©2014

LONGHE ZHANG

ALL RIGHTS RESERVED SUPRAMOLECULAR BLOCK COPOLYMERS

VIA IONIC INTERACTIONS

A Dissertation

Presented to

The Graduate Faculty of The University of Akron

In Partial Fulfillment

of the Requirements for the Degree

Doctor of Philosophy

Longhe Zhang

August, 2014

SUPRAMOLECULAR BLOCK COPOLYMERS

VIA IONIC INTERACTIONS

Longhe Zhang

Dissertation

Approved: Accepted:

______Advisor Department Chair Dr. Robert A. Weiss Dr. Robert A. Weiss

______Co-advisor Dean of the College Dr. Kevin A. Cavicchi Dr. Stephen Z. D. Cheng

______Committee Member Dean of the Graduate School Dr. Alamgir Karim Dr. George R. Newkome

______Committee Member Date Dr. Coleen Pugh

______Committee Member Dr. Wiley J. Youngs

ii

ABSTRACT

Supramolecular block copolymers, which are the supramolecular analog of covalently-bonded block copolymers, consist of individual polymer blocks connected by non-covalent bonds. They can be produced by self-assembly of telechelic oligomers or polymers with complementary end-groups, such that a variety of block combinations may be achieved by simple mixing of the appropriate polymers.

Supramolecular block copolymers are advantageous for fabricating nanostructured functional materials, since they can exhibit morphologies mimicking conventional covalently-bonded block copolymers and the reversible nature of the supramolecular

bonds between blocks allows for unique responses to external stimuli.

In the first part, a supramolecular multiblock copolymer was synthesized by

mixing two telechelic oligomers, α,ω-sulfonated polystyrene, derived from reversible

addition−fragmentation chain-transfer (RAFT) polymerization, and α,ω-amino- polyisobutylene, prepared by cationic polymerization. Proton transfer from the sulfonic acid to the amine formed ionic bonds that produced a multiblock copolymer that formed free-standing flexible films. Small angle X-ray scattering characterization showed a lamellar morphology, whose domain spacing was consistent with the formation of a multiblock copolymer based on comparison to the chain dimensions. A reversible order-disorder transition occurred between 190°C and 210°C, but the sulfonic acid and amine functional groups decomposed at those elevated temperatures.

For high non-linear strains, the dynamic modulus, G’, decreased by nearly an order of

iii

magnitude and the loss modulus, but recovered to the original values once the strain

was reduced to 1%.

In the second part, two groups of RAFT agents that contain either quaternary

ammonium or quaternary phosphonium groups were prepared. At first, a series of

trithiocarbonate RAFT agents containing quaternary ammonium functionality in the

“R-group” were synthesized. The synthetic route involves the optimized synthesis of

4-(bromomethyl)-N,N,N-trialkyl benzyl ammonium bromide compounds, which were subsequently reacted with the alkyl trithiocarbonate anion to directly produce the

trithiocarbonate RAFT agent. However, quaternary ammonium group partially

degraded when the RAFT agents were used in polymerizations at 120 oC. This issue

was overcome by using lower polymerization temperature. On the other hand,

quaternary phosphonium-containing, trithiocarbonate RAFT agents were also

synthesized via similar synthetic method. Thermal stabilities of RAFT-PR3 were

enhanced compared to their ammonium analogues, which significantly improved the

retention of the cationic end-functionality of the polystyrene obtained at 120 oC. For

both classes of RAFT agents, the crude polystyrene can be further purified via column

chromatography to afford high purity hemi-telechelic cationomers.

In the third part, matrix-assisted laser desorption ionization time-of-flight mass

spectrometry (MALDI-ToF MS) was used to quantify the sulfonation level and

sulfonation distribution of sulfonated polystyrene ionomers prepared by homogeneous

solution sulfonation. The sulfonation levels obtained by MALDI ToF-MS and acid- base titration were compared, and the sulfonate distributions determined by MALDI-

ToF MS were compared with theoretical random distributions. The results indicate that the sulfonation reaction used produces a sample with a random sulfonate distribution.

iv

DEDICATION

To my parents, my brother,

and my wife Zhouying He

v

ACKNOWLEDGEMENTS

I am very grateful to all people who have provided me with advice,

encouragement and support over my graduate career.

First of all, I would like to thank my advisors, Professor Robert A. Weiss and

Professor Kevin A. Cavicchi, for providing me guidance and inspiration throughout

this journey. It has been a great pleasure working with both of you. I have been

enjoying my research because of all the freedom, opportunities and trust that you gave

me. You are always willing to share your experience, knowledge and wisdom. From

your advice and suggestions, I have learned many things that are essential for being

an excellent scientist.

I would like to thank my committee members, Professor Alamgir Karim,

Professor Coleen Pugh and Professor Wiley Youngs. I appreciate all the time you

spent on this dissertation and all the advice you provided to me.

I am also grateful to the group members in both labs and all other faculty

members and students in the University of Akron. There are too many people to list

here. Without your constant friendship and support, I cannot enjoy my life and

research in Akron. I would especially like to thank Yuqing Liu who offered me a lot of help in research and daily life when he was in Akron.

I also need to thank all the friends who are not in Akron but kindly provide great support for my research.

vi

TABLE OF CONTENTS Page

LIST OF TABLES ...... xi

LIST OF FIGURES ...... xii

LIST OF ABBREVIATIONS ...... xx

CHAPTER

I. INTRODUCTION ...... 1

II.BACKGROUND* ...... 4

2.1 Block copolymers ...... 4

2.1.1 Phase behavior ...... 6

2.1.2 Small angle X-ray scattering ...... 11

2.1.3 Linear viscoelastic behavior ...... 12

2.2 Supramolecular block polymer ...... 15

2.2.1 Supramolecular chemistry ...... 15

2.2.2 Supramolecular polymer ...... 18

2.2.3 Supramolecular block copolymer ...... 20

2.3 Ionomers ...... 24

2.3.1 General introduction of ionomers ...... 24

2.3.2 Synthesis of ionomers ...... 26

2.4 Cationic polymerization technique ...... 29

2.5 RAFT polymerization technique...... 31

2.5.1 Controlled free radical polymerization ...... 31

2.5.2 RAFT polymerization ...... 33

vii

2.5.3 Complex macromolecular architecture via RAFT polymerization ...... 36

2.5.4 End-functional polymer via RAFT approach ...... 39

III. SUPRAMOLECULAR MULTIBLOCK POLYSTYRENE- POLYISOBUTYLENE COPOLYMER VIA IONIC INTERACTIONS...... 42

3.1 Introduction ...... 42

3.2. Experimental Section ...... 46

3.2.1 Materials ...... 46

3.2.2 Synthesis of end-functionalized (telechelic) oligomers ...... 47

3.2.3 Preparation of the Supramolecular Block Copolymer ...... 54

3.2.4 Characterization ...... 55

3.3 Results and discussion ...... 57

3.3.1 Analytical Evidence for the Formation of a Supramolecular Block Copolymer...... 59

3.3.2 Morphology of the blend ...... 63

3.3.3 Mechanical Properties ...... 72

3.3.4 Nonlinear Rheological Behavior...... 74

3.4 Conclusions ...... 76

3.5 Acknowledgements ...... 77

IV. SYNTHESIS OF QUATERNARY AMMONIUM-CONTAINING, TRITHIOCARBONATE RAFT AGENTS AND HEMI-TELECHELIC CATIONOMERS ...... 78

4.1 Introduction ...... 78

4.2 Experimental section ...... 81

4.2.1 Materials ...... 81

4.2.2 Instrumentation ...... 81

4.2.3 Synthesis of 4-(bromomethyl)benzyltrimethylammonium bromide (Br-Ph-NMe3) ...... 82

viii

4.2.4 Synthesis of 4-(bromomethyl)benzyltriethylammonium bromide (Br- Ph-NEt3) ...... 83

4.2.5 Synthesis of 4-(bromomethyl)benzyltributylammonium bromide (Br- Ph-NBu3) ...... 84

4.2.6 Synthesis of S-1-dodecyl-S’-(methylbenzyltriethylammonium bromide) trithiocarbonate RAFT Agents (RAFT-NEt3) ...... 85

4.2.7 Synthesis of benzyl dodecyl trithiocarbonate (BDTC) ...... 87

4.2.8 RAFT Bulk Polymerization of Styrene at 120 °C ...... 88

4.2.9 Bulk RAFT Polymerization of Styrene at 65 °C ...... 88

4.2.10 RAFT Polymerization of Acrylate Monomers at 65 °C ...... 89

4.3 Results and discussion ...... 90

4.3.1 Synthesis of 4-(bromomethyl)benzyl-N,N,N-trialkylammonium bromide compounds ...... 90

4.3.2 Synthesis of quaternary ammonium-containing RAFT agents ...... 99

4.3.3 Bulk styrene polymerization ...... 105

4.4 Conclusions ...... 126

4.5 Acknowledgements ...... 127

V. SYNTHESIS AND CHARACTERIZATION OF QUATERNARY PHOSPHONIUM-CONTAINING, TRITHIOCARBONATE RAFT AGENTS* ...... 128

5.1 Introduction ...... 128

5.2 Experimental section ...... 130

5.2.1 Materials ...... 130

5.2.2 Instrumentation ...... 130

5.2.3 Synthesis of 4-(bromomethyl)benzyltri-n-butylphosphonium bromide (Br-Ph-PBu3) ...... 131

5.2.4 Synthesis of 4-(bromomethyl)benzyltriphenylphosphonium bromide (Br-Ph-PPh3) ...... 132

5.2.5 Synthesis of S-1-dodecyl-S’-(methylbenzyltributylphosphonium bromide) trithiocarbonate RAFT agents (RAFT-PBu3) ...... 132

ix

5.2.6 Synthesis of S-1-dodecyl-S’-(methylbenzyltriphenylphosphonium bromide) trithiocarbonate RAFT agents (RAFT-PPh3) ...... 133

5.2.7 RAFT bulk polymerization of styrene at 120 °C ...... 134

5.3 Results and discussion ...... 135

5.3.1 Synthesis of the RAFT agents...... 135

5.3.2 Thermal stability of the synthesized RAFT agents ...... 140

5.3.3 Bulk styrene polymerizations using RAFT-PR3 ...... 147

5.4 Conclusions ...... 162

5.5 Acknowledgements ...... 162

VI. SULFONATION DISTRIBUTION IN SULFONATED POLYSTYRENE ...... 163 IONOMERS MEASURED BY MALDI-ToF MS ...... 163

6.1 Introduction ...... 163

6.2 Experimental section ...... 166

6.2.1 Materials ...... 166

6.2.2 MALDI-ToF MS Analysis ...... 167

6.3 Results and discussion ...... 169

6.4 Conclusion ...... 180

6.5 Acknowledgements ...... 180

VII. SUMMARY ...... 181

REFERENCES ...... 187

x

LIST OF TABLES Table Page

2.1 Relative peak positions for various block copolymer microstructures ...... 11

2.2 Characteristics of non-covalent interactions in supramolecular chemistry ...... 16

4.1 Reaction conditions of 4-(bromomethyl)benzyl-N,N,N-trialkylammonium bromide compounds ...... 91

5.1 Degradation temperature when weight loss=5% ...... 141

6.1 Sulfonation level determined by MALDI-ToF MS...... 174

xi

LIST OF FIGURES

Figure Page

2.1 Schematic representation of different block copolymer architectures...... 5

2.2 Composition profiles of component A in weak and strong segregation limits of a AB diblock copolymer melt. ΦA and f represent the local and overall volume fraction of A blocks, respectively. Reproduced with permission from ref.11 ...... 8

2.3 Phase diagrams of a diblock copolymer melt determined by (a) self-consistent field theory and (b) experimental data of model poly(styrene-b-isoprene) diblock copolymers. Theory predicts five morphologies, including lamellae (L), hexagonally packed cylinders (C), body-centered cubic spheres (S), gyroid (G) and close-packed spheres (CPS). The experimental phase diagram also contains a perforated layers phase (PL), which was finally proved to be a metastable phase. (c) indicates the strong dependence of the block copolymer morphology on the composition. Reproduced with permission from ref.23...... 10

2.4 Temperature dependence of the storage modulus of a poly(ethylenepropylene)- poly(ethylethylene) diblock copolymer (Mn=81200 Da, PDI=1.05, wt%(PEP)=53%). Reproduced with permission from ref.33 ...... 13

2.5 Dependence of the storage modulus on frequency in the terminal region for disordered state and different ordered state: body-centered cubic spheres (cubic), hexagonally packed cylinders (cylinders), and lamellae. Reproduced with permission from ref.38 ...... 14

2.6 Scheme of different supramolecular interactions. (a) direction-ion interaction in tetrabutylammonium chloride; (b) ion-dipole interaction in sodium complex of [15]crown-5 and Ruthemium(II) complex of 2,2'-bipyridine; (c) dipole-dipole interactions in acetone; (d) A hydrogen bond formed between a secondary amine and carbonyl group; (e) π- π interactions. Adapted from ref. 39...... 17

2.7 Schematic representation of chain extension via the dimerization of UPy units. The chemical structure of UPy functional group is shown in the inset box. Reproduced with permission from ref. 47...... 19

2.8 UPy-functionalized poly(ethylene-co-butylene) (3.5k Da) is a flexible elastomers at room temperature and viscoelastic liquid at elevated temperature. Adapted from ref.48...... 19

xii

2.9 Proposed phase diagram for the PIP(NR2)2 (M = 18k)/PαMSt(COOH)2 (M =10k). The UCST represents phase separation of the constituent telechelic polymers. MST is the order-disorder transition (ODT) of the BCP structure. Tg is the glass transition, and Ti is the temperature at which the ionic bonds forming the BCP dissociate. A microphase-separated BCP morphology occurs within the left bottom to right top diagonally hatched area. The telechelic ionomers mixture phase is macroscopically phase separated in the right bottom to left top diagonally hatched area. The stippled area is where the mixtures exist as a disordered copolymer phase, and the clear area is where the mixture is a homogeneous mixture of the constituent telechelic ionomers. . The dashed lines are continuations of the phase diagrams which due to ionic aggregation or disruption of the end groups are not possible to observe. Reproduced with permission from ref. 63...... 22

2.10 Schematic of ionomer microstructure. (±) denotes the ion-pairs that are covalently attached to the polymer backbone. Reproduced with permission from ref.67...... 25

2.11 A general mechanism of controlled free radical polymerizations. Reproduced with permission from ref 115...... 33

2.12 Schematic of typical thiocarbonylthio RAFT agents and the formation of intermediate radicals. Reproduced with permission from ref.118...... 34

2.13 Mechanism of RAFT polymerizations. Reproduced with permission from ref.118...... 35

2.14 Guideline for selection of Z-group (top) and R-group (bottom) for different monomers. Dash line indicates partial control. For Z-group, fragmentation rate increases while the addition rate decreases from left to right. For R-group, fragmentation rates decreases from left to right. Reproduced with permission from ref. 122...... 36

2.15 Examples of complex macromolecular architecture prepared by RAFT polymerization. Reproduced with permission from ref. 117...... 37

2.16 A simplified scheme for the synthesis of block copolymers via sequential RAFT polymerization. Reproduced with permission from ref. 117...... 38

2.17 An illustration of the polymer species generated after two-step RAFT polymerization for making diblock copolymers. In this illustration, [M]/[RAFT]/[I] for both steps is set as 72/10/1. The initiator efficiency and the monomer conversion are both set as 1. Reproduced with permission from ref.123...... 38 2.18 Three routes that can introduce end-functionality to the polymer via RAFT polymerization. R’ refers to the new ω-group transformed from thiocarbonylthio group. Reproduced with permission from ref.117...... 39

2.19 Transformation of the thiocarbonylthio group. R’· = radicals, [H] = hydrogen donor, M = monomer. Reproduced with permission from ref. 135...... 40 xiii

2.20 Reaction of the thiol group transformed from thiocarbonylthio group. Reproduced with permission from ref.134...... 41

3.1 Synthesis of PS(SO3)2...... 48

3.2 Synthesis of PIB(NH2)2...... 50

3.3 1H NMR spectrum of Br-PIB-Br...... 52

3.4 1H NMR spectrum of PI-PIB-PI...... 53

1 3.5 H NMR spectrum of H2N-PIB-NH2...... 54

3.6 (a) Synthesis of (PIB-b-PS)n SMBCP (b) Flexibility and clarity of SMBCP film. .... 58

3.7 FTIR spectra of (a) H2N-PIB-NH2, (b) HO3S-PS-SO3H, (c) stoichiometric blend + - - + and (d) H9C4H3N O3S -PS-SO3 N H3C4H9 in the spectral region of 1100 – 1000 cm-1...... 60

1 3.8 H NMR of (a) stoichiometric blend and (b) H2N-PIB-NH2. Protons a and b were shifted upfield due to the ion complexation...... 61

3.9 DSC heating thermograms of (a) H2N-PIB-NH2, (b) 50/50 blend, and (c) HO3S- PS-SO3H. The heating rate was 10ºC/min...... 63

3.10 Azimuthally averaged SAXS pattern for the room temperature blend...... 64

3.11 (a) Temperature-resolved SAXS profiles of SMBCP measured at 20°C intervals during heating from 110°C to 250°C and then cooling to 150°C. (b) Temperature dependence of domain spacing, (c) inverse maximum intensity for the first-order scattering peak and (d) full-width at the half-maximum (FWHM) for the SAXS of the SMBCP. The filled symbols and open symbols represent heating and cooling data, respectively. Calculations for some temperatures were not possible, because of an ill-defined scattering peak...... 65

3.12 Polarized optical micrographs of SMBCP at (a) 185 °C, (b) 190 °C and (c) 195 °C. Each picture was taken after holding the sample at the temperature indicated for 30 min. The scale bar in each photograph is 250 μm...... 67 3.13 FTIR spectra of (a) HO3S-PS-SO3H, (b) SMBCP, (c) SMBCP that exhibited macrophase separation after heating at 190 oC. For (c), shoulder at 1039 is still observed but much smaller compared to original shoulders and shoulder at 1080 due to ammonium was disappeared. Also, no peak at 1050 is observed for (c), indicating no sulfonic acid was reformed during the heating, i.e. no reverse proton transfer occurred...... 68

3.14 1H NMR spectrum of SMBCP that exhibited macrophase separation after heating at 190 oC. The resonance at 3.81 ppm and 2.69 ppm were still observed but much smaller compared to the original peaks...... 69

xiv

3.15 Frequency dependence of (a) the storage modulus, G' and (b) the loss modulus, G'' as a function of temperature for the SMBCP. Strain amplitude = 5%...... 70

3.16 Dynamic storage and loss tensile moduli of SMBCP as a function of temperature. The curves marked as measured tensile data used a frequency of 6.3 rad/s and the curves marked as calculated from shear data were measured at a frequency of 10 rad/s and converted to tensile values by assuming E = 3G...... 71

3.17 Engineering tensile stress-strain curve of the SMBCP at room temperature...... 74

3.18 G’ (filled circle) and G” (open circles) for three consecutive strain sweep-time sweep cycles of the SMBCP at 150°C. For the strain sweep, γ varied from 0.01% to 100%. For the time sweep, γ = 1% and ω = 10 rad/s...... 76

4.1 Synthesis of 4-(bromomethyl)-N,N,N-trialkylbenzyl ammonium compounds (alkyl=methyl, ethyl, butyl) and subsequent RAFT agents………………………. 79

4.2 Reaction routes reported by (a) O’Reilly176 and (b) Samakande113...... 80

1 4.3 H NMR spectrum in D2O of Br- Ph-NMe3...... 94

1 4.4 H NMR spectrum in CDCl3 of Br- Ph-NEt3...... 95

1 4.5 H NMR spectrum in CDCl3 of Br- Ph-NBu3...... 96

4.6 ESI mass spectrum of Br- Ph-NMe3...... 97

4.7 ESI mass spectrum of Br-Ph-NEt3...... 98

4.8 ESI mass spectrum of Br- Ph-NBu3. The peak at m/z 237.1 is due to the di- substituted compounds...... 99

4.9 Chemical Structure of RAFT-NR3 and the potential sulfide by-product...... 100

4.10 NMR spectrum in CDCl3 of RAFT-NMe3...... 100

4.11 NMR spectrum in CDCl3 of RAFT-NEt3...... 101

4.12 NMR spectrum in CDCl3 of RAFT-NBu3...... 102

4.13 ESI mass spectrum of RAFT-NMe3...... 103

4.14 ESI mass spectrum of RAFT-NEt3...... 104

4.15 ESI mass spectrum of RAFT-NBu3...... 104

4.16 1H NMR spectrum of the reaction mixture taken from bulk polymerization of o styrene using RAFT-NEt3 at 120 C for 6 hr...... 106

xv

4.17 Pseudo first-order kinetic plot for the RAFT bulk polymerization of styrene at 120 °C with RAFT-NEt3. Target molecular weight is 25 kDa. The solid line is a linear fit to the data...... 107

4.18 Plot of Mn (SEC) and PDI versus monomer conversion for the RAFT bulk polymerization of styrene at 120 °C with RAFT-NEt3. Target molecular weight is 25 kDa...... 108

4.19 SEC traces as a function of polymerization time for the RAFT bulk polymerization of styrene at 120 °C with RAFT-NEt3. Target molecular weight is 25 kDa...... 109

4.20 Pseudo first-order kinetic plot for the RAFT bulk polymerization of styrene at 120 °C with RAFT-NMe3. Target molecular weight is 25 kDa. The solid line is a linear fit to the data...... 110

4.21 Pseudo first-order kinetic plot for the RAFT bulk polymerization of styrene at 120 °C with RAFT-NBu3. Target molecular weight is 25 kDa. The solid line is a linear fit to the data...... 111

4.22 NMR spectrum of PS from styrene bulk polymerization mediated by RAFT-NEt3 at 120 °C for 1h. One drop of the polymerization mixture was taken out and 1 diluted with CDCl3 and ran H NMR subsequently. Peak “x" is attributed to the thermal degradation product...... 112

4.23 The end functionality as a function of polymerization time at 120 °C...... 113

4.24 TLC test of polystyrene prepared from bulk polymerization at 120 °C for 1 h and 65 °C for 24 h. Toluene was used as the developing solvent...... 114

4.25 TGA curves for RAFT-NR3 (R=methyl, ethyl and n-butyl), Br- Ph-NEt3 and BDTC RAFT. Experiments were conducted using a nitrogen atmosphere with a heating rate of 20 °C/min...... 115

4.26 Isothermal TGA curves for RAFT-NEt3 at 65 °C and 120 °C. Experiments were conducted using a nitrogen atmosphere...... 116

4.27 Pseudo first-order kinetic plot for the RAFT bulk polymerization of styrene at 65 °C with RAFT-NEt3. Target molecular weight is 25 kDa. The solid line is a linear fit to the data...... 117

4.28 Plot of Mn (SEC) and PDI versus monomer conversion for the RAFT bulk polymerization of styrene at 65 °C with RAFT-NEt3. Target molecular weight is 25 kDa...... 118

4.29 SEC traces as a function of polymerization time for the RAFT bulk polymerization of styrene at 65 °C with RAFT-NEt3. Target molecular weight is 25 kDa...... 119

xvi

4.30 1H NMR spectra of a) crude polystyrene from styrene bulk polymerization at 65 °C, b) the purified polystyrene (CHCl3/acetone/methanol fraction), c) toluene fraction...... 121

4.31 The ion exchange tests that demonstrated the ionic functionality. The vials were shaken to perform ion exchange and then placed in the hood for overnight. The contents of each solution before shaking were as follows: Vial #1: toluene (top layer) & water+ methyl orange (bottom layer); Vial #2: toluene + purified polystyrene (Rf=0) (top layer) & water (bottom layer); Vial #3: toluene + purified polystyrene (Rf=0) (top layer) & water + methyl orange (bottom layer); Vial #4: toluene + polystyrene impurities (Rf=1) (top layer) & water +methyl orange (bottom layer)...... 122

4.32 1H NMR spectra of a) crude PMA, b) the purified PMA (acetone fraction), c) CHCl3/acetone fraction...... 124

4.33 1H NMR spectrum of purified PBA (acetone fraction)...... 125

4.34 1H NMR spectrum of PDMAEA...... 126

5.1 Synthesis of quaternary phosphonium containing trithiocarbonate RAFT agents, RAFT-PR3 (R=n-butyl and phenyl)...... 129

1 5.2 H NMR spectrum in CDCl3 of Br-Ph-PBu3...... 136

31 5.3 P NMR spectrum in CDCl3 of Br-Ph-PBu3...... 136

1 5.4 H NMR spectrum in CDCl3 of Br-Ph-PPh3...... 137

31 5.5 P NMR spectrum in CDCl3 of Br-Ph-PPh3...... 137

1 5.6 H NMR spectrum in CDCl3 of RAFT-PBu3...... 138

31 5.7 P NMR spectrum in CDCl3 of RAFT-PBu3...... 138

1 5.8 H NMR spectrum in CDCl3 of RAFT-PPh3...... 139

31 5.9 P NMR spectrum in CDCl3 of RAFT-PPh3...... 139

o 5.10 Temperature-ramp TGA traces (20 C/min) for RAFT-PR3 (R=n-butyl and phenyl), Br-PR3, and RAFT-NBu3 and BDTC RAFT. Tests were performed in nitrogen atmosphere. Data of RAFT-NBu3 and BDTC were adapted from Ref.194 with permission from The Royal Society of Chemistry...... 142

o 5.11 Isothermal TGA traces (120 C) for RAFT-PR3, RAFT-NBu3 and BDTC RAFT. Tests were performed in nitrogen atmosphere...... 143

1 5.12 H NMR spectrum in CDCl3 of RAFT-PBu3 before (top) and after (bottom) the isothermal TGA test, i.e. 120 oC for 6 h in nitrogen environment...... 144

xvii

1 5.13 H NMR spectrum in CDCl3 of RAFT-PPh3 before (top) and after (bottom) the isothermal TGA test, i.e. 120 oC for 6 h in nitrogen environment...... 145

1 5.14 H NMR spectrum in CDCl3 of BDTC before (top) and after (bottom) the isothermal TGA test, i.e. 120 oC for 6 h in nitrogen environment...... 146

1 5.15 H NMR spectrum in CDCl3 of RAFT-NBu3 before (top) and after (bottom) the isothermal TGA test, i.e. 120 oC for 6 h in nitrogen environment...... 147

5.16 Pseudo first-order kinetic plot for the RAFT bulk polymerization of styrene at 120 °C with RAFT-PBu3. Target molecular weight is 25 kDa. The solid line is a linear fit to the data...... 149

5.17 Plot of Mn (SEC) and PDI versus monomer conversion for the RAFT bulk polymerization of styrene at 120 °C with RAFT-PBu3. Target molecular weight is 25 kDa...... 150

5.18 SEC traces as a function of polymerization time for the RAFT bulk polymerization of styrene at 120 °C with RAFT-PBu3. Target molecular weight is 25 kDa...... 151 5.19 Pseudo first-order kinetic plot for the RAFT bulk polymerization of styrene at 120 °C with RAFT-PPh3. Target molecular weight is 25 kDa...... 153

5.20 Plot of Mn (SEC) and PDI versus monomer conversion for the RAFT bulk polymerization of styrene at 120 °C with RAFT-PPh3. Target molecular weight is 25 kDa...... 154

5.21 SEC traces as a function of polymerization time for the RAFT bulk polymerization of styrene at 120 °C with RAFT-PPh3. Target molecular weight is 25 kDa...... 155

5.22 TLC test of polystyrene obtained from bulk polymerization using (a) RAFT- o PBu3, (b) RAFT-PPh3, and (c) BDTC at 120 C...... 156

5.23 1H NMR spectra of a) crude polystyrene from styrene bulk polymerization using o RAFT-PBu3 at 120 C, b) the purified polystyrene (CHCl3/acetone/methanol fraction), c) toluene fraction...... 158

5.24 1H NMR spectra of a) crude polystyrene from styrene bulk polymerization using o RAFT-PPh3 at 120 C, b) the purified polystyrene (CHCl3/acetone/methanol fraction), c) toluene fraction...... 159

5.25 The ion exchange test that visualizes the cationic functionality. The vials were vortexed to facilitate ion exchange and then placed in the hood for one day. The contents of each vial before mixing were as follows: Vial #1: toluene (top layer) & water+D&C Green 5 (bottom layer); Vial #2: toluene+PS-PBu3 (top layer) & water (bottom layer); Vial #3: toluene+PS-PBu3 (top layer) & water+ D&C Green 5 (bottom layer); Vial #4: toluene+PS-PPh3 (top layer) & water (bottom layer); Vial #5: toluene+PS-PPh3 (top layer) & water+ D&C Green 5 (bottom layer) ...... 161 xviii

6.1 Schematics of chain structures with associative ionic groups. (a) Chains with one or two sulfonate groups can result in chain extension. Chains without ionic functionality are inactive. (b) Chains with three or more sulfonate groups may form a network structure. Chains with two sulfonate groups can participate in a network if both sulfonate groups are incorporated into different ionic clusters. (c) Multiple associations of monofunctional chains will form a micelle-like structure or dangling branches from a multiple sulfonated chain...... 165

6.2 Expanded views of mass spectra acquired using different matrices; (a) lower MW region; (b) higher MW region. "Sx" refers to polystyrene chains with x sulfonate groups...... 170

6.3 MALDI-ToF MS spectrum of LiSPS2.5...... 171

6.4 MALDI-ToF MS spectrum of LiSPS3.7 (no cationization salt)...... 172

6.5 MALDI-ToF MS spectrum of LiSPS6.5 (no cationization salt)...... 172

6.6 Expanded view of mass spectra of LiSPS2.5, LiSPS3.7, and LiSPS6.5. "Sx" refers to polystyrene chains with x sulfonate groups. Note that the degree of polymerization of S3, S2, S1, and S0 is 20, 21, 22, and 23, respectively...... 173

6.7 Sulfonation level versus degree of polymerization for N = 15-47...... 176

6.8 Sulfonation distribution measured by MALDI-TOF MS and binomial distribution predictions (Equation (1) for: (a) LiSPS2.5; (b) LiSPS3.7; and (c) LiSPS6.5). The dashed lines have no physical significance. They are only included to make it clear that the points denoting the predictions of the binomial distribution, equation (1), are discrete values (the points connected by the lines)...... 178

6.9 Sulfonation distribution from the N = 38 fraction of LiSPS2.5, LiSPS3.7, and LiSPS6.5. The dashed lines have no physical significance. They are only included to make it clear that the points denoting the predictions of the binomial distribution, equation (1), are discrete values (the points connected by the lines). .. 179

xix

LIST OF ABBREVIATIONS

ADMET acyclic diene metathesis

AIBN azobisisobutyronitrile

ARES advanced rheometric expansion systems

ATRP atom transfer radical polymerization

BCP block copolymers

BDTC benzyl dodecyl trithiocarbonate

CFRP controlled free radical polymerization

DBX α,α’-dibromo-p-xylene

DMA dynamic mechanical analysis

DSC differential scanning calorimetry

ESI electron spray ionization

FTIR Fourier transform infrared

MALDI-ToF MS matrix-assisted laser desorption ionization time-of-flight mass spectrometry

mCPBA m-chloroperoxybenzoic acid

NMP nitroxide-mediated polymerization

NMR nuclear magnetic resonance

OOT order-order transition

ODT order-disorder transition

PαMSt poly(α-methyl styrene)

PBD polybutadiene

xx

PEP poly(ethylenepropylene)

PIB polyisobutylene

PIP polyisoprene

PS polystyrene

PMA poly(methyl acrylate)

PBA poly(butyl acrylate)

PDMAEA poly(2-dimethylaminoethyl acrylate)

RAFT reversible addition-fragmentation transfer

SEC size exclusion chromatograph

SMBCP supramolecular multiblock copolymer

SSL strong segregation limit

SCFT self-consistent field theory

SAXS small angle X-ray scattering

SPS sulfonated polystyrene

Tg glass transition temperature

TGA thermogravimetric analysis

UPy 2-ureido-4[1H]-pyrimidinone

UCST upper critical solution temperature

WSL weak segregation limit

xxi

CHAPTER I

INTRODUCTION

Since the advent of supramolecular chemistry,1 much effort has been directed

towards the development of supramolecular polymers, as the incorporation of

supramolecular interactions allows the design of more complex macromolecular

structures with tailored properties that cannot be achieved by covalent bonds.

Supramolecular block copolymers,2,3 an important class of supramolecular polymers, can be produced by replacing the covalent junctions between individual polymer blocks with non-covalent interactions. Most studies on supramolecular block polymers used hydrogen bonding or metal coordination as the supramolecular interactions while Coulombic interactions have received much less attention.

However, it should be noted that Coulombic interactions have long been employed to construct supramolecular materials, such as dye-surfactant complexes, 4 polymer-

surfactant complexes,5 ionomers,6,7 and polyelectrolyte complexes.8,9 Therefore, it is

very interesting to study the supramolecular block copolymers via Coulombic interactions.

The objective of this dissertation was to prepare and characterize the

supramolecular block copolymers via direct ion-ion interactions. In the first part

(Chapter III), a supramolecular multiblock copolymer, was synthesized by mixing two telechelic oligomers, α,ω-disulfonated polystyrene, HO3S-PS-SO3H, derived from a

polymer prepared by RAFT polymerization, and α,ω-diamino-polyisobutylene, H2N-

1

PIB-NH2, prepared by cationic polymerization. During solvent casting, proton transfer from the sulfonic acid to the amine formed ionic bonds that produced a multiblock copolymer that formed free-standing flexible films with a modulus of 90 MPa, a yield

point at 4% strain and a strain energy density of 15 MJ/m3. Small angle X-ray

scattering characterization showed a lamellar morphology, whose domain spacing was

consistent with the formation of a multiblock copolymer based on comparison to the

chain dimensions. A reversible order-disorder transition occurred between 190°C and

210°C, but the sulfonic acid and amine functional groups decomposed at those elevated temperatures based on companion optical microscopy and spectroscopy measurements. For high non-linear strains, the dynamic modulus, G’, decreased by nearly an order of magnitude and the loss modulus, G”, decreased by a factor of 1.4, but both recovered to their original values once the strain was reduced to within the linear response region.

In the second part (Chapter IV and Chapter V), two groups of reversible

addition−fragmentation chain-transfer (RAFT) agents that contain either quaternary

ammonium or quaternary phosphonium groups were prepared. At first, a series of

trithiocarbonate RAFT agents containing quaternary ammonium functionality in the

“R-group” (RAFT-NR3) were synthesized via a facile and high-yield approach. The

synthetic route first involves the optimized synthesis of 4-(bromomethyl)-N,N,N-

trialkyl benzyl ammonium bromide compounds, which were subsequently reacted

with the alkyl trithiocarbonate anion to directly produce the trithiocarbonate RAFT

agent. However, the quaternary ammonium group partially degraded when the RAFT

agents were used in polymerizations at 120 oC. This issue was overcome by using

lower polymerization temperature. Quaternary phosphonium-containing,

trithiocarbonate RAFT agents (RAFT-PR3) were also synthesized via a similar

2

synthetic method. The thermal stabilities of RAFT-PR3 were enhanced compared to comparable quaternary ammonium-containing RAFT agents, which significantly improved the retention of the cationic end-functionality of the polystyrene obtained at

120 oC. For both classes of RAFT agents, the polystyrene product was further purified

via column chromatography to afford high purity hemi-telechelic cationomers.

In the third part (Chapter VI), matrix-assisted laser desorption ionization time-

of-flight mass spectrometry (MALDI-ToF MS) was used to quantify the sulfonation level and sulfonation distribution of sulfonated polystyrene ionomers prepared by homogeneous solution sulfonation. The sulfonation levels obtained by MALDI ToF-

MS and acid-base titration were compared, and the sulfonate distributions determined by MALDI-ToF MS were compared with theoretical random distributions. The results indicate that the sulfonation reaction used produces a sample with a random sulfonate distribution.

3

CHAPTER II

BACKGROUND*

2.1 Block copolymers

This section deals with three aspects of block copolymer physics, including phase behavior, small angle X-ray scattering and linear viscoelastic behavior.

Block copolymers are macromolecules made up of linear (diblocks, triblocks,

multiblocks) or nonlinear (graft, star-block, miktoarm star, cyclic, etc.) structure of

chemically distinct blocks, as shown in Figure 2.1.10 Even when the different blocks

are immiscible, macroscopic phase separation is prevented by the covalent bonds

tethering the blocks, which gives rise to a wide variety of microstructures in bulk state,

thin films and solutions.11,12,13,14,15 Due to their unique self-assembly behavior, block

copolymers have attracted considerable academic and industrial interest over the past few decades. Block copolymers have found commercialized applications, such as

thermoplastic elastomers 16 and compatibilizers for polymer blends, 17 and potential

applications in emerging technologies such as drug delivery,18 nanopatterning and

electronics,19,20,21 and photonic crystals.22

∗ Portions of this section are reproduced from: Zhang, L.; Brostowitz, N. R.; Cavicchi, K. A.; Weiss, R. A. Macromol. React. Eng. 2014, 8, 81-99. 4

AB diblock

ABC triblock Cyclic AB diblock ABA triblock

(AB)n multiblock

AB Graft

(AB) star ABC miktoarm star 4

Figure 2.1 Schematic representation of different block copolymer architectures.

5

2.1.1 Phase behavior

Diblock copolymers represent the simplest combination among the possible

block copolymer architectures. The phase behavior of diblock copolymers in the melt state has been extensively studied experimentally and theoretically, which are summarized in a few review articles11, 23 and books.12 Three material factors

determine the phase behavior of a diblock copolymer melt AB: the total degree of

polymerization N, the composition of the copolymer (overall volume fraction of one component) f, and the Flory-Huggins (segment-segment) interaction parameter χ.11,12

For an AB diblock copolymer melt, χ is defined as23,24

( 1 2( + )) = 푧∆푤 푧 푤퐴퐵 − ⁄ 푤퐴퐴 푤퐵퐵 χAB ≡ where is the number of nearest푘푇 neighboring monomers푘푇 to a monomer, is the thermal푧 energy, is the exchange energy that is required to switch the position푘푇 of one A and one B∆푤 monomer when they are nearest neighbors and represents the

푖푗 interaction energy between monomer A and B. When 푤0 , i.e. χ 0 ,

AB monomers A and B oppose mixing with each other. When ∆푤 specific ≥ interactions≥ are

absent, most monomer pairs have positive χ and therefore AB diblock exhibit phase

separation. For most polymer pairs, the temperature dependence of χ is

+ T α χ ≈ β whereα is positive, β is a constant and both parameters are dependent on specific

composition and architecture of the block copolymers. Therefore, χ decreases as

temperature increases.

To describe the phase equilibria of block copolymer melts, two quantities are

used, i.e. the composition of the copolymer f and the product χN.11,25 The quantity χN represents the degree of segregation, as a larger χ indicates larger repulsion forces 6

between A and B blocks and a larger N diminishes the translational and

configurational entropy contribution to the Gibbs energy of mixing, i.e. leads to a

reduction of A-B monomer contacts in the interface and therefore promotes local

ordering.

Three regimes have been identified based on the magnitude of χN, namely

weak segregation limit (WSL), strong segregation limit (SSL) and intermediate

segregation region (ISR).26 For a diblock copolymer with symmetric composition,

mean-field theory predicts that a critical point occurs at (χN)ODT=10.5, where the

system undergoes an order-disorder transition (ODT).25 When χN is smaller than

(χN)ODT, the system is dominated by entropic terms and exhibits a disordered state.

For WSL, the value of χN is approximately between 10.5 and 12, for which the

composition profile is sinusoidal, as shown in Figure 2.2. For much larger χN, i.e.

χ 100, the system enters the regime of SSL, where the phase segregation is so

strongAB ≥ that the component in each microdomain is essentially pure and the interface is

very narrow, as shown in Figure 2.2.

7

Figure 2.2 Composition profiles of component A in weak and strong segregation limits of a AB diblock copolymer melt. ΦA and f represent the local and overall volume fraction of A blocks, respectively. Reproduced with permission from ref.11 .

Representative classical theories for WSL and SSL were developed by Leibler

and Semenov, respectively.25, 27 These theories successfully predict three classical

phases (lamellar, hexagonal cylinder, and spherical) but fail to compute the existence

of other complex microstructures in the regime of ISR which ranges between WSL

and SSL. Later, Matsen and co-workers developed the most general self-consistent field theory (SCFT) that is capable to account for non-classical phases of diblock copolymers, such as bicontinuous double gyroid.28,29

The phase diagram of a symmetric AB diblock copolymer was determined by

SCFT28,29 and experimental data of a model diblock copolymer poly(styrene-b-

isoprene)30, is shown in Figure 2.3(a) and (b). It is clear that both phase diagrams

compare very well to each other, which indicated the validity of SCFT methods.

Phases that were observed in both phase diagrams included lamellar (L), hexagonally

packed cylinders (C), spheres in body-centered cubic lattice (S), and bicontinuous 8

double gyroid structure (G). It is interesting that a complex structure of perforated

layers was observed in experiments but ruled out by the SCFT approach. This phase

was later proved to be a metastable phase, which also demonstrated the significance of the SCFT methods.

Figure 2.3(c) shows that the equilibrium microstructure of the diblock copolymers is dictated by the composition of the block copolymer, f. When the volume fraction of A block, fA is very small, A block forms spherical phases. When fA

increases, A block formed cylinders, and then gyroids. When the volume fractions of

both blocks are close to each other, the system forms lamellar structure. Therefore,

there is a strong dependence of the morphology on the composition of the two blocks

for diblock copolymers.

9

Figure 2.3 Phase diagrams of a diblock copolymer melt determined by (a) self- consistent field theory and (b) experimental data of model poly(styrene-b-isoprene) diblock copolymers. Theory predicts five morphologies, including lamellae (L), hexagonally packed cylinders (C), body-centered cubic spheres (S), gyroid (G) and close-packed spheres (CPS). The experimental phase diagram also contains a perforated layers phase (PL), which was finally proved to be a metastable phase. (c) indicates the strong dependence of the block copolymer morphology on the composition. Reproduced with permission from ref.23.

10

2.1.2 Small angle X-ray scattering

Small angle scattering has been an indispensable means for characterizing the microstructure of block copolymers. The theoretical background of small angle scattering can be found in several books and reviews. 31, 32 Briefly, the scattering vector can be defined as

sin q = | | = 4π θ 퐪 where is the scattering angle and is the wavelengthλ of the X-ray radiation.

2θFor block copolymers, the relativeλ position of the higher order reflections to the first order reflections indicates the type of microstructure, as summarized in Table

2. 1.32 Therefore, the microstructure of the block copolymer can be determined when sufficient peaks are observed.

Table 2. 1 Relative peak positions for various block copolymer microstructures

Microstructure Ratio q/q*

Lamellar 1: 2: 3: 4: 5: 6

Hexagonally packed Cylinder 1: 3: 4: 7: 9: 12

Gyroid 1: √4 3√: 7√ 3:√ 8√3: 10 3: 11 3

Body centered cubic (BCC) 1: � 2⁄: 3�: ⁄4: �5: ⁄6 � ⁄ � ⁄

√ √ √ √ √

11

2.1.3 Linear viscoelastic behavior

Besides small angle X-ray scattering (SAXS), rheology provides another

versatile technique to study the phase behavior of block copolymers. In particular, linear viscoelastic behavior of block copolymer melts have been frequently used to identify order-order transitions (OOT) and the order-disorder transition (ODT). In addition, the linear viscoelastic behavior in the low frequency region can distinguish between different ordered microstructures, but it cannot unambiguously assign the microstructure. Both experimental and theoretical aspects of linear viscoelasticity of block copolymers were summarized by Fredrickson and Bates.26

Low frequency viscoelastic behavior of block copolymers can be used to locate an OOT and the ODT. Two techniques have been used, i.e. isochronal

temperature scans and isothermal frequency scans.

When subjected to low frequency isochronal temperature scans with slow

heating rate (1-2 oC/min), the dynamic elastic modulus of block copolymer melts

exhibits a sharp decrease when the system undergoes an ODT, as shown in Figure

12,26,30,33 2.4. This method of determining the temperature of ODT, i.e. TODT, has been

applied to spherical, 34 cylindrical, 35 gyroid30 and lamellar33, 36 microstructures.

Different types of order-order transitions can also be detected using this method.30,37

The frequency dependence of the storage modulus in the terminal region reflects the

state of order in block copolymer melts, as summarized by Bates and coworkers in

Figure 2.5. 38 Thus, the microstructure of a block copolymer melt can also be

determined by frequency sweep.

12

Figure 2.4 Temperature dependence of the storage modulus of a poly(ethylenepropylene)-poly(ethylethylene) diblock copolymer (Mn=81200 Da, PDI=1.05, wt%(PEP)=53%). Reproduced with permission from ref.33 .

13

Figure 2.5 Dependence of the storage modulus on frequency in the terminal region for disordered state and different ordered state: body-centered cubic spheres (cubic), hexagonally packed cylinders (cylinders), and lamellae. Reproduced with permission from ref.38 .

14

2.2 Supramolecular block polymer

This section covers the general concepts of supramolecular chemistry, supramolecular polymer and supramolecular block copolymer.

2.2.1 Supramolecular chemistry

The field of supramolecular chemistry concerns the self-organization of molecular complexes using non-covalent intermolecular bonding.1 A variety of non- covalent interactions are used in supramolecular chemistry, including ionic interactions, metal coordination, hydrogen bonding, van der Waals interactions, stacking interactions, hydrophobic interactions, etc. Examples of supramolecular휋 − 휋 interactions are shown in Figure 2.6.39 These non-covalent interactions represent the energies and directionality holding supramolecular species. The strength and other characteristics of different non-covalent interactions are summarized in

Table 2.2.40 These intermolecular forces are, in general, weaker than covalent bonds, and therefore supramolecules and supramolecular assemblies are more dynamic in nature compared to covalently-bonded molecules.

15

Table 2.2 Characteristics of non-covalent interactions in supramolecular chemistry [40]

Range of action Interaction Strength (kJ/mol) Character (nm)

100-350, Non-selective, Ionic (ion-ion) comparable to Long Non-directional covalent bonding Metal coordination 50-200 Short Directional or ion-dipole Dipole-dipole 5-50 Short Directional Selective, Hydrogen Bonding 10-120 Short directional Cation- 5-80 Short interactionsπ Anion- interactions 1-10 Short

π stacking 0-50 Short Directional 휋 − 휋 <5 but depending on Non-selective, Van der Waals Short surface area Non-directional Related to solvent- Non-selective, Hydrophobic solvent interaction Short Non-directional energy

16

Figure 2.6 Scheme of different supramolecular interactions. (a) direction-ion interaction in tetrabutylammonium chloride; (b) ion-dipole interaction in sodium complex of [15]crown-5 and Ruthemium(II) complex of 2,2'-bipyridine; (c) dipole- dipole interactions in acetone; (d) A hydrogen bond formed between a secondary amine and carbonyl group; (e) π- π interactions. Adapted from ref. 39.

17

2.2.2 Supramolecular polymer

Supramolecular polymers involve the association of low or high molar mass

unimers (building blocks) via reversible non-covalent interactions, for which the

unimers can be covalent molecules, oligomers, macromolecules, and supramolecular

assemblies such as micelles.41,42 As an intersection of supramolecular chemistry and polymer science, supramolecular polymer has been intensively studied since early

1990s.1, 43, 44, 45 A prominent example of supramolecular polymer was reported by

Meijer and coworkers, who exploited the chemistry of 2-ureido-4[1H]-pyrimidinone

(UPy).44,45,46,47 UPy groups form exceptionally strong dimers via quadruple hydrogen bonds, as shown in Figure 2.7.48 The introduction of UPy to polymer chains leads to a very interesting supramolecular material, which, via the dynamic nature of the hydrogen bonds, exhibits the mechanical properties of conventional polymers at room

temperature and excellent processability at elevated temperature, as illustrated in

Figure 2.8. Due to the dynamic nature of the supramolecular bonds, supramolecular polymers have advantages such as recyclable and better processability. In addition,

the dynamic and reversible nature of supramolecular bonds provides great

opportunities for fabricating smart materials that are self-healable49,50 or responsive to

external stimuli such as heat,48,51 light52 or pH53.

18

Figure 2.7 Schematic representation of chain extension via the dimerization of UPy units. The chemical structure of UPy functional group is shown in the inset box. Reproduced with permission from ref. 47.

Figure 2.8 UPy-functionalized poly(ethylene-co-butylene) (3.5k Da) is a flexible elastomers at room temperature and viscoelastic liquid at elevated temperature. Adapted from ref.48.

19

2.2.3 Supramolecular block copolymer

Supramolecular block copolymers,54, 55, 56 which are also called pseudo-block

copolymer57 or block-type supramacromolecules,58 can be produced by replacing the

covalent linkage between different blocks in block copolymers with non-covalent

interactions. This enables supramolecular block copolymers to exhibit the intrinsic

phase behavior of covalently-bonded block copolymers and the reversibility of

supramolecular materials. Such a combination makes supramolecular block

copolymers highly desirable for smart or nanostructured materials.50,51,53, 59, 60, 61 In addition, supramolecular block copolymers can be produced by self-assembly, i.e. simple "mixing and matching", of functional polymeric building blocks. Thus, combination of the blocks is versatile and a library of functional oligomers or polymers can be drawn upon to produce new supramolecular block copolymers. This

"supramolecular block copolymerization" demonstrates the great capability to produce multiphase polymer systems using simpler and less expensive chemistry than traditional strategies of copolymerization.

Supramolecular block copolymers have been synthesized using a variety of supramolecular interactions, including ionic interactions,58,62,63 hydrogen bonding64

and metal-ligand coordination.60, 65 , 66 In this section, supramolecular block copolymers based on ionic interactions will be reviewed. 67

Jerome and coworkers used proton transfer between either polyisoprene, PIP, or polybutadiene, (PBD), end-capped with tertiary amine groups and polystyrene, PS,

or poly(α-methyl styrene), PαMST, end-capped with sulfonic acid or carboxylic acid

groups to prepare supramolecular BCPs. 62, 63, 68, 69, 70 An order-disorder transition

(ODT), similar to what is observed in conventional BCPs, was found for the

supramolecular BCPs, but because of the strong ionic interaction that coupled the

20

blocks, the interface between the blocks remained sharp even when the system was approaching the ODT.63 Dissociation of the ionic bond occurred at higher temperature,

where a phase separation (UCST behavior) of the constituent telechelic ionomers by

spinodal decomposition was observed. The authors proposed a phase diagram such as

shown in

Figure 2.9 to explain the experimental observations for these supramolecular BCPs

(the explanation of the phase diagram is included in the caption of

Figure 2.9. A subsequent study on the viscoelastic behavior of these supramolecular

block copolymers supported the proposed phase behavior.69

Other supramolecular BCPs have also been prepared using proton transfer to

provide ionic bonds between telechelic polymers, e.g., PS-poly(ethylene oxide),71- 74

PS-polyisobutylene, 75 , 76 and PS-PIP.51,58 Some applications of ionically bonded supramolecular BCPs include interfacial modification for immiscible polymer blends

{PS-poly(dimethyl siloxane)}, 77 antireflective coatings {PS-poly(methyl

methacrylate)}, 78 structured nanoparticles (PS-PBD),61 nanoporous films {(PS-

poly(3-hexylthiophene)}, 79 (PS-PEO) 80 , 81 and vesicles {PS-

poly(isopropylacrylamide)}.82

21

Figure 2.9 Proposed phase diagram for the PIP(NR2)2 (M = 18k)/PαMSt(COOH)2 (M =10k). The UCST represents phase separation of the constituent telechelic polymers. MST is the order-disorder transition (ODT) of the BCP structure. Tg is the glass transition, and Ti is the temperature at which the ionic bonds forming the BCP dissociate. A microphase-separated BCP morphology occurs within the left bottom to right top diagonally hatched area. The telechelic ionomers mixture phase is macroscopically phase separated in the right bottom to left top diagonally hatched area. The stippled area is where the mixtures exist as a disordered copolymer phase, and the clear area is where the mixture is a homogeneous mixture of the constituent telechelic ionomers. . The dashed lines are continuations of the phase diagrams which due to ionic aggregation or disruption of the end groups are not possible to observe. Reproduced with permission from ref. 63.

Architectures other than linear chains can also be prepared via ionic

interactions or simple metal coordination. Weiss and co-workers studied

supramolecular graft copolymers formed by ionic bonds or transition metal

coordination (ion-dipole), e.g., hemi-telechelic amino terminated PIP/SPS (acid or

zinc salt)66 and telechelic PBD with Cu-sulfonate groups/SPS with random vinyl

pyridine groups.83 Similarly, Orfano et al. grafted hemi-telechelic sulfonated PS or

22

sulfonated PIP to poly(vinyl-2-pyridine) using ionic bonds. 84 Noro et al. prepared supramolecular graft-block copolymers by developing ionic bonds between a hemi- telechelic amino terminated PIP and a BCP of PS-b-poly(4-styrene sulfonic acid).58

Bazuin and co-workers used ionic bonds between a SPS ionomer and amino

terminated telechelic or hemi-telechelic PS to prepare a graft BCP of short

polystyrene chains attached to a polystyrene backbone.85,86 Jo et al. used ionic bonds

to graft poly(L-lysine) hydrobromide to a poly(lactic-glycolic acid) chain, which

produced spherical micelles in water.87 Polymer vesicles have also been prepared by

graft-block copolymers of hemi-telechelic carboxylic acid-poly(N-vinylpyrrolidone)

ionically bonded to poly(4-vinyl pyridine). 88 Supramolecular star-BCPs have also

been prepared by a number of research groups (hemi-telechelic tri(dimethylamino)-

polystyrene and hemi-telechelic sulfonic acid-terminated PIP),89 (hemi-telechelic tert- amine--end-functionalized poly(phenyl vinyl sulfoxide) and hemi-telechelic carboxylic acid-PS or randomly carboxylated PS),90 six-arm poly(tert-butyl acrylate),

PS or a BCP were prepared by ionically bonding of carboxylic acid functional RAFT agent to a core consisting of six tertiary amine groups and subsequent RAFT polymerization.91

23

2.3 Ionomers

This section covers the general concepts, structural characteristics and

synthesis of ionomers.

2.3.1 General introduction of ionomers

Ionomers are hydrophobic polymers containing a small fraction (<15 mol%)

of chemically bonded ionic groups. 92 In ionomers, the ionic groups will phase

separate from the low dielectric polymer backbones and form nanodomains, resulting

in nanostructured materials. The key feature of ionomers is that a relatively modest

concentration of acid or ionic groups can provide substantial changes in the physical,

mechanical, optical, dielectric and dynamic properties of a polymer. For acid

functionalization, interchain, physical crosslinks are formed by hydrogen bonding of

two acid groups. If the acid groups are fully or partially neutralized to form an ionic

compound, i.e. a salt, ionic or dipole-dipole interactions between two or more ionic groups, depending on the valency of the cation, also form physical crosslinks that

significantly alter the material properties.

As indicated above, a distinctive characteristic of ionomers is their nano-phase

separation structure, as illustrated in Figure 2.10. Small angle X-ray scattering

(SAXS) data and analyses indicate the nanodomains are of the order of 1 nm in

diameter with an average separation of 2 – 5 nm.6 Since the ion-pairs are covalently

attached to the polymer chain, the ionic nanodomains behave as multifunctional,

physical crosslinks that significantly affect the mechanical properties and dynamics of

these materials. Analysis of nanodomains’ structure derived from scattering

experiments suggests that a 3-4 nm diameter nanodomain may contain as many as 10-

30 ion-pairs; that is, the functionality of the “crosslink junction” can be of the order of

24

10-30. 93 In addition, the crosslink junction is a distinct phase of a viscoelastic material that exhibits a distribution of relaxation times. Thus, for ionomers the dynamics are affected by not only the chain motions, but also by the relaxation of the ionic or dipolar interactions that actually dominate the relaxation and flow behavior of these materials. The solution behavior and melt rheology of ionomers has been previously reviewed.94,95

Figure 2.10 Schematic of ionomer microstructure. (±) denotes the ion-pairs that are covalently attached to the polymer backbone. Reproduced with permission from ref.67.

25

Due to the presence of the ionic domain, ionomers are inherently supramolecular polymers in that physical interactions such as ion-ion interactions

(ionic bonds), ion-dipole interactions (e.g., metal coordination) and dipole-dipole

interactions are responsible for their unique properties. Ionic interactions are

comparable to covalent bonding in strength and are non-specific in that ion-pairs will

form between a variety of anions and cations. Ion-dipole and dipole-dipole

interactions exhibit orientation dependence in order to optimize the alignment of two

interacting species. 4,40,96

2.3.2 Synthesis of ionomers

Ionomers can be divided into two categories in terms of the position of ionic

functionality, i.e. random ionomer and telechelic ionomers.

Random ionomers can be synthesized by direct copolymerization of an ionic

and non-ionic monomer or post-polymerization modification (e.g. sulfonation, quaternization). The latter approach is often used due to the difficulty of dissolving both the ionic and non-ionic monomers and the compatibility of the ionic monomer with the polymerization technique, such as cationic or anionic polymerization. 97

However, the problems associated with post-polymerization modification98-100 such as,

aggressive reaction conditions that require hazardous chemicals and/or produce side-

reactions (e.g. sulfonation, chloromethylation), incomplete functionalization, and

purification drive the investigation of the direct copolymerization of ionic and non-

ionic monomers. Advances in the direct copolymerization of ionomers have been

enabled due to intensive research in two related areas: controlled free radical

polymerization (CFRP) 101 and ionic liquids 102 CFRP produces polymers with

controlled molecular weight and monomer distribution and has a higher tolerance for

26

chemical functionalities compared to anionic and cationic polymerizations (see

Section 2.5 for more detailed introduction of CFRP). For example, Okamura et al.103 used nitroxide mediated free radical polymerization to prepare sulfonated polystyrene ionomers by the copolymerization of styrene and styrene sulfonate ester, where the latter was subsequently deprotected to obtain the sulfonate group.

A number of ionic liquids have recently been investigated as monomers for the direct polymerization of ionomers. First, the low melting point of ionic liquids is useful for the bulk polymerization of ionomers. For example, Aitken et al.104 reported the synthesis of polyolefins with pendant imidazolium groups by acyclic diene metathesis (ADMET) polymerization. Second, many ionic liquid monomers are oleophilic, allowing their direct copolymerization with hydrophobic monomers. For example, Cheng et al. 105 prepared neutral-cationic-neutral triblock copolymers of poly(vinyl trialkylbenzyl phosphonium chloride) and poly(n-butyl acrylate) by NMP and Chen et al.106 prepared cationic ionomers by free radical copolymerization of n- butyl methacrylate and methacrylate-based imidazolium monomers. Counter-ion exchange has also been widely used to tune the solubility of sulfonate monomers to allow their copolymerization with hydrophobic monomers. The work in that area was recently reviewed by Cavicchi.107

The other category of ionomers, i.e. telechelic ionomers, has more simple architecture and has often been used as model ionomer systems.108 They can also be used as construct supramolecular building blocks, as discussed in Section 2.2.3.

Telechelic ionomers were conventionally prepared by living ionic polymerization techniques.109 Despite their precise control of the molecular weight, molecular weight dispersity and end-functionality of the polymers, ionic polymerizations suffer from several drawbacks, such as laborious procedures that require rigorous purification of

27

all of the solvents and reagents. Recently, CFRP approaches have been employed to prepare telechelic ionomers. For example, Feng et al. 110 demonstrated that ω- sulfonated polymers can be generated by oxidation of trithiocarbonate-containing polystyrene prepared by RAFT polymerization using m-chloroperoxybenzoic acid.

These polymers were purified by column chromatography with >95% end-group functionality. Similarly, cationic functional initiators for NMP, atom transfer radical polymerization and chain transfer agents for RAFT polymerization have been reported.111-113

28

2.4 Cationic polymerization technique

This section provides a brief introduction of cationic polymerization of

carbon-carbon double bonds in vinyl monomers.115

2.4.1 Monomer

Cationic polymerization has high selectivity towards vinyl monomers.

Essentially, only monomers with electron-donating substituent groups, e.g. alkyl, phenyl, alkoxy, can undergo cationic polymerizations. The electron-donating groups increase the electron density on the C-C double bonds and therefore enhance its reactivity towards a cationic species. On the other hand, the electron-donating substituents stabilize the cationic species by delocalizing the positive charges. Thus, vinyl monomers that are suitable for cationic polymerization include styrene, isoprene, isobutylene, methyl vinyl ether, etc.

2.4.2 Initiation

Cationic polymerization can be initiated by various methods, including protonic acids, Lewis acids, halogen, electroinitiation, ionizing radiation or photoinitiation. Among these methods, protonic acids and Lewis acids are more

frequently used and therefore will be discussed in this section.

Protonic acids are able to protonate the C-C double bonds to initiate cationic

polymerization. To be an effective cationic initiator, a protonic acid should be

sufficiently strong to generate reasonable amount of protonated species. On the other

hand, the anion of the protonic acid should not be too nucleophilic, since a highly

nucleophilic anion will terminated the initiation.

29

Lewis acids are the most important initiation methods for cationic

polymerization. Pure Lewis acid can barely initiate polymerization with a reasonable

rate. Usually either a proton donor, i.e. water, alcohol, and hydrogen halide, or a

carbocation donor, such as alkyl halide, is used together with Lewis acid as an

initiating system. The initiation process can be expressed as follows.

where I is the Lewis acid, ZY is the proton donor or carbocation donor, and M is the

monomer. This initiating system is advantageous over protonic acid, because the

counterion (IZ)- can be much less nucleophilic compared to the counterion of the protonic acid. In this way, the termination caused by the anion is reduced and the lifetime of the propagating species is longer.

2.4.3 Living cationic polymerization

It is difficult to achieve cationic polymerization in a living/controlled manner, primarily due to the chain transfer and chain termination that are likely to occur during cationic polymerization. For example, β-proton from the propagating carbocation could be transferred to basic species presented in the reaction systems, such as monomer and counterion. Chain transfer to monomer is the dominant factor limiting the chain growth of cationic polymerization, which can be overcome by using lower polymerization temperature, since chain transfer to monomer requires higher activation energy than propagation. Besides chain transfer to monomer or counterion,

chain transfer to polymers, solvent or impurity also prevents cationic polymerization

from a controlled manner.

30

To achieve living cationic polymerization, lower temperature is used to

suppress the chain transfer to monomer. In addition, deliberately designed initiating

systems, i.e. Lewis acids with a counterion of appropriate nucleophilicity, can be used,

as they allow a fast reversible conversion between the propagating species and the

dormant species, which leads to longer lifetime of the propagating species and

therefore more controlled polymerization.

2.5 RAFT polymerization technique

This section covers the general introduction of RAFT polymerization and its

application in engineering designed macromolecular structures.

2.5.1 Controlled free radical polymerization

Currently, there are three major types of controlled (or “living”) free radical

polymerization (CFRP) techniques that are frequently used in polymer synthesis, including atom transfer radical polymerization (ATRP), nitroxide-mediated

polymerization (NMP), and reversible addition-fragmentation transfer (RAFT) polymerization.101 These techniques were all developed in the 1990s and since then have received great attention because CFRP can offer substantial advantages for building well-defined macromolecular systems that provide tailored nanostructures and properties for emerging applications.114

For conventional free radical polymerizations, the lifetime of the radicals is

very short due to the diffusion-controlled bimolecular termination, i.e. coupling or

disproportionation. This means the radical intermediate will be dead in a very short time, which makes it difficult to control the characteristics of the resulting polymer, i.e. molecular weight and dispersity, architecture, and end-group functionalization.101

31

Different from conventional free radical polymerization, the termination process are

greatly suppressed by introducing dormant states for the propagating species in CFRP

approaches. Dormant species and propagating species can be switched via either reversible termination or reversible transfer, which significantly prolong the lifetime of the propagating species.101,115 Fast initiation is also very important, which allows

all propagating radicals to grow for approximately the same time.

A schematic representation of the mechanism of CFRP is illustrated in Figure

2.11. At first, the initiator should be able to quickly produce a reactive and a stable

radical. The reactive radicals initiate the polymerization while the stable radicals

couple with the propagating species to form dormant species that do not participate in

the propagation. The coupling of stable radicals to the propagating species is

reversible, i.e. there is an equilibrium between the dormant and the propagating

species. This equilibrium is the key to make the polymerization controllable. The

dormant species should not be too stable to convert back to propagating species, but

should be stable enough that the concentration of the propagating species can be

lowered and thus the bimolecular termination can be effectively suppressed. In this

case, the lifetime of the radicals is significantly increased and the polymerization is

close to the state of “living polymerization”.

32

Figure 2.11 A general mechanism of controlled free radical polymerizations. Reproduced with permission from ref 115.

2.5.2 RAFT polymerization

RAFT polymerization is one of the most versatile CFRP techniques because of its ability to polymerize a wide variety of monomers and its high tolerance to monomer functionality and reaction conditions.116,117 Since its first report published in

1998 by the Commonwealth Scientific and industrial Research Organization

(CSIRO),116 RAFT polymerization has been extensively studied.117,118,119,120,121

33

Figure 2.12 Schematic of typical thiocarbonylthio RAFT agents and the formation of intermediate radicals. Reproduced with permission from ref.118.

The structural features of a typical RAFT agent include a reactive C=S double bond, a Z-group modifying the addition and fragmentation rates, and a free radical leaving group R, as shown in Figure 2.12. When subjected to a free radical polymerization, a practically useful RAFT agent acts as a mediating agent that reversibly couples with the propagating radicals to form the dormant species.

Therefore, the polymerization can achieve controlled manner. A generally accepted mechanism for RAFT polymerization is shown in Figure 2.13. The key feature of the mechanism is the addition-fragmentation equilibria. First, the propagating radicals are added to the thiocarbonylthio compounds to form a radical thiocarbonylthio intermediate, which subsequently fragments to a polymeric thiocarbonylthio compound and a new radical (a R-group radical or another polymeric radicals). The leaving radicals should be able to reinitiate the polymerization and form new propagating radicals. Due to the regulation of the equilibria, the termination is effectively suppressed and all of the propagating chains share equal probability to grow, which leads to a polymer with low polydispersity and controlled molecular weights. In addition, most of the polymer chains retain the thiocarbonylthio groups when the polymerization is terminated. This provides opportunities for further RAFT polymerization or post-polymerization modifications. 34

Figure 2.13 Mechanism of RAFT polymerizations. Reproduced with permission from ref.118.

Based on the mechanism shown in Figure 2.13, it is clear that the choice of R and Z-group is crucial for a successful RAFT polymerization. For R-group, the leaving R-group radicals should be able to reinitiate the polymerization. An appropriate Z-group should be chosen so that the intermediate 2 and 4 can rapidly fragment, i.e. high k , and the intermediate 2 favors a formation of 3, i.e. k k .

β β add The general guidelines for choosing R- and Z-group are summarized in≥ Figure

2.14.118,122

35

Figure 2.14 Guideline for selection of Z-group (top) and R-group (bottom) for different monomers. Dash line indicates partial control. For Z-group, fragmentation rate increases while the addition rate decreases from left to right. For R-group, fragmentation rates decreases from left to right. Reproduced with permission from ref. 122.

2.5.3 Complex macromolecular architecture via RAFT polymerization

One of the most important missions for polymer chemists is to produce precisely designed macromolecular architectures that are potentially useful for novel applications. This could not be done without ionic polymerizations until the

development of CFRP techniques, which have been demonstrated as versatile and

powerful tools for constructing complex macromolecular architectures.101,114 RAFT

polymerization, as one of the most successful CFRP techniques, has been employed to

prepare block copolymers, star polymers, comb polymers, etc., as shown in Figure

2.15.117

36

Figure 2.15 Examples of complex macromolecular architecture prepared by RAFT polymerization. Reproduced with permission from ref. 117.

The synthesis of block copolymers using RAFT polymerization has been

extensively studied117,118,119,120,121 and the guidelines123 were recently given. The most straightforward route for making block copolymer is sequential polymerization of two types of monomers, as shown in Figure 2.16. Ideally, a diblock copolymer is expected

to be the main product after the second polymerization. However, a more detailed

examination of the stepwise polymerization reveals that the sequential polymerization

is more complicated, as shown in Figure 2.17. This implies a small fraction of defects

is unavoidable for the block copolymer produced by this means, which can be

minimized by carefully selecting the order of monomer addition and the amount of

radical initiators.123

37

Figure 2.16 A simplified scheme for the synthesis of block copolymers via sequential RAFT polymerization. Reproduced with permission from ref. 117.

Figure 2.17 An illustration of the polymer species generated after two-step RAFT polymerization for making diblock copolymers. In this illustration, [M]/[RAFT]/[I] for both steps is set as 72/10/1. The initiator efficiency and the monomer conversion are both set as 1. Reproduced with permission from ref.123.

Besides diblock copolymers, linear triblock and multiblock copolymer have also been prepared via RAFT approach. 124 , 125 , 126 Polymers with nonlinear

architectures, e.g. graft copolymer,127 comb copolymer,127,128 star copolymer,127,129,130

dendrimers131,132 and hyperbranched polymers,126,133 were all reported recently.

38

2.5.4 End-functional polymer via RAFT approach

RAFT polymerization allows for the introduction of a wide variety of

functional groups through either RAFT agent design or postpolymerization

modification of the retained thiocarbonylthio functionality, as shown in Figure

2.18.117,122,134

Figure 2.18 Three routes that can introduce end-functionality to the polymer via RAFT polymerization. R’ refers to the new ω-group transformed from thiocarbonylthio group. Reproduced with permission from ref.117.

The functional groups introduced via R- or Z-group will be retained as α- or

ω-functional group after RAFT polymerization, though a small fraction of polymer chains will lose these functional groups during polymerization. 135 The R-group is

usually favored as the functional group, because the Z-group will be lost if the

thiocarbonylthio group is degraded/removed in the polymerization or if the functional

group is to be modified via organic transformation.122 Theoretically, all of the functional groups compatible with thiocarbonylthio can be introduced via R- or Z-

group.

39

Figure 2.19 Transformation of the thiocarbonylthio group. R’· = radicals, [H] = hydrogen donor, M = monomer. Reproduced with permission from ref. 135.

End-group functionality can also be introduced by transforming the thiocarbonylthio groups. Different strategies have been established, as summarized in

Figure 2.19134 and other recent reviews.117,136 In particular, the thiocarbonylthio group can react with nucleophilies and generate thiol groups, which allows for more reactions, as shown in Figure 2.20.134

40

Figure 2.20 Reaction of the thiol group transformed from thiocarbonylthio group. Reproduced with permission from ref.134.

41

CHAPTER III

SUPRAMOLECULAR MULTIBLOCK POLYSTYRENE-POLYISOBUTYLENE

COPOLYMER VIA IONIC INTERACTIONS*

3.1 Introduction

Supramolecular block copolymers, which are the supramolecular analog of

covalently-bonded block copolymers, consist of individual polymer blocks connected

by non-covalent bonds.2 They can be produced by self-assembly of telechelic

oligomers or polymers with complementary end-groups, e.g., hydrogen bonding or

acid-base interactions, such that a variety of block combinations may be achieved by

simple mixing of the appropriate polymers.3,60, 137 , 138 , 139 Supramolecular block

copolymers are advantageous for fabricating nanostructured functional materials.

While they can exhibit morphologies similar to conventional covalently-bonded block copolymers, the reversible nature of the non-covalent bonding between blocks allows for unique responses to external stimuli, where the equilibrium between associated and unassociated groups can be sensitive to temperature, pH or mechanical stress.60, 140 Such stimuli-responsive equilibria afford additional opportunities for

controlling the phase behavior and properties of such systems. Various applications of

these materials have been explored, such as self-healing,50 thermally tunable

nanostructures,51,57 nanoporous materials,59 ,79,80 and nanostructured assemblies.60,141

Much of the recent work on supramolecular block copolymers has focused on hydrogen bonding or metal-ligand coordination for the supramolecular bonds. 2,3,60,137,

42

139,139 Ionic interactions have also been recognized as a means for constructing

supramolecular structures and nanostructured materials, e.g., dye-surfactant complexes,4 polymer-surfactant complexes,5 ionomers,6,7 and polyelectrolyte

complexes.8,9 The utility of ionic interactions in supramolecular chemistry arises from a combination of properties as follows. First, ionic bonds are stronger and less- directional compared to other physical interactions;4 second, they may form larger aggregated structures depending on the steric environment of the ion pair; third,

Coulombic interactions are asymmetric and sensitive to the local constant of the medium they are in;138 and fourth, ionic interactions are easily tunable through the

choice of anion (e.g. sulfonate vs. carboxylate) and cation (e.g. primary or secondary

amine, quaternary ammonium), many of which are accessible through straightforward

chemistry.94,95,107,142

The use of Coulombic interactions in polymer blends, as well as the entire

field of ionomers that dates back to the 1950’s, predates the field of supramolecular

polymers, which gained traction in the early 1990s.1,43,143,144 The majority of that

work involved improving the miscibility of polymer blends using proton transfer

between acidic and basic species on different polymers 145, and although the term

“supramolecular” was not used in the reports of that research, the materials were

essentially supramolecular graft-block copolymers or networks.

The first examples of linear (AB)n supramolecular multiblock copolymers

prepared through ionic interactions were reported by Jerome and co-

workers62,63 ,68 ,69 ,70 nearly 30 years ago. They studied mixtures of telechelic ionomers

of polystyrene (PS) or poly(α-methylstyrene) (PαMS) with carboxylic or sulfonic acid

end-groups (block A) and polyisoprene (PI) with tertiary amine end-groups (block B).

The polymers were solution blended and proton transfer from the acid to the amine

43

caused self-assembly of a block copolymer structure. The thermodynamic phase behavior of these systems was studied directly and indirectly by SAXS and oscillatory

63,69 shear rheology, respectfully. Mixtures of the R2N-PI-NR2 and PS-COOH macrophase separated; while mixtures with the dicarboxylic acid ionomers were microphase separated at room temperature. When heated these systems underwent macrophase separation at ca. 130 °C due to the dissociation of the ammonium carboxylate ion-pairs. The analogous polymers using sulfonated telechelics formed much stronger ion-pairs and were found to dissociate at higher temperature (160 –

180 °C). Order-disorder transitions could be observed for these block copolymers on heating prior to macrophase separation depending on the overall molecular weight of the block copolymer.

These ammonium sulfonate bonded polymers showed little variation of the domain spacing with temperature, contrary to what has been commonly observed in covalently bound systems where the domain spacing decreases with increasing temperature.12 An opposite result was obtained by Huh et al. in a mixture of hemi- telechelic primary amine terminated polyisoprene (PI-NH2) and telechelic sulfonate

51 terminated polystyrene (HO3S-PS-SO3H). Here the domain spacing increased by ca.

30 nm when the system was heated from 100 to 200 °C. This was attributed to ion- pair dissociation, resulting in the formation of PI and PS homopolymers, which swelled their respective domains. Noro et al. prepared blends using monofunctional

PI-NH2 and PS-SO3H. Depending on the stoichiometry different nano-objects (e.g. spherical micelles, worm-like micelles and lamellar sheets) were observed.58 However, macrophase separation was observed in all of these blends, which was attributed to the difficulty of ion-pair formation during casting from THF/toluene/water (66/33/1)

44

solutions, which are more polar than the toluene solutions used in the previous two examples.

Although these materials possess interesting properties due to the strong interactions of ammonium sulfonate ion-pairs, the potential to prepare nanostructured block copolymers by mixing and casting of acid/base telechelic polymer blends, and the flexibility to tune structure through the blend composition, these materials have not been widely studied since they were first reported. This is especially striking when compared to the attention given to other supramolecular block copolymer systems, such as those produced through hydrogen bonding or metal complexation. One reason may be that hydrogen bonding and metal complexation are the more widely used methods in small molecule supramolecular chemistry, and thus are naturally chosen for supramolecular polymers, perhaps without consideration for why they are preferred in small molecule systems. Ionic interactions produce higher order aggregates, such as ion-pair multiplets, that could impede the design of supramolecular small molecules; however, aggregation is better tolerated in polymeric systems where the ionic groups are compatible with the overall polymer self-assembly and microphase separation.

Most efforts on ammonium sulfonate linked block copolymers have concentrated on various architectures,7 such as zwitterionic,146,147 star,89,148 and graft block copolymers,66,83,84 or the synthesis of nanobjects/nanostructures. 51,61,79,80 In contrast there has been little investigation of the viscoelastic behavior or the mechanical properties of these types of polymers, which are important if these polymers are to be used in bulk material applications. Another reason for the more limited investigation of these types of systems is the difficulty in synthesis of the telechelic ionomers. For example, end-sulfonated polystyrene has been typically

45

prepared by anionic polymerization followed by termination with propane sultone, a

known carcinogen. One of the authors recently reported a method to prepare end-

sulfonated polystyrene (PS-SO3H) using RAFT polymerization and oxidation of the trithiocarbonate groups with m-chloroperxoybenzoic acid, which simplifies the polymer preparation.110 In this report low molecular weight telechelic polystyrene

(HO3S-PS-SO3H) and polyisobutylene (H2N-PIB-NH2) prepared by RAFT and

cationic polymerization, respectively, were investigated to prepare supramolecular

multiblock copolymers. Blends of these polymers were solution cast to obtain a clear,

flexible, free-standing film with an ordered lamellar microstructure and small domain size. The temperature dependent morphological and linear viscoelastic behavior were measured and compared to previously studied linear ammonium-sulfonate linked supramolecular block copolymers. In addition, the room temperature mechanical properties, and the nonlinear rheological behavior were also measured and discussed.

3.2. Experimental Section

3.2.1 Materials

Hexane (anhydrous, 95%), tetrahydrofuran (THF) (anhydrous, 99.9%), 1-

methyl-2-pyrrolidinone (NMP) (anhydrous, 99.5%), hydrazine monohydrate (98%),

ethanol (99.5%), 2,6-lutidine (redistilled, 99.5%), titanium tetrachloride (TiCl4)

(99.9%), dichloromethane (DCM) (99.8%), (3-bromopropoxy)benzene (96%), phthalimide potassium salt (98%), chloroform and chloroform-d (CDCl3) were

purchased from Sigma-Aldrich and used as received. Heptane, methanol (99.9%), toluene (99.5%) and anhydrous magnesium sulfate (MgSO4) were purchased and used as received from Fisher Scientific. Isobutylene (BOC Gases) and methyl chloride

(Alexander Chemical Corp.) were dried by passing the gases through columns of

46

CaSO4/CaCl2/molecular sieves and condensed within a N2-atmosphere glovebox

immediately prior to use. Styrene (99%, stabilized, Acros Organics) was purified by

passing over a column of basic alumina. Silica gel (Dynamic Adsorbents 60Å, 32-63

µM, flash grade) was used for column chromatography. m-Chloroperoxybenzoic acid

(mCPBA, 70-75%, balance 3-Chlorobenzoic acid and water) was supplied from Acros

Organics. The difunctional RAFT agent, didodecyl-1,2-phenylene-bis(methylene)-

bistrithiocarbonate, bis-RAFT, was synthesized as previously reported. 149 The

difunctional cationic initiator, 1,3-di(1-chloro-1-methylethyl)-5-tert-butylbenzene

(bDCC) was synthesized as previously reported150 and stored at 0°C. The source and purity of all other reagents used for carbocationic polymerization and PIB end-group modification have been described elsewhere151,152. All of the other chemicals, unless

otherwise indicated, were obtained commercially and directly used without further

purification.

3.2.2 Synthesis of end-functionalized (telechelic) oligomers

Telechelic sulfonated polystyrene, HO3S-PS-SO3H, (Mn = 6500 Da; PDI =

1.19) was prepared by RAFT polymerization using a difunctional RAFT agent,

didodecyl-1,2-phenylene-bis(methylene)-bistrithiocarbonate149, followed by reaction

with mCPBA,110 as shown in Figure 3.1. Styrene (2.495 g, 24.0 mmol) and bis-RAFT

(0.137 g, 0.208 mmol) were added to a 10 mL flask equipped with a magnetic stir bar and sealed with a rubber septum. The sealed flask was purged with dry nitrogen for 15 min and placed in a preheated oil bath at 130 ºC. The polymerization was allowed to proceed for 6 h and then the temperature was quenched by placing the flask in ice water. The polymer, PS-RAFT, was isolated by precipitation in methanol (150 mL).

The precipitate was isolated by filtration, washed three times with methanol, and dried

47

in a vacuum oven at room temperature. The number average molecular weight, Mn,

and molecular weight dispersity, Đ, determined by size exclusion chromatography,

SEC, were 6900 Da and 1.11, respectively.

130 oC + 6 h

(1)

(1) + Toluene, 0 oC 6 h

Figure 3.1 Synthesis of PS(SO3)2.

The PS-RAFT was directly treated with mCPBA to oxidize the

trithiocarbonate group to sulfonic acid.110 PS-RAFT (1.022 g, 0.150 mmol) was

dissolved in 5 ml toluene in a 15 ml flask in an ice bath, mCPBA (1.2563 g, 5.10

mmol) was then added and the reaction was allowed to proceed for 6 h at 0 °C. The

HO3S-PS-SO3H product was isolated by precipitating the reaction mixture in

methanol (150 mL), filtered, washed three times with methanol, and dried under

vacuum. The HO3S-PS-SO3H was further purified by silica gel column

chromatography by eluting first with chloroform and then with THF. The purified

HO3S-PS-SO3H was eluted as a THF-soluble fraction. The molecular weight of

HO3S-PS-SO3H could not be accurately determined by SEC due to interactions

between the sulfonic acid groups and the SEC column. Therefore, the molecular

weight of HO3S-PS-SO3H, 6510 Da was estimated by subtracting the difference 48

between the molecular weight of the end-groups of HO3S-PS-SO3H and PS-RAFT,

which is 392 Da, from Mn of the PS-RAFT (6900 Da). The end-functionality,

+ - - determined by acid-base titration, was ca. 95%. A control sample, H9C4H3N O3S -

- + PS-SO3 N H3C4H9, was prepared by neutralizing HO3S-PS-SO3H with a 10-fold

excess of n-butylamine in toluene and then dried in vacuum. The excess amount of n-

butylamine ensured complete neutralization of the sulfonic acid in the HO3S-PS-

SO3H. The free n-butylamine (vapor pressure of 9.1 kPa at 20 °C) that was not

ionically bound to the HO3S-PS-SO3H was removed by vacuum.

Telechelic primary-amine-terminated polyisobutylene, H2N-PIB-NH2, (Mn=

6,200 Da; Đ = 1.02) was prepared by living carbocationic polymerization of

isobutylene, in situ quenching with (3-bromopropoxy)benzene, and subsequent

Gabriel synthesis to convert the bromide group to primary amine group,151 as shown

in Figure 3.2. The difunctional cationic initiator, 1,3-di(1-chloro-1-methylethyl)-5- tert-butylbenzene (bDCC) was synthesized as previously reported150 and stored at

0 °C. Prechilled hexane (134 mL) and methyl chloride (201 mL) (-70 °C) were added to a four-neck 1 L round bottom flask equipped with a temperature probe, infrared probe, and an overhead mechanical stirrer, inside a glove box with a dry nitrogen atmosphere. bDCC (3.48 g, 12.1 mmol) and 2,6-lutidine (0.16 mL, 1.3 mmol) were added to the flask, and the mixture was stirred for 15 min, after which isobutylene

(107 mL, 1.337 mol) was charged to the flask. The cationic polymerization was initiated by injecting TiCl4 (0.76 mL, 7.0 mmol) to the reactor.

49

Figure 3.2 Synthesis of PIB(NH2)2.

Complete conversion of the isobutylene monomer was determined by infrared

monitoring, and at that time the quenching agent (3-bromopropoxy)benzene (7.62 mL,

48.4 mmol) and an additional amount of TiCl4 (4.54 mL, 41.4 mmol) were added to the reaction mixture to obtain primary bromide-terminated PIB. After 5 h, excess methanol was charged to the flask to terminate the reaction. The PIB was precipitated by the addition to methanol and then re-dissolved in hexane. The resulting solution was washed three times with DI water. The organic phase was collected, dried over

MgSO4, filtered, and dried in a rotary evaporator to yield the primary bromide- terminated PIB product as a viscous liquid (Br-PIB-Br).

The Br-PIB-Br (50 g) was dissolved in 200 mL THF, and the resulting solution was diluted with 100 mL NMP. Potassium phthalimide (30 g) was added, and the reaction mixture was refluxed at 85 °C for 4 h. The reaction mixture was cooled to room temperature and then washed three times with DI water. The organic phase was collected, dried over MgSO4, filtered, and rotary evaporated to yield the phthalimide-

terminated PIB product as a viscous liquid (PI-PIB-PI). 50

The PI-PIB-PI (50 g) was dissolved in heptanes (200 mL), and the resulting

solution was diluted with ethanol (200 mL). Hydrazine hydrate (25 g) was added to

the flask, and the reaction mixture was refluxed at 105 °C for 5 h. After cooling to

room temperture, the reaction mixture was collected, dried over MgSO4, filtered and

dried in a rotary evaporator to yield the primary amine-terminated PIB product as a

viscous liquid (H2N-PIB-NH2). The product was further dried in a vacuum oven at

room temperature. Quantitative end group transformation was confirmed by 1H NMR, as shown in Figure 3.3 to Figure 3.5. From Figure 3.3 to Figure 3.4, resonance due to

the proton a completely shifted from 3.6 ppm to 3.9 ppm, which indicated the

bromide group was completely transformed to phthalimide group. From Figure 3.4 to

Figure 3.5, resonance due to the proton a completely shifted from 3.9 ppm to 3.1 ppm

when the end-group was transformed from phthalimide group to primary amine group.

51

Figure 3.3 1H NMR spectrum of Br-PIB-Br.

52

Figure 3.4 1H NMR spectrum of PI-PIB-PI.

53

011814-DI-PIB-NH2 b h g O NH2 j hi f ac n d e g

e, j CHCl 3 i,f d c a b

7 6 5 4 3 2 1 Chemical Shift (ppm)

1 Figure 3.5 H NMR spectrum of H2N-PIB-NH2.

3.2.3 Preparation of the Supramolecular Block Copolymer

The supramolecular block copolymer was prepared by dissolving equimolar amount of HO3S-PS-SO3H and H2N-PIB-NH2 separately in toluene to make two 10%

w/v solutions. The H2N-PIB-NH2 solution was then added dropwise to the stirred

HO3S-PS-SO3H solution, and the blend solution was stirred for another 18 h. The

clear solution was then cast into a Teflon dish and the toluene was slowly evaporated

in the hood for several days. The final film was dried to constant weight in a vacuum

oven at room temperature.

54

3.2.4 Characterization

A Waters Breeze size exclusion chromatograph (SEC) equipped with three

Styragel columns at 35°C and a refractive index detector (Waters 2414) was used to measure number average molecular weight, Mn, and molecular weight dispersity, Đ of

the polymers. The Styragel column set consists of a Styragel® HR 4 THF column (4.6

× 300 mm) with an effective molecular weight range 5 k to 600 k Da, a Styragel®

HR 3 THF column (4.6 × 300 mm) with an effective molecular weight range 0.5 k to

30 k Da, and a Styragel® HR 4E THF column (4.6 × 300 mm) with an effective molecular weight range 0.05 k to 100 k Da. The SEC was calibrated using PS

standards of narrow molar mass dispersity (Ð) with the molecular weight of 1300 Da

to 400 k Da (Alfa Aesar). Optical microscopy was performed with an Olympus BX51

microscope equipped with a digital camera (Olympus Q-Color-5) and an INSTEC

1 HCS302 heating stage. H NMR spectra of CDCl3 solutions were acquired using a

Varian 500 MHz NMR spectrometer.

Differential scanning calorimetry (DSC) was performed with a Q200 DSC from TA Instruments using a heating rate of 10°C/min and a dry nitrogen atmosphere.

Except for the H2N-PIB-NH2, the samples were first heated to 150 °C and then cooled

to room temperature and reheated to 150 °C. The glass transition temperature (Tg) was

defined as the midpoint in the heat capacity change at the glass transition during the

second heating scan. For the H2N-PIB-NH2, the sample was cooled to -90 °C and then

reheated to 50°C.

Fourier transform infrared (FTIR) spectra were obtained with a NICOLET-

380 FTIR spectrophotometer using the transmission mode. The samples were first

dissolved in toluene, and the resulting solution was cast on KBr pellets and finally

55

dried in a vacuum oven before FTIR analysis. An average of 512 scans was taken for

each run at a resolution of 4 cm-1.

Room temperature small-angle X-ray scattering (SAXS) experiments were measured with a Rigaku SAXS instrument using an 18 kW rotating anode X-ray generator (MicroMax002+) and CuKα (λ = 0.154 nm) radiation operated at 45 kV and

0.88 mA at the University of Akron. The instrument was calibrated with silver behenate. The exposure time for each sample was 300 s. The background scattering was subtracted and analyzed using Saxsgui software (JJ X-Ray Systems ApS).

Elevated temperature SAXS was measured between 110°C and 250°C using a

Rigaku S-MAX3000 pinhole SAXS camera and a MicroMax-002+ Microfocus sealed tube X-ray source at Drexel University. The sample was first packed into a circular cavity of 2 mm diameter in an aluminum sample holder of 1 mm thickness, and then the cavity was sealed by Kapton films using epoxy resin. The sandwiched sample was mounted to a Linkam high temperature control stage to control the temperature from

110 °C to 250 °C. The heating rate was 20 °C/min, and samples were annealed at each temperature for 15 min before data collection. The scattering data were collected using a Gabriel two dimensional multiwire X-ray area detector. Data analysis was done using a Matlab-based Graphical User Interface (GUI).

Linear and nonlinear oscillatory shear, melt-rheology measurements were performed with an Advanced Rheometric Expansion System (ARES) G2 from TA

Instruments. A parallel plate fixture with a diameter of 8 mm was used, and the linear response region for each sample was determined by a strain sweep. For linear measurements, a strain amplitude of γ = 3 - 5% was used for isochronal temperature sweeps and isothermal frequency sweeps. For the isochronal temperature data, measurements were conducted from 40 °C to 250 °C with γ = 3% and frequency, ω =

56

10 rad/s. For isothermal frequency data, measurements were performed with a

frequency range of 0.01 - 100 rad/s over a temperature range of 110 – 250 °C with γ =

5%. Nonlinear oscillatory shear rheology experiments consisted of three consecutive

cycles measured at a constant temperature of 150 °C. A single cycle consisted of a

strain-sweep from γ = 0.01% to 100% at constant frequency, ω = 10 rad/s, followed by a 5-min period within the linear viscoelastic region with ω = 10 rad/s and γ = 1%.

In general, non-linear viscoelastic response was observed at the higher oscillatory frequency, and the ensuing period of a linear viscoelastic deformation was used to follow the structure recovery of the polymer.

Dynamic mechanical properties in simple extension and static tensile

properties of solid samples were measured using a TA Instruments Q800 dynamic

mechanical analyzer (DMA) equipped with a tensile film fixture. The samples were

compression-molded into rectangular bars (~20 mm × 4 mm × 0.5 mm). The complex

stress was measured from -100 °C to 95 °C using a displacment of 1 μm, a heating

rate of 2 °C/min, a preload force of 0.001N, and a frequency of 1 Hz. The DMA software was used to calculate the dynamic modulus, E’, the loss modulus, E”, and tan δ = E”/E’. The tensile properties (modulus, elongation at break, stress at break) were measured with the DMA using a preload force of 0.001N to keep the sample taut, which provided an initial displacement of 0.1%, and an elongation rate of 1%/min.

3.3 Results and discussion

Mixing the two telechelic oligomers, one a viscous liquid (PIB) and the other a brittle plastic (PS), produced a clear, flexible, self-standing film, as seen in Figure 3.6.

The flexible film is a demonstration of the formation of a high molecular weight

57

supramolecular multiblock copolymer (SMBCP), due to the multiple proton transfer

reactions between the end groups of the two telechelic oligomers. The absence of macrophase separation of an otherwise immiscible mixture of polystyrene and polyisobutylene, inferred from the transparency of the sample, also supports the

formation of the ionic bonds, which prevent large scale phase separation. The next

section provides detailed evidence for the formation of a multiblock copolymer.

Figure 3.6 (a) Synthesis of (PIB-b-PS)n SMBCP (b) Flexibility and clarity of SMBCP film.

58

3.3.1 Analytical Evidence for the Formation of a Supramolecular Block Copolymer

The formation of the ionic bond between the sulfonic acid and primary amine

groups was verified by FTIR and 1H NMR analyses. Figure 3.7 shows the infrared

-1 + - spectral region 1100 – 1000 cm for H2N-PIB-NH2, HO3S-PS-SO3H, H9C4H3N O3S -

- + -1 PS-SO3 N H3C4H9, and the stoichiometric PS/PIB blend. The vibration at 1050 cm

in Figure 3.7b is the symmetric stretching vibration of the SO3- ion from the SO3H

153 groups in the HO3S-PS-SO3H. For the blend (Figure 3.7c) the absorbance due to

-1 -1 the symmetric SO3- stretching shifted from 1050 cm to a shoulder at 1039 cm as a

result of the neutralization of the sulfonic acid to the ammonium salt. The 1028 cm-1

peak is due to the polystyrene. The presence of the ammonium sulfonate confirms the

formation of ammonium sulfonate ionic bonds in the blend.

59

1039

d

1039 1080 c

1050 Absorbance b

1070 a

1100 1050 1000 -1 Wavenumber (cm )

Figure 3.7 FTIR spectra of (a) H2N-PIB-NH2, (b) HO3S-PS-SO3H, (c) stoichiometric + - - + blend and (d) H9C4H3N O3S -PS-SO3 N H3C4H9 in the spectral region of 1100 – 1000 cm-1.

A similar result was obtained when the HO3S-PS-SO3H was neutralized with a

+ - - + low molar mass alkyl amine, n-butylamine, to H9C4H3N O3S -PS-SO3 N H3C4H9

- -1 (Figure 3.7d). As in Figure 3.7c the symmetric SO3 stretching vibration at 1050 cm

shifted to ~1039 cm-1, which indicates complete proton transfer as the sulfonic acid

groups were converted to ammonium sulfonate. The weak, broad peak at 1070 cm-1 in

Figure 3.7a is from the C—N stretching vibration of the primary amine end-groups

60

154 -1 from the H2N-PIB-NH2 sample. That peak shifted to 1080 cm in the blend sample due to the protonation of the primary amine by the sulfonic acid155, which provides additional proof of the ionic bond formation.

The proton transfer to form an ionic bond was also confirmed by 1H NMR spectroscopy as shown in Figure 3.8. The protons on the methylene groups adjacent to the amine in the H2N-PIB-NH2 (a and b in Figure 3.8b) shifted upfield in the spectrum for the blend, Figure 3.8a, which indicates protonation of the primary amine by the sulfonic acid.

O NH3SO3 multi.esp (a) b’ a’ n

b’ a’

3.81 2.76

O NH2

011814-DI-PIB-NH2 ab (b) n

b a

4.13 3.02 7 6 5 4 3 2 1 Chemical Shift (ppm)

1 Figure 3.8 H NMR of (a) stoichiometric blend and (b) H2N-PIB-NH2. Protons a and b were shifted upfield due to the ion complexation.

61

DSC thermograms of the telechelic oligomers and the blend are shown in

Figure 3.9. The Tg of the HO3S-PS-SO3H was 96 °C, which is considerably higher than the 85 ºC prediction of the Fox-Loshaek equation 156 for an unfunctionalized

polystyrene of the same molecular weight. The higher value is a consequence of the

supramolecular behavior of the HO3S-PS-SO3H telechelic ionomer, where hydrogen bonding between sulfonic acid groups produces chain extension of the oligomer and reduces the concentration of chain-ends. The Tg of the H2N-PIB-NH2 was -78 °C,

which is a little lower than the -73 °C prediction of the Fox-Loshaek equation. In this case, the deviation may be due to the flexible ether bond in the end groups of the

H2N-PIB-NH2.

The blend exhibited two glass transitions that corresponded to a PS-rich phase

(77 °C) and a PIB-rich phase (-58 °C). The decrease of the PS Tg and the increase of

the PIB Tg relative to the neat oligomers are consistent with the formation of an

SMBCP with nanoscale domains, as these types of Tg shifts have previously been

observed in similar SMBCPs and confined polymer blends.62, 157

62

exotherm

Figure 3.9 DSC heating thermograms of (a) H2N-PIB-NH2, (b) 50/50 blend, and (c) HO3S-PS-SO3H. The heating rate was 10ºC/min.

3.3.2 Morphology of the blend

Evidence of block copolymer formation and microphase separation of the blend was provided by small angle X-ray scattering (SAXS) as shown in Figure 3.10 and Figure 3.11. The as-cast film exhibited three distinct SAXS peaks with values of q/q* of ~1: 2: 3, where q* is the wavevector for the first-order reflection. This sequence of peaks is characteristic of an ordered lamellar microstructure, which would be expected for a block copolymer with nearly equal size blocks. From q*, a domain spacing of 14.4 nm was determined.

63

q*

2q*

3q* 4q* Relative Log Intensity (a.u.) Log Intensity Relative 0.0 0.5 1.0 1.5 2.0 2.5 3.0 q (nm-1)

Figure 3.10 Azimuthally averaged SAXS pattern for the room temperature blend.

The SAXS patterns measured on heating and cooling are shown in Figure

3.11(a) and the domain spacing, the reciprocal first order peak intensity, I(q*)-1, and

first order peak width are shown in Figure 3.11(b), (c), and (d). The sharp increase in

the peak width and the increase in I(q*)-1 from 190 to 210 °C are both indicative of an

order-disorder transition of the SMBCP. Upon cooling the system orders between 210

to 190 °C indicating that the ODT is thermoreversible. The domain spacing of a

multiblock copolymer is dependent on the degree of segegration (χN),158 the number

of blocks, 159 and the unperturbed chain dimensions of the AB repeat unit. The

2 1/2 unperturbed end-to-end distances for the PS and PIB blocks are (o,PIB) = 5.9

2 1/2 2 nm and (o,PS) = 5.3 nm based on the o/M values reported at 140°C by

Fetters et al. from SANS measurements of the bulk homopolymer radii of gyration.160

This gives an unperturbed end-to-end distance of the AB diblock repeat of 11.2 nm at

140 °C. Averaging the domain spacings measured at 130 and 150°C gives D = 13.8

nm, which gives a ratio of stretching ratio of 1.2. This domain spacing is also 64

significantly smaller than the unperturbed domain spacing of the AB diblock

copolymer (22.4 nm) providing further evidence that a multiblock copolymer is

formed.

(b) 17 (a) 16 15 14

150 oC 13 Domain Spacing (nm) o 170 C 2.0 190 oC (c) 210 oC 1.5 230 oC o 1.0 250 C (a.u.) -1

o m 230 C I 0.5 210 oC 190 oC 0.0 RelativeIntensityLog 170 oC (d) o 0.20

150 C ) 130 oC -1 0.15 110 oC 0.3 0.6 0.9 1.2 1.5 0.10 -1 0.05

q (nm ) FWHM(nm 0.00 2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7 -1 1000/T (K )

Figure 3.11 (a) Temperature-resolved SAXS profiles of SMBCP measured at 20°C intervals during heating from 110°C to 250°C and then cooling to 150°C. (b) Temperature dependence of domain spacing, (c) inverse maximum intensity for the first-order scattering peak and (d) full-width at the half-maximum (FWHM) for the SAXS of the SMBCP. The filled symbols and open symbols represent heating and cooling data, respectively. Calculations for some temperatures were not possible, because of an ill-defined scattering peak.

The stretching ratio of the domain spacing to the unperturbed end-to-end distance of the diblock repeat unit of an infinite multiblock copolymer, D/aN1/2, as a function of χN was previously calculated by self-consistent mean field theory.158 To compare to this theoretical value, the χN for the PS-PIB multiblock must be estimated.

Based on the reported densities of PS and PIB at 413K by Fetters et al. the geometric average molar volume of the PS and PIB repeat units is 84 cm3/mol, which gives a 65

degree of polymerization of 167 for the AB diblock repeat unit of the (AB)n

multiblock.160 χ = 305K/T – 0.56 was determined by interpolation of the

concentration dependent χ at 35 °C for a PS/PIB blend161 and the χ at 250 °C from the TODT of PS-PIB-PS triblock copolymer based on the calculated phase

diagram.162,163 This gives a χN at 140 °C of 30, where a D/aN1/2 of 1.24 is predicted,

which shows good agreement with the calculated D/aN1/2 of 1.2 based on the SAXS

data.

The domain spacing vs. temperature data show two differences compared to

what would be predicted for a covalently bonded analog. First, the domain spacing

increases with increasing temperature above 150°C, which is the opposite of what

would be expected for a covalently bonded block copolymer. This is likely due to

dissociation of the ionic bonds with increasing temperature. The domain spacing for

159 an (AB)n multiblock is known to increase with decreasing n, with a gradual

increase at large n. The domain spacing was also observed to increase ca. 14% upon

cooling from the temperature peak of 250 °C. This may be due to both the ion-pair

dissociation and the thermal decomposition of the ammonium sulfonate ion-pair,

which has been reported to occur above 180 °C.69,164 Ion-pair dissociation is expected to be a reversible process, but could be kinetically hindered, while thermal decomposition would be an irreversible process. Both would reduce the average n of the multiblock, resulting in larger domain spacing.

66

a b c

Figure 3.12 Polarized optical micrographs of SMBCP at (a) 185 °C, (b) 190 °C and (c) 195 °C. Each picture was taken after holding the sample at the temperature indicated for 30 min. The scale bar in each photograph is 250 μm.

The decomposition of the ionic functional groups at elevated temperatures was investigated by optical microscopy, FTIR, and 1H NMR .Optical micrographs are

shown in Figure 3.12 for samples annealed at 185, 190 and 195 °C for 30 min at each

temperature. The formation of droplets at 195 °C is likely due to the macrophase

separation of the sample due to the decomposition of the ionic bonds. This also

corresponds to the temperature at where the ODT occurs during SAXS. The

reversibility of the SAXS data with temperature suggests that they are independent

events i.e., that the blend exhibits a true ODT, similar to that of conventional block

copolymers, but the ionic groups begin to decompose in the same temperature region.

The decomposition rate may also be slower in the SAXS experiments where the

sample is annealed under vacuum, while in the optical microscopy it is only protected

from air by a glass slide. After annealing at 195 °C, the FTIR intensities of the

sulfonate and ammonium ion peak at 1039 cm-1 and 1080 cm-1 were significantly

reduced (Figure 3.13), indicating a decrease in the concentration of the ionic bonds in

- the blend. In addition, neither the symmetric stretching vibration of the SO3 of the

sulfonic acid at 1050 cm-1, nor the N-H stretching vibration at 1070 cm-1 for a primary amine were observed in the heat-treated sample, which indicates that the decrease in the ionic bond concentration was not due to reversion of the ionic bond to its sulfonic

67

acid and amine components. Those two observations, the decrease in the concentration of ionic bonds and the absence of sulfonic acid and amine end-groups, confirm the decomposition of the end-groups at 195 °C. 1H NMR analysis (Figure

3.14) also confirmed that the concentration of ammonium ion decreased and no primary amine was formed when the sample was heated for 30 min. at 195 °C.

1039

c

1039 1080 b

1050 Absorbance a

1100 1050 1000 -1 Wavenumber (cm )

Figure 3.13 FTIR spectra of (a) HO3S-PS-SO3H, (b) SMBCP, (c) SMBCP that exhibited macrophase separation after heating at 190 oC. For (c), shoulder at 1039 is still observed but much smaller compared to original shoulders and shoulder at 1080 due to ammonium was disappeared. Also, no peak at 1050 is observed for (c), indicating no sulfonic acid was reformed during the heating, i.e. no reverse proton transfer occurred.

68

multi-195.esp

3.81 2.69

7 6 5 4 3 2 1 Chemical Shift (ppm)

Figure 3.14 1H NMR spectrum of SMBCP that exhibited macrophase separation after heating at 190 oC. The resonance at 3.81 ppm and 2.69 ppm were still observed but much smaller compared to the original peaks.

The blend was also characterized by oscillatory rheology as shown in Figure

3.15. Bates and coworkers have shown that the viscoelastic properties of block copolymers at low frequencies are sensitive to the microstructure.26,33 At low frequency (ω), the dynamic and loss moduli, G’(ω) and G”(ω), in Figure 3.15 scaled as ω0.5, which is characteristic for a block copolymer with a lamellar texture.

69

106

105

0.5 104

3

10 o

G' (Pa) 110 C 2 10 130oC 150oC 101 170oC 190oC 100 0.01 0.1 1 10 100 (rad/s) 6 10 o 110 C 190oC 130oC o o 210 C 105 150 C 230oC 170oC 104

103

G'' (Pa) G'' 2 10 0.5 1

101

100 0.01 0.1 1 10 100  (rad/s)

Figure 3.15 Frequency dependence of (a) the storage modulus, G' and (b) the loss modulus, G'' as a function of temperature for the SMBCP. Strain amplitude = 5%.

Above 190 °C, i.e. for the data at 210 °C and 230 °C, the slope of G” vs. ω changed to ~1, i.e., G”/ω = η’ = constant, and the G’ values of the blend above 190 °C were too low to measure, which indicates that the liquid lost its elasticity at the same

70

time the viscosity became frequency independent. Those results are consistent with an order-disorder transition (ODT) between 190 °C and 210 °C.

The temperature dependence of the linear viscoelastic tensile behavior of the blend between -100 °C and 200 °C is shown in Figure 3.16. The data between 70 °C and 200 °C were measured by simple shear and converted to tensile moduli by assuming E = 3G. The poor overlap of the two sets of data between 80-100 °C is a consequence of the glass transition of the PS-rich phase (see discussion below), during which the modulus was very sensitive to temperature. The frequencies used for tension and shear were slightly different (see caption of Figure 3.16), 1.0 Hz (6.3 rad/s) and 1.6 Hz (10 rad/s), but the differences in temperature associated with those two frequencies is very small – estimated by the WLF equation to be ~1 °C.

104

103

102

101

100 (MPa) -1 Measured

E" 10

, Tensile Data -2

E' 10 10-3 Calculated from Shear Data 10-4

10-5 -100 0 100 200 o T ( C)

Figure 3.16 Dynamic storage and loss tensile moduli of SMBCP as a function of temperature. The curves marked as measured tensile data used a frequency of 6.3 rad/s and the curves marked as calculated from shear data were measured at a frequency of 10 rad/s and converted to tensile values by assuming E = 3G.

71

The high E’ (~4 x 103 MPa) at -100 °C is consistent with the fact that both PS

and PIB are glasses below about -80 °C. The decrease in E’ to ~5 x 102 MPa and the

peak in E” at about -50 °C are due to the glass transition of the PIB-rich phase. The

plateau in E’ that persists to about 60 °C is due to the physical network formed by the

glassy PS nanodomains. The modulus for this plateau is relatively high for an

elastomer, which is a consequence of the high HO3S-PS-SO3H concentration, 50

mol%, which produced a very high effective crosslink density.

Starting at ~60 °C, E’ decreased by several orders of magnitude to ~0.1 MPa

at ~100 °C and a second peak in E” occurred at 65 °C. Those results were due to the

glass transition of the PS-rich phase. Above the glass transition of the PS-rich phase E’

decreased relatively slowly to about 103 Pa, and another abrupt decrease in E’

occurred at above 190 °C due to the onset of viscous flow. Even though the SMBCP

was a liquid at 185 °C, it was still quite elastic (tanδ = 2) due to the retention of the

lamellar microstructure above the Tg’s of both phases. The abrupt decrease in E’ above 190 °C is consistent with the occurrence of an ODT.

3.3.3 Mechanical Properties

An example tensile stress-strain curve of the blend at room temperature is

shown in Figure 3.17. The blend exhibited the characteristics of a soft plastic with a

modulus of 90 MPa, a yield point of 4% strain, failure at 7% strain and a strain energy

to failure (i.e., the area under the stress – strain curve) of 15 MJ/m3. The high modulus

was a consequence of the relatively high polystyrene fraction. If a lower HO3S-PS-

SO3H concentration were used, by either using shorter HO3S-PS-SO3H oligomers or

longer H2N-PIB-NH2 oligomers, the blend would be expected to be softer. The

relatively high modulus of the material and the yielding phenomenon demonstrate the

72

chain-like supramolecular structure of the material. The low elongation to break may, however, be a consequence of the supramolecular bonds, which though strong are still weaker than covalent bonds. In effect, the ionic bond acts to some extent as an imperfection of the chain, since it cannot support as much stress without breaking as a covalent bond. The shorter the block lengths are, the higher is the concentration of supramolecular bonds, or “defects”, in the chain.

Much of the elongation and strain energy of high molecular weight polymers is due to the coiled nature of the chains, which can accommodate a great deal of strain by uncoiling before the bonds are markedly stressed. For the supramolecular block copolymer, uncoiling of the oligomers may stress the junction points, i.e., the supramolecular bonds, producing bond fracture, which eventually leads to material failure at a relatively low strain compared with a covalently bonded block copolymer.

An example of this is the lower tensile fracture strain of an oligomeric ionomer melt undergoing unaxial extensional flow compared with that for a high molecular weight, entangled polymer of the same backbone165.

73

3.5 3.0 2.5 2.0 1.5

Stress (MPa) 1.0 0.5 0.0 0 2 4 6 8

Strain (%)

Figure 3.17 Engineering tensile stress-strain curve of the SMBCP at room temperature.

3.3.4 Nonlinear Rheological Behavior.

Three consecutive shear strain-time cycles for the SMBCP are shown in

Figure 3.18. The temperature was held constant at 150°C, which was above the Tg’s of both blocks, but below the ODT and the decomposition temperature of the ionic bonds. At 150°C, the SMBCP was a viscoelastic liquid, i.e., G” > G’. In the strain- sweep part of the cycle, the frequency was fixed at ω = 10 rad/s and the strain amplitude was varied from γ = 0.01% to 100%. The strain-sweep was followed immediately after measuring the viscoelastic behavior at γ = 100% by a time-sweep of

300 s during which the frequency was fixed at ω = 10 rad/s and γ was lowered to within the linear region and fixed at 1%.

For each cycle, the magnitude of G' dropped nearly an order of magnitude when the strain amplitude was greater than ~2%. The decrease of G’, the elastic

74

component of the viscoelastic response, in the non-linear region may have been due to disruption of the supramolecular ionic bonds due to the high stresses developed166,167

or as a consequence of shear alignment induced by the large strain amplitude

oscillatory flow as has been observed with covalent block copolymers. For the first

recovery (time) cycle, when the stress was decreased by lowering the strain amplitude

to 1%, G’ rapidly recovered to about 90% of its original value within 100 s. The

failure of G’ to completely recover may have been due to some residual shear

alignment or a different equilibrium for the supramolecular structure, such as a

different number of blocks per chain. However, for the subsequent strain-time cycles,

following the first cycle, G’ recovered nearly 100% with 300 s following a non-linear

strain sweep.

G” represents the viscous response of the material (note that the dynamic

viscosity, η’ = G”/ω). In the experiment shown in Figure 3.18, G” was much less

sensitive to strain than was G’, decreasing only by a factor of ~1.4 for 100% strain.

When the strain was decreased to within the linear region, G” recovered quickly, like

G’ to about 90% of the initial value for the first cycle and nearly 100% for subsequent

cycles.

75

Time (102s) Time (102s) Time (102s) 0 1 2 3 0 1 2 3 0 1 2 3

104 G', G" (Pa) G" G',

3 10            

10-1 100 101 102 10-1 100 101 102 10-1 100 101 102         

Figure 3.18 G’ (filled circle) and G” (open circles) for three consecutive strain sweep- time sweep cycles of the SMBCP at 150°C. For the strain sweep, γ varied from 0.01% to 100%. For the time sweep, γ = 1% and ω = 10 rad/s.

3.4 Conclusions

Supramolecular multiblock copolymers were prepared by solvent casting blends of telechelic sulfonate terminated polystyrene and telechelic primary amine terminated polyisobutylene. Multiblock copolymer formation was supported by small angle X-ray scattering results, where the domain spacing of the nanostructure was consistent with theoretical calculations for infinite multiblocks. This demonstrates that sulfonated telechelics prepared by a simpler RAFT polymerization route behave analogously to polymers prepared previously by anionic polymerization. This approach also offers a route to prepare block copolymers using homopolymers prepared by two separate polymerization routes. The synthesis of gram scale quantities of both polymers allowed the preparation of free-standing, thick films and bulk mechanical characterization. The strain to failure of these materials was low and

76

attributed to the mechanically driven dissociation of the ionic bonds. While the low strain to failure may be an undesirable material property, the mechanical response of the ionic bonds may be useful for programming weaker bonds for mechano-plastic or self-healing polymers.

3.5 Acknowledgements

This work was supported by grants from the Chemical, Bioengineering,

Environmental and Transport Systems Program (CBET-1066517), the Polymer

Program (DMR-1309853) of the National Science Foundation, and the Office of

Naval Research and Northrop Grummund (Award No. N00014-07-1057).

77

CHAPTER IV

SYNTHESIS OF QUATERNARY AMMONIUM-CONTAINING,

TRITHIOCARBONATE RAFT AGENTS AND HEMI-TELECHELIC

CATIONOMERS*

4.1 Introduction

Hemi-telechelic cationomers, defined as polymers containing a single cationic end-group, have attracted interest in polymer science and engineering as model ionomers,147, 168 supramolecular building blocks,51,58,73,77,79,80,81,90 nanoclay

dispersants,111,113,169,170,171,172,173 nanostructured ionic liquids,174 aqueous polymeriza-

tions,112,175 stimuli-responsive polymers, 176 and antimicrobial polymers.177-179

These types of polymers were first prepared by living anionic and cationic

polymerization.164,180, 181,182 While this provides precise control over the molecular

weight distribution and end-group functionality, ionic polymerization can be hindered

by the rigorous purification and reaction conditions required and the sensitivity to

many functional groups.183 More recently, a number of initiators and chain transfer

agents containing quaternary ammonium functional groups have been generated for

controlled free radical polymerization (CFRP), including atom transfer radical

polymerization,172 nitroxide mediated polymerization,111,112 and reversible addition

fragmentation chain transfer polymerization (RAFT).113, 173, 175, 176, 184 CFRP offers

control over the molecular weight distribution and functionality for a wide range of

78

monomers with less stringent experimental conditions and are therefore attractive for

the synthesis of functional polymers, such as hemi-telechelic, cationomers.101

This chapter presents a facile, high yield synthetic method for making

quaternary ammonium-containing trithiocarbonate RAFT agents as shown

schematically in Figure 4.1. By placing the quaternary ammonium group on the

initiating fragment (R-group) of the RAFT agent, it is stable under conditions that

cleave the trithiocarbonate group. Furthermore, the trithiocarbonate group is available

for further modification to produce hetero-telechelic polymers.136 As shown in Figure

4.2, similar RAFT agents have previously been prepared by Samakande et al.113 and

Moughton and O’Reilly176 by reacting 1,4-dibromoxylylene or 4-(chloromethyl)- benzyl alcohol, respectively, with an alkyl trithiocarbonate anion and subsequent modification to produce the tri-n-alkyl benzyl ammonium group. While the reaction of benzyl halides with alkyltrithiocarbonate anions is a robust method for the synthesis of trithiocarbonate RAFT agents,185, 186 in the examples above the yields

were reduced due to the inherent by-products of the syntheses and the purification steps required.

Br Br + NR3 Br NR3 Br R=Me, Et, Bu 1 S C12H25 NEt3 1+ C12H25SH + CS2 S S NR3Br R=Me, Et, Bu

Figure 4.1 Synthesis of 4-(bromomethyl)-N,N,N-trialkylbenzyl ammonium compounds (alkyl=methyl, ethyl, butyl) and subsequent RAFT agents.

79

(a) (b)

65

Figure 4.2 Reaction routes reported by (a) O’Reilly176 and (b) Samakande113.

As will be shown in this chapter, the key improvement in the synthesis of the

RAFT agents in Figure 4.1 was the use of a 4-(bromomethyl)-N,N,N-trialkyl benzyl ammonium bromide compounds, which were reacted with the alkyl trithiocarbonate anion to directly produce the trithiocarbonate RAFT agents. These compounds have been prepared by the mono-quaternization of 1,4-dibromoxylylene by driving the precipitation of the mono-substituted compound out of the reaction solution187-189.

Two previous examples in the polymer literature of the synthesis of these types of

compounds are 4-(bromomethyl)-N,N,N-trimethyl benzyl ammonium bromide used to

prepare at dithiocarbamate inifiter170 and 4-(bromomethyl)-N-dodecyl-N,N-

179 dimethylbenzyl ammonium bromide, which was used as an initiator in the cationic

polymerization of poly(methyloxazoline)s. However, these compounds have not

previously been investigated for the synthesis of trithiocarbonate RAFT agents. In this

article, in addition to presenting the reaction conditions for the synthesis of 4- 80

(bromomethyl)-N,N,N-trialkylbenzyl ammonium compounds and subsequent RAFT

agents, the purification of the polymers produced with these RAFT agents to produce

high purity hemi-telechelic cationomers is discussed.

4.2 Experimental section

4.2.1 Materials

Azobisisobutyronitrile (98%, Sigma-Aldrich) was recrystallized from

methanol and dried under vacuum before use. Styrene (99%, stabilized, Acros

Organics), methyl acrylate (99%, stabilized, Acros Organics), n-butyl acrylate (Alfa

Aesar, 98%) and 2-dimethylaminoethyl acrylate (TCI America, 97%) were purified

by passing through a column of basic alumina. Silica gel (Dynamic Adsorbents 60 Å,

32–63 μM, standard grade) was used for column chromatography. Silica gel thin layer

chromatography plates (250 µm, with fluorescent indicator activated at 2540 ) were

supplied from J. T. Baker. All other chemicals used in this article were obtainedÅ

commercially in high purity and used as received.

4.2.2 Instrumentation

Both 1H NMR (300 MHz) and 13C NMR (75 MHz) spectra were collected on a

Varian Gemini spectrometer. The molecular weight and molecular weight distribution

of the polymer products was characterized by size exclusion chromatography (SEC) with a Waters Breeze system equipped with a column set at 35 oC and a refractive

index detector (Waters 2414). The column set consists of a Styragel® HR 4 THF

column (4.6 × 300 mm) with an effective molecular weight range 5 k to 600 k Da, a

Styragel® HR 3 THF column (4.6 × 300 mm) with an effective molecular weight

range 0.5 k to 30 k Da, and a Styragel® HR 4E THF column (4.6 × 300 mm) with an

81

effective molecular weight range 0.05 k to 100 k Da. The SEC was calibrated using

PS standards of narrow molar mass dispersity (Ð) with the molecular weight of 1300

Da to 400 k Da (Alfa Aesar). Thermogravimetric analysis (TGA) was performed on a

TA Instruments TGA Q50 from room temperature to 700 °C with a heat rate of

20 °C/min. All mass spectrometry experiments were performed on an HCT Ultra II

quadrupole ion trap mass spectrometer (Bruker Daltonics, Billerica, MA) equipped

with an electrospray (ESI) source.

4.2.3 Synthesis of 4-(bromomethyl)benzyltrimethylammonium bromide (Br-Ph-NMe3)

In a modified procedure of Rammo and Schneider187, by adding 50%

trimethylamine aqueous solution to solid potassium hydroxide in a flask,

trimethylamine gas was immediately generated and directly bubbled into 30 mL

toluene in the other flask until 0.236 g (4.0 mmol) was absorbed. The resulting

trimethylamine/toluene solution was slowly dropped into a stirred solution of α,α'-

dibromo-p-xylene (1.321 g, 5.0 mmol) in 30 mL toluene at room temperature. Upon

completion of the addition, the reaction was stirred overnight, c.a. 12 h. The white

precipitate formed was filtered, washed with 30 mL toluene and transferred to a

Soxhlet extractor for 48 h with acetone to extract the desired product from the crude

reaction product. The product, 4-(bromomethyl)benzyltrimethylammonium bromide,

(Br-Ph-NMe3), was recovered as a white powder by drying the acetone solution under

1 reduced pressure and vacuum (1.098 g, 85% yield). H NMR (300 MHz, D2O) δ

+ (ppm): 7.61 (d, J = 8.2 Hz, 2H, (CHar)2CCH2N (CH3)3), 7.53 (d, J = 7.9 Hz, 2H,

+ BrCH2C(CHar)2(CHar)2C), 4.64 (s, 2H, (CHar)2CCH2N (CH3)3), 4.49 (s, 2H,

+ 13 BrCH2C(CHar)2), 3.09 (s, 9H, (CHar)2CCH2N -(CH3)3). C NMR (75 MHz, D2O) δ

+ (ppm): 141.02 ((CHar)2CCH2N (CH3)3), 133.31 (BrCH2C(CHar)2(CHar)2C), 129.74

82

+ ((CHar)2CCH2N (CH3)3), 127.76 (BrCH2C(CHar)2-(CHar)2C), 69.06 ((CHar)2C-

+ + CH2N (CH3)3), 52.49 ((CHar)2CCH2N (CH3)3), 32.93 (BrCH2C(CHar)2(CHar)2C). MS

- (ESI-MS) calcd for C11H17BrN (m/z), 242.1; found, 241.9 [M-Br ], 182.8 [M-

- NMe3Br ].

4.2.4 Synthesis of 4-(bromomethyl)benzyltriethylammonium bromide (Br-Ph-NEt3)

Triethylamine (5.720 g, 56.6 mmol) in 60 mL ethyl acetate was slowly added

dropwise to a stirred solution of α,α'-dibromo-p-xylene (13.202 g, 50.0 mmol) in 120 mL ethyl acetate at 50 °C. The reaction mixture was stirred for 48 hours. The product,

4-(bromomethyl)benzyltriethylammonium bromide (Br-Ph-NEt3), gradually

precipitated as a fine white powder. Then Br-Ph-NEt3 was recovered by filtration,

washed with 60 mL ethyl acetate and dried in a vacuum oven (18.085 g, 99% yield).

1 No further purification was required. H NMR (300 MHz, CDCl3) δ (ppm): 7.63 (d, J

+ = 8.1 Hz, 2H, (CHar)2CCH2N (CH2CH3)3), 7.46 (d, J = 8.3 Hz, 2H,

+ BrCH2C(CHar)2(CHar)2C), 4.95 (s, 2H, (CHar)2CCH2N (CH2CH3)3), 4.49 (s, 2H,

+ BrCH2C(CHar)2), 3.47 (q, J = 7.3 Hz, 6H, (CHar)2CCH2N (CH2CH3)3), 1.48 (t, J = 7.1

+ 13 Hz, 9H, (CHar)2CCH2N ( CH2CH3)3). C NMR (75 MHz, CDCl3) δ (ppm): 140.52

+ ((CHar)2CCH2N (CH2CH3)3), 133.13 (BrCH2C(CHar)2(CHar)2C), 129.94

+ ((CHar)2CCH2N (CH2CH3)3), 127.35 (BrCH2C-(CHar)2(CHar)2C), 61.07

+ + ((CHar)2CCH2N (CH2CH3)3), 53.12 ((CHar)2CCH2N (CH2CH3)3), 31.93

+ (BrCH2C(CHar)2(CHar)2C), 8.54 ((CHar)2CCH2N (CH2CH3)3). MS (ESI-MS) calcd for

- C14H23BrN (m/z), 284.1; found, 284.1 [M-Br ].

83

4.2.5 Synthesis of 4-(bromomethyl)benzyltributylammonium bromide (Br-Ph-NBu3)

Tributylamine (0.925 g, 5.0 mmol) in 30 mL acetonitrile was slowly added

dropwise to a stirred solution of α,α'-dibromo-p-xylene (6.626 g, 25.1 mmol) in 50 mL acetonitrile at 50 °C. Upon the completion of the addition, the reaction mixture was stirred for 48 hours and then was allowed to cool down. Unreacted α,α'-dibromo- p-xylene partially precipitated out and was removed by filtration. The filtrate was collected and the acetonitrile was removed under vacuum. Toluene (250 mL) was added to the dried reaction mixture. The product, 4-

(bromomethyl)benzyltributylammonium bromide (Br-Ph-NBu3), was hardly soluble in toluene and was recovered as a white powder. The white product, which contains trace amount of di-substituted compound (~1 mol% by 1H NMR), was collected by

filtration and dried in a vacuum oven (1.851 g, 82% yield). 1H NMR (300 MHz,

+ CDCl3) δ (ppm): 7.59 (d, J= 8.2 Hz, 2H, (CHar)2CCH2N (CH2CH2 CH2CH3)3), 7.47

+ (d, J= 7.9 Hz, 2H, BrCH2C(CHar)2(CHar)2C), 4.99 (s, 2H, (CHar)2CCH2N (CH2CH2

CH2CH3)3), 4.48 (s, 2H, BrCH2C(CHar)2), 3.35 (m, 6H,

+ (CHar)2CCH2N (CH2CH2CH2CH3)3), 1.78 (m, 6H, (CHar)2CCH2-

+ + N (CH2CH2CH2CH3)3), 1.43 (m, 6H, (CHar)2CCH2N (CH2CH2CH2CH3)3), 1.01 (t, J=

+ 13 7.3 Hz , 9H, (CHar)2CCH2N (CH2CH2CH2CH3)3). C NMR (75 MHz, CDCl3) δ

+ (ppm): 140.54 ((CHar)2CCH2N (CH2CH2CH2CH3)3), 133.34 (BrCH2C-

+ (CHar)2(CHar)2C), 129.74 ((CHar)2CCH2N (CH2CH2CH2CH3)3), 127.49

+ (BrCH2C(CHar)2(CHar)2C), 62.69 ((CHar)2CCH2N (CH2CH2CH2CH3)3), 59.09

+ ((CHar)2CCH2N (CH2CH2CH2CH3)3), 32.09 (BrCH2C(CHar)2(CHar)2C), 24.44

+ + ((CHar)2CCH2N (CH2CH2CH2CH3)3), 19.94 ((CHar)2CCH2N (CH2CH2CH2CH3)3),

+ 13.64 ((CHar)2CCH2N ( CH2CH2CH2CH3)3). MS (ESI-MS) calcd for C20H35BrN

(m/z), 368.2; found, 368.1 [M-Br-]. 84

4.2.6 Synthesis of S-1-dodecyl-S’-(methylbenzyltriethylammonium bromide)

trithiocarbonate RAFT Agents (RAFT-NEt3)

Triethylamine (1.127 g, 11.2 mmol) was added dropwise to a stirred solution

of 1-dodecanethiol (2.030 g, 10.1 mmol) and CS2 (1.202 g, 15.8 mmol) in CH2Cl2 (30

mL) at room temperature. After stirring for 1 hour, Br-Ph-NEt3 (3.650 g, 10.0 mmol)

was directly added into the reaction mixture. The reaction mixture was stirred for

another 24 hours and then dried in vacuum. Toluene was added to the dried reaction

mixture while stirring. The by-product, triethylamine hydrochloride, precipitated and

was removed by filtration. The product, S-1-dodecyl-S’-

(methylbenzyltriethylammonium bromide) trithiocarbonate (RAFT-NEt3), was

recovered as a yellow solid by evaporating the toluene under vacuum (5.503 g, 98%

1 yield). H NMR (300 MHz, CDCl3) δ (ppm): 7.57 (d, J= 7.9 Hz, 2H,

+ (CHar)2CCH2N (CH2CH3)3), 7.41 (d, J= 7.0 Hz, 2H, S=CSCH2C(CHar)2(CHar)2C),

+ 4.88 (s, 2H, (CHar)2CCH2N (CH2CH3)3), 4.63 (s, 2H, S=CSCH2C(CHar)2), 3.42 (m,

+ 8H, (CHar)2CCH2N (CH2CH3)3 and CH3C9H18CH2CH2CS=S), 1.70 (m, 2H,

+ CH3C9H18CH2CH2CS=S), 1.47 (t, J = 7.1 Hz, 9H, (CHar)2CCH2N ( CH2CH3)3), 1.21-

1.38 (br m, 18H, CH3C9H18CH2CH2CS=S), 0.89 (t, J = 6.4 Hz, 3H,

13 CH3C9H18CH2CH2CS=S). C NMR (75 MHz, CDCl3) δ (ppm): 222.96 (SC=S),

+ 138.72 ((CHar)2CCH2N (CH2CH3)3), 132.93 (S=CSCH2C(CHar)2(CHar)2C), 130.14

+ ((CHar)2CCH2N (CH2CH3)3), 126.55 (S=CSCH2C(CHar)2(CHar)2C), 61.07

+ + ((CHar)2CCH2N (CH2CH3)3), 53.09 ((CHar)2CCH2N (CH2CH3)3), 40.12 (CH3C9H18-

CH2CH2CS=S), 37.32 (CH3C9H18CH2CH2CS=S), 31.94 (S=CSCH2C(CHar)2(CHar)2C),

22.66-29.62 (CH3C9H18CH2CH2CS=S), 14.17 (CH3C9H18CH2CH2CS=S), 8.78

+ ((CHar)2-CCH2N (CH2CH3)3). MS (ESI-MS) calcd for C27H48NS3 (m/z), 482.3; found,

482.3 [M-Br-].

85

RAFT agents S-1-dodecyl-S’-(methylbenzyltrimethylammonium bromide)

trithiocarbonate (RAFT-NMe3) and S-1-dodecyl-S’-(methylbenzyltributylammonium

bromide) trithiocarbonate (RAFT-NBu3) were synthesized by analogous procedure.

For RAFT-NMe3, acetone was used as solvent instead of CH2Cl2.

1 RAFT-NMe3. Yellow solid. Yield=91%. H NMR (300 MHz, CDCl3) δ

+ (ppm): 7.63 (d, J= 8.2 Hz, 2H, (CHar)2CCH2N (CH3)3), 7.43 (d, J= 7.6 Hz, 2H,

+ S=CSCH2C(CHar)2(CHar)2C), 5.09 (s, 2H, (CHar)2CCH2N (CH3)3), 4.65 (s, 2H,

+ S=CSCH2C(CHar)2), 3.41 (m, 11H, (CHar)2CCH2N (CH3)3) and

CH3C9H18CH2CH2CS=S), 1.69 (m, 2H, CH3C9H18CH2CH2CS=S), 1.22-1.54 (br m,

13 18H, CH3C9H18CH2CH2CS=S), 0.89 (t, J= 6.6 Hz, 3H, CH3C9H18CH2CH2CS=S). C

+ NMR (75 MHz, CDCl3) δ (ppm): 222.85 (SC=S), 138.63 ((CHar)2CCH2N (CH3)3),

+ 133.27 (S=CSCH2C(CHar)2(CHar)2C), 129.96 ((CHar)2CCH2N (CH3)3), 126.66

+ (S=CSCH2C(CHar)2(CHar)2C), 68.46 ((CHar)2C-CH2N (CH2CH3)3), 52.77

+ ((CHar)2CCH2N (CH3)3), 40.32 (CH3C9H18CH2CH2CS=S), 37.23

(CH3C9H18CH2CH2CS=S), 31.93 (S=CSCH2C(CHar)2(CHar)2C), 22.66-29.73

(CH3C9H18CH2CH2CS=S), 14.28 (CH3C9H18CH2CH2CS=S). MS (ESI-MS) calcd for

- C24H42NS3 (m/z), 440.3; found, 440.2 [M-Br ].

1 RAFT-NBu3. Yellow liquid. Yield=79.1%. H NMR (300 MHz, CDCl3) δ

+ (ppm): 7.52 (d, J= 6.7 Hz, 2H, (CHar)2CCH2N (CH2CH2CH2CH3)3), 7.42 (d, J= 7.9

Hz, 2H, S=CSCH2C(CHar)2(CHar)2C), 4.94 (s, 2H,

+ (CHar)2CCH2N (CH2CH2CH2CH3)3), 4.64 (s, 2H, S=CSCH2C(CHar)2), 3.33 (m, 8H,

+ (CHar)2CCH2N (CH2CH2CH2CH3)3 and CH3C9H18CH2CH2CS=S), 1.78 (m, 8H,

+ (CHar)2CCH2N (CH2CH2CH2CH3)3 and CH3C9H18CH2CH2CS=S), 1.42 (m, 6H,

+ (CHar)2CCH2N (CH2CH2CH2CH3)3), 1.21-1.36 (br m, 18H, CH3C9H18CH2CH2CS=S),

+ 1.00 (t, J= 7.3 Hz, 3H, (CHar)2CCH2N (CH2CH2CH2CH3)3), 0.88 (t, J= 7.1 Hz, 3H,

86

13 CH3C9H18CH2CH2CS=S). C NMR (75 MHz, CDCl3) δ (ppm): 222.91 (SC=S),

+ 138.61 ((CHar)2CCH2N (CH2CH2CH2CH3)3), 132.92 (S=CSCH2C(CHar)2(CHar)2C),

+ 130.08 ((CHar)2CCH2N (CH2CH2CH2CH3)3), 126.80 (S=CSCH2C(CHar)2(CHar)2C),

+ 62.93 ((CHar)2CCH2N (CH2CH2CH2CH3)3), 58.77 ((CHar)2CCH2-

+ N (CH2CH2CH2CH3)3), 40.18 (CH3C9H18CH2CH2CS=S), 37.34 (CH3C9H18-

CH2CH2CS=S), 31.87 (S=CSCH2-C(CHar)2(CHar)2C), 22.66-29.62

+ (CH3C9H18CH2CH2CS=S), 24.60 ((CHar)2CCH2-N (CH2CH2CH2CH3)3), 19.75

+ ((CHar)2CCH2N (CH2CH2CH2CH3)3), 13.66 (CH3-C9H18CH2CH2CS=S). MS (ESI-

- MS) calcd for C33H60NS3 (m/z), 566.4; found, 566.2 [M-Br ].

4.2.7 Synthesis of benzyl dodecyl trithiocarbonate (BDTC)

A nonionic RAFT agent was prepared by a similar to that route described above. Triethylamine (10.110 g, 100 mmol) was added dropwise to a stirred solution of 1-dodecanethiol (10.115 g, 50 mmol) and CS2 (7.724 g, 101.6 mmol) in CH2Cl2

(30 mL) at room temperature. After stirring for 3 hour, benzyl chloride (6.340 g, 50.1

mmol) was directly added into the reaction mixture. The reaction mixture was stirred

for another 24 hours. Upon the completion of the reaction, the reaction mixture was

poured into a separation funnel and washed with deionized water for 3 times. The

yellow CHCl3 layer was collected, dried over MgSO4, filtered, and rotary evaporated

to provide yellow oil. The crude product was recrystallized in CHCl3 to give a yellow

1 solid. Yield=13.69 g (74.4%). H NMR (300 MHz, CDCl3) δ (ppm): 7.32 (m, 5H,

S=CSCH2C6H5), 4.62 (s, 2H, S=CSCH2C6H5), 3.38 (t, J= 7.6 Hz, 2H,

CH3C9H18CH2CH2CS=S), 1.71 (m, J= 7.4 Hz, 2H, CH3C9H18CH2CH2CS=S), 1.21-

1.47 (br m, 18H, CH3C9H18CH2CH2CS=S), 0.89 (t, J= 6.6 Hz, 3H,

CH3C9H18CH2CH2CS=S).

87

4.2.8 RAFT Bulk Polymerization of Styrene at 120 °C

Bulk styrene polymerizations using the abovementioned quaternary

ammonium functionalized RAFT agents were carried out at 120 °C. For all of these

polymerizations, the theoretical molecular weight targeted was 25000 g/mol. A

typical RAFT polymerization procedure and subsequent kinetic studies were

performed as follows. A master batch of styrene (35 mL, 31.819 g, 305.5 mmol) and

RAFT-NEt3 (0.733 g, 1.3 mmol) was prepared. Aliquots of 5 mL were added to

separate flasks each equipped with a magnetic stirrer and sealed with a rubber septum.

The sealed flasks were subsequently purged with dry nitrogen for 15 min and placed

in preheated oil bath at 120 °C. After a given time, the polymerization flask was

quenched in ice water. To determine the conversion of monomer, one drop of the

1 reaction mixture was taken out and diluted with CDCl3 and characterized by H NMR spectroscopy. The polymers were recovered by precipitation of the reaction mixtures into methanol twice, filtering, washed three times by methanol and dried in a vacuum oven at room temperature.

4.2.9 Bulk RAFT Polymerization of Styrene at 65 °C

Styrene (4.545 g, 43.6 mmol) , RAFT-NEt3 (0.470 g, 0.834 mmol) and AIBN

(0.014 g, 0.083 mmol) were placed in a 10 mL flask equipped with a magnetic stir bar

and sealed with a rubber septum. The sealed flask was purged with dry nitrogen for 15

min and placed in a preheated oil bath at 65 °C. The polymerization was stirred for 24

h and then quenched by placing the flask in ice water. The polymer was isolated by

precipitating the reaction mixture into methanol twice, filtered, washed three times

with methanol, and dried in a vacuum oven at room temperature. (Mn SEC: 3200 Da,

Ð=1.15.) A kinetic study of bulk polymerization of styrene at 65 °C using RAFT-

88

NEt3 was performed following the similar procedure as described for the kinetics

study of bulk styrene polymerization at 120 °C.

4.2.10 RAFT Polymerization of Acrylate Monomers at 65 °C

Three acrylate monomers were polymerized using RAFT-NEt3. In a typical

polymerization, methyl acrylate (MA) (2.150 g, 25.0 mol), RAFT-NEt3 (0.281 g, 0.5

mmol) and AIBN (0.008 g, 0.05 mmol) were dissolved in 2.26 mL chlorobenzene and

placed in a 15 mL flask equipped with a magnetic stir bar and sealed with a rubber

septum. The sealed flask was purged with dry nitrogen for 15 min and placed in

preheated oil bath at 65 °C. The polymer was isolated by precipitating the reaction

mixture to methanol/water (1/1 v/v). The precipitated polymer was dissolved in

methanol and dried over Na2SO4, the methanol was removed by rotary evaporation

and the polymer was dried in a vacuum oven at room temperature. A broad SEC

elution peak was observed (Mn =3200 Da, Ð = 1.45), which is attributed to a strong interaction between the polymer and the SEC column. Mn(NMR)= 3100 Da.

A similar procedure was applied to the RAFT polymerization of n-butyl

acrylate (BA) and 2-dimethylaminoethyl acrylate (DMAEA). For both

polymerizations reactant ratios of [M]/[RAFT]/[AIBN] =50:1:0.1 were used.

Chlorobenzene was used as solvent to make a 50% v/v monomer solution. The

polymerization was run at 65 °C for 3 hours for both monomers.

Poly(n-butyl acrylate). PBA was isolated by precipitating the reaction

mixture to in methanol/water (1/1 v/v). The precipitated polymer was dissolved in

acetone and dried over MgSO4. Acetone was removed by rotary evaporation and the

polymer was dried in a vacuum oven at 50 °C. Mn SEC=7000 Da, Ð = 1.08, Mn

NMR= 7600 Da).

89

Poly(2-dimethylaminoethyl acrylate). PDMAEA was isolated by

precipitating the reaction mixture to cold hexane. The precipitated polymer was dried

in a vacuum oven at 50 °C. No polymer peak was observed in SEC, which is due to

the strong interaction between the polymer chains and the SEC column. Mn(NMR)=

3300 Da.

4.3 Results and discussion

4.3.1 Synthesis of 4-(bromomethyl)benzyl-N,N,N-trialkylammonium bromide

compounds

A series of 4-(bromomethyl)benzyl-N,N,N-trialkylammonium bromide

compounds (Br-Ph-NR3) were synthesized as shown in Figure 4.1 by selective

quaternization of α,α'-dibromo-p-xylene. The results are summarized in Table 4.1.

Mono-quaternization of 1,ω-dihaloalkanes or α,α'-dihalo-p-xylene with tertiary amines have been reported previously and a common way to drive mono-substitution is by precipitation of the mono-substituted compounds in an appropriate reaction medium to prevent further reaction.187- 190 Although some (ω-

bromoalkyl)trialkylammonium bromide compounds are straightforward to prepare190

and are even commercially available, we focused on the mono-quaternization of α,α'-

dibromo-p-xylene (DBX) to place the quaternary ammonium group in the initiator fragment (R-group) in a benzyl trithiocarbonate RAFT agent as shown in Figure 4.1.

90

Table 4.1 Reaction conditions of 4-(bromomethyl)benzyl-N,N,N-trialkylammonium bromide compounds Sample Solvent Temp Time Stoichiometry Ratio of mono- Yield of /°C /h Amine: DBX to di-substituted Mono- DBXa substituted DBX

Br-Ph-NMe3 Toluene RT 24 1.1: 1 84: 16 66% Toluene RT 21 0.8: 1 90: 10 85% p-Xylene RT 1 2: 3 NA 59%c, 187

Br-Ph-NEt3 Toluene RT 12 1: 1 100: 0 30% Toluene 60 36 1: 1 100: 0 77% Toluene RT 36 1.2: 1 100: 0 40% EtOAc RT 36 1: 1 100: 0 67% EtOAc 50 48 1.1: 1 100: 0 99% EtOAc 50 24 1.1: 1 100: 0 97% p-Xylene/ RT 0.5 1: 1 NA 40% c, 188 ether (3/1)

b Br-Ph-NBu3 EtOAc RT 64 1:1 67: 33 < 30% Toluene RT 64 1:1 83: 17 < 30%b MeCN 50 36 1:1 75: 25 >70%b MeCN 50 48 1:4 90: 10 NA MeCN 50 48 1:5 99: 1 81% b Toluene RT Several7: 10 NA NAc, 189

days a Calculated by 1H NMR. b All these entries were a mixture of mono- and di- substituted compounds. The portion of di-substituted DBX is indicated in the table. c Highest literature reported yield.

91

The mono-quaternization of DBX using trimethylamine has previously been

investigated.187, 188, 190 Compared to other tertiary amines, trimethylamine should

produce quaternary ammonium salts that are the least lipophilic and the easiest to

precipitate out from non-polar solvents during quaternization. However, it proved to

be very difficult to obtain clean mono-substituted compounds directly from the

quaternization of DBX,190 as the precipitates usually contained both mono- and di-

quaternized compounds, which was probably due to the high reactivity of

trimethylamine. Quaternization of DBX using trimethylamine was reported by

Rammo and Schneider,187 in which trimethylamine was directly bubbled into the

solution of DBX in p-xylene. However, their yield was relatively low (59%) and this

may because bubbling feeds the gas too quickly to the reaction mixture. We tried to

lower the feeding rate by first dissolving trimethylamine gas in toluene and then

slowly adding the trimethylamine/toluene solution dropwise to DBX solution. Several

trials were performed as summarized in Table 4.1. We were not able to prevent the

formation of di-substituted compound but we improved the selectivity and yield compared to previous reports by using 20% excess of DBX. The mono-substituted compounds were extracted by Soxhlet extraction with acetone from the crude reaction product.

When triethylamine rather than trimethylamine was used to mono-quaternize

DBX, reaction conditions were found where the mono-substituted compound precipitated as a pure products even if triethylamine was excess. A series of trials were performed and summarized in Table 4.1. The results showed that the mono- substituted products were obtained with very high yield and high purity, i.e. not contaminated by the di-substituted compounds. Therefore, no further purification was needed for removing di-substituted compounds or unreacted DBX. Our procedure and

92

results should be more favorable compared to that was used by Covitz,188 which

involved relatively laborious synthesis and purification procedures.

Mono-quaternization of DBX with tri-n-butylamine proved to be harder compared to trimethylamine and triethylamine. Due to the much greater lipophilicity of the mono-substituted compound, there is a competition between precipitation and further reaction producing di-substitution. Akhavan-Tafti and co-workers reported that when toluene was used, the mono-substituted compound precipitated as a pure compound.189 However, trying a similar procedure afforded a precipitate containing

more than 15% of di-substituted compound. In addition, due to the relatively low

polarity of toluene and the steric hindrance of the butyl group, the quaternization

proceeded very slowly and the overall yield was very low, as shown in Table 4.1.

When ethyl acetate was used instead of toluene, the yield was still low and even more

di-substituted compound was formed. Since the reaction performed in low polar

solvent can afford neither pure mono-substituted compound nor high yield, we tried to

improve the reaction yield by using a higher reaction temperature and a more polar

solvent, acetonitrile. The use of acetonitrile eliminated any precipitation but increased

the reaction rate significantly. The final optimization of this reaction was performed

by using a large excess of DBX, in which the di-substitution was suppressed to a

satisfactory level. Since it was difficult to separate the mono and di-substituted

compounds, the final product (Br-NBu3 containing 1 mol% di-substituted compounds)

was used in the RAFT agent synthesis without further purification.

All of the obtained mono-substituted compounds were characterized by 1H

NMR, 13C NMR and electrospray ionization mass spectroscopy (ESI-MS), which

confirmed the high purity of the obtained mono-substituted compounds. The 1H NMR

93

spectra and the ESI- MS spectra of the three mono-substituted compounds are shown in Figure 4.3 to Figure 4.8.

041013-BR-NME3-ACETONEEXTRACTED-APH-DC e a b c H2O

d e

d a b c 3.88 2.001.94 8.88

8 6 4 2 Chemical Shift (ppm)

1 Figure 4.3 H NMR spectrum in D2O of Br- Ph-NMe3.

94

a b f f c e 010312-BR-N-FROM-ETAC-APH-BC-DCBr Br N d a e d CHCl3

H2O c b

1.861.98 2.002.02 6.21 9.32

6 4 2 Chemical Shift (ppm)

1 Figure 4.4 H NMR spectrum in CDCl3 of Br- Ph-NEt3.

95

h 060613-BR-NBU3-APH-APH-DC g h a b f Br c e Br N d CHCl3

a H2O d e g cb f

3.86 2.002.06 6.21 8.786.195.85

6 4 2 Chemical Shift (ppm)

1 Figure 4.5 H NMR spectrum in CDCl3 of Br- Ph-NBu3.

96

243.9 241.9 242.1

Experimental Theoretical -NMe3 182.8

200 400 600 800 1000 1200 1400 1600 1800 m/z

Figure 4.6 ESI mass spectrum of Br- Ph-NMe3.

97

284.1 284.1 284.1

Experimental Theoretical

100 200 300 400 500 600 700 800 900 m/z

Figure 4.7 ESI mass spectrum of Br-Ph-NEt3.

98

368.1 368.1 368.2

Experimental Theoretical

237.1

500 1000 1500 2000 2500 m/z

Figure 4.8 ESI mass spectrum of Br- Ph-NBu3. The peak at m/z 237.1 is due to the di- substituted compounds.

4.3.2 Synthesis of quaternary ammonium-containing RAFT agents

The synthesis of the RAFT agents is shown in the second step of Figure 4.1.

By using triethylamine as the base to form the trithiocarbonate anion, the RAFT agents were synthesized in one-pot at room temperature without using phase transfer catalysts, which can be difficult to remove. CS2 was added in excess to promote the formation of trithiocarbonate groups and prevent the formation of the sulfide side product (Figure 4.9). 191, 192 The 1H NMR spectra and ESI-MS spectra of all three

RAFT-NR3 confirmed the success of the synthesis, as shown in Figure 4.10 to Figure

4.15. In the 1H NMR spectra, proton h exhibits different chemical shifted for different substituent alkyl chains. The resonances due to the alkyl chains were all observed with correct integration ratio. In the ESI spectra, the observed molar mass were very

99

close to the theoretical predication for all three RAFT agents, which also indicated the purity of these RAFT agents.

Trithiocarbonate Sulfide

Figure 4.9 Chemical Structure of RAFT-NR3 and the potential sulfide by-product.

S e f a042313-MONO-NME3-RAFT-APH-DCa c b S S g i Br b d N h d+i

CHCl3

h e a f g c

4.37 2.032.08 10.75 3.0022.89

6 4 2 Chemical Shift (ppm)

Figure 4.10 NMR spectrum in CDCl3 of RAFT-NMe3.

100

042313-MONO-NET3-RAFT-PURIFIEDBYTOLUENE-APH-DC b

S e f j c g i a S S Br b d N h j+H2O

CHCl3 i h e a g f d c

2.022.16 1.991.96 8.12 31.63 3.00

7 6 5 4 3 2 1 Chemical Shift (ppm)

Figure 4.11 NMR spectrum in CDCl3 of RAFT-NEt3.

101

071613-RAFT-NBU3-APH-DC b

c e f a g i k b d l h j l

c+j d+i k CHCl3 h e a gf

4.30 2.001.94 8.18 35.6611.47

6 4 2 Chemical Shift (ppm)

Figure 4.12 NMR spectrum in CDCl3 of RAFT-NBu3.

102

440.2

440.2 440.2

Experimental Theoretical

200 400 600 800 1000 1200 1400 1600 1800 m/z

Figure 4.13 ESI mass spectrum of RAFT-NMe3.

103

482.3 482.3 482.3

Experimental Theoretical

200 400 600 800 1000 1200 1400 1600 1800 m/z

Figure 4.14 ESI mass spectrum of RAFT-NEt3.

566.2 566.4 566.2

Experimental Theoretical

200 400 600 800 1000 1200 1400 1600 1800 m/z

Figure 4.15 ESI mass spectrum of RAFT-NBu3.

104

4.3.3 Bulk styrene polymerization

Based on the successful synthesis of the three cationic RAFT agents, their

abilities to control polymerizations were first investigated through the bulk RAFT

polymerization of styrene at 120 °C with each RAFT agent. Here, no added initiator

was needed since styrene undergoes self-initiated thermal polymerization with

reasonable polymerization rate at above 100 °C. 125,193 The NMR spectra and SEC traces, the pseudo first-order kinetic plot and the number average molecular weight

(Mn) and polydispersity index (Ð) versus monomer conversion for the polymerization

with RAFT-NEt3 are shown in Figure 4.17 to Figure 4.19. The conversion was

determined by 1H NMR by comparing the integral area of the monomer vinyl

resonance at 5.25 ppm (one proton, Ph-CH=CHHcis) and the integral area due to

polystyrene between 6.3-7.8 ppm (5 protons, ArH) after subtracting the contribution

from the monomer according to the following equation.

( . . 6 . ) 5 = . + ( . . 6 . ) 5 퐴6 3−7 8 − 퐴5 25 ⁄ 푐표푛푣푒푟푠푖표푛 5 25 6 3−7 8 5 25 As shown in Figure 4.16, in퐴 the region퐴 from −6.3-퐴7.8 ppm,⁄ there are 5 protons

from polystyrene, i.e. aromatic proton e, and 6 protons from styrene monomer, i.e.

proton c and aromatic proton d.

105

a H d c 043013-PS-NET3BR-6HR b H H n

d e b a

e

c

7 6 5 4 3 2 1 66 Chemical Shift (ppm)

Figure 4.16 1H NMR spectrum of the reaction mixture taken from bulk o polymerization of styrene using RAFT-NEt3 at 120 C for 6 hr.

106

1.2 NMR data 1.0 Linear fit ) t 0.8 /[M] 0 0.6

ln([M] 0.4 0.2 0.0 0 1 2 3 4 5 6 7 Time (hr)

Figure 4.17 Pseudo first-order kinetic plot for the RAFT bulk polymerization of styrene at 120 °C with RAFT-NEt3. Target molecular weight is 25 kDa. The solid line is a linear fit to the data.

107

20000 2.0 SEC Mn Mn, theory 16000 D 1.8 12000 1.6 D (g/mol) n 8000 1.4 M 4000 1.2

0 1.0 0.0 .2 .4 .6

Monomer Conversion

Figure 4.18 Plot of Mn (SEC) and PDI versus monomer conversion for the RAFT bulk polymerization of styrene at 120 °C with RAFT-NEt3. Target molecular weight is 25 kDa.

108

1.2 1 hr 1.0 2 hr 3 hr .8 4 hr 5 hr .6 6 hr .4 .2

Normalized RI Intensity 0.0 12 14 16 18 20 22

Elution Time (min)

Figure 4.19 SEC traces as a function of polymerization time for the RAFT bulk polymerization of styrene at 120 °C with RAFT-NEt3. Target molecular weight is 25 kDa.

The Mn and Ð were determined by SEC calibrated with PS standards. These plots are as expected for a controlled RAFT polymerization. First, the linearity of the pseudo first order kinetic plot (Figure 4.17) implies a constant radical concentration during the polymerization up to 58% monomer conversion. Second, Mn increases

linearly with monomer conversion (Figure 4.18) and agrees well with the theoretical

Mn. Third, a relatively narrow molecular weight dispersity (<1.25) (Figure 4.19) is

observed throughout the polymerization. Linear pseudo-first order kinetic plots were

also observed for the styrene polymerizations conducted with RAFT-NMe3 and

RAFT-NBu3 (Figure 4.20 and Figure 4.21).

109

.7 NMR data .6 Linear fit ) t .5

/[M] .4 0 .3

ln([M] .2 .1 0.0 0 1 2 3 4 5 6 7

Time (hr)

Figure 4.20 Pseudo first-order kinetic plot for the RAFT bulk polymerization of styrene at 120 °C with RAFT-NMe3. Target molecular weight is 25 kDa. The solid line is a linear fit to the data.

110

1.0 NMR data .8 Linear fit ) t .6 /[M] 0 .4 ln([M] .2

0.0 0 1 2 3 4 5 6 7

Time (hr)

Figure 4.21 Pseudo first-order kinetic plot for the RAFT bulk polymerization of styrene at 120 °C with RAFT-NBu3. Target molecular weight is 25 kDa. The solid line is a linear fit to the data.

While the RAFT polymerizations were demonstrated to be well-controlled up to ca. 60% monomer conversion, the quaternary ammonium functionality in the R-

group of the RAFT agents was unstable at this polymerization temperature. As shown

in the 1H NMR spectrum of the in Figure 4.22, the intensity of the resonance a at 4.88

+ ppm, (CHar)2CCH2N (CH2CH3)3, and resonance b at 3.47 ppm,

+ (CHar)2CCH2N (CH2CH3)3, was lower than expected after 1h of polymerization and

decreased as polymerization proceeded. The quaternary ammonium end functionality

as a function of polymerization time for all three RAFT agents is shown in Figure

4.23. The end-group functionality was determined by the integral area of the –CH2–

111

+ peak, (CHar)2CCH2N R3, in the quaternary ammonium group to the methyl peak in the

C12H25 group.

043013-PS-NET3BR-1HR

b d a N d Br S S n c S

a x b c

0.94 2.77 3.00

4 3 2 1 Chemical Shift (ppm)

Figure 4.22 NMR spectrum of PS from styrene bulk polymerization mediated by RAFT-NEt3 at 120 °C for 1h. One drop of the polymerization mixture was taken out 1 and diluted with CDCl3 and ran H NMR subsequently. Peak “x" is attributed to the thermal degradation product.

112

70 RAFT-NMe 60 3 RAFT-NEt3

50 RAFT-NBu3 40 30 20 10

End group functionality (%) 0 0 1 2 3 4 5 6 7

Polymerization Time (h)

Figure 4.23 The end functionality as a function of polymerization time at 120 °C.

In addition, the thin layer chromatography (TLC) of the polystyrene prepared with RAFT-NEt3 (1h) qualitatively showed that a large fraction of the polymer

products lacked ionic end-groups (Figure 4.24). When toluene was used as developing

solvent, two spots were seen at retardation factors (Rf) of 0 and 1. The spot at Rf = 0 is attributed to the PS with a quaternary ammonium end-group as it is more polar than the non-ionic PS. The non-ionic PS is attributed to the spot at Rf = 1. The intensity of the spot of Rf = 1 is greater than the spot at Rf = 0, consistent with the degradation of

the quaternary ammonium end-group during polymerization to a non-ionic species.

113

120 oC

65 oC

Rf=0, polymers with Rf=1, polymers w/o ionic end group ionic end group

Figure 4.24 TLC test of polystyrene prepared from bulk polymerization at 120 °C for 1 h and 65 °C for 24 h. Toluene was used as the developing solvent.

The loss of the cationic end-group was attributed to the thermal degradation of

171 the quaternary ammonium groups. Thermal gravimetric analysis (TGA) were

performed to investigate the thermal stability of the quaternary ammonium

functionality of all three RAFT-NR3 agents (R=methyl, ethyl and n-butyl), as shown in Figure 4.25. A RAFT agent, benzyl dodecyl trithiocarbonate (BDTC), with the same structure other than the quaternary ammonium end-group was synthesized and used as a control for the thermal degradation study. The degradation of BDTC occurs as a single step above 200 °C while RAFT-NEt3 begins to degrade at ~150 °C.

Compared to RAFT-NEt3, RAFT-NMe3 and RAFT-NBu3 showed better and worse

thermal stability, respectively, which was in accordance with the trend in Figure 4.23.

Quaternary ammonium salts usually degrade through two mechanisms, i.e. reverse

nucleophilic substitution or Hoffman elimination, for which the counterion abstracts

the β-hydrogen of the alkyl chains. In addition, the length of the substituent alkyl

chains will influence the thermal stability of the quaternary ammonium salts.

114

Typically, longer alkyl chains will decrease the thermal stability. Therefore, RAFT-

NMe3 exhibits the highest thermal stability among the three RAFT agents since it has the shortest alkyl chain and no β-hydrogen is present.

100 RAFT-NEt3

Br-NEt3 80 BDTC RAFT-NMe3 RAFT-NBu 60 3

40

20 Mass Remaining (%) 0 0 100 200 300 400 500 600 700 o Temperature ( C)

Figure 4.25 TGA curves for RAFT-NR3 (R=methyl, ethyl and n-butyl), Br- Ph-NEt3 and BDTC RAFT. Experiments were conducted using a nitrogen atmosphere with a heating rate of 20 °C/min.

Isothermal tests (Figure 4.26) were also run for RAFT-NEt3 at 120 °C, the temperature for the previously discussed polymerization, and 65 °C, a typical temperature for RAFT polymerization with AIBN initiation. Except the weight loss in the initial stage, which is probably due to moisture, it is clear that RAFT-NEt3 is stable at 65 °C while it degrades at 120 °C. As the BDTC RAFT agent is stable up to

200 °C, the weight loss at 120 °C in Figure 4.26 is attributed to the thermal degradation of the quaternary ammonium functional group.

115

100 98 96 94 92 90 120 oC 88 65 oC Weight Remaining (%) 86 0 50 100 150 200 250 300 350

Time (min)

Figure 4.26 Isothermal TGA curves for RAFT-NEt3 at 65 °C and 120 °C. Experiments were conducted using a nitrogen atmosphere.

To preserve the quaternary ammonium group and obtain a polymer with high

end functionality, a styrene polymerization was conducted at 65 °C. The kinetics of

the polymerizations at 65 °C was studied and exhibited living polymerization

characteristics up to 26% monomer conversion, as shown in Figure 4.27 to Figure

4.29.

116

0.4 NMR data Linear fit 0.3 ) t /[M] 0 0.2

ln([M] 0.1

0.0 0 1 2 3 4 5 6 7 Time (hr)

Figure 4.27 Pseudo first-order kinetic plot for the RAFT bulk polymerization of styrene at 65 °C with RAFT-NEt3. Target molecular weight is 25 kDa. The solid line is a linear fit to the data.

117

12000 2.0 SEC Mn Mn, theory D 1.8 8000 1.6 D (g/mol) n 1.4 M 4000 1.2

0 1.0 0.00 .05 .10 .15 .20 .25 .30 Monomer Conversion

Figure 4.28 Plot of Mn (SEC) and PDI versus monomer conversion for the RAFT bulk polymerization of styrene at 65 °C with RAFT-NEt3. Target molecular weight is 25 kDa.

118

1.4 1 hr 1.2 2 hr 3 hr 1.0 4 hr 5 hr .8 6 hr .6 .4 .2 Normalized RI Intensity 0.0 12 14 16 18 20

Elution Time (min)

Figure 4.29 SEC traces as a function of polymerization time for the RAFT bulk polymerization of styrene at 65 °C with RAFT-NEt3. Target molecular weight is 25 kDa.

TLC tests qualitatively showed that, for the bulk styrene polymerizations with

RAFT-NEt3, the polystyrene product obtained at 65 °C contained a much higher ionic end functionality compared to that from 120 °C (Figure 4.24). Although the quaternary ammonium group is stable at 65 °C, there are still some polymer impurities presented in the polystyrene obtained at 65 °C, due to those chains initiated by AIBN instead of the cationic R-group in the RAFT agent,135 and, if any, those chains with degraded end-groups. Removing these impurities is significant since it will improve the end functionality of the cationomers and also block copolymers obtained by sequential polymerization using these cationic RAFT agents. Thus, the crude polystyrene was purified by silica gel column chromatography by eluting first

1 with toluene, followed with a mixture of CHCl3/acetone/methanol (1: 1: 0.1). The H

119

NMR spectra in Figure 4.30 show that this purification procedure effectively separates the ionic and non-ionic polymer species. For the non-ionic polystyrene

obtained from the toluene fraction, the resonances at 4.7 ppm and 3.4 ppm are missing

due to the absence of the quaternary ammonium groups. For the polymer eluted with

1 CHCl3/acetone/methanol, the end functionality estimated from H NMR is >99%

compared to the 89% functionality in crude polystyrene.

120

a b N d Br S S n c S

1.ESP

(a) a d b c 1.78 7.16 3.00 PS-0.ESP(b)

a d b c 2.07 8.51 3.00

PS-1.ESP(c)

d c

6 4 2 Chemical Shift (ppm)

Figure 4.30 1H NMR spectra of a) crude polystyrene from styrene bulk polymerization at 65 °C, b) the purified polystyrene (CHCl3/acetone/methanol fraction), c) toluene fraction.

The ionic functionality of the obtained polymers was also confirmed by

mixing the polymer dissolved in toluene with methyl orange, a dye containing an anionic sulfonate group, dissolved in water. Four tests are shown in Figure 4.31.

Under quiescent conditions the two solvent separate into a lower aqueous layer and an upper toluene layer. In the control samples (vial #1-2), consisting of only toluene,

121

water, and either methyl orange (vial #1) or polystyrene (vial #2), the dye partitions

primarily to the water layer and the polymer partitions primarily to the toluene layer.

Vial #3 contains the purified PS, which is able to extract the methyl orange into the

toluene layer. The non-ionic polystyrene fraction is dissolved in vial #4, and is unable

to extract the dye into the organic phase.

Figure 4.31 The ion exchange tests that demonstrated the ionic functionality. The vials were shaken to perform ion exchange and then placed in the hood for overnight. The contents of each solution before shaking were as follows: Vial #1: toluene (top layer) & water+ methyl orange (bottom layer); Vial #2: toluene + purified polystyrene (Rf=0) (top layer) & water (bottom layer); Vial #3: toluene + purified polystyrene (Rf=0) (top layer) & water + methyl orange (bottom layer); Vial #4: toluene + polystyrene impurities (Rf=1) (top layer) & water +methyl orange (bottom layer).

Besides styrene, three acrylate monomers were also polymerized at 65 °C using RAFT-NEt3. For PMA, the crude product was purified by column

chromatography by eluted with CHCl3/acetone (3/1 v/v) and followed with acetone.

The NMR spectra of the crude product, the impurities and the purified PMA were

122

shown in Figure 4.32. Similar to the purification of polystyrene, column chromatography effectively separated the ionic and non-ionic species in the crude

PMA. The end-functionality of the purified PMA was 93%, as estimated by 1H NMR spectroscopy. PBA was also purified by column chromatography by eluting the column with CHCl3/acetone (10/1 v/v) and followed with acetone. The end

functionality is greater than 99%, as estimated by the 1H NMR spectroscopy in Figure

4.33. The column purification of PDMAEA was also attempted but appropriate

solvent conditions were not found for separation due to the strong interaction of the

polymer with the silica gel. The crude PDMAEA showed 85% end functionality

based on 1H NMR results (Figure 4.34). A similar cationic RAFT agent has also been used to prepare poly(N-isopropyl acrylamide)-b-poly(tert-butyl acrylate) diblock copolymers,176 which implies that these cationic RAFT agents should be useful in the polymerization of a range of polymer chemistries.

123

c d b N e c d f g i k m a Br e S S d c n jk i f g i S l m a b OO O O j l h h g a, f, k l 082313-PMA-CRUDE-APH m e+CHCl3 MeOH b d c j i 1.020.38 3.00 h g a, f, k l 082313-PMA-RF=0 m CHCl 3 b c d j e i 2.031.91 1.860.79 2.087.07 2.99 f, k l 082313-PMA-RF=1 m h g

j i

7 6 5 4 3 2 1 Chemical Shift (ppm)

Figure 4.32 1H NMR spectra of a) crude PMA, b) the purified PMA (acetone fraction), c) CHCl3/acetone fraction.

124

c d b N e f n p a Br g l S S d c n m i S o k 102213-PBA-NET3-PURIFIED-1.ESP OO O O a,f,i,j,n h i j CHCl3 h k

o

g p b d e c l m

1.972.13 2.001.06 116.51 2.006.13 56.47 17.91 176.60

7 6 5 4 3 2 1 Chemical Shift (ppm)

Figure 4.33 1H NMR spectrum of purified PBA (acetone fraction).

125

c d b N e f k m o a Br g S S d c n l i S n OO O O h i j 101913-PDMAEA-NET3.espe+CHCl3 N N j n

i h a,f,m o b

d kc l

2.10 1.700.81 32.39 1.785.32 31.09 106.70 42.60 3.0016.79

7 6 5 4 3 2 1 Chemical Shift (ppm)

Figure 4.34 1H NMR spectrum of PDMAEA.

4.4 Conclusions

In summary, we optimized the synthesis of a series of 4-(bromomethyl)-

N,N,N-trialkylbenzyl ammonium compounds (alkyl=methyl, ethyl and n-butyl) and

the corresponding RAFT agents. The quaternary ammonium functional groups

thermally degraded in bulk polymerization at 120 °C, which can be overcome by

using lower polymerization temperature, such as 65 °C. High purity α-N,N,N-trialkyl

benzyl ammonium hemi-telechelic cationomers were obtained by purifying the crude

polymers with silica gel column chromatography. The straightforward synthesis and

the broad applicability of RAFT polymerization should make these RAFT agents and

resultant hemi-telechelic cationomers useful in the range of applications including supramolecular polymers, dispersants, aqueous polymerizations, stimuli-responsive polymers, and antimicrobial polymers, as discussed in the introduction.

126

4.5 Acknowledgements

This material is based upon work supported by the National Science

Foundation under Grant No. CHE-1012237 (KAC), DMR-1309853 (RAW) and CHE-

1012636 (CW).

127

CHAPTER V

SYNTHESIS AND CHARACTERIZATION OF QUATERNARY

PHOSPHONIUM-CONTAINING, TRITHIOCARBONATE RAFT AGENTS*

5.1 Introduction

Hemi-telechelic cationomers, polymers bearing a cationic group at one end,

are an interesting class of end-functional polymers. In a previous publication the authors demonstrated the synthesis of quaternary ammonium-containing RAFT agents useful for the synthesis of hemi-telechelic cationomers. 194 While telechelic

cationomers with quaternary ammonium cations have been popular in a number of

applications due to their affordability and accessibility, 7,168,169,174,175,176,178 quaternary

phosphonium cations have been attracting attention for their higher thermal

stability, 195 , 196 improved antimicrobial activity, 197 and improved gene

delivery,198,199,200,201 compared to quaternary ammonium salts.

(Hemi)telechleic quaternary phosphonium cationomers have been synthesized

with various methods, such as anionic polymerization,181 cationic

polymerization, 202 , 203 , 204 , 205 polycondensation, 206 group transfer polymerization, 207

free radical polymerization with a functional chain transfer agent,208 and atom transfer

radical polymerization. 209 However, to our knowledge, the syntheses of hemi-

telechelic, quaternary phosphonium cationomers via reversible addition fragmentation

chain transfer (RAFT) polymerization have not been reported. RAFT polymerization

128

is one of the most versatile controlled polymerization techniques due to its ability to

polymerize a wide variety of monomers and its high tolerance to monomer

functionality and reaction conditions.117 In addition, the functional groups present in

the initial RAFT agent (CTAs) are retained in the final polymer. This allows the

synthesis of telechelic polymers with a wide range of α- or ω-end groups via the

design of RAFT agents. 117, 122, 210

In this chapter, the syntheses of quaternary phosphonium-containing,

trithiocarbonate RAFT agents (RAFT-PR3) (Figure 5.1) and the bulk thermally

initiated polymerization (i.e. thermal polymerization) of styrene are presented. These

results are compared to the previous investigation of quaternary ammonium-

containing trithiocarbonate RAFT agents where the quaternary ammonium groups degraded during thermal polymerization of polystyrene at 120°C due to their lower thermal stability.194 It is shown that the quaternary phosphonium has much higher

thermal stability, resulting in much higher end-group fidelity compared to quaternary

ammonium-containing RAFT agents in the thermal polymerization of polystyrene at

120 °C.

Br Br + PR3 Br PR3 Br R=Bu, Ph 1 S C12H25 NEt3 1+ C12H25SH + CS2 S S PR3Br R=Bu, Ph

Figure 5.1 Synthesis of quaternary phosphonium containing trithiocarbonate RAFT agents, RAFT-PR3 (R=n-butyl and phenyl).

129

5.2 Experimental section

5.2.1 Materials

Azobisisobutyronitrile (98%, Sigma-Aldrich) was recrystallized from methanol and dried under vacuum prior to use. Styrene (99%, stabilized, Acros

Organics) was purified by passing over a column of basic alumina. Silica gel

(Dynamic Adsorbents 60Å, 32–63 μM, flash grade) was used for column chromatography. Thin layer chromatography plates (250 μm, with fluorescent

indicator activated at 2540 Å) were supplied from J. T. Baker. Benzyl dodecyl

trithiocarbonate was synthesized according to previous report.194 All other chemicals

used in this article were obtained commercially with high purity and used as received.

5.2.2 Instrumentation

1H NMR, 13C NMR and 31P NMR spectra were collected using either a Varian

Gemini 300 MHz or a Varian 500 MHz spectrometer. Thermogravimetric analysis

(TGA) was performed on a TA Instruments TGA Q50 from room temperature to 700

oC at a heating rate of 20 oC /min in nitrogen atmosphere. All mass spectrometry

experiments were acquired on an HCT Ultra II quadrupole ion trap mass spectrometer

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

molecular weight and molecular weight distribution of the polymer products was

characterized by size exclusion chromatography (SEC) with a Waters Breeze system

equipped with a column set at 35 oC and a refractive index detector (Waters 2414).

The column set consists of a Styragel® HR 4 THF column (4.6 × 300 mm) with an

effective molecular weight range 5 k to 600 k Da, a Styragel® HR 3 THF column (4.6

× 300 mm) with an effective molecular weight range 0.5 k to 30 k Da, and a

Styragel® HR 4E THF column (4.6 × 300 mm) with an effective molecular weight

130

range 0.05 k to 100 k Da. The SEC was calibrated using PS standards of narrow molar

mass dispersity (Ð) with the molecular weight of 1300 Da to 400 k Da (Alfa Aesar).

5.2.3 Synthesis of 4-(bromomethyl)benzyltri-n-butylphosphonium bromide (Br-Ph-

PBu3)

Tri-n-butylphosphine (1.008 g, 4.9 mmol) was added to a stirred solution of

α,α'-dibromo-p-xylene (2.643 g, 10.0 mmol) in 20 mL ethyl acetate. The reaction

mixture was allowed to stir for 48 hours. The product, Br-Ph-PBu3, gradually

precipitated as a fine white powder. Br-Ph-PBu3 was recovered by filtration, washed

with 40 mL ethyl ether and dried in a vacuum oven (2.222 g, 96% yield). Due to the

lipophicity of the mono-substituted product, the obtained product contained ca. 1 mol%

(calculated by 1H NMR) di-substituted compound, i.e. 1,4-bis(tri-n-

1 butylphosphoniummethyl)benzene dibromide. H NMR (300 MHz, CDCl3) δ (ppm):

+ 7.49 (dd, J = 8.1 Hz, 2.2 Hz, 2H, (CHar)2CCH2P (CH2CH2CH2CH3)3), 7.36 (d, J =8.1

Hz, 2H, BrCH2C(CHar)2(CHar)2C), 4.45 (s, 2H, BrCH2C(CHar)2), 4.38 (d, J = 15.4 Hz,

+ 2H, (CHar)2CCH2P (CH2CH2CH2CH3)3), 2.41 (m, 6H, (CHar)2CCH2-

+ + P (CH2CH2CH2CH3)3), 1.45 (m, 12H, (CHar)2CCH2P (CH2CH2CH2CH3)3), 0.91 (t,

+ 13 J=6.8 Hz, 9H, (CHar)2CCH2P (CH2CH2CH2CH3)3). C NMR (125 MHz, CDCl3) δ

(ppm): 137.9 (d, JCP = 3.7 Hz), 130.5 (d, JCP = 5.1 Hz), 129.7, 129.0 (d, JCP = 8.8 Hz),

32.6, 26.7 (d, JCP = 45.1 Hz), 23.8 (d, JCP = 14.0 Hz), 23.5 (d, JCP = 4.7 Hz), 18.8 (d,

31 JCP = 46.5 Hz), 13.3. P NMR (121.5 MHz, CDCl3) δ (ppm): 31.4. MS (ESI-MS)

- calcd for C20H35BrP (m/z), 385.1; found, 385.3 [M-Br ].

131

5.2.4 Synthesis of 4-(bromomethyl)benzyltriphenylphosphonium bromide (Br-Ph-

PPh3)

Triphenylphosphine (1.357 g, 5.2 mmol) was added to a stirred solution of

α,α'-dibromo-p-xylene (1.317 g, 5.0 mmol) in 20 mL ethyl acetate. The reaction

mixture was stirred for 48 hours, during which Br-Ph-PPh3 gradually precipitated as a

white powder. Br-Ph-PPh3 was recovered by filtration, washed with 40 mL ethyl ether and dried in a vacuum oven (2.357 g, 90% yield). The obtained product was free of

1 di-substituted compound. H NMR (300 MHz, CDCl3) δ (ppm): 7.75 - 7.51 (m, 15H,

+ (CHar)2CCH2P Ph3), 7.06 (m, 4H, BrCH2C(CHar)2(CHar)2C), 5.46 (d, J = 14.7 Hz, 2H,

+ 13 (CHar)2CCH2P Ph3), 4.32 (s, J = 1.2 Hz, 2H, BrCH2C(CHar)2). C NMR (125 MHz,

CDCl3) δ (ppm): 138.1 (d, JCP = 4.2 Hz), 134.9 (d, JCP = 3.3 Hz), 134.4 (d, JCP = 9.8

Hz), 132.0 (d, JCP = 5.6 Hz), 130.1 (d, JCP = 12.6 Hz), 129.3 (d, JCP = 3.3 Hz), 127.7 (d,

31 JCP = 8.8 Hz),118.1 (d, JCP = 85.6 Hz), 32.8, 30.6 (d, JCP = 47.0 Hz). P NMR (121.5

MHz, CDCl3) δ (ppm): 23.5. MS (ESI-MS) calcd for C26H23BrP (m/z), 445.1; found,

444.9 [M-Br-].

5.2.5 Synthesis of S-1-dodecyl-S’-(methylbenzyltributylphosphonium bromide)

trithiocarbonate RAFT agents (RAFT-PBu3)

Triethylamine (0.525 g, 5.2 mmol) was added dropwise to a stirred solution of

1-dodecanethiol (0.8642 g, 4.3 mmol) and CS2 (1.6087 g, 21.1 mmol) in CH2Cl2 (5

mL) at room temperature. After stirring for 3 hours, Br-Ph-PBu3 (2.0691 g, 4.4 mmol)

was directly added into the reaction mixture. The reaction mixture was stirred for

another 30 hours, diluted with an additional amount of CH2Cl2 (20 mL) and washed

with deionized water in a separatory funnel for four times. A small amount of

concentrated sodium bromide aqueous solution was added to the separatory funnel to

132

facilitate the phase separation. The organic phase was collected, dried over MgSO4,

and rotary evaporated to produce a viscous yellow liquid. The product was further

o 1 dried in a vacuum oven at 50 C (2.760 g, 97% yield). H NMR (300 MHz, CDCl3) δ

+ (ppm): 7.47 (m, 2H, (CHar)2CCH2P (CH2CH2CH2CH3)3), 7.35 (m, 2H,

S=CSCH2C(CHar)2(CHar)2C), 4.61 (s, 2H, S=CSCH2C(CHar)2), 4.34 (d, J = 14.9 Hz,

+ 2H, (CHar)2CCH2P (CH2CH2CH2CH3)3), 3.41 (t, J = 7.0 Hz, 2H,

+ CH3C9H18CH2CH2CS=S), 2.43 (m, (CHar)2CCH2P (CH2CH2CH2CH3)3), 1.72 (m, 2H,

+ CH3C9H18CH2CH2CS=S), 1.21-1.55 (br m, 30H, (CHar)2CCH2P (CH2CH2CH2CH3)3

+ and CH3C9H18CH2CH2CS=S), 0.92 (m, 12H, (CHar)2CCH2P (CH2CH2CH2CH3)3 and

13 CH3C9H18CH2CH2CS=S). C NMR (125 MHz, CDCl3) δ (ppm): 223.2, 136.0 (d, JCP

= 3.7 Hz), 130.5 (d, JCP = 4.7 Hz), 130.2 (d, JCP = 3.3 Hz), 128.0 (d, JCP = 8.4 Hz),

40.5, 37.2, 31.9 (m), 29.6-22.6, 26.9 (d, JCP = 45.1 Hz), 24.0 (d, JCP = 15.4 Hz), 23.7

31 (d, JCP = 5.1 Hz), 19.0 (d, JCP = 46.5 Hz), 14.1, 13.4. P NMR (121.5 MHz, CDCl3) δ

- (ppm): 31.6. MS (ESI-MS) calcd for C33H60PS3 (m/z), 583.4; found, 583.5 [M-Br ].

5.2.6 Synthesis of S-1-dodecyl-S’-(methylbenzyltriphenylphosphonium bromide)

trithiocarbonate RAFT agents (RAFT-PPh3)

RAFT-PPh3 was synthesized following an analogue procedure as described

above and was obtained as a yellow solid upon drying (2.966 g, 96% yield.) 1H NMR

+ (300 MHz, CDCl3) δ (ppm): 7.79 - 7.52 (m, 15H, (CHar)2CCH2P Ph3), 7.07 (m, 4H,

+ S=CSCH2C(CHar)2(CHar)2C), 5.46 (d, J = 14.4 Hz, (CHar)2CCH2P Ph3), 4.49 (d, J =

1.5 Hz, S=CSCH2C(CHar)2(CHar)2C), 3.35 (t, J = 7.3 Hz, 2H,

CH3C9H18CH2CH2CS=S), 1.68 (m, 2H, CH3C9H18CH2CH2CS=S), 1.44-1.18 (br m,

13 18H, CH3C9H18CH2CH2CS=S), 0.87 (t, J = 6.8 Hz, 3H, CH3C9H18CH2CH2CS=S). C

NMR (125 MHz, CDCl3) δ (ppm): 223.1, 135.5 (d, JCP = 4.2 Hz), 134.9 (d, JCP = 2.8

133

Hz), 134.3 (d, JCP = 9.8 Hz), 131.7 (d, JCP = 3.7 Hz), 130.1 (d, JCP = 12.6 Hz), 129.4,

126.8 (d, JCP = 8.4 Hz), 117.9 (d, JCP = 86.1 Hz), 40.6, 37.1, 31.8 (m), 30.4 (d, JCP =

31 46.5 Hz), 29.6-22.6, 14.1. P NMR (121.5 MHz, CDCl3) δ (ppm): 23.3. MS (ESI-MS)

- calcd for C39H48PS3 (m/z), 643.3; found, 643.4 [M-Br ].

5.2.7 RAFT bulk polymerization of styrene at 120 °C

Bulk styrene polymerizations with either RAFT-PBu3 or RAFT-PPh3 were

carried out at 120°C. For all these polymerizations, the theoretical molecular weight targeted was 25,000 g/mol. The following procedure is typical: A master batch was prepared by dissolving RAFT-PBu3 (0.8370 g, 1.26 mmol) in styrene (30.68 g, 294.6

mmol). Aliquots (4 mL) of the master batch were charged to separate flasks each equipped with a magnetic stir bar and sealed with a rubber septum. The flasks were

purged with dry nitrogen for 20 min, placed into a preheated oil bath at 120 oC and

finally quenched by placing the flasks in ice water at selected time points to terminate

the polymerizations. A drop of the reaction mixture was taken out to determine the

monomer conversion by 1H NMR using the following equation,

( . . 6 . ) 5 = . + ( . . 6 . ) 5 퐴6 3−7 8 − 퐴5 25 ⁄ 푐표푛푣푒푟푠푖표푛 5 25 6 3−7 8 5 25 where . denotes the integral area퐴 of peak퐴 at 5.25− ppm퐴 due⁄ to the styrene vinyl

5 25 peak (one퐴 proton, Ph-CH=CHHcis), ( . . 6 . ) as a whole denotes the integral

6 3−7 8 5 25 area due to polystyrene between 6.3 퐴to 7.8 ppm− 퐴(5 protons, ArH). The polymer was

isolated by precipitation twice into methanol and dried in a vacuum oven at 50 oC.

134

5.3 Results and discussion

5.3.1 Synthesis of the RAFT agents

RAFT-PR3 agents were synthesized in two steps, as shown in Figure 5.1.

First,α,α’-dibromo-p-xylene was selectively quaternized with PBu3 or PPh3 by

precipitating the mono-substituted compounds in ethyl acetate under mild conditions.

Compared to their quaternary ammonium analogues,194 mono-substituted compounds

containing phosphonium salts exhibited less lipophicity and precipitated from the

reaction solution more readily. Following previously reported procedures produced

precipitates containing either di-substituted compounds in the case of tributylphosphine or by-products in the case of triphenylphosphine.211 In the second

step, the corresponding trithiocarbonate RAFT agents were prepared in high yield via

the reaction of the alkyl trithiocarbonate anion and the benzyl bromide functionality

presented in the mono-substituted compounds. The 1H NMR and 31P NMR spectra of

all of the mono-substituted compounds and the corresponding RAFT agents are

shown in Figure 5.2 to Figure 5.9. Each 31P NMR spectrum shows single peak, which

indicates the high purity of the mono-substituted compounds and the corresponding

1 RAFT agents. In H NMR, proton a of both Br-Ph-PBu3 and Br-Ph-PPh3 are

completely shifted when bromide group was replaced by a trithiocarbonate group.

This indicated the mono-substituted compounds were quantitatively converted to the

corresponding RAFT agents.

135

120313-BR-PBU3-1-0.esp h a b g h c e Br f P Br d f,g

a 1% di-substituted by-products b d e c CHCl3

2.001.960.04 1.941.95 6.02 12.26 9.00

7 6 5 4 3 2 1 Chemical Shift (ppm)

1 Figure 5.2 H NMR spectrum in CDCl3 of Br-Ph-PBu3.

111613-BR-PBU3-1-0

80 60 40 20 0 -20 Chemical Shift (ppm)

31 Figure 5.3 P NMR spectrum in CDCl3 of Br-Ph-PBu3.

136

120413-BR-PPH3.espCHCl3 a Br Br b P e c d

e b,c a

d H2O

4.0115.54 2.03 2.00

7 6 5 4 3 2 1 Chemical Shift (ppm)

1 Figure 5.4 H NMR spectrum in CDCl3 of Br-Ph-PPh3.

120313-BR-PPH3-PNMR

80 60 40 20 0 -20 Chemical Shift (ppm)

31 Figure 5.5 P NMR spectrum in CDCl3 of Br-Ph-PPh3.

137

b, j,k 111613-RAFT-PBU3.esp

S a c e S S Br b d P f g h i j k a,l l

e f g h id c CHCl3

2.202.16 2.002.02 2.29 6.13 2.60 32.4111.97

7 6 5 4 3 2 1 Chemical Shift (ppm)

1 Figure 5.6 H NMR spectrum in CDCl3 of RAFT-PBu3. 111613-RAFT-PBU3-PNMR

80 60 40 20 0 -20 Chemical Shift (ppm)

31 Figure 5.7 P NMR spectrum in CDCl3 of RAFT-PBu3. 138

120313-RAFT-PPH3.esp b

S a c e S S Br b d P f g h i

CHCl3 a i f,g e h d c

15.35 4.06 2.02 1.96 2.02 3.0018.732.06

7 6 5 4 3 2 1 Chemical Shift (ppm)

1 Figure 5.8 H NMR spectrum in CDCl3 of RAFT-PPh3. 111613-RAFT-PPH3-PNMR

80 60 40 20 0 -20 Chemical Shift (ppm)

31 Figure 5.9 P NMR spectrum in CDCl3 of RAFT-PPh3. 139

5.3.2 Thermal stability of the synthesized RAFT agents

Following the successful syntheses of the RAFT-PR3, their thermal stabilities

were evaluated via temperature-ramped TGA (Figure 5.10) and isothermal TGA

(Figure 5.11). Td,5%, the temperatures at which the weight loss of the sample is 5%, of

all the tested samples are summarized in Table 5.1. The two mono-substituted

precursor compounds, i.e. Br-Ph-PBu3 and Br-Ph-PPh3, showed relatively high

o thermal stability. Td,5% of both were higher than 260 C. However, their corresponding

RAFT agents showed lower thermal stabilities and Td,5% of both were approximately

200 oC. A nonionic RAFT agent, benzyl dodecyl trithiocarbonate (BDTC), which has

the same structure as other RAFT agents except the cationic end group, was

synthesized as a control sample to study the thermal stabilities. BDTC showed a

o o single-step degradation starting from 200 C and Td,5% of BDTC is 228 C, higher

than both RAFT-PR3. Since the precursor bromomethyl benzyl quaternary

phosphonium salts exhibited high thermal stability, as shown in the TGA traces of Br-

Ph-PBu3 and Br-Ph-PPh3, the initial weight loss of RAFT-PR3 should be due to the

thermal degradation of the trithiocarbonates. Compared to their quaternary

194 ammonium analog (RAFT-NBu3), RAFT-PBu3 exhibited enhanced thermal

o stabilities with a Td,5% of the RAFT-PBu3 86 C higher than that of the RAFT-NBu3.

Therefore, by replacing the quaternary ammonium group with quaternary

phosphonium groups significantly improves the thermal stability of the cationic

functionality. Both the Br-Ph-PR3 and RAFT-PR3 show more complex decay profiles

than the BDTC. In addition, the Br-Ph-PR3 has a higher thermal stability than the

RAFT-PR3. Therefore, the initial decay in the RAFT-PR3 is likely due to the

incorporation of the trithiocarbonate group. However, the decomposition pathway is

140

complex and would need further analysis, such as TGA coupled with GC-MS, to understand fully.

Table 5.1 Degradation temperature when weight loss=5%

o Sample Td,5%/ C

Br-Ph-PBu3 282

Br-Ph-PPh3 267

RAFT-PBu3 200

RAFT-PPh3 201

RAFT-NBu3 114

BDTC 228

141

100 Br-Ph-PBu3 Br-Ph-PPh3 80 RAFT-PBu3 RAFT-PPh3 RAFT-NBu3 60 BDTC

40

20 Mass Remaining (%) 0 0 100 200 300 400 500 600 700 o Temperature ( C)

o Figure 5.10 Temperature-ramp TGA traces (20 C/min) for RAFT-PR3 (R=n-butyl and phenyl), Br-PR3, and RAFT-NBu3 and BDTC RAFT. Tests were performed in nitrogen atmosphere. Data of RAFT-NBu3 and BDTC were adapted from Ref.194 with permission from The Royal Society of Chemistry.

Due to its dynamic nature, temperature-ramp TGA overestimates the thermal

stability of the analyzed samples. 212 Therefore, isothermal TGA was employed to

better assess the thermal stability of the RAFT agents at a desired temperature.

Isothermal TGA of four RAFT agents was performed at 120 oC, as shown in Figure

5.11. BDTC and both RAFT-PR3 exhibited less than 5% weight loss while RAFT-

NBu3 showed ~50% weight loss after 6 hours. This result was consistent with the result of temperature-ramp TGA. Since TGA only reflects the change in mass but not in chemical structure, each sample was collected after the isothermal TGA tests and characterized by 1H NMR spectroscopy (Figure 5.12 to Figure 5.15). The results 142

showed the chemical structures of BDTC and RAFT-PBu3 were well retained while

1 RAFT-PPh3 slightly degraded. In contrast, the H NMR showed RAFT-NBu3 severely degraded.

100

80

60

RAFT-PBu3

RAFT-PPh3 RAFT-NBu 40 3 BDTC Mass Remaining (%) 20 0 100 200 300

Time (min)

o Figure 5.11 Isothermal TGA traces (120 C) for RAFT-PR3, RAFT-NBu3 and BDTC RAFT. Tests were performed in nitrogen atmosphere.

143

111613-RAFT-PBU3.ESP

2.202.16 2.002.02 2.29 32.412.606.13 11.97

121813-RAFT-PBU3-TGA

2.052.07 2.002.06 2.15 33.622.276.26 12.77

7 6 5 4 3 2 1 Chemical Shift (ppm)

1 Figure 5.12 H NMR spectrum in CDCl3 of RAFT-PBu3 before (top) and after (bottom) the isothermal TGA test, i.e. 120 oC for 6 h in nitrogen environment.

144

120313-RAFT-PPH3.ESP

4.0615.35 2.02 1.96 2.02 3.0018.732.06

121713-RAFT-PPH3-120-ISOTHERMAL-

4.3417.94 2.17 1.95 1.94 3.0020.511.90

7 6 5 4 3 2 1 Chemical Shift (ppm)

1 Figure 5.13 H NMR spectrum in CDCl3 of RAFT-PPh3 before (top) and after (bottom) the isothermal TGA test, i.e. 120 oC for 6 h in nitrogen environment.

145

011513-DODECANE-BENZYL-RAFT-AGENT-APH

4.46 1.97 2.00 2.7417.381.89

121813-RAFT-PH-TGA

4.23 1.95 2.00 2.9519.152.10

7 6 5 4 3 2 1 Chemical Shift (ppm)

1 Figure 5.14 H NMR spectrum in CDCl3 of BDTC before (top) and after (bottom) the isothermal TGA test, i.e. 120 oC for 6 h in nitrogen environment.

146

071613-RAFT-NBU3-APH-DC

121713-RAFT-NBU3-120-ISOTHERMAL

6 4 2 Chemical Shift (ppm)

1 Figure 5.15 H NMR spectrum in CDCl3 of RAFT-NBu3 before (top) and after (bottom) the isothermal TGA test, i.e. 120 oC for 6 h in nitrogen environment.

5.3.3 Bulk styrene polymerizations using RAFT-PR3

Based on the excellent thermal stability of both RAFT-PR3, their capability to control the bulk styrene polymerization at 120 oC was examined. The plots of pseudo first-order kinetics, evolution of number average molecular weight (Mn) and molecular weight dispersity (Ð) with monomer conversion, and SEC traces for the polymerizations using RAFT-PBu3 are shown in Figure 5.16 to Figure 5.18, while those using RAFT-PPh3 are shown in Figure 5.19 to Figure 5.21. The pseudo first-

order kinetic plot in Figure 5.16 is linear, which is consistent with a controlled free

radical polymerization. In Figure 5.17, Mn increases with increasing monomer

conversion. At higher monomer conversion region, Mn is in good agreement with the 147

theoretical predication while in the lower conversion region, i.e. shorter

polymerization time, Mn is slightly higher than theoretical value. This is attributed to the “hybrid behavior” caused by the relatively slow consumption of the initial RAFT agents.117 Since the theoretical curve is calculated based on the full consumption of

the RAFT agents, the observed Mn will be higher than the theoretical values if the

RAFT agents were not completely consumed. As expected for a controlled

polymerization, relatively narrow Ð (Ð<1.3) is observed for all polymerizations.

These SEC results were obtained using an older column set that had previously been

treated many times with a 2 wt% solution of tri-n-octylamine in THF.213 Using a new

set of Styragel columns no elution of the phosphonium terminated polymers was

observed. Therefore, the older column set has been conditioned for ionic polymers.

An SEC of a 3350 MW polystyrene standard showed a Ð of 1.07 and 1.13 for the new

and old columns respectively, so there is a slight broadening of the older columns. In

Figure 5.18, all the SEC curves show single, symmetric peaks and these peaks

systematically shift to the lower retention time as the polymerizations proceed. All

these observations in Figure 5.16 to Figure 5.18 are expected for a controlled free

radical polymerization.

148

0.5 NMR data 0.4 Linear fit ) t 0.3 /[M] 0 0.2 ln([M] 0.1

0.0 0 1 2 3 4 5 6 7

Time (h)

Figure 5.16 Pseudo first-order kinetic plot for the RAFT bulk polymerization of styrene at 120 °C with RAFT-PBu3. Target molecular weight is 25 kDa. The solid line is a linear fit to the data.

149

12 2.0 SEC Mn Mn, theory D 1.8 8 1.6 D (kg/mol)

n 1.4

M 4 1.2

0 1.0 0.1 0.2 0.3 0.4

Monomer Conversion

Figure 5.17 Plot of Mn (SEC) and PDI versus monomer conversion for the RAFT bulk polymerization of styrene at 120 °C with RAFT-PBu3. Target molecular weight is 25 kDa.

150

1 hr 1.0 2 hr 3 hr 4 hr 5 hr 6 hr 0.5

Normalized RI Intensity 0.0 14 15 16 17 18 19 20

Elution Time (min)

Figure 5.18 SEC traces as a function of polymerization time for the RAFT bulk polymerization of styrene at 120 °C with RAFT-PBu3. Target molecular weight is 25 kDa.

For the bulk styrene polymerization mediated by RAFT-PPh3, however, the

polymerization showed retardation at the beginning of the polymerization and an

increased rate after 4 hours, as shown in Figure 5.19. It has been discussed that the retardation phenomena in RAFT polymerizations are due to slow fragmentation of adduct intermediate radicals, slow reinitiation of the fragmentation radicals, or the reversible/irreversible cross-terminations.117,214,215,216,217, 218 Considering that RAFT-

194 PBu3 and previously studied quaternary ammonium RAFT agents have the same

structure as RAFT-PPh3 except the R-group, the retardation in the present case was likely related to the R-group, such as poor reinitiation of R-group in the pre- equilibrium stage or intermediate radicals were stabilized due to the 151

triphenylphosphonium groups. In Figure 5.20, despite the departure from the theoretical prediction, the molecular weights of the polystyrene grow linearly as the

monomer conversion increases, which implies the RAFT polymerization is still

somewhat controlled. The Ð keeps narrowing as the monomer conversion increases

and is lowered to 1.29 when the monomer conversion is 35%. The SEC curves in

Figure 5.21 show single peaks for all the polymerization times. Although the SEC

curves of the low molecular weight samples overlap with the solvent peak after 20

min, the systematic shift of the peaks to the lower retention time indicates the built-up

of the molecular weights as the polymerization proceeds.

152

0.5 NMR data 0.4 ) t 0.3 /[M] 0 0.2

ln([M] 0.1 0.0

0 2 4 6 8

Time (h)

Figure 5.19 Pseudo first-order kinetic plot for the RAFT bulk polymerization of styrene at 120 °C with RAFT-PPh3. Target molecular weight is 25 kDa.

153

12 3.0 SEC Mn Mn, theory D 2.5 8

2.0 D (kg/mol) n

M 4 1.5

0 1.0 0.1 0.2 0.3 0.4

Monomer Conversion

Figure 5.20 Plot of Mn (SEC) and PDI versus monomer conversion for the RAFT bulk polymerization of styrene at 120 °C with RAFT-PPh3. Target molecular weight is 25 kDa.

154

1 hr 1.0 2 hr 3 hr 4 hr 5 hr 6 hr 0.5 8 hr

Normalized RI Intensity 0.0 14 15 16 17 18 19 20

Elution Time (min)

Figure 5.21 SEC traces as a function of polymerization time for the RAFT bulk polymerization of styrene at 120 °C with RAFT-PPh3. Target molecular weight is 25 kDa.

Polystyrene obtained using both RAFT agents, were first characterized via

thin layer chromatography (TLC) to qualitatively examine the end-group fidelity.

When a nonpolar solvent, e.g. toluene, was used, the ion-containing polymers exhibited much lower Rf values compared to nonionic polymers. As shown in Figure

5.22, both polystyrene samples exhibit high end-functionality, since the fraction of

Rf=0 is much larger than that of Rf=1. This is in contrast to the TLC result of the

polystyrene based on quaternary ammonium-containing RAFT agents, for which a large fraction of cationic groups were degraded during the polymerization at 120

oC.194

155

(a) (b) (c)

Figure 5.22 TLC test of polystyrene obtained from bulk polymerization using (a) o RAFT-PBu3, (b) RAFT-PPh3, and (c) BDTC at 120 C.

Both crude polystyrenes prepared with the two phosphonium-containing

RAFT agents were further purified via column chromatography by eluting toluene

1 and then CHCl3/acetone/methanol (1:1:0.1). The H NMR spectra of the crude

polystyrene, purified polystyrene and the toluene fraction from the polymerization using RAFT-PBu3 are shown in Figure 5.23. Compared to the crude polystyrene, the

CHCl3/acetone/methanol fraction shows >99% end-functionality while the toluene

fraction lacks the peaks from the benzyl phosphonium groups. The 1H spectra of the

polystyrene based on RAFT-PPh3 is shown in Figure 5.24. The crude PS-PPh3

exhibited high end functionality while column purification further enhanced the end

functionality. It should be noted that the chemical shift and the integral area of the 156

methylene of the benzyl group in the crude PS-PPh3 are both expected for the benzyl phosphonium groups “inherited” from RAFT-PPh3. This is consistent with the benzyltriphenylphosphonium functional group being retained during the polymerization. Therefore, though the styrene polymerization via RAFT-PPh3 exhibited retardation, polymers with PDI<1.3 with high end functionality was obtained at higher monomer conversion.

157

c a P e b Br S S n S d

1.ESP (a) a,e

b c d

1.95 1.83 5.59 12.00 2.ESP(b) a,e

b

c d

2.13 2.09 7.19 12.00 3.ESP(c) e

d

6 4 2 Chemical Shift (ppm)

Figure 5.23 1H NMR spectra of a) crude polystyrene from styrene bulk o polymerization using RAFT-PBu3 at 120 C, b) the purified polystyrene (CHCl3/acetone/methanol fraction), c) toluene fraction.

158

b a P d Br S S n c S

4.ESP (a) a d

b c

5.ESP 12.80 1.84 1.84 3.00 (b) a d

b c

16.63 2.21 2.11 3.00

6.ESP (c) d

c

6 4 2 Chemical Shift (ppm)

Figure 5.24 1H NMR spectra of a) crude polystyrene from styrene bulk polymerization using RAFT-PPh3 at 120 oC, b) the purified polystyrene (CHCl3/acetone/methanol fraction), c) toluene fraction.

159

An ion-exchange experiment was performed to visualize the presence of the cationic end groups in both PS-PBu3 and PS-PPh3. D&C Green 5 was used as the dye since it contains sodium sulfonate groups that can ion exchange with the triphenylphosphonium bromide groups presented in the obtained end-functional polystyrene. As shown in Figure 5.25, vial #1 shows that D&C Green 5 partitions to aqueous phase while vial #2 and #4 show that both PS-PBu3 and PS-PPh3 partitions to toluene phase. However, when the dye was mixed with the polymer solution, ion exchange occurred and the dye was extracted into the toluene phase, as implied by the color change of the toluene layer from yellow, which is a characteristic color of polymers obtained from RAFT polymerization, to blue-green, which is a characteristic color of the dye. In addition, the color of the aqueous phase of via #3 and #5 was colorless, which also indicated the dye disappeared from the aqueous phase.

160

Figure 5.25 The ion exchange test that visualizes the cationic functionality. The vials were vortexed to facilitate ion exchange and then placed in the hood for one day. The contents of each vial before mixing were as follows: Vial #1: toluene (top layer) & water+D&C Green 5 (bottom layer); Vial #2: toluene+PS-PBu3 (top layer) & water (bottom layer); Vial #3: toluene+PS-PBu3 (top layer) & water+ D&C Green 5 (bottom layer); Vial #4: toluene+PS-PPh3 (top layer) & water (bottom layer); Vial #5: toluene+PS-PPh3 (top layer) & water+ D&C Green 5 (bottom layer)

161

5.4 Conclusions

Two quaternary phosphonium-containing RAFT agents were synthesized and examined in terms of thermal stability and capability to control bulk styrene polymerization. It was found that the thermal stability of cationic groups was significantly improved when quaternary phosphonium was employed compared to the ammonium analogues. This allows their use at higher temperatures, such as the thermal RAFT polymerization of PS as shown.

5.5 Acknowledgements

This material is based upon work supported 5 by the National Science

Foundation under Grant No. CHE-1012237 (KAC), DMR-1309853 (RAW) and the financial support from Hubei Provincial Department of Education (QT).

162

CHAPTER XI

SULFONATION DISTRIBUTION IN SULFONATED POLYSTYRENE

IONOMERS MEASURED BY MALDI-ToF MS*

6.1 Introduction

Ionomers are predominantly hydrophobic polymers containing a small fraction of chemically bonded ionic groups, usually < 15 mol%.92 Over the past few decades,

lightly sulfonated polystyrene (SPS) has served as a model ionomer system for the

study of melt rheology,95 solution behavior,94 morphology,93, 219 , 220

dynamics,165,221,222,223 and wetting behavior224,225 of ionomers. In addition to its use as

a model ionomer system, SPS has been used in applications such as adhesives,226

drilling fluid, 227 compatibilizing agents for polymer blends, 228 golf balls, 229 fluid

viscosification,230 organogels,231 and propellants.232

SPS ionomers are most commonly synthesized following the procedure

developed by Makowski et al.,233 which involves a homogeneous solution sulfonation using acetyl sulfate as the sulfonating agent. The sulfonation level, which is usually defined in terms of mol% sulfonation (i.e. the average number of sulfonate groups in

100 styrene units) can be conveniently determined by elemental analysis, acid-base

titration or 1H NMR.234

The Makowski sulfonation method is considered to proceed randomly along

the chain, primarily at the para-position of the phenyl ring, and one would expect that

there is a inhomogeneous, but random distribution of sulfonate groups on the

163

polystyrene chains. 235 That distribution has been estimated with a binomial distribution function:95,224,235

! ( ) = (1 )( ) ( )! ! (1) 푁 푥 푁−푥 푃 푥 푁−푥 푥 푝 − 푝 where P( ) is the probability that for an average sulfonation level p, a chain with N repeat units푥 have x sulfonate groups. However, no experimental studies have confirmed this sulfonation distribution for solution SPS. An extensive solid state

NMR investigation by VanderHart et al.235 of a low molecular weight SPS, N = 38 and p = 0.025, where Equation (1) predicts that ~40% of the chains should be unsulfonated, failed to detect phase separation of unsulfonated chains.

The quantification of the ion-distribution in SPS ionomers remains an open question in the field of ionomers, and a rather important one given the large number of research groups that use that material in their research. One problem with SPS as a model ionomer system is the inhomogeneous ion distribution at low sulfonation levels.

The interesting properties of SPS evolve from intermolecular association of the sulfonate groups, which provides a network structure dominated by the formation of ionic nanodomains. The influence of the ion-distribution on the microstructure is unknown, but may be important. For example, at low sulfonation levels, there may be a significant fraction of chains that are completely unsulfonated or have only one or two sulfonate groups. Unsulfonated and monosubstituted chains are ineffective for carrying load in a physically crosslinked ionomer network (see Figure 6.1). At best, multiple associations of monosubstituted chains will produce a micelle-like structure or dangling branches from a multiply sulfonated chain. Disubstituted chains can produce chain-extension by simple associations or ionic bonding between sulfonate groups. If multiple associations occur, such as when the sulfonate groups are incorporated into nanodomain aggregates (ionic clusters), they can produce a network 164

structure. Even chains with only two sulfonate groups can participate in such a network if both are incorporated into different clusters. Chains containing three or more sulfonate groups can easily form network structures by simple association.

Unsulfonated (a) polymer chains Ionic group

(b)

(c) Micelle Ionic pair

Ionic aggregates

Figure 6.1 Schematics of chain structures with associative ionic groups. (a) Chains with one or two sulfonate groups can result in chain extension. Chains without ionic functionality are inactive. (b) Chains with three or more sulfonate groups may form a network structure. Chains with two sulfonate groups can participate in a network if both sulfonate groups are incorporated into different ionic clusters. (c) Multiple associations of monofunctional chains will form a micelle-like structure or dangling branches from a multiple sulfonated chain.

165

Matrix-assisted laser desorption ionization time-of-flight mass spectrometry

(MALDI-ToF MS)236,237,238 provides an opportunity to reveal the ion heterogeneity in

SPS. MALDI-ToF MS can yield quantitative information on individual chains, which

makes it useful in characterizing mixtures of polymer systems with complex chemical

structures. 239 , 240 For example, MALDI-ToF MS has been successfully applied,

without the use of calibrants, for the quantitation of polystyrene end groups

introduced by reacting poly(styryl) lithium with ethylene oxide; 241 the degree of

ethylene oxide oligomerization determined by MALDI-ToF MS was in excellent agreement with NMR results.241 In the present work, the sulfonation level and the

sulfonation distribution of SPS ionomers were measured by MALDI-ToF MS. The

sulfonation levels obtained by MALDI ToF-MS were compared to the values deduced

by conventional acid-base titration which normally shows very good agreement with

elemental analysis.221,242 The measured sulfonate distributions were also compared to

the predictions of Equation (1). To our knowledge, this is the first report of

experimentally measuring the sulfonation distribution of randomly sulfonated SPS.

6.2 Experimental section

6.2.1 Materials

A narrow molecular weight distribution polystyrene (PS) with a weight

average molecular weight of 4000 Da and a polydispersity index of < 1.06 was

obtained from Pressure Chemicals, Inc. All other reagents and solvents used were

obtained from Fisher Scientific. Sulfonated polystyrene (SPS) was prepared by

reacting a ~5% solution of PS in 1,2-dicholorethane (reagent grade) at 50°C with

acetyl sulfate according to the procedure of Makowski et al.233 The acetyl sulfate was

prepared by the reaction of concentrated sulfuric acid (reagent grade) and a 60%

166

excess of acetic anhydride (reagent grade) at 0°C. The excess acetic anhydride was

used to scavenge any water present. The resulting sulfonic acid derivative of SPS, i.e.

HSPS, was precipitated with methanol (99.9%), washed several times with fresh

methanol and dried under vacuum. HSPS was neutralized in a toluene/MeOH (9:1 v/v)

mixture with 50% excess of a methanol solution of the selected alkali hydroxide or

silver nitrate and the resulting salt of HSPS was recovered by steam stripping, filtered,

washed and thoroughly dried in vacuum. This procedure produces 100%

neutralization of the ionomer, as has been demonstrated by the disappearance of the

infrared absorbance bands specific for the sulfonic acid group. 243 The sulfonated

polystyrene (SPS) ionomers prepared are represented as MSPSp, where M is the

metal cation and p is the average sulfonation level. LiSPS, NaSPS, KSPS, RbSPS, and

AgSPS with sulfonation levels of 2.5, 3.7, and 6.7 mol% were prepared.

6.2.2 MALDI-ToF MS Analysis

Matrix-assisted laser desorption ionization time-of-flight (MALDI-ToF) mass spectra were acquired on a Bruker UltraFlex-III ToF/ToF mass spectrometer (Bruker

Daltronics, Inc., Billerica, MA) equipped with a Nd:YAG laser (at 355 nm). All spectra were measured in positive reflectron mode. The instrument was calibrated prior to each measurement with poly(methyl methacrylate) (PMMA) as a positive external standard. For SPS samples, the following MALDI matrices were used: 1,8- dihydroxy-9,10-dihydroanthracen-9-one (dithranol, DIT), 2,5-dihydroxybenzoic acid

(DHB), and trans-2-(3-(4-tert-butylphenyl)-2-methyl-2-propenyliedene)malononitrile

(DCTB). Each matrix was dissolved in THF at a concentration of 20 mg/mL. MSPSp samples were dissolved in THF at a concentration of 10 mg/mL. Sodium trifluoroacetate, silver trifluoroacetate, lithium trifluoroacetate, or potassium

167

trifluoroacetate were tested as cationizing agents to aid in the ionization process. The cationizing salts were dissolved in THF at a concentration of 10 mg/mL. Several

MALDI sample preparation techniques were attempted to optimize experimental conditions: 1) dry-droplet method244 where matrix, sample, and salt solutions were mixed in a ratio of 10:5:1 (v/v) respectively; 2) a solvent-free approach245,246,247 where the matrix, sample, and salt are mixed mechanically, and a small amount of the mixture is deposited onto the target plate; 3) a sandwich style approach 248 where matrix and salt solutions were mixed together (10:1 v/v) then a 1 μL aliquot was deposited and allowed to evaporate which was followed by a 1 μL aliquot of sample then another layer of matrix/salt. In addition to the above three methods, MALDI samples just consisting of matrix and MSPSp were prepared (10:5 v/v) via the dry- droplet method. All MALDI-ToF MS samples were deposited in wells of a 384-well ground-steel target plate. After evaporation of the solvent, the plate was inserted into the MALDI source. The attenuation of the Nd:YAG laser was adjusted to minimize unwanted polymer fragmentation and to maximize the sensitivity. All data were acquired using Bruker FlexControl (version 3.3) software with 10,000 laser shots, no delayed extraction, and ions were detected in an m/z range of 1,500-7,000. However, a m/z range of 1,500-5,500 was selected for data analysis due to the severe widening of the ionic peaks and low intensities above 5,500 Da.

MALDI-ToF MS measurements were repeated three times for each sample to verify their reproducibility. The peak heights of the most abundant isotope peaks were used to represent the amount of the polymer. Peak heights rather than the peak areas were used due to the limited resolution at high m/z range. Baseline corrections of the obtained mass spectra were done by the Bruker FlexAnalysis (version 3.3) software.

168

6.3 Results and discussion

Due to the random nature of the sulfonation reaction, SPS is essentially a

mixture of chains with varying sulfonate functionality, including unsulfonated

polystyrene. Therefore, differences in ionization efficiencies between all the

components in the mixture need to be accounted for and the MALDI-ToF MS

experimental conditions need to be optimized. Three SPS polymers were prepared

from a narrow molecular weight distribution polystyrene (Pressure Chemical Co.; Mw

= 4,000 Da, polydispersity index = 1.06) by sulfonation according to the procedure of

Makowski et al.233 The sulfonation levels (p) were 2.5, 3.7, and 6.5 mol% (by

titration). The samples are denoted as MSPSp, where M is the metal cation and p is the average sulfonation level.

Three MALDI matrices, viz. 1,8-dihydroxy-9,10-dihydroanthracen-9-one

(dithranol, DIT), trans-2-(3-(4-tert-butylphenyl)-2-methyl-2-propenyliedene)malo- nonitrile (DCTB), and 2,5-dihydroxybenzoic acid (DHB), and five different monovalent metal cations (Li, Na, K, Rb, and Ag) were examined in order to obtain the most effective experimental conditions for MALDI-ToF MS analysis. When DIT was used as the matrix, the mass spectra severely biased to mono-sulfonated polystyrene. Compared to DIT, DCTB gave higher signals of non-sulfonated polystyrene but the mono-sulfonated species still dominated the spectra. This issue was, to a large extent, addressed when DHB was used as the matrix (Figure S1).

When alkali cations other than lithium were used, either the ion peaks overlapped or the unsulfonated chains were not detected by MALDI-ToF MS. AgSPS failed to provide good quality mass spectra. No extra cationizing agent was added since the one tested, lithium trifluoroacetate (LiTFA), was found to severely deteriorate the

169

spectra, most likely due to high excess salt concentration that results in poor signal-to- noise ratio.

(a) S1 S0 DCTB

S2

DIT

DHB

1780 1800 1820 1840 m/z

(b) S1 DCTB S0

S3 S2

DIT

DHB

2400 2420 2440 2460 m/z

Figure 6.2 Expanded views of mass spectra acquired using different matrices; (a) lower MW region; (b) higher MW region. "Sx" refers to polystyrene chains with x sulfonate groups.

170

Therefore, optimization studies indicated that LiSPS and DHB without a

cationizing agent, using the dry-droplet sample preparation method,244 provided the

best quality MALDI-ToF MS spectra for analysis of the sulfonation distribution.

Representative MALDI-ToF MS spectra of LiSPS2.5, LiSPS3.7, and LiSPS6.5 are shown in Figure 6.3 to Figure 6.5. The spectra for LiSPS3.7 and LiSPS6.5 were

qualitatively similar to the LiSPS2.5 spectrum.

2000 3000 4000 5000 m/z

Figure 6.3 MALDI-ToF MS spectrum of LiSPS2.5.

171

2000 3000 4000 5000 6000 m/z

Figure 6.4 MALDI-ToF MS spectrum of LiSPS3.7 (no cationization salt).

2000 3000 4000 5000 6000 m/z

Figure 6.5 MALDI-ToF MS spectrum of LiSPS6.5 (no cationization salt).

172

[S1+Li]+ LiSPS6.5 [S2+Li]+ + [S3+Li]+ [S0+Li] [S0+Na]+

LiSPS3.7

LiSPS2.5

2400 2420 2440 2460 2480 m/z

Figure 6.6 Expanded view of mass spectra of LiSPS2.5, LiSPS3.7, and LiSPS6.5. "Sx" refers to polystyrene chains with x sulfonate groups. Note that the degree of polymerization of S3, S2, S1, and S0 is 20, 21, 22, and 23, respectively

The expanded mass spectra of these three samples, Figure 6.6, demonstrate the

variation in the sulfonation distribution for the three polymers. The major peaks

labeled as "Sx" indicate polystyrene chains with x sulfonate groups. It should be noted

that all ionic peaks detected are Li+ adduct peaks, except for the minor component

peak labeled as [S0 + Na]+, which is the sodium adduct of unsulfonated chains. Peaks

from sodium ions appear due to the trace sodium ions present as impurities in sample holder and glassware used.236,249 As the sulfonation level increased, the distribution

shifts towards higher sulfonation levels.

173

The average sulfonation levels of the three ionomers were calculated from the

mass spectra, and the results are compared with the titration results in Table 1. The

sulfonation levels determined by MALDI-ToF MS for LiSPS3.7 and LiSPS6.5 are in

good agreement with the respective titration values, though the MALDI-ToF MS

value for LiSPS6.5 is about 4% higher than the titration value. The sulfonation level

of LiSPS2.5 given by MALDI-ToF MS is ~25% higher than the titration value. With

the MALDI-ToF MS experimental conditions applied, two effects are likely to exist.

First, the sulfonated species have significantly higher ionization efficiency than the

unsulfonated species. Second, the ionization efficiency continues to increase as the

sulfonate functionality increases, albeit at a significantly smaller extent compared to

moving from the unsulfonated to monosulfonated polystyrene.

Table 6. 1 Sulfonation level determined by MALDI-ToF MS. Sample MALDI-ToF MS Titration

LiSPS2.5 3.12 ± 0.20 2.47 ± 0.11

LiSPS3.7 3.74 ± 0.31 3.70 ± 0.10

LiSPS6.5 6.82 ± 0.06 6.57 ± 0.12

The disparity in the sulfonation level by the two techniques for LiSPS2.5 is

attributed to the first effect. As will be discussed later in this paper, the molar ratios of

the unsulfonated species calculated by MALDI-ToF MS for LiSPS2.5, LiSPS3.7, and

LiSPS6.5 were about 33 mol%, 25 mol%, and 9 mol%, respectively. These values agree reasonably well with the binomial distribution values calculated from Equation

(1) at 38 mol%, 24 mol%, and 8 mol%, respectively. The first effect becomes prominent since the molar ratio of unsulfonated species is much higher in LiSPS2.5.

174

Consequently, MALDI-ToF MS tends to underestimate the amount of unsulfonated species, so that the MALDI-ToF MS sulfonation level for LiSPS2.5 is overestimated.

For LiSPS6.5, the MALDI-ToF MS value is slightly higher than the titration value due to the second effect.

The above-mentioned arguments can be further supported by the variation of

MALDI-ToF MS sulfonation levels with degree of polymerization, N, as shown in

Figure 6.7. Since MALDI-ToF MS provides quantitative information on individual chains, it is possible to determine the effect of chain length on the sulfonation distribution. For LiSPS2.5 and LiSPS3.7, the sulfonation level slightly decreases and eventually plateaus with increasing molecular weight of the chains, giving rise to average sulfonation levels of 3.1 mol% and 3.7 mol%, respectively. The independence of sulfonation level on chain length confirms that the sulfonation process is random.

175

8

6

4

2 LiSPS2.5 LiSPS3.7 Sulfonation Level (mol%) Level Sulfonation LiSPS6.5 0 15 20 25 30 35 40 45

Degree of Polymerization

Figure 6.7 Sulfonation level versus degree of polymerization for N = 15-47.

For chains with a lower degree of polymerization there is a higher probability for completely unsulfonated chains. The considerable ionization efficiency difference between the unsulfonated and sulfonated species will therefore over-predict the degree of sulfonation at low molecular weight. For LiSPS6.5 on the other hand, except for the lowest molecular weight chains, which also had the largest experimental error, the sulfonation level was reasonably constant with molecular weight and near the average value of ~6.8 mol%. The errors in the MALDI ToF-MS measurements for LiSPS6.5 were relatively small (< 1%) compared with the two lower sulfonation levels where the error for the average sulfonation level was ~6-8%

(see Table 6. 1). This is consistent with the conclusion that the larger fractions of unsulfonated chains for the lower average sulfonation levels result in an overestimation of the sulfonation level. The fraction of unsulfonated chains in

176

LiSPS6.5 predicted by Equation (1) is about 30% of that for LiSPS3.7 and 20% of

that for LiSPS2.5.

The sulfonation distributions for the three ionomers calculated from the

MALDI-ToF MS data are compared with the predictions from Equation (1) in Figure

6.8. For each ionomer, the concentration of monofunctional species (x = 1) is

overestimated and the multifunctional species (x ≥ 2) are generally underestimated.

This is most likely due to mass discrimination effects in the MALDI ToF-MS data, i.e. the low molecular weight species have stronger desorption/ionization efficiency than the higher molecular weight species (with the same degree of sulfonation).250 Multi-

sulfonated species mainly appear in the high molecular weight region where the

signals are partially suppressed due to mass discrimination. However, because of the

relatively low polydispersity of the parent polystyrene (Mw = 4,150 Da; polydispersity

index = 1.08 from MALDI ToF-MS and Mw = 4,000 Da; PDI = 1.07 from GPC), the

mass discrimination effects were not that severe. The overestimation of the

monofunctional species can be explained in the similar way. For unsulfonated species,

both mass discrimination and ionization discrimination occur and offset each other,

resulting in better agreement with the theoretical prediction for random sulfonation.

177

(a) (b) .5 .5 MALDI values MALDI values Binomial Distr. Binomial Distr. .4 .4

.3 .3 P(x) P(x) .2 .2

.1 .1

0.0 0.0 0 1 2 3 4 0 1 2 3 4 x x (c) .4 MALDI values Binomial Distr. .3

.2 P(x)

.1

0.0 0 1 2 3 4 5 6 x

Figure 6.8 Sulfonation distribution measured by MALDI-TOF MS and binomial distribution predictions (Equation (1) for: (a) LiSPS2.5; (b) LiSPS3.7; and (c) LiSPS6.5). The dashed lines have no physical significance. They are only included to make it clear that the points denoting the predictions of the binomial distribution, equation (1), are discrete values (the points connected by the lines).

Equation (1) ignores any effect that the molecular weight distribution may

have on the sulfonation level, which based on the discussion above can be significant

for a low molecular weight sample, especially for the low molecular weight fraction

of a low average molecular weight sample. That may also increase the disparity

between the experimental and theoretical sulfonation distributions. For the parent

polystyrene used (Mw = 4,000 g/mol), the number average degree of polymerization is

N = 38. The MALDI-ToF MS data for N = 38 were isolated and analyzed to 178

determine their sulfonation distributions. Such an analysis removes the effect of mass discrimination on the spectra since only a single N is considered. Figure 6.9 shows the results for each of the three ionomers, and in this case, with the exception of the datum point for the unsulfonated fraction of LiSPS2.5, the MALDI-ToF MS and theoretical predictions are in good agreement. This result again supports the random nature of the sulfonation reaction. The unsulfonated fraction in the LiSPS2.5 is expected to be the most problematic one to measure correctly by MALDI-ToF MS, because of the ionization discrimination towards unsulfonated species as explained earlier.

.5 LiSPS2.5 .4 LiSPS3.7 LiSPS6.5 .3 P(x) .2

.1

0.0 0 1 2 3 4 5 6

x

Figure 6.9 Sulfonation distribution from the N = 38 fraction of LiSPS2.5, LiSPS3.7, and LiSPS6.5. The dashed lines have no physical significance. They are only included to make it clear that the points denoting the predictions of the binomial distribution, equation (1), are discrete values (the points connected by the lines).

179

6.4 Conclusion

In summary, the sulfonation distribution of an SPS ionomer was, for the first

time, measured experimentally by MALDI-ToF MS. Deviation of the MALDI-ToF

MS results from a random sulfonation prediction decreased for ionomer fractions with

increasing molecular weight and with increasing sulfonation level for the low

molecular weight ionomers discussed herein. The results from the MALDI-ToF MS analysis deviated from a theoretical random distribution due to errors associated with mass discrimination effects from the molecular weight distribution and ionization discrimination for chains without sulfonation. The experimental and theoretical distributions for the number average molecular weight fraction (N = 38) of the low molecular weight ionomers were in good agreement. That result, as well as the independence of the sulfonation level on the chain length, indicates that the solution sulfonation procedure described by Makowski et al.233 is indeed random, which

implies the validity of using a binomial distribution to describe the sulfonation

distribution. Most work on SPS ionomers has involved polymers with much higher

molecular weights, where it is unlikely that the mass and ionization discrimination

effects described in this paper will be as noticeable as seen herein for a low molecular

weight ionomer. Thus, a binomial distribution appears to be reasonable for describing

random ionomers, especially when the polymers are not readily or quantitatively

analyzable by MALDI-ToF MS.

6.5 Acknowledgements

The authors gratefully acknowledge support from the National Science

Foundation (grants CBET-1066517 to RAW and CHE-1012636 to CW).

180

CHAPTER VII

SUMMARY

This dissertation mainly deals with the preparation of supramolecular block copolymers based on ionic interactions. In the first part, a supramolecular multiblock

polystyrene-polyisobutylene copolymer (SMBCP) were synthesized via proton

transfer between the end groups of α,ω-sulfonated polystyrene and α,ω-amino polyisobutylene. The morphology, phase behavior, viscoelastic behavior and mechanical property of the obtained SMBCP were studied and discussed. In the second part, cationic RAFT agents carrying either quaternary ammonium group or quaternary phosphonium group were synthesized via a facile, high yield method.

These cationic RAFT agents can be used in RAFT polymerization to obtain hemi- telechelic cationomers, which are the cationic building blocks in constructing supramolecular ionic block copolymers. In the third part, matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-ToF MS) was used to experimentally quantify the sulfonation distribution of sulfonated polystyrene (SPS) synthesized by Makowski method, a homogeneous solution sulfonation reaction. Our work for the first time experimentally determined the sulfonation distribution of three low molecular weight SPS with low sulfonation level. More importantly, the results revealed that the Makowski method is indeed a random sulfonation reaction, which

181

implied the validity of using a binomial distribution to describe the sulfonation distribution of random ionomers that were not analyzable experimentally.

In the first part, a supramolecular polystyrene-polyisobutylene multiblock

copolymer was prepared based on ionic interactions. To obtain multiblock structure,

telechelic α,ω-sulfonated polystyrene (HO3S-PS-SO3H) and telechelic α,ω-amino polyisobutylene (H2N-PIB-NH2) were first synthesized. HO3S-PS-SO3H was prepared

by RAFT polymerization using a difunctional RAFT agent, didodecyl-1,2-phenylene-

bis(methylene)-bistrithiocarbonate, and then oxidizing the trithiocarbonate group to

sulfonic acid group with mCPBA. The resulting HO3S-PS-SO3H has a molecular

weight of 6500 Da and end-group functionality of 95%. H2N-PIB-NH2 was prepared

by living carbocationic polymerization of isobutylene (IB) and postpolymerization modification of end-groups. A difunctional initiator, 1,3-di(1-chloro-1-methylethyl)-

5-tert-butylbenzene, was used in the IB polymerization. (3-bromopropoxy)benzene was used as a quenching agent to terminate the polymerization, which resulted in a telechelic primary bromide terminated PIB (Br-PIB-Br). The primary bromide was converted to primary amine groups by Gabriel synthesis and the quantitative end group transformation was confirmed by 1H NMR.

The SMBCP was prepared by mixing equimolar HO3S-PS-SO3H and H2N-

PIB-NH2 in toluene and then solution casting to a Teflon dish. Upon the complete

drying of the film, SMBCP was obtained as a flexible and transparent film, which

indicated the macrophase separation between PS and PIB was effectively suppressed

by the formation of the ionic bonds. FTIR and 1H NMR also proved the ionic bonds formed via proton transfer between sulfonic acid and primary amine. The DSC curve of SMBCP showed two Tg that were corresponding to PS-rich phase and PIB-rich

phase. This two Tg were closer compared to the two Tg from neat oligomers. Both

182

features were consistent with the formation of block copolymer structures with nanodomains.

The direct evidence of block copolymer formation and morphology was demonstrated by small angle X-ray scattering (SAXS). The as-cast SMBCP exhibited a lamellar morphology with a domain spacing of 14.4 nm, which was much smaller compared to the unperturbed domain spacing calculated for the diblock PS-PIB copolymer with same molecular weights (22.4 nm). This indicated the formation of a multiblock structure.

The phase behavior of the SMBCP was studied by temperature dependent

SAXS (T-SAXS) and frequency sweep melt rheology. In the frequency sweep, the slope of both moduli versus frequency was 0.5, which is indicative of lamellar structure. From 190 oC to 210 oC, the slope changed from 0.5 to 1, which implies the order-disorder transition. This result is consistent with T-SAXS, which showed that the first order peak was greatly broadened and weakened from 190 oC to 210 oC.

Surprisingly, the domain spacing was observed to increase ca. 14% in the cooling process of T-SAXS study, which was attributed to the thermal degradation of the sulfonate ammonium ion pairs, as confirmed by FTIR and 1H NMR.

The linear viscoelastic behavior of the SMBCP was also studied. Three transitioins, viz. glass transition of PIB-rich phase, glass transition of PS-rich phase, and order-disorder transition, were all observed. A rubbery plateau was seen between the Tg of PIB-rich phase and PS-rich phase, which was due to the physical network formed by glassy PS nanodomains.

The tensile stress-strain curve of the SMBCP exhibited the characteristics of a soft plastics with a modulus of 90 MPa, a yield point of 4% strain, failure at 7% strain

183

and a toughness of 15 MJ/m3. The elongation at break is small, which is due to the

stretched polymer chains and the weaker ionic bonds compared to covalent bonds.

The nonlinear rheological behavior of the SMBCP was studied by three strain

sweep-time sweep cycles. The storage modulus of SMBCP decreased at large strain in

the strain sweep and then recovered quickly to about 90% of its original value in the time sweep. The decrease of G’ could be due to the disruption of the ionic bonds under high stress, or the shear alignment induced by large strain amplitude oscillatory flow.

In the second part, two types of cationic trithiocarbonate RAFT agents were synthesized and further applied to prepare hemi-telechelic cationomers. At first, quaternary ammonium-containing trithiocarbonate RAFT agents (RAFT-NR3) were

synthesized by preparing mono-substituted 4-(bromomethyl)-N,N,N-trialkyl benzyl

ammonium bromide compounds, which were further reacted with the alkyl

trithiocarbonate anion to directly yield the trithiocarbonate RAFT agents. The reaction

conditions for making three type of 4-(bromomethyl)-N,N,N-trialkyl benzyl

ammonium bromide compounds (alkyl=Me, Et, and Bu) were optimized. Based on

the successful synthesis of RAFT-NR3, they were used to control bulk RAFT

o polymerization of styrene at 120 C. Although all three types of RAFT-NR3 were

capable to control the styrene polymerization, it was found the quaternary ammonium

groups exhibited severe thermal degradation after polymerization, which significantly

reduced the end-group functionality of the resulting polymers. Tempeature ramp TGA

and isothermal TGA also confirmed the moderate thermal stability of RAFT-NR3. To

o overcome this issue for RAFT-NR3, a much lower temperature, i.e. 65 C, was

selected as polymerization temperature, which effectively addressed the thermal

184

stability issue. By using RAFT-NR3, different hemi-telechelic cationomers can be produced with high end-group functionality.

On the other hand, quaternary phosphonium-containing trithiocarbonate RAFT agents (RAFT-PR3) were also synthesized to explore their usage in high temperature

polymerization, since quaternary phosphonium groups were well-known to be much

more thermally stable compared to their ammonium analogues. Two RAFT-PR3

(R=Bu and Ph) were synthesized via similar synthetic routes as RAFT-NR3. Both

temperature ramp TGA and isothermal TGA demonstrated that RAFT-PR3 had greatly improved thermal stabilities compared to RAFT-NR3 and were thermally stable at 120

o o C. RAFT-PBu3 was capable to control the bulk styrene polymerization at 120 C and

yield polystyrene with high end-group functionality before purification due to the excellent thermal stability of quaternary phosphonium groups. RAFT-PPh3, however,

showed relatively poor control capability due to the retardation in the initial stage of

polymerization and the departure of the molecular weight from the theoretical

prediction. Nevertheless, as the monomer conversion increased, the polystyrene obtained via RAFT-PPh3 exhibited low molecular weight dispersity and high end-

group functionality, which means the obtained polystyrene cationomers are still useful

polymers for constructing supramolecular structures and other applications requiring

quaternary phosphonium functionality.

In the third part, the sulfonation distribution and sulfonation level of lightly

sulfonated polystyrene (SPS) ionomers, which were prepared by a homogeneous solution sulfonation, were experimentally quantified by matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-ToF MS). Three SPS ionomers were prepared from a narrow molecular weight distribution polystyrene

(Pressure Chemical Co.; Mw = 4,000 Da, Đ = 1.06) by sulfonation according to the

185

procedure of Makowski et al. and subsequent neutralized to alkali or silver salt of SPS ionomers. The sulfonation level were 2.5, 3.7 and 6.5 by acid-base titration of the acid form of SPS ionomers. Different conditions for MALDI-ToF MS tests were perfomed and optimization studies showed lithium salt of SPS ionomer, i.e. LiSPS, and DHB without a cationizing agent provided the best quality of MALDI-ToF MS spectra without severe biased to monosulfonated species or unsulfonated species. The sulfonation levels for three SPS ionomers were calculated based on the MALDI-ToF

MS spectra and the results were compared to sulfonation level obtained by acid-base titration. The sulfonation distribution for the three SPS ionomers were also calculated from the MALDI-ToF MS data. It was found the molecular weight discrimination affected the sulfonation distribution. To avoid this effect, the data from SPS of degree of polymerization of 38 were selected and analyzed, which provided the sulfonation distribution that were very close to the theoretical binomial distribution. This indicated that the Makowski method is indeed random and the validity of using a binomial distribution to describe the sulfonation distribution.

186

REFERENCES

1 Lehn, J. M. Supramolecular Chemistry-Concepts and Perspectives; VCH: Weinheim, 1995.

2 Haino, T. Polym. J. 2013, 45, 363-383.

3 Yang, S. K.; Ambade, A. V.; Weck, M. Chem. Soc. Rev 2011, 40, 129-137.

4 Faul, C. F. J.; Antonietti, M. Adv. Mater. 2003, 15, 673-683.

5 Ruokolainen, J.; Makinen, R.; Torkkeli, M.; Makela, T.; Serimaa, R.; ten Brinke, G.; Ikkala, O. Science 1998, 280, 557-560.

6 Grady, B. P.; Cooper, S. L. In Ionomers: Synthesis, Structure, Properties and Application; Mauritz, K. A.; Wilkes, G. L.; Tant, M. R. (Eds.); Chapman & Hall: New York, 1997.

7 Zhang, L.; Brostowitz, N. R.; Cavicchi, K. A.; Weiss, R. A. Macromol. React. Eng. 2014, 8, 81-99.

8 Antonietti, M.; Conrad, J.; Thunemann, A. Macromolecules 1994, 27, 6007- 6011.

9 MacKnight, W. J.; Ponomarenko, E. A.; Tirrell, D. A. Acc. Chem. Res. 1998, 31, 781-788.

10 Hadjichristidis, N.; Pitsikalis, M.; Iatrou H. Adv. Polym. Sci. 2005, 189, 1-124.

11 Bates, F. S.; Fredrickson, G. H. Annu. Rev. Phys. Chem. 1990, 41, 525-527.

12 Hamley, I. W. The Physics of Block Copolymers; Oxford Univ Press: Oxford, 1999.

13 Albert, J. N. L.; Epps, T. H., III. Materials Today 2010, 13, 24-33.

14 Jain, S.; Bates, F. S. Science 2003, 300, 460-464.

15 Lodge, T. P.; Pudil, B.; Hanley, K. J. Macromolecules 2002, 35, 4707-4717.

187

16 Spontak, R. J.; Patel, N. P. Curr. Opin. Colloid Interface Sci. 2000, 5, 333-340.

17 Macosko, C. W.; Guegan, P.; Khandpur, A. K. Macromolecules 1996, 29, 5590-5598.

18 Kataoka, K.; Harada, A.; Nagasaki, Y. Adv. Drug Deliv. Rev. 2001, 47, 113- 131.

19 Kim, H.-C.; Park, S.-M.; Hinsberg, W. D. Chem. Rev. 2010, 110, 146-177.

20 Thurn-Albrecht, T.; Schotter, J.; Kastle, G. A.; Emley, N.; Shibauchi, T.; Krusin-Elbaum, L.; Guarini, K.; Black, C. T.; Tuominen, M. T.; Russell, T. P. Science 2000, 290, 2126-2129.

21 Park, M.; Harrison, C.; Chaikin, P. M.; Register, R. A.; Adamson, D. H. Science 1997, 276, 1401-1404.

22 Edrington, A. C.; Urbas, A. M.; DeRege, P.; Chen, C. X.; Swager, T. M.; Hadjichristidis, N.; Xenidou, M.; Fetters, L. J.; Joannopoulos, J. D.; Fink, Y.; Thomas, E. L. Adv. Mater. 2001, 13, 421-425.

23 Bates, F. S.; Fredrickson, G. H. Physics Today 1999, 52, 32-38.

24 Lodge, T. P. Polymer Chemistry; CRC Press: Boca Raton, 2007.

25 Leibler, L. Macromolecules 1980, 19, 1602-1617.

26 Fredrickson, G. H.; Bates, F. S. Annu. Rev. Mater. Sci. 1996, 26, 501-550.

27 Semenov, A. N. Soviet Phys. JETP 1985, 61, 733.

28 Matsen, M. W.; Schick, M. Phys. Rev. Lett. 1994, 72, 2660-2663.

29 Matsen, M. W.; Bates, F. S. Macromolecules 1996, 29, 1091-1096.

30 Khandpur, A. K.; Forster, S.; Bates, F. S.; Hamley, I. W.; Ryan, A. J.; Bras, W.; Almdal, K.; Mortensen, K. Macromolecules 1995, 28, 8796.

31 Roe, R. J. Methods of X-ray and Neutron Scattering in Polymer Science; Oxford University Press: New York, 2000.

32 Hamley, I. W.; Castelletto, V. Prog. Polym. Sci. 2004, 29, 909-948.

33 Rosedale, J. H.; Bates, F. S. Macromolecules 1990, 23, 2329-2338.

34 Koppi, K. A.; Tirrell, M.; Bates, F. S.; Almdal, K.; Mortensen, K. J. Rheol. 1994, 38, 999-1027.

188

35 Almdal, K.; Bates, F. S.; Mortensen, K. J. Chem. Phys. 1992, 96, 9122-9132.

36 Rosedale, J. H.; Bates, F. S.; Almdal, K.; Mortensen, K.; Wignall, G. D. Macromolecules 1995, 28, 1429-1443.

37 Kim, J. K.; Lee, H. H.; Gu, Q.-J.; Chang, T.; Jeong, Y. H. Macromolecules 1998, 31, 4045-4048.

38 Kossuth, M. B.; Morse, D. C.; Bates, F. S. J. Rheol. 1999, 43, 167-196.

39 Steed, J. W.; Turner, D. R.; Wallace, K. J. Core Concepts in Supramolecular Chemistry and Nanochemistry; John Wiley & Sons: West Sussex, 2007.

40 Steed, J. W.; Atwood, J. L. Supramolecular Chemistry, 2nd ed; John Wiley & Sons: West Sussex, 2009.

41 Ciferri, A. Macromolecular Rapid Communications 2002, 23, 511-529.

42 Ciferri, A. Supramolecular Polymers, 2nd ed; CRC Press: Boca Raton, 2005.

43 Lehn, J.-M. Angew. Chem. Int. Ed. 1990, 29, 1304-1319.

44 Brunsveld, L.; Folmer, B. J. B.; Meijer, E. W.; Sijbesma, R. P. Chem. Rev. 2001, 101, 4071-4097.

45 Sijbesma, R. P.; Beijer, F. H.; Brunsveld, L.; Folmer, B. J. B.; Hirschberg, J. H. K. Ky.; Lange, R. F. M.; Lowe, J. K. L., Meijer, E. W. Science, 1997, 278, 1601-1604.

46 Beijer, F. H.; Sijbesma, R. P.; Kooijman, H.; Spek, A. L.; Meijer, E. W. J. Am. Chem. Soc. 1998, 120, 6761-6769.

47 Folmer, B. J. B.; Sijbesma, R. P.; Versteegen, R. M.; van der Rijt, J. A. J.; Meijer, E. W. Adv. Mater. 2000, 12, 874-878.

48 Bosman, A. W.; Sijbesma, R. P.; Meijer, E. W. Mater. Today 2004, 7, 34-39.

49 Cordier, P.; Tournilhac, F.; Soulie-Ziakovic, C.; Leibler, L. Nature 2008, 451, 977-980.

50 Hentschel, J.; Kushner, A. M.; Ziller, J.; Guan, Z. B. Angew. Chem. Int. Ed. 2012, 51, 10561-10565.

51 Huh, J.; Park, H. J.; Kim, K. H.; Kim, K. H.; Park, C.; Jo, W. H. Adv. Mater. 2006, 18, 624-629.

52 Burnworth, M.; Tang. L.; Kumpfer, J. R.; Duncan, A. J.; Beyer, F. L.; Flore, G. L.; Rowan, S. J.; Weder, C. Nature 2011, 472, 334-338. 189

53 Gohy, J.-F.; Lohmeijer, B. G. G.; Varshney, S. K.; Decamps, B.; Leroy, E.; Boileau, S.; Schubert, U. S. Macromolecules 2002, 35, 9748-9755.

54 Yang, S. K.; Ambade, A. V.; Weck, M. J. Am. Chem. Soc. 2010, 132, 1637- 1645.

55 Park, T.; Zimmerman, S. C. J. Am. Chem. Soc. 2006, 128, 13986-13987.

56 Yang, X. W.; Hua, F. J.; Yamato, K.; Ruckenstein, E.; Gong, B.; Kim. W.; Ryu, C. Y. Angew. Chem. Int. Ed. 2004, 43, 6471-6474.

57 Binder, W. H.; Bernstorff, S.; Kluger, C.; Petraru, L.; Kunz, M. J. Adv. Mater. 2005, 17, 2824-2828.

58 Noro, A.; Tamura, A.; Wakao, S.; Takano, A.; Matsushita, Y. Macromolecules 2008, 41, 9277-9283.

59 Fustin, C.-A.; Lohmeijer, B. G. G.; Duwez, A.-S.; Jonas, A. M.; Schubert, U. S.; Gohy, J.-F. Adv. Mater. 2005, 17, 1162-1165.

60 Moughton, A. O.; O'Reilly, R. K. Macromol. Rapid. Commun. 2010, 31, 37- 52.

61 Huh, J.; Jung, J. Y.; Lee, J. U.; Cho, H.; Park, S.; Park, C.; Jo, W. H. ACS Nano 2011, 5, 115-122.

62 Horrion, J.; Jerome, R.; Teyssie, Ph. J. Polym. Sci. Polym. Lett. Ed. 1986, 24, 69-76.

63 Russell, T. P.; Jerome, R.; Charlier, P.; Foucart, M. Macromolecules 1988, 21, 1709-1717.

64 Binder, W. H.; Zirbs, R. Adv. Polym. Sci. 2007, 207, 1-78.

65 Fustin, C.-A.; Guillet, P.; Schubert, U. S.; Gohy, J.-F. Adv. Mater. 2007, 19, 1665-1673.

66 Weiss, R. A.; Sasongko, S.; Jerome, R. Macromolecules 1991, 24, 2271-2277.

67 Zhang, L.; Brostowitz, N. R.; Cavicchi, K. A.; Weiss, R. A. Macromol. React. Eng. 2014, 8, 81-99.

68 Horrion, J.; Jerome, R.; Teyssie, Ph. J. Polym. Sci. Polym. Chem. 1990, 28, 153-171.

69 Charlier, P.; Jerome, R.; Teyssie, Ph.; Utracki, L. A. Macromolecules, 1992, 25, 2651-2656.

190

70 Horrion, J.; Jerome, R.; Teyssie, Ph.; Vanderschueren, J.; Corapci, M. Polym. Bull. 1989, 21, 627-633.

71 Iwasaki, K.; Hirao, A.; Nakahama, S. Macromolecules 1993, 26, 2126-2131.

72 Haraguchi, M.; Nakagawa, T.; Nose, T. Polymer 1995, 36, 2567-2572.

73 Inomata, K.; Haraguchi, M.; Nose, T. Polymer 1996, 37, 4223-4228.

74 Inomata, K.; Liu, L.-Z.; Nose, T.; Chu, B. Macromolecules 1999, 32, 1554- 1558.

75 Zhang, L.; Storey, R. F.; Cavicchi, K. A.; Weiss, R. A. Preparation of Supramolecular Diblock Copolymers via Ionic Interactions, presented at IUPAC World Polymer Congress, Blacksburg, VA, June, 2012

76 Brostowitz, N.; Cavicchi, K. A.; Weiss, R. A. Supramolecular ionic copolymers. presented at IUPAC World Polymer Congress, Blacksburg, VA, June, 2012.

77 Fleischer, C. A.; Morales, A. R.; Koberstein, J. T. Macromolecules 1994, 27, 379-385.

78 Gao, J. P.; Li, X.; Li, B. Y.; Han, Y. C. Polymer 2010, 51, 2683-2689.

79 Takahashi, A.; Rho, Y.; Higashihara, T.; Ahn, B.; Ree, M.; Ueda, M. Macromolecules 2010, 43, 4843-4852.

80 Yu, H. Z.; Stoffelbach, F.; Detrembleur, C.; Fustin, C.-A.; Gohy, J.-F. Eur. Polym. J. 2012, 48, 940-944.

81 Song, G.; Cho, S. M.; Jung, H. J.; Kim, R. H.; Bae, I.; Ahn, H.; Ryu, D. Y.; Huh, J.; Park, C. Chem. Eur. J. 2012, 18, 15662-15668.

82 Qian, J.; Wu, F. P. Macromolecules 2008, 41, 8921-8926.

83 Sen, A.; Weiss, R. A.; Garton, A. In Multiphase Polymers: Blends and Ionomers, (Eds: L. A. Utracki, R. A. Weiss RA) ACS Symposium Series, Vol. 395, American Chemical Society, Washington D.C., 1989, Ch. 14.

84 Orfanou, K.; Topouza, D.; Sakellariou, G.; Pispas, S. J. Polym. Sci. Polym. Chem. 2003, 41, 2454-2461.

85 Plante, M.; Bazuin, C. G. Macromolecules 1997, 30, 2613-2617.

86 Bazuin, C. G.; Plante, M.; Varshney, S. K. Macromolecules 1997, 30, 2618.- 2622

191

87 Jo, Y. M.; Park, C. W.; Jung, B.; Yang, H.-M.; Kim, J.-D. Macromol, Chem. Phys. 2010, 211, 2434-2442.

88 Qian, J.; Wu, F. P. Chem. Mater. 2009, 21, 758-762.

89 Pispas, S.; Floudas, G.; Pakula, T.; Lieser, G.; Sakellarious, S.; Hadjichristidis, N. Macromolecules 2003, 36, 759-763.

90 Sugiyama, K.; Karasawa, Y.; Higashihara, T.; Zhao, Y. L.; Hirao, A. Monatsh Chem. 2006, 137, 869-880.

91 Tao, K.; Lu, D. R.; Bai, R. K. Macromol. Rapid Commun. 2008, 29, 1477- 1483.

92 Eisenberg, A.; Kim, J.-S. Introduction to Ionomers; John Wiley & Sons: New York, 1998.

93 Yarusso, D. J.; Cooper, S. L. Macromolecules 1983, 16, 1871–1880

94 Chakrabarty, K.; Shao, P.; Weiss, R. A. In Ionomers: Synthesis, Structure, Properties and Application, (Eds: M. R. Tant, K. A. Mauritz, G. L. Wilkes), Blackie Acad. Press: London 1997, Ch.4.

95 R. A. Register, R. K. Prud’homme. In Ionomers: Synthesis, Structure, Properties and Application, (Eds: M. R. Tant, K. A. Mauritz, G. L. Wilkes), Blackie Acad. Press: London 1997, Ch.5.

96 Craig, S. L. Angew. Chem. Int. Ed. 2009, 48, 2645-2647.

97 Bouix, M.; Gouzi, J.; Charleux, B.; Vairon J.-P.; Guinot, P. Macromol. Rapid Commun. 1998, 19, 209-213.

98 Kucera, F.; Jancar, J. Polym. Eng. Sci. 1998, 38, 783-792.

99 Jones, G. D. Industrial & Engineering Chemistry 1952, 44, 2686-2693.

100 Sen, A. K.; Roy, S.; Juvekar, V. A. Polym. Int. 2007, 56, 167-174.

101 Matyjaszewski, K.; Davis, T. P. Handbook of Radical Polymerization; John Wiley and Sons: Hoboken, 2002.

102 Kirchner, B. Ionic Liquids; Springer-Verlag: Heidelberg, Germany, 2009.

103 Okamura, H.; Takatori, Y.; Tsunooka M.; Shirai, M. Polymer 2002, 43, 3155- 3162.

104 Aitken, B. S.; Lee, M.; Hunley, M. T.; Gibson H. W.; Wagener, K. B. Macromolecules, 2010, 43, 1699-1701. 192

105 Cheng, S.; Beyer, F. L.; Mather, B. D.; Moore R. B.; Long, T. E. Macromolecules, 2011, 44, 6509-6517.

106 Chen, H.; Choi, J.-H.; Salas-de la Cruz, D.; Winey K. I.; Elabd, Y. A. Macromolecules, 2009, 42, 4809-4816.

107 Cavicchi, K. A. ACS Appl. Mater. Interfaces 2011, 4, 518-526

108 Vanhoorne, P.; Jerome, R. In Ionomers: Characterization, Theory, and Applications, (Eds: S. Schlick), CRC Press: Boca Raton, FL, 1996, Ch.9.

109 Jerome, R.; Mazurek, M. In Ionomers: Synthesis, Structure, Properties and Application, (Eds: M. R. Tant, K. A. Mauritz, G. L. Wilkes), Blackie Acad. Press: London 1997, Ch.1.

110 Feng, L.; Cavicchi, K. A.; Katzenmeyer, B. C.; Wesdemiotis, C. J. Polym. Sci. Polym. Chem. 2011, 49, 5100-5108.

111 Weimer, M. W.; Chen, H.; Giannelis, E. P.; Sogah, D. Y. J. Am. Chem. Soc. 1999, 121, 1615-1616.

112 Huang, W.; Charleux, B.; Chiarelli, R.; Marx, L.; Rassat, A.; Vairon, J.-P. Macromol. Chem. Phys. 2002, 203, 1715-1723.

113 Samakande, A.; Sanderson, R. D.; Hartmann, P. C. Synth. Commun. 2007, 37, 3861-3872.

114 Braunecker, W. A.; Matyjaszewski, K. Prog. Polym. Sci. 2007, 32, 93-146.

115 Odian, G. Principles of Polymerization, 4th ed; John Wiley & Sons: Hoboken, 2004.

116 Chiefari, J.; Chong, Y. K.; Ercole, F.; Krstina, J.; Jeffrey, J.; Le, T. P.; Mayadunne R. T. A.; Meijs, G. F.; Moad, C. L.; Moad, G.; Rizzardo, E.; Thang, S. H. Macromolecules 1998, 31, 5559-5562.

117 Barner-Kowollik, C. In Handbook of RAFT Polymerization; Barner- Kowollik, C. Eds; Wiley-VCH: Weinheim, 2008, Chap. 1.

118 Moad, G.; Rizzardo, E.; Thang, S. H. Aust. J. Chem. 2005, 58, 379-410.

119 Moad, G.; Rizzardo, E.; Thang, S. H. Aust. J. Chem. 2006, 59, 669-692.

120 Moad, G.; Rizzardo, E.; Thang, S. H. Aust. J. Chem. 2009, 62, 1402-1472.

121 Moad, G.; Rizzardo, E.; Thang, S. H. Aust. J. Chem. 2012, 65, 985-1076.

193

122 Keddie, D. J.; Moad, G.; Rizzardo, E.; Thang, S. H. Macrmolecules 2012, 45, 5321-5342.

123 Keddie, D. J. Chem. Soc. Rev. 2014, 43, 496-505.

124 Mayadunne, R. T. A.; Rizzardo, E.; Chiefari, J.; Krstina, J.; Moad, G.; Postma, A.; Thang, S. H. Macromolecules 2000, 33, 243-245.

125 Liu, Y.; Cavicchi, K. A. Macromol. Chem. Phys. 2009, 210, 1647-1653.

126 Xu. J.; Tao, L.; Boyer, C.; Lowe, A. B.; Davis, T. P. Macromolecules 2010, 43, 20-24.

127 Quinn, J. F.; Chaplin, R. P.; Davis, T. P. J. Polym. Sci. A Polym. Chem. 2002, 40, 2956-2966.

128 Bernard, J.; Favier, A.; Davis, T. P.; Barner-Kowollik, C.; Stenzel, M. H. Polymer 2006, 47, 1073-1080.

129 Mayadunne, R. T. A.; Jeffery, J.; Moad, G.; Rizzardo, E. Macromolecules 2003, 36, 1505-1513.

130 Barner-Kowollik, C.; Davis, T. P.; Stenzel, M. H. Aust. J. Chem. 2006, 59, 719-727.

131 Zheng, Q.; Pan, C.-Y. Eur. Polym. J. 2006, 42, 807-814.

132 Hong, C.-Y.; You, Y.-Z.; Liu, J.; Pan, C.-Y. J. Polym. Sci. A Polym. Chem. 2005, 43, 6379-6393.

133 Liu, B.; Kazlauciunas, A.; Guthrie, J. T.; Perrier, S. Macromolecules 2005, 38, 2131-2136.

134 Moad, G.; Rizzardo, E.; Thang, S. H. Polym. Int. 2011, 60, 9-25.

135 Moad, G.; Rizzardo, E.; Thang, S. H. Acc. Chem. Res. 2008, 41, 1133-1142.

136 Willcock, H.; O’Reilly, R. K. Poly. Chem. 2010, 1, 149-157.

137 Fox, J. D.; Rowan, S. J. Macromolecules 2009, 42, 6823-6835.

138 Hoogenboom, R.; Fournier, D.; Schubert, U. S. Chem. Commun. 2008, 155- 162.

139 Stuparu, M. C.; Khan, A.; Hawker, C. J. Polym. Chem. 2012, 3, 3033-3044.

140 Wojtecki, R. J.; Meador, M. A.; Rowan, S. J. Nat. Mater. 2011, 10, 14-27.

194

141 Mugemana, C.; Guillet, P.; Fustin, C.-A.; Gohy, J.-F. Soft Matter 2011, 7, 3673-3678.

142 Hunt, J. N.; Feldman, K. E.; Lynd, N. A.; Deek, J.; Campos, L. M.; Spruell, J. M.; Hernandez, B. M.; Kramer, E. J.; Hawker, C. J. Adv. Mater. 2011, 23, 2327-2331.

143 Kato, T.; Fréchet, J. M. J. Macromolecules 1989, 22, 3819-3821.

144 Lange, R. F. M.; Meijer, E. W.; Macromolecules 1995, 28, 782-783.

145 Gao, Z., Molnar, A., Eisenberg, A. In Ionomers: Synthesis, Structure, Properties and Application; (Eds: M. R. Tant, K. A. Mauritz, G. L. Wilkes), Blackie Acad. Press: London 1997, Ch.10.

146 Schadler, V.; Spickermann,J.; Rader, H. J.; Wiesner, U. Macromolecules 1996, 29, 4865-4870.

147 Schadler, V.; Wiesner, U. Macromolecules 1997, 30, 6698-6701.

148 Lu, D.; Wang, Y.; Wang, Hu.; Bai, R. Eur. Polym. J. 2010, 46, 1417-1424.

149 Liu, Y. Synthesis and Characterization of Well-defined, Amphiphilic, Ionic Copolymers; Ph.D. Dissertation: University of Akron, 2011, p. 44.

150 Storey, R.F.; Chisholm, B.J.; Lee, Y. Polymer 1993, 34, 4330-4335.

151 Morgan, D.L.; Martinez-Castro, N.; Storey, R.F. Macromolecules 2010, 43(21), 8724–8740.

152 Kucera, L. R.; Brei, M. R.; Storey, R. F. Polymer 2013, 54, 3796-3805.

153 Zundel, G. Hydration and Intermolecular Interactions; Academic Press: New York, 1969.

154 Lin-Vien, D.; Colthup, N. B.; Fateley, W. G.; Grasselli, J. G. The Handbook of Infrared and Raman Characteristic Frequencies of Organic Molecules; Academic Press: Boston, 1991.

155 Boyd, S. L.; Boyd, R. J.; Bessonette, P. W.; Kerdraon, D. I.; Aucoln, N. T. J. Am. Chem. Soc. 1995, 117, 8816-8822.

156 Fox, T.J.; Loshaek, S.J. J. Polym. Sci. 1955, 371, 371-390.

157 Roth, C. B.; Torkelson, J. M. Macromolecules 2007, 40, 3328-3336.

158 Matsen, M. W.; Schick, M. Macromolecules 1994, 27, 7157-7163.

195

159 Spontak, R. J.; Smith, S. D. J. Polym. Sci. B Polym. Phys. 2001, 39, 947-955.

160 Fetters, L. J.; Lohse, D. J.; Witten, T. A.; Zirkel, A. Macromolecules 1994, 27, 4639.

161 Petri, H. –M.; Horst, R.; Wolf, B. A. Polymer 1996, 37, 2709-2713.

162 Antony, P.; Puskas, J. E.; Kontopoulou, M. Polym. Eng. Sci. 2003, 43, 243- 253.

163 Matsen, M. W.; Thompson, R. B. J. Chem. Phys. 1999, 111, 7139-7146.

164 Charlier, P.; Jerome, R.; Teyssie, Ph.; Prud’homme, R. E. J. Polym. Sci. A Polym. Chem. 1993, 31, 129-134.

165 Ling, G.; Wang, Y.; Weiss, R. A. Macromolecules 2012, 45, 481−490.

166 Aboudzadeh, M. A.; Munoz, M. E.; Santamaria, A.; Fernandez-Berridi, M. J.; Irusta, L.; Mecerreyes, D. Macromolecules 2012, 45, 7599-7606.

167 Qiao, X.; Weiss, R. A. Macromolecules 2013 46, 2417-2424.

168 Wu, J.-L.; Wang, Y.-M.; Hara, M.; Granville, M.; Jerome, R. J. Macromolecules 1994, 27, 1195.

169 Samakande, A.; Sanderson, R. D.; Hartmann, P. C. Polymer 2009, 50, 42.

170 Di, J.; Sogah, D. Y. Macromolecules 2006, 39, 1020.

171 Zhao, H.; Argoti, S.D.; Farrell, B. P.; Shipp, D. A. J. Polym. Sci., Part A: Polym. Chem. 2004, 42, 916.

172 Bottcher, H.; Hallensleben, M. L.; Nuβ, S.; Wurm, H.; Bauer, J.; Behrens, P. J. Mater. Chem. 2002, 12, 1351.

173 Zhang, B.-Q.; Pan, C.-Y.; Hong, C.-Y.; Luan, B.; Shi, P.-J. Macromol. Rapid Commun. 2006, 27, 97.

174 Zare, P.; Stojanovic, A.; Herbst, F.; Akbarzadeh, J.; Peterlik, H.; Binder, W. H. Macromolecules 2012, 45, 2074.

175 Baussard, J. -F.; Habib-Jiwan, J. -L.; Laschewsky, A.; Mertoglu, M.; Storsberg, J. Polymer 2004, 45, 3615.

176 Moughton, A. O.; O'Reilly, R. K. Chem. Commun. 2010, 46, 1091.

177 Waschinski, C. J.; Herdes, V.; Schueler, F.; Tiller, J. C. Macromol. Biosci., 2005, 5, 149. 196

178 Waschinski, C. J.; Tiller, J. C. Biomacromolecules 2005, 6, 235.

179 Fik, C. P.; Krumm, C.; Muennig, C.; Baur, T. I.; Salz, U.; Bock, T.; Tiller, J. C. Biomacromolecules 2012, 13, 165.

180 Tezuka, Y.; Goethals, E. J. Eur. Polym. J. 1982, 18, 991.

181 Lindsell, W. E.; Radha, K.; Soutar, I.; Stewart, M. J. Polymer 1990, 31, 1374.

182 Guinot, P.; Bryant, L.; Chow, T. Y.; Saegusa, T. Macromol. Chem. Phys. 1996, 197, 1.

183 Rempp, P.; Merrill, E. W. Polymer Synthesis, 2nd Ed; Huthig & Wepf: New York, 1990.

184 Coady, D. J.; Norris, B. C.; Lynch, V. M.; Bielawski, C. W. Macromolecules 2008, 41, 3775.

185 Aoyagi, N.; Ochiai, B.; Mori, H.; Endo, T. Synlett 2006, 636.

186 Skey, J.; O'Reilly, R. K. Chem. Commun. 2008, 35, 4183-4185.

187 Rammo, J.; Schnei der, H.-J. Inorg. Chim. Acta. 1996, 251, 125.

188 Covitz, F. H. J. Am. Chem. Soc. 1967, 89, 5403.

189 Akhavan-Tafti, M. H.; Arghavani, Z. US Patent, 5650099, 1997.

190 Bartsch, R. A.; Zhao, W. Y.; Zhang, Z.-Y. Synth. Commun. 1999, 29, 2393.

191 Strube, O. I.; Nothdurft, L.; Drache, M.; Schmidt-Naake, G. Macromol. Chem. Phys. 2011, 212, 574-582.

192 Movassagh, B.; Soleiman-Beigi, M. Monatsh. Chem. 2008, 139, 927.

193 Khuong, K. S.; Jones, W. H.; Pryor, W. A.; Houk, K. N. J. Am. Chem. Soc. 2005, 127, 1265.

194 Zhang, L.; Cool, L. R.; Wesdemiotis, C.; Weiss, R. A.; Cavicchi, K. A. Polym. Chem. 2014, 5, 1180-1190.

195 Xie, W.; Xie, R.; Pan, W.-P.; Hunter, D.; Koene, B.; Tan, L.-S.; Vaia, R. Chem. Mater. 2002, 14, 4837-4845.

196 Hemp, S. T.; Zhang, M.; Allen M. H. Jr.; Cheng, S.; Moore, R. B.; Long, T. E. Macromol. Chem. Phys. 2013, 214, 2099-2107.

197

197 Kanazawa, A.; Ikeda, T.; Endo, T. J. Polym. Sci., Part A: Polym. Chem. 1993, 31, 335-343.

198 Floch, V.; Loisel, S;. Guénin, E.; Hervé, A.-C.; Clément, J. C.; Yaouanc, J.-J.; des Abbayes, H.; Férec, C. J. Med. Chem. 2000, 43, 4617-4628.

199 Guénin, E.; Hervé, A.-C.; Floch, V.; Loisel, S.; Yaouanc, J.-J.; Clément, J.-C.; Férec, C.; des Abbayes, H. Angew. Chem. Int. Ed. 2000, 39, 629-631.

200 Hemp, S. T.; Smith, A. E.; Bryson, J. M.; Allen, M. H. Jr.; Long, T. E. Biomacromolecules 2012, 13, 231-238.

201 Ornelas-Megiatto, C.; Wich, P. R.; Fréchet, J. M. J. J. Am. Chem. Soc. 2012, 134, 1902-1905.

202 Brzezinska, K.; Chwialkowska, W.; Kubisa, P.; Matyjaszewski, K.; Penczek, S. Makromol. Chem. 1977, 178, 2491-2494.

203 Kubisa, P.; Penczek, S. Makromol. Chem. 1979, 180, 1821-1823.

204 Goethals, E. J. Makromol Chem. Macromol. Symp. 1986, 6, 53-66.

205 Atmaca, L.; Yagci, K. Y. Polymer 2000, 41, 6035-6041.

206 Unal, S.; McKee, M. G.; Massa, D. J.; Long, T. E. Polymer Preprints 2005, 46, 1028-1029.

207 Shen, W.-P.; Zhu, W.-D.; Yang, M.-F.; Wang, L. Makromol. Chem. 1989, 190, 3061-3066.

208 Callahan, J.; Kopecek, J. Biomacromolecules 2006, 7, 2347-2356.

209 Coessens, V.; Matyjaszewski, K. J. Macromol. Sci., Pure Appl. Chem. 1999, 36, 653-666.

210 Tasdelen, M. A.; Kahveci, M. U.; Yagci, Y. Prog. Polym. Sci. 2011, 36, 455.

211 Villemin, D.; Didi, M. A.; Makhoukhi, B. Lett. Org. Chem. 2007, 4, 509-511.

212 Baranyai, K. J.; Deacon, G. B.; MacFarlane, D. R.; Pringle, J. M.; Scott, J. L. Aust. J. Chem. 2004, 57, 145-147.

213 Liu, Y.; Pollock, K. L.; Cavicchi, K. A. Polymer 2009, 50, 6212-6217.

214 Moad, G.; Chiefari, J.; Chong, Y. K.; Krstina, J.; Mayadunne, R. T. A.; Postma, A.; Rizzardo, E.; Thang, S. H. Polym. Int. 2000, 49, 993.

198

215 Barner-Kowollik, C.; Vana, P.; Quinn, J. F.; Davis, T. P. J. Polym. Sci., Part A: Polym. Chem. 2002, 40, 1058.

216 McLeary, J. B.; Calitz, F. M.; McKenzie, J. M.; Tonge, M. P.; Sanderson, R. D.; Klumperman, B. Macromolecules 2005, 38, 3151.

217 McLeary, J. B.; Calitz, F. M.; McKenzie, J. M.; Tonge, M. P.; Sanderson, R. D.; Klumperman, B. Macrmolecules 2004, 37, 2383.

218 Tang, C.; Kowalewski, T.; Matyjaszewski, K. Macromolecules 2003, 36, 8587.

219 Ding, Y. S.; Hubbard, S. R.; Hodgson, K. O.; Register, R. A.; Cooper, S. L. Macromolecules 1988, 21, 1698-1703.

220 Chu, B.; Wu, D. Q.; Lundberg, R. D.; MacKnight, W. J. Macromolecules 1993, 26, 994-999.

221 Colby, R. H.; Zheng, X.; Rafailovich, M. H.; Sokolov, J; Peiffer, D. G.; Schwarz, S. A.; Strzhemechny, Y.; Nguyen, D. Phys. Rev. Lett. 1998, 81, 3876-3879.

222 Weiss, R. A.; Zhao, H. Y. J. Rheol. 2009 53, 191-212.

223 Castgna, A. M.; Wang, W. Q.; Winey, K. I.; Runt, J. Macromolecules 2010, 43, 10498-10504.

224 Feng, Y.; Karim, A.; Weiss, R. A.; Han, C. C. Macromolecules 1998, 31, 484- 493.

225 Zhai, X.; Weiss, R. A. Langmuir 2008 24, 12928-12935.

226 Lundberg, R. D.; Agarwal, P. K.; Weiss, R. A. U.S. Patent 4387174, 1983.

227 Lundberg, R. D. In Ionomers: Synthesis, Structure, Properties and Application; Mauritz, K. A.; Wilkes, G. L.; Tant, M. R. (Eds.); Chapman & Hall: New York, 1997.

228 Weiss, R. A. U. S. Patent 5422398, 1995.

229 Rajagopalan, M. U. S. Patent 6245862, 2001.

230 Peiffer, D. G.; Lundberg, R. D.; Walker, T. O. U. S. Patent 4465801, 1984.

231 Agarwal, P. K.; Lundberg, R. D.; Graessley, W. W.; Ver Strate, G. U. S. Patent 4536310, 1985.

232 Willer, R. L.; Hartwell, J. A. U. S. Patent 5028283, 1991. 199

233 Makowski, H. S.; Lundberg, R. D.; Singhal, G. H. U.S. Patent 3870841, 1975.

234 Baigl, D., Seery, T. A. P.; Williams, C. E. Macromolecules 2002, 35, 2318- 2326.

235 VanderHart, D. L.; Feng, Y.; Han, C. C.; Weiss, R. A. Macromolecules 2000, 33, 2206-2227.

236 Nielen, M.W.F. Mass Spectrom. Rev. 1999, 18: 309-344.

237 Montaudo, G.; Lattimer, R. P. (Eds.). Mass Spectrometry of Polymers; CRC Press: Boca Raton, 2003.

238 Li, L. (Ed.). MALDI Mass Spectrometry for Synthetic Polymer Analysis; John Wiley & Sons: Hoboken, 2010.

239 Montaudo, G.; Samperi, F.; Montaudo, M.S. Prog. Polym. Sci. 2006, 31, 277- 357.

240 Arnould, M.A.; Polce, M.J.; Quirk, R.P.; Wesdemiotis, C. Int. J. Mass Spectrom. 2004, 238, 245-255.

241 Quirk, R.P.; Mathers, R.T.; Wesdemiotis, C.; Arnould, M.A. Macromolecules 2002, 35, 2912-2918.

242 Zhou, N. C.; Xu, C.; Burghardt, W. R.; Composto, R. J.; Winey, K. I. Macromolecules 2006, 39, 2373-2379.

243 Fitzgerald, J. J.; Weiss, R. A. Coulombic Interactions in Macromolecular Systems; Eisenberg, A.; Bailey, F. E. (Eds.). ACS Symposium Series, American Chemical Society: Washington, 1986, vol. 302, pp. 35-53.

244 Owens, K. G.; Hanton, S. D. MALDI Mass Spectrometry for Synthetic Polymer Analysis; Li, L. (Ed.). John Wiley & Sons: Hoboken, 2010, pp. 129-158.

245 Skelton, R.; Dubois, F.; Zenobi, R. Anal. Chem. 2000, 72, 1707-1710.

246 Trimpin, S.; Rouhanipour, A.; Az, R.; Räder, H.J.; Müllen, K. Rapid Commun. Mass Spectrom. 2001, 15, 1364-1373.

247 Trimpin, S. MALDI Mass Spectrometry for Synthetic Polymer Analysis; Li, L. (Ed.). John Wiley & Sons: Hoboken, 2010, pp. 159-186.

248 Li, X.; Guo, L.; Casiano-Maldonado, M.; Zhang, D.; Wesdemiotis, C. Macromolecules 2011, 44, 4555-4564.

249 Danis, P. O.; Karr, D. E. Org. Mass. Spectrom. 1993, 28, 923-925. 200

250 Zenobi, R. MALDI Mass Spectrometry for Synthetic Polymer Analysis; Li, L. (Ed.). John Wiley & Sons: Hoboken, 2010, Chapter 2.

201