Synthesis and Functionalization of 1,4-Polyketones and Enantioselective Polyester Catalyst Development Using Molecular Lego Scaffolds

Home , Knorr pyrrole synthesis

SYNTHESIS AND FUNCTIONALIZATION OF 1,4-POLYKETONES AND ENANTIOSELECTIVE POLYESTER CATALYST DEVELOPMENT USING MOLECULAR LEGO SCAFFOLDS

A Dissertation Submitted to the Temple University Graduate Board

In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY

by Evan M. Samples May 2019

Examining Committee Members:

Graham E. Dobereiner, Ph.D, Dissertation Supervisor, Temple University Ann M. Valentine, Ph.D., Temple University Stephanie L. Wunder, Ph.D., Temple University Deanna L. Zubris, Ph.D., External Member, Villanova University

Evan M. Samples

ii ABSTRACT

Objectives of the present study are aimed towards improving upon alternating copolymerization techniques for polyketones and aliphatic polyesters, and the majority of this work focused on post-polymerization modifications to alternating polyketones. These materials are currently understudied in the literature, but the aptly spaced, repeating carbonyl functionality creates an easily functionalized material. Complementary work described herein relates to efforts currently underway to prepare highly enantioselective catalysts for the alternating copolymerization of epoxides with cyclic anhydrides.

Aliphatic polyesters currently suffer from a lack of chemical diversity, and, with greener chemistries on the forefront of research efforts, polyesters made from environmentally benign and/or renewable materials are desirable.

Additional limitations of aliphatic polyesters include difficulty obtaining stereoregular polyesters. In collaboration with the Schafmeister laboratory we are developing catalysts for the alternating copolymerization of polyesters to address these limitations. The model catalysts are carefully designed scaffolds of spiroligomers encasing a Lewis acidic transition metal at its center ([spiro]MX). The spiroligomer bulk around the metal center imparts significant chirality onto the catalyst thereby controlling which enantiomer of a given monomer is polymerized leading to stereoregular polyesters.

Additionally, the use of more than one monomer increases the available chemical space with which to create novel polyesters. To date, three [spiro]MX catalysts have been prepared all of which are catalytically active for poly(propylene maleate) synthesis.

iii A core objective of this work is the study of functionalization methods to create novel materials from inexpensive polyketones. The chemical modifications performed on polyketones to date have been limited, and the utility of the functionalized materials often goes unmentioned. Efforts to functionalize polyketones in this study were aimed at creating electrically conducting polymeric materials which would be used as hole transport materials in photovoltaic devices. Polyketones were decorated with pendant

(tri)arylamine functionality creating several novel polymeric materials, and electrochemical experiments supported the formation of radical cations at the triarylamine nitrogen of the pendants. Further, the functionalization of the polyketones provided enhanced ultraviolet stability of the functionalized polymers.

Concurrent to the functionalization of polyketones, we investigated the effects

Lewis acids had on the synthesis of the polyketone itself. Through previous research conducted in the Dobereiner laboratory we know that a Lewis acid will interact with carbonyls of molecules during catalytic reactions. The addition of Lewis acids to the synthesis of the polyketones is thought to have similar interactions altering the polymerization. This study explored the bulk properties of the polyketone synthesized in the presence of several Lewis acids. As a result of this study specific polymer properties

(e.g. molecular weight) could be targeted through careful selection of the Lewis acid and the amount added during polymerization.

iv DEDICATION

This dissertation is dedicated to my parents

Ron and Judy Samples

and to my sisters

Emily and Sarah

for their unwavering love and support.

v ACKNOWLEDGMENTS

After graduating from college, I never expected to go to graduate school for a

PhD, and I definitely did not expect to find myself in an organometallic laboratory.

Working with Dr. Graham Dobereiner has been a great experience from start to finish.

Graham has taught me so much about chemistry, and I am grateful to have him as a mentor and a friend. It has been a great experience building up a research lab over the years and watching the group grow from just Derek and I to a full-fledged research group. Thanks to Dr. Ann Valentine for serving on my committee and providing scholarly advice during my time at Temple. Thank you to Dr. Stephanie Wunder for serving as a member on my advisory committee, and for being an invaluable resource for polymer chemistry. I would like to thank Dr. Deanna Zubris for serving as an external committee member, and for offering valuable research insight during my first years of graduate studies. Dr. Bradford Wayland has been an enthusiastic teacher and mentor during my inorganic chemistry studies and I have learned much from him. I would also like to thank Dr. Charles Debrosse for his spectroscopic expertise and maintenance of the

NMR facilities. Dr. Debrosse was always willing to help with NMR experiments and spectral analyses.

My colleagues and labmates (past and present) in the Dobereiner lab have provided a great work environment. Dr. Kushan Weerasiri and Alan Shaffer were great labmates who were always helpful when I needed a different view of my chemistry.

Additional gratitude is due to Alexa Wallace and Jeremy Schuck for being such hard working undergraduates that I have had the pleasure of mentoring through the years. The

vi chemistry department at Temple University is very collaborative, and so I must thank Dr.

Kallie Willets, Dr. Padmanabh Joshi, and Dr. Drew Wilson of the Willets laboratory for their electrochemical expertise which played a significant role during my research.

I have made many friends at Temple who have made the years fly by.

Specifically, I want to acknowledge Derek Wozniak, Lauren Martin, Dr. Brenden

Derstine, Dr. Christiana Teijaro, Dr. Colleen Keohane, Dr. Andrew Steele, and Alex

Koval for all the happy hours, Wednesday wings, journal club shenanigans, and road trips we have had together. Not only are these people great friends, but they are incredibly helpful and supportive.

Mostly, I need to acknowledge and thank my amazing family. My mother and father have been hugely supportive during my college career and further in my graduate studies. They have always pushed me to be my best and have always been enthusiastic about my research even when they do not understand it. In addition, my sisters are incredibly smart and funny. They always have rock solid advice for me in any situation, and I know that I can count on them for anything.

vii TABLE OF CONTENTS

Page

ABSTRACT ...... III

DEDICATION...... V

ACKNOWLEDGMENTS ...... VI

LIST OF TABLES ...... X

LIST OF FIGURES ...... XI

CHAPTER 1 INTRODUCTION ...... 1

1.1 Statement of Dissertation Objectives ...... 1 1.2 Background to Polyketones ...... 2 1.3 Mechanism of Polyketone Formation with Cationic Palladium Species ...... 4 1.4 Post-polymerization Functionalization of Polyketones ...... 7 1.5 Alternating Copolymers and Stereo-control ...... 11 1.6 Background to Aliphatic Polyesters...... 14 1.7 Summary ...... 17 1.8 References Cited ...... 19 CHAPTER 2 POLYKETONE FUNCTIONALIZATION FOR HOLE TRANSPORT ...... 25

2.1 Introduction ...... 25 2.2 Results and Discussion ...... 29 2.2.1 Polyketone Synthesis ...... 29 2.2.2 Functionalization of Polyketone ...... 31 2.2.2.1 Reductive Amination ...... 31 2.2.2.2 Paal-Knorr Cyclization of 1,4-polyketone ...... 34 2.2.3 Properties of Functionalized Polyketones ...... 41 2.2.3.1 Azobenzene-derived Polyketone ...... 41 2.2.3.2 Triarylamine-functionalized Polyketones ...... 46 2.2.3.3 Substituted Pyrrole Model Compounds ...... 58 2.3 Conclusion ...... 65 2.4 Experimental ...... 66 2.5 References Cited ...... 76

viii CHAPTER 3 LEWIS ACID EFFECTS ON POLYKETONE SYNTHESIS ...... 85

3.1 Introduction ...... 85 3.2 Results and Discussion ...... 87 3.3 Conclusion ...... 101 3.4 Experimental ...... 101 3.5 References Cited ...... 105 CHAPTER 4 SPIROLIGOMER POLYESTER CATALYST DEVELOPMENT ...... 109

4.1 Introduction ...... 109 4.2 Results and Discussion ...... 115 4.2.1 Catalyst Synthesis ...... 116 4.2.2 Alternating Copolymerization ...... 118 4.2.3 Reductive Degradation...... 126 4.3 Conclusion ...... 128 4.4 Experimental ...... 128 4.5 References Cited ...... 133 CHAPTER 5 1-D PEROVSKITE SYNTHESIS AND HOLE TRANSPORTER OXIDATION ...... 136

5.1 Introduction ...... 136 5.1.1 Perovskite Solar Cell Device Fabrication ...... 138 5.1.2 Hole Transporting Materials (HTM) ...... 140 5.2 Results and Discussion ...... 142 5.2.1 Spiro-OMeTAD Oxidation ...... 142 5.2.2 Perovskite Synthesis ...... 146 5.3 Conclusion ...... 151 5.4 Experimental ...... 152 5.5 References Cited ...... 154

ix LIST OF TABLES

Table Page

Table 2.1: Various amines, solvents, and reaction tempertaures used during reductive

amination reactions with a polyketone. Molecular sieves were used in each case.

...... 33

Table 2.2: Percent incorporation for each tricopolymer. Amine nucleophilicity had little

influence on the actual polymer composition...... 58

Table 2.3: Absorption maxima for the neutral and oxidized polymers and molecular

models. Absorption maxima for the neutral and oxidized polymers and molecular

models...... 61

Table 4.1: Selected PPM copolymerization results, used to determine optimum reaction

conditions and most appropriate metal/ligand combination...... 120

Table 4.2: Comparison of [salcy]MX catalysts to [spiro]MX catalysts for PPM

sysnthesis...... 125

Table 5.1: Dopants screened in spiro-OMeTAD oxidation...... 143

x LIST OF FIGURES

Figure Page

Figure 1.1: Alternating copolymerization of carbon monoxide and α-olefins...... 3

Figure 1.2: A general mechanism for polyketone synthesis using a cationic palladium

catalyst. L-L: bidentate phosphorus or nitrogen ligand. R: alkyl (-H, -(CH 2)n-, or –

CH 3) or aryl (C 6H5). S: solvent monomer or counterion. P: growing polymer

chain...... 4

Figure 1.3: Possible end groups for polyketones...... 5

Figure 1.4: Proposed chain transfer mechanisms for the copolymerization of CO with

ethylene. I, P, M, and H represent initiation, propagation, methanolysis, and

protonolysis, respectively. Reprinted with permission from Drent, E.; Budzelaar,

Chemical Society...... 7

Figure 1.5: Various strategies for post-polymerization functionalization of polyketones.

C) Ar is p-tertbutylbenzene and A) R is alkyl or aryl substitution...... 9

Figure 1.6: Carilite curing chemistry through Paal-Knorr reaction with a difunctional

amine...... 10

Figure 1.7: Paal-Knorr cyclization of a 1,4-polyketone. Sulfur may be in the form of

Lawesson’s reagent or phosphorus pentasulfide...... 11

Figure 1.8: Regioregular polyketones are formed when the olefin inserts in either a 1,2-

or a 2,1-insertion pathway. L-L: bidentate phosphorus or nitrogen ligand. P:

growing polymer chain...... 12

xi Figure 1.9: Stereochemistry found in 1,4-polyketones. Reproduced from Nakano, K. et

al., Dalton Transactions 2003, (21), 4039-4050. with permission from The Royal

Society of Chemistry...... 13

Figure 1.10: ROP of lactide to give polylactic acid (PLA)...... 14

Figure 1.11: ROP of meso -lactide may yield either syndio- or heterotactic PLA while

ROP of an enantiopure lactide ((S,S)-lactide is shown) will yield a syndiotactic

polymer. Polymerization of the racemate leads to predominantly heterotactic

PLA. Nakano, K. et al., Dalton Transactions 2003, (21), 4039-4050. with

permission from The Royal Society of Chemistry...... 15

Figure 1.12: Chiral metal-alkoxides are efficient ROP catalysts to produce entantiopure

polyesters via kinetic resolution of racemic monomers...... 16

Figure 1.13: Ring opening copolymerization of epoxides with anhydrides. The spheres

represent sites of potential chemical functionalization of the monomers, and sites

of potential stereo- and regio-selectivity in the polyester...... 17

Figure 2.1: State-of-the-art HTM Spiro-OMeTAD compared to polymeric HTMs...... 25

Figure 2.2: Contemporary examples of polymeric HTM candidates highlighting the

synthetic complexity...... 27

Figure 2.3: A) The polyketone catalyst used herein to copolymerize styrene with CO. B)

The catalyst used by Brookart et al. to copolymerize styrene with CO...... 29

Figure 2.4: A general pathway to reduce the carbonyls of a polyketone. The aryl groups

of the amine lead to redox active pendants ideal for hole transport...... 31

Figure 2.5: Control reductive amination reaction...... 32

xii 1 Figure 2.6: H NMR in CDCl 3 of N-(sec-butyl)-N-methylaniline. The expansions show

(left, arrow) the chiral methine and (right) the adjacent methylene protons...... 32

Figure 2.7: A general reaction for reductive amination of the polyketone...... 33

Figure 2.8: Gel permeation chromatograph (GPC) of the control reaction between the

polyketone and STAB-H. The polymer suffered significant reduction (23%) in

molecular weight as evidenced by its longer retention time...... 34

Figure 2.9: The general Paal-Knorr reaction for a 1,4-diketone. Sulfur may be in the form

of Lawesson’s reagent or phosphorus pentasulfide...... 35

Figure 2.10: Arylamines used to functionalize PK6 via catalyzed Paal-Knorr cyclization.

...... 37

Figure 2.11: A step-wise process to building up amine-pendant complexity using Paal-

Knorr cyclization and subsequent Buchwald-Hartwig amination...... 38

Figure 2.12: Buchwald-Hartwig amination to prepre substituted amino-triarylamines used

in Paal-Knorr cyclizations...... 39

Figure 2.13: Triarylamine-functionalized polymers via Paal-Knorr cyclization. %

incorporation was determined by 1H NMR...... 40

Figure 2.14: Azobenzene-derived polymer is easily prepared from 4-aminoazobenzene to

yield 13 as a bright orange polymer. Structural differences are compared to

Nozaki’s material...... 41

Figure 2.15: Reversible trans →cis photoisomerization of polymer 13...... 42

Figure 2.16: The UV-visible spectra of 13 before UV exposure (solid trace), after UV

light exposure (dotted trace), and after visible light exposure (dashed trace). PK6

under identical conditions is shown inset. Reprinted with permission from

American Chemical Society...... 43

Figure 2.17: GPC chromatograms during UV exposure of A) PK6 and B) polymer 13.

Reprinted with permission from Samples, E. M. et al., Macromolecules 2018, 51

Figure 2.18: UV-induced polymer degradation reported as a the percentage of polymer

intact (M n/M n0 ) as a function of time. Reprinted with permission from Samples, E.

Chemical Society...... 45

Figure 2.19: UV degradation of selected functionalized polyketones...... 46

Figure 2.20: UV-Vis spectra for polymers 12a-e (A) prior to and (B) after reaction with

AgSbF 6 in dichloromethane. 12a, purple, dashed; 12b, blue, dash-dot; 12c, green,

square dot; 12d, red, round dot; 12e, blue, solid. Reprinted with permission from

American Chemical Society...... 47

Figure 2.21: A compilation of the cyclic voltammograms collected for the triarylamine-

functionalized polyketones 12a-e...... 49

Figure 2.22: Differential pulse voltammogram for PK6...... 51

Figure 2.23: Differential pulse voltammogram for 12a...... 51

Figure 2.24: Differential pulse voltammogram for 12b...... 52

Figure 2.25: Differential pulse voltammogram for 12c...... 52

Figure 2.26: Differential pulse voltammogram for 12d...... 53

Figure 2.27: Differential pulse voltammogram for 12e...... 53

xiv Figure 2.28: DPV voltammograms contrasting Spiro-OMeTAD and polymer 12b

oxidation. The dotted lines are to highlight the anodic differences between the two

materials...... 55

Figure 2.29: A random tricopolymer was prepared by reaction with aniline and 11b. The

ratio of aniline:11b minimally affected the ratio n:o ...... 57

Figure 2.30: DPV voltammogram for each of the tricopolymers 1, 2, 3, and polymers 3

and 12b. Reprinted with permission from Samples, E. M. et al., Macromolecules

Figure 2.31: Paal-Knorr reaction between 2,5-hexanedione and 4-bromoaniline yielding

14...... 59

Figure 2.32: Amination procedure to prepare model compounds 15a-e...... 60

Figure 2.33: Normalized UV-Vis spectra for 15a-e. For each spectrum the blue (solid)

trace represents the neutral molecule and the red (dotted) trace represents the

oxidized species after reaction with AgSbF 6 in dichloromethane...... 63

Figure 2.34: Voltammogram comparison for 12b and 15b. Reprinted with permission

from Samples, E. M. et al., Macromolecules 2018, 51 (22), 9323-9332. Copyright

2018 American Chemical Society...... 64

Figure 2.35: DPV voltammograms for each of the molecular models 15a-e...... 65

Figure 2.36: Potential waveform used for DPV experiments. Potential pulses of 50 mV

amplitude for 50 ms duration over 500 ms period were applied...... 68

1 Figure 2.37: Polymer 3 annotated H NMR spectrum in CD 2Cl 2. 1,3,5-trimethoxy

benzene was used as an internal standard to calculate the percent conversion of

PK6...... 74

xv Figure 3.1: The general structure of a cationic polymerization catalyst. Where ligands L 1,2

may be monodentate or bidentate, and an open coordination site is designated by

□...... 85

Figure 3.2: 1) Example of a SCC used for polyketone synthesis. 2) Catalyst precursor

used herein to copolymerize α-olefins with CO...... 86

Figure 3.3: Synthetic route to polyketone synthesis with olefins larger than ethene or

propene. No spiroketal formation was observed using the present catalytic system.

...... 88

Figure 3.4: The average catalytic activity of a 24 h reaction for each of the LAs used

during PK6 synthesis...... 89

Figure 3.5: Polyketone yield as a function of BCF loading...... 90

Figure 3.6: Averaged isolated masses of PK6...... 90

Figure 3.7: Averaged molecular weights of PK6 with different LAs...... 91

Figure 3.8: Molecular weight distributions (Ð) for PK6 using different LAs. PK6 without

LA has a Ð of 1.6...... 92

Figure 3.9: Cooley and coworkers’ proposed activation process with BCF beginning with

fluoroaryl transfer to palladium. Reprinted with permission from Barlow, G. K. et

Society...... 93

Figure 3.10: The relative rates of propagation and termination determine the polymer

molecular weight...... 94

Figure 3.11: Internal chelation of the polyketone chains and potential disruption of

chelation by Lewis acid additives...... 95

xvi Figure 3.12: Ionic strength influence on PK6 yield. High-valent species reduced the yield

more than low-valent species. *BCF data are not included since this Lewis acid does not contribute to the

ionic strength. **The dotted line is to illustrate the partition between high- and low-valent additives...... 96

Figure 3.13: Catalytic activity comparison of polymerizations with additive

alkylammonium triflate or BCF...... 97

Figure 3.14: Catalytic activity comparison of PK6 with alkylammonium triflate or LAs.

...... 98

Figure 3.15: Isolated masses of PK5, PK6, and PK8 when polymerized in the presence of

either the alkylammonium triflates or BCF...... 98

Figure 3.16: Glass transition temperatures as a function of BCF loading...... 99

Figure 3.17: Glass transition temperatures fall off precipitously as a result of the reduced

molecular weight observed with the addition of BCF...... 100

Figure 3.18: As the molecular weight decreases the T g is more affected by the relative

ratio of alkyl side chain to alternating ketone backbone...... 101

Figure 4.1: A) Step-growth polymerization between two bifunctional monomers to

prepare polyesters. B) ROP of lactone. C) Alternating copolymerization of

epoxides and cyclic anhydrides is a facile route to prepare novel APs...... 110

Figure 4.2: A simplified mechanism for the alternating copolymerization of an epoxide

with cyclic anhydrides with PPNCl as a cocatalyst. Reprinted with permission

from Longo, J.M. et al. Chemical Reviews 2016, 116 (24), 15167-15197.

xvii Figure 4.3: Examples of semi-crystalline APs prepared from enantiopure epoxides. A)

poly((S)-propylene succinate) T m = 79 °C. B) poly((S)-propylene maleate) T m =

117 °C. C) poly((S)-propylene phthalate) T m = 150 °C...... 113

Figure 4.4: A cartoon representation of the proposed spiroligomer-based catalyst. This

representation illustrates four spiroligomer segments connected at the bottom with

a transition metal (M) ligated within the chiral pocket created by the spiroligomer

segments. Figure credit to Dr. Christian Schafmeister...... 114

Figure 4.5: Comparison of [salcy]MX catalysts to the [spiro]MX catalysts developed in

the Schafmeister laboratory. The [salcy]MX catalysts are active at both axial faces

of the complex; the bottom face of the [spiro]MX catalyst is proposed to be

inaccessible to substrate due to the spiroligomer/alkyl tether...... 115

Figure 4.6: The general synthesis for a salen ligand...... 116

Figure 4.7: Salen scaffolds used in model copolymerization reactions...... 117

Figure 4.8: Metalation of salens 1-4 with various metal salts led to a library of catalysts

for PPM synthesis...... 118

Figure 4.9: Alternating copolymerization of propylene oxide and maleic anhydride to

yield poly(propylene maleate) (PPM)...... 119

Figure 4.10: 1H NMR spectra for PPM prepared with: catalyst 6 without PPNCl

(spectrum 1); catalyst 14 without PPNCl (spectrum 2); and 14 with PPNCl

(spectrum 3). *indicates residual diethyl ether...... 122

Figure 4.11: [spiro]MX catalysts screened for PPM synthesis...... 123

Figure 4.12: Polyether synthesis during induction period leading to consistent ether

content in PPM during different reaction times. Reprinted from Brocas et al.,

permission from Elsevier...... 124

Figure 4.13: Reductive degradation of the polyester provides a mixture of diols which can

be resolved via chiral GC...... 126

Figure 4.14: A) PPM catalyzed with 14 exhibited no selectivity for PO. B) PPM

catalyzed by 6. *Propylene glycol peaks are not well resolved with the current

method, and relative ratios of R- and S-PG are qualitative...... 127

Figure 4.15: A) The relative ratios of R- and S-PG from PPM catalyzed by 17. B) The

sample was spiked with a standard S-PG to aid in peak indentity...... 127

Figure 5.1: Perovskite crystal structure. MA and FA represent methylammonium and

formamidinium, respectively. Reprinted with permission from Calió, L. et al.,

Angewandte Chemie International Edition 2016, 55 (47), 14522-14545.

Figure 5.2: Device architecture of a) mesoscopic and b) planar heterojunction PSCs.

Reprinted with permission from Angewandte Chemie International Edition 2016,

Figure 5.3: Doping spiro-OMeTAD with LiTFSI leads to a usable HTM for PSCs...... 141

Figure 5.4: 12-c-4 chelates Li + effectively “removing” the ions from solution and

preventing oxidation of spiro-OMeTAD...... 142

Figure 5.5: Solid-state oxidation of spiro-OMeTAD with LiTFSI. Absorptions at ca. 400

nm and 520 correspond to spiro-OMeTAD and spiro(TFSI), respectively...... 144

Figure 5.6: Spiro-OMeTAD oxidation with varying amounts of CE...... 145

xix Figure 5.7: Spiro-OMeTAD oxidation with varying amounts of 12-c-4 deposited on glass

slides over the course of 3 h. Vast differences and inconsistencies in absorption at

520 nm highlights inconsistencies in sample preparation...... 146

Figure 5.8: Crystals of 1 prepared from different solvents were analyzed via UV-Vis to

develop a finger printing method of perovskite identification...... 148

Figure 5.9: TGA was used to determine if hydrated perovskites were prepared...... 149

Figure 5.10: A 1-D perovskite structure was only attained once with hydrazinium iodide.

...... 150

Figure 5.11: DMSO solvated crystal structure of PbI3 which was recovered from an

attempt to prepare 1...... 150

CHAPTER 1

INTRODUCTION

1.1 Statement of Dissertation Objectives

Copolymerization is an effective method for combining and maximizing favorable properties of different monomers while minimizing their individual deficiencies.1 This thesis focuses on two classes of copolymers - polyketones and polyesters. Polyketones are produced from copolymerization of carbon monoxide and an

α-olefin, and these materials offer a multitude of attractive properties from chemical resistance to post-polymerization functionalization. The alternating polymerization mechanism of palladium-catalyzed copolymerization is known to produce perfectly alternating microstructures, and the resulting polyketone is ideal for post-polymerization functionalization through a myriad of chemical transformations. While polyketones have many positive attributes, they are derived from fossil fuels. As petroleum reserves diminish, there is rising interest in renewable and biodegradable copolymers. Polyesters, especially aliphatic polyesters, are compelling alternatives to petroleum-based materials and are easily prepared from low cost feedstocks. 2 These plastics are sought after for their biodegradability/biocompatibility which enable their use as artificial tissues, drug delivery systems, and medical devices. 3

This dissertation describes the synthesis and reactivity of alternating polyketones and polyesters. The major objective of this work is the post-polymerization functionalization of polyketones to novel amine-functionalized polymers, and to adjust 1 macromolecular properties through the addition of Lewis acid cocatalysts during copolymerization. Further, work to develop stereoselective spiroligomer-based catalysts for polyester syntheses is described.

1.2 Background to Polyketones

Polyketones are obtained by alternating copolymerization of carbon monoxide and an α-olefin, and represent a class of thermoplastics that have attracted attention for their properties and post-polymerization functionalization potential. The first instance of copolymerization of carbon monoxide and an olefin was achieved by Farbenfabriken

Bayer in 1941. 4 Initially, nickel(II) catalysts were used in combination with high

temperatures (200 °C) and high partial pressures of carbon monoxide (200 atm) and

ethylene. 4 Since then major improvements have been made to catalyst, monomer selectivity, and environmental impact. The use of methanol and water as solvents has eliminated the need for chlorinated hydrocarbons, and a large variety of monomers have been found to copolymerize with carbon monoxide leading to novel materials with vastly different properties. 5, 6

Polyketones (Figure 1.1) offer a myriad of valuable properties as the result of the alternating polymerization mechanism achieved using homogenous transition metal catalysts. Chemical resistance, mechanical resilience, and hydrocarbon impermeability have all been achieved.7, 8, 9 Polyketones have been also used as adhesives, coatings, and fibers, 4, 10, 11 and have potential use in liquid crystals, drug delivery, ion conduction, and membrane applications. 8, 12 Furthermore, some polyketones exhibit a high degree of

2 crystallinity due to the keto functionality resulting in high melting temperatures (220-260

°C) – polyethylene and syndiotactic polystyrene have melting temperatures of 115-140

°C and 270 °C, respectively.

Figure 1.1: Alternating copolymerization of carbon monoxide and α-olefins.

Only a few transition metals have successfully been used to prepare polyketones.

Nickel(II) catalysts were the first used and are a viable alternative to the more costly

palladium systems, but are limited to o-methoxy-modified diphosphine ligands like 1,2- bis(bis(2-methoxyphenyl)phosphino)-ethane. 4 Palladium-based catalysts, however, are active with nitrogen or phosphorus-based ligands. Rhodium carbonyls can also produce low molecular weight polyketones from ethylene, but catalytic activity is low due to slow olefin insertion into the Rh-C bond. 4, 13 To date, the most effective catalyst for preparing

polyketones with simple olefins (i.e. ethylene and propylene) remains the dicationic

palladium species stabilized with bidentate phosphine ligands such as 1,3-

bis(diphenylphosphino)propane (DPPP) in the complex [Pd(DPPP)(MeCN) 2](BF 4)2;

catalysts for styrene copolymerization are generally stabilized with bidentate nitrogen

ligands. 4, 7, 14, 15, 16

1.3 Mechanism of Polyketone Formation with Cationic Palladium Species

The mechanism of polymerization for olefins with carbon monoxide was investigated thoroughly by Drent, 4 Sen, 13 and Brookhart 17 and a general reaction

mechanism is shown in Figure 1.2.18

Figure 1.2: A general mechanism for polyketone synthesis using a cationic

palladium catalyst. L-L: bidentate phosphorus or nitrogen ligand. R: alkyl (-H, -

(CH 2)n-, or –CH 3) or aryl (C 6H5). S: solvent monomer or counterion. P: growing

polymer chain.

During the copolymerization in the presence of methanol there are two competing initation pathways as well as two termination pathways, which are distinguished by the resulting end groups (Figure 1.3): keto-ester 1, diester 2, and diketone 3; the combination of initiation and termination pathway determines the end group of the polyketones. 4, 13

Figure 1.3: Possible end groups for polyketones.

Initiation can occur when carbon monoxide inserts into the Pd-OMe bond followed by olefin insertion (Cycle A, Figure 1.4). Termination of this pathway via protonolysis leads to predominantly keto-ester end groups. If initiation occurs by insertion into a Pd-H, each of the three types of end groups are possible (Cycle B, Figure

1.4). Keto-esters result from methanolysis of Pd-acyl bonds and diketone groups are a result of protonolysis of the Pd-alkyl bond. 4, 13, 19

The catalytically-active species for the propagating polymer chains is thought to be a square planar d 8 cationic palladium species (L-L)PdP + where L-L and P represent a

bidentate ligand and growing polymer chain, respectively. 13 The fourth coordination site

may be filled by solvent, monomer, or counter ion (S in Figure 1.2). For the propagating

polymer there are two alternating steps – CO insertion into the Pd-alkyl bond (Eq. 1) and

olefin insertion into the resulting Pd-acyl species (Eq. 2).

(1)

(2)

Differences in carbon monoxide and olefin binding affinity lead to a perfectly alternating polymer product with no double insertion of olefin. Work by Sen 20, 21 revealed

that since palladium has a higher binding affinity for carbon monoxide even at low partial

pressures of carbon monoxide double enchainment of ethylene is kinetically unfavorable.

Conversly, double insertion of another carbon monoxide molecule is thermodynamically

unfavorable. 4

Figure 1.4: Proposed chain transfer mechanisms for the copolymerization of CO

with ethylene. I, P, M, and H represent initiation, propagation, methanolysis, and protonolysis, respectively. Reprinted with permission from Drent, E.; Budzelaar, P.

Society.

1.4 Post-polymerization Functionalization of Polyketones

Despite many strengths, polyketones’ have had only limited commercial

availability from Shell Chemical as Carilon thermoplastics and Carilite resins (available

from 1996 – 2000), and, more recently from Hyosung as “Poketone” and Akro-Plastic as

“Akrotek PK”; these offerings are a terpolymer of ethylene/propylene/CO. One drawback 7 when using palladium-based catalysts is catalyst precipitation from solution as palladium black, complicating polymer purification. 4, 13 Another drawback is polyketone instability to prolonged high temperatures or to ultraviolet light. 7, 22, 23 However, facile post- polymerization reactions at carbonyls yield a vast range of polymer structures with diverse functionality (Figure 1.5). 4 These modified structures could serve as high performance materials for niche applications where costly purification and thermal/light sensitivity can be tolerated.

Figure 1.5: Various strategies for post-polymerization functionalization of

polyketones. C) Ar is p-tertbutylbenzene and A) R is alkyl or aryl substitution.

24 Polyketones react with LiAlH 4 reduction to form polyols D, undergo acid-

promoted cyclization to generate conjugated furans C,6 and condense with hydroxylamine to yield polyketoximes B25, 26 or with hydrazine to form N-heterocycles. 27

Paal-Knorr cyclizations A represent one class of reactions which will yield three different

products depending on the reagents used (Figure 1.7). Use of protic acids provide a furan

product 6, using primary amines yields pyrroles 28, 29, 30, 31 , and sulfonating agents (e.g. 9

Lawesson’s Reagent) yield thiophenes. 24, 32 Brookhart and coworkers were able to prepare conjugated furan polymers from a styrene derivative/CO polyketone. 6 Sen was

able to convert up to 75% of the carbonyls of a polyketone to thiophenes using this

chemistry, but found the product to be unstable when stored exposed to air due to

crosslinking between thiols. 32 The Piccioni group has exploited Paal-Knorr reactions to develop polymeric surfactants and fluorescent polyamines from the ethylene/propylene terpolymers. 29 Other functional materials from the Piccioni group range from wood adhesives 10 to metal ion scavengers 33 , and illustrate the utility and modularity of

polyketones with the Paal-Knorr cyclization. Industrial 31, 34, 35, 36, 37, 38 research groups have also modified polymer properties via Paal-Knorr pyrrole synthesis; for example, condensation of polyketones and alkyl diamines effectively cross-links oligomers of

Shell’s Carilon thermoplastics to form Shell’s Carilite Resins which were patented for use as wood adhesives (Figure 1.6).13, 30, 39

Figure 1.6: Carilite curing chemistry through Paal-Knorr reaction with a

difunctional amine.

Figure 1.7: Paal-Knorr cyclization of a 1,4-polyketone. Sulfur may be in the form of

Lawesson’s reagent or phosphorus pentasulfide.

1.5 Alternating Copolymers and Stereo-control

Copolymerization is an effective method for combining and maximizing favorable properties of different monomers while minimizing their individual deficiencies. 1 Additionally, controlling the stereochemistry of a copolymer can

dramatically affect the polymer’s properties – even changes within a homopolymer’s

microstructure can have a marked influence on the resulting material. 40, 41 Stereoregular polymers like isotactic polypropylene have had a major impact on the world for its mechanical properties, but there are few examples of polymers with main-chain chirality. 42

Olefins for polyketone synthesis are classified into two groups: aliphatic and

vinylarenes. When higher olefins are used in copolymerization with carbon monoxide the

stereocontrol of the catalyst can be assessed. With the substitution of higher olefins it is

11 then possible to distinguish between either a regioregular or regioirregular copolymer.

Regioregular polyketones are obtained with either 1,2- or 2,1-insertion of the olefin in the

Pd-acyl bond during chain propagation (Figure 1.8). The difference between 1,2- and 2,1- insertion arises between aliphatic and vinylarenes as aliphatic olefins like propene follow the 1,2-insertion pathway whereas vinylarenes like styrene follow the 2,1-insertion pathway. 43 In order to influence the insertion pathway and synthesize a stereoregular polyketone, chiral ligands must be used.

Figure 1.8: Regioregular polyketones are formed when the olefin inserts in either a

1,2- or a 2,1-insertion pathway. L-L: bidentate phosphorus or nitrogen ligand. P:

growing polymer chain.

Precise design of homogeneous metal catalysts has allowed for the development

of stereoregular polymers. For polyketones with olefins larger than ethylene (e.g.

propylene), backbone stereocenters yield either isotactic and syndiotactic

12 stereochemistries (Figure 1.9). Using achiral catalysts will yield a polyketone mixture of the A, B, and atactic tacticities for aliphatic olefins 44 and syndio-rich polyketones for vinylarenes since the catalyst does not differentiate between the two enantiomers of the incoming monomer. To date there have not been any reports of syndiotactic polyketones with aliphatic monomers, but work by Drent and Consiglio has led to syndiotactic poly(styrene-alt -CO). 43

Figure 1.9: Stereochemistry found in 1,4-polyketones. Reproduced from Nakano, K.

et al., Dalton Transactions 2003, (21), 4039-4050. with permission from The Royal

Society of Chemistry.

Though the bulk of the research covered in this dissertation is centered around polyketone synthesis and functionalization, their stereochemistry is not discussed further.

The catalytic system used in Chapters 2 and 3 for polyketone synthesis are well- researched 7, 14, 15, 16, 45 to yield perfectly alternating copolymers, and the tacticity of the

13 polyketones is not researched herein. Chapter 4 discusses efforts to develop highly enantioselective catalysts for aliphatic polyester copolymerization.

1.6 Background to Aliphatic Polyesters

Polyesters are attractive materials for their biocompatibility and biodegradability which leads to their applications in medical devices like sutures and screws or as artificial tissue. 3, 46, 47 Polyesters may be synthesized through condensation polymerization,3 bacterial fermentation 3, ring-opening polymerizations of cyclic esters,3, 43 and alternating copolymerization of epoxides and cyclic anhydrides.3, 43, 48 The most common polyester synthesis is the ring-opening polymerization (ROP) of lactide (the cyclic ester of lactic acid) to polylactic acid (PLA, Figure 1.10).

Figure 1.10: ROP of lactide to give polylactic acid (PLA).

A variety of metal-alkoxides have been successful for ROP of lactide, and nucleophilic addition of the alkoxide to the carbonyl cleaves the acyl-oxygen bond allowing retention of the monomer configuration at stereocenters (Figure 1.11). 43, 49, 50

This stereocenter retention and the two diastereomers of lactide allow for the possibility

14 of three types of stereoregularity in the polyester: isotactic, syndiotactic, and heterotactic

(Figure 1.11).

Figure 1.11: ROP of meso -lactide may yield either syndio- or heterotactic PLA while

ROP of an enantiopure lactide ((S,S)-lactide is shown) will yield a syndiotactic

polymer. Polymerization of the racemate leads to predominantly heterotactic PLA.

Nakano, K. et al., Dalton Transactions 2003, (21), 4039-4050. with permission from

The Royal Society of Chemistry.

The simplest method to produce isotactic PLA is to polymerize one of the optically pure lactides (i.e. (S,S)-lactide) using a metal-alkoxide. Kinetic resolution of the racemate is also possible as Spassky and co-workers demonstrated with the enantiomerically pure, chiral aluminum complex ( R)-1 (R = Me, Figure 1.12). 43 Catalyst

2 was used by Coates et al. to prepare heterotactic polylactic acid via kinetic resolution of the racemic lactide. Coates was also the first to report the synthesis of syndio tactic polylactic acid using ( R)-1 from the meso -lactide. 43, 49, 50

Figure 1.12: Chiral metal-alkoxides are efficient ROP catalysts to produce

entantiopure polyesters via kinetic resolution of racemic monomers.

Although polyester like PLA and poly( ε-caprolactone) are readily prepared via

ROP, the scope of the polymer architecture is rather limited by the availability of diverse

monomers. 3 Coates et al. have actively pursued alternating ring opening

copolymerization routes to circumvent the limited structural diversity of polyester

synthesis. 2, 3, 48, 51, 52 Initial work was performed with derivatives of 2 to copolymerize 16 diglycolic acid and cyclohexene oxide with various other epoxides and anhydrides. 3

Though the stereochemistry was not reported for these initial copolymers, it was noted that consecutive enchainment of either the epoxide or the anhydride was not observed. 3

Unsaturated polyesters or biorenewable polyesters have been realized through this polymerization method with metal-salen catalysts, and the stereochemistry has been investigated thoroughly (Figure 1.13). 2, 48, 53

Figure 1.13: Ring opening copolymerization of epoxides with anhydrides. The

spheres represent sites of potential chemical functionalization of the monomers, and

sites of potential stereo- and regio-selectivity in the polyester.

1.7 Summary

Polyketone functionalization to novel amine-containing polymers is a major focus

of this dissertation. The modularity of polyketones has not yet been fully realized, and the

methods in the literature focus heavily on the ethylene or styrene copolymers.

Additionally, the use of Lewis acid cocatalysts affects the synthesis of polyketones, and a

screening of these effects is explored with various Lewis acids. Minor directions for this

dissertation will look into perovskite synthesis and polyester catalyst development.

Chapter two examines polyketone synthesis itself as well as functionalization and resultant properties of some of the functionalized materials. Ultra-violet stability of the polymer is assessed. The redox dependence on polymer pendant substitution of triarylamine-derived polymers are compared to molecular analogues.

Chapter three explores the role of Lewis acids during polyketone synthesis.

Several metal triflates are screened and the resulting polyketone is analyzed via 1H NMR,

GPC, and DSC. The same analyses are performed when BCF is used, and are contrasted

to those polymers prepared with metal triflates. The observed effects on polyketone yield

and molecular weight are discussed, and the generality of these results are tested with

other polyketones.

Chapter four reports the development of spiroligomer-based polyester catalysts.

Several salen scaffolds were synthesized and metalated with various transition metals.

Using these metal-salens poly(propylene maleate) synthesis is used as a model

polymerization system for monomer conversion, molecular weight, and enantiomeric

excess. Catalytic activity of spiroligomer complexes were compared to the smaller salens.

Chapter five discusses lead-based perovskites and hole transport materials and

their use in developing perovskite solar cells (PSCs). Additionally, the requirements for

efficient HTMs and perovskite solar cells fabrication is discussed.

1.8 References Cited

1. Arrington, K. J.; Waugh, J. B.; Radzinski, S. C.; Matson, J. B., Photo- and Biodegradable Thermoplastic Elastomers: Combining Ketone-Containing Polybutadiene with Polylactide Using Ring-Opening Polymerization and Ring-Opening Metathesis Polymerization. Macromolecules 2017, 50 (11), 4180-4187.

2. Van Zee, N. J.; Coates, G. W., Alternating Copolymerization of Propylene Oxide with Biorenewable Terpene-Based Cyclic Anhydrides: A Sustainable Route to Aliphatic Polyesters with High Glass Transition Temperatures. Angewandte Chemie International Edition 2015, 54 (9), 2665-2668.

3. Jeske, R. C.; DiCiccio, A. M.; Coates, G. W., Alternating Copolymerization of Epoxides and Cyclic Anhydrides: An Improved Route to Aliphatic Polyesters. Journal of the American Chemical Society 2007, 129 (37), 11330-11331.

4. Bianchini, C.; Meli, A., Alternating copolymerization of carbon monoxide and olefins by single-site metal catalysis. Coordination Chemistry Reviews 2002, 225 (1–2), 35-66.

5. Kosaka, N.; Oda, T.; Hiyama, T.; Nozaki, K., Synthesis and Photoisomerization of Optically Active 1,4-Polyketones Substituted by Azobenzene Side Chains. Macromolecules 2004, 37 (9), 3159-3164.

6. Cheng, C.; Guironnet, D.; Barborak, J.; Brookhart, M., Preparation and characterization of conjugated polymers made by postpolymerization reactions of alternating polyketones. Journal of the American Chemical Society 2011, 133 (25), 9658- 61.

7. Sen, A., Mechanistic aspects of metal-catalyzed alternating copolymerization of olefins with carbon monoxide. Accounts of Chemical Research 2002, 26 (6), 303-310.

8. Zehetmaier, P. C.; Vagin, S. I.; Rieger, B., Functionalization of aliphatic polyketones. MRS Bulletin 2013, 38 (03), 239-244.

9. Beller, M.; Steinhoff, B. A.; Zoeller, J. R.; Cole ‐Hamilton, D. J.; Drent, E.; Wu, X.; Neumann, H.; Ito, S.; Nozaki, K., In Applied Homogeneous Catalysis with 19

Organometallic Compounds , Cornils, B.; Herrmann, W. A.; Beller, M.; Paciello, R., Eds. Wiley-VCH: Weinheim, 2017; pp 91-190.

10. Hamarneh, A. I.; Heeres, H. J.; Broekhuis, A. A.; Sjollema, K. A.; Zhang, Y.; Picchioni, F., Use of soy proteins in polyketone-based wood adhesives. International Journal of Adhesion and Adhesives 2010, 30 (7), 626-635.

11. Kinneberg, P. A.; Armer, T. A.; Van Breen, A. W.; Barton, R. E. C.; Klei, E. Polyketone-based structural adhesive. US 4880904, November 14, 1989, 1989.

12. Bartsch, G. C.; Malinova, V.; Volkmer, B. E.; Hautmann, R. E.; Rieger, B., CO- alkene polymers are biocompatible scaffolds for primary urothelial cells in vitro and in vivo. BJU International 2007, 99 (2), 447-453.

13. Drent, E.; Budzelaar, P. H. M., Palladium-Catalyzed Alternating Copolymerization of Alkenes and Carbon Monoxide. Chemical Reviews 1996, 96 (2), 663-682.

14. Drent, E.; Budzelaar, P. H. M., Palladium-Catalyzed Alternating Copolymerization of Alkenes and Carbon Monoxide. Chemical Reviews 1996, 96 (2), 663.

15. Nozaki, K.; Hiyama, T., Stereoselective alternating copolymerization of carbon monoxide with alkenes. Journal of Organometallic Chemistry 1999, 576 (1), 248-253.

16. Beckmann, U.; Eichberger, E.; Rufinska, A.; Sablong, R.; Klaui, W., Nickel(II) catalysed co-polymerisation of CO and ethene: Formation of polyketone vs. polyethylene – The role of co-catalysts. Journal of Catalysis 2011, 283 (2), 143-148.

17. Brookhart, M.; Rix, F. C.; DeSimone, J. M.; Barborak, J. C., Palladium(II) catalysts for living alternating copolymerization of olefins and carbon monoxide. Journal of the American Chemical Society 1992, 114 (14), 5894-5895.

18. Ayusman, S., Mechanistic aspects of Metal-catalyzed Alternating Copolymerization of Olefins with Carbon Monoxide. Accounts of Chemical Research 1993, 26 , 303.

19. Vavasori, A.; Ronchin, L., Polyketones: Synthesis and Applications. In Encyclopedia of Polymer Science and Technology , 2017.

20. Chen, J. T.; Sen, A., Mechanism of transition-metal-catalyzed double carbonylation reactions. Synthesis and reactivity of benzoylformyl complexes of palladium(II) and platinum(II). Journal of the American Chemical Society 1984, 106 (5), 1506-1507.

21. Sen, A.; Chen, J. T.; Vetter, W. M.; Whittle, R. R., Synthesis, characterization, and reactivity of .alpha.-ketoacyl complexes of platinum(II) and palladium(II). Crystal structures of trans-Pt(PPh3)2(Cl)(COCOPh) and cis-Pt(PPh3)2(COPh)(CO2Me). Journal of the American Chemical Society 1987, 109 (1), 148-156.

22. Koizumi, T.; Hasegawa, Y.; Takata, T.; Endo, T., Synthesis and photodegradation of polyacrylonitrile having ketone group obtained from radical copolymerization of 2,2- diphenyl-4-methylene-1,3-dioxolane with acrylonitrile. Journal of Polymer Science Part A: Polymer Chemistry 1994, 32 (16), 3193-3195.

23. Gooden, R.; Hellman, M. Y.; Hutton, R. S.; Winslow, F. H., Solid-state photochemistry of poly(ethylene-co-carbon monoxide). Model studies of polyethylene photochemistry. Macromolecules 1984, 17 (12), 2830-2837.

24. Zhaozhong, J.; Sanganeria, S.; Sen, A., Polymers Incorporating Backbone Thiophene, Furan, and Alcohol Functionalities Formed Through Chemical Modifications of Alternating Olefin-Carbon Monoxide Copolymers. Journal of Polymer Science Part A: Polymer Chemistry 1994, 32 , 841-847.

25. Green, M. J.; Lucy, A. R.; Lu, S.-Y.; Paton, R. M., Functionalisation of alkene– carbon monoxide alternating copolymers via transketalisation reactions. Journal of the Chemical Society, Chemical Communications 1994, (18), 2063-2064.

26. Lu, S.-Y.; Paton, R. M.; Green, M. J.; Lucy, A. R., Synthesis and characterization of polyketoximes derived from alkene-carbon monoxide copolymers. European Polymer Journal 1996, 32 (11), 1285-1288.

27. Sen, A.; Jiang, Z.; Chen, J. T., Novel nitrogen-containing heterocyclic polymers derived from the alternating ethylene-carbon monoxide copolymer. Macromolecules 1989, 22 (4), 2012-2014.

28. Chen, J.-T.; Yeh, Y.-S.; Sen, A., Synthesis and characterization of N-substituted poly(ethylenepyrrole): functionalization of ethylene–carbon monoxide alternating copolymers. Journal of the Chemical Society, Chemical Communications 1989, (14), 965-967.

29. Zhang, Y.; Broekhuis, A. A.; Stuart, M. C. A.; Picchioni, F., Polymeric amines by chemical modifications of alternating aliphatic polyketones. Journal of Applied Polymer Science 2008, 107 (1), 262-271.

30. Mul, W. P.; Dirkzwager, H.; Broekhuis, A. A.; Heeres, H. J.; van der Linden, A. J.; Guy Orpen, A., Highly active, recyclable catalyst for the manufacture of viscous, low molecular weight, CO–ethene–propene-based polyketone, base component for a new class of resins. Inorganica Chimica Acta 2002, 327 (1), 147-159.

31. Kiovsky, T. E.; Kromer, R. C. Polymeric pyrrollic derivative. US 3979374A, June 13, 1975, 1975.

32. Jiang, Z.; Sanganeria, S.; Sen, A., Polymers incorporating backbone thiophene, furan, and alcohol functionalities formed through chemical modifications of alternating olefin–carbon monoxide copolymers. Journal of Polymer Science Part A: Polymer Chemistry 1994, 32 (5), 841-847.

33. Toncelli, C.; Haijer, A.; Alberts, F.; Broekhuis, A. A.; Picchioni, F., The Green Route from Carbon Monoxide Fixation to Functional Polyamines: A Class of High- Performing Metal Ion Scavengers. Industrial & Engineering Chemistry Research 2015, 54 (38), 9450-9457.

34. Wong, P. K. Pyridine derivatives. US 5047501A, December 29, 1987, 1987.

35. Wong, P. K. Polymeric polyketone derivatives. EP 0324998A2, December 29, 1987.

36. Sinai-Zingde, G. D. Polymer formed by reaction of a polyketone and an amino acid. US 5605988A, March 16, 1992, 1992.

37. Brown, S. L. Polymers and processes for their preparation. EP 0400903A2, May 31, 1989, 1989.

38. Brown, S. L. Poly pyrrole from co/olefin terpolymer. US 5081207A, May 31, 1989, 1989.

39. Broekhuis, A. A.; Freriks, J. Polymeric amines. US 5952459, March 24, 1997, 1997.

40. Brookhart, M.; Wagner, M. I., Synthesis of a Stereoblock Polyketone through Ancillary Ligand Exchange. Journal of the American Chemical Society 1996, 118 (30), 7219-7220.

41. Hauptman, E.; Waymouth, R. M.; Ziller, J. W., Stereoblock Polypropylene: Ligand Effects on the Stereospecificity of 2-Arylindene Zirconocene Catalysts. Journal of the American Chemical Society 1995, 117 (46), 11586-11587.

42. Schätz, A.; Scarel, A.; Zangrando, E.; Mosca, L.; Carfagna, C.; Gissibl, A.; Milani, B.; Reiser, O., High Stereocontrol and Efficiency in CO/Styrene Polyketone Synthesis Promoted by Azabis(oxazoline)−Palladium Complexes. Organometallics 2006, 25 (17), 4065-4068.

43. Nakano, K.; Kosaka, N.; Hiyama, T.; Nozaki, K., Metal-catalyzed synthesis of stereoregular polyketones, polyesters, and polycarbonates. Dalton Transactions 2003, (21), 4039-4050.

44. Abu-Surrah, A. S.; Rieger, B., High molecular weight 1-olefin/carbon monoxide copolymers: a new class of versatile polymers. Topics in Catalysis 1999, 7 (1), 165-177.

45. Pérez ‐Foullerat, D.; Meier, U. W.; Hild, S.; Rieger, B., High ‐Molecular ‐Weight Polyketones from Higher α‐Olefins: A General Method. Macromolecular Chemistry and Physics 2004, 205 (17), 2292-2302.

46. Lasprilla, A. J. R.; Martinez, G. A. R.; Lunelli, B. H.; Jardini, A. L.; Filho, R. M., Poly-lactic acid synthesis for application in biomedical devices — A review. Biotechnology Advances 2012, 30 (1), 321-328.

47. Ikada, Y.; Tsuji, H., Biodegradable polyesters for medical and ecological applications. Macromolecular Rapid Communications 2000, 21 (3), 117-132.

48. DiCiccio, A. M.; Coates, G. W., Ring-Opening Copolymerization of Maleic Anhydride with Epoxides: A Chain-Growth Approach to Unsaturated Polyesters. Journal of the American Chemical Society 2011, 133 (28), 10724-10727.

49. Ovitt, T. M.; Coates, G. W., Stereoselective Ring-Opening Polymerization of meso-Lactide: Synthesis of Syndiotactic Poly(lactic acid). Journal of the American Chemical Society 1999, 121 (16), 4072-4073.

50. Ovitt, T. M.; Coates, G. W., Stereochemistry of Lactide Polymerization with Chiral Catalysts: New Opportunities for Stereocontrol Using Polymer Exchange Mechanisms. Journal of the American Chemical Society 2002, 124 (7), 1316-1326.

51. Fieser, M. E.; Sanford, M. J.; Mitchell, L. A.; Dunbar, C. R.; Mandal, M.; Van Zee, N. J.; Urness, D. M.; Cramer, C. J.; Coates, G. W.; Tolman, W. B., Mechanistic Insights into the Alternating Copolymerization of Epoxides and Cyclic Anhydrides Using a (Salph)AlCl and Iminium Salt Catalytic System. Journal of the American Chemical Society 2017, 139 (42), 15222-15231.

52. Sanford, M. J.; Peña Carrodeguas, L.; Van Zee, N. J.; Kleij, A. W.; Coates, G. W., Alternating Copolymerization of Propylene Oxide and Cyclohexene Oxide with Tricyclic Anhydrides: Access to Partially Renewable Aliphatic Polyesters with High Glass Transition Temperatures. Macromolecules 2016, 49 (17), 6394-6400.

53. Longo, J. M.; Sanford, M. J.; Coates, G. W., Ring-Opening Copolymerization of Epoxides and Cyclic Anhydrides with Discrete Metal Complexes: Structure–Property Relationships. Chemical Reviews 2016, 116 (24), 15167-15197.

CHAPTER 2

POLYKETONE FUNCTIONALIZATION FOR HOLE TRANSPORT

2.1 Introduction

There is renewed interest in discovering replacements for Spiro-OMeTAD used as hole transport material (HTM) in perovskite solar cells (PSC) and OLEDs. 1 Polymeric options offer several desirable features like processing ease and manipulation of melt temperatures. Several hole transport polymers already exist, but only a few examples are used extensively. Example materials are poly(3-hexylthiophene-2,5-diyl) (P3HT) and poly(triarylamines) (pTAAs; Figure 2.1) but the power conversion efficiency (PCE) of polymer-derived devices fall short of molecular Spiro-OMeTAD (20% PCE). 1, 2, 3, 4

Figure 2.1: State-of-the-art HTM Spiro-OMeTAD compared to polymeric HTMs.

A potential reason for low PCE is the recombination of holes from the photoactive layer due to insufficient coverage of the HTM to the photoactive material.

Other challenges in HTM design include maximizing the extent of pore filling in perovskite solar cell (PSC) assembly. 5, 6 The tackiness of an aliphatic polyketone may help mitigate poor contact between the two layers of the device, and exploit the functional versatility of the polyketone to include redox-relevant pendant groups would help bridge the gap between pore filling and PCE.

6, 7 It is cited that HTMs with high glass transition temperatures T g are favorable to prevent microcrystalline states within the material. These microcrystalline states create phase boundaries between polycrystallites and prevent hole mobility leading to charge recombination and device failure. 6 To put these temperatures in perspective Spiro-

8 OMeTAD has a T g of 125 °C and P3HT and PTAA have T gs of 1-15 °C and 46-98 °C, respectively. These temperature ranges depend on molecular weight and annealing temperature for the polymers. 9, 10, 11 The modularity of a polyketone polymer allows for

the customization of hydrocarbon tail length, thus manipulating T g, opens new pathways

to redox active materials.

Figure 2.2: Contemporary examples of polymeric HTM candidates highlighting the

synthetic complexity.

While pTAAs are widespread throughout optoelectronic research (Figure 2.2).12,

13, 14, 15, 16, 17, 18 , most polymer designs (Spiro-OMeTAD included) require laborious

syntheses of specialized monomers. Additionally, there seems to be no combination of

properties hypothesized to produce a perfect HTM, but there are recurring properties that many groups note. Amorphous HTMs with high glass transition temperatures (T g) to prevent inducing crystallinity during operation and high hole mobility (~10 -3 cm 2/V s) are

underlying themes in HTM research. 13, 19, 20, 21 Polymeric HTMs provide good thermal

stability, good solubility, tunable optoelectronic properties, and high conductivity. 21

Given that the overwhelming majority of molecular and polymeric hole transport materials contain (tri)arylamines 6 this type of functionality was an obvious target for the pendant groups to be used with the polyketone. Polyketones themselves are ideal candidates for HTM design owing to their low cost of monomer feedstock (carbon monoxide and olefin) and the ease of producing perfectly-alternating copolymer structures.22, 23, 24 . The regular spacing of carbonyls is ideal for for a host of post- polymerization modifications which can be used to modify the polyketone to contain redox active pendants. This chapter reports a synthetic approach to developing HTMs based on a polyketone backbone. The design incorporates triarylamines linked to the polymer backbone via pyrrole linkages in a post-polymerization modification of the polyketone. The modified polyketones were then characterized and compared with molecular analogues. The electrochemical data obtained from these materials is also reported. While several novel materials were synthesized, the notable failures are also mentioned.

2.2 Results and Discussion

2.2.1 Polyketone Synthesis

As mentioned in Chapter 1 vinylarenes are typically copolymerized with CO using palladium catalysts stabilized with bidentate bisnitrogen ligands. Styrene/CO copolymerization is far more prevalent in current polyketone research than its aliphatic counterparts, so initially a styrene/CO catalyst was synthesized following work from

Brookhart et al (Figure 2.3). 25

Figure 2.3: A) The polyketone catalyst used herein to copolymerize styrene with CO.

B) The catalyst used by Brookart et al. to copolymerize styrene with CO.

3 Catalyst A was prepared from [( η -C3H5)Pd(µ-Cl)] 2. Copolymerizations were

attempted under air-free conditions with 1 atmosphere of CO, but the copolymer was

never isolated; polystyrene was the sole product. This inactivity is likely due to only

using 1 atm of CO during the reaction. Diffusion of the CO into the solution of monomer

and catalyst was likely slow compared to the homopolymerization of styrene from

adventitious impurities or oxygen.

Since vinylarene copolymerizations were not successful, aliphatic olefin copolymerization was the next target. Palladium catalysts are useful here again and are stabilized with diphosphine ligands. Depending on catalyst, preparations of poly(1- alkene-alt -CO) can yield mixtures of interconvertible polyketone and polyspiroketal. 26, 27

22, 27, 28, 29 [Pd(dppp)(MeCN) 2](BF 4)2 was used to ensure generation of an exclusively

alternating polyketone structure. 30

[Pd(dppp)(MeCN) 2](BF 4)2 produced a polyketone from 1-hexene and carbon

monoxide (henceforth PK6 ) in good yield (~ 40% olefin conversion), and the product

1 13 was characterized by H NMR (methylene protons, 3.25 – 2.75 ppm in CDCl 3) C NMR

-1 (carbonyl carbons, 212 ppm), ATR-FTIR (carbonyl νCO , 1704 cm ) and GPC

4 4 (Molecular weights of the copolymer M n = 2.2 x 10 and M w = 3.7 x 10 with dispersity of 1.6).

2.2.2 Functionalization of Polyketone

2.2.2.1 Reductive Amination

Figure 2.4: A general pathway to reduce the carbonyls of a polyketone. The aryl

groups of the amine lead to redox active pendants ideal for hole transport.

Functionalizing a polyketone through reductive amination is a logical route to installing arylamino functionality for hole transport onto the polyketone scaffold (Figure

2.4). Aldehydes and ketones may be reduced to form amines through reductive amination, and Abdel-Magid and Mehrman released a thorough review on the use of sodium triacetoxyborohydride (STAB-H) in reductive aminations. 31 General reaction

conditions were noted in the review and were used as guidelines for reactions with the

polyketone and various amines. A model reaction using 2-butanone and N-methylaniline

produced N-(sec-butyl)-N-methylaniline, a tertiary amine (Figure 2.5; 1H NMR spectrum

Figure 2.6).

Figure 2.5: Control reductive amination reaction.

1 Figure 2.6: H NMR in CDCl 3 of N-(sec-butyl)-N-methylaniline. The expansions

show (left, arrow) the chiral methine and (right) the adjacent methylene protons.

Similar reactions were performed with PK6 (Figure 2.7). To perform this reaction a solution of THF and 1 eq of PK6 were added to a flask with activated molecular sieves.

N-methylaniline (1.1 eq) was then injected into the flask containing the polyketone/THF

32 solution followed by acetic acid (1 eq) and STAB-H (1.4 eq). Reaction with 4- methoxybenzylamine as well as with diethylamine resulted in no reaction. Additional reaction conditions were attempted (Table 2.1) but without success. Additionally, a reduction in polymer molecular weight for each reaction was observed, suggesting degradation of the polymer. A control reaction confirmed a STAB-H promoted reduction in molecular weight even in the absence of amine (Figure 2.8).

Figure 2.7: A general reaction for reductive amination of the polyketone.

Table 2.1: Various amines, solvents, and reaction tempertaures used during

reductive amination reactions with a polyketone. Molecular sieves were used in each

case.

Temp. Amine Solvent °C Diethylamine THF rt THF 50

DCM rt

DCE 60

N-methylaniline THF rt 4-methoxy-benzylamine DCE 70 DCE 75

Control (no amine) DCE 70

Figure 2.8: Gel permeation chromatograph (GPC) of the control reaction between

the polyketone and STAB-H. The polymer suffered significant reduction (23%) in

molecular weight as evidenced by its longer retention time.

2.2.2.2 Paal-Knorr Cyclization of 1,4-polyketone

In light of the polymer degradation observed under reducing conditions, other

functionalization strategies were necessary. Paal-Knorr cyclizations were an appealing

strategy for incorporating amine functionality, leading to isolable materials for HTMs.

The presence of the 1,4-diketone moiety allows the polyketone to be

functionalized with an amine via Paal-Knorr condensation (Figure 2.9).32 Paal-Knorr cyclizations represent a class of reactions that will yield three different products depending on the reagents used (Figure 2.9). Protic acids provide a furan product 33 , primary amines yields pyrroles 34, 35, 36, 37 , and sulfonating agents (e.g. Lawesson’s 34

Reagent) yield thiophenes. 38, 39 Brookhart and coworkers were able to prepare conjugated

furan polymers from a styrene derivative/CO polyketone. 33 Sen converted up to 75% of

the carbonyls of a polyketone to thiophenes using this chemistry, but found the product to

be unstable to long-term storage due to crosslinking between thiols. 39

These reactions are favorable and spontaneous between a 1,4-diketone and a primary amine, but generally require very high temperatures (e.g. > 100 °C). Catalysts can be used to promote milder Paal-Knorr reactions,34, 35 and montmorillonite clay,

iodine, iron phosphate, microwaves, and various Lewis acids have all been used as

catalysts for small-molecule Paal-Knorr condensations. 40, 41, 42 Polymer-based reactions are expected to be more challenging than molecular counterparts due to the steric bulk of the alkyl side chain and the polymer backbone.

Figure 2.9: The general Paal-Knorr reaction for a 1,4-diketone. Sulfur may be in the

form of Lawesson’s reagent or phosphorus pentasulfide.

First reactions were attempted using 1.1 equivalents of amine and polyketone (1 eq) in a high-boiling solvent (e.g. toluene or ortho-dichlorobenzene). Reactions were also attempted neat in the amine when amine was liquid at elevated temperatures. However, conversion of the PK6 was low – < 6 % conversion was observed when 4-bromoaniline was used. Catalyzing these reactions was desirable to lower reaction temperatures as well as increase percent conversion of PK6 . A catalytic Paal-Knorr reaction was adapted from

40, 42 literature precedent using Sc(OTf)3 (Figure 2.10). It is worth mentioning that long

polymer chains will undergo cyclization at random places along the polymer backbone.

This will, invariably, leave randomly-spaced ketones trapped between the pyrroles

(Figure 2.10); 86% conversion of the 1,4-diketone moieties to pyrroles is the statistical

maximum. 34, 43

Figure 2.10: Arylamines used to functionalize PK6 via catalyzed Paal-Knorr

cyclization.

In order to develop a useful HTM, large triarylamine pendants were desired for their inherent radical stability which is necessary for an organic p-type semi-conductor 6, 7,

13, 21, 44 and the steric bulk of the aryl rings provides delocalization of the incurred charge, but in a previous report Sen 34 found amine reactivity to depend heavily on steric bulk. A

step-wise functionalization of the polyketone was therefore pursued, in order to prepare a

37 functionalized material with a high percent conversion to redox active pendants. A condensation of halogenated anilines and PK6 using a Paal-Knorr condensation was first

attempted, followed by a Buchwald-Hartwig amination of resulting N-(haloaryl)pyrrole

units with anilines (Figure 2.11). The latter reaction failed using p-toluidine or 2- pyrrolidinone likely due to the steric bulk offered by the alkyl side chains of PK6 and the

low effective concentration of R-X bonds in solution. An alternative path was chosen

where functionalized amines were used directly in Paal-Knorr cyclizations.

Figure 2.11: A step-wise process to building up amine-pendant complexity using

Paal-Knorr cyclization and subsequent Buchwald-Hartwig amination.

Buchwald-Hartwig amination was used to join Boc-protected 4-bromoaniline 9 and commercially available diarylamines 10a-e to form amines 11a-e (Figure 2.12). 11a- e were then deprotected with 30% HCl and were stored as hydrochloride salts and neutralized as needed.

Figure 2.12: Buchwald-Hartwig amination to prepre substituted amino-

triarylamines used in Paal-Knorr cyclizations.

When necessary the triarylamine ·HCl was basified by washing a methylene

chloride solution of the salt with aqueous sodium carbonate. The organic layer was then

dried with magnesium sulfate and filtered to remove the solids. The dried organic layer

was then used for subsequent Paal-Knorr reaction with PK6 . Triarylamino-derived

polymers 12a-e are listed in Figure 2.13. NMR was used to ascertain the percent incorporation of functionalized units (see experimental section). No obvious correlation

39 between percent incorporation and amine properties (sterics, electronics) is clear from the data collected.

Figure 2.13: Triarylamine-functionalized polymers via Paal-Knorr cyclization. %

incorporation was determined by 1H NMR.

Additionally, an azo-benzene functionalized polymer was also prepared in this same manner (Figure 2.14). Covalently incorporating the azobenzene functionality post- polymerization can introduce photodegradation resistance, birefringence, and nonlinear optical effects 45, 46, 47, 48, 49 while avoiding the expense of producing azobenzene-derived

monomers. 46, 48, 50 Nozaki and coworkers had previously prepared a similar polyketone

with copolymerization of a 4-(5-hexenyloxy)azobenzene. 46 No comment on the

appearance of the material was made. The high molar extinction coefficient of

azobenzene-functionalized 13 resulted in a brightly colored orange polymer despite low

percent incorporation (2%).

Figure 2.14: Azobenzene-derived polymer is easily prepared from 4-

aminoazobenzene to yield 13 as a bright orange polymer. Structural differences are

compared to Nozaki’s material.

2.2.3 Properties of Functionalized Polyketones

2.2.3.1 Azobenzene-derived Polyketone

Using chromophores is an effective way to protect a polymer from UV-induced

degradation, and covalently linking the azo moiety to the polymer scaffold offers greater

polymer stability and maximum photoresponsiveness of the azobenzene chromophore.

Using azobenzene offers significant UV degradation mitigation while also creating an

optically active material. The photoisomerization of stilbenes and azobenzenes is

known, 51, 52 so the optical activity of 13 was expected to exhibit reversible trans → cis isomerization via irradiation with UV and visible light (Figure 2.15). Azo π – π* and azo

n – π* excitations trigger trans → cis and cis → trans isomerizations, respectively. 46

Figure 2.15: Reversible trans →cis photoisomerization of polymer 13.

To induce photoisomerization, 13 was dissolved in ethyl acetate and irradiated for

60 mins with a medium pressure mercury lamp (400 W; 280 – 600 nm with a Pyrex

filter). Absorption of the polymer π-π* absorption band (ca. 338 nm) decreases after 60 minutes of irradiation, indicating conversion to the cis isomer (Figure 2.16). 53

Irradiating this sample subsequently with visible light (5.88 W blue LED, 440-460 nm) promoted the cis →trans isomerization (dashed trace), with a partial recovery of the π-π* trans absorption band. The unfunctionalized polymer PK6 was subjected to the same

conditions and is shown inset of Figure 2.16.

Figure 2.16: The UV-visible spectra of 13 before UV exposure (solid trace), after UV

light exposure (dotted trace), and after visible light exposure (dashed trace). PK6

under identical conditions is shown inset. Reprinted with permission from Samples,

Chemical Society.

The loss of the faint spectral feature for PK6 at 350 nm and the inability to revive that feature is due to degradation; polyketones are susceptible to photodegradation through a Norrish type II mechanism.22, 54, 55 UV degradation was also assessed by measuring the change in polymer molecular weight after a period of irradiation (Figure

2.17). 43

Figure 2.17: GPC chromatograms during UV exposure of A) PK6 and B) polymer

13. Reprinted with permission from Samples, E. M. et al., Macromolecules 2018, 51

Based on gel permeation chromatography analysis, a sample of PK6 irradiated by

a 400 W mercury lamp rapidly decomposes, reaching a 33% reduction in molecular

weight within 1 hour. In comparison, the molecular weight of 13 is essentially unchanged after 1 hour of irradiation, with a more gradual decomposition thereafter (Figure 2.18), consistent with prior findings where azobenzene chromophores added to polymers prevent UV degradation. 53, 56

Figure 2.18: UV-induced polymer degradation reported as a the percentage of

polymer intact (M n/M n0 ) as a function of time. Reprinted with permission from

American Chemical Society.

This protection from UV-induced degradation was general for other functionalized polyketones (Figure 2.19). Radicals formed during the Norrish Reaction attack the carbonyls along the polymer chain. Therefore, functionalizing PK6 would inherently increase the resistance to UV degradation. It is interesting to note in Figure

2.19 that polymer 12b increases in molecular weight after being degraded for some time.

This may be attributed to polymer-polymer oligomerization or crosslinking reactions that occur in the presence of radicals formed by the Norrish II reaction. 57 Nevertheless, the azo- chromophore appears to be the most effective additive to prevent degradation.

Figure 2.19: UV degradation of selected functionalized polyketones.

2.2.3.2 Triarylamine-functionalized Polyketones

Triarylamine-derived polymers 12a-e all form stable radical cations upon

oxidation with AgSbF 6. Before oxidation, the polymers resemble the other amine- functionalized polymers (dark brown, glassy solids). The UV-Vis spectra obtained from dilute solutions of 12a-e had absorption maxima in the UV (250 – 350 nm) similar to other pTAAs and are assignable to π-π* transitions from the triarylamine pendants

15, 16, 17 (Figure 2.20). To oxidize the polymers, 2 mL of AgSbF 6 in methylene chloride (26

mM) was added dropwise to dilute solutions of each polymer (~0.01 w/v% in DCM).

Figure 2.20: UV-Vis spectra for polymers 12a-e (A) prior to and (B) after reaction

with AgSbF 6 in dichloromethane. 12a, purple, dashed; 12b, blue, dash-dot; 12c, green, square dot; 12d, red, round dot; 12e, blue, solid. Reprinted with permission

from Samples, E. M. et al., Macromolecules 2018, 51 (22), 9323-9332. Copyright

2018 American Chemical Society.

Each polymer solution initially became dark yellow; within a few minutes each developed its own color ( a, tan/yellow; b, yellow-grey; c, tan/yellow; d, blue-grey; e,

teal). All of the oxidized triarylamines absorbed at ~466 nm, and polymers bearing –OMe

groups ( 12b, 12d, 12e) showed additional absorptions at wavelengths >650 nm.

Electrochemical studies were performed on polymers 12a-e. Initially, solution

phase cyclic voltammetry (CV) was used (Figure 2.21), but this technique was not

sufficiently sensitive to unambiguously conclude that the oxidation was a result of the

functionalization.

Figure 2.21: A compilation of the cyclic voltammograms collected for the

triarylamine-functionalized polyketones 12a-e.

Differential pulse voltammetry (DPV) was used in place of CV. Changes in current ( ∆I) for a given polymer indicate the relative amount of charge transferred, but a quantitative comparison of polymer electroactivities is complicated by effective triarylamine concentration; for a given polymer mass, a higher % incorporation of a triarylamine will result in a higher absolute ∆I value. Each functionalized polymer 12a-e showed distinct electrochemical signatures absent in the parent polyketone PK6 between

0.5 and 1.2 V (vs. Ag/AgCl). Electrochemical behavior of the triarylamine-functionalized 49 polyketones depends upon the structure of the polymer’s pendant group, indicating the oxidations are occurring at or around the triarylamine moieties (Figure 2.22 - Figure

2.27). Further, the oxidation features of the triarylamine polymers are similar to the behavior of molecular model compounds (section 2.2.3.3 below).

Voltammetric features of polymers 12a-e resemble oxidation behavior of other

triarylamines. For example, the strong influence of pendant functional groups upon

oxidation potentials is consistent with profiles for other polymeric triarylamines

(pTAAs). 53 Oxidation of the pTAA nitrogen centers typically occurs 0.6 – 1.2 V vs.

Ag/AgCl, depending on aryl substituent. Monomeric triphenylamine itself undergoes oxidation at ca. 1.0 V 58 (vs. Ag/AgCl unless otherwise noted), and polymers incorporating a N,N-diphenylamino substituent show oxidations between 0.8 V and 1.1

V. 76 59 The features of 12a and 12e voltammograms just above 1.0 V are therefore

consistent with N-centered triarylamine oxidation. Methyl-substituted (pTAAs) oxidize at

slightly lower potentials (ca. 0.8 V 60, 61 ); similarly, methyl-substituted 12c and 12d show

oxidations at ca. 0.85 V. One of the two features of 12b between 0.6 – 0.9 V could be N-

centered triarylamine oxidation; methoxy-substituted (pTAAs) oxidize at ca. 0.65 V. 61

Figure 2.22: Differential pulse voltammogram for PK6.

Figure 2.23: Differential pulse voltammogram for 12a.

Figure 2.24: Differential pulse voltammogram for 12b.

Figure 2.25: Differential pulse voltammogram for 12c.

Figure 2.26: Differential pulse voltammogram for 12d.

Figure 2.27: Differential pulse voltammogram for 12e.

Complex oxidative features were elucidated with DPV, and each of the triarylamine-functionalized polymers yielded a characteristic voltammogram suggesting more complex behavior than found in other molecular and polymeric triarylamines. In the case of unsubstituted or p-methyl-substituted p-TAAs, oxidation can induce intra- and

intermolecular coupling between the aryl rings, forming carbazole units or dimers, which

themselves can be oxidized. 62 For intermolecular coupling, this can cause cross-linking

of polymer chains and precipitation of the polymer, but no solubility changes were

observed upon oxidation of 12a-e.

Polymer 12b is of particular interest for its N,N-bis (p -dimethoxyphenyl)amino

substitution pattern, a common motif among triarylamines used in devices. 6, 7 For

example, the widely-employed Spiro-OMeTAD hole transport material features four such

units per molecule. 6, 7, 63 McGehee and coworkers 63 have previously studied Spiro-

OMeTAD oxidation with a silver oxidant as well as electrochemical oxidation using

solution-phase DPV. In doing the same, we found that for Spiro-OMeTAD there are three

major oxidation features between 0.25 and 1.25 V vs. Ag/AgCl (Figure 2.28).

Figure 2.28: DPV voltammograms contrasting Spiro-OMeTAD and polymer 12b

oxidation. The dotted lines are to highlight the anodic differences between the two

materials.

12b exhibits oxidative features at similar potentials. Higher currents were observed for Spiro-OMeTAD than 12b since the effective concentration of oxidizable pendants is low. While the functionalization of 12b is random, a fraction of pendants are likely close enough to one another to influence oxidation behavior, either as neighbors along a single polyketone chain or via through-space interactions between polymer chains.

It has been shown previously that π stacking has a marked influence on lowering oxidation potentials of thiophene oligomers, 64 poly(dibenzofulvene)s, 65, 66, 67 and similar pTAAs. 68 Therefore, precise redox potentials would be influenced by π stacking of the 55 triarylamine residues of our materials. 64, 67, 69 This phenomenon is evidenced when single molecules are contrasted against the polymeric forms (section 2.2.3.3 below).

Additionally, pyrrole oxidation is possible and may further complicate electrochemical behavior. A strongly oxidizing environment could form radical cations centered at the pyrrole rings, which could couple to oxidized triarylamine pendants within the same N- arylpyrrole unit or to radical cations nearby. 70 Cai et al. 59 propose complex oxidation behavior at N-(triarylamino)pyrrole units because of oxidations at both the pyrrole and

triarylamine centers.

Further insight was gained about the voltammogram of 12b when a random

tricopolymer was synthesized from PK6 , aniline and the deprotected 11b. The resulting tricopolymers trico1 , trico2 and trico3 are shown in Figure 2.29. Despite varying the stoichiometry of aniline and 11b , the previously-optimized conditions consistently provided an approximate 3:1 phenyl to triarylamine ratio. The overall % incorporation

(the amount of 1,4-diketones converted to an N-aryl pyrrole) ranged from 10-25%, allowing for a comparison of overall triarylamine incorporation into the polymer. With increased % incorporation of the triarylamine unit, there is an increase in ∆I of the shoulder at 0.75 V. This feature may reflect a localized decrease in oxidation potential due to proximity effects between triarylamines. Figure 2.30 compares the voltammograms of trico1 -3 and 12b with the N-phenylpyrrole-functionalized poly(1- hexene-alt -CO) ( 3). While triarylamine-containing polymers demonstrate a pronounced oxidation feature near 0.9 V, 3 exhibits a ∆Imax at a higher potential (1.15 V). Since 3 does not exhibit features 0.8 – 1.0 V, the pyrrole ring and ketone functional groups are

56 not alone responsible for the oxidation features observed with triarylamine polymers. 3 does show a minor feature at 0.5 V, similar to 12b, and so this feature may be due to backbone or N-arylpyrrole oxidation.

Figure 2.29: A random tricopolymer was prepared by reaction with aniline and 11b.

The ratio of aniline:11b minimally affected the ratio n:o .

Table 2.2: Percent incorporation for each tricopolymer. Amine nucleophilicity had

little influence on the actual polymer composition.

% incorp ratio added PhNH 2:11b (total) n:o trico1 2:1 10 3.1:1 trico2 1:1 25 2.8:1 trico3 1:2 16 2.7:1

Figure 2.30: DPV voltammogram for each of the tricopolymers 1, 2, 3, and polymers

3 and 12b. Reprinted with permission from Samples, E. M. et al., Macromolecules

2.2.3.3 Substituted Pyrrole Model Compounds

The redox character of the triarylamine polymer pendants was compared to molecular models without the polymer scaffold. Several N-arylpyrrolyl compounds were synthesized to contain the same redox moieties as the polymers 12a-e. Compounds 15a-e

58 were prepared via Buchwald-Hartwig amination of the N-(p-bromophenyl)pyrrole (14 ,

Figure 2.31 ) with commercially-available diarylamines (Figure 2.32).

Figure 2.31: Paal-Knorr reaction between 2,5-hexanedione and 4-bromoaniline

yielding 14.

Figure 2.32: Amination procedure to prepare model compounds 15a-e.

The effects of aryl substitution of 15a-e upon oxidation behavior were compared to polymers 12a-e to understand the origins of anodic features in the DPV electrochemistry as well as the UV-Vis behavior of the chemically oxidized polymers.

Once prepared, 15a-e were easily oxidized with a dilute solution of AgSbF 6 (26 mM in methylene chloride).

The oxidized polymers 12a-e and the oxidized molecular analogues 15a-e all absorb strongly at ~450 nm (

Table 2.3). UV-Vis spectra for 15a-e are shown in Figure 2.33. UV-Vis spectra

for polymers 12b, 12d, and 12e and molecules 15b and 15e feature an additional absorption band at ~750 nm – an absorption observed in other –OMe substituted triarylamines. 17 The molecular oxidized triarylamine absorptions show negligible change

versus the analogous polymer immobilized triarylamines. The substitution pattern and

oxidized absorption maxima do not follow a clear trend. The –OMe substituted materials

were expected to be more red shifted than the alkyl-substituted materials (red absorption

wavelength in the order of dimethoxy > methoxy-methyl > methoxy), 17, 71 but this relationship is not observed, perhaps due to the additional influence of the pyrrole moiety.

Table 2.3: Absorption maxima for the neutral and oxidized polymers and molecular

models. Absorption maxima for the neutral and oxidized polymers and molecular

models.

Neutral ( λmax , nm) Oxidized ( λmax , nm) Polymer Molecule Polymer Molecule (12a-e) (15a-e) (12a-e) (15a-e) Diphenyl ( 12a , 15a ) 293 303 463 445 465, Dimethoxy ( 12b , 15b ) 295 298 445, 767 767 Dimethyl ( 12c , 15c ) 302 304 465 441 Methoxy-methyl ( 12 d, 465, 300 300 470 15d ) 739 465, Methoxy ( 12e, 15e ) 300 302 448, 766 785

The electrochemical oxidation (DPV) features of the triarylamine model compounds differ from the corresponding polymers. For example the voltammograms for

12b and 15b (Figure 2.34) clearly indicate substantial changes in properties of

triarylamine upon polymer incorporation. The more complex features of polymer 12b

(indicated by *) are absent in molecule 15b. Potentially, this is due to poor oxidative

stability of molecular pyrroles relative to moieties isolated in the polymer scaffolds. 72

Additionally, low concentrations of molecule 15b do not permit interactions between triarylamines, as proposed for the 12b polymer.

Figure 2.33: Normalized UV-Vis spectra for 15a-e. For each spectrum the blue

(solid) trace represents the neutral molecule and the red (dotted) trace represents

the oxidized species after reaction with AgSbF 6 in dichloromethane.

Oxidation of the polymeric triarylamines occurred at lower potentials than molecular analogues. This is not unexpected; appending chromophores to polymer backbones is known to reduce the anodic potentials.64, 65, 66, 67, 68 Previous studies have also reported a dependence on substitution pattern for the relative lifetimes of the formed radicals before intermolecular coupling reactions quench radical species. 62, 73

Figure 2.34: Voltammogram comparison for 12b and 15b. Reprinted with

permission from Samples, E. M. et al., Macromolecules 2018, 51 (22), 9323-9332.

A dependence on substitution and oxidizing potential was observed for all of the molecular model compounds (Figure 2.35). Oxidative events occurring between 0.25 and

0.75 V are reasonably assigned to the oxidation of the pyrrole ring itself 72 with the later features assigned to the oxidation of the triarylamines. Pyrrole oxidation potentials are strongly dependent on substituent and factor strongly into the potentials required for pyrrole electropolymerization. 72

Figure 2.35: DPV voltammograms for each of the molecular models 15a-e.

2.3 Conclusion

The goal of the present chapter was to develop a polymer scaffold that could function as a hole transport material in optoelectronic devices like perovskite solar cells.

Reductive amination of the carbonyls of PK6 with simple substrates was unsuccessful

leading to the use of Paal-Knorr reactions. Consequently, the modularity of the

polyketone scaffold was highlighted through the Paal-Knorr cyclization reaction. Several

arylamines were successfully appended to the PK6 , and it was encouraging to

successfully attach sterically hindered triarylamines to the polymer with good conversion

(15-33 %). Unfortunately, there was not an obvious correlation between triarylamine

nucleophilicity and conversion of PK6 .

Polymers 12a-e UV spectra were drastically changed upon reaction with AgSbF 6, and their oxidized spectra support the formation of radical cationic species. Cyclic voltammetry was initially used to probe the oxidation potentials of 12a-e, but CV was not

sufficient to unambiguously conclude that the oxidation was a result of the

functionalization. During the DPV studies of polymers 12a-e it was found that the entire pyrrole moiety was involved in the oxidation of the pendants beginning with the pyrrole ring itself (~0.5 V vs. Ag/AgCl). Further oxidation potentials depended on triarylamine environment (i.e. stacked or isolated pendants) as well as substitution. In comparison to compounds 15a-e the polymeric triarylamines were more facile to oxidize – the pyrrole oxidation was shifted from ~0.64 V to ~0.44 V. Electrochemical studies of polymers 12a-

e support the functionalization of the polyketone and provided encouraging results for the

modest functionalization of PK6 with the triarylamines. 12a-e produced stable and

observable radical cations, and, specifically, 12b yielded similar oxidation potentials to

the State-Of-The-Art hole transport material Spiro-OMeTAD.

Further functionality was observed from polymer 13 . This material was the least

converted polyketone, but provided significant protection from ultraviolet-induced

radical degradation; PK6 retained only ~20 % its original molecular weight after four

hours of irradiation whereas 13 retained >50 % its original molecular weight.

Additionally, the modest azobenzene functionalization of 13 led to an observable and

reversible trans → cis isomerization induced by the same UV light which degraded PK6 .

2.4 Experimental

General Remarks. All reactions were carried out in an inert atmosphere glovebox or using standard Schlenk techniques unless otherwise noted. NMR spectra are referenced to deuterated solvent (e.g. CDCl 3 corrected to 7.26 ppm) or external standards

(31 P NMR, 85% phosphoric acid solution) and chemical shifts reported as ppm.

Chemicals. Chemicals were used from Sigma Aldrich Chemical Co., Oakwood

Chemicals, Airgas, Kodak, and Tokyo Chemical Industry Co. without further purification

with the exception of p-bromoaniline (Kodak) which was sublimed prior to use. Solvents

used during air-free reactions were purified on a solvent purification system.

Apparatus. All NMR spectra were collected at room temperature on a Bruker

400 spectrometer (400 MHz) and a Bruker Avance 500 (500 MHz). Infrared spectra were collected as ATR-FTIR spectra using a Thermo Scientific Nicolet is5 with a id5 ATR attachment. UV-Vis spectra were collected using a Shimadzu UV-1800 double beam spectrophotometer. Polymerizations were carried out in a 100 mL Parr Instrument Co.

4590 micro stirred reactor. GPC was performed with a Shimadzu LC 20-AT chromatograph using three PolarGel-M (300x7.5 mm) columns, and THF as an eluent.

Calibration was done using standard PMMA samples (Mp range: 202 – 71,800 Da).

Solvents were purified by Pure Process Technology purification system.

Electrochemical measurements. Differential pulse voltammetry (DPV) measurements were performed using 650E CH instruments potentiostat.

Polymers/molecular analogs of interest were dissolved in dichloromethane with 0.1M tetrabutyl ammonium hexafluorophosphate (NBu 4PF 6) as a supporting electrolyte. Glassy carbon, silver wire, and platinum wire were used as a working, reference, and counter

67 electrode, respectively. DPV was conducted by applying potential pulses of 50 mV amplitude for 50 ms duration over 500 ms period (potential waveform Figure 2.36)

Figure 2.36: Potential waveform used for DPV experiments. Potential pulses of 50

mV amplitude for 50 ms duration over 500 ms period were applied.

Synthesis of [Pd(CH 3CN) 4](BF 4)2. Palladium sponge (0.802 mg, 7.53 mmol)

was roughened with a mortar and pestle and added to a 100 mL round bottom flask.

Nitrosomiumtetrafluoroborate (NOBF 4, 1.78 g, 15.3 mmol) was added to the flask

followed by 50 mL of acetonitrile. The flask was sealed with a pierced septum and the

reaction was allowed to stir overnight. The dark, clear yellow solution was decanted from

any unreacted palladium, and the mother liqueour was reduced under vacuum. Addition

68 of diethyl ether precipitated the product which was collected on a glass frit and washed further with diethyl ether yielding 3.07 g (90 %) of [Pd(CH 3CN) 4](BF 4)2 as a pale yellow

1 11 powder. H NMR (500 MHz, CD 3NO 2) δ 2.65 (s, 3H). B NMR (160 MHz, CD 3NO 2) δ -

1.25.

Synthesis of [(Ph) 2P(CH 2)3P(Ph) 2Pd(NCMe) 2](BF 4)2. To a 100 mL flask

[(NCMe) 4Pd](BF 4)2 (0.533 g, 1.2 mmol), diphenylphosphinopropane (0.494 g, 1.2 mmol), and a magnetic stir bar were added. Acetonitrile (50 mL) was then added to the flask to produce a clear, bright yellow solution. The reaction was left to stir rapidly for 24 h. Once complete, the solvent was removed under vaccum to produce

[(Ph) 2P(CH 2)3P(Ph) 2Pd(NCMe) 2](BF 4)2 as a bright yellow film (0.831 g, 89.3 % yield).

1 H NMR (400 MHz, CDCl 3) δ 7.67 (dd, 7H), 7.49 (d, 3H), 2.93 (s, 4H), 2.30 (s, 2H),

31 1.90 (s, 6H). P NMR (202 MHz, CD 2Cl 2) δ 12.47.

General copolymerization of α-olefin and carbon monoxide (PK6). The procedure was adapted from Abu-Surrah et al. 74 A pressure reactor was charged with

[(Ph) 2P(CH 2)3P(Ph) 2Pd(NCMe) 2](BF 4)2 (0.1 mol%, 0.0457 g, 0.0595 mmol) followed by

dichloromethane (10 mL), 1-hexene (7.43 mL, 59.5 mmol), and methanol (358 µL). The

reactor was sealed and removed from the glovebox. The reactor was pressurized with

carbon monoxide (1000 psi), and the reaction stirred at room temperature for 48 h. Once

complete, the reaction solution was depressurized and quenched with an excess of

methanol followed by removal of catalyst residues via silica gel plug. Solvent was

removed under vacuum to yield the product as a translucent gel (2.05 g, 31 % yield). 1H

NMR (500 MHz, CDCl 3): δ 3.09 (br), 2.90 (br), 2.65 (br), 2.57 (br), 2.38 (br), 2.36 (br),

1.61 (br), 1.25 (br), 0.88 (br).

Tert -butyl(4-bromophenyl)carbamate (9). Exclusion of air was not necessary.

To a 250 mL round bottom flask equiped with a stir bar freshly sublimed 4-bromoaniline

(12.6 g, 73.1 mmol) was added followed by THF (80 mL). Once a homogeneous solution

was formed, a solution of di-tert -butyl-dicarbamate (17.7 g, 81.1 mmol) in THF (70 mL)

was added dropwise via addition funnel over a course of 30 min. After addition was

complete the addition funnel was exchanged for a condenser, and the reaction was

refluxed for 24 h or until judged complete via TLC (1:10 ethyl acetate:hexanes). The

product was then purified via flash chromatography to afford 3 as a white solid (4.18 g,

1 86.4 % yield). H NMR (500 MHz, CD 2Cl 2): δ 7.43 – 7.23 (m, 4H), 6.57 (s, 1H), 1.49 (s,

9H).

N-(4-bromophenyl)-2,5-dimethylpyrrole (14). Prepared via the work of Chen et al. without exclusion of air. 42 2,5-hexanedione (5 mmol), 4-bromoaniline (5 mmol), and

Sc(OTf) 3 (0.05 mmol, 1 mol%) were added to a round bottom flask equiped with a stir

bar. The reaction stirred at 30 °C for 30 min or until complete. After the required time a

tan solid was collected via filtration and washed with cold water to yield an off-white

1 solid (1.04 g, 67 % yield). H NMR (400 MHz, CD 2Cl 2): 7.60 (d, 2H), 7.09 (d, 2 H),

5.83 (s, 2H), 2.00 (s, 6H).

General cross-coupling procedure for triarylamines (11 and 15). To a Schlenk

flask 9 (1.37 g, 5.03 mmol), diphenylamine (0.425 g, 2.52 mmol),

tris(dibenzylideneacetone)dipalladium (0.184 g, 0.201 mmol, 4 mol%), tri(tert-

70 butyl)phosphine (0.163 g, 0.805 mmol, 16 mol%), potassium tert-butoxide (1.13 g, 10.1 mmol), and toluene (15 mL) were added. The flask was removed from the glovebox and heated to 111 °C under nitrogen for 24 h or until judged complete by TLC. Once completed, the reaction was cooled to room tempertaure and extracted via aqueous workup. The crude product was purified via flash chromatography (1:10 ethyl acetate:hexanes) and dried under vacuum to afford 11a as an opalescent foam (0.624 g 70 % yield). 1H

NMR (500 MHz, CD 2Cl 2) δ 7.33 – 7.19 (m, 6H), 7.11 – 6.87 (m, 8H), 6.52 (s, 1H), 1.49

(s, 9H).

1 15a. H NMR (400 MHz, CD 2Cl 2): = 2.04 (s, 6H), 5.81 (s, 2H), 7.02-7.14 (m,

10H), 7.27-7.31 (m, 4H)

13 C NMR (101 MHz, CD 2Cl 2): = 148.12, 129.92, 129.40, 125.23, 123.85,

123.65, 105.99, 100.18, 13.28.

1 15b. H NMR (400 MHz, CD 2Cl 2): = 2.08 (s, 6H), 3.79 (s, 6H), 5.78 (s, 2H),

6.85-6.96 (m, 8H), 7.11 (d, 4H, J = 9.1 Hz)

13 C NMR (101 MHz, CD 2Cl 2): = 156.91, 141.07, 129.08, 127.61, 119.91,

115.29, 105.76, 100.16, 55.99, 13.26.

1 15c. H NMR (400 MHz, CD 2Cl 2): = 2.02 (s, 6H), 2.32 (s, 6H), 5.79 (s, 2H),

6.99-7.03 (m, 8H), 7.11 (d, 4H, J = 8.0 Hz)

13 C NMR (101 MHz, CD 2Cl 2): = 45.57, 130.53, 129.19, 125.56, 105.87,

100.18, 21.08, 13.26

1 15d. H NMR (400 MHz, CD 2Cl 2): = 2.02 (s, 6H), 2.31 (s, 3H), 3,79 (s, 3H),

5.79 (s, 2H), 6.87 (d, 2H, J = 9.1 Hz), 6.96 (s, 4H), 7.02 (d, 2H, J = 8.5 Hz), 7.10 (d, 4H,

J = 9.0 Hz)

13 C NMR (101 MHz, CD 2Cl 2): = 157.07, 145.66, 140.97, 133.29, 132.12,

130.47, 129.14, 128.00, 125.05, 121.11, 115.33, 105.82, 100.17, 55.99, 21.16, 13.27

1 15e. H NMR (400 MHz, CD 2Cl 2): = 2.03 (s, 6H), 3.80 (s, 3H), 5.79 (s, 2H),

6.89 (d, 2H, J = 9.0 Hz), 6.98-7.02 (m, 5H), 7.09-7.13 (m, 4H), 7.24-7.28 (m, 2H)

13 C NMR (101 MHz, CD 2Cl 2): = 157.27, 148.35, 148.08, 140.79, 132.59,

129.77, 129.24, 128.31, 124.15, 123.13, 122.11, 115.40, 105.89, 55.99, 13.27

4-aminotriphenylamine·HCl. 11a (0.624 g, 1.73 mmol) is added to a round

bottom flask and a pre-mixed solution of 30% HCl in ethyl acetate (5 mL) is slowly

added to dissolve the solid. Once homogeneous, the reaction was concentrated to induce

precipitation. The solid was filtered and washed with water to afford a beige solid (0.342

1 g, 66.5 % yield). H NMR (500 MHz, (CD 3)2SO) δ 9.67 (s, 3H), 7.35 – 7.27 (m, 4H),

7.22 – 7.15 (m, 2H), 7.10 – 6.97 (m, 8H).

Preparation of pyrrole-functionalized polyketone (1-8, 12a-e, and 13).

Exclusion of air was not necessary. Traditional Paal-Knorr reactions. To a solution of

PK6 (2.7 mmol, 1 eq) in toluene (40 mL), 4-bromoaniline (2.9 mmol, 1.1 eq) was added.

A reflux condenser was attached to the reaction flask and the assembly was lowered into

a 100 °C preheated oil bath. The reaction was allowed to reflux for 24 h. After this time

the polymer was extracted with an acidic aqueous workup to remove excess 4-

bromoaniline. The resulting dark brown polymer was dried under vacuum and analyzed

72 via 1H NMR for percent conversion. For this uncatalyzed reaction a percent conversion of

5.6 % was obtained.

Lewis acid-catalyzed Paal-Knorr reactions. PK6 (0.36 g, 3.2 mmol) was dissolved in methylene chloride (10 mL), and, once dissolved, freshly distilled aniline

(0.275 mL, 3.0 mmol) and Sc(OTf) 3 (0.14 g, 5 mol%) were added. The reaction was heated to 40 °C for 24 h. When the reaction was halted, catalyst solids were removed via filtration and solvents removed under vacuum to yield crude polymer. Purification of the pyrrole-functionalized polymer was achieved by repeated precipitation from methylene chloride with methanol to yield 3 as a dark brown solid. If triarylamines were used ( 11a-

e), the triarylamine·HCl was extracted with base (2 M Na 2CO 3 (aq) ) prior to use.

Calculation of Paal-Knorr conversion. Proton NMR was used to calculate the

percent conversion of PK6 . The polymer 3 1H NMR spectrum is shown in Figure 2.37.

For each Paal-Knorr functionalized polyketone, % conversion was calculated as follows:

Since the polyketone backbone signals are contained between 3.2 ppm and 0.7 ppm, any

aromatic signals can be attributed to the functionalization of the polyketone – residual

amine reagent is identifiable by discrete, sharp peaks in the 1H NMR, and the functionalized polymers are purified until all unreacted amine is removed. Using a known mass of 3 (20.4 mg) and a known mass of an internal standard (1,3,5-trimethoxybenzene,

TMB, 6.7 mg; MW 168.19)) the aromatic protons were integrated for the internal standard and the polymer.

1 Figure 2.37: Polymer 3 annotated H NMR spectrum in CD 2Cl 2. 1,3,5-trimethoxy

benzene was used as an internal standard to calculate the percent conversion of

PK6.

Equation 1 was used with the integrations of 3 to determine the ratio which

represents the ratio between the analyte a (the aromatic units of 3) and the internal

standard is . Using Equation 2 with nis moles of TMB we can find the moles na for the analyte. The moles of analyte were converted to milligrams using the molecular weight for the functionalized unit of the polymer (the repeat unit n in Figure 2.37); 3 has a MW

= 197.28. The milligrams of pyrrolic units in the NMR sample were compared to the 74 mass of the total sample to obtain the percent conversion. For this example we used 20.4

-5 mg of 3 and 6.7 mg ( nis = 3.9e ) of TMB. We obtained = 0.306. After using

-5 Equation (2) we obtained na = 1.2e . Using the repeat unit n molecular weight of 197.28

we found the polyketone had been functionalized to 12 %.

= (3)

= (4)

Photoisomerization of polyaminoazopyrrole (13). 13 was dissolved in ethyl acetate and sealed in a cuvette. The cuvette was affixed to the water jacket which was used to cool the lamp, and the solution was irradiated for 60 mins with a medium pressure mercury lamp (400 W; 280 – 600 nm with a Pyrex filter) to induce photoisomerization. A UV-vis spectrum was collected immediately after irradiation.

Irradiating the sample with visible light from a 5.88 W blue LED (440-460 nm; 447.5

λmax ) for 60 mins induced the cis → trans isomerization; UV-vis spectrum was collected

immediately after irradiation. The assembly consisted of 7 LEDs arranged on a disk, and

the disk was mounted to a heat sink. The 5.88 W LEDs (440-460 nm; 447.5 λmax ) were powered by 24 V power supply and a LUXdrive 700 mA FlexBlock constant current driver (part no. A011-D-V-700). The LED assembly is described in Tucker et al. 75

2.5 References Cited

1. Saliba, M.; Orlandi, S.; Matsui, T.; Aghazada, S.; Cavazzini, M.; Correa-Baena, J. P.; Gao, P.; Scopelliti, R.; Mosconi, E.; Dahmen, K. H.; De Angelis, F.; Abate, A.; Hagfeldt, A.; Pozzi, G.; Graetzel, M.; Nazeeruddin, M. K., A molecularly engineered hole-transporting material for efficient perovskite solar cells. Nature Energy 2016, 1.

2. Guo, Y.; Liu, C.; Inoue, K.; Harano, K.; Tanaka, H.; Nakamura, E., Enhancement in the efficiency of an organic-inorganic hybrid solar cell with a doped P3HT hole- transporting layer on a void-free perovskite active layer. Journal of Materials Chemistry A 2014, 2 (34), 13827-13830.

3. Ahn, N.; Son, D.-Y.; Jang, I.-H.; Kang, S. M.; Choi, M.; Park, N.-G., Highly Reproducible Perovskite Solar Cells with Average Efficiency of 18.3% and Best Efficiency of 19.7% Fabricated via Lewis Base Adduct of Lead(II) Iodide. Journal of the American Chemical Society 2015, 137 (27), 8696-8699.

4. Huang, C.; Fu, W.; Li, C.-Z.; Zhang, Z.; Qiu, W.; Shi, M.; Heremans, P.; Jen, A. K. Y.; Chen, H., Dopant-Free Hole-Transporting Material with a C3h Symmetrical Truxene Core for Highly Efficient Perovskite Solar Cells. Journal of the American Chemical Society 2016, 138 (8), 2528-2531.

5. Melas-Kyriazi, J.; Ding, I. K.; Marchioro, A.; Punzi, A.; Hardin, B. E.; Burkhard, G. F.; Tétreault, N.; Grätzel, M.; Moser, J.-E.; McGehee, M. D., The Effect of Hole Transport Material Pore Filling on Photovoltaic Performance in Solid-State Dye- Sensitized Solar Cells. Advanced Energy Materials 2011, 1 (3), 407-414.

6. Agarwala, P.; Kabra, D., A review on triphenylamine (TPA) based organic hole transport materials (HTMs) for dye sensitized solar cells (DSSCs) and perovskite solar cells (PSCs): evolution and molecular engineering. Journal of Materials Chemistry A 2017, 5 (4), 1348-1373.

7. Calió, L.; Kazim, S.; Grätzel, M.; Ahmad, S., Hole-Transport Materials for Perovskite Solar Cells. Angewandte Chemie International Edition 2016, 55 (47), 14522- 14545.

8. Leijtens, T.; Ding, I. K.; Giovenzana, T.; Bloking, J. T.; McGehee, M. D.; Sellinger, A., Hole Transport Materials with Low Glass Transition Temperatures and 76

High Solubility for Application in Solid-State Dye-Sensitized Solar Cells. ACS Nano 2012, 6 (2), 1455-1462.

9. Tung, N. T.; Nguy ễn, N.; Nguyen, T., Glass transition of PCBM, P3HT and their blends in quenched state. Adv Nat Sci Nanosci Nanotechnol 3(4):045001 . 2012 ; Vol. 3, p 045001.

10. Xie, R.; Lee, Y.; Aplan, M. P.; Caggiano, N. J.; Müller, C.; Colby, R. H.; Gomez, E. D., Glass Transition Temperature of Conjugated Polymers by Oscillatory Shear Rheometry. Macromolecules 2017, 50 (13), 5146-5154.

11. Müller, C., On the Glass Transition of Polymer Semiconductors and Its Impact on Polymer Solar Cell Stability. Chemistry of Materials 2015, 27 (8), 2740-2754.

12. Hsiao, S.-H.; Hsiao, Y.-H., Synthesis and electrochemical properties of new redox-active polyimides with (1-piperidinyl)triphenylamine moieties. High Performance Polymers 2017, 29 (4), 431-440.

13. Ishida, N.; Wakamiya, A.; Saeki, A., Quantifying Hole Transfer Yield from Perovskite to Polymer Layer: Statistical Correlation of Solar Cell Outputs with Kinetic and Energetic Properties. ACS Photonics 2016, 3 (9), 1678-1688.

14. Kisselev, R.; Thelakkat, M., Synthesis and Characterization of Poly(triarylamine)s Containing Isothianaphthene Moieties. Macromolecules 2004, 37 (24), 8951-8958.

15. Matsui, T.; Petrikyte, I.; Malinauskas, T.; Domanski, K.; Daskeviciene, M.; Steponaitis, M.; Gratia, P.; Tress, W.; Correa-Baena, J.-P.; Abate, A.; Hagfeldt, A.; Grätzel, M.; Nazeeruddin, M. K.; Getautis, V.; Saliba, M., Additive-Free Transparent Triarylamine-Based Polymeric Hole-Transport Materials for Stable Perovskite Solar Cells. ChemSusChem 2016, 9 (18), 2567-2571.

16. Schlotthauer, T.; Schroot, R.; Glover, S.; Hammarström, L.; Jäger, M.; Schubert, U. S., A multidonor–photosensitizer–multiacceptor triad for long-lived directional charge separation. Physical Chemistry Chemical Physics 2017, 19 (42), 28572-28578.

17. Schroot, R.; Schubert, U. S.; Jäger, M., Block Copolymers for Directional Charge Transfer: Synthesis, Characterization, and Electrochemical Properties of Redox-Active Triarylamines. Macromolecules 2015, 48 (7), 1963-1971.

18. Sprick, R. S.; Hoyos, M.; Navarro, O.; Turner, M. L., Synthesis of poly(triarylamine)s by C–N coupling catalyzed by (N-heterocyclic carbene)-palladium complexes. Reactive and Functional Polymers 2012, 72 (5), 337-340.

19. Bach, U.; Lupo, D.; Comte, P.; Moser, J. E.; Weissörtel, F.; Salbeck, J.; Spreitzer, H.; Grätzel, M., Solid-state dye-sensitized mesoporous TiO2 solar cells with high photon- to-electron conversion efficiencies. Nature 1998, 395 , 583.

20. Poplavskyy, D.; Nelson, J., Nondispersive hole transport in amorphous films of methoxy-spirofluorene-arylamine organic compound. Journal of Applied Physics 2003, 93 (1), 341-346.

21. Zhang, W.; Cheng, Y.; Yin, X.; Liu, B., Solid-State Dye-Sensitized Solar Cells with Conjugated Polymers as Hole-Transporting Materials. Macromolecular Chemistry and Physics 2011, 212 (1), 15-23.

22. Sen, A., Mechanistic aspects of metal-catalyzed alternating copolymerization of olefins with carbon monoxide. Accounts of Chemical Research 2002, 26 (6), 303-310.

23. Zehetmaier, P. C.; Vagin, S. I.; Rieger, B., Functionalization of aliphatic polyketones. MRS Bulletin 2013, 38 (03), 239-244.

24. Beller, M.; Steinhoff, B. A.; Zoeller, J. R.; Cole ‐Hamilton, D. J.; Drent, E.; Wu, X.; Neumann, H.; Ito, S.; Nozaki, K., In Applied Homogeneous Catalysis with Organometallic Compounds , Cornils, B.; Herrmann, W. A.; Beller, M.; Paciello, R., Eds. Wiley-VCH: Weinheim, 2017; pp 91-190.

25. Brookhart, M.; Rix, F. C.; DeSimone, J. M.; Barborak, J. C., Palladium(II) catalysts for living alternating copolymerization of olefins and carbon monoxide. Journal of the American Chemical Society 1992, 114 (14), 5894-5895.

26. Nozaki, K.; Kawashima, Y.; Nakamoto, K.; Hiyama, T., Alternating Copolymerization of Various Alkenes with Carbon Monoxide Using a Cationic Pd(II) Complex of (R,S)-BINAPHOS as an Initiator. Polymer Journal 1999, 31 (11), 1057. 78

27. Nozaki, K.; Hiyama, T., Stereoselective alternating copolymerization of carbon monoxide with alkenes. Journal of Organometallic Chemistry 1999, 576 (1), 248-253.

28. Drent, E.; Budzelaar, P. H. M., Palladium-Catalyzed Alternating Copolymerization of Alkenes and Carbon Monoxide. Chemical Reviews 1996, 96 (2), 663.

29. Beckmann, U.; Eichberger, E.; Rufinska, A.; Sablong, R.; Klaui, W., Nickel(II) catalysed co-polymerisation of CO and ethene: Formation of polyketone vs. polyethylene – The role of co-catalysts. Journal of Catalysis 2011, 283 (2), 143-148.

30. Pérez ‐Foullerat, D.; Meier, U. W.; Hild, S.; Rieger, B., High ‐Molecular ‐Weight Polyketones from Higher α‐Olefins: A General Method. Macromolecular Chemistry and Physics 2004, 205 (17), 2292-2302.

31. Abdel-Magid, A. F.; Mehrman, S. J., A Review on the Use of Sodium Triacetoxyborohydride in the Reductive Amination of Ketones and Aldehydes. Organic Process Research & Development 2006, 10 (5), 971-1031.

32. Bianchini, C.; Meli, A., Alternating copolymerization of carbon monoxide and olefins by single-site metal catalysis. Coordination Chemistry Reviews 2002, 225 (1–2), 35-66.

33. Cheng, C.; Guironnet, D.; Barborak, J.; Brookhart, M., Preparation and characterization of conjugated polymers made by postpolymerization reactions of alternating polyketones. Journal of the American Chemical Society 2011, 133 (25), 9658- 61.

34. Chen, J.-T.; Yeh, Y.-S.; Sen, A., Synthesis and characterization of N-substituted poly(ethylenepyrrole): functionalization of ethylene–carbon monoxide alternating copolymers. Journal of the Chemical Society, Chemical Communications 1989, (14), 965-967.

35. Zhang, Y.; Broekhuis, A. A.; Stuart, M. C. A.; Picchioni, F., Polymeric amines by chemical modifications of alternating aliphatic polyketones. Journal of Applied Polymer Science 2008, 107 (1), 262-271.

36. Mul, W. P.; Dirkzwager, H.; Broekhuis, A. A.; Heeres, H. J.; van der Linden, A. J.; Guy Orpen, A., Highly active, recyclable catalyst for the manufacture of viscous, low molecular weight, CO–ethene–propene-based polyketone, base component for a new class of resins. Inorganica Chimica Acta 2002, 327 (1), 147-159.

37. Kiovsky, T. E.; Kromer, R. C. Polymeric pyrrollic derivative. US 3979374A, June 13, 1975, 1975.

38. Zhaozhong, J.; Sanganeria, S.; Sen, A., Polymers Incorporating Backbone Thiophene, Furan, and Alcohol Functionalities Formed Through Chemical Modifications of Alternating Olefin-Carbon Monoxide Copolymers. Journal of Polymer Science Part A: Polymer Chemistry 1994, 32 , 841-847.

39. Jiang, Z.; Sanganeria, S.; Sen, A., Polymers incorporating backbone thiophene, furan, and alcohol functionalities formed through chemical modifications of alternating olefin–carbon monoxide copolymers. Journal of Polymer Science Part A: Polymer Chemistry 1994, 32 (5), 841-847.

40. Banik, B. K.; Samajdar, S.; Banik, I., Simple synthesis of substituted pyrroles. Journal of Organic Chemistry 2004, 69 (1), 213-6.

41. Samadi, M.; Behbahani, F. K., The Application of Iron(III) Phosphate in the Synthesis of N-substituted Pyrroles. Journal of the Chilean Chemical Society 2015, 60 (2), 2881-2884.

42. Chen, J.; Wu, H.; Zheng, Z.; Jin, C.; Zhang, X.; Su, W., An approach to the Paal– Knorr pyrroles synthesis catalyzed by Sc(OTf)3 under solvent-free conditions. Tetrahedron Letters 2006, 47 (30), 5383-5387.

43. Flory, P. J., Intramolecular Reaction between Neighboring Substituents of Vinyl Polymers. Journal of the American Chemical Society 1939, 61 (6), 1518-1521.

44. Sreenath, K.; Suneesh, C. V.; Ratheesh Kumar, V. K.; Gopidas, K. R., Cu(II)- Mediated Generation of Triarylamine Radical Cations and Their Dimerization. An Easy Route to Tetraarylbenzidines. The Journal of Organic Chemistry 2008, 73 (8), 3245- 3251.

45. Bandara, H. M. D.; Burdette, S. C., Photoisomerization in different classes of azobenzene. Chemical Society Reviews 2012, 41 (5), 1809-1825.

46. Kosaka, N.; Oda, T.; Hiyama, T.; Nozaki, K., Synthesis and Photoisomerization of Optically Active 1,4-Polyketones Substituted by Azobenzene Side Chains. Macromolecules 2004, 37 (9), 3159-3164.

47. Natansohn, A.; Rochon, P., Photoinduced Motions in Azo-Containing Polymers. Chemical Reviews 2002, 102 (11), 4139-4176.

48. Qiu, F.; Cao, Y.; Xu, H.; Jiang, Y.; Zhou, Y.; Liu, J., Synthesis and properties of polymer containing azo-dye chromophores for nonlinear optical applications. Dyes and Pigments 2007, 75 (2), 454-459.

49. Wen, H.; Zhang, W.; Weng, Y.; Hu, Z., Photomechanical bending of linear azobenzene polymer. RSC Advances 2014, 4 (23), 11776-11781.

50. Takenaka, Y.; Osakada, K.; Nakano, M.; Ikeda, T., A Liquid Crystalline Polyketone Prepared from Allene Having an Azobenzene Substituent and Carbon Monoxide. Macromolecules 2003, 36 (4), 1414-1416.

51. Wang, X., Trans–Cis Isomerization. In Azo Polymers , Springer Berlin Heidelberg: Berlin, Heidelberg, 2017; pp 19-56.

52. R. King, N.; A. Whale, E.; Davis, F. J.; Gilbert, A.; Mitchell, G. R., Effect of media polarity on the photoisomerisation of substituted stilbene, azobenzene and imine chromophores. Journal of Materials Chemistry 1997, 7 (4), 625-630.

53. Samples, E. M.; Schuck, J. M.; Joshi, P. B.; Willets, K. A.; Dobereiner, G. E., Synthesis and Properties of N-Arylpyrrole-Functionalized Poly(1-hexene-alt-CO). Macromolecules 2018, 51 (22), 9323-9332.

54. Koizumi, T.; Hasegawa, Y.; Takata, T.; Endo, T., Synthesis and photodegradation of polyacrylonitrile having ketone group obtained from radical copolymerization of 2,2- diphenyl-4-methylene-1,3-dioxolane with acrylonitrile. Journal of Polymer Science Part A: Polymer Chemistry 1994, 32 (16), 3193-3195.

55. Gooden, R.; Hellman, M. Y.; Hutton, R. S.; Winslow, F. H., Solid-state photochemistry of poly(ethylene-co-carbon monoxide). Model studies of polyethylene photochemistry. Macromolecules 1984, 17 (12), 2830-2837.

56. Yager, K. G.; Barrett, C. J., Azobenzene Polymers for Photonic Applications. In Smart Light ‐Responsive Materials , John Wiley & Sons, Inc.: 2008.

57. Arrington, K. J.; Waugh, J. B.; Radzinski, S. C.; Matson, J. B., Photo- and Biodegradable Thermoplastic Elastomers: Combining Ketone-Containing Polybutadiene with Polylactide Using Ring-Opening Polymerization and Ring-Opening Metathesis Polymerization. Macromolecules 2017, 50 (11), 4180-4187.

58. Thelakkat, M., Star-Shaped, Dendrimeric and Polymeric Triarylamines as Photoconductors and Hole Transport Materials for Electro-Optical Applications. Macromolecular Materials and Engineering 2002, 287 (7), 442-461.

59. Cai, S.; Wang, S.; Wei, D.; Niu, H.; Wang, W.; Bai, X., Multifunctional polyamides containing pyrrole unit with different triarylamine units owning electrochromic, electrofluorochromic and photoelectron conversion properties. Journal of Electroanalytical Chemistry 2018, 812 , 132-142.

60. Choi, K.; Yoo, S. J.; Sung, Y.-E.; Zentel, R., High Contrast Ratio and Rapid Switching Organic Polymeric Electrochromic Thin Films Based on Triarylamine Derivatives from Layer-by-Layer Assembly. Chemistry of Materials 2006, 18 (25), 5823- 5825.

61. Bender, T. P.; Graham, J. F.; Duff, J. M., Effect of Substitution on the Electrochemical and Xerographic Properties of Triarylamines: Correlation to the Hammett Parameter of the Substituent and Calculated HOMO Energy Level. Chemistry of Materials 2001, 13 (11), 4105-4111.

62. Yurchenko, O.; Heinze, J.; Ludwigs, S., Electrochemically Induced Formation of Independent Conductivity Regimes in Polymeric Tetraphenylbenzidine Systems. ChemPhysChem 2010, 11 (8), 1637-1640.

63. Nguyen, W. H.; Bailie, C. D.; Unger, E. L.; McGehee, M. D., Enhancing the Hole-Conductivity of Spiro-OMeTAD without Oxygen or Lithium Salts by Using Spiro(TFSI)2 in Perovskite and Dye-Sensitized Solar Cells. Journal of the American Chemical Society 2014, 136 (31), 10996-11001. 82

64. Knoblock, K. M.; Silvestri, C. J.; Collard, D. M., Stacked Conjugated Oligomers as Molecular Models to Examine Interchain Interactions in Conjugated Materials. Journal of the American Chemical Society 2006, 128 (42), 13680-13681.

65. Luo, J.; Wang, Y.; Nakano, T., Free-Radical Copolymerization of Dibenzofulvene with (Meth)acrylates Leading to π-Stacked Copolymers. Polymers 2018, 10 (654).

66. Nageh, H.; Wang, Y.; Nakano, T., Cationic polymerization of dibenzofulvene leading to a π-stacked polymer. Polymer 2018, 144 , 51-56.

67. Nakano, T.; Yade, T., Synthesis, Structure, and Photophysical and Electrochemical Properties of a π-Stacked Polymer. Journal of the American Chemical Society 2003, 125 (50), 15474-15484.

68. Wang, H.; Wang, Y.; Ye, X.; Hayama, H.; Sugino, H.; Nakano, H.; Nakano, T., Stacked poly(vinyl ketone)s with accumulated push-pull triphenylamine moieties in the side chain. Polymer Chemistry 2017, 8 (4), 708-714.

69. Wang, H.; Wang, Y.; Ye, X.; Hayama, H.; Sugino, H.; Nakano, H.; Nakano, T., pi-Stacked poly(vinyl ketone)s with accumulated push-pull triphenylamine moieties in the side chain. Polymer Chemistry 2017, 8 (4), 708-714.

70. Pittelli, S. L.; Shen, D. E.; Österholm, A. M.; Reynolds, J. R., Chemical Oxidation of Polymer Electrodes for Redox Active Devices: Stabilization through Interfacial Interactions. ACS Applied Materials & Interfaces 2018, 10 (1), 970-978.

71. Jeon, N. J.; Lee, H. G.; Kim, Y. C.; Seo, J.; Noh, J. H.; Lee, J.; Seok, S. I., o- Methoxy Substituents in Spiro-OMeTAD for Efficient Inorganic–Organic Hybrid Perovskite Solar Cells. Journal of the American Chemical Society 2014, 136 (22), 7837- 7840.

72. Diaz, A. F.; Martinez, A.; Kanazawa, K. K.; Salmón, M., Electrochemistry of some substituted pyrroles. Journal of Electroanalytical Chemistry and Interfacial Electrochemistry 1981, 130 , 181-187.

73. Seo, E. T.; Nelson, R. F.; Fritsch, J. M.; Marcoux, L. S.; Leedy, D. W.; Adams, R. N., Anodic Oxidation Pathways of Aromatic Amines. Electrochemical and Electron

Paramagnetic Resonance Studies. Journal of the American Chemical Society 1966, 88 (15), 3498-3503.

74. Abu-Surrah, A. S.; Wursche, R.; Rieger, B., Polyketone materials: control of glass transition temperature and surface polarity by co- and terpolymerization of carbon monoxide with higher 1-olefins. Macromolecular Chemistry and Physics 1997, 198 (4), 1197-1208.

75. W., T. J.; Yuan, Z.; F., J. T.; J., S. C. R., Visible ‐Light Photoredox Catalysis in Flow. Angewandte Chemie International Edition 2012, 51 (17), 4144-4147.

CHAPTER 3

LEWIS ACID EFFECTS ON POLYKETONE SYNTHESIS

3.1 Introduction

1,4-Polyketones have attracted attention for the low cost of monomer feedstock

(carbon monoxide and α-olefin), ease of producing perfectly alternating copolymer structures, and a collection of valuable properties which were described in Chapters 1 and

2. Copolymerization is used widely to design materials with desired properties; this helps minimize the individual monomers’ deficiencies. 1 Typical polyketones use only ethylene or propylene as comonomers; higher olefins have been polymerized previously 2, 3 , but interest has declined. These polymers are prepared almost exclusively by homogeneous cationic palladium catalysts. (Figure 3.1). 4

Figure 3.1: The general structure of a cationic polymerization catalyst. Where

ligands L 1,2 may be monodentate or bidentate, and an open coordination site is

designated by □.

A single component catalyst (SCC) 4 like catalyst 1 in Figure 3.2 directly initiates polymerization by insertion of CO into the Pd-C bond in the presence of monomer, without an induction period. 4 Other precatalysts must be first activated in situ with a 85 cocatalyst, analogous to the alkylaluminum or Lewis acid cocatalysts used in group IV metallocene polymerization. 5, 6, 7, 8, 9, 10 For polyketones, active catalyst species are

formed in situ by ligand exchange, protonolysis or halide abstraction from a precursor

(e.g. (P^P)PdMeCl) with a silver salt in a weakly coordinating solvent like acetonitrile. 4

While not typically used to activate polyketone catalysts, oxidants help initiate polymerization by regenerating active Pd II species, preventing reduction and precipitation of Pd(0).11

Figure 3.2: 1) Example of a SCC used for polyketone synthesis. 2) Catalyst

precursor used herein to copolymerize α-olefins with CO.

Lewis acids can also promote polyketone synthesis, although this area is far less commonly explored. Cooley and coworkers 12 found a significant increase in polymer yield, subtle changes to polymer structure, as well as differences in rate of polymerization when tris(pentafluorophenyl)borane (BCF) was used to activate the catalyst precursor

12, 13 (dppp)Pd(OAc) 2 in the copolymerization of ethene and propene. The observed rate differences depended on the initial cocatalyst concentration, which was partially consumed to form the active species (dppp)Pd(OAc)(C6F5) (Figure 3.9). This work shows

86 the potential of Lewis acids to improve polyketone synthesis. Since it remains the only significant example of a potentially important promoting effect, we became interested in potentially broadening the scope of Lewis acid effects, as well as gaining a better understanding of the underlying mechanistic effects.

In this chapter the Lewis acid activation of catalyst 2 (Figure 3.2) is investigated during the copolymerization of higher α-olefins with CO. By screening various Lewis

acids (LAs) we found we can tune the polyketones’ bulk properties (i.e. molecular

weight, dispersity, and overall yield of the polymers) without significantly changing the

polymers’ structure. In addition to activating a catalyst precursor, we propose that the

relative rates of propagation (kp) or termination (kt) of the growing polymer chains are affected by coordination of the LA to the carbonyls of the growing polymer chain.

3.2 Results and Discussion

In seeking control of yield, chain length, and dispersity of the polyketones,

polyketones were prepared from 1-pentene, 1-hexene, and 1-octene in the presence of

several LAs. The effects of the LAs with 1-hexene was observed, then the generality of

the observed trends was tested with the other two olefins. Figure 3.3 shows the general

method for synthesizing the polymers. [Pd(dppp)(MeCN) 2](BF 4)2 (2) was used to ensure generation of an exclusively alternating polyketone structure 13, 14, 15, 16, 17, 18 and is insensitive to olefin chain lengths. The palladium precursor was prepared according to literature precedent. 18

Stereoregular aliphatic polyketones are known to exist in equilibrium between alternating ketone units and spiroketal units (Figure 3.3). 19 No evidence of spiroketal structures were present in 13 C NMR spectrum (114 ppm) supporting the generation of

alternating polyketones. Polyketones discussed herein were prepared using 1-pentene

(PK5 ), 1-hexene ( PK6 ), and 1-octene ( PK8 ) and polymerized in a stainless steel pressure

reactor for 24 hrs as previously described.18 Since BCF is known to activate palladium

8 precatalysts, catalyst activity (g polymer /mol catalyst ·h) for each of the metal triflate additives was compared to BCF.

Figure 3.3: Synthetic route to polyketone synthesis with olefins larger than ethene or

propene. No spiroketal formation was observed using the present catalytic system.

Polymerizations with additive LAs adhered to several trends. For PK6 the

presence of metal triflates reduced catalyst activity relative to reactions without additive,

yet the polymer yield increased with increasing LA concentration (Figure 3.4). BCF leads

to higher polymer yields than the other LAs. Interestingly, addition of BCF did not affect

catalyst rate until BCF:Pd = 3:1, while adding BCF increased yield in all cases (Figure

3.5). Among the other LAs tested, the next-highest yielding additive was zinc triflate

(Zn(OTf) 2). Even as Zn(OTf) 2 out-performed the other LAs the catalyst activity was 88 initially ca. 1.5 times lower than BCF and quickly reached a difference of ca. 2 times lower catalyst activity at LA:Pd = 7:1 (Figure 3.6).

Figure 3.4: The average catalytic activity of a 24 h reaction for each of the LAs used

during PK6 synthesis.

Figure 3.5: Polyketone yield as a function of BCF loading.

Figure 3.6: Averaged isolated masses of PK6.

Relative to the control (additive-free) reaction, the metal triflates screened all produced lower yields of PK6 (Figure 3.6), lower molecular weights (Figure 3.7), and

higher molecular weight distributions (Ð, Figure 3.8). The dispersity and molecular

weights were relatively less influenced at the same loadings of BCF. Zinc triflate

produced higher molecular weights and slightly higher Ð than BCF, while offering a

lower catalyst activity. Copper triflate provided the lowest Ð and the lowest molecular

weights despite comparable activity and yields to the other LA reactions.

Figure 3.7: Averaged molecular weights of PK6 with different LAs.

Figure 3.8: Molecular weight distributions (Ð) for PK6 using different LAs. PK6

without LA has a Ð of 1.6.

Cooley and coworkers cited aryl transfer to the palladium center from BCF as the initial activation step, leading to a complex mixture of catalytically active species for ethylene/CO polymerization (Figure 3.9). 4, 12 A similar mixture of active species is

reasonably assumed to be present after abstraction of an acetonitrile ligand from 2

lending to the increased polyketone yields. 20, 21

Figure 3.9: Cooley and coworkers’ proposed activation process with BCF beginning

with fluoroaryl transfer to palladium. Reprinted with permission from Barlow, G.

Society.

While BCF is an effective cocatalyst for the activation of certain catalysts, the

initial hypothesis for this project was to shift the relative rates of propagation ( kp) and

termination ( kt) to favor chain growth (Figure 3.10). Internal chelation of the growing

polymer chain inhibits insertion.4, 11 After each insertion of olefin the polymer chain end

acts as a chelating ligand through the β-carbonyl group, which effectively blocks the

fourth coordination site of the metal (3, Figure 3.11). An incoming CO must displace the

chelating polymer chain; then, the resulting 6-membered chelate 4 can easily be broken

by the next incoming olefin. 11, 22 We proposed that adding LAs during polymerization would bind carbonyls of the growing polymer chain (5, Figure 3.11), 23 preventing the 93 closure of the fourth coordination site on the palladium center and ultimately enhancing

23 kp.

Figure 3.10: The relative rates of propagation and termination determine the

polymer molecular weight.

Unexpectedly, metal triflate-promoted reactions increased molecular weight

distributions and reduced molecular weight, suggesting LAs are instead increasing the

rate of termination (kt). Mechanistically, the effect of triflate salts is clearly different from

BCF, which causes a significant increase in yield with only a slight change to molecular

weight and dispersity relative to the control (co-catalyst free) reaction.

Figure 3.11: Internal chelation of the polyketone chains and potential disruption of

chelation by Lewis acid additives.

Ionic strength (µ) could also be influencing polymerization rates. Berchtold et al. demonstrated a solvent polarity rate dependence in norbornene polymerization;8 in our case the addition of the metal triflates increases the ionic strength of the system. Figure

3.12 graphs PK6 yield against ionic strength. The trivalent cations are roughly grouped in

Figure 3.12 below the dotted line and the lower-valent additives are clustered above the dotted line. Zn(OTf) 2 minimally affected the yield, molecular weight, and dispersity compared to the other additives and showed similar behavior to BCF. In the absence of any LA, µ = 1.04x10 -3 mM.

Figure 3.12: Ionic strength influence on PK6 yield. High-valent species reduced the

yield more than low-valent species. *BCF data are not included since this Lewis acid does not contribute to

the ionic strength. **The dotted line is to illustrate the partition between high- and low-valent additives.

Polymerizations were conducted in the presence of tetra-n-butylammonium

trifluoromethanesulfonate (( n-bu) 4NO(O) 2SCF 3; alkylammonium triflate) in order to

elucidate the influence of ionic strength independent of Lewis acid effects. When

compared to polymerizations using BCF (BCF5-8), at a given loading, the catalytic

activity of the alkylammonium triflate polymerizations (triflate5-8) was significantly

reduced and barely reached activities above 1000 g polymer /mol catalyst ·h (Figure 3.13). Figure

3.14 shows the catalytic activity comparison between the alkylammonium triflate and

LAs during PK6 synthesis. As µ increases the catalytic activity differences between the alkylammonium triflates and LAs diminish suggesting that ionic strength alone is not a 96 determining factor for affecting polyketone yield. As observed with the metal triflates above, the addition of the alkylammonium triflates reduced the yields of the resulting polyketones (Figure 3.15). Reduced polymerization activity may be a result of the triflate anion coordinating with the palladium metal center and preventing coordination of monomer.

Figure 3.13: Catalytic activity comparison of polymerizations with additive

alkylammonium triflate or BCF.

Figure 3.14: Catalytic activity comparison of PK6 with alkylammonium triflate or

LAs.

Figure 3.15: Isolated masses of PK5, PK6, and PK8 when polymerized in the

presence of either the alkylammonium triflates or BCF.

Differential scanning calorimetry (DSC) was performed on PK5 , PK6 , and PK8

to estimate glass transition temperatures (T g). The T g for each was compared in reactions

both with and without additive BCF. As can be seen in Figure 3.16, in all 3 polyketone

systems the value for the T g initially drops, and subsequently stabilizes around a constant value – approximately -15 °C, -34 °C and -42 °C for PK5 , PK6 and PK8 , respectively.

This can be attributed to the decrease in molecular weight that is observed concomitantly

with increasing amounts of BCF (Figure 3.7). Furthermore, the observed T g for the 3

polyketone systems follow the overall trend of PK5 > PK6 > PK8 (Figure 3.17). This

behavior is expected for linear polymers with aliphatic side chains of increasing length,

which increase polymer free volume and therefore decrease the glass transition

temperature. 24, 25

Figure 3.16: Glass transition temperatures as a function of BCF loading.

Figure 3.17: Glass transition temperatures fall off precipitously as a result of the

reduced molecular weight observed with the addition of BCF.

The stark decrease and concomitant leveling of the molecular weight and T g for each of the polyketones may be the result of increased side chain length relative to the polyketone itself. T gs are known to depend on the bulk properties of the polymer as well

as the identity of the side chains. 26 Using PK8 as an example, decreases in molecular weight will lead to shorter ketone segments in relation to the hexyl side chain. The T g would then be more influenced by the alkyl side chain rather than the alternating ketones in the backbone (Figure 3.18).

100

Figure 3.18: As the molecular weight decreases the T g is more affected by the

relative ratio of alkyl side chain to alternating ketone backbone.

3.3 Conclusion

Adding Lewis acids provides important mechanistic insight into the polyketone polymerization reaction. The enhanced activity seen with BCF – likely due to abstraction of acetonitrile ligands – is not observed in metal triflates. The molecular weight and dispersities for the polyketones indicate that the triflates are likely affecting the rate of termination ( kt) and may stem from catalyst poisoning when triflates are used. The

observed trends were general for several olefins. Besides offering some information on

mechanisms, our findings can now be used to tailor polyketones with specific molecular

weights and dispersities by judicious selection of metal triflate identity and concentration.

3.4 Experimental

General Remarks. All reactions were carried out in an inert atmosphere glovebox or using standard Schlenk techniques unless otherwise noted. NMR spectra are

101 referenced to deuterated solvent (e.g. CDCl 3 corrected to 7.26 ppm) and chemical shifts

reported as ppm.

Apparatus. 4590 Micro Stirred Reactor from Parr Instrument Company was used

for the copolymerization reactions of α-olefin and carbon monoxide. For the reactions

involving Lewis acids, a Parr reactor with a larger volume was used.

All NMR spectra were collected at room temperature on a Bruker Avance 400

(400 MHz) and a Bruker Avance 500 (500 MHz). GPC was performed with a Shimadzu

LC 20-AT chromatograph using three PolarGel-M (300x7.5 mm) columns, and THF as

an eluent. Calibration was done using standard PMMA samples (M p range: 202 – 71,800

Da). Solvents were purified by Pure Process Technology purification system.

Thermal Analyses. Differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) were completed with TA Instruments DSC 2920

Differential Scanning Calorimeter and TA Instruments Hi-Res TGA 2950

Thermogravimetric Analyzer, respectively. DSC measurements were performed by first cooling samples to -80 °C or -100 °C (for samples with a lower T g) at a rate of 10

°C/min, and subsequently heating to +100 °C at the same rate; after this first cycle, the

samples were cooled and heated a second time, using the same parameters. Between each

temperature ramp, the samples were left to rest at the destination temperature for 2

minutes.. For TGA, the heating rate was 10 ºC per minute from room temperature to

800ºC under nitrogen atmosphere.

102

Chemicals. Chemicals were used as received from Sigma Aldrich Chemical Co. and Fischer Scientific. Solvents used in air-free reactions were purified by a solvent purification system.

18 Preparation of [(Ph) 2P(CH 2)3P(Ph) 2Pd(NCMe) 2](BF 4)2 (2). To a 100 mL

18 flask [(NCMe) 4Pd](BF 4)2 (0.533 g, 1.2 mmol), diphenylphosphinopropane (0.494 g, 1.2

mmol), and a magnetic stir bar were added. Acetonitrile (50 mL) was then added to the

flask to produce a clear, bright yellow solution. The reaction was left to stir rapidly for 24

h. Once complete, the solvent was removed under vaccum to produce

1 [(Ph) 2P(CH 2)3P(Ph) 2Pd(NCMe) 2](BF 4)2 as a bright yellow film (0.831 g, 89.3 %). H

NMR (400 MHz, CDCl 3) δ 7.67 (dd, 7H), 7.49 (d, 3H), 2.93 (s, 4H), 2.30 (s, 2H), 1.90 (s,

31 6H). P NMR (202 MHz, CD 2Cl 2) δ 12.47.

Preparation of tetrabutylammonium trifluoromethansulfonate. The procedure was adapted from Rousseau et al. 27 Approximately 10 mL of water were added to a 50

mL round bottom flask followed by 4.06 g (12.6 mmol) tetrabutyl ammonium bromide.

With rapid stirring 1.0 mL (11.3 mmol) trifluoromethanesulfonic acid was added to the

flask dropwise until a thick, white slurry was formed. The slurry was chilled in an ice

bath and filtered through a glass frit. The solid was washed with water and dried under

1 vacuum to yield (n-bu) 4NO(O) 2SCF 3 as a white solid (4.01 g, 91 %). H NMR (500 MHz,

CDCl 3) δ 3.31 – 3.12 (m, 2H), 1.63 (p, J = 7.8 Hz, 2H), 1.43 (h, J = 7.3 Hz, 2H), 1.01 (t,

J = 7.4 Hz, 3H).

Copolymerization of α-olefin and carbon monoxide (PK5, PK6, PK8). The procedure was adapted from Abu-Surrah et al. 28 A pressure reactor was charged with 2

103

(0.1 mol%, 0.0457 g, 0.0595 mmol) followed by dichloromethane (10 mL), the selected

α-olefin (59.5 mmol), and methanol (358 µL). The reactor was sealed and removed from the glovebox. The reactor was pressurized with carbon monoxide (1000 psi), and the reaction stirred at room temperature for 24 h. Once complete, the reaction solution was depressurized and quenched with an excess of methanol followed by removal of catalyst residues via silica gel plug. Solvent was removed under vacuum to yield the product as a translucent gel.

Copolymerization of α-olefin and carbon monoxide with Lewis acid.

Employing the same copolymerization strategy as described above. A stock solution of catalyst, monomer, and initiator was prepared with 2 (8.71 mg, 0.01124 mmol) followed

by dichloromethane (25 mL), the selected α-olefin (71.3 mmol), and methanol (73.5 µL,

1.82 mmol). Seven 4-mL vials were equipped with stir bars. The first vial was set as the

control sample (without Lewis acid) and the remaining were charged with desired

amounts of the Lewis acid. 3 mL of the stock solution was added to each vial and a

septum punctured with a 16-gauge needle was used to seal each vial. The vials were

placed into the reactor, and the reactor was sealed and removed from the glovebox. The

reactor was pressurized with carbon monoxide (1000 psi), and the reactions were stirred

at room temperature for 24 h. Once complete, the reaction solutions were slowly

depressurized and quenched with an excess of methanol followed by removal of catalyst

residues by silica plug and purified further by precipitation with methanol. Polymers

were dried under vacuum.

104

3.5 References Cited

2. Abu-Surrah, A. S.; Wursche, R.; Rieger, B., Polyketone materials: control of glass transition temperature and surface polarity by co- and terpolymerization of carbon monoxide with higher 1-olefins. Macromolecular Chemistry and Physics 1997, 198 , 1197-1208.

3. Abu-Surrah, A. S.; Rieger, B., High molecular weight 1-olefin/carbon monoxide copolymers: a new class of versatile polymers. Topics in Catalysis 1999, 7 (1), 165-177.

4. Anselment, T. M. J.; Vagin, S. I.; Rieger, B., Activation of late transition metal catalysts for olefin polymerizations and olefin/CO copolymerizations. Dalton Transactions 2008, (34), 4537-4548.

5. Dash, A. K.; Jordan, R. F., Vinyl C−H Activation Reactions of Vinyl Esters Mediated by B(C6F5)3. Organometallics 2002, 21 (5), 777-779.

6. Barnes, D. A.; Benedikt, G. M.; Goodall, B. L.; Huang, S. S.; Kalamarides, H. A.; Lenhard, S.; McIntosh, L. H.; Selvy, K. T.; Shick, R. A.; Rhodes, L. F., Addition Polymerization of Norbornene-Type Monomers Using Neutral Nickel Complexes Containing Fluorinated Aryl Ligands. Macromolecules 2003, 36 (8), 2623-2632.

7. Beckmann, U.; Eichberger, E.; Rufi ńska, A.; Sablong, R.; Kläui, W., Nickel(II) catalysed co-polymerisation of CO and ethene: Formation of polyketone vs. polyethylene – The role of co-catalysts. Journal of Catalysis 2011, 283 (2), 143-148.

8. Berchtold, B.; Lozan, V.; Lassahn, P.-G.; Janiak, C., Nickel(II) and palladium(II) complexes with α-dioxime ligands as catalysts for the vinyl polymerization of norbornene in combination with methylaluminoxane, tris(pentafluorophenyl)borane, or triethylaluminum cocatalyst systems. Journal of Polymer Science Part A: Polymer Chemistry 2002, 40 (21), 3604-3614.

105

9. Hagihara, H.; Shiono, T.; Ikeda, T., Living Polymerization of Propene and 1- Hexene with the [t-BuNSiMe2Flu]TiMe2/B(C6F5)3 Catalyst. Macromolecules 1998, 31 (10), 3184-3188.

10. Chen, E. Y.-X.; Marks, T. J., Cocatalysts for Metal-Catalyzed Olefin Polymerization: Activators, Activation Processes, and Structure−Activity Relationships. Chemical Reviews 2000, 100 (4), 1391-1434.

11. Drent, E.; Budzelaar, P. H. M., Palladium-Catalyzed Alternating Copolymerization of Alkenes and Carbon Monoxide. Chemical Reviews 1996, 96 (2), 663-682.

12. Barlow, G. K.; Boyle, J. D.; Cooley, N. A.; Ghaffar, T.; Wass, D. F., Mechanistic Studies of Alkene/CO Polymerization with Palladium Complexes Promoted by B(C6F5)3. Organometallics 2000, 19 , 1470-1476.

13. Drent, E.; Budzelaar, P. H. M., Palladium-Catalyzed Alternating Copolymerization of Alkenes and Carbon Monoxide. Chemical Reviews 1996, 96 (2), 663.

14. Sen, A., Mechanistic aspects of metal-catalyzed alternating copolymerization of olefins with carbon monoxide. Accounts of Chemical Research 2002, 26 (6), 303-310.

15. Nozaki, K.; Hiyama, T., Stereoselective alternating copolymerization of carbon monoxide with alkenes. Journal of Organometallic Chemistry 1999, 576 (1), 248-253.

17. Pérez ‐Foullerat, D.; Meier, U. W.; Hild, S.; Rieger, B., High ‐Molecular ‐Weight Polyketones from Higher α‐Olefins: A General Method. Macromolecular Chemistry and Physics 2004, 205 (17), 2292-2302.

18. Samples, E. M.; Schuck, J. M.; Joshi, P. B.; Willets, K. A.; Dobereiner, G. E., Synthesis and Properties of N-Arylpyrrole-Functionalized Poly(1-hexene-alt-CO). Macromolecules 2018, 51 (22), 9323-9332.

106

19. Kosaka, N.; Oda, T.; Hiyama, T.; Nozaki, K., Synthesis and Photoisomerization of Optically Active 1,4-Polyketones Substituted by Azobenzene Side Chains. Macromolecules 2004, 37 (9), 3159-3164.

20. Li, L.; Marks, T. J., New Organo-Lewis Acids. Tris( β-perfluoronaphthyl)borane (PNB) as a Highly Active Cocatalyst for Metallocene-Mediated Ziegler−Natta α-Olefin Polymerization. Organometallics 1998, 17 (18), 3996-4003.

21. Bergquist, C.; Bridgewater, B. M.; Harlan, C. J.; Norton, J. R.; Friesner, R. A.; Parkin, G., Aqua, Alcohol, and Acetonitrile Adducts of Tris(perfluorophenyl)borane: Evaluation of Brønsted Acidity and Ligand Lability with Experimental and Computational Methods. Journal of the American Chemical Society 2000, 122 (43), 10581-10590.

22. Vavasori, A.; Ronchin, L., Polyketones: Synthesis and Applications. In Encyclopedia of Polymer Science and Technology , 2017.

23. Chen, D.; Gau, M. R.; Dobereiner, G. E., Palladium and Platinum Acyl Complexes and Their Lewis Acid Adducts. Experimental and Computational Study of Thermodynamics and Bonding. Organometallics 2015, 34 (16), 4069-4075.

24. Ediger, M. D.; Angell, C. A.; Nagel, S. R., Supercooled Liquids and Glasses. The Journal of Physical Chemistry 1996, 100 (31), 13200-13212.

25. Tschoegl, N. W.; Knauss, W. G.; Emri, I., The Effect of Temperature and Pressure on the Mechanical Properties of Thermo- and/or Piezorheologically Simple Polymeric Materials in Thermodynamic Equilibrium – A Critical Review. Mechanics of Time-Dependent Materials 2002, 6 (1), 53-99.

26. Allcock, H. R.; Lampe, F. W.; Mark, J. E., Contemporary Polymer Chemistry . 3 ed.; Pearson Education, Inc.: New Jersey, 2003.

27. Rousseau, K.; Farrington, G. C.; Dolphin, D., Tetraalkylammonium trifluoromethanesulfonates as supporting electrolytes. The Journal of Organic Chemistry 1972, 37 (24), 3968-3971.

28. Abu-Surrah, A. S.; Wursche, R.; Rieger, B., Polyketone materials: control of glass transition temperature and surface polarity by co- and terpolymerization of carbon 107 monoxide with higher 1-olefins. Macromolecular Chemistry and Physics 1997, 198 (4), 1197-1208.

108

CHAPTER 4

SPIROLIGOMER POLYESTER CATALYST DEVELOPMENT

4.1 Introduction

Polyesters are attractive materials for their biocompatibility and biodegradability which leads to their applications in medical devices like sutures and screws or as artificial tissue. 1, 2, 3 Polyesters may be synthesized through condensation polymerization,1 bacterial fermentation,1 ring-opening polymerizations of cyclic esters,1, 4 and alternating copolymerization of epoxides and cyclic anhydrides, and are largely divided into two classes: semiaromatic and aliphatic. Semiaromatic polyesters like poly(ethylene terephthalate) (PET) have good mechanical properties and barrier properties and serve as commodity plastics (blow-molded bottles and fibers for example)5. Aliphatic polyesters

(APs) can be made from biorenewable monomers and exhibit good biocompatibility.1, 4, 6,

Polyesters are commonly synthesized via step-growth polymerization of diacids or diesters with diols, but this route requires the removal of small molecule by-products like water (Figure 4.1A). Step-growth polymerizations also require long reaction times to achieve high conversion and high molecular weights, often resulting in dispersities ≥ 2. 7

Therefore, step-growth polymerization is inherently more energy intensive. Chain-growth polymerization, in contrast, does not create any small molecule by-products, and low dispersity polymers are achievable at lower conversion of monomer. APs are often produced from chain-growth ring-opening polymerization (ROP) of lactones. Poly(lactic 109 acid), for example, is obtained from ROP of lactide (the cyclic ester of lactic acid, Figure

4.1B) using a myriad of catalysts, including organocatalysts and metal alkoxides. 4, 7

While no by-products are formed from these reactions, detrimental side reactions (e.g. transesterification) plague these reactions at high monomer conversion. 7 ROP of lactones

also suffers from lack of functional diversity, therefore limiting the end-user applications

of the materials.

Figure 4.1: A) Step-growth polymerization between two bifunctional monomers to

prepare polyesters. B) ROP of lactone. C) Alternating copolymerization of epoxides

and cyclic anhydrides is a facile route to prepare novel APs.

110

Alternating copolymerization of epoxides with cyclic anhydrides has been realized as a facile route to circumvent the lack of chemical diversity of APs (Figure

4.1C). To date, there have been approximately 20 epoxides and 20 anhydrides successfully copolymerized resulting in 400 different APs. 1, 6, 7, 8, 9, 10, 11

Copolymerization of epoxides and cyclic anhydrides has been accomplished by an array of metal complexes including zinc, magnesium, manganese, chromium, and aluminum.

In each case the addition of a nucleophilic cocatalyst like 4-(dimethylamino)pyridine

(DMAP) or bis(triphenylphosphine)iminium salts ([PPN]X) increases catalyst activity. 7

The earliest report of a controlled copolymerization of phthalic anhydride and propylene oxide used an aluminum porphyrin catalyst and a tetraalkylammonium halide cocatalyst. 12

Salen-type metal complexes have since become one of the most thoroughly studied catalysts for AP copolymerizations, and success with chromium, cobalt, aluminum, and iron have all been reported. 1, 6, 7, 8, 9, 10, 11 Because the salen-type metal

complexes have relatively planar structures, the metal center is exposed at two axial

faces. These faces are both capable of participating in polymerization, therefore

complicating the understanding of the polymerization mechanism (Figure 4.2).7, 12 In a general proposal, the epoxide is activated at the metal center which is then attacked by a carboxylate, forming a metal-alkoxide (Figure 4.2, steps 1 and 2). The alkoxide can then react with an anhydride forming the polyester (Figure 4.2, steps 3 and 4).7, 9 If the metal

alkoxide selectively attacks an anhydride, it prevents the formation of polyethers; only a

few catalysts, such as the chromium-salen system 5 (Figure 4.8), meet this requirement. 7

111

Figure 4.2: A simplified mechanism for the alternating copolymerization of an

epoxide with cyclic anhydrides with PPNCl as a cocatalyst. Reprinted with

permission from Longo, J.M. et al. Chemical Reviews 2016, 116 (24), 15167-15197.

Due to the nature of chain-growth ROP, control of polymer regiochemistry is possible. Ring opening may either occur at the less sterically hindered methylene carbon resulting in retention of the stereocenter, or at the methine carbon leading to inversion of the stereocenter (Figure 4.2 step 1). Using enantiopure epoxides produces a stereoregular, isotactic polmer. As discussed in Chapter 1, tacticity of a polymer has a marked influence

11, 13 on its physical properties like T g and T m. Coates and coworkers prepared several isotactic, semicrystalline high T m polyesters from enantiopure propylene oxide (Figure

4.3), and also investigated the regioregularity of poly(propylene maleate) (PPM) using

112 cobalt- and chromium-salen catalysts. 6 Their findings identified cobalt complexes as

producing a more regioregular PPM than the analogous chromium complexes.

Figure 4.3: Examples of semi-crystalline APs prepared from enantiopure epoxides.

A) poly((S)-propylene succinate) T m = 79 °C. B) poly((S)-propylene maleate) T m =

117 °C. C) poly((S)-propylene phthalate) T m = 150 °C.

The ability of the metal-salen to activate an epoxide and a cyclic anhydride at

either axial face leads to a lack of stereoselectivity of the polymer; stereoregular

polyesters from copolymerization is achieved from an enantioenriched epoxide, rather

than kinetic resolution of a racemic monomer. The aim of the current Chapter

investigates the development of advanced polyester catalysts based on the incorporation

of a salen-type metal catalyst scaffold into a much larger spiroligomer framework

developed in the Schafmeister laboratory. The proposed catalysts would enclose the

metal in a chiral “pocket” formed of spiroligomer segments (Figure 4.4). High degrees of

113 chirality from the spiroligomer segments mimic binding pockets of metalloenzymes leading to control of stereochemistry of the growing polymer chain.

Figure 4.4: A cartoon representation of the proposed spiroligomer-based catalyst.

This representation illustrates four spiroligomer segments connected at the bottom

with a transition metal (M) ligated within the chiral pocket created by the

spiroligomer segments. Figure credit to Dr. Christian Schafmeister.

114

4.2 Results and Discussion

Before benchmarking larger spiroligomer-based catalysts, we began by comparing the activity of known salen catalysts to simple scaffolds that can be later incorporated into a spiroligomer cleft. The salen scaffolds used herein were functionalized using chemical techniques developed in the Schafmeister laboratory, and the resulting [spiro]MX (Figure 4.5) complexes were compared to known [salcy]MX catalysts for the synthesis of PPM.

Figure 4.5: Comparison of [salcy]MX catalysts to the [spiro]MX catalysts developed

in the Schafmeister laboratory. The [salcy]MX catalysts are active at both axial

faces of the complex; the bottom face of the [spiro]MX catalyst is proposed to be

inaccessible to substrate due to the spiroligomer/alkyl tether.

115

4.2.1 Catalyst Synthesis

Several [salcy]MX scaffolds were synthesized according to literature precedent 14 following the general reaction scheme in Figure 4.6, leading to a small library of salen ligands (Figure 4.7). The salens were then metalated with chromium, cobalt, or manganese salts, and these catalyst precursors were then screened in a model polymerization to determine each catalyst’s baseline reactivity for a given monomer set.

Figure 4.6: The general synthesis for a salen ligand.

With salens 1-4 in hand metalation was straight forward. Following the work of

Coates and coworkers 6 chromium was easily coordinated to 1-4 by reacting the ligand in a 1:1 ratio with anhydrous CrCl 2 in tetrahydrofuran under a nitrogen atmosphere (Figure

4.8). The reaction slurry was then exposed to air to oxidize. Catalysts 5, 6, 10, 14, and 16 were isolated as air-stable solids. Catalyst 7, 8, 9, 11, 12, 13, and 15 were prepared in a similar fashion using the respective anhydrous manganese and cobalt salts.

116

Figure 4.7: Salen scaffolds used in model copolymerization reactions.

117

Figure 4.8: Metalation of salens 1-4 with various metal salts led to a library of

catalysts for PPM synthesis.

Metalation was confirmed using ATR-IR spectroscopy, UV-Vis spectroscopy,

-1 and HPLC-MS. The imine stretch (-C=N-, νcm ≈ 1630) of 1-4 shifts to lower wavenumbers upon metalation. 15, 16 The mass spectra were collected in acetonitrile and

the masses for the corresponding acetonitrile adducts were observed.

4.2.2 Alternating Copolymerization

The alternating copolymerization of maleic anhydride and propylene oxide was

chosen for this study as it is well precedented 1, 6, 11 and benchmarks of polymerization efficiency and selectivity (ether content, enantiomeric excess) are easily obtained from analysis of the polyester (Figure 4.9).

118

Figure 4.9: Alternating copolymerization of propylene oxide and maleic anhydride

to yield poly(propylene maleate) (PPM).

Coates and coworkers reported 6 as a highly active catalyst for PPM synthesis. 6

The racemic catalyst also produced no poly(ether) – consecutive propylene oxide enchainments – in the product ( m in Figure 4.9). Racemic 6 was not prepared during this study, but scalemic 6 proved to be a competent catatlyst for PPM synthesis nonetheless reaching 60% conversion in less than 24 h (Entry 1, Table 4.1). Baseline catalytic activity of catalysts prepared in this study for PPM synthesis was contrasted with 6.

119

Table 4.1: Selected PPM copolymerization results, used to determine optimum

reaction conditions and most appropriate metal/ligand combination.

Entry Catalyst Temp (°C) Time (h) Conv. (%) d Ether (%) d Mn (Da) e Ð (Mw/Mn) e 1 6 45 a 23 60 <1 - - 2 7 45 a 16 0 <1 - - 3 7 45 a 22 46 18 - - 4 9 45 a 16 33 2 - - 5 10 45 a 40 28 11 - - 6 12 45 a 16 11 <1 - - 7 14 45 a 16 15 51 - - 8 14 100 a 16 87 6 - - 9 16 100 a 26 22 28 - - 10 16 100 a 26 17 46 - - 11 6 60 b 2 88 7 2152 1.4 12 6 60 b 18 >99 20 5638 1.6 13 9 60 b 2 55 <1 1539 1.4 14 12 60 b 2 32 <1 851 1.6 15 14 60 b 2 68 9 1632 1.7 16 14 60 b,c 2 12 49 244 2.5 17 14 60 b 18 >99 17 1878 9.6 18 16 60 b 2 79 10 1806 1.3 19 16 60 b,c 2 16 44 520 1.4 20 16 60 b 18 >99 11 7487 1.5 a Reaction conditions: [MA]:[PO]:[cat] = 200:200:1; [MA]:[PO] = 4 mmol in 1 mL toluene. b Reaction conditions: [MA]:[cat]:[PPNCl] = 400:1:1; carried out neat in 1 mL

PO. c No PPNCl used. d Estimated based on the chemical shifts due to ether linkages in the

1H NMR spectrum. e Determined by gel permeation chromatography calibrated with

PMMA standards in THF at 30 °C

120

Conditions in entries 1-10, Table 4.1 required long reaction times to reach moderate monomer conversions. At 0.5 mol % catalyst loading, these reactions also required a significant mass of the [spiro]MX catalysts. A marked increase in catalyst activity was achieved in entries 11-20 by using 0.25 mol% catalyst, neat PO as the solvent and addition of 1 eq PPNCl ([Ph 3P═N═PPh 3]Cl) cocatalyst. However, reduced

catalyst loadings and running reactions in neat PO led to the formation of significant

consecutive ether enchainment (entry 12, Table 4.1). The addition of PPNCl helped to

reduce the ether content of the polyester as evidenced by entries 15 and 16. The ratio n:m

in PPM was determined through analysis of the 1H NMR of PPM; an example is shown in Figure 4.10.

121

Figure 4.10: 1H NMR spectra for PPM prepared with: catalyst 6 without PPNCl

(spectrum 1); catalyst 14 without PPNCl (spectrum 2); and 14 with PPNCl

(spectrum 3). *indicates residual diethyl ether.

No PPM was detected when 7 was used, and monomer conversion was low when the [salcy]CoX catalysts were used (entries 3, 4, and 12, respectively; Table 4.1).

Addtionally, for [salcy]CoX catalysts metalation required an additional step to oxidize the metal center to Co(III). Therefore, at this stage we focused our efforts on Cr catalysts.

122

Figure 4.11: [spiro]MX catalysts screened for PPM synthesis.

Three spiroligomer ligands were prepared in the Schafmeister laboratory (Figure

4.11). Catalysts 17-19 , prepared via the metalation procedure outlined earlier, were screened for PPM synthesis (Table 4.2). Catalysts 16 and 17 have analogous diimine linkers, and a comparison of their monomer conversions shows that 17 is not as active as

14 within the 2 h reaction, although they produced the same ether content (entries 1 and

4; Table 4.2). When 17 is allowed to react for 24 h (entry 5; Table 4.2), quantitative monomer conversion was obtained with no increase in ether content of the product. The ether content (entries 4 and 5; Table 4.2) may occur before a sufficient concentration of the carboxylate ion is formed. The catalyst may homopolymerize PO (Figure 4.12)17 until the ring opening of MA forms the on-pathway carboxylate species.

123

Figure 4.12: Polyether synthesis during induction period leading to consistent ether

content in PPM during different reaction times. Reprinted from Brocas et al.,

from Elsevier.

124

Table 4.2: Comparison of [salcy]MX catalysts to [spiro]MX catalysts for PPM

sysnthesis.

a b c Entry Catalyst Time (h) Conv. (%) Ether (%) Mn(Da)

1 16 2 79 10 1806

2 14 2 68 9 1718

3 5 2 88 7 2152

4 17 d 2 45 11 1065

5 17 e 24 >99 8 7784

6 18 e 24 24 30 1560

7 19 e 24 24 29 1680

a Determined from analysis of crude 1H NMR spectrum. b Estimated based on the chemical shifts due to ether linkages in the 1H NMR spectrum. c Determined by gel permeation chromatography calibrated with PMMA standards in THF at 30 °C. d

Reaction conditions: [MA]:[cat]:[PPNCl] = 400:1:1; carried out in 1 mL PO. e Reaction

conditions: [MA]:[cat]:[PPNCl] = 1430:1:1; carried out in 1 mL PO.

125

4.2.3 Reductive Degradation

The polyesters may be degraded to obtain enantiomeric excess ( ee ) of the polymer backbone. The reaction between the PPM and lithium aluminum hydride (LAH) cleaves the polymer at the ester bond, providing a mixture of diols ( 20 and 21 , Figure 4.13). The ee of the resulting propylene glycol (PG, 20 ) was analyzed by chiral GC.

Figure 4.13: Reductive degradation of the polyester provides a mixture of diols

which can be resolved via chiral GC.

Polyesters prepared with 6 and 14 were degraded following (Figure 4.13) and

analyzed via chiral GC. Unsurprisingly, 14 showed no selectivity for either the R or S enantiomer of PO (as represented by the relative ratios of R- and S-PG), but 6 exhibited a preference for the S-PO, as seen in Figure 4.14. Diol peaks for 20 elute at ~14.2 min are not well resolved with the current method, and the relative ratios of R- and S-PG are qualitative.

126

R-PG S-PG A

Figure 4.14: A) PPM catalyzed with 14 exhibited no selectivity for PO. B) PPM

catalyzed by 6. *Propylene glycol peaks are not well resolved with the current

method, and relative ratios of R- and S-PG are qualitative.

S-PG

Figure 4.15: A) The relative ratios of R- and S-PG from PPM catalyzed by 17. B)

The sample was spiked with a standard S-PG to aid in peak indentity.

At present, only 17 has been evaluated for PO selectivity. Qualitatively, there is a preference for 17 to polymerize R-PO (Figure 4.15). In light of the poor resolution of the

127

PG peaks new conditions are being explored, and a more appropriate GC column purchased.

4.3 Conclusion

In collaboration with the Schafmeister laboratory we have prepared a series of

[salcy]MX and [spiro]MX catalysts. The [salcy]MX catalysts 5-16 were screened for

baseline synthesis in the alternating copolymerization of propylene oxide and maleic

anhydride, and optimized reaction conditions were developed. Three [spiro]MX catalysts

17-19 were synthesized and tested for PPM synthesis. Preliminary data suggest that the

meso -[spiro]CrCl 17 exhibits a higher catalytic activity for PPM synthesis than catalysts

18 or 19 , while providing one of the lowest consecutive ether enchainment rates observed

in our study. At present, the ee of the respective PPM products prepared with 17-19 is not

known beyond a qualitative observation that 17 has a slight preference for R-PO. Future

efforts are aimed at improving the resolution of the chiral GC assay. Additionally,

tethering multiple spiro-blocks to form larger, more complex ligand scaffolds is currently

underway.

4.4 Experimental

General Remarks. All reactions were carried out in an inert atmosphere

glovebox or using standard Schlenk techniques unless otherwise noted. NMR spectra are

referenced to deuterated solvent (e.g. CDCl 3 corrected to 7.26 ppm) and chemical shifts

reported as ppm.

128

Apparatus. All NMR spectra were collected at room temperature on a Bruker

Avance 400 (400 MHz) and a Bruker Avance 500 (500 MHz). GPC was performed with a Shimadzu LC 20-AT chromatograph using three PolarGel-M (300x7.5 mm) columns, and THF as an eluent. Calibration was done using standard PMMA samples (M p range:

202 – 71,800 Da). Solvents were purified by Pure Process Technology purification system. Chiral gas chromatography was performed on an Hewlett Packard 6890 GC equipped with a Supelco Astec Chiraldex® B-TA column (30 m x 0.25 mm, d f 0.12 µm) and a flame ionization detector. Chiral GC heating was increased from 30 °C to 160 °C by 5 °C min -1.

Chemicals. Chemicals were used from Sigma Aldrich Chemical Co. and Fischer

Scientific. Solvents used during air-free reactions were purified by a solvent purification system. [spiro]MX catalysts 17-19 were prepared by Yanfeng Fan of the Schafmeister laboratory.

Preparation of salen ligands (1-4) – a general procedure. Salen ligands 1-4 were prepared according to literature precedent. 14 Exclusion of air was not necessary for

the synthesis of 1-4. 1,2-diaminocyclohexane (0.060 g, 0.52 mmol) was added to a 50 mL

round bottom flask and dissolved with ~5 mL of methanol, and the round bottom flask

was warmed in a 65 °C oil bath. 3,5-ditertbutylsalicylaldehyde (0.25 g, 1.1 mmol) was

dissolved into ~5 mL warm methanol. The aldehyde solution was added dropwise to the

solution of diamine forming a dark yellow solution. A reflux condenser was attached to

the round bottom flask, and the reaction was refluxed at 85 °C for ca. 4 h. After this time

a yellow precipitate was visible. The solid was filtered and rinsed with chilled methanol,

129

1 then dried under vacuum yielding 0.207 g (72 %) of 1. H NMR (500 MHz, CDCl 2) δ

13.69 (s, 2H), 8.30 (s, 3H), 7.29 (d, J = 2.4 Hz, 3H), 6.99 (d, J = 2.5 Hz, 3H), 3.38 – 3.27

(m, 4H), 1.97 (d, J = 13.5 Hz, 3H), 1.88 (d, J = 8.9 Hz, 3H), 1.81 – 1.67 (m, 4H), 1.38 (s,

18H), 1.21 (s, 18H). ATR-FTIR: 1629 (-C=N-).

1 2. 0.086 g (70%). H NMR (500 MHz, CDCl 2) δ 13.59 (s, 2H), 8.39 (s, 2H), 7.32

(d, J = 2.4 Hz, 3H), 7.26 – 7.14 (m, 11H), 7.00 (d, J = 2.5 Hz, 2H), 4.77 (s, 2H), 1.41 (s,

18H), 1.21 (s, 18H). ATR-FTIR: 1626 (-C=N-).

1 3. Prepared previously by Joe Becica. H NMR (500 MHz, CDCl 2) δ 13.68 (s,

2H), 8.40 (s, 2H), 7.36 (d, J = 2.5 Hz, 2H), 7.09 (d, J = 2.5 Hz, 2H), 3.92 (s, 4H), 1.41 (s,

19H), 1.27 (s, 18H). ATR-FTIR: 1626 (-C=N-).

1 4. 0.194 g (48 %). H NMR (500 MHz, CDCl 2) δ 13.61 (s, 2H), 8.70 (s, 2H), 7.45

(d, J = 2.4 Hz, 3H), 7.36 (dd, J = 5.8, 3.4 Hz, 3H), 7.30 – 7.23 (m, 5H), 1.42 (s, 18H),

1.32 (s, 18H). ATR-FTIR: 1615 (-C=N-).

Preparation of [salcy]MX (5-16) – a general procedure. 5 was prepared

6 according to a modified literature procedure. In a glovebox, anhydrous CrCl 2 (0.0494 g,

0.402 mmol) and 1 (0.200 g, 0.366 mmol) were added to a flame dried schlenk flask equipped with a stir bar and were diluted with ~10 mL anhydrous THF. The slurry was allowed to stir for at least 3 h under dry N 2. After 3 h, the reaction flask was removed

from the glovebox and the stopper of the flask was removed and replaced with a drying

tube charged with Drierite® agent. The brown solution was allowed to stir for an

additional 3 h while oxidizing under dry air. Subsequently, the reaction mixture was

diluted with diethyl ether (50 mL) and washed with saturated ammonium chloride (3 x 10

130 mL) and brine (3 x 10 mL). The organic layer was collected and dried over anhydrous sodium sulfate, filtered, and evacuated to dryness, yielding a dark brown powder. The crude product was dissolved in minimal DCM and an excess of hexanes was added.

Precipitation was induced by concentration under vacuum, and 5 was collected as a dark

brown solid (0.129 g, 55% yield). ATR-FTIR: 1604 (-C=N-); MS (EI) m/z 596.4

([salcy]M +), 637.4 ([salcy]M-NCMe).

[diphen]CrCl, 10. (0.090 g, 99% yield). ATR-FTIR: 1607 (-C=N-); MS (EI) m/z

694.4 ([diphen]M +), 735.4 ([diphen]M-NCMe).

[phen]CrCl, 14. (0.126 g, 99% yield). ATR-FTIR: 1601 (-C=N-); MS (EI) m/z

590.3 ([phen]M +), 632.3 ([phen]M-NCMe).

Representative copolymerization procedure. The following procedure was modified from a previously reported method. 6 In a glovebox, maleic anhydride (0.196 g,

2 mmol), 0.005 mmol catalyst, and PPNCl (0.003 g, 0.005 mmol) were added to a 20 mL

vial equipped with a stir bar. Propylene oxide (1.0 mL) was used as the solvent. The vial

was sealed with a Teflon lined cap and removed from the glovebox. The vial was placed

in a 60 °C pre-heated aluminum heating block. After the desired reaction time an aliquot

was removed for NMR analysis. The reaction mixture was concentrated under vacuum,

then dissolved in a minimum amount of methylene chloride and precipitated with diethyl

ether. The precipitated polymer was dried at 60 °C under vacuum. 1H NMR (500 MHz,

CDCl 3) δ 6.42 – 6.17 (m, 2H), 5.26 (td, J = 6.3, 4.0 Hz, 1H), 4.35 – 4.15 (m, 2H), 1.32

(dd, J = 6.6, 1.6 Hz, 3H). Ether content of PPM was estimated based on the chemical

131 shifts reported for consecutive PO enchainment: 3.41-3.53 [-OCHCH 2-], 1.13-1.18 [-

18 CH 3].

Representative reductive degradation procedure. The following procedure was modified from a previously reported method. 10 In a 20 mL vial the polyester (0.076 g,

0.48 mmol) was dissolved in 10 mL THF. Once the polyester was dissolved, lithium aluminum hydride (0.183 g, 4.8 mmol) was added. The vial was capped, taped, and placed into a 60 °C pre-heated aluminum heating block for 12 h. After 12 h the reaction was cooled to 0 °C and diluted with 10 mL diethyl ether. Excess LAH was quenched with

0.2 mL water, then 0.2 mL 15% NaOH, followed by 0.5 mL water. The reaction was dried with MgSO 4, and the mixture was filtered through Celite. The organic layer was

dried under vacuum to yield 0.031 g of a clear, slightly yellow oil. Some mass was lost

during drying.

Chiral gas chromatography. After reductive degradation of the polyester the resulting glycols were analyzed by chiral GC diluting 1 (one) drop of the glycol solution with methylene chloride in a 1 mL mass spectrum vial.

132

4.5 References Cited

1. Jeske, R. C.; DiCiccio, A. M.; Coates, G. W., Alternating Copolymerization of Epoxides and Cyclic Anhydrides: An Improved Route to Aliphatic Polyesters. Journal of the American Chemical Society 2007, 129 (37), 11330-11331.

2. Lasprilla, A. J. R.; Martinez, G. A. R.; Lunelli, B. H.; Jardini, A. L.; Filho, R. M., Poly-lactic acid synthesis for application in biomedical devices — A review. Biotechnology Advances 2012, 30 (1), 321-328.

3. Ikada, Y.; Tsuji, H., Biodegradable polyesters for medical and ecological applications. Macromolecular Rapid Communications 2000, 21 (3), 117-132.

4. Nakano, K.; Kosaka, N.; Hiyama, T.; Nozaki, K., Metal-catalyzed synthesis of stereoregular polyketones, polyesters, and polycarbonates. Dalton Transactions 2003, (21), 4039-4050.

5. Odian, G., Principles of Polymerization . 4 ed.; John Wiley & Sons, Inc.: Hoboken, New Jersey, 2004.

6. DiCiccio, A. M.; Coates, G. W., Ring-Opening Copolymerization of Maleic Anhydride with Epoxides: A Chain-Growth Approach to Unsaturated Polyesters. Journal of the American Chemical Society 2011, 133 (28), 10724-10727.

7. Longo, J. M.; Sanford, M. J.; Coates, G. W., Ring-Opening Copolymerization of Epoxides and Cyclic Anhydrides with Discrete Metal Complexes: Structure–Property Relationships. Chemical Reviews 2016, 116 (24), 15167-15197.

8. Gudeangadi, P. G.; Sakamoto, T.; Shichibu, Y.; Konishi, K.; Nakano, T., Chiral Polyurethane Synthesis Leading to π-Stacked 2/1-Helical Polymer and Cyclic Compounds. ACS Macro Letters 2015, 4 (9), 901-906.

9. Fieser, M. E.; Sanford, M. J.; Mitchell, L. A.; Dunbar, C. R.; Mandal, M.; Van Zee, N. J.; Urness, D. M.; Cramer, C. J.; Coates, G. W.; Tolman, W. B., Mechanistic Insights into the Alternating Copolymerization of Epoxides and Cyclic Anhydrides Using a (Salph)AlCl and Iminium Salt Catalytic System. Journal of the American Chemical Society 2017, 139 (42), 15222-15231. 133

10. Van Zee, N. J.; Coates, G. W., Alternating Copolymerization of Propylene Oxide with Biorenewable Terpene-Based Cyclic Anhydrides: A Sustainable Route to Aliphatic Polyesters with High Glass Transition Temperatures. Angewandte Chemie International Edition 2015, 54 (9), 2665-2668.

11. DiCiccio, A. M.; Longo, J. M.; Rodríguez-Calero, G. G.; Coates, G. W., Development of Highly Active and Regioselective Catalysts for the Copolymerization of Epoxides with Cyclic Anhydrides: An Unanticipated Effect of Electronic Variation. Journal of the American Chemical Society 2016, 138 (22), 7107-7113.

12. Aida, T.; Inoue, S., Catalytic reaction on both sides of a metalloporphyrin plane. Alternating copolymerization of phthalic anhydride and epoxypropane with an aluminum porphyrin-quaternary salt system. Journal of the American Chemical Society 1985, 107 (5), 1358-1364.

13. Longo, J. M.; DiCiccio, A. M.; Coates, G. W., Poly(propylene succinate): A New Polymer Stereocomplex. Journal of the American Chemical Society 2014, 136 (45), 15897-15900.

14. Coletti, A.; Galloni, P.; Sartorel, A.; Conte, V.; Floris, B., Salophen and salen oxo vanadium complexes as catalysts of sulfides oxidation with H2O2: Mechanistic insights. Catalysis Today 2012, 192 (1), 44-55.

15. F. Choudhary, N.; G. Connelly, N.; B. Hitchcock, P.; Jeffery Leigh, G., New compounds of tetradentate Schiff bases with vanadium(IV) and vanadium(V) †. Journal of the Chemical Society, Dalton Transactions 1999, (24), 4437-4446.

16. Bermejo, M. R.; Castiñeiras, A.; Garcia-Monteagudo, J. C.; Rey, M.; Sousa, A.; Watkinson, M.; McAuliffe, C. A.; Pritchard, R. G.; Beddoes, R. L., Electronic and steric effects in manganese Schiff-base complexes as models for the water oxidation complex in photosystem II. The isolation of manganese-(II) and -(III) complexes of 3- and 3,5- substituted N,N ′-bis(salicylidene)ethane-1,2-diamine (H2salen) ligands. Journal of the Chemical Society, Dalton Transactions 1996, (14), 2935-2944.

17. Brocas, A.-L.; Mantzaridis, C.; Tunc, D.; Carlotti, S., Polyether synthesis: From activated or metal-free anionic ring-opening polymerization of epoxides to functionalization. Progress in Polymer Science 2013, 38 (6), 845-873.

134

18. Hua, Z.; Qi, G.; Chen, S., Ring-opening copolymerization of maleic anhydride with propylene oxide by double-metal cyanide. Journal of Applied Polymer Science 2004, 93 (4), 1788-1792.

135

CHAPTER 5

1-D PEROVSKITE SYNTHESIS AND HOLE TRANSPORTER OXIDATION

5.1 Introduction

In photovoltaic (PV) devices, an electron donor (n-type) and an electron acceptor

(p-type) material convert solar energy into electrical energy. 1 These devices are currently dominated by silicon-based materials, but growing energy demands and pushes towards greener energy alternatives necessitate the development of novel PV devices.

- - Perovskites are described by the formula ABX 3 where X is an anion (Cl , Br , etc.) and A and B are cations of different sizes where A is larger than B (Figure 5.1). 1, 2 A variety of perovskites are available from insulators to conductors, but those used for solar cell development and other optoelectronic devices are semiconducting hybrid inorganic- organic perovskites. 1, 2, 3 In perovskite solar cells (PSCs), perovskites typically consist of

+ a monocationic organic molecule ( A, methylammonium (CH 3NH 3 ) or foramidinium

+ - - - 2+ 2+ 1, 2, 3, 4 ((NH 2)2CH )), a halide ( X, Cl , Br , or I ), and a metallic ion like Pb or Sn (B).

136

Figure 5.1: Perovskite crystal structure. MA and FA represent methylammonium

and formamidinium, respectively. Reprinted with permission from Calió, L. et al.,

2016 John Wiley & Sons Inc.

Inorganic/organic hybrid perovskites, such as the archetypal methylammonium

2 lead iodide (CH 3NH 3PbI 3), are semi-conducting pigments, making the materials good

light absorbers. Low power conversion efficiencies (PCE) were originally attained when

CH 3NH 3PbBr 3 was used in a dye-sensitized solar cell, likely because the devices architecture required a liquid electrolyte which dissolved the perovskite sensitizer within minutes leading to a meager PCE of 2 %. 1, 2 Continuing research efforts led to the

breakthrough that liquid electrolytes should be replaced with a solid hole-transporting

material (HTM). Simultaneously, Grätzel 5 and Snaith 6 developed methods to use solid

HTMs to increase PSC power conversions to 9.7 % and 10%, respectively. Since these 137 reports, PSC research has gained significant momentum, and research groups focus on different facets of PSC device fabrication such as design optimization of the perovskite material, deposition techniques, device architecture, and the use of different n- and p-type charge transporters. 1

5.1.1 Perovskite Solar Cell Device Fabrication

There have been two main device architectures to emerge from PSC research: mesoscopic nanostructure and planar structures. In a mesoscopic device (Figure 5.2a) the perovskite is infiltrated into a mesoporous metal oxide (e.g. TiO 2). This device is

constructed by adding a ‘hole blocking’ (TiO 2) layer onto an electrode followed by the a

layer of the mesoporous perovskite scaffold. An organic HTM is deposited on top of the

mesoporous perovskite layer and capped with a gold electrode. 1, 3 A planar heterojunction

(Figure 5.2b) device is prepared in a similar way to the mesoscopic device, but without

1, 3 the use of the mesoporous TiO 2 support. The perovskite layer (i) is thus sandwiched between a thin layer of TiO 2 (n-type conductor) and the HTM (p-type conductor).

138

Figure 5.2: Device architecture of a) mesoscopic and b) planar heterojunction PSCs.

Reprinted with permission from Angewandte Chemie International Edition 2016, 55

The reproducibility of the chosen fabrication technique can be critical to cell efficiency. Initial devices were prepared using a single step deposition of the perovskite onto the electrode from a polar solvent such as dimethylformamide (DMF) or γ- butyrolactone (GBL). 3 This led to uncontrolled morphological changes to the perovskite layer, and poor PCE reproducibility. Spin-coating later became a widely used method of device fabrication. For either mesoscopic or planar devices the perovskite layer may be deposited onto the device in a one- or two-step process. 1, 2, 3 In a one-step method the inorganic component (PbX2) and the organic component (CH 3NH 3X) are mixed in a 1:1 ratio in a polar solvent like DMF or GBL. This solution is the spin-coated onto the n- or p-type conductors. The construction is then annealed at 100 °C to form the perovskite layer, but this technique led to incomplete coverage and uncontrolled morphology of the

1 perovskite. The two-step deposition a PbX 2 film is spin-coated onto the mesoporous support or directly onto the HTM, then the construction is immersed into a solution of the

1 organic component (CH 3NH 3X). This device is also annealed at 100 °C.

139

A myriad of variables contribute to the overall performance of the PSC. For example, the mesoporous TiO 2 layer thickness, porosity, and even pore size affect PCEs.

Additionally, the arrangement of the n- and p-type layers (e.g. p-i-n or n-i-p arrangements) within the devices has led to advances in their own right. 7, 8 Though these devices have shown marked increases in PCE (PCE up to 24 %), instability of the devices

(CH 3NH 3PbI 3 decomposes in the presence of moisture and elevated temperatures) and

inherent toxicity of lead currently inhibit these devices from gaining commercial

attention. 3, 9

5.1.2 Hole Transporting Materials (HTM)

Solid state HTMs may be classified into three categories: inorganic, polymeric, and small molecule HTMs. 1, 10 Inorganic HTMs like copper iodide (CuI) and nickel oxide

(NiO) were explored for their low cost, intrinsically high stability, and high hole mobility

> 10 -3 cm 2V-1s-1), but the solvents used to deposit these materials during device fabrication led to dissolution of the perovskite layer. 1, 10 Polymeric HTMs were discussed

in Chapter 2, and these materials have several advantages such as good thermal stability,

good solubility, tunable optoelectronic properties, and high conductivity. 11

Twisted spirobifluorene-based molecules are currently the most widely studied molecules for HTMs 1, most prominent the 2,2 ′,7,7 ′-Tetrakis(N,N-di-p- methoxyphenylamine)-9,9 ′-spiro-bifluorene (spiro-OMeTAD). Despite spiro-OMeTAD’s widespread HTM use 1, 2, 3, 12 devices prepared with this HTM suffer from poor reproducibility. This stems from the fact that spiro-OMeTAD actually has a low intrinsic hole conductivity and mobility. 13 Additionally, spiro-OMeTAD needs to be doped (p- 140 type) to improve its conductivity. 12, 13, 14 Spiro-OMeTAD is commonly doped with

lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) which does not directly oxidize

13 spiro-OMeTAD, but facilitates its oxidation in the presence of O2 and light (Figure 5.3).

Spiro-OMeTAD oxidation, therefore, requires exposure to air making it difficult to control the extent of spiro-OMeTAD oxidation. The extent of oxidation and concentration of oxidized Spiro-OMeTAD species has been shown to affect PCEs. 13

Figure 5.3: Doping spiro-OMeTAD with LiTFSI leads to a usable HTM for PSCs.

This chapter reports on efforts to elucidate the role of lithium during the oxidation of spiro-OMeTAD, and the challenges associated with an open system oxidation.

Furthermore, attempts to prepare novel perovskites with substitution of A site of the

ABX 3 formula are discussed herein.

141

5.2 Results and Discussion

5.2.1 Spiro-OMeTAD Oxidation

Crown ethers (CE) are used to chelate metal ions to improve solubility of inorganic salts in nonpolar solvents, and differently sized CEs solvate different metal ions 15 ; for example 12-crown-4 ether is used to chelate Li + ions. To gain insight of the

role lithium plays during spiro-OMeTAD oxidation 12-crown-4 ether (12-c-4) was

employed to sequester the Li + ions during the reaction of spiro-OMeTAD and LiTFSI.

The addition of 12-c-4 was hypothesized to inhibit the rate spiro-OMeTAD oxidation

(Figure 5.4).

Figure 5.4: 12-c-4 chelates Li + effectively “removing” the ions from solution and

preventing oxidation of spiro-OMeTAD.

Initial experiments to oxidize spiro-OMeTAD were carried out in solutions of tetrahydrofuran (THF), dichloromethane (DCM), dimethylformamide (DMF), and chlorobenzene with varying ratios spiro-OMeTAD:LiTFSI and monitored via UV-Visible spectroscopy. However, varying rates of spiro(TFSI) formation were observed over the course of 24 h, and the results were irreproducible. This was attributed to inconsistent

142 rates of oxygen diffusion through the solutions of spiro-OMeTAD/LiTFSI. Attempts were made to mitigate large differences in oxygen diffusion by either capping the solutions during experimental time frame or by mixing the reaction solution between UV- vis scans; neither of these attempts provided reproducible data. Additionally, we investigated the necessity of lithium in the dopant to oxidize spiro-OMeTAD. Sodium

F tetrakis(3,5-bis(trifluoromethyl)phenyl)borate (NaBAr 4) and sodium tetrafluoroborate

(NaBF 4) and the analogous lithium salts were tested; only salts containing lithium led to spiro-OMeTAD oxidation (Table 5.1).

Table 5.1: Dopants screened in spiro-OMeTAD oxidation.

F F Dopant LiTFSI LiNO 3 Li(Otf) LiBAr 4 LiBF 4 NaBAr 4 NaBF 4 Oxidation a yes yes yes yes yes no no a UV-vis absorption at 520 nm.

Solid-state UV-vis has been used previously to observe spiro-OMeTAD oxidation. 13 A 1:1 solution of spiro-OMeTAD and LiTFSI was deposited onto a microscope slide, and the concentration of spiro(TFSI) (520 nm) was observed to increase over the course of several hours (Figure 5.5).

143

Spiro-OMeTAD

Spiro(TFSI)

Figure 5.5: Solid-state oxidation of spiro-OMeTAD with LiTFSI. Absorptions at ca.

400 nm and 520 correspond to spiro-OMeTAD and spiro(TFSI), respectively.

In order to suppress oxidation of spiro-OMeTAD, 12-c-4 was added in increasing ratios. Continuing with solid-state UV-vis, solutions of spiro-OMeTAD and LiTFSI (1:1) were prepared in THF. 12-c-4 was then added to the spiro-OMeTAD/LiTFSI solution to desired ratios (i.e. 1:1, 1:10, etc). These reaction solutions were cast onto a glass slide and analyzed via UV-vis. Figure 5.6 shows the absorption data for spiro(TFSI) at 520 nm over the course of three hours for different ratios of spiro-OMeTAD:12-c-4. In each case oxidation was not completely suppressed. Surprisingly, the addition of 12-c-4 had a negligible effect on the final oxidation of spiro-OMeTAD – each ratio of spiro-

OMeTAD:12-c-4 showed similar final absorptions.

144

Figure 5.6: Spiro-OMeTAD oxidation with varying amounts of CE.

Figure 5.7 shows the spectra from solid-state UV-vis analyses for spiro-OMeTAD oxidation with various amounts of 12-c-4. There was no appreciable difference in the extent of spiro-OMeTAD oxidation until a spiro-OMeTAD:12-c-4 of 1:30 was used. Like with solution state experiments irreproducibility plagued these experiments. Differences in observed oxidation (as well as oxidation suppression) were observed, and are likely a result of differences in film thickness. Despite precisely preparing solutions of spiro-

OMeTAD, LiTFSI, and 12-c-4, preparing identical films on the glass slides was not possible.

145

Spiro(TFSI)

Figure 5.7: Spiro-OMeTAD oxidation with varying amounts of 12-c-4 deposited on

glass slides over the course of 3 h. Vast differences and inconsistencies in absorption

at 520 nm highlights inconsistencies in sample preparation.

5.2.2 Perovskite Synthesis

As mentioned previously, CH 3NH 3PbI 3 has been the most studied organic/inorganic hybrid perovskite researched, and with its 3-D structure it has unusual optoelectronic properties useful for the design and fabrication of PSCs. 2 With the general

formula ABX 3 perovskites may be chemically versatile. Several successful substitutions

have been made at B and X sites of the perovskite, but substitution at A is limited by the

volume of the perovskite lattice. Other than methylammonium, formamidinium 16, 17, 18, 19 and cesium cations 20, 21, 22 are the only successful substitutions at site A. The 3-D structure of the perovskite is not maintained when larger organic cations like

146

+ + 23 CH 3CH 2NH 3 or (CH 3)2NH 2 are substituted for A. However, organic cations

isoelectronic with methylammonium may yield a 3-D perovskite structure when

substituted at A. We attempted to prepare perovskites with substitution at A with

hydrazinium iodide (NH 2NH 3I) and hydroxylammonium iodide (HONH 3I) to yield

(NH 2NH 3)PbI 3 1 and (HONH 3)PbI 3 2.

Perovskites 1 and 2 were prepared by vapor diffusion a non-solvent (methylene

chloride) into solutions of PbI 2 and hydrazinium iodide or hydroxylammonium iodide. A method previously described by Dr. Gregory Imler featured vapor diffusion of these solutions in the presence of added water to prepare hydrated perovskite precursors. It was initially thought that these hydrated materials would lead to a 3-D perovskite structure

24, though annealing to remove the water as this process is spontaneous for CH 3NH 3PbI 3.

25 Exact quantities of water were not reported, however, and attempts to add “small”

quantities (a few microliters) of water only lead to phase separation. We then relied upon

the natural presence of water in the solvents used ( vide infra ) to provide enough water to

form hydrated perovskites.

X-ray diffraction is a common technique to analyze the crystal lattice of the

perovskites, but developing other analyses for a faster identification of the materials

would be advantageous. UV-Visible spectroscopy was attempted to confirm that the same

material was prepared from each time materials were synthesized. Figure 5.8 shows the

UV-vis spectra for crystals of 1 prepared in DMF or DMSO. The identical spectra were

encouraging that the same material had been prepared, but the x-ray analyses of these

materials concluded that a perovskite structure was not prepared. Thermogravimetric

147 analysis (TGA) was used to determine if hydrated materials were formed. Figure 5.9 shows a thermogram for a sample of 1; no loss of water was detected suggesting the hydrated materials were not formed.

Figure 5.8: Crystals of 1 prepared from different solvents were analyzed via UV-Vis

to develop a finger printing method of perovskite identification.

148

Figure 5.9: TGA was used to determine if hydrated perovskites were prepared.

Attempts to prepare 2 were not successful, and only sporadic success was achieved with the hydrazinium iodide. A 1-D perovskite structure was prepared as concluded through x-ray diffraction, but this was not reproducible. Figure 5.10 shows the crystal structure of 1 as a 1-D perovskite structure, and Figure 5.11 shows DMSO-ligated

PbI 2 – a common phenomenon was to recrystallize starting materials.

149

Figure 5.10: A 1-D perovskite structure was only attained once with hydrazinium

iodide.

Figure 5.11: DMSO solvated crystal structure of PbI3 which was recovered from an

attempt to prepare 1.

150

5.3 Conclusion

The state-of-the-art HTM spiro-OMeTAD must be doped in order to make a competent hole transporter, and LiTFSI is commonly used to facilitate spiro-OMeTAD’s reaction with atmospheric oxygen. The role of lithium during the oxidation of spiro-

OMeTAD was investigated, but the experiments performed herein only highlighted the complexity of the doping process. Solution studies revealed that irreproducible kinetic profiles may be attributed to varying rates of oxygen diffusion through solutions of spiro-

OMeTAD/LiTFSI. Using a crown ether to sequester the Li + ions was thought to suppress spiro-OMeTAD oxidation, but oxidation suppression was not observed until a vast excess of 12-c-4 was used (1:30 sprio-OMeTAD:CE). At such high quantities of 12-c-4 all of the Li + ions are reasonably assumed to be sequestered from interacting with spiro-

OMeTAD, but oxidation of the HTM was observed nonetheless suggesting that lithium is not required for the doping process. However, when screening other dopants we observed no oxidation of spiro-OMeTAD if sodium salts were used (e.g. NaBF 4 vs. LiBF 4).

Concurrent to spiro-OMeTAD oxidation was the preparation of novel perovskites

1 and 2. Attempts to prepare perovskite crystals with novel substitution of the A site with the hydrazinium iodide and hydroxylammonium iodide cations were not successful.

Analytical methods other than x-ray diffraction were used to aid in identification of the perovskites, but the irreproducibility of 1 and 2 precluded the use of UV-vis and TGA.

Using vapor diffusion to grow crystals of 1 and 2 often led to the recrystallization of the individual reagents. Ultimately, 2 was never successfully prepared, and an irreproducible

1-D structure of 1 resulted. 151

5.4 Experimental

Chemicals. Chemicals and solvents were used as received from Sigma Aldrich

Chemical Co. and Fischer Scientific.

Apparatus. UV-Visible analyses were performed with a Shimadzu UV-1800 series spectrophotometer. X-ray diffraction was performed on a Bruker APEX-II CCD detector. Thermogravimetric analyses were performed on a TA Instruments Hi-Res TGA

2950 Thermogravimetric Analyzer under N 2 with a heating rate of 10 ºC per minute from room temperature to 800ºC. Solid-state UV-vis samples were prepared using a Laurell

Technologies Corp. CZ-650 coater.

Spiro-OMeTAD Oxidation. Separate solutions of spiro-OMeTAD (10.0 mg,

0.27 mM) and LiTFSI (200 mg, 0.69 mM) were prepared in THF. Once the solutions were prepared, the LiTFSI solution was added to the spiro-OMeTAD solution. This combined solution was used for spin-coating microscope slides which were cleaned with methanol. The spiro-OMeTAD/LiTFSI solution was dropped onto the microscope slide using a Pasteur pipette, and the slide was spun for 15 s @ 2500 rpm. This casting method was repeated three times to coat one slide. When using 12-c-4, the crown ether was added to the spiro-OMeTAD/LiTFSI solution directly using a microliter syringe. Microscope slides were prepared as previously mentioned.

Preparation of HONH 3I. Aqueous hydroiodic acid (57 wt. %) was added dropwise to a cooled (0 °C) hydroxylamine solution (50 wt. % in water) until a white precipitate formed (pH < 3). The product was filtered and washed with cold ethanol.

152

Preparation of NH 2NH 3I. Hydrazinium iodide was prepared according to literature procedure 26 by Dr. Gregory Imler.

Preparation of (NH 2NH 3)PbI 3 (1) and (HONH 3)PbI 3 (2). Slow vapor diffusion of methylene chloride into equimolar solutions of PbI 2 and the iodide in dimethylformamide produced the crystalline products.

Acknowledgements. The work conducted in this chapter would not have been possible without Alexa Wallace’s (’17) contribution to the oxidation of spiro-OMeTAD and Eromon Asikhia’s (’18) synthetic contributions to perovskite synthesis. Dr. Gregory

Imler (’14) was integral in perovskite synthesis methodology.

153

5.5 References Cited

1. Calió, L.; Kazim, S.; Grätzel, M.; Ahmad, S., Hole-Transport Materials for Perovskite Solar Cells. Angewandte Chemie International Edition 2016, 55 (47), 14522- 14545.

2. Green, M. A.; Ho-Baillie, A.; Snaith, H. J., The emergence of perovskite solar cells. Nature Photonics 2014, 8, 506.

3. Grätzel, M., The light and shade of perovskite solar cells. Nature Materials 2014, 13 , 838.

4. Pathak, S.; Sepe, A.; Sadhanala, A.; Deschler, F.; Haghighirad, A.; Sakai, N.; Goedel, K. C.; Stranks, S. D.; Noel, N.; Price, M.; Hüttner, S.; Hawkins, N. A.; Friend, R. H.; Steiner, U.; Snaith, H. J., Atmospheric Influence upon Crystallization and Electronic Disorder and Its Impact on the Photophysical Properties of Organic–Inorganic Perovskite Solar Cells. ACS Nano 2015, 9 (3), 2311-2320.

5. Kim, H.-S.; Lee, C.-R.; Im, J.-H.; Lee, K.-B.; Moehl, T.; Marchioro, A.; Moon, S.-J.; Humphry-Baker, R.; Yum, J.-H.; Moser, J. E.; Grätzel, M.; Park, N.-G., Lead Iodide Perovskite Sensitized All-Solid-State Submicron Thin Film Mesoscopic Solar Cell with Efficiency Exceeding 9%. Scientific Reports 2012, 2, 591.

6. Lee, M. M.; Teuscher, J.; Miyasaka, T.; Murakami, T. N.; Snaith, H. J., Efficient Hybrid Solar Cells Based on Meso-Superstructured Organometal Halide Perovskites. Science 2012, 338 (6107), 643-647.

7. Burschka, J.; Pellet, N.; Moon, S.-J.; Humphry-Baker, R.; Gao, P.; Nazeeruddin, M. K.; Grätzel, M., Sequential deposition as a route to high-performance perovskite- sensitized solar cells. Nature 2013, 499 , 316.

8. Chen, S.; Lei, L.; Yang, S.; Liu, Y.; Wang, Z.-S., Characterization of Perovskite Obtained from Two-Step Deposition on Mesoporous Titania. ACS Applied Materials & Interfaces 2015, 7 (46), 25770-25776.

9. Wang, R.; Mujahid, M.; Duan, Y.; Wang, Z.-K.; Xue, J.; Yang, Y., A Review of Perovskites Solar Cell Stability. Advanced Functional Materials 0 (0), 1808843. 154

10. Agarwala, P.; Kabra, D., A review on triphenylamine (TPA) based organic hole transport materials (HTMs) for dye sensitized solar cells (DSSCs) and perovskite solar cells (PSCs): evolution and molecular engineering. Journal of Materials Chemistry A 2017, 5 (4), 1348-1373.

11. Zhang, W.; Cheng, Y.; Yin, X.; Liu, B., Solid-State Dye-Sensitized Solar Cells with Conjugated Polymers as Hole-Transporting Materials. Macromolecular Chemistry and Physics 2011, 212 (1), 15-23.

12. Leijtens, T.; Lim, J.; Teuscher, J.; Park, T.; Snaith, H. J., Charge Density Dependent Mobility of Organic Hole-Transporters and Mesoporous TiO2 Determined by Transient Mobility Spectroscopy: Implications to Dye-Sensitized and Organic Solar Cells. Advanced Materials 2013, 25 (23), 3227-3233.

13. Nguyen, W. H.; Bailie, C. D.; Unger, E. L.; McGehee, M. D., Enhancing the Hole-Conductivity of Spiro-OMeTAD without Oxygen or Lithium Salts by Using Spiro(TFSI)2 in Perovskite and Dye-Sensitized Solar Cells. Journal of the American Chemical Society 2014, 136 (31), 10996-11001.

14. MacDiarmid, A. G., “Synthetic Metals”: A Novel Role for Organic Polymers (Nobel Lecture). Angewandte Chemie International Edition 2001, 40 (14), 2581-2590.

15. Wade, L. G. J., Organic Chemistry . 6 ed.; Pearson Education, Inc.: New Jersey, 2006.

16. Mitzi, D. B.; Liang, K., Synthesis, Resistivity, and Thermal Properties of the Cubic Perovskite NH2CH=NH2SnI3and Related Systems. Journal of Solid State Chemistry 1997, 134 (2), 376-381.

17. Pellet, N.; Gao, P.; Gregori, G.; Yang, T.-Y.; Nazeeruddin, M. K.; Maier, J.; Grätzel, M., Mixed-Organic-Cation Perovskite Photovoltaics for Enhanced Solar-Light Harvesting. Angewandte Chemie International Edition 2014, 53 (12), 3151-3157.

18. Jeon, N. J.; Noh, J. H.; Yang, W. S.; Kim, Y. C.; Ryu, S.; Seo, J.; Seok, S. I., Compositional engineering of perovskite materials for high-performance solar cells. Nature 2015, 517 , 476.

155

19. Eperon, G. E.; Stranks, S. D.; Menelaou, C.; Johnston, M. B.; Herz, L. M.; Snaith, H. J., Formamidinium lead trihalide: a broadly tunable perovskite for efficient planar heterojunction solar cells. Energy & Environmental Science 2014, 7 (3), 982-988.

20. Chung, I.; Lee, B.; He, J.; Chang, R. P. H.; Kanatzidis, M. G., All-solid-state dye- sensitized solar cells with high efficiency. Nature 2012, 485 , 486.

21. Kulbak, M.; Cahen, D.; Hodes, G., How Important Is the Organic Part of Lead Halide Perovskite Photovoltaic Cells? Efficient CsPbBr3 Cells. The Journal of Physical Chemistry Letters 2015, 6 (13), 2452-2456.

22. Wu, K.; Liang, G.; Shang, Q.; Ren, Y.; Kong, D.; Lian, T., Ultrafast Interfacial Electron and Hole Transfer from CsPbBr3 Perovskite Quantum Dots. Journal of the American Chemical Society 2015, 137 (40), 12792-12795.

23. Stoumpos, C. C.; Kanatzidis, M. G., The Renaissance of Halide Perovskites and Their Evolution as Emerging Semiconductors. Accounts of Chemical Research 2015, 48 (10), 2791-2802.

24. Imler, G. H.; Li, X.; Xu, B.; Dobereiner, G. E.; Dai, H.-L.; Rao, Y.; Wayland, B. B., Solid state transformation of the crystalline monohydrate (CH3NH3)PbI3(H2O) to the (CH3NH3)PbI3 perovskite. Chemical Communications 2015, 51 (56), 11290-11292.

25. Hao, F.; Stoumpos, C. C.; Liu, Z.; Chang, R. P. H.; Kanatzidis, M. G., Controllable Perovskite Crystallization at a Gas–Solid Interface for Hole Conductor-Free Solar Cells with Steady Power Conversion Efficiency over 10%. Journal of the American Chemical Society 2014, 136 (46), 16411-16419.

26. Gilbert, E. C.; Decius, J. C., The Iodides of Hydrazine1. Journal of the American Chemical Society 1958, 80 (15), 3871-3872.

156