.1 Synthesis and Characterization of Maleated Soybean Oil

The maleation of seed oils, which proceeds via an ene mechanism, is a well-known

synthetic route to afford functionalized fatty acids (Figure 6.2). Elevated temperatures in

the range of 200-230 ºC are typically required for this reaction to take place due to the low

reactivity of SBO unsaturation sites. As a way to circumvent these conditions, free radicals

have been used to expedite the maleation process.215 However, both free radicals and high

temperatures can lead to substantial increases in product viscosity. This is due to the

propensity for triglycerides to react with one another though dimerization and

oligomerization processes. Similar reactions conditions are used to prepare “blown” or

“bodied” oil. As such, a preliminary N2 sparge was conducted prior to the reaction to

remove any dissolved gases; the reaction was subsequently carried out under an inert

atmosphere by running nitrogen through the vapor space of the round bottomed flask. This

effectively minimized the coupling reactions that could occur between the SBO molecules, which was confirmed by MALDI-ToF-MS data, shown in Figure 6.3.

Figure 6.2. Reaction scheme for MA-SBO.

141 20190228_SBO 0:J1 MS Raw 3000 SBO Inte ns. [a.u.] 2500 1001.604

2000 1043.326 1370.321 1398.349 1500

1000

500

04 x1 0 20190228_MASBO 0:M1 MS Raw 1001.748 MA-SBO

Inte ns. [a.u.] 3

1071.714

1097.729 2

1 1195.730

1674.235

0 1000 1500 2000 2500 3000 3500 m/z Figure 6.3. MALDI-ToF MS of SBO (top) and MA-SBO (bottom), showing no high molecular weight species in MA-SBO.

The maleation reaction was carried out with a slight molar excess of MA to SBO

(approximately 1.20 equivalents of MA to 1.00 equivalent of SBO). This was to account for any volatilized MA (bp = 202 ºC), which may have evaporated and solidified on the condenser. Higher degrees of functionalization were also attempted to increase the functionality of the final RD. However, it was found that the viscosity tends to increase exponentially with the amount of MA that is added (Figure 6.5, a). To model the effect of ring-opening on diluent viscosity, reactions were conducted with ethylene glycol monopropyl ether as the nucleophile, shown in Figure 6.4. Because the ring opening reaction produces a carboxylic acid group, the concern was that viscosity might increase with ring-opening due to more hydrogen bonding in the system. However, the ring opening reaction slightly decreased the viscosity, but the exponential increase was still observed for

142 systems containing a higher degree of functionality (Figure 6.5, b). This indicated that a near stoichiometric equivalence ratio was necessary for the preparation of effective RDs.

O HO O O O O O O O O O O HO O O O O O O 120 ºC, 12 h O

O O

Figure 6.4. Reaction scheme of ring-opening model reaction.

Unfunctionalized a. 12000 b. Functionalized 12000 10738 11210 10000 10000

8000 8000

6000 6000

4000 4000

Viscosity (mPa.s 3657 Viscosity (mPa.s) 3110 2000 2000

0 653 0 461 54.3 54.3

0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 Degree of maleation (i.e. moles of MA in feed) Degree of maleation (i.e. moles of MA in feed)

Figure 6.5. Viscosity of (a.) MA-SBO with varying degrees of maleation, and (b.) after functionalization of MS-SBO with ethylene glycol monopropyl ether.

The model ring opening reactions were also useful for determining the degree of maleation, which was accomplished via AN calculations.49 As previously mentioned, one must be conscientious of MA’s volatility at high reaction temperatures. Evaporated MA alters the degree of maleation that is achieved relative to the actual amount charged at the beginning of the reaction. The results of the experimental AN values were compared against theoretical AN values and are provided in Table 6.1. In each case, the theoretical

AN was always higher, as expected. The theoretical value corresponds to the actual amount of MA that was added at the beginning of the reaction and was calculated using Equation 143 (16). This suggests that a small amount of MA did not graft onto SBO during the course of

the reaction. Any unreacted MA was subsequently removed in vacuo. Regardless, the

values always remained close to theory. Back calculations of the experimental AN were

used to determine the actual amount of MA that was successfully grafted onto SBO.

Table 6.1. Comparison of experimental and theoretical AN to determine the degree of maleation. aTheoretical AN was determine from Equation (16).

Moles of MA Actual moles of MA Theoretical ANa Experimental AN charged initially grafted onto SBO 1.00 52.17 49.03 0.94 1.20 61.07 56.68 1.10 1.50 74.24 69.44 1.39 2.00 94.64 93.17 1.96

The progression of the maleation reaction was followed using FT-IR.

Disappearance of MA’s characteristic absorbance band at 840 cm-1 was used to determine

the completion of the reaction. This absorbance band corresponds to the unsaturated group

of MA that is consumed during the ene reaction that occurs. The final MA-SBO product

contains a succinic anhydride group grafted onto the fatty acid chain, as shown in Figure

6.2. Several other absorbance bands were used in conjunction to confirm the reaction

between MA and SBO (Figure 6.6 and Figure 6.7). In the carbonyl region of the FT-IR spectra, unmodified SBO shows the presence of only one absorption band at 1745 cm-1.

This corresponds to C=O ester of the triglyceride. Following the maleation reaction, two

new absorption bands at 1783 cm-1 and 1864 cm-1 were observed. These are attributed to the asymmetric and symmetric C=O stretching, respectively, of the grafted succinic anhydride moiety. Additionally, the formation of a new absorption band at 916 cm-1 was observed, which can be assigned to the C-C stretch of succinic anhydride.216 As this new

144 band appears, the reduction of C=C stretching of the MA was observed at 860 and

890 cm-1.216,217

Figure 6.6. FT-IR spectra overlay of SBO (top) and the MA-SBO (bottom).

-C=O (anhydride) -C=O (ester)

-C=O (anhydride)

-C=C

-C=C

-C-C

Figure 6.7. FT-IR spectra of real-time reaction monitoring for the maleation reaction between 1600-1950 cm-1 (top) and 780-1000 cm-1 (bottom). 145 The 1H NMR spectrum shows new resonances that are not present in SBO (Figure

6.8). These new resonances between 2.8-3.2 ppm correspond to the methylene and methine protons of the succinic anhydride group. An additional resonance at 6.4 ppm is observed, which corresponds to the conjugated double bond protons that form during the ene reaction.

In the 13 C NMR spectrum, new resonances between 43-45 ppm appear due to the carbon atoms of the succinic anhydride (Figure 6.9). Further analysis by 2D NMR spectroscopy experiments confirmed these findings. Distinct couplings that show the grafted succinic anhydride is present. In the 2D HSQC NMR spectra, new couplings are found between the protons at 2.8-3.2 ppm and the carbons between 43-45 ppm (Figure 6.10, left). These same couplings are also observed in the 2D HMBC NMR spectra in the same region (Figure

6.10, right). The HSQC indicates that the coupled protons and carbons are covalently bound; the HMBC shows correlations between protons and their neighboring carbon groups, but not the carbon to which they are directly bound. Additional correlations in 2D

HMBC NMR spectra show that the cyclic protons and the fatty acid chain protons correlate directly with the carbonyl carbons in the anhydride (Figure 6.11). Lastly, coupling between the protons in the conjugated double bonds (6.4 ppm) and the cyclic carbons in the succinic anhydride (43 ppm) clearly shows that the grafting occurs in the fatty acid chain, adjacent to the diallylic methylene groups (Figure 6.12). Further comparisons between MA-SBO

and SBO in terms of 1H, 13C and 2D NMR spectroscopy can be found in the Appendix, 0.

146 MA_SBO_yellowcap_750data_04223019_proton_for2d.esp J 1.31 1.28 1.26 0.89 K E 0.88 L N A M D G J C B

C E F H I 2.31

A I 1.34 D 1.35 2.32 2.06 2.05 1.61 1.36 F,H 1.62

B, G 0.90 2.77 2.30 0.87 5.34 1.37 2.02

A 2.01 5.35 2.07 5.33 2.78 4.15 2.76 4.29 4.29 4.16 5.38 4.14 4.30 5.33 1.38 4.14 4.31 5.39 N 5.26

5.26 K, L, M 7.27 5.27 0.98 2.81 5.40 2.08 0.99 2.75 2.09 3.23 2.93 5.90 3.19 6.41 5.44 3.14 3.00 5.81 5.51 0.81 5.61

0.13 0.31 7.62 4.34 0.650.46 0.59 3.83 7.090.66 8.81 7.33 56.99 9.00

7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ChemicalShift(ppm) Figure 6.8. 1H NMR spectra of MA-SBO.

O K L N P E F CARBON_2019Apr16_01.espA M

R 77.24 76.99 76.74 29.32 27.18

D G J 29.67

B Q C E F 25.61

H I 22.54 H P J A 31.50

K, R 14.03

G, N 22.65 130.19 31.87 24.82

129.98 D 127.87 34.01 128.05 O, Q B A C, M 62.08 34.17 68.89 173.20 172.77 I L 13.95 20.52 14.23 131.93 127.09 127.73 22.04 44.41 42.46 42.79 45.47

176 168 160 152 144 136 128 120 112 104 96 88 80 72 64 56 48 40 32 24 16 ChemicalShift(ppm) Figure 6.9. 13C NMR spectra of MA-SBO.

147 HSQC HMBC

K L M

Figure 6.11. 2D NMR spectra of MA-SBO with HSQC (left) and HMBC (right).

HMBC, MA-SBO HMBC, SBO

K L N M

N coupling K, L, M with L, M couplings

Figure 6.10. 2D HMBC NMR spectra of MA-SBO (left) which shows coupling between the anhydride carbonyl carbons and the protons in the 5-membered ring that are not present in SBO (right). 148 HMBC, MA-SBO HMBC, SBO

O K L M

O coupling with K, L, M

Figure 6.12. 2D HMBC NMR spectra of MA-SBO (left) which shows coupling between the conjugated double bonds and the anhydride ring; the unmodified SBO (right) does not exhibit this coupling behavior.

The MALDI-MS analysis of the MA-SBO compared to the SBO is shown in Figure

6.13. The addition of maleic anhydride is indicated by the peaks at higher molecular

weight, corresponding to the grafting of the succinic anhydride to the SBO. This also confirms that no dimerization took place during the functionalization reaction because no high molecular weight products are observed.

149 4 x1 0 901.736 20190228_SBO 0:J1 MS Raw 909.793 6 SBO Inte ns. [a.u.]

5 SBO+H

877.734 SBO+Na 4

881.769 3 897.702

2 885.372

1 893.720 917.728 933.764 962.619 869.353 1001.604 1043.326

04 x1 0 907.355 20190228_MASBO 0:M1 MS Raw

5 MA-SBO Inte ns. [a.u.]

4 SBO+H SBO+Na 901.730 1001.748 877.730 MASBO+H 3 874.571 899.709 MASBO+Na 975.732

2

962.615 1053.739 1 1017.739 1029.740

0 880 900 920 940 960 980 1000 1020 1040 m/z Figure 6.13. MALDI-MS of SBO (top) and MA-SBO (bottom).

6.4.2 Synthesis and Characterization of Reactive Diluents

The synthetic preparation of the reactive diluents involved the addition of various alkyl, allyl, or methacrylate functionalities onto the MA-SBO. Trimethylolpropane diallyl ether (TMPDAE), 2-hydroxyethyl methacrylate (HEMA), allyl glycidyl ether (AGE) with diethylene glycol monobutyl ether, and glycidyl methacrylate (GMA) with diethylene glycol monobutyl ether were used in this study. The idealized products are shown in Figure

6.14. These molecules were chosen due to their ability to react with the anhydride (or carboxylic acid of the esterified succinate) and the additional unsaturated bonds that were introduced, which are capable of crosslinking via an autoxidative curing mechanism with alkyd resins.

150 O

O O O HO O O O O O O OH O O O O O O O O O O

O O

TMPDAE-functionalized MA-SBO HEMA-functionalized MA-SBO

OH OH O O O O O O O O O O O O O O O O O O O

O O O O O O

O O

AGE-functionalized MA-SBO GMA-functionalized MA-SBO Figure 6.14. Reaction products for MA-SBO-based reactive diluents.

These reactions were monitored with FT-IR spectroscopy. Disappearance of the

anhydride absorption bands at 1783 cm-1 and 1864 cm-1 was used to follow the progression of the reaction. It should be noted that these absorbance bands, particularly the band at

1783 cm-1 never fully disappeared, which was likely due to the excess MA that was charged during the preceding maleation reaction. In addition to these changes, the appearance of a shoulder around 1700 cm-1 was observed. This corresponds to the carboxylic acid group as the succinic anhydride ring opens to form the half-acid, half-ester moiety. Spectra from real-time reaction monitoring for the synthesis of the AGE RD are shown in Figure 6.15.

As expected, one can see the succinic anhydride absorbance bands concurrently disappear as the carboxylic acid shoulder forms. For this particular example, where the carboxylic acid is subsequently reacted with AGE, a decrease in intensity in this carboxylic acid shoulder is found as it is consumed during the epoxy-carboxylic acid addition reaction.

Disappearance of the absorption band at 920 cm-1, which corresponds to the cyclic C-C in

the succinic anhydride moiety,216 occurs when the ring is opened by reaction with AGE

(Figure 6.18). Differences observed from 900 to 1200 cm-1 correspond to the 151 functionalization added from TMPDAE (Figure 6.16), HEMA (Figure 6.17), AGE (Figure

6.18), and GMA (Figure 6.19).

-C=O (ester)

-C=O (anhydride)

-C=O (acid)

-C=O (anhydride)

Figure 6.15. Real-time FT-IR monitoring of the reaction of AGE and DEGMBE with MA-SBO from 1600 to 1900 cm-1.

-C=O (ester) -C=O (anhydride) -C-C (anhydride)

Figure 6.16. FT-IR spectra overlay of MA-SBO (top) and the TMPDAE- functionalized MA-SBO (bottom).

152 Figure 6.17. FT-IR spectra overlay of MA-SBO (top) and the HEMA-functionalized MA-SBO (bottom).

Figure 6.18. FT-IR spectra overlay of MA-SBO (top) and AGE with DEGMBE (bottom).

-C=O (ester) -C=O (anhydride)

-C=O (acid)

Figure 6.19. FT-IR spectra overlay of MA-SBO (top) and GMA with DEGMBE (bottom).

153 In addition to the FT-IR analysis, 1H NMR spectroscopy was also used to monitor and characterize the functionalization reactions. Unfortunately, because many resonances tend to overlap with one another (especially between 3.0-4.5 ppm), it was difficult to follow the progress of the acyl substitution reactions. These are shown in Figure 6.20 and Figure

6.21. The addition reactions that were conducted with glycidyl ether compounds, on the other hand, could be easily followed via 1H NMR spectroscopy. The disappearance of

epoxide resonances at 2.61, 2.80, and 3.15 ppm for AGE and at 2.67, 2.85, and 3.25 ppm

for GMA were used; once these resonances were completely consumed, the reaction was

deemed complete. The final 1H NMR spectrum of AGE- and GMA-functionalized MA-

SBO are shown in Figure 6.22 and Figure 6.23, respectively. The AN also decreased after

the addition reactions were conducted, which further confirmed the functionalization.

However, it should be noted that these AN values never fully reached zero. This was likely

a result of the in-situ formation of hydroxyl groups which are also capable of reacting with

epoxies.218

N L E B, C K K TH_062A_done.esp M N

L 1.25 H E J J G I A

E C C G E D C C H 1.30 E F C B E D C A CHCl3 A E, I J, M, L K H

D 0.87 7.26 F G 0.88 2.31 5.34 0.85

N 5.35 2.03 0.86 3.31 0.90 2.05 2.28 2.33 2.01 0.84 1.60 3.93 1.37 3.30 3.94 3.92 2.76 2.29 0.91 3.92 5.37 3.32 3.61 5.27 4.15 4.17 5.32 5.27 5.26 4.28 5.15 2.74 2.78 0.82 5.37 5.39 5.85 5.87 5.84 5.82 5.91

1.55 11.37 9.41 4.43 2.99 6.99 8.69 64.14 12.00

7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0 Chemical Shift (ppm) Figure 6.20. 1H-NMR spectrum of TMPDAE-functionalized MA-SBO.

154 B, C K TH_064B1_8h.esp

J 1.25 H J J J L I

E C C G E 1.30 D C C H E F C B E K D C A E, I A CHCl3 H J D F G 0.88 0.87 2.31 7.26

L L 5.34 2.03 5.36 2.05 2.28 2.33 2.01 0.86 1.94 1.60 2.76 0.91 0.85 4.32 1.99 4.28 5.37 4.26 4.17 4.15 5.32 4.30 2.75 2.79 5.28 5.39 5.26 5.30 5.59 0.97 6.13 5.40

0.58 0.69 8.27 6.70 3.41 7.15 11.20 64.99 9.00

7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0 Chemical Shift (ppm) Figure 6.21. 1H-NMR spectrum of HEMA-functionalized MA-SBO.

B, C, L Q TH_070B1.esp R P M N J J M M L 1.25 H A J J M M K I

E C C G E D C C H 1.30 E F CHCl3 C B E

7.26 D C A E, I, R A H J, N D 0.89

F G,K 0.88

M, P 2.31

Q 5.34 5.36 0.91 2.03 2.28 2.05 2.33 1.60 2.01 1.58 0.86 0.92 2.77 2.29 0.85 3.47 4.17 4.15 4.26 5.37 3.45 4.28 3.62 3.64 3.59 3.60 5.32 3.63 2.78 2.75 4.13 3.67 5.30 4.30 5.38 5.28 5.39 5.91 5.87 5.89 5.89 5.92

0.81 10.40 9.88 11.40 3.72 7.99 9.84 9.98 61.63 12.00

7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0 Chemical Shift (ppm) Figure 6.22. 1H-NMR spectrum of AGE-functionalized MA-SBO.

155 Q B, C, L

TH_072A_complete.esp P N J J J M M L 1.25 H A J J M M K I

E C C G 1.30 CHCl3 E D C C H E F C B E A D C Q

7.26 A E, I J, N H M D 0.89

F G,K 0.88 2.31

P P 5.34 2.03 0.91 5.36 2.28 2.05 1.60 2.01 0.86 2.33 1.58 0.85 2.77 1.95 0.92 4.17 4.15 4.26 5.37 4.28 3.59 3.62 5.32 3.47 3.45 4.13 3.64 3.60 2.74 2.79 4.11 3.58 5.39 5.28 5.30 5.26 6.14 5.60

0.82 0.99 8.53 9.20 9.23 3.55 7.74 12.80 9.78 60.81 12.00

7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0 Chemical Shift (ppm) Figure 6.23. 1H-NMR spectrum of GMA-functionalized MA-SBO.

6.4.3 Formulations and Viscosity Measurements

Viscosity measurements of the formulations were conducted to analyze the effectiveness of the RDs. The viscosity of the formulations was found to decrease significantly with the addition of RDs, as shown in Figure 6.24. AGE shows the greatest ability to reduce viscosity of the formulations. Compared to the viscosity of the alkyd, the addition of 20 wt.% of the AGE reactive diluent results in approximately a 60 % decrease in formulation viscosity. All of these diluents are shown to work effectively to decrease viscosity of the coatings.

156 Alkyd Formulation Viscosity 26,058 25,000

20,000

15,000 16,365

10,000

Viscosity (mPa.s) Viscosity 5487 10,654 5,000 2959 6,911 1456 0 0 5 10 15 20 25 30 Diluent Concentration, wt.%

SBO TMPDAE HEMA DEGMBE + AGE DEGMBE + GMA

Figure 6.24. Viscosity of the formulations with increasing amounts of RDs.

6.4.4 Crosslink Density

The crosslink density of these coatings can be evaluated to gain a better

understanding of the differences in properties between the different RDs. Basing a

crosslink density on glass transition temperature has been previously investigated and

several equations and models have been proposed.185–189,192,219,220 DiMarzio187 developed an equation based on the change in Tg between the uncrosslinked polymer and the

crosslinked polymer which has been shown to accurately predict crosslink density in

triglyceride-based polymer systems.192 This is shown in Equation (18),

�, � = (18) 1 − �ν where Tg is the glass transition temperature of the crosslinked polymer (in Kelvin), Tg,u is the glass transition temperature of the uncrosslinked polymer (in Kelvin), ν is the crosslink

-3 -5 3 -1 192 density in mol m , and K2 is a material constant, found to be 2.0 x 10 m mol .

Using Equation (18),187,192 the crosslink density was calculated for each sample, and is shown in Table 6.2. The addition of RD decreases the crosslink density, especially 157 at high loadings. This has been shown previously to be the case when large amounts of

RDs are added to a coating system.11,165 The alkyd polyester backbone (particularly phthalic anhydride) is the only rigid feature in the polymer structure. The molecular weight is built up through autoxidative crosslinking of the pendant fatty acid chains on the alkyd and the MA-SBO. These long, flexible, alkyl chains have significant rotational and segmental freedom from the increased free volume within the crosslinked network.

Additionally, the SBO contains a proportion of fatty acids that are not reactive (e.g. stearic acid and oleic acid). This leads to more dangling chain ends that are simply plasticizing the formulation rather than contributing to crosslink density. Because of this, when the RD makes up a significant portion of the formulation, the crosslink density decreases accordingly.11,165 Comparing the functionalized RDs to the unmodified SBO shows that

the crosslinking is much greater for the functionalized samples, indicating that these added

functionalities promote more crosslinking reactions. Further, the differences in crosslink

density between the different sets of RDs results from the functionalities present in the

modification. For instance, HEMA-functionalized MA-SBO has a higher overall ν ,

possibly due to the methacrylate functionality that can participate in the autoxidative curing

process.

Soxhlet extraction was used to measure the gel content of these coatings, shown in

Figure 6.25. The gel content is an indication of crosslink density, as any unreacted materials

will be dissolved during the extraction, leaving only the crosslinked portions of the film.

Soxhlet extraction on the films using SBO as the RD was found to yield an insignificant

amount of material, indicating a very low gel content. This is in agreement with previous

results on the unmodified SBO coatings, which yielded poor coating properties and were

158 also too soft and tacky to undergo tensile testing. The gel content of the RD-containing samples was found to decrease slightly compared to the control. Most of the RD systems show a gel content within error of each other, indicating little change in the overall crosslink density of the coatings with increasing RD content. Overall, the incorporation of

RD led to a slight decrease in crosslink density when compared against the control. This was in good in agreement with the crosslink density data obtained from DSC experiments

(Table 6.2).

Table 6.2. The crosslink density, ν calculated from Equation (18) using the Tg,u and Tg, measured by DSC. -3 Sample Tg,u (°C) Tg (°C) ν (mol m ) Control -18 ± 2 19 ± 2 6.3 x 103 10% SBO -25 ± 3 -6.5 ± 0.7 3.4 x 103 20% SBO -14 ± 1 -8.7 ± 0.9 1.0 x 103 30% SBO -15 ± 2 -8.5 ± 0.9 1.4 x 103 10% TMPDAE -21 ± 2 5.1 ± 0.5 4.7 x 103 20% TMPDAE -14 ± 1 5.3 ± 0.5 3.5 x 103 30% TMPDAE -15 ± 2 6.8 ± 0.7 3.8 x 103 10% HEMA -27 ± 3 3.2 ± 0.3 5.4 x 103 20% HEMA -20 ± 2 6.0 ± 0.6 4.6 x 103 30% HEMA -20 ± 2 4.2 ± 0.4 4.3 x 103 10% AGE -28 ± 3 6.2 ± 0.6 6.2 x 103 20% AGE -14 ± 1 3.5 ± 0.4 3.1 x 103 30% AGE -15 ± 2 4.4 ± 0.4 3.6 x 103 10% GMA -28 ± 3 4.4 ± 0.4 5.8 x 103 20% GMA -19 ± 2 6.0 ± 0.6 4.5 x 103 30% GMA -17 ± 2 4.2 ± 0.4 3.7 x 103

Soxhlet Extraction Data 63 61 59 57 55 53 51 content (wt.%) content - 49

Gel 47 45 43

Control 10% AGE 20% AGE 30% AGE 10% HEMA20% HEMA30% HEMA 10% GMA20% GMA30% GMA 10% TMPDAE20% TMPDAE30%TMPDAE Figure 6.25. Gel content of the coatings analyzed by Soxhlet extraction. 159 6.4.5 Tensile Properties

Tensile properties are another indication of the crosslink density and overall

mechanical strength of the coatings. The unmodified SBO samples could not be tested for

tensile properties because they had not achieved complete through-cure following two

weeks, indicating overall poor crosslinking and coating properties. The results for the

functionalized RD samples are shown in Figure 6.26. For the functionalized RDs, the strain at break generally shows a decrease in every series as RD content is increased. The Young’s modulus was found to increase as a function of HEMA content, while the other RDs either had a decreasing modulus or showed no significant change. This would indicate that the addition of HEMA is creating a more flexible coating. The tensile strength generally decreases for each RD series. These properties are in agreement with the findings from crosslink density and gel content, which also indicated that high loadings of the RDs have a detrimental effect on mechanical properties. Strain at Break Modulus 1.8 2 1.6 1.8 1.4 1.6 1.2 1.4 1.2 1 1 0.8 0.8 0.6

Modulus, MPa Modulus, 0.6 0.4 0.4

Strain at Break, mm/mm at Break, Strain 0.2 0.2 0 0

Control Control 10% AGE20% AGE30% AGE 10% AGE20% AGE30% AGE 10% HEMA20% HEMA30% HEMA 10% GMA20% GMA30% GMA 10% HEMA20% HEMA30% HEMA 10% GMA20% GMA30% GMA 10% TMPDAE20% TMPDAE30% TMPDAE 10% TMPDAE20% TMPDAE30% TMPDAE

Tensile Strength 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 Tensile Strength, MPa Strength, Tensile 0.2 0

Control 10% AGE20% AGE30% AGE 10% HEMA20% HEMA30% HEMA 10% GMA20% GMA30% GMA 10% TMPDAE20% TMPDAE30% TMPDAE Figure 6.26. Tensile properties, including strain at break (top, left), modulus (top, right) and tensile strength (bottom). 160 6.4.6 Coatings Tests

General coatings tests were carried out to observe the performance of coatings

made with these RDs. The coating test results are summarized in Table 6.3 and Figure 6.27.

The functionalized RDs performed much better than unmodified SBO in these tests.

Overall, the modified RDs were found to have properties similar to the control alkyd.

However, hardness and adhesion were shown to decrease as the RD content was increased.

Previous studies11,14,165 have also shown a decrease coatings properties with high loadings of RDs; this is due to the reduction in crosslink density of the coatings due to the low molecular weight of these additives.

The drying time of these coatings was also analyzed (Table 6.3), showing that the drying time was generally increased with the addition of RDs. However, at 10 wt.% loading of TMPDAE or AGE, the drying time was improved, even compared to the control. The allyl ether functionality in these RDs is likely speeding up the curing process by the addition of more functional crosslinking sites. However, when the loading was increased over 10 wt.%, the large amount of a lower molecular weight component led to more plasticizing of the coating and a longer time for formation of a solid film. This is consistent with results shown by Wutticharoenwong and Soucek10 for similarly functionalized RDs.

161 Table 6.3. Coating properties. MEK Tack-Free Dry Hard Sample Double Time Time Rubs (h) (h) Control > 200 19 ± 1 > 24 10% SBO 120 ± 30 > 24 > 24 20% SBO 110 ± 20 > 24 > 24 30% SBO 90 ± 10 > 24 > 24 10% TMPDAE > 200 15 ± 1 20 ± 1 20% TMPDAE > 200 > 24 > 24 30% TMPDAE > 200 > 24 > 24 10% HEMA > 200 > 24 > 24 20% HEMA > 200 > 24 > 24 30% HEMA > 200 > 24 > 24 10% AGE > 200 20 ± 1 22 ± 1 20% AGE > 200 20 ± 1 > 24 30% AGE > 200 > 24 > 24 10% GMA > 200 > 24 > 24 20% GMA > 200 > 24 > 24 30% GMA > 200 > 24 > 24

162 Pendulum Hardness Pencil Hardness 25 16 14 20 12

15 10 8 10 6

Pencil Hardness Pencil 4 5 Pendulum Hardness, s Hardness, Pendulum 2 0 0

Control Control 10% SBO20% SBO30% SBO 10% AGE20% AGE30% AGE 10% SBO20% SBO30% SBO 10% AGE20% AGE30% AGE 10% HEMA20% HEMA30% HEMA 10% GMA20% GMA30% GMA 10% HEMA20% HEMA30% HEMA 10% GMA20% GMA30% GMA 10% TMPDAE20% TMPDAE30% TMPDAE 10% TMPDAE20% TMPDAE30% TMPDAE

Pull-Off Adhesion 450 2

- 400 350 300 250 200 150 Off Adhesion, lb in Off Adhesion, - 100

Pull 50 0

Control 10% SBO20% SBO30% SBO 10% AGE20% AGE30% AGE 10% HEMA20% HEMA30% HEMA 10% GMA20% GMA30% GMA 10% TMPDAE20% TMPDAE30% TMPDAE

Figure 6.27. Pendulum hardness (top, left), pencil hardness (top, right) and pull-off adhesion (bottom).

6.5 Discussion

The development and testing of SBO-based reactive diluents are described here.

Facile maleation and subsequent ring-opening functionalization reactions were carried out in the absence of solvents. The maleation reaction resulted in succinic anhydride groups being grafted to SBO. The reaction was confirmed by 2D HSQC and HMBC NMR spectroscopy, which showed couplings between the protons and carbons on the anhydride with the conjugated double bonds formed following the ene reactions. New couplings were also observed between the carbonyl carbons in the anhydride and the protons in the cyclic succinic anhydride. To our knowledge, this is the first time a detailed 2D spectroscopic analysis has been carried out on maleated soybean oil.

163 The maleation reaction results in grafting of a succinic anhydride moiety, which

can take place in a few different fashions,215,221,222 as shown in Figure 6.28. It is likely that

the final product is some combination of these configurations, along with additions onto

linoleic and oleic acid chains as well. The presence of multiple signals, between 2.8-3.2

ppm, in the HSQC and HMBC 2D NMR spectra suggest multiple modes of addition, even

with only a slight molar excess of MA to SBO, i.e. 1.2:1.0. In an effort to enhance crosslink

density and increase RD functionality, maleation reactions were carried out using higher

concentrations of MA. Unfortunately, it was found that at higher degrees of

functionalization, the viscosity increased exponentially, which created a modified SBO that

would not be useful as a diluent. This viscosity build-up could be partially due to the

secondary Diels-Alder reaction that can occur following the conjugation of the double

bonds.215,221,222 Additionally, polarity and oligomerization effects can contribute to the

exponential increase in viscosity.192 This same viscosity increase has been observed previously in other functionalized seed oils, including epoxidized,223 norbornylized,224 and maleated.225

164 O R O where R is the remaining soybean oil triglyceride

O O O O O O O O ene ene O ene

O O O O O O O O O O O O R R R O O O

O O Diels-Alder O

O O O O R O

O O O

Figure 6.28. Various configurations following the maleation reaction.

Following the maleation reaction, the MA-SBO can be modified with any

nucleophile by a ring-opening reaction. Optionally, the ring-opened anhydride can be

functionalized further by exploiting the reaction between carboxylic acids and epoxies.

This allows for a wide variety of functionalization possibilities for these bio-based RDs.

Here, TMPDAE, HEMA, AGE and GMA were chosen for their allyl ether and

methacrylate functionalities, respectively. DEGMBE was chosen based on the good diluent

properties exhibited by the model compounds and highlights the utility of these

compounds. Because these RDs were designed for use in seed oil-based alkyd coatings,

the functionalities were chosen to improve diluency and provide additional reactive sites

for autoxidative curing. The final coating formulations contained no solvent and are

comprised mainly of renewable, bio-based materials.

SBO was chosen for these RDs because it is an abundant, renewable resource and

is a low-cost seed oil. Unfortunately, as shown in this study, SBO alone cannot be used as

165 a reactive diluent. While it is capable of reducing viscosity, unmodified SBO leads to tacky

films with poor thermo-mechanical properties. This is primarily due to the low

concentration of diallylic methylene groups in SBO which are responsible for autoxidative

curing processes (SBO is considered a semi-drying oil rather than a drying oil, such as

linseed or tung oil).3 Through simple maleation and acyl substitution reactions, RDs suitable for high solids alkyd coatings could be prepared. These additional functionalizations provide new crosslinking sites that can participate in autoxidative curing reactions, mitigating the plasticizing effects found when unmodified SBO is used as a RD.

To confirm the crosslinking reactions with the added functionalities, those formulations were compared to ones made with unmodified SBO as a RD. The SBO worked more efficiently as a diluent, but the drying time was significantly longer than for the functionalized counterparts. The tensile samples were cast at 4 mil thickness were still not through-cured after 2 weeks. The crosslink density and coating performance were also found to be significantly diminished with the use of unmodified SBO. Comparing SBO with the functionalized MA-SBO RDs indicates that the added functionality significantly improves the reactivity, while still maintaining sufficient diluent properties.

Evaluation of crosslink density by the gel content of the coatings showed a slight decrease from the control with inclusion of the RDs. However, there was not a significant change at higher loadings, except in AGE and GMA, which are also functionalized with

DEGMBE. While this additional functionalization promotes improved diluency, the plasticization from this flexible chain also contributes to diminished mechanical properties.

These same trends were found in the calculated crosslink density, and the modulus and tensile strength. This indicates that there is an inverse relationship between the diluency

166 and mechanical properties that can be achieved for these systems, as was shown by the mechanical properties of the functionalized MA-SBO samples when compared to the unmodified SBO, which were too soft to undergo tensile testing.

The coating properties were also found to be consistent with the mechanical performance of these coatings. At higher loadings of RDs, the coatings were shown to soften and have reduced adhesion to the substrate. The unmodified SBO formulations showed even more significant reductions in these properties, even at only 10 wt.%, and the chemical resistance was also found to be much lower. However, with only 10 wt.% of the functionalized RDs, many performance properties remained similar to those of the control alkyd. These results were consistent with the mechanical properties. The drying time of these samples was also much slower than the functionalized samples. At higher film thicknesses, these samples did not cure at all. However, the functionalized RDs were found to have a positive impact on drying, especially at low concentrations. TMPDAE- and AGE- functionalized MA-SBO at 10 wt.% dried faster than the control alkyd with no RDs, which is likely due to the added allyl ether functionality. This functionality reacts easily with the autoxidative crosslinking of the alkyds because the hydrogens between the allyl group and the oxygen which are easily abstracted, in the same way as the diallylic methylene groups in the alkyd’s fatty acid chains. The other RD systems, HEMA- and GMA-functionalized

MA-SBO have methacrylate functionality, which does not participate in the autoxidative curing process as easily as the electron donating allyl ether groups.

At low loading levels, the functionalized RDs reported here work effectively to decrease viscosity and improve drying time without having detrimental effects on coating performance. Significant improvements were found when compared to unmodified SBO

167 as a RD. Additionally, these SBO-based RDs can be synthesized without solvents, using facile and well-known reaction techniques. They perform effectively to reduce the viscosity of alkyd coatings, leading to a highly plant-based coating system that contains no

VOCs. The maleation and subsequent functionalization reactions are incredibly versatile which creates the opportunity for several other functionalities to be explored to in future studies.

6.6 Conclusions

Novel SBO-based RDs were developed by a facile and solvent-free synthetic route.

The methodology used here could be applied to a variety of unsaturated seed oils and diverse nucleophilic groups could be used to add functionality. Each RD was shown to effectively decrease viscosity and react within the crosslinking of the alkyd coating without detrimental effects on the coating properties. Comparison to unmodified SBO showed that the functionalities led to significant improvement in performance and crosslinking. AGE- functionalized MA-SBO had the best diluent ability and resulted in a 60% reduction in viscosity with only 20 wt.% addition. The drying time was also found to be significantly improved by TMPDAE- and AGE-functionalized MA-SBO due to reaction of the allyl ether functionality into the coating system.

6.7 Acknowledgments

We wish to thank Mangaldeep Kundu and Dr. Venkat Reddy Dudipala for the 2D

NMR spectroscopy experiments and analysis, as well as Savannah Snyder and Dr. Chrys

Wesdemiotis for the MALDI-MS experiments and analysis. We also wish to thank the

168 Kresge Foundation and donors to the Kresge Challenge Program at The University of

Akron for funds used to purchase the NMR instrument used in this work.

169 CHAPTER VII

SUMMARY

In the first part of this dissertation, efforts were focused on understanding the effect of using alkoxysilane- and fluorine-functionalized alkyds or reactive diluents on performance, corrosion resistance and stratification of alkyd coatings. These specific functionalities were chosen to improve adhesion, hardness, hydrophobicity and corrosion protection of the coatings. Following synthesis of the components, formulations were made with increasing amounts of modified alkyds or reactive diluents. The coatings were analyzed for performance, tensile properties, corrosion resistance and weatherability.

Additionally, the distribution of the alkoxysilane- and fluorine-functionalized components in the cross-section of the coating was analyzed by ESEM-EDX spectroscopy. At higher concentrations of RDs, adhesion, hardness and corrosion resistance improved significantly, which was found to be due to a slight stratification of the alkoxysilane groups toward the substrate and fluorine toward the surface. The functionalized alkyds had enhanced corrosion resistance and tensile properties when compared to the RD systems. This work demonstrated the reactive diluents, used in loadings as low as 5 wt.% each, were found to partially stratify and provide improved corrosion protection and coating performance, even compared to modified alkyds.

In the second part of this dissertation, each of these modified alkyds and reactive diluents were studied in a different way. The alkoxysilane-functional components were

170 studied for their performance and drying abilities when cured in varying relative humidity conditions. The fluorinated components were compared for stratifying ability and corrosion performance to gain a better understanding of the difference in properties based on the molecular weight and viscosity of the two components. The alkoxysilane study showed that at high humidity, alkoxysilane functional components provided improved performance when compared to a control alkyd. The control alkyd showed slower drying time and poor performance when cured at high humidity, due to the retardation of the autoxidative curing under these conditions. The alkoxysilane-functional reactive diluents dried faster and formed harder coatings than the control alkyds, while the alkoxysilane- functional alkyds showed significant improvements in tensile strength and adhesion following curing at high humidity.

The fluorinated alkyd and reactive diluent systems demonstrated the effect of viscosity and molecular weight on stratification ability and coating performance. It was found that the fluorinated reactive diluents showed more stratification and with that, the coating surface was more hydrophobic. However, the lower molecular weight and reduction in crosslinking sites on this reactive diluent had a detrimental effect on the tensile properties and corrosion resistance of the coatings. The fluorinated alkyd systems did not stratify, but provided significantly improved corrosion protection, chemical resistance and hardness. Only a small amount of fluorinated alkyd as an additive in the coating was needed to provide these improved properties.

The final chapter focused on synthesis and performance of novel soybean oil-based reactive diluents. The chemistry used here provides a facile and solvent-free route to a variety of functionalization possibilities for these reactive diluents. Using maleic anhydride

171 and soybean oil, an ene reaction was carried out to obtain maleated soybean oil, which was extensively characterized by FT-IR, 1H-, 13C-, and 2D NMR spectroscopy. This could then be functionalized with a nucleophilic group in a simple anhydride ring-opening reaction.

The modifications explored here were various allyl ether and methacrylate groups, including trimethylolpropane diallyl ether (TMPDAE), 2-hydroxyethyl methacrylate

(HEMA), allyl glycidyl ether (AGE) with diethylene glycol monobutyl ether, and glycidyl methacrylate (GMA) with diethylene glycol monobutyl ether. These functionalized soybean oil compounds worked to effectively decrease the viscosity of the alkyd coatings by over 70 % in some cases. The coating performance did not diminish with the inclusion of these diluents at low concentrations, and the drying time was even decreased in some cases.

In summary, this study clearly demonstrated the differences in properties obtained when using functionalized alkyds or reactive diluents. The use of alkoxysilane and fluorine functionalized reactive diluents at only 10 wt.% of the total resin content provided significantly improved adhesion and corrosion resistance. The use of alkoxysilane- functionalized reactive diluents or alkyds in high humidity curing conditions resulted in faster drying, and improved coating and tensile properties, when compared to a control alkyd cured in these same conditions. Fluorine-functionalized alkyds used in small amounts provided corrosion protection and chemical resistance, despite a lack of stratification. Lastly, a novel route to functionalized soybean oil reactive diluents was developed, which opens up a variety of potential paths to further optimize and functionalize high solids alkyd coatings.

172 REFERENCES

(1) Hofland, A. Alkyd Resins: From down and out to Alive and Kicking. Prog. Org.

Coat. 2012, 73 (4), 274–282. https://doi.org/10.1016/j.porgcoat.2011.01.014.

(2) Elliott, W. T. Alkyd Resins. In Surface Coatings; Springer Dordrecht, 1993; pp 76–

109.

(3) Organic Coatings: Science and Technology, 3rd ed.; Wicks, Z. W., Ed.; Wiley-

Interscience: Hoboken, N.J, 2007.

(4) Encyclopedia of Polymeric Nanomaterials; Kobayashi, S., Müllen, K., Eds.; Springer

Berlin Heidelberg: Berlin, Heidelberg, 2015. https://doi.org/10.1007/978-3-642-

29648-2.

(5) Holmberg, K. High Solids Alkyd Resins; Marcel Dekker, Inc.: New York, 1987.

(6) Kodali, D. R. Linseed Oil Based Polymeric Vehicle. 5,288,805.

(7) Yadav, S. K.; Schmalbach, K. M.; Kinaci, E.; Stanzione, J. F.; Palmese, G. R. Recent

Advances in Plant-Based Vinyl Ester Resins and Reactive Diluents. Eur. Polym. J.

2018, 98, 199–215. https://doi.org/10.1016/j.eurpolymj.2017.11.002.

(8) Lligadas, G.; Ronda, J. C.; Galià, M.; Cádiz, V. Renewable Polymeric Materials from

Vegetable Oils: A Perspective. Mater. Today 2013, 16 (9), 337–343.

https://doi.org/10.1016/j.mattod.2013.08.016.

173 (9) La Scala, J. J.; Sands, J. M.; Orlicki, J. A.; Robinette, E. J.; Palmese, G. R. Fatty Acid-

Based Monomers as Styrene Replacements for Liquid Molding Resins. Polymer

2004, 45 (22), 7729–7737. https://doi.org/10.1016/j.polymer.2004.08.056.

(10) Wutticharoenwong, K.; Soucek, M. D. Synthesis of Tung-Oil-Based Reactive

Diluents. Macromol. Mater. Eng. 2010, 295 (12), 1097–1106.

https://doi.org/10.1002/mame.201000099.

(11) Wutticharoenwong, K.; Dziczkowski, J.; Soucek, M. D. Tung Based Reactive

Diluents for Alkyd Systems: Film Properties. Prog. Org. Coat. 2012, 73 (4), 283–

290. https://doi.org/10.1016/j.porgcoat.2011.03.017.

(12) Biermann, U.; Butte, W.; Holtgrefe, R.; Feder, W.; Metzger, J. O. Esters of Calendula

Oil and Tung Oil as Reactive Diluents for Alkyd Resins. Eur. J. Lipid Sci. Technol.

2010, 112 (1), 103–109. https://doi.org/10.1002/ejlt.200900142.

(13) Nalawade, P. P.; Mehta, B.; Pugh, C.; Soucek, M. D. Modified Soybean Oil as a

Reactive Diluent: Synthesis and Characterization. J. Polym. Sci. Part Polym. Chem.

2014, 52 (21), 3045–3059. https://doi.org/10.1002/pola.27352.

(14) Nalawade, P. P.; Soucek, M. D. Modified Soybean Oil as a Reactive Diluent: Coating

Performance. J. Coat. Technol. Res. 2015, 12 (6), 1005–1021.

https://doi.org/10.1007/s11998-015-9691-2.

(15) Ren, X.; Meng, L.; Soucek, M. Environmentally Friendly Coatings. In Biobased and

Environmental Benign Coatings; Tiwari, A., Galanis, A., Soucek, M. D., Eds.; John

Wiley & Sons, Inc.: Hoboken, NJ, USA, 2016; pp 183–223.

https://doi.org/10.1002/9781119185055.ch8.

174 (16) Alam, M.; Akram, D.; Sharmin, E.; Zafar, F.; Ahmad, S. Vegetable Oil Based Eco-

Friendly Coating Materials: A Review Article. Arab. J. Chem. 2014, 7 (4), 469–479.

https://doi.org/10.1016/j.arabjc.2013.12.023.

(17) Canvin, D. T. THE EFFECT OF TEMPERATURE ON THE OIL CONTENT AND

FATTY ACID COMPOSITION OF THE OILS FROM SEVERAL OIL SEED

CROPS. Can. J. Bot. 1965, 43 (1), 63–69. https://doi.org/10.1139/b65-008.

(18) Baughman, W. F.; Jamieson, G. S. THE CHEMICAL COMPOSITION OF SOYA

BEAN OIL. J. Am. Chem. Soc. 1922, 44 (12), 2947–2952.

https://doi.org/10.1021/ja01433a035.

(19) Montero de Espinosa, L.; Meier, M. A. R. Plant Oils: The Perfect Renewable

Resource for Polymer Science?! Eur. Polym. J. 2011, 47 (5), 837–852.

https://doi.org/10.1016/j.eurpolymj.2010.11.020.

(20) Orsavova, J.; Misurcova, L.; Ambrozova, J.; Vicha, R.; Mlcek, J. Fatty Acids

Composition of Vegetable Oils and Its Contribution to Dietary Energy Intake and

Dependence of Cardiovascular Mortality on Dietary Intake of Fatty Acids. Int. J. Mol.

Sci. 2015, 16 (12), 12871–12890. https://doi.org/10.3390/ijms160612871.

(21) Lewinska, A.; Zebrowski, J.; Duda, M.; Gorka, A.; Wnuk, M. Fatty Acid Profile and

Biological Activities of Linseed and Rapeseed Oils. Molecules 2015, 20 (12), 22872–

22880. https://doi.org/10.3390/molecules201219887.

(22) Bayrak, A.; Kiralan, M.; Ipek, A.; Arslan, N.; Cosge, B.; Khawar, K. M. Fatty Acid

Compositions of Linseed ( Linum Usitatissimum L.) Genotypes of Different Origin

Cultivated in Turkey. Biotechnol. Biotechnol. Equip. 2010, 24 (2), 1836–1842.

https://doi.org/10.2478/V10133-010-0034-2.

175 (23) Jokić, S.; Sudar, R.; Svilović, S.; Vidović, S.; Bilić, M.; Velić, D.; Jurković, V. Fatty

Acid Composition of Oil Obtained from Soybeans by Extraction with Supercritical

Carbon Dioxide. Czech J. Food Sci. 2013, 31 (No. 2), 116–125.

https://doi.org/10.17221/8/2012-CJFS.

(24) Patton, T. C. Alkyd Resin Technology: Formulating Techniques and Allied

Calculations; John Wiley & Sons, Inc.: New York, 1962.

(25) Martens, C. R. Alkyd Resins; Reinhold Publishing Corporation: New York, 1961.

(26) Coşge, B.; Gürbüz, B.; Kiralan, M. Oil Content and Fatty Acid Composition of Some

Safflower (Carthamus Tinctorius L.) Varieties Sown in Spring and Winter. 2007, 5.

(27) Chen, Y.-H.; Chen, J.-H.; Chang, C.-Y.; Chang, C.-C. Biodiesel Production from

Tung (Vernicia Montana) Oil and Its Blending Properties in Different Fatty Acid

Compositions. Bioresour. Technol. 2010, 101 (24), 9521–9526.

https://doi.org/10.1016/j.biortech.2010.06.117.

(28) White, K.; Lorenz, N.; Potts, T.; Roy Penney, W.; Babcock, R.; Hardison, A.; Canuel,

E. A.; Hestekin, J. A. Production of Biodiesel Fuel from Tall Oil Fatty Acids via High

Temperature Methanol Reaction. Fuel 2011, 90 (11), 3193–3199.

https://doi.org/10.1016/j.fuel.2011.06.017.

(29) McHughen, A. Flax (Linum Usitatissimum L.): In Vitro Studies. In Legumes and

Oilseed Crops I. Biotechnology in Agriculture and Forestry, vol 10.; Bajaj, Y. P. S.,

Ed.; Springer Berlin Heidelberg, 1990; Vol. 10, pp 502–514.

(30) Deffar, D.; Soucek, M. D. Synergistic Effect of Driers on Soybean Oil-Based Ceramer

Coatings. J. Coat. Technol. 2001, 73 (8), 95–104.

https://doi.org/10.1007/BF02698390.

176 (31) Soucek, M. D.; Khattab, T.; Wu, J. Review of Autoxidation and Driers. Prog. Org.

Coat. 2012, 73 (4), 435–454. https://doi.org/10.1016/j.porgcoat.2011.08.021.

(32) Sailer, R. A.; Soucek, M. D. Oxidizing Alkyd Ceramers. Prog. Org. Coat. 1998, 33

(1), 36–43. https://doi.org/10.1016/S0300-9440(98)00005-8.

(33) Mallégol, J.; Gardette, J.-L.; Lemaire, J. Long-Term Behavior of Oil-Based Varnishes

and Paints I. Spectroscopic Analysis of Curing Drying Oils. J. Am. Oil Chem. Soc.

1999, 76 (8), 967–976. https://doi.org/10.1007/s11746-999-0114-3.

(34) Mallégol, J.; Gardette, J.-L.; Lemaire, J. Long-Term Behavior of Oil-Based Varnishes

and Paints. Fate of Hydroperoxides in Drying Oils. J. Am. Oil Chem. Soc. 2000, 77

(3), 249–255. https://doi.org/10.1007/s11746-000-0041-5.

(35) Sailer, R. A.; Wegner, J. R.; Hurtt, G. J.; Janson, J. E.; Soucek, M. D. Linseed and

Sunflower Oil Alkyd Ceramers. Prog. Org. Coat. 1998, 33 (2), 117–125.

https://doi.org/10.1016/S0300-9440(98)00033-2.

(36) Mallégol, J.; Gardette, J.-L.; Lemaire, J. Long-Term Behavior of Oil-Based Varnishes

and Paints. Photo- and Thermooxidation of Cured Linseed Oil. J. Am. Oil Chem. Soc.

2000, 77 (3), 257–263. https://doi.org/10.1007/s11746-000-0042-4.

(37) Mallégol, J.; Gonon, J.; Lemaire, J.; Gardette, J.-L. Long-Term Behaviour of Oil-

Based Varnishes and Paints 4. Influence of Film Thickness on the Photooxidation.

Polym. Degrad. Stab. 2001, 72, 191–197.

(38) Falla, N. A. R. Linoleic Based Coatings: A Study of the Dry Film Structure. J. Coat.

Technol. 1992, 64 (815), 55.

(39) Meneghetti, S. M. P.; de Souza, R. F.; Monteiro, A. L.; de Souza, M. O. Substitution

of Lead Catalysts by Zirconium in the Oxidative Polymerization of Linseed Oil. Prog.

177

Org. Coat. 1998, 33 (3–4), 219–224. https://doi.org/10.1016/S0300-9440(98)00058-

7.

(40) Muizebelt, W. J.; Donkerbroek, J. J.; Nielen, M. W. F.; Hussem, J. B.; Biemond, M.

E. F.; Klaasen, R. P.; Zabel, K. H. Oxidative Crosslinking of Alkyd Resins Studied

with Mass Spectrometry and NMR Using Model Compounds. J. Coat. Technol. 1998,

70 (1), 83–93. https://doi.org/10.1007/BF02720501.

(41) Warzeska, S. T.; Zonneveld, M.; van Gorkum, R.; Muizebelt, W. J.; Bouwman, E.;

Reedijk, J. The Influence of Bipyridine on the Drying of Alkyd Paints: A Model

Study. Prog. Org. Coat. 2002, 44 (3), 243–248. https://doi.org/10.1016/S0300-

9440(02)00057-7.

(42) Oyman, Z. O.; Ming, W.; van der Linde, R.; van Gorkum, R.; Bouwman, E. Effect of

[Mn(Acac)3] and Its Combination with 2,2′-Bipyridine on the Autoxidation and

Oligomerisation of Ethyl Linoleate. Polymer 2005, 46 (6), 1731–1738.

https://doi.org/10.1016/j.polymer.2004.12.045.

(43) de Boer, J. W.; Wesenhagen, P. V.; Wenker, E. C. M.; Maaijen, K.; Gol, F.; Gibbs,

H.; Hage, R. The Quest for Cobalt-Free Alkyd Paint Driers. Eur. J. Inorg. Chem.

2013, 2013 (21), 3581–3591. https://doi.org/10.1002/ejic.201300205.

(44) Hancock, R. A.; Leeves, N. J.; Nicks, P. F. STUDIES IN AUTOXIDATION. PART

II. AN EVALUATION OF THE BY-PRODUCTS FORMED BY THE

AUTOXIDATION OF FATTY ACID MODIFIED ALKYD RESINS UNDER THE

INFLUENCE OF DIFFERENT PROMOTERS. 11.

(45) Bieleman, J. H. Progress in the Development of Cobalt-Free Drier System. Macromol

Symp 2002, 187, 811–822.

178 (46) Sward. Paint Testing Manual: Physical and Chemical Examination of Paints,

Varnishes, Lacquers, and Colors; ASTM International, 1972.

(47) Gardner, P. N. Method of and Means for Determining the Viscosities of Liquids.

2414864, January 28, 1947.

(48) Callahan, W. B.; Coe, F. B. Alkyd Resin Manufacturing Process by Using Water or

Steam to Reverse or Retard Gelation. 4,045,392, August 30, 1977.

(49) D01 Committee. Test Methods for Acid Value of Organic Coating Materials; D 1639-

90; ASTM International, 1996.

(50) Weiss, K. D. Paint and Coatings: A Mature Industry in Transition. Prog. Polym. Sci.

1997, 22 (2), 203–245. https://doi.org/10.1016/S0079-6700(96)00019-6.

(51) Kangas, S. L.; Jones, F. N. Binders for Higher-Solids Coatings. 1. Synthesis of Model

Alkyd Resins. J. Coat. Technol. 1987, 59 (744), 89–97.

(52) Kangas, S. L.; Jones, F. N. Binders for Higher-Solids Coatings. 2. Properties of Model

Alkyd Resins. J. Coat. Technol. 1987, 59 (744), 99–103.

(53) CHEN, D.-S.; Jones, F. N. Binders for Higher-Solids Coatings. III: Model Alkyd

Resins Modified by Liquid Crystalline Groups. JCT J. Coat. Technol. 1988, 60 (756),

39–45.

(54) Chen, D.-S.; Jones, F. N. Binders for Higher-Solids Coatings. IV. Liquid–Crystalline

Acrylic Lacquers. J. Appl. Polym. Sci. 1988, 36 (1), 141–163.

https://doi.org/10.1002/app.1988.070360112.

(55) Muizebelt, W. J.; Hubert, J. C.; Nielen, M. W. F.; Klaasen, R. P.; Zabel, K. H.

Crosslink Mechanisms of High-Solids Alkyd Resins in the Presence of Reactive

179 Diluents. Prog. Org. Coat. 2000, 40 (1–4), 121–130. https://doi.org/10.1016/S0300-

9440(00)00121-1.

(56) Tuck, N. Waterborne and Solvent Based Alkyds and Their End User Applications;

John Wiley & Sons, 2000.

(57) Whittingstall, P. PAINT EVALUATION USING RHEOLOGY. 8.

(58) Cognard, P. Chapter 2 - Equipment for the Application of Adhesives and Sealants:

Mixing, Metering, Coating or Applying the Adhesives. In Handbook of Adhesives

and Sealants; Cognard, P., Ed.; Adhesives and Sealants; Elsevier Science Ltd, 2006;

Vol. 2, pp 51–xxxvii. https://doi.org/10.1016/S1874-5695(06)80013-8.

(59) Zabel, K. H.; Klaasen, R. P.; Muizebelt, W. J.; Gracey, B. P.; Hallett, C.; Brooks, C.

D. Design and Incorporation of Reactive Diluents for Air-Drying High Solids Alkyd

Paints. Prog. Org. Coat. 1999, 10.

(60) Popadyuk, A.; Breuer, A.; Bahr, J.; Tarnavchyk, I.; Voronov, A.; Chisholm, B. J.

Sucrose Octaesters as Reactive Diluents for Alkyd Coatings. J. Coat. Technol. Res.

2018, 15 (3), 481–488. https://doi.org/10.1007/s11998-017-0016-5.

(61) Milligan, C. L.; Ward, R. J.; Martin, R. A. HEXADENOL AND HEXADIENAL

DERVATIVES. 3,520,839, July 21, 1970.

(62) Höhlein, P.; Meixner, J.; Pedain, J. COMPOSITIONS CONTAINING

OCTADIENYL ETHERS AS REACTIVE THINNERS. 5,252,648, October 12,

1993.

(63) Lligadas, G.; Ronda, J. C.; Galià, M.; Cádiz, V. Renewable Polymeric Materials from

Vegetable Oils: A Perspective. Mater. Today 2013, 16 (9), 337–343.

https://doi.org/10.1016/j.mattod.2013.08.016.

180 (64) Hoffmann, H. M. R. The Ene Reaction. Angew. Chem. Int. Ed. Engl. 1969, 8 (8), 556–

577. https://doi.org/10.1002/anie.196905561.

(65) Keung, E. C.; Alper, H. The Ene Synthesis. J. Chem. Educ. 1972, 49 (2), 97.

https://doi.org/10.1021/ed049p97.

(66) Nahm, S. H.; Cheng, H. N. Transition-State Geometry and Stereochemistry of the Ene

Reaction between Olefins and Maleic Anhydride. J. Org. Chem. 1986, 51 (26), 5093–

5100. https://doi.org/10.1021/jo00376a008.

(67) Biswas, A.; Sharma, B. K.; Willett, J. L.; Erhan, S. Z.; Cheng, H. N. Room-

Temperature Self-Curing Ene Reactions Involving Soybean Oil. Green Chem. 2008,

10 (3), 290. https://doi.org/10.1039/b712385j.

(68) Eren, T.; Ksefo?lu, S. H. Hydroxymethylation and Polymerization of Plant Oil

Triglycerides. J. Appl. Polym. Sci. 2004, 91 (6), 4037–4046.

https://doi.org/10.1002/app.13608.

(69) Can, E.; Wool, R. P.; Küsefoğlu, S. Soybean- and Castor-Oil-Based Thermosetting

Polymers: Mechanical Properties. J. Appl. Polym. Sci. 2006, 102 (2), 1497–1504.

https://doi.org/10.1002/app.24423.

(70) Benn, F. R.; Dwyer, J.; Chappell, I. The Ene Reaction of Maleic Anhydride with

Alkenes. J. Chem. Soc. Perkin Trans. 2 1977, No. 5, 533.

https://doi.org/10.1039/p29770000533.

(71) Sclavons, M.; Laurent, M.; Devaux, J.; Carlier, V. Maleic Anhydride-Grafted

Polypropylene: FTIR Study of a Model Polymer Grafted by Ene-Reaction. Polymer

2005, 46 (19), 8062–8067. https://doi.org/10.1016/j.polymer.2005.06.115.

181 (72) Zhao, Y.; Han, L.; Li, D.; Xu, W. The Mechanism of Grafting of Maleic Anhydride

onto Isotactic Polybutene-1 Assisted by Comonomers. Colloid Polym. Sci. 2017, 295

(3), 463–469. https://doi.org/10.1007/s00396-017-4018-8.

(73) Wen, J.; Wilkes, G. L. Organic/Inorganic Hybrid Network Materials by the Sol−Gel

Approach. Chem. Mater. 1996, 8 (8), 1667–1681.

https://doi.org/10.1021/cm9601143.

(74) Danks, A. E.; Hall, S. R.; Schnepp, Z. The Evolution of ‘Sol–Gel’ Chemistry as a

Technique for Materials Synthesis. Mater. Horiz. 2016, 3 (2), 91–112.

https://doi.org/10.1039/C5MH00260E.

(75) Ebelmen, M. Ann Chim Phys 1846, 16, 129.

(76) Hench, L. L.; West, J. K. The Sol-Gel Process. Chem. Rev. 1990, 90 (1), 33–72.

https://doi.org/10.1021/cr00099a003.

(77) Brinker, C. J.; Scherer, G. W. Sol-Gel Science: The Physics and Chemistry of Sol-Gel

Processing; Academic Press, 2013.

(78) Young, S. K. Overview of Sol-Gel Science and Technology; ARL-TR-2650; Army

Research Laboratory, 2002.

(79) Julbe, A.; Balzer, C.; Barthez, J. M.; Guizard, C.; Larbot, A.; Cot, L. Effect of Non-

Ionic Surface Active Agents on TEOS-Derived Sols, Gels and Materials. J. Sol-Gel

Sci. Technol. 1995, 4 (2), 89–97. https://doi.org/10.1007/BF00491675.

(80) Brinker, C. J. Hydrolysis and Condensation of Silicates: Effects on Structure. J. Non-

Cryst. Solids 1988, 100 (1–3), 31–50. https://doi.org/10.1016/0022-3093(88)90005-

1.

182 (81) Schmidt, H.; Scholze, H.; Kaiser, A. Principles of Hydrolysis and Condensation

Reaction of Alkoxysilanes. J. Non-Cryst. Solids 1984, 63, 1–11.

(82) Ni, H.; Skaja, A. D.; Sailer, R. A.; Soucek, M. D. Moisture-Curing Alkoxysilane-

Functionalized Isocyanurate Coatings. Macromol. Chem. Phys. 2000, 201 (6), 722–

732. https://doi.org/10.1002/(SICI)1521-3935(20000301)201:6<722::AID-

MACP722>3.0.CO;2-D.

(83) Sailer, R. A.; Soucek, M. D. Viscoelastic Properties of Alkyd Ceramers. J. Appl.

Polym. Sci. 1999, 73 (10), 2017–2028. https://doi.org/10.1002/(SICI)1097-

4628(19990906)73:10<2017::AID-APP20>3.0.CO;2-1.

(84) Wold, C. R.; Soucek, M. D. Viscoelastic and Thermal Properties of Linseed Oil-

Based Ceramer Coatings. Macromol. Chem. Phys. 2000, 201 (3), 382–392.

https://doi.org/10.1002/(SICI)1521-3935(20000201)201:3<382::AID-

MACP382>3.0.CO;2-9.

(85) Teng, G.; Soucek, M. D. Epoxidized Soybean Oil-Based Ceramer Coatings. J. Am.

Oil Chem. Soc. 2000, 77 (4), 381–387. https://doi.org/10.1007/s11746-000-0062-0.

(86) Teng, G.; Wegner, J. R.; Hurtt, G. J.; Soucek, M. D. Novel Inorganic/Organic Hybrid

Materials Based on Blown Soybean Oil with Sol–Gel Precursors. Prog. Org. Coat.

2001, 42 (1–2), 29–37. https://doi.org/10.1016/S0300-9440(01)00126-6.

(87) Tuman, S. J.; Soucek, M. D. Novel Inorganic/Organic Coatings Based on Linseed Oil

and Sunflower Oil with Sol-Gel Precursors. JCT J. Coat. Technol. 1996, 68 (854).

(88) Wold, C. R.; Soucek, M. D. Mixed Metal Oxide Inorganic/Organic Coatings. J. Coat.

Technol. 1998, 70 (7), 43–51. https://doi.org/10.1007/BF02720514.

183 (89) Zou, K.; Soucek, M. D. UV-Curable Organic-Inorganic Hybrid Film Coatings Based

on Epoxidized Cyclohexene Derivatized Linseed Oil. Macromol. Chem. Phys. 2004,

205 (15), 2032–2039. https://doi.org/10.1002/macp.200400115.

(90) Baghdachi, J.; Li, D.; LaForest, J. Isocyanate-Free Moisture Cure Coatings. J. Coat.

Technol. 2002, 74 (9), 81–87. https://doi.org/10.1007/BF02697977.

(91) Chang, W.-H.; Hartman, M. E.; Scriven, R. L. Ambient Temperature, Moisture-

Curable Acrylic-Silane Coating Compositions Having Improved Potlife. 4,043,953,

August 23, 1977.

(92) Emmerling, W.; Padola, T.; Unger, L.; Majolo, M. Moisture-Curing Alkoxysilane-

Terminated Polyurethanes. 5,554,709, September 10, 1996.

(93) Ni, H.; Skaja, A. D.; Soucek, M. D. Acid-Catalyzed Moisture-Curing

Polyurea/Polysiloxane Ceramer Coatings. Prog. Org. Coat. 2000, 40 (1–4), 175–184.

https://doi.org/10.1016/S0300-9440(00)00116-8.

(94) Munekata, S. Fluoropolymers as Coating Material. Prog. Org. Coat. 1988, 16 (2),

113–134. https://doi.org/10.1016/0033-0655(88)80010-4.

(95) Fluorinated Surfaces, Coatings, and Films; Castner, D. G., Grainger, D. W., Eds.;

ACS symposium series; American Chemical Society ; Oxford University Press:

Washington, DC : [New York], 2001.

(96) Sheppard, W. A.; Sharts, C. M. Organic Fluorine Chemistry. 1969.

(97) Banks, R. E.; Smart, B. E.; Tatlow, J. C. Organofluorine Chemistry: Principles and

Commercial Applications; Springer Science & Business Media, 2013.

(98) van de Grampel, R. D.; Ming, W.; van Gennip, W. J. H.; van der Velden, F.; Laven,

J.; Niemantsverdriet, J. W.; van der Linde, R. Thermally Cured Low Surface-Tension

184 Epoxy Films. Polymer 2005, 46 (23), 10531–10537.

https://doi.org/10.1016/j.polymer.2005.08.024.

(99) McKeen, L. W. Fluorinated Coatings and Finishes Handbook: The Definitive User’s

Guide; William Andrew, 2015.

(100) Anton, D. Surface-Fluorinated Coatings. Adv. Mater. 1998, 10 (15), 1197–1205.

https://doi.org/10.1002/(SICI)1521-4095(199810)10:15<1197::AID-

ADMA1197>3.0.CO;2-F.

(101) Barbucci, A.; Pedroni, E.; Perillon, J. L.; Cerisola, G. Organic Coatings for Pre-

Treated Galvanized Steel Technology. Prog. Org. Coat. 1996, 29 (1–4), 7–11.

https://doi.org/10.1016/S0300-9440(96)00665-0.

(102) Thanamongkollit, N.; Soucek, M. D. Synthesis and Properties of Acrylate

Functionalized Alkyds via a Diels–Alder Reaction. Prog. Org. Coat. 2012, 73 (4),

382–391. https://doi.org/10.1016/j.porgcoat.2011.02.004.

(103) Funke, W. New Aspects of Paint Films with Inhomogeneos Structures. Prog. Org.

Coat. 1973, 2, 289–313.

(104) Alyamac, E.; Gu, H.; Soucek, M. D. Interface-Driven Phase-Separated Coatings. J.

Coat. Technol. Res. 2014, 11 (5), 665–683. https://doi.org/10.1007/s11998-014-

9591-x.

(105) Zhao, J.; Millians, W.; Tang, S.; Wu, T.; Zhu, L.; Ming, W. Self-Stratified

Antimicrobial Acrylic Coatings via One-Step UV Curing. ACS Appl. Mater.

Interfaces 2015, 7 (33), 18467–18472. https://doi.org/10.1021/acsami.5b04633.

185 (106) Yagci, M. B.; Bolca, S.; Heuts, J. P. A.; Ming, W.; de With, G. Self-Stratifying

Antimicrobial Polyurethane Coatings. Prog. Org. Coat. 2011, 72 (3), 305–314.

https://doi.org/10.1016/j.porgcoat.2011.04.021.

(107) Bodkhe, R. B.; Stafslien, S. J.; Daniels, J.; Cilz, N.; Muelhberg, A. J.; Thompson,

S. E. M.; Callow, M. E.; Callow, J. A.; Webster, D. C. Zwitterionic Siloxane-

Polyurethane Fouling-Release Coatings. Prog. Org. Coat. 2015, 78, 369–380.

https://doi.org/10.1016/j.porgcoat.2014.07.011.

(108) Kuczyńska, H.; Langer, E.; Kamińska-Tarnawska, E.; Kulikov, D. A.; Indeikin, E.

A. Study of Self-Stratifying Compositions. J. Coat. Technol. Res. 2009, 6 (3), 345–

352. https://doi.org/10.1007/s11998-008-9148-y.

(109) Verkholantsev, V. V. Self‐stratifying Coatings for Industrial Applications. Pigment

Resin Technol. 2003, 32 (5), 300–306. https://doi.org/10.1108/03699420310497445.

(110) Baghdachi, J.; Perez, H.; Talapatcharoenkit, P.; Wang, B. Design and Development

of Self-Stratifying Systems as Sustainable Coatings. Prog. Org. Coat. 2015, 78, 464–

473. https://doi.org/10.1016/j.porgcoat.2014.06.017.

(111) Webster, D. C.; Ekin, A.; Sommer, S. ANCHORED POLYSILOXANE-

MODIFIED POLYURETHANE COATINGS AND USES THEREOF. US 8,299,200

B2, October 30, 2012.

(112) Beaugendre, A.; Degoutin, S.; Bellayer, S.; Pierlot, C.; Duquesne, S.; Casetta, M.;

Jimenez, M. Self-Stratifying Coatings: A Review. Prog. Org. Coat. 2017, 110, 210–

241. https://doi.org/10.1016/j.porgcoat.2017.03.011.

(113) Zahedi, S.; Zaarei, D.; Ghaffarian, S. R. Self-Stratifying Coatings: A Review. J.

Coat. Technol. Res. 2018, 15 (1), 1–12. https://doi.org/10.1007/s11998-017-9996-4.

186 (114) Walbridge, D. J. Self-Stratifying Coatings — an Overview of a European

Community Research Project. Prog. Org. Coat. 1996, 28 (3), 155–159.

https://doi.org/10.1016/0300-9440(95)00593-5.

(115) Toussaint, A. Self-Stratifying Coatings for Plastic Substrates (BRITE EURAM

PROJECT RI 1B 0246 C(H)). Prog. Org. Coat. 1996, 28 (3), 183–195.

https://doi.org/10.1016/0300-9440(95)00597-8.

(116) Sikka, M.; Singh, N.; Karim, A.; Bates, F. S.; Satija, S. K.; Majkrzak, C. F. Entropy-

Driven Surface Segregation in Block Copolymer Melts. Phys. Rev. Lett. 1993, 70 (3),

307–310. https://doi.org/10.1103/PhysRevLett.70.307.

(117) Glaris, P.; Coulon, J. F.; Dorget, M.; Poncin-Epaillard, F. Surface Migration of

Fluorinated Additive during the Curing of Epoxy Resin. Compos. Part B Eng. 2015,

73, 10–15. https://doi.org/10.1016/j.compositesb.2014.12.020.

(118) Thomas, R.; Durix, S.; Sinturel, C.; Omonov, T.; Goossens, S.; Groeninckx, G.;

Moldenaers, P.; Thomas, S. Cure Kinetics, Morphology and Miscibility of Modified

DGEBA-Based Epoxy Resin – Effects of a Liquid Rubber Inclusion. Polymer 2007,

48 (6), 1695–1710. https://doi.org/10.1016/j.polymer.2007.01.018.

(119) Ye, X.; Zuo, B.; Deng, M.; Hei, Y.; Ni, H.; Lu, X.; Wang, X. Surface Segregation

of Fluorinated Moieties on Poly(Methyl Methacrylate-Ran-2-Perfluorooctylethyl

Methacrylate) Films during Film Formation: Entropic or Enthalpic Influences. J.

Colloid Interface Sci. 2010, 349 (1), 205–214.

https://doi.org/10.1016/j.jcis.2010.05.042.

187 (120) Benjamin, S.; Carr, C.; Walbridge, D. J. Self-Stratifying Coatings for Metallic

Substrates. Prog. Org. Coat. 1996, 28 (3), 197–207. https://doi.org/10.1016/0300-

9440(95)00596-X.

(121) Jones, D. A. Principles and Prevention of Corrosion (2nd Edition), 2nd ed.;

Pearson, 1995.

(122) Koch, G.; Varney, J.; Thompson, N.; Moghissi, O.; Gould, M.; Payer, J.

International Measures of Prevention, Application, and Economics of Corrosion

Technologies Study; NACE International: 2016.; NACE International, 2016.

(123) Sørensen, P. A.; Kiil, S.; Dam-Johansen, K.; Weinell, C. E. Anticorrosive Coatings:

A Review. J. Coat. Technol. Res. 2009, 6 (2), 135–176.

https://doi.org/10.1007/s11998-008-9144-2.

(124) Torre, G. C. la; Espinosa-Medina, M. A.; Martinez-Villafane, A.; Gonzalez-

Rodriguez, J. G.; Castano, V. M. Study of Ceramic and Hybrid Coatings Produced by

the Sol-Gel Method for Corrosion Protection. Open Corros. J. 2009, 2 (1), 197–203.

https://doi.org/10.2174/1876503300902010197.

(125) Gervasio, D.; Song, I.; Payer, J. H. Determination of the Oxygen Reduction

Products on ASTM A516 Steel during Cathodic Protection. 14.

(126) Buchheit, R. G.; Cunningham, M.; Jensen, H.; Kendig, M. W.; Martinez, M. A. A

Correlation Between Salt Spray and Electrochemical Impedance Spectroscopy Test

Results for Conversion-Coated Aluminum Alloys. CORROSION 1998, 54 (1), 61–

72. https://doi.org/10.5006/1.3284829.

188 (127) Bonora, P. L.; Deflorian, F.; Fedrizzi, L. Electrochemical Impedance Spectroscopy

as a Tool for Investigating Underpaint Corrosion. Electrochimica Acta 1996, 41 (7–

8), 1073–1082. https://doi.org/10.1016/0013-4686(95)00440-8.

(128) White, C. C.; White, K. M.; Pickett, J. E. Service Life Prediction of Polymers and

Plastics Exposed to Outdoor Weathering; William Andrew, 2017.

(129) Handbook of Material Weathering - 5th Edition

https://www.elsevier.com/books/handbook-of-material-weathering/wypych/978-1-

895198-62-1 (accessed Mar 29, 2019).

(130) Wood, K. The Effect of Fluoropolymer Architecture on the Exterior Weathering of

Coatings. Macromol Symp 2002, 187, 469–479.

(131) Pappas, S. P. Weathering of Coatings - Formulation and Evaluation. Prog. Org.

Coat. 1989, 17 (2), 107–114. https://doi.org/10.1016/0033-0655(89)80017-2.

(132) Light Stabilizers Make the UV Protection of Environmentally Friendly Coatings

Easier https://www.coatingsworld.com/contents/view_experts-opinion/2011-12-

08/light-stabilizers-make-the-uv-protection-of-environmentally-friendly-coatings-

easier/ (accessed Mar 29, 2019).

(133) Ramirez, M.; Miller, K. R.; Soucek, M. D. Linking of Oligoesters Hydrolysis to

Polyurethane Coatings. J. Appl. Polym. Sci. 2014, 131 (9), n/a-n/a.

https://doi.org/10.1002/app.40198.

(134) Ramirez-Huerta, M.; Zvonkina, I. J.; Soucek, M. D. Phenyl Carbamate End-Capped

Oligoesters: A Model for Hydrolytic Stability of Ester-Based Urethanes. J. Coat.

Technol. Res. 2016, 13 (5), 781–793. https://doi.org/10.1007/s11998-016-9798-0.

189 (135) Soucek, M. D.; Johnson, A. H. New Intramolecular Effect Observed for Polyesters:

An Anomeric Effect. J. Coat. Technol. Res. 2004, 1 (2), 111–116.

https://doi.org/10.1007/s11998-004-0005-3.

(136) Dziczkowski, J.; Soucek, M. D. Factors Influencing the Stability and Film

Properties of Acrylic/Alkyd Water-Reducible Hybrid Systems Using a Response

Surface Technique. Prog. Org. Coat. 2012, 73 (4), 330–343.

https://doi.org/10.1016/j.porgcoat.2011.03.005.

(137) Johnson, A. H.; Wegner, J.; Soucek, M. D. Hydrolytic Stability of Oligoesters:

Comparison of Steric with Anchimeric Effects. Eur. Polym. J. 2004, 40 (12), 2773–

2781. https://doi.org/10.1016/j.eurpolymj.2004.07.028.

(138) Soucek, M. D.; Johnson, A. H.; Meemken, L. E. Effect of Additional Hydroxyl

Functionalities on the Hydrolytic Stability of Oligoesters. Macromol. Chem. Phys.

2004, 205 (1), 35–41. https://doi.org/10.1002/macp.200350084.

(139) Hsiao, Y.-C.; Hill, L. W.; Pappas, S. P. Reversible Amine Solubilization of Cured

Siloxane Polymers. J. Appl. Polym. Sci. 1975, 19 (10), 2817–2820.

https://doi.org/10.1002/app.1975.070191017.

(140) Pappas, S. P.; Just, R. L. Aminolysis of Crosslinked Polysiloxanes: Tautomeric

Catalysis by 2-Pyridone. J. Polym. Sci. Polym. Chem. Ed. 1980, 18 (2), 527–531.

https://doi.org/10.1002/pol.1980.170180213.

(141) Bauer, D. R. Interpreting Weathering Acceleration Factors for Automotive

Coatings Using Exposure Models. Polym. Degrad. Stab. 2000, 69 (3), 307–316.

https://doi.org/10.1016/S0141-3910(00)00074-4.

190 (142) Ekstedt, J.; Kungliga Tekniska högskolan; Institutionen för byggvetenskap. Studies

on the Barrier Properties of Exterior Wood Coatings, Stockholm, 2002.

(143) Brunner, S.; Richner, P.; Müller, U.; Guseva, O. Accelerated Weathering Device

for Service Life Prediction for Organic Coatings. Polym. Test. 2005, 24 (1), 25–31.

https://doi.org/10.1016/j.polymertesting.2004.08.001.

(144) Skaja, A.; Fernando, D.; Croll, S. Mechanical Property Changes and Degradation

during Accelerated Weathering of Polyester-Urethane Coatings. J. Coat. Technol.

Res. 2006, 3 (1), 41–51. https://doi.org/10.1007/s11998-006-0004-7.

(145) Yang, X. F.; Tallman, D. E.; Bierwagen, G. P.; Croll, S. G.; Rohlik, S. Blistering

and Degradation of Polyurethane Coatings under Different Accelerated Weathering

Tests. Polym. Degrad. Stab. 2002, 77 (1), 103–109. https://doi.org/10.1016/S0141-

3910(02)00085-X.

(146) Kienle, R. H.; Ferguson, C. S. Alkyd Resins as Film-Forming Materials. Ind. Eng.

Chem. 1929, 21 (4), 349–352. https://doi.org/10.1021/ie50232a022.

(147) Salata, R. R.; Pellegrene, B.; Soucek, M. D. Migration of Fluorinated Alkyd and

Fluorinated Tung Oil Additives for Partially Self-Stratifying Coatings. Prog. Org.

Coat. 2019, 133, 406–417.

(148) Carr, C.; Wallstöm, E. Theoretical Aspects of Self-Stratification. Prog. Org. Coat.

1996, 28 (3), 161–171. https://doi.org/10.1016/0300-9440(95)00594-3.

(149) Salata, R. R.; Pellegrene, B.; Soucek, M. D. Synthesis and Properties of a High

Solids Triethoxysilane-Modified Alkyd Coatings. Prog. Org. Coat. 2019, 133, 340–

349.

(150) Witucki, G. L. A Silane Primer: Chemistry and Applications of AIkoxy Silanes. 4.

191 (151) Sabata, A.; Ooij, W. J. V.; Koch, R. J. The Interphase in Painted Metals Pretreated

by Functional Silanes. J. Adhes. Sci. Technol. 1993, 7 (11), 1153–1170.

https://doi.org/10.1163/156856193X00024.

(152) Walker, P. Organo Silanes as Adhesion Promoters for Organic Coatings. P Walk. J

Coat. Technol 52670 49-611980 Nov 1980.

(153) Sathyanarayana, M. N.; Yaseen, M. Role of Promoters in Improving Adhesions of

Organic Coatings to a Substrate. Prog. Org. Coat. 1995, 26 (2), 275–313.

https://doi.org/10.1016/0300-9440(95)00572-2.

(154) Rider, A. N.; Brack, N.; Andres, S.; Pigram, P. J. The Influence of Hydroxyl Group

Concentration on Epoxy–Aluminium Bond Durability. J. Adhes. Sci. Technol. 2004,

18 (10), 1123–1152. https://doi.org/10.1163/1568561041581298.

(155) D01 Committee. Test Methods for Hydroxyl Value of Fatty Oils and Acids; D 1957-

86; ASTM International, 2001.

(156) D01 Committee. Test Methods for Hardness of Organic Coatings by Pendulum

Damping Tests; D 4366-16; ASTM International. https://doi.org/10.1520/D4366-16.

(157) D01 Committee. Test Method for Film Hardness by Pencil Test; D 3363-05; ASTM

International. https://doi.org/10.1520/D3363-05R11E02.

(158) D01 Committee. Practice for Assessing the Solvent Resistance of Organic Coatings

Using Solvent Rubs; D5402-15; ASTM International. https://doi.org/10.1520/D5402-

15.

(159) D01 Committee. Test Method for Pull-Off Strength of Coatings Using Portable

Adhesion Testers; D 4541-17; ASTM International. https://doi.org/10.1520/D4541-

17.

192 (160) D01 Committee. Test Methods for Measuring Adhesion by Tape Test; D 3359-17;

ASTM International. https://doi.org/10.1520/D3359-17.

(161) D01 Committee. Test Method for Resistance of Organic Coatings to the Effects of

Rapid Deformation (Impact); D 2794-93; ASTM International.

https://doi.org/10.1520/D2794-93R19.

(162) E12 Committee. Practice for Identification of Instrumental Methods of Color or

Color-Difference Measurement of Materials; ASTM International.

https://doi.org/10.1520/E0805-12AR17.

(163) E12 Committee. Test Method for Specular Gloss; ASTM International.

https://doi.org/10.1520/D0523-14R18.

(164) D01 Committee. Practice for Xenon-Arc Exposures of Paint and Related Coatings;

ASTM International. https://doi.org/10.1520/D6695-16.

(165) Sinha, A.; Islam Khan, N.; Das, S.; Zhang, J.; Halder, S. Effect of Reactive and

Non-Reactive Diluents on Thermal and Mechanical Properties of Epoxy Resin. High

Perform. Polym. 2018, 30 (10), 1159–1168.

https://doi.org/10.1177/0954008317743307.

(166) Sharma, G.; Bala, R. Digital Color Imaging Handbook; CRC Press, 2002.

(167) Faucheu, J.; Wood, K. A.; Sung, L.-P.; Martin, J. W. Relating Gloss Loss to

Topographical Features of a PVDF Coating. J. Coat. Technol. Res. 2006, 3 (1), 29–

39. https://doi.org/10.1007/s11998-006-0003-8.

(168) Kanai, T.; Mahato, T. K.; Kumar, D. Synthesis and Characterization of Novel

Silicone Acrylate–Soya Alkyd Resin as Binder for Long Life Exterior Coatings. Prog.

Org. Coat. 2007, 58 (4), 259–264. https://doi.org/10.1016/j.porgcoat.2006.11.002.

193 (169) Kämpf, G.; Sommer, K.; Zirngiebl, E. Studies in Accelerated Weathering: Part I.

Determination of the Activation Spectrum of Photodegradation in Polymers. Prog.

Org. Coat. 1991, 19 (1), 69–77. https://doi.org/10.1016/0033-0655(91)80011-7.

(170) Nikolic, M.; Nguyen, H. D.; Daugaard, A. E.; Löf, D.; Mortensen, K.; Barsberg, S.;

Sanadi, A. R. Influence of Surface Modified Nano Silica on Alkyd Binder before and

after Accelerated Weathering. Polym. Degrad. Stab. 2016, 126, 134–143.

https://doi.org/10.1016/j.polymdegradstab.2016.02.006.

(171) Arkles, B. Hydrophobicity, Hydrophilicity and Silane Surface Modification. 84.

(172) Castellano, M.; Conzatti, L.; Turturro, A.; Costa, G.; Busca, G. Influence of the

Silane Modifiers on the Surface Thermodynamic Characteristics and Dispersion of

the Silica into Elastomer Compounds. J. Phys. Chem. B 2007, 111 (17), 4495–4502.

https://doi.org/10.1021/jp0702144.

(173) Shi, A.; Koka, S.; Ullett, J. Performance Evaluation on the Weathering Resistance

of Two USAF Coating Systems (Standard 85285 Topcoat versus Fluorinated APC

Topcoat) via Electrochemical Impedance Spectroscopy. Prog. Org. Coat. 2005, 52

(3), 196–209. https://doi.org/10.1016/j.porgcoat.2004.11.005.

(174) Birkmeyer, W. J.; Lewarchik, R. J.; Klanica, J. A.; Anderson, C. C. Coating

Compositions Containing Organic Alcoholic Reactive Diluents. US4314918A,

February 9, 1982.

(175) Pathan, S.; Ahmad, S. Synergistic Effects of Linseed Oil Based Waterborne Alkyd

and 3-Isocynatopropyl Triethoxysilane: Highly Transparent, Mechanically Robust,

Thermally Stable, Hydrophobic, Anticorrosive Coatings. ACS Sustain. Chem. Eng.

2016, 4 (6), 3062–3075. https://doi.org/10.1021/acssuschemeng.6b00024.

194

(176) Zvonkina, I.; Soucek, M. Inorganic–Organic Hybrid Coatings: Common and New

Approaches. Curr. Opin. Chem. Eng. 2016, 11, 123–127.

https://doi.org/10.1016/j.coche.2016.01.008.

(177) Wang, D.; Bierwagen, Gordon. P. Sol–Gel Coatings on Metals for Corrosion

Protection. Prog. Org. Coat. 2009, 64 (4), 327–338.

https://doi.org/10.1016/j.porgcoat.2008.08.010.

(178) Tohge, N. Formation of Si02-Based Coatings by the Sol-Gel Method and Their

Effects on Water Vapour Permeability of Polyimide Films. 3.

(179) Gilberts, J.; Tinnemans, A. H. A.; Hogerheide, M. P.; Koster, T. P. M. UV Curable

Hard Transparent Hybrid Coating Materials on Polycarbonate Prepared by the Sol-

Gel Method. 7.

(180) BRINKER, C.; FRYE, G.; ASHLEY, C.; SMITH, D.; DAVIS, P.; HIETALA, S.

SOL-GEL ROUTES TO CONTROLLED PORE-SIZE AMORPHOUS OXIDES. In

ABSTRACTS OF PAPERS OF THE AMERICAN CHEMICAL SOCIETY; AMER

CHEMICAL SOC 1155 16TH ST, NW, WASHINGTON, DC 20036, 1991; Vol. 201,

pp 348–INOR.

(181) Brinker, C. J.; Hurd, A. J.; Schunk, P. R.; Frye, G. C.; Ashley, C. S. Review of Sol-

Gel Thin Film Formation. J. Non-Cryst. Solids 1992, 147–148, 424–436.

https://doi.org/10.1016/S0022-3093(05)80653-2.

(182) Greenbank, G. R.; Holm, G. E. Some Factors Concerned in the Autoxidation of

Fats. Ind. Eng. Chem. 1924, 16 (6), 598–601. https://doi.org/10.1021/ie50174a024.

(183) Minemoto, Y.; Adachi, S.; Matsuno, R. Effect of Relative Humidity during Storage

on the Autoxidation of Linoleic Acid Encapsulated with a Polysaccharide by Hot-

195 Air-Drying and Freeze-Drying. Food Sci. Technol. Res. 2001, 7 (1), 91–93.

https://doi.org/10.3136/fstr.7.91.

(184) Wool, R. P. Properties of Triglyceride-Based Thermosts. In Bio-Based Polymers

and Composites.

(185) DiMarzio, E. A. The Entropy Theory of Glass Formation after 40 Years. Comput.

Mater. Sci. 1995, 4, 317–324.

(186) DiMarzio, E. A. The Glass Temperature of Polymer Blends. Polymer 1990, 31,

2294–2298.

(187) DiMarzio, E. A. On the Second-Order Transition of a Rubber. J. Res. Natl. Bur.

Stand. - Phys. Chem. 1964, 68A (6), 611–617.

(188) DiMarzio, E. A.; Guttman, C. M. The Glass Temperature of Polymer Rings.

Macromolecules 1987, 20, 1403–1407.

(189) Hale, A.; Macosko, C. W.; Bair, H. E. Glass Transition Temperature as a Function

of Conversion in Thermosetting Polymers. Macromolecules 1991, 24 (9), 2610–2621.

https://doi.org/10.1021/ma00009a072.

(190) Arkles, B. Commercial Applications of Sol-Gel-Derived Hybrid Materials. MRS

Bull. 2001, 26 (05), 402–408. https://doi.org/10.1557/mrs2001.94.

(191) Lange, R. M.; Bartley, S. L.; Ollinger, C. G.; Hogan, J. M. MALEATED

VEGETABLE OILS AND DERIVATIVES, AS SELF-EMULSIFYING

LUBRICANTS IN METALWORKING. US 2009/0209441 A1, August 20, 2009.

(192) Wool, R.; Sun, X. S. Bio-Based Polymers and Composites; Elsevier, 2011.

196 (193) Thames, S. F.; Rawlins, J. W.; Mendon, S. K.; Johnson, E. N.; Yu, Z.

FUNCTIONALIZED VEGETABLE OIL DERIVATIVES, LATEX

COMPOSITIONS AND TEXTLE FINISHES. US 2006/0236467 A1, 2006.

(194) Anton, D. R.; Kirchner, J. R.; Jr, W. W. B. Coatings Containing Fluorinated Esters.

US5859126A, January 12, 1999.

(195) Kirchner, J. R.; Jr, W. W. B. Coatings Containing Fluorinated Esters.

US5670573A, September 23, 1997.

(196) Thomas, R. R.; Glaspey, D. F.; DuBois, D. C.; Kirchner, J. R.; Anton, D. R.; Lloyd,

K. G.; Stika, K. M. Solution Properties of Partially Fluorinated Unsaturated Esters

and Surface Properties of Polyester Coatings Containing Partially Fluorinated

Unsaturated Esters as Cross-Linkers. Langmuir 2000, 16 (17), 6898–6905.

https://doi.org/10.1021/la0003600.

(197) Yoon, S. C.; Ratner, B. D. Surface Structure of Segmented Poly(Ether Urethanes)

and Poly(Ether Urethane Ureas) with Various Perfluoro Chain Extenders. An x-Ray

Photoelectron Spectroscopic Investigation. Macromolecules 1986, 19 (4), 1068–

1079. https://doi.org/10.1021/ma00158a023.

(198) Yoon, S. C.; Ratner, B. D.; Ivan, B.; Kennedy, J. P. Surface and Bulk Structure of

Segmented Poly(Ether Urethanes) with Perfluoro Chain Extenders. 5. Incorporation

of Poly(Dimethylsiloxane) and Polyisobutylene Macroglycols. Macromolecules

1994, 27 (6), 1548–1554. https://doi.org/10.1021/ma00084a041.

(199) Chapman, T. M.; Marra, K. G. Determination of Low Critical Surface Tensions of

Novel Fluorinated Poly(Amide Urethane) Block Copolymers. 2. Fluorinated Soft-

197 Block Backbone and Side Chains. Macromolecules 1995, 28 (6), 2081–2085.

https://doi.org/10.1021/ma00110a047.

(200) Camille M. Kassis; Jack K. Steehler; Douglas E. Betts; Zhibin Guan; Timothy J.

Romack; Joseph M. DeSimone, * and; Linton*, R. W. XPS Studies of Fluorinated

Acrylate Polymers and Block Copolymers with Polystyrene

https://pubs.acs.org/doi/full/10.1021/ma951782x (accessed Mar 26, 2019).

https://doi.org/10.1021/ma951782x.

(201) Jianguo Wang; Guoping Mao; Christopher K. Ober, * and; Kramer, E. J. Liquid

Crystalline, Semifluorinated Side Group Block Copolymers with Stable Low Energy

Surfaces: Synthesis, Liquid Crystalline Structure, and Critical Surface Tension

https://pubs.acs.org/doi/full/10.1021/ma961412o (accessed Mar 26, 2019).

https://doi.org/10.1021/ma961412o.

(202) Richard R. Thomas, *; Douglas R. Anton, †; William F. Graham, ‡; Michael J.

Darmon, ‡ and; Stika§, K. M. Films Containing Reactive Mixtures of

Perfluoroalkylethyl Methacrylate Copolymers and Fluorinated Isocyanates: Synthesis

and Surface Properties https://pubs.acs.org/doi/full/10.1021/ma970652g (accessed

Mar 26, 2019). https://doi.org/10.1021/ma970652g.

(203) Richard R. Thomas, *; Douglas R. Anton, †; William F. Graham, ‡; Michael J.

Darmon, ‡; Bryan B. Sauer, §; Katherine M. Stika, ǁ and; Swartzfagerǁ, D. G.

Preparation and Surface Properties of Acrylic Polymers Containing Fluorinated

Monomers https://pubs.acs.org/doi/full/10.1021/ma9618686 (accessed Mar 26,

2019). https://doi.org/10.1021/ma9618686.

198 (204) Schmidt, D. L.; Coburn, C. E.; DeKoven, B. M.; Potter, G. E.; Meyers, G. F.;

Fischer, D. A. Water-Based Non-Stick Hydrophobic Coatings. Nature 1994, 368

(6466), 39. https://doi.org/10.1038/368039a0.

(205) Weisz, P. B. Basic Choices and Contraints on Long-Term Energy Supplies. Phys

Today 2004, 47–52.

(206) Williams, C.; Hillmyer, M. Polymers from Renewable Resources: A Perspective

for a Special Issue of Polymer Reviews. Polym. Rev. 2008, 48 (1), 1–10.

https://doi.org/10.1080/15583720701834133.

(207) Habib, F.; Bajpai, M. SYNTHESIS AND CHARACTERIZATION OF

ACRYLATED EPOXIDIZED SOYBEAN OIL FOR UV CURED COATINGS. 10.

(208) Grishchuk, S.; Karger-Kocsis, J. Hybrid Thermosets from Vinyl Ester Resin and

Acrylated Epoxidized Soybean Oil (AESO). Express Polym. Lett. 2011, 5 (1), 2–11.

https://doi.org/10.3144/expresspolymlett.2011.2.

(209) Zhang, Y.; Li, Y.; Wang, L.; Gao, Z.; Kessler, M. R. Synthesis and Characterization

of Methacrylated Eugenol as a Sustainable Reactive Diluent for a Maleinated

Acrylated Epoxidized Soybean Oil Resin. ACS Sustain. Chem. Eng. 2017, 5 (10),

8876–8883. https://doi.org/10.1021/acssuschemeng.7b01673.

(210) Khot, S. N.; Lascala, J. J.; Can, E.; Morye, S. S.; Williams, G. I.; Palmese, G. R.;

Kusefoglu, S. H.; Wool, R. P. Development and Application of Triglyceride-Based

Polymers and Composites. J. Appl. Polym. Sci. 2001, 82 (3), 703–723.

https://doi.org/10.1002/app.1897.

(211) Campanella, A.; Zhan, M.; Watt, P.; Grous, A. T.; Shen, C.; Wool, R. P.

Triglyceride-Based Thermosetting Resins with Different Reactive Diluents and Fiber

199 Reinforced Composite Applications. Compos. Part Appl. Sci. Manuf. 2015, 72, 192–

199. https://doi.org/10.1016/j.compositesa.2015.02.009.

(212) Lu, J.; Khot, S.; Wool, R. P. New Sheet Molding Compound Resins from Soybean

Oil. I. Synthesis and Characterization. Polymer 2005, 46 (1), 71–80.

https://doi.org/10.1016/j.polymer.2004.10.060.

(213) Dai, J.; Ma, S.; Wu, Y.; Han, L.; Zhang, L.; Zhu, J.; Liu, X. Polyesters Derived

from Itaconic Acid for the Properties and Bio-Based Content Enhancement of

Soybean Oil-Based Thermosets. Green Chem. 2015, 17 (4), 2383–2392.

https://doi.org/10.1039/C4GC02057J.

(214) EBECRYL® 860 https://www.allnex.com/en/product/ecbddca9-4099-41c3-940d-

2b8b0110169b/ebecryl-860 (accessed Jan 30, 2019).

(215) Tran, P.; Seybold, K.; Graiver, D.; Narayan, R. Free Radical Meleation of Soybean

Oil via a Single-Step Process. J. Am. Oil Chem. Soc. 2005, 82 (3), 189–194.

https://doi.org/10.1007/s11746-005-5171-7.

(216) Yang, L.; Zhang, F.; Endo, T.; Hirotsu, T. Microstructure of Maleic Anhydride

Grafted Polyethylene by High-Resolution Solution-State NMR and FTIR

Spectroscopy. Macromolecules 2003, 36 (13), 4709–4718.

https://doi.org/10.1021/ma020527r.

(217) Santos, E. A. G. de los; Gonzalez, M. J. L.; Gonzalez, M. C. Modification of

Polypropylene with Maleic Anhydride: Ultrasonic Irradiation Effects. J. Appl. Polym.

Sci. 1998, 68 (1), 45–52. https://doi.org/10.1002/(SICI)1097-

4628(19980404)68:1<45::AID-APP5>3.0.CO;2-N.

200 (218) Shechter, L.; Wynstra, J. Glycidyl Ether Reactions with Alcohols, Phenols,

Carboxylic Acids, and Acid Anhydrides. Ind. Eng. Chem. 1956, 48 (1), 86–93.

https://doi.org/10.1021/ie50553a028.

(219) Stutz, H.; Illers, K.-H.; Mertes, J. A Generalized Theory for the Glass Transition

Temperature of Crosslinked and Uncrosslinked Polymers. J. Polym. Sci. Part B

Polym. Phys. 1990, 28 (9), 1483–1498.

https://doi.org/10.1002/polb.1990.090280906.

(220) DiBenedetto, A. T. Prediction of the Glass Transition Temperature of Polymers: A

Model Based on the Principle of Corresponding States. J. Polym. Sci. Part B Polym.

Phys. 1987, 25 (9), 1949–1969. https://doi.org/10.1002/polb.1987.090250914.

(221) Vendamme, R.; Olaerts, K.; Gomes, M.; Degens, M.; Shigematsu, T.; Eevers, W.

Interplay Between Viscoelastic and Chemical Tunings in Fatty-Acid-Based Polyester

Adhesives: Engineering Biomass toward Functionalized Step-Growth Polymers and

Soft Networks. Biomacromolecules 2012, 13 (6), 1933–1944.

https://doi.org/10.1021/bm300523e.

(222) Teeter, H. M.; Geerts, M. J.; Cowan, J. C. Polymerization of Drying Oils. III. Some

Observations on Reaction of Maleic Anhydride with Methyl Oleate and Methyl

Linoleate. J. Am. Oil Chem. Soc. 1948, 25 (5), 158–162.

https://doi.org/10.1007/BF02631517.

(223) Muturi, P.; Wang, D.; Dirlikov, S. Epoxidized Vegetable Oils as Reactive Diluents

I. Comparison of Vernonia, Epoxidized Soybean and Epoxidized Linseed Oils. Prog.

Org. Coat. 1994, 25 (1), 85–94. https://doi.org/10.1016/0300-9440(94)00504-4.

201 (224) Chen, J.; Soucek, M. D.; Simonsick, W. J.; Celikay, R. W. Synthesis and

Photopolymerization of Norbornyl Epoxidized Linseed Oil. Polymer 2002, 43 (20),

5379–5389. https://doi.org/10.1016/S0032-3861(02)00404-4.

(225) Gaddam, S. K.; Palanisamy, A. Anionic Waterborne Polyurethane Dispersions

from Maleated Cotton Seed Oil Polyol Carrying Ionisable Groups. Colloid Polym.

Sci. 2016, 294 (2), 347–355. https://doi.org/10.1007/s00396-015-3787-1.

202 APPENDICES

203 APPENDIX A:

NMR SPECTRA OF SOYBEAN OIL AND MALEATED SOYBEAN OIL

SBO_greencap_750data_04222019_proton_for2d.esp 1.26 E 1.31

O

O A O D G J B O C E F H I O J A O 0.90

C I 0.89 2.05 2.06 2.31 D F,H B, G 1.35 1.36 2.32 1.61 2.77

A 1.62 0.90 2.30 5.34 0.88 1.61 1.37 2.04 2.07 5.35 2.02 5.34 2.01 5.38 2.78 2.77 4.29 2.33 4.15 4.29 4.16 5.36 4.30 4.15 4.31 5.33 4.14 5.39 5.32 5.27 1.38 0.98 2.81 7.27 2.08 0.08 0.99

9.69 4.73 4.41 6.94 11.07 7.35 58.93 9.00

8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0 -0.5 ChemicalShift(ppm) Figure A.1. 1H NMR spectrum of SBO.

O P R SBO_CARBON_01.espA O O G D J 76.99 76.74 B O Q F 77.24 O C E H I

A O F R

27.18 H P E 31.50 22.54 G 25.61 J 29.32 14.03 130.19 C D

Q 127.88

B A 24.82 34.01 173.17 62.08

34.17 I 172.76 68.89 173.21 14.23 127.10 131.92 20.52 127.74 31.76

176 168 160 152 144 136 128 120 112 104 96 88 80 72 64 56 48 40 32 24 16 ChemicalShift(ppm) Figure A.2. 13C NMR spectrum of soybean oil.

204 1H NMR, MA-SBO HMBC, MA-SBO

O O O O K L M

O coupling with K, L, M

M K, L

Figure A.3. 1H NMR and 2D HMBC NMR spectra of MA-SBO.

HSQC, MA-SBO HSQC, SBO

O O O K L M

K, L

M

Figure A.4. 2D HSQC NMR spectra of MA-SBO (left) and SBO (right). 205 HMBC, MA-SBO HMBC, SBO

O O O K L N M

N coupling with unsaturated carbons

Figure A.5. 2D HMBC NMR spectra of MA-SBO (right) and SBO (left).

206