Accessing Synthetically-Challenging Isoindole-Based Materials for Assessment in Organic Photovoltaics Via Chemical and Engineering Methodologies

Accessing Synthetically-Challenging Isoindole-Based Materials for Assessment in Organic Photovoltaics via Chemical and Engineering Methodologies

Jeremy Dang

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Department of Chemical Engineering and Applied Chemistry University of Toronto

Accessing Synthetically-Challenging Isoindole-Based Materials for Assessment in Organic Photovoltaics via Chemical and Engineering Methodologies

Jeremy Dang

Doctor of Philosophy

Department of Chemical Engineering and Applied Chemistry University of Toronto

2016 Abstract

Isoindoles are a broad class of compounds that comprise a very small space within the domain of established photoactive materials for organic photovoltaics (OPVs). Given this scarcity, combined with the performance appeal of presently and well known isoindole-based compounds such as the phthalocyanines, it is a worthy undertaking to develop new materials in this domain.

This thesis aims to bring to light the suitability of five novel, or underexplored, classes of isoindole-based materials for OPVs. These classes are the boron subphthalocyanine (BsubPc) polymers, oxygen-bridged dimers of BsubPcs (μ-oxo-(BsubPc) 2), boron subnaphthalocyanines

(BsubNcs), group XIII metal complexes of 1,3-bis(2-pyridylimino)isodinoline (BPI), and the boron tribenzosubporphyrins (BsubPys).

The synthesis of these materials was proven to be challenging as evident in their low isolated yields, lengthy synthetic and purification processes, and/or batch-to-batch variations. This outcome was not surprising given their undeveloped chemical nature. The photo- and electro- physical properties were characterized and shown to be desirable for all classes other than the

group XIII metal complexes of BPI for OPVs. μ-Oxo-(BsubPc) 2 and BsubNcs show promise in this application while BsubPc polymers and BsubPys will be subjects of future exploration.

The results from the work herein aid to develop and strengthen the fundamental understanding of the structure-property relationships of isoindole derivatives. On a broader scale, the work demonstrates their versatility as functional materials for OPVs and their possible expansion to other organic electronic technologies like organic light emitting diodes and organic field effect transistors.

iii

Acknowledgments

I would like to begin by expressing my deepest gratitude to my supervisor, Professor Timothy P. Bender, for his support and guidance throughout my entire degree. You have taught me valuable lessons, given me the freedom to explore my research interests, and most importantly, broadened my perspective on the real-world applications of science. I will always appreciate your mentorship and your encouragement in having a well-balanced lifestyle. Thank you for the opportunity.

I would also like to thank my committee members, Professor Dwight S. Seferos and Professor Christopher M. Yip, for their guidance in my Ph.D. journey. It has been a long and difficult one, but truly a humbling and worthwhile experience. Thank you for your insights and feedback in this process.

To the Benderites/Weekday Benders/Bender Minions/Bending Bad Crew/Average Benders, thank you for a wonderful time. I will always remember and cherish the good moments… and even the bad ones. I'd like to specifically give thanks to Dr. Benoit Lessard, Anjuli Szawiola, Hasan Raboui, Katie Sampson, David Josey, Trevor Plint, Dr. Brett Kamino, Dr. John Grande, Dr. Jessica Virdo, Dr. Jeffrey Castrucci, Dr. Graham Morse, Dr. Andrew Paton, Dr. Catherine Bonnier, Mabel Fulford, Mike Gretton, Jane Cho, Ahmed Balawi, Mona Khatibi, Alex Peltekoff, Alaa Sifate, Stephanie Nyikos, Cynthia Cheung, Devon Holst, Richard Garner, Esmeralda Bukuroshi, and Aleksa Dovijarski. A departing message to all present and future Benderites, this is a great group filled with great people. Keep our unique dynamic identity strong within the group, as well as within the department.

To my parents and brothers, I appreciate your efforts in trying to comprehend my research work. I am aware that this was not a simple task. I'd like to thank you for the endless support.

Lastly, I'd like to thank my dear Anna. While I'm able to convey my scientific findings in this dissertation, I find that my words cannot do the same in expressing my admiration and appreciation for you. You exemplify the image of a hard-working, driven person and you have constantly aspired me to strive for more. I admittedly don't say it enough, but I love you very much and I look forward to the next chapter of our lives.

Table of Contents

List of Tables viii List of Figures ix List of Schemes xiii List of Abbreviations xiv

Chapter 1: Introduction 1 1.1 The Solar Energy Paradigm 1 1.2 Brief History and Key Discoveries 2 1.3 Organic Photovoltaics (OPVs) - The Future of PV Technologies? 3 1.4 Operating Principles of OPVs 4 1.5 Device Architectures 6 1.6 Methods of Fabrication 8 1.7 Device Performance Definition 9 1.8 Optoelectronic Effects on Performance 11 1.9 Commonly Employed Organic Photoactive Materials 14 1.10 Isoindole-Based Compounds in OPVs 17 1.11 Thesis Statement 25 1.12 References 28 Chapter 2: A boron subphthalocyanine polymer: poly(4-methylstyrene)-co -poly(phenoxy-boron-subphthalocyanine) 36 2.1 Abstract 37 2.2 Introduction 37 2.3 Results and Discussion 38 2.4 Conclusions 51 2.5 References 52

Chapter 3: Process for the synthesis of symmetric and unsymmetric oxygen bridged dimers of boron subphthalocyanines (μ-oxo(BsubPcs) 2s) 55 3.1 Abstract 56 3.2 Introduction 56 3.3 Results and Discussion 59 3.4 Conclusions 76 3.5 References 77

Chapter 4: Growth of μ-oxo-boron subphthalocyanine (μ-oxo(BsubPcs) 2) thin films via organic chemical vapor deposition (OCVD) 79 4.1 Abstract 80 4.2 Introduction 80 4.3 Results and Discussion 83 4.4 Conclusions 90 4.5 References 91

Chapter 5: The mixed alloyed chemical composition of chloro-(chloro) n-boron subnaphthalocyanines dictate their physical properties and performance in organic photovoltaic devices 94 5.1 Abstract 95 5.2 Introduction 95 5.3 Results and Discussion 100 5.4 Conclusions 119 5.5 References 121 Chapter 6: Phenoxy boron subnaphthalocyanines: synthesis, properties, and their applications in organic planar heterojunction photovoltaics 125 6.1 Abstract 126 6.2 Introduction 126 6.3 Results and Discussion 129 6.4 Conclusions 137 6.5 References 138

Chapter 7: Boron, aluminum, gallium, and indium complexes of 1,3-bis(2-pyridylimino)isoindoline (BPI) 141 7.1 Abstract 142 7.2 Introduction 142 7.3 Results and Discussion 144 7.4 Conclusions 151 7.5 References 153 Chapter 8: A synthetic and engineering process to boron tribenzosubporphyrins 155 8.1 Abstract 156 8.2 Introduction 156 8.3 Results and Discussion 159 8.4 Conclusions 176 8.5 References 177 Chapter 9: Summary and Future Work 179 9.1 Summary 179 9.2 Future Work 185 9.3 References 199 Appendix: General Experimental Section 203 Appendix A: Supplementary Information for Chapter 2 226 Appendix B: Supplementary Information for Chapter 3 206 Appendix C: Supplementary Information for Chapter 4 249 Appendix D: Supplementary Information for Chapter 5 254 Appendix E: Supplementary Information for Chapter 6 298 Appendix F: Supplementary Information for Chapter 7 335 Appendix G: Supplementary Information for Chapter 8 349

vii

List of Tables

Table 1.1. Number of publications for each of the listed organic photoactive material. 15

Table 1.2. Summary of reported BsubNc synthesis. 23

Table 3.1. Photophysical and electronic properties of µ-oxo-(BsubPc) 2, F12 BsubPc-O-BsubPc, Cl 6BsubPc-O-BsubPc, and Cl 12 BsubPc-O-BsubPc. 68

Table 3.2. Average device parameters of PHJ OPVs of the following configuration: ITO/PEDOT:PSS/α-6T/acceptor/TPBi/Ag or ITO/PEDOT:PSS/MoO x/donor/C 60/70 /TPBi/Ag. 75

Table 5.1. Photophysical properties of Cl-Cl nBsubNcs. 112

Table 5.2. Electrochemical properties of Cl-Cl nBsubNcs. 113

Table 5.3. Electrochemical properties (DPV) and ultraviolet photoelectron spectroscopy (UPS) characteristics of Cl-Cl nBsubNcs. 114

Table 5.4. Average device parameters of PHJ OPVs of the following configuration: ITO/PEDOT:PSS/α-6T(55 nm)/Cl-Cl nBsubNc(25 nm)/BCP(10 nm)/Ag(80 nm) whereby the Cl-Cl nBsubNc layer is made of nitrobenzene-Cl-Cl nBsubNc, p-cymene-Cl-Cl nBsubNc, literature-Cl-Cl nBsubNc, or commercial-Cl-Cl nBsubNc. 118

Table 6.1. X-ray diffraction results. 132

Table 6.2. Photophysical and electronic properties of phenoxy Cl nBsubNcs. 133

Table 8.1. Device parameters of PHJ OPVs of the following configuration: ITO/PEDOT:PSS/MeO-n-BsubPy/Cl-Cl 12 BsubPc/BCP/Ag. 170

Table 9.1. Electronic properties of μ-oxo-(BsubPc) 2 and the unsymmetric variants. 187

Table 9.2. Device performance characteristics reported by Torres et al . and Jones et al . 188

Table 9.3. Attempted syntheses of Br-BsubNc. 195

Table 9.4. Attempted syntheses of F-BsubNc. 196

viii

List of Figures

Figure 1.1. Global energy consumption (in TW) per year since 1986. 1

Figure 1.2. Operating mechanism for Schottky cells. 6

Figure 1.3. Operating mechanism for donor-acceptor (D-A) heterojunction OPV cells. 6

Figure 1.4. Left to right : Device architecture of a Schottky cell, planar heterojunction (PHJ), bulk heterojunction (BHJ), and a tandem cell. 7

Figure 1.5. An example of a (a) vacuum deposition and (b) solution processing process. 9

Figure 1.6. A typical current density vs. voltage (J-V) curve of an OPV cell under illumination. 11

Figure 1.7. Solar irradiance spectrum (black, sea level) overlaid with the absorption spectrum (pink) of Cl-BsubPc. 13

Figure 1.8. Number of publications on polymer-based ( red ) and small-molecule-based (blue ) materials for OPVs since the year 2000. 17

Figure 1.9. Chemical structure of a generic metallophthalocyanine and boron subphthalocyanine showing the axial (R ax ) and peripheral (R p) positions. 18

Figure 1.10. Chemical structure of μ-oxo-(BsubPc) 2. 21

Figure 1.11. The five classes of isoindole-based compounds studied in this dissertation showing their progress and future direction. 27

Figure 2.1. GPC chromatograms and extracted UV-vis absorption spectra (inset) of BsubPc polymers 11a (a) and 11b (b). 45

Figure 2.2. UV-vis absorption (blue) and photoluminescence (PL, red) spectra of 3,4-dimethylphenoxy-BsubPc and polymers 11a and 11b (as indicated). 47

Figure 2.3. GPC chromatograms and extracted UV-vis absorption spectra ( left : lower R T; right : higher R t) of BsubPc polymers 11a (a) and 11b (b) within two years of storage. 49

Figure 3.1. 2-D structure and ellipsoid plot (35% probability) of μ-oxo-(BsubPc) 2 (CCDC deposition number: 914863). 57

Figure 3.2. Absorption spectrum (pink) of μ-oxo-(BsubPc) 2 overlaid with the solar irradiance (black, sea level) spectrum. 58

Figure 3.4. Absorption (blue) and photoluminescence (red) spectra of (a) μ-oxo-(BsubPc) 2, (b) F 12 BsubPc-O-BsubPc, (c) Cl 6BsubPc-O-BsubPc, and (d) Cl 12 BsubPc-O-BsubPc in toluene solutions at room temperature. 66

Figure 3.5. UV-vis absorption (solid) and PL (λ exc = 530 nm, dash) spectra (concentration -6 = 2.30 x 10 M) of (a) µ-oxo-(BsubPc) 2, (b) F12 BsubPc-O-BsubPc, (c) Cl 6BsubPc-O-BsubPc, and (d) Cl 12 BsubPc-O-BsubPc in DMF (violet), 10:1 (blue), 1:1 (green), 1:10 (orange), 1:50 (red), and 1:200 (black) DMF/H 2O ( v/v ). 71

Figure 3.6. Cyclic voltammograms of (a) μ-oxo-(BsubPc) 2, (b) F12 BsubPc-O-BsubPc, (c) Cl 6BsubPc-O-BsubPc, and (d) Cl 12 BsubPc-O-BsubPc in DCM with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference ( E1/2,red = -0.012 V vs. Ag/AgCl) at room temperature. 72

Figure 3.7. Chemical structure of donor and acceptor materials used in this study. 74

Figure 3.8. Schematic of the device structure for testing μ-oxo-(BsubPc) 2 and Cl-BsubPc as an (a) electron donor and (b) electron acceptor. 75

Figure 4.1. Chemical and crystal packing structure of μ-oxo-(BsubPc) 2 (CCDC deposition Number: 914863). 81

Figure 4.2. Schematic diagram of (a) the OCVD apparatus and (b) the OCVD process within the reactor. 85

Figure 4.3. Picture of the glass substrate at three different locations and (b) a close-up picture (black and white camera) showing the crystallites following an OCVD (Method 1.1). 86

Figure 4.4. (a) Picture of the glass substrate and (b) a close-up picture (black and white camera) following an OCVD (Method 1.8). 88

Figure 5.1. Generic chemical structures of boron subphthalocyanine (BsubPc) and boron subnaphthalocyanine (BsubNc) and the generic synthesis of BsubPcs and BsubNcs, where X = Cl, Br, Ph, or other. 96

Figure 5.2. Mass spectrum of a sublimed sample of Cl-Cl nBsubNc made via an adaptation of the Kennedy method indicating the presence of Cl-Cl 1BsubPc, Cl-Cl 2BsubNc, and Cl-Cl 3BsubNc. 103

Figure 5.4. Overlay of the literature-Cl-Cl nBsubNc (blue), commercial-Cl-Cl nBsubNc (red), p-cymene-Cl-Cl nBsubNc (green) and nitrobenzene-Cl-Cl nBsubNc (purple) normalized Cl 2p core-level XPS spectra. 110

Figure 5.5. Overlay of the literature-Cl-Cl nBsubNc (blue), commercial-Cl-Cl nBsubNc (red), p-cymene-Cl-Cl nBsubNc (green) and nitrobenzene-Cl-Cl nBsubNc (purple) normalized UPS valence band spectra. 116

Figure 5.6. (a) J/V characteristics and (b) external quantum efficiency (EQE) spectra of p-cymene-Cl-Cl nBsubNc, literature-Cl-Cl nBsubNc, and commercial-Cl-Cl nBsubNc. 117

Figure 6.1. Chemical structure of boron subphthalocyanine (BsubPc) and boron subnaphthalocyanine (BsubNc). 127

Figure 6.2. Ellipsoid plot (50% probability) showing the structure and atom numbering scheme of PhO-Cl nBsubNc. 131

Figure 6.3. Overlay of the PhO-Cl nBsubNc (blue) and F 5-Cl nBsubNc (red) (a) normalized absorption spectra and (b) normalized photoluminescence (PL, λ ex = 630 nm) spectra. 133

Figure 6.4. Cyclic voltammograms of (a) PhO-Cl nBsubNc and (b) F 5-Cl nBsubNc in DCM with 0.1M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference (E 1/2,red = -0.012 V vs. Ag/AgCl) at room temperature. 134

Figure 6.5. (a) J/V characteristics and (b) external quantum efficiency (EQE) spectra of PhO-Cl nBsubNc, F5-Cl nBsubNc, and Cl-Cl nBsubNc. 135

Figure 7.1. From left to right: structure of 1,3-bis(2-pyridylimino)isoindoline (BPI), phthalocyanine (Pc), and boron subphthalocyanine (BsubPc). 143

Figure 7.2. UV-vis absorption spectrum of metal-free BPI. 144

Figure 7.3. Ellipsoid plot (50% probability) showing the structure and atom numbering scheme for the (a) BPI·GaCl 2 complex (CCDC deposition number: 904845) and the (b) BPI·BF 2 complex (CCDC deposition number: 904844). 146

Figure 7.4. UV-vis absorption spectra of BPI, BPI·BF 2, BPI·AlCl 2, BPI·GaCl 2, and BPI·InCl 2 in degassed dichloromethane solutions at room temperature. 148

Figure 7.5. Fluorescence emission spectra (λ ex = 352 nm, highest and longest λ max of absorption as indicated) for BPI, BPI·BF 2, BPI·AlCl 2, BPI·GaCl 2, and BPI·InCl 2 in degassed dichloromethane solutions at room temperature. 150

Figure 8.1. From left to right: chemical structure of a BsubPc, n-BsubPy, and mAr-BsubPy. 156

Figure 8.2. Absorption spectrum of a n-BsubPy (green) and mPh-BsubPy (orange) overlaid with the solar irradiance (black, sea level) spectrum. 159

Figure 8.3. A picture of the train sublimation experiment (Method 5.1) with a close-up image of the maroon red band ( n-BsubPy). 164

Figure 8.4. Pictures of the OVPD experiment (Method 5.5) showing the seven ITO on glass substrates, each with a pre-deposited layer of PEDOT:PSS. 165

Figure 8.5. Cartoon image showing the film thickness at various locations across Substrate #2-6. 166

Figure 8.6. Cyclic (left) and differential pulse (right) voltammetry traces of MeO-n-BsubPy in DCM with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference at room temperature. 167

Figure 8.7. Energy level (eV) band diagram of MeO-n-BsubPy, Cl-Cl 12 BsubPc, C 60 , and C70 . 168

Figure 8.8. (a) J/V characteristics and (b) external quantum efficiency spectra. PHJ OPVs of the following configuration: ITO/PEDOT:PSS/MeO-n-BsubPy/Cl-Cl 12 BsubPc/BCP/Ag. 169

Figure 8.9. Cyclic (left) and differential pulse (right) voltammetry traces of MeO-mPh-BsubPy in DCM with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference at room temperature. 173

Figure 8.10. Pictures of the OVPD experiment (Method 6.12) showing the seven ITO on glass substrates, each with a pre-deposited layer of PEDOT:PSS. 175

Figure 9.1. The five classes of isoindole-based compounds studied in this dissertation showing their progress and future work. 184

Figure 9.2. Band diagram of energy levels (eV) for α-6T, Cl-BsubPc, μ-oxo-(BsubPc) 2, Cl-Cl 6BsubPc, Cl-Cl 12 BsubPc, C 60 , and C 70 . 187

Figure 9.3. The five proposed synthetic routes to Cl-Cl 6BsubNc. 191

Figure 9.4. Cyclic voltammogram of BPI in DCM with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference (E 1/2,red = -0.012 V vs. Ag/AgCl) at room temperature. 197

xii

List of Schemes

Scheme 1.1 Synthesis of Cl-BsubPc, Cl-BsubNc, unsymmetric Pc (black), and unsymmetric Nc (red). 20

Scheme 2.1. Attempted synthesis of BsubPc-containing styrenic monomers via the pre-polymerization functionalization approach. 39

Scheme 2.2. The synthesis of BsubPc-containing copolymers via the post-polymerization functionalization approach. 40

Scheme 2.3. The synthesis of BsubPc-containing copolymers 13 and 15 via the post-polymerization functionalization approach. 50

Scheme 3.1. Synthesis of μ-oxo-(BsubPc) 2. 60

Scheme 3.2. Synthesis of F 12 BsubPc-O-BsubPc, Cl 6BsubPc-O-BsubPc, and Cl 12 BsubPc-O-BsubPc. 64

Scheme 4.1. Synthetic route to μ-oxo-(BsubPc) 2. 81

Scheme 6.1. Synthesis of BsubNc from 2,3-dicyanonaphthalene. 128

Scheme 6.2. Synthesis of PhO-Cl nBsubNc and F 5-Cl nBsubNc. 130

Scheme 7.1. Synthesis of BPI and the proposed synthesis of its group XIII complexes and its post-metallated derivatives. 143

Scheme 7.2. Synthesis of BPI·BF 2, BPI·AlCl 2, BPI·GaCl 2, and BPI·InCl 2. 145

Scheme 8.1. Synthesis of HO-n-BsubPy (left route) and HO-mAr-BsubPy (right routes) and the ellipsoid plot (50% probability) of HO-n-BsubPy (CCDC deposition number: 281754) and Ph 3SiO-mPh-BsubPy (CCDC deposition number: 676629). 158

Scheme 9.1. Proposed synthesis of (F12 BsubPc) 2O, (Cl 6BsubPc) 2O, and (Cl 12 BsubPc) 2O. 189

Scheme 9.2. Attempted synthesis of α,α,α',α'-tetrabromo-α,α'-dichloro-o-xylene. 192

Scheme 9.3. Attempted synthesis of 1,4-dichloro-2,3-dicyanonaphthalene from 2,3-dicyanonaphthalene. 192

Scheme 9.4. Attempted synthesis of Cl-Cl 6BsubNc from Cl-Cl nBsubNc. 193

xiii

List of Abbreviations

BHJ - bulk heterojunction BPI - 1,3-bis(2-pyridylimino)isoindoline BsubNc - boron subnaphthalocyanine BsubPc - boron subphthalocyanine BsubPy - boron tribenzosubporphyrin CV - cyclic voltammetry CVD - chemical vapor deposition DPV - differential pulse voltammetry

Eg - energy band gap ETL - electron transport layer EQE - external quantum efficiency FF - fill factor FMO - frontier molecular orbital GC - gas chromatography GHG - greenhouse gas GPC - gel permeation chromatography HOMO - highest occupied molecular orbital HPLC - high performance liquid chromatography HRMS - high resolution mass spectroscopy ITO - indium tin oxide

JSC - short-circuit current density LRMS - low resolution mass spectroscopy LUMO - lowest unoccupied molecular orbital NLO - non-linear optic NMP - nitroxide mediated polymerization NMR - nuclear magnetic resonance OE - organic electronic OFET - organic field effect transistor OLED - organic light emitting diode OPV - organic photovoltaic

xiv

OSC - organic solar cell OCVD - organic chemical vapor deposition OVPD - organic vapor phase deposition Pc - phthalocyanine PCE - power conversion efficiency PDI - polydispersity index PEDOT:PSS - poly(3,4-ethylenedioxythiophene) poly(styrenesulfonate) PHJ - planar heterojunction PL - photoluminescence

Rf - retention factor

Rt - retention time TLC - thin layer chromatography UPS - ultraviolet photoelectron spectroscopy UV-vis or UV/Vis - ultraviolet-visible

VOC - open-circuit voltage XPS - X-ray photoelectron spectroscopy

xv 1

Chapter 1 Introduction

1.1 The Solar Energy Paradigm For three billion years and counting, photosynthetic organisms have been exploiting sunlight for the production of chemical fuel. 1 This phenomenon has inspired scientists to mimic nature for the generation of electrical energy, a pursuit that has grown stronger and stronger as the global energy consumption and demand continues to rise (Figure 1.1). 2 Presently, the world annual energy consumption is above 17 terawatt (TW) 2 and this number is expected to increase to 28 TW by 2050 3. Of the current energy usage, close to 70% of it is derived from the burning of fossil fuels – natural gas, crude oil, and coal. 4 The high fossil fuel usage creates a concern as their reserves are non-renewable and finite and will therefore, eventually deplete. To illustrate this concern, assuming that the consumption rate stays constant with time and that new reserves are not discovered, it will take about 50 years to fully exhaust all of the world's supply of crude oil. 4

Figure 1.1 . Global energy consumption (in TW) per year since 1986. 2

Another matter to consider is that the by-product of burning fossil fuels is the emission of greenhouse gases (GHGs) and their negative environmental impact. Since the Industrial Revolution, atmospheric concentrations of carbon dioxide, methane, and nitrous oxide have increased by about 40, 250, and 15%, respectively. Their further increase will likely lead to an extinction event for species particularly those in marine ecosystems, melting of the ice sheets leading to shortages of fresh water and a rise in the sea level, a disruption to agricultural productivity, and climate changes at the global scale. 5,6

The challenge with meeting the continuously growing electricity demand while reducing GHG emissions have sparked interests into pursuing alternative, sustainable, and renewable energy resources. A viable solution is the Sun as the surface of the Earth intercepts about 120,000 TW of solar energy per annum, 7 an amount that is almost 7000 times the current global energy consumption. 2 Moreover, pressure on fossil fuel reserves would be alleviated along with their associated negative environmental impact. Unfortunately, current photovoltaic (PV) technologies are not cost-competitive in comparison to fossil fuels, hindering their implementation on a large scale at the residential, commercial, industrial, and transportation level. 5 Further research and development in this field are necessary to improve power output ( i.e. light-to-electricity efficiency) while reducing their cost before PV technologies can be widely employed.

1.2 Brief History and Key Discoveries The origin of PVs began in 1839 with the discovery of the photovoltaic effect, a phenomenon where an electrical current is created upon the irradiation of light onto a material. 8 Photoconductivity was also observed for other materials following this discovery. 9-11 In 1883, the first PV cell using selenium and gold was developed, producing efficiency below 1%. 12 Similar efficiencies were also observed for a PV cell based on copper-cuprous oxide in 1927 13 and for silicon in 1941 14 . Despite the low power output, the latter finding led to the discovery of the first practical silicon-based PV cell made by Bell Laboratories with an efficiency of 6% in 1954. 15 Several years following this, efficiencies quickly rose to nearly 15%. 16 At the time, PV cells were too costly and were primarily used by the space program to power space vehicles and satellites. 17 Improvements and innovations in cell design and geometry, fabrication methods,

material purification techniques, new material integration, etc. over the next six decades have led to PV cells today with efficiencies as high as 46.0% 18 .

The development of PV technologies is divided into three waves of innovation with respect to the fabrication process and material usage. First generation cells are made of crystalline silicon (c-Si) wafers. These wafers generate cells with the best performance, except that they are very energy-intensive to produce. Despite this, these cells dominate about 80% of the PV market. To address the high-cost issue, second generation cells were established. These cells are thin film- based (few microns vs. hundreds of microns) consisting of inorganic materials such as amorphous silicon (a-Si), cadmium telluride (CdTe), or copper indium diselenide (CuInSe 2). While this generation greatly benefits from a materials-cost perspective, their performance is poorer compared to the first generation cells. Moreover, another drawback is the use of rare and toxic elements. 6 Third generation cells evolved in an effort to improve performance while reducing cost via the preservation of the thin film approach. Technologies of this generation include organic photovoltaics (OPVs), dye-sensitized solar cells, perovskite based solar cells, tandem cells, and quantum dot based solar cells. Their overall performances lag behind the prior two generations, however, their development have seen major strides in recent years. 6,19

1.3 Organic Photovoltaics (OPVs) – the Future of PV Technologies? As described earlier, c-Si-based PV cells are the best performing devices that dominate the market but suffer from high material-costs. 6 Although intensive research and development have significantly lowered their manufacturing cost, 20 they are still not as cost-effective as conventional fossil fuels and this factor will continue to hinder their large scale deployment. OPV technology has the potential to overcome this cost issue. Unlike the first two generations, OPVs make use of carbon-containing (organic) semiconductors to harness solar energy. These organic chromophores are much better light absorbers than c-Si or the inorganic materials found in the second generation cells and thus, there is a materials-saving associated with OPVs; thin films at the nanometer scale is often sufficient to absorb all photons at certain wavelengths. Another cost-saving feature is the potential to fabricate the cells on a large, high-throughput scale using printing technologies such as a roll-to-roll process. 20,21 Other benefits are that they

can be made flexible to adopt various desired shapes and sizes, they are lightweight to serve aesthetic purposes or applications where weight is a major concern, they are easy to install due to their light-weight and flexible features, they are durable compared to the brittle nature of c-Si, and their properties can be tailored for a specific purpose through chemical design and synthesis. 20-22 The disadvantage of OPVs is that their performance efficiencies are lower than c- Si cells. The highest reported efficiency of an OPV device is 11.5% 23 versus 27.6% 24 for c-Si. With that being said, although it is natural to target the highest possible efficiencies, it is actually not necessary. Having PV devices with moderate performance at a lower cost would suffice. For instance, covering less than 1% of the world’s deserts with 10% efficient cells would generate an amount of electricity equivalent to the current world energy consumption. Although this appears readily feasible, the mere 1% translates to a very large absolute area of ~10 5 square kilometers. 5

The use of OPVs as a primary energy resource is promising, but the technology is still in its infancy and will require much more research and development before possibly realizing this long-term endeavour. However, in the short term, the technology can be used to fulfill specialized applications with flexibility and lightweight appeal that are not feasible through traditional PVs. For example, tree-shaped architectures called "Solar Trees," were built and showcased to power light sources by the German Pavilion at the 2015 World Exhibition in Milan. 25 Before further development into OPVs can be made, it is necessary to give a brief overview on the operating principles, device architectures, methods of fabrication, evaluation of performance, and optoelectronic effects on performance.

1.4 Operating Principles of OPVs The conversion of incident light into electrical energy by an OPV cell is described by the following sequence of events (Figure 1.2 & 1.3): 1) photon absorption and exciton generation; 2) exciton diffusion; 3) exciton dissociation; 4) charge carrier collection. Photon absorption (step 1) by an organic semiconductor is achieved if the energy of the incident photons is equal to the energy band gap (E g) of the material, defined by the energy difference between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO). Collectively, the HOMO and LUMO is called the frontier molecular orbital (FMO). When this

condition is met, an electron is promoted from the HOMO to the LUMO leaving behind a hole, the absence of an electron, in the HOMO. Although the two charge carriers are situated in different molecular orbitals, they are still Coulombically-bound and the pair is referred to as an exciton. The exciton migrates (step 2) by diffusion towards an interface such as the donor- acceptor interface in a heterojunction solar cell or the electrode interface in a Schottky solar cell (more details on device architecture to follow). An exciton is an excited state and therefore, has a finite and short lifetime on the order of nanoseconds whereby it diffuses a distance of ~10 nm. If the exciton does not reach an interface, it decays in a process known as recombination whereby the electron re-unites with its hole partner at the HOMO level. If the exciton does reach an interface, dissociation (step 3) occurs by means of charge transfer in an energetically favourable process. A single charge carrier type ( e.g. electron) is transferred at a given interface such that an electron flows “downhill” via the LUMO state towards the cathode while a hole flows “uphill” via the HOMO state towards the anode. Following dissociation, the unbound charge carriers are separated (free) and transported towards their respective electrodes (step 4). This flow of charge carriers in the opposite direction is induced by the potential difference across the terminal electrodes. Once collected, the generated current is subsequently used to power a load ( e.g. light bulb) in the external circuit via the relaxation/recombination of the LUMO-situated excited-state electron with the HOMO-situated hole. 26,27

A major limitation responsible for the lower efficiencies of OPVs versus inorganic PVs lies in the exciton dissociation process (step 3). The dissociation of an exciton by charge transfer at an interface is more difficult due to strong Coulombic attraction between the opposite charges caused by a weaker charge screening effect ( i.e. lower dielectric constant) inherent within organic materials. In contrast, very weakly bound charge carriers are formed within inorganic semiconductors due to a stronger charge screening effect ( i.e. higher dielectric constant) and for this reason, these charge carriers are referred to as being "free" in this medium. 21,26

Figure 1.2 . Operating mechanism for Schottky cells. (1) Photon absorption by the organic layer to form an exciton. (2) Exciton diffusion to an organic-electrode interface. (3) Exciton dissociation at the interface. (4) Charge transport and collection at the respective electrodes. Exciton diffusion towards the cathode is shown in step (2), but random processes also occurs towards the anode, whereby holes are transferred at the organic-anode interface.

Figure 1.3 . Operating mechanism for donor-acceptor (D-A) heterojunction OPV cells. (1) Photon absorption by the donor layer to form an exciton. (2) Exciton diffusion to the D-A interface. (3) Exciton dissociation at the interface. (4) Charge transport and collection at the respective electrodes. Exciton formation is shown on the donor side, but a similar mechanism applies on the acceptor side whereby holes are transferred at the interface to the donor material.

1.5 Device Architectures The simplest OPV is called the Schottky junction solar cell and consists of a single organic semiconductor sandwiched between a pair of electrodes (Figure 1.2 & 1.4). Cells of this type

suffer from low power outputs due to poor exciton dissociation efficiency at the organic- electrode interface, where the electric field generated across the terminal electrodes does not break the excitons efficiently. This results in a greater tendency of the excitons to recombine rather than dissociate. In this device architecture, the organic semiconductor has a dual functionality, behaving as both an electron donor, also known as a hole transporter or p-type material, at the anode interface and an electron acceptor, also known as an electron transporter or n-type material, at the cathode interface. 26

Cathode Acceptor Donor

Cathode Cathode Interlayer e- Cathode Acceptor Acceptor Acceptor Donor/Acceptor Donor Donor Donor Anode Anode Anode Anode Glass Substrate Glass Substrate Glass Substrate Glass Substrate

Figure 1.4 . Left to right : Device architecture of a Schottky cell, planar heterojunction (PHJ), bulk heterojunction (BHJ), and a tandem cell.

The incorporation of a second organic semiconductor such that one behaves as the donor and the other as the acceptor addresses the exciton dissociation issue observed in the Schottky cell (Figure 1.3 & 1.4). 28 The electric field generated at the donor-acceptor interface aids to split the exciton ( i.e. stronger driving force) and allow the charge carriers to be extracted and collected. Moreover, with a second active material more photons can be absorbed, leading to more exciton generation. The end result is that more charge carriers are extracted ( i.e. higher current) to do more work. Solar cell of this nature falls under one of two main architecture types: planar heterojunction (PHJ) and bulk heterojunction (BHJ) solar cell. In PHJ cells, the two materials are deposited in sequential layers creating a single planar interface. In BHJ cells, the two materials are codeposited as blends or mixtures such that both materials penetrate one another to create spatially distributed interfaces. The PHJ cells benefit from having a continuous, ordered pathway for charge carriers to be transported to their respective electrodes with minimal unwanted charge

recombination events while BHJ cells benefit from an increase in the donor-acceptor interfacial surface area and thus, enhanced exciton dissociation. 21,26

The PHJ and BHJ cells are the most commonly studied architectures as these devices have a nice balance of performance efficiencies and ease of fabrication. More complex architectural variants using additional photoactive, buffer, blocking, interfacial, and/or transport layers with higher power outputs are known, but they are more energy-intensive to produce. For instance, tandem solar cells benefit from a significant increase in current generation by making use of multiple cells connected in series, whereby each cell absorbs a different portion of the solar spectrum (Figure 1.4). With each additional cell, costs in the form of material use and labour are increased.21

1.6 Methods of Fabrication OPV devices are fabricated through one of two methods: vacuum deposition or solution processing. Vacuum deposition (Figure 1.5) involves heating a sample under vacuum to a temperature whereby it evaporates (solid ‰gas transition). The evaporated sample then condenses (gas ‰solid transition) onto a cooler surface of a solid substrate. This technique affords either PHJ or BHJ cells, whereby the latter is attainable via the simultaneous co- deposition of two organic materials. Multi-layer architectures are also attainable due to the solvent-free nature of the process, resulting in no or limited intermixing between the adjacent layers. To be compatible for this process, a material must be sublimable. As such, high molecular weight compounds such as polymers are not suited for this and will degrade under extreme heat; this technique is feasible with small molecules and low molecular weight oligomers/polymers. 27,29

Solution processing (Figure 1.5) involves applying a solution, made from the dissolution of a sample in an organic solvent or solvents, onto a substrate followed by the removal of the solvent(s). As a consequence of being a solvent-based technique, PHJ cells are not attainable due to intermixing of layers and thus, only BHJ cells are afforded. 30 However, this limitation is the subject of current/future investigation through a method called orthogonal processing. 31 To be

compatible for solution processing, a material must be soluble in an organic solvent such as toluene or chlorobenzene. This non-stringent condition permits both small molecules and polymers to be feasible with this simple and cheaper technique. 30

Substrate Substrate holder Deposited sample

Vacuum chamber

Sample holder (source) Heater (a)

Sample

Substrate Spin Spin

Substrate holder

(b)

Figure 1.5 . An example of a (a) vacuum deposition and (b) solution processing process.

1.7 Device Performance Definition The performance of an OPV cell under illumination is defined by its power conversion efficiency (PCE):

· · =

where J SC is the short-circuit current density, V OC is the open-circuit voltage, FF is the fill factor, and P in is the input power (i.e. solar power). The J SC is the current produced when the terminals (i.e. electrodes) of the circuit are directly connected and represents the maximum current flow that can be attained. This value is extracted from the y-intercept of a J-V curve (Figure 1.6) and is dependent on the efficiency of the four steps, from the initial photon absorption to the final charge collection step, described in the operating principles of OPVs. The V OC is the electrical potential difference that is developed across the terminals when the circuit are disconnected and represents the maximum voltage that can be produced. This value is also extracted from a J-V curve, but from the x-intercept and is largely dependent on the electronic nature of the materials under study. The FF is the ratio of the maximum power output from a cell to the product of the

JSC and VOC :

· = ·

where J max and V max are the corresponding current and voltage point for the cell's maximum power output. Graphically, the J max and V max is the point on the curve where the area of the enclosed rectangle is the greatest. From the J-V plot, the FF is a measure of the "squareness" of the curve in which a value of 1 is a perfect square. The FF is indicative of the degree of resistance in the circuit, specifically the balance of charge collection (desired) to charge recombination (undesired). Beyond this, it is unclear what factors influence the FF. The operating regime of an OPV is at any voltage between zero and the maximum attainable voltage, the V OC . At these extremes, power is not generated since power is the product of voltage and current and at these points, either the voltage (V = 0 V, J = J SC ) or current output (V = V OC , J =

0) is zero. Despite this, both the J SC and V OC are informative metrics that represent the maximum current and voltage output of a device. 32-34

It is necessary to emphasize that the performance of a cell is not entirely dependent on the materials’ properties. There are other factors to consider such as the deposition method, thickness of each deposited layer (for PHJ cells), and blending ratios of donor and acceptor

materials (for BHJ cells). Optimization in the fabrication conditions alone is a widely studied area of research. 35-37 Moreover, performance is also affected by the stability of the device against heat (thermal), water (hydrolytic), air (oxidative), and light (photo). 22 These considerations are also key, in addition to the development of low-cost, high-performing materials, in the promising future of OPV technologies.

Figure 1.6 . A typical current density vs. voltage (J-V) curve of an OPV cell under illumination.

The V OC , J SC , J max , V max , and FF are shown. The inner dashed rectangular region represents the maximum power output from the cell. Data obtained from Bender et al .38

1.8 Optoelectronic Effects on Performance The performance of an OPV device is largely influenced by the intrinsic nature of the organic photoactive materials. These properties include, but are not limited to, the molar absorptivity, energy band gap (E g), frontier molecular orbital (FMO) energy level alignment, electrochemical behaviour, charge carrier mobility, film morphology, and photoluminescence quantum yield. 22 Due to the numerous factors that contribute to the performance of a device, it is a very challenging task to design and develop new and improved organic photoactive materials for OPV applications. However, this is a necessary mission in research and development for unlocking

higher potentials. In consideration of this importance, a brief overview of some of the inherent properties, particularly the optical and electronic properties, will be given.

Arguably one of the most important traits of a photoactive material is the ability to capture light and to do so with high intensity, given that the aim of an OPV cell is to convert light into electricity. This property is defined by a compound's molar extinction coefficient or molar absorptivity (ε), whereby the higher the value of ε the greater the amount of incident light that is absorbed at a given wavelength. 21 For instance, chloro boron subphthalocyanine (Cl-BsubPc) has a high molar absorptivity of 8.86x10 4 L·mol -1·cm -1 (at 565 nm) 39 while common organic solvents like toluene, methanol, and acetonitrile have very low molar absorptivities in the 10 -3 L·mol - 1·cm -1 range 40 . An accompanying benefit with using compounds with high extinction coefficients is the materials-saving component from using less of the material in the fabrication process. 6

Not only is it key for a compound to absorb light strongly, it must do so within the spectral region where the flux of solar energy is the highest and must also do so in a broad manner ( i.e. good spectral coverage). 21 This promotes more photons to be absorbed, leading to more excitons to be generated and therefore, more current to be extracted. Figure 1.7 shows the solar irradiance spectrum at sea level, showing that the irradiance or power density is particularly the highest between 450 and 700 nm. This range falls within the visible region (400-700 nm) of the electromagnetic radiation spectrum and combined with the inherent trait of coloured compounds to absorb color/energy of visible light, makes them desirable subjects for light harvesting purposes. This specific feature drove the early work on natural and synthetic dyes such as magnesium phthalocyanine, 41 chlorophyll-a, 42 hydroxy squarylium, 43 and crystal violet 44 .

The photovoltaic process is initiated when the energy of the incident light is equal to the E g of the 27 photoactive material. Thus, a compound with an E g in the 450-700 nm window (i.e. 1.77-2.76 eV) would fall within the region of highest solar irradiance, leading to a higher current output. A compound's absorption spectrum (absorption vs. wavelength) shows what portion of the solar spectrum is covered and also enables the E g to be determined from the onset of the absorption band at the longest wavelength ( i.e. lowest energy transition). This is exemplified in Figure 1.7

using Cl-BsubPc, which shows the compound's absorption band within the 450-700 nm region.

Organic compounds with suitable E gs are characterized by a highly conjugated or delocalized π- electron system. Extending the conjugation length as seen going from benzene to pentacene causes the E g to decrease, permitting excitons to be generated with less difficulty as incident photons of lower energy can induce the HOMO-to-LUMO electronic transition. 27,45

Figure 1.7 . Solar irradiance spectrum (black, sea level) overlaid with the absorption spectrum (pink) of Cl-BsubPc. The red shaded area represents the area of highest solar irradiance (450-700 nm).

Related to the E g, the alignment of the FMO energy levels plays a critical role in the flow of charge carriers to the respective electrodes. It is energetically favourable for electrons to flow "downhill" via the LUMO state towards the cathode and for holes to flow "uphill" via the HOMO state towards the anode. 26,27 As such, careful planning must go into selecting and designing photoactive materials with the appropriate energy levels ( i.e. offset energy levels) to facilitate the directionally-specific charge carrier movement for photocurrent generation. As an example in Figure 1.3, the HOMO and LUMO levels of the donor are both higher lying than the corresponding HOMO and LUMO levels of the acceptor to ensure efficient, exergonic charge

transfer between the two materials. The magnitude of the driving force in the biased flow of 27 charges is governed by the built-in potential or V OC across the OPV device. Although the V OC is fundamentally defined as the potential difference across the terminals, it is closely linked to the difference between the HOMO level of the donor and the LUMO level of the acceptor in a heterojunction cell. The former is related to the donor's ionization energy, the energy required to remove an electron, while the latter is related to the acceptor's electron affinity, the energy 46,47 gained from the receipt of an electron. From this, enhancement in the V OC is possible via tuning of the donor's HOMO level and acceptor's LUMO level. The FMO energy levels are commonly determined using cyclic voltammetry (CV), an electrochemical technique that measures potential for each oxidation and reduction process. 48 This is a practical and sensible technique as the reduction-oxidation (redox) state of the photoactive materials is changed upon an exciton dissociation reaction. For example in a heterojunction cell, the donor becomes oxidized while the acceptor becomes reduced following a charge transfer reaction.

Tuning these optoelectronic properties have been the engine for driving the design and development of new materials for OPV applications. Recall that the power output from a cell is dependent on the J SC and V OC . From an optical perspective, attempts to enhance the J SC s can be carried out using materials with high ɛs and with good absorption coverage of the solar spectrum (i.e. broad absorption in the visible region). From an electronic perspective, attempts to enhance

JSC s and V OC s can be done using materials with low E gs and with the appropriate position of the FMO energy levels.

1.9 Commonly Employed Organic Photoactive Materials Numerous organic compounds have been explored as photoactive materials in OPVs. Some of the most commonly studied compounds, as evident by the quantity of manuscripts in the scientific literature, are those based on thiophene, carbazole, fluorene, pyrrole, phenylene vinylene, porphyrin, perylene, pentacene, fullerene, and phthalocyanine (Table 1.1, Entry 1-10). These publication numbers encompass both small molecule and polymeric derivatives and it is worth noting that the latter type dominates the space of photoactive materials in OPVs as illustrated by the volume of papers published per year (Figure 1.8).

Table 1.1 . Number of publications for each of the listed organic photoactive material. Entry Class of Organic Photoactive Materials Number of Publications a 1 5991

Thiophene 2 795

Carbazole 3 505

Fluorene 4 511

Pyrrole 5 1013

Phenylene Vinylene 6 852

Porphyrin 7 769

Perylene 8 505

Pentacene 9 6395

Fullerene 10 1552

Phthalocyanine (Pc)

11 145

Boron Subphthalocyanine (BsubPc) 12 23

Naphthalocyanine (Nc) 13 11

Boron Subnaphthalocyanine (BsubNc) 14 27

Tetrabenzoporphyrin a Search strings used via Web of Science (Accessed on December 28, 2015): ‘TS=((organic solar cell*) OR (organic photovoltaic*)) AND TS=((X 1) OR (X 2) OR ...(X n))’, where X n = molecule’s generic names, derivatives’ names, and common acronyms. For the fullerene example, X 1 = fullerene, X 2 = C60, X 3 = C70, X 4 = PCBM, X 5 = PC60BM, X 6 = PC61BM, X 7 = PC70BM, X 8

= PC71BM, X 9 = Buckminster*.

Figure 1.8 . Number of publications on polymer-based ( red ) and small-molecule-based ( blue ) materials for OPVs since the year 2000. The following search strings were used via Web of Science (Accessed on December 28, 2015): ‘TS=((organic solar cell*) OR (organic photovoltaic*)) AND TS=(polymer)’ for polymer-based OPVs and ‘TS=((organic solar cell*) OR (organic photovoltaic*)) AND TS=((small molecul*) OR (small-molecul*))’ for small- molecule-based OPVs.

1.10 Isoindole-Based Compounds in OPVs Within the domain of organic photoactive materials, isoindole-based compounds (Table 1.1, Entry 10-14) are showing signs of promise for OPV applications. This is strongly exemplified by the phthalocyanines (Pcs, Figure 1.9), synthetic analogues that chemically resemble the naturally-occurring porphyrins. Pcs are a class of highly conjugated (18 π-electron system), aromatic macrocycles composed of four isoindoline units bridged by sp 2 hybridized nitrogen atoms. They have versatile metal-binding properties as determined by their ability to coordinate

with nearly all metal elements of the periodic table to form metallophthalocyanines. Pcs are highly derivatizable, whereby substituents can be introduced at the axial position(s) of the metal center and at the periphery (Figure 1.9). As a result of this flexibility, their physical properties can be highly tailored. 49 For example, Pcs can be synthesized with improved water-50 or organic- solubility, 51 bathochromic or hypsochromic shifted absorption spectra, 52 broader light absorption 53 54 properties, and tunable E gs . Pcs have desirable optical and electronic properties for OPVs such as being an intensely green-blue coloured dye with a strong absorption band in the red 4 5 -1 -1 55 region (ɛ ~10 -10 L·mol ·cm , ~700 nm) of the visible spectrum, low E g, and p-type conductivity. 49 These properties have permitted them to be successfully integrated into OPV devices, where PCEs up to nearly 6% 56 and beyond 7% 57 in a single-junction and tandem cell have been achieved, respectively. In addition to OPVs, the employment of Pcs has also extended to other organic electronic (OE) technologies like organic light emitting diodes (OLEDs) 58-60 and organic field effect transistors (OFETs) 61-63 .

Figure 1.9 . Chemical structure of a generic metallophthalocyanine and boron subphthalocyanine showing the axial (R ax ) and peripheral (R p) positions. The highlighted region represents a single isoindole unit.

Other than the Pcs, other isoindole-based compounds are less established and known (Table 1.1, Entry 11-14). However, to further illustrate the potential merits of isoindoles for OPVs, another case example will be discussed. Related to the Pcs, the subphthalocyanines (subPcs) are lower

homologues made of three diiminoisoindoline units as opposed to four (Figure 1.9). Unlike the versatile metal-binding properties of Pcs, the subPcs can only complex with boron. For this reason, subphthalocyanines (subPcs) and boron subphthalocyanines (BsubPcs) are used interchangeably in the literature. In contrast to the planar nature of Pcs, BsubPcs are uniquely bowl-shaped despite their highly conjugated, aromatic 14 π-electron system. 64 This bowl-shape structure is not an energetically favourable conformation and requires the templating action of boron for its formation; 65 it is not possible to synthesize a metal/boron-free subPc. However, like the Pcs, BsubPcs can be chemically functionalized at the axial and/or peripheral position, allowing for their physiochemical properties to be modified (Figure 1.9). 64 For example, BsubPcs can be prepared with high organic-solubility, 66 bipolar electrochemical stability, 67 better 68 69 hydrolytic stability, and lower E gs . Derivatizations at the axial position (R ax ) have minor effects on the optical and electronic (HOMO, LUMO, E g) behaviours while the effects are 70 significant at the periphery (R p). This differential effect is a powerful tool to independently modify the optoelectronics or other non-related properties such as solubility and solid state packing arrangement. BsubPcs display optical and electronic properties that are suitable for OPVs like having a strong absorption in the visible light region as observed by their high ɛs (~10 4 L·mol -1·cm -1, 560-600 nm), 39,71,72 absorption spectra that aligns closely with the maximum 73-75 74-76 solar irradiance (Figure 1.7), suitable E gs, and n-type conductivity . Contrary to their CV- measured n-type behaviours, BsubPcs exhibit bipolar characteristics as demonstrated by their successful integration as a donor ( p-type material) 77-79 and as an acceptor ( n-type material) 80-82 in heterojunction devices. With their use, PCEs as high as 8.40% have been attained. 83 In comparison to Pcs, BsubPcs have larger Egs and deeper HOMO levels (i.e. lower in energy), producing devices with lower J SC s and higher V OC s when utilized as a donor material, respectively. 77,84 Outside of OPVs, BsubPcs have also found purpose in OLEDs, 75,85,86 OFETs, 87,88 and non-linear optics (NLO) 89 .

Within the family of BsubPcs, Cl-BsubPc is the most widely studied material for OPV applications. It is also the first known BsubPc derivative and was discovered by Meller and Ossko in 1972, via the cyclotrimerization reaction of phthalonitrile with boron trichloride with 40% yield (Scheme 1.1). 90 In the 1990s, it was shown that Cl-BsubPc could be used as an

intermediate for the synthesis of unsymmetric Pcs via a reaction with a substituted 1,3- diiminoisoindoline (Scheme 1.1). 91,92 This ring expansion reaction was originally discovered for the analogous tri-tert -butylated Br-BsubPc compound by Kobayashi et al .93 Also within this decade, Wӧhrle et al . developed a more efficient method (64% yield) for synthesizing Cl-BsubPc by using a 1.0 M solution of boron trichloride in n-hexane as the boron reagent. This approach also avoided the use of hazardous gaseous boron reagents, making the process more practical. 94 In the 2000s, both Kennedy et al .95 and Torres et al .96 showed that 1.0 M solution of boron trichloride in n-heptane and in p-xylene were also feasible, producing Cl-BsubPc in 51% and 82% yield, respectively.

Scheme 1.1 . Synthesis of Cl-BsubPc (black), Cl-BsubNc (red), unsymmetric Pc (black), and unsymmetric Nc (red).

In the Bender group, our research interest is centered on the design, synthesis, and engineering of functional materials like the BsubPcs for OE applications. In our earliest work, we adapted and optimized the process of Kennedy et al .95 for the multi-gram scale synthesis of Cl-BsubPc with yield of 63%. 72 This reproducibly provided us ample amounts of the material to enable the generation of new derivatives of BsubPcs with modified physical properties and solid state arrangement via the introduction of new functional groups at the axial position as previously described. Examples include phenoxy, 66,70,75,97,98 sulfonyl, 68 phthalimido, 67 phenyl, 99 and carboxyl 100 groups at the axial position of BsubPc. During this pursuit of other derivatives beyond Cl-BsubPc, our group 67 as well as other groups 14,101-103 have identified the formation of

an oxygen-bridged dimer of BsubPc, μ-oxo-(BsubPc) 2 (Figure 1.10), as a minor product of BsubPc reactions. This dimer was first reported in the literature in 1996, and was shown to have unique photophysical properties within the family of BsubPcs such as having a significantly blue-shifted Q absorption band. 71 Since the 20 years following its discovery, the chemistry behind its synthesis remains largely undeveloped. As will be described in Chapter 3, our group has surveyed a number of different synthetic routes and developed a viable process for producing gram scale quantities of μ-oxo-(BsubPc) 2 with sufficiently high purity. This enabled its physical properties to be characterized such as its absorption, photoluminescence, and electrochemical properties as well as its performance characteristics in OPV devices, all of which other than the absorption properties were unknown in the literature.

Figure 1.10 . Chemical structure of μ-oxo-(BsubPc) 2.

In 2012, our group published a review article describing the synthesis, structural features, photo- and electro-physical properties, and OPV device performance characteristics of BsubPcs with the bulk of the focus on Cl-BsubPc. 64 A follow up review is currently in progress to encompass new findings within the BsubPc space since the 2012. This will also include a review on the boron subnaphthalocyanines (BsubNcs, Table 1.1, Entry 13), a class of compounds that are related to the BsubPcs but are much less recognized. The BsubNcs are non-planar, cone-shaped, and aromatic in structure just like the BsubPcs. They have a more extended π-conjugation system than the BsubPcs, causing their Q absorption band to be heavily red-shifted by about 100 nm to the 650-700 nm range. 55,95,104 Their ability to capture light within this region, combined with

their high extinction coefficients 95,104 that are comparable with those of the BsubPcs, has made them appealing targets for OPVs. 80,83,105-111

The synthesis of BsubNcs is highly analogous to that of the BsubPcs, whereby 2,3- dicyanonaphthalene or its substituted derivative is treated with a boron source and template agent such as boron trichloride or boron tribromide (Scheme 1.1). Also like the BsubPcs, the BsubNcs can undergo a ring expansion reaction to form unsymmetric naphthalocyanines (Nc, Scheme 1.1). 112 The first report on the synthesis of a BsubNc was made by Hanack and Rauschnabel in

1995 (Table 1.2, Entry 1), where they treated 2,3-dicyanonaphthalene with PhBCl 2 in boiling naphthalene to produce Ph-BsubNc in only an analytical amount. They also made a more soluble tri-tert -butylated Ph-BsubNc derivative by using 6-tert -butyl-2,3-dicyanonaphthalene, which was also formed in analytical amounts. 113 This synthesis was reported again in the following year in a review article by Torres, Hanack, and coworkers. 71 The very low yields were significantly improved by Kobayashi et al ., who used BBr 3 as the boron source to synthesize Br-BsubNc in a 34.6% yield (Table 1.2, Entry 2). 55 Immediately following this work, both Kennedy et al. 95 and 104 Torres et al . reported on the synthesis of Cl-BsubNc in 53% and 35%, respectively, using BCl 3 as the boron template (Table 1.2, Entry 3 & 4). The yield was further enhanced by Giribabu and coworkers, who employed microwave irradiation on a reaction mixture of 6-tert -butyl-2,3- dicyanonaphthalene, boron trichloride, and 1-chloronaphthalene to afford tris-tert -butylated Cl- BsubNc in 82% yield (Table 1.2, Entry 5). 114 In 2014, Takao, Ohno, and coworkers made two peripherally fluorinated Cl-BsubNc derivatives, Cl-F6BsubNc (26% yield) and Cl-F12 BsubNc (44% yield), from their respective fluorinated 2,3-dicyanonaphthalene (Table 1.2, Entry 6). They also treated the two products with AgBF 4 to substitute the axial chloride with a fluoride via an axial ligand exchange reaction. 115 Very recently, Mizutani et al . made two other peripherally halogenated Cl-BsubNc derivatives (Table 1.2, Entry 7), Cl-Cl 6BsubNc (22% yield) and Cl- 116 I6BsubNc (73% yield), and also carried out the axial chloride-to-chloride transformation.

Although the synthesis of BsubNc have improved with the employment of BCl 3 or BBr 3 as the boron source and template agent, 55,95,104,114-116 the use of boron trihalides has been reported to cause halogenation at the periphery of the BsubPc/BsubNc scaffold although no evidence have

been provided to support this. 117,118 Nevertheless, this outcome obviously affects the purity of the product and makes the isolation of the peripherally-unsubstituted, non-halogenated BsubNc very challenging or entirely unfeasible. This effect has been shown to be reduced, but not completely eliminated, with the use of an electron-rich solvent or cosolvent such as 1-methylnaphthalene 117 or 2,3-dimethyl-6-tert -butylnaphthalene 55 to scavenge halogenating reactants produced in situ .

Table 1.2 . Summary of reported BsubNc synthesis. Entry Starting Material Boron Solvent Temperature (°C) Yield (%) Reference Source 1 PhBCl 2 NAP 218 Anal. 71,113

2 BBr 3 2,3-DM-6- 180 34.6 55 tBuNAP

3 BCl 3 1,2-DCB 180 53 95

4 BCl 3 ClBZH: 130 35 104 TOL (1:1)

5 BCl 3 1-ClNAP microwave 82 114

6 BCl 3 p-XYL: 150-180 Cl-F6BsubNc: 26 115 1,2-DCB Cl-F12 BsubNc: 44 (1:1)

7 BCl 3 p-XYL: Cl-Cl 6BsubNc: 160 Cl-Cl 6BsubNc: 22 116 1,2-DCB Cl-I6BsubNc: 180 Cl-I6BsubNc: 73 (1:1)

*NAP = naphthalene; 2,3-DM-6-tBuNAP = 2,3-dimethyl-6-tert -butylnaphthalene; 1,2-DCB = 1,2-dichlorobenzene; ClBZH = chlorobenzene; TOL = toluene; 1-ClNAP = 1-chloronaphthalene; p-XYL = para -xylene.

A review of the papers outlined in Table 1.2 reveals that characterizations for the BsubNcs produced is limited or entirely absent (in the case of Entry 5). 1H NMR data has been numerically reported for BsubNcs in Entry 1, 3, 4, 6, and 7 with spectra published only for the latter two entries to graphically illustrate the purity of the sample. Moreover, high resolution mass spectrometry (HRMS) data were only reported for the BsubNc compounds in Entry 6 and 7; low resolution mass spectrometry data was detailed for the Cl-BsubNc compounds made via Entry 3 and 4. References from Entry 2 and 4 have reported on the use of elemental analysis (EA), however, EA data was only supplied by the former entry and the values were not within an acceptable range ( i.e. +/- 0.4% for most journals) from the calculated values for the respective proposed molecular formula (C 36 H18 N6BBr). Overall, adequate characterizations are only reported in Entry 6 and 7. In the shadow of the known peripheral halogenation outlined above, this lack of analytical data calls into question the ultimate purity of the reported BsubNc samples. As will be described in great details in Chapter 5, we have presented evidence to prove the occurrence of this halogenation side reaction, which was found exclusive only to the bay position of the BsubNc structure. We also described our unsuccessful efforts to separate and isolate the halogenated Cl-BsubNc species, but provided sufficient characterizations of the product mixtures, and tested their performance within OPV devices.

At the present moment, there are a total of nine papers that have incorporated a BsubNc into OPV devices with the first report made in 2009. 80,83,105-111 It is worthwhile to highlight that Cnops et al . reported a BsubNc-based device with a PCE of 8.4%, a value that is among the highest for a PHJ OPV cell. 83 Within the nine investigations, only the Cl-BsubNc compound has been studied; no other BsubNc derivative has been explored. In Chapter 6, we will describe the synthesis of a set of phenoxy-substituted BsubNc compounds and their integration into PHJ OPV devices, marking the first time a non-Cl-BsubNc compound has been studied for OPV applications.

The adaptability of organic semiconductors among various OE technologies is not uncommon. For example, Pcs 56,58,61 and BsubPcs 75,83,87 have found applications in OPVs, OLEDs, and OFETs. Given this and the fact that other isoindoles such as naphthalocyanines, 119

tetrabenzoporphyrins, 120,121 indolizino[3,4,5-ab ]isoindoles, 122 and 1,3,4,5,6,7-hexaphenyl-2-{3’- (9-ethylcarbazolyl)}-isoindole 123 have also found application in OEs, further augments the potential worth of isoindole-based compounds.

1.11 Thesis Statement The aim of this thesis is to strengthen the fundamental understanding of the structure-property- performance relationships of isoindole-based photoactive materials for OPVs. To make this advancement, five classes of novel isoindole-based compounds are studied in this dissertation and they are the: 1) BsubPc polymers (Chapter 2);

2) Oxygen-bridged dimers of BsubPc (μ-oxo-(BsubPc) 2) (Chapter 3 & 4); 3) Boron subnaphthalocyanines (BsubNcs) (Chapter 5 & 6); 4) Group XIII complexes of 1,3-bis(2-pyridylimino)isoindoline (BPI) (Chapter 7); and 5) Boron tribenzosubporphyrins (BsubPys) (Chapter 8). The BsubPc polymers and group XIII complexes of BPI were unprecedented in the literature while the μ-oxo-(BsubPc) 2 dimers, BsubNcs, and BsubPys were known but were unexplored or underexplored in OPV devices. As a consequence of their undeveloped nature, the syntheses of these materials were found to be challenging. To emphasize again, the space of known isoindole- based compounds for OPVs is a small one. Thus, this inspired the development of my thesis statement:

Can the synthetic difficulties of certain isoindole-based compounds be overcome using chemical and/or engineering methodologies so as to assess their potential as functional materials within organic photovoltaic devices?

In our laboratory/research group, the general course of action that is taken in studying any new material is to first synthesize and purify the compound(s) (Figure 1.10). This is followed by an assessment of its basic photo- and electro-physical properties. If these properties are not promising for OPV use, then its continual investigation comes to a halt. If they are found suitable and promising for OPVs, then its chemical process is optimized in an effort to obtain sufficient

amounts of the compound(s) (ideally >250 mg) for the next step. Lastly, the compound(s) is integrated within OPV devices and its performance is evaluated.

The progress/advancement of each research stream is mapped using circles in Figure 1.10 with the size of the circle indicating the amount of effort spent for that specific step. For example, a considerate amount of effort was put into the chemical process optimization of BsubNcs. For the BsubPc polymers and group XIII complexes of BPI, advancements were not made beyond the assessment of their basic properties. The other three streams saw progress to the final step of OPV device integration and performance assessment. A process engineering and film characterization step were incorporated for the μ-oxo-(BsubPc) 2 and BsubPys to bypass the major issues with their chemical process optimization step. The future direction for the BsubPc polymers, μ-oxo-(BsubPc) 2s, and BsubPys are also displayed (as dashed lines) in Figure 1.10. The full details in the advancement of the five classes of isoindole-based materials are described in the following eight chapters.

BsubPc Polymers BsubPc Dimers BsubNcs BPI Complexes BsubPys

N N N N B N

R = H Rn = F 12 , X = OPh, X = Cl Cl 6, F5Ph Cl 12 Initial Synthesis

Assessment of Basic Properties

Chemical Process X Optimization

Process Engineering

Film Characterization

Assessment of OPV Device Performance Accomplished Future Work

Figure 1.11. The five classes of isoindole-based compounds studied in this dissertation showing their progress and future direction. The isoindole subunit is emphasized in blue for each class of compounds.

1.12 References 1. Freeman, S. In Biological Science ; 2nd ed.; Pearson Prentice Hall: Upper Saddle River, NJ, 2005. 2. U.S. Energy Information Administration. International Energy Statistics. https://www.eia.gov/cfapps/ipdbproject/IEDIndex3.cfm?tid=44&pid=44&aid=2 (accessed December 28, 2015). 3. Larcher, D.; Tarascon, J. M. Nature Chemistry 2015 , 7, 19-29. 4. Central Intelligence Agency. The World Factbook. http://www.cia.gov/library/publications/resources/the-world-factbook/ (accessed December 28, 2015). 5. Danny Harvey, L. D. Energy and the New Reality 2: Carbon-free Energy Supply ; Earthscan: London, 2010. 6. El Chaar, L.; Lamont, L. A.; El Zein, N. Renewable & Sustainable Energy Reviews 2011 , 15 , 2165-2175. 7. Lewis, N. S. MRS Bulletin 2007 , 32 , 808-820. 8. Becquerel, A. E. Comptes Rendus de l’Academie des Sciences 1839 , 9, 561-567. 9. Smith, W. Nature 1873 , 7, 303. 10. Volmer, M. Annalen der Physik 1913, 40 , 775-796. 11. Adams, W. G.; Day, R. E. Proceedings of the Royal Society of London 1876 , 25 , 113- 117. 12. Fritts, C. E. American Journal of Science 1883 , 26 , 465-472. 13. Grondahl, L. O. Reviews of Modern Physics 1933 , 5, 0141-0168. 14. Ohl, R. S. Light-sensitive electric device. U.S. Patent 2,402,662, June 25, 1946. 15. Chapin, D. M.; Fuller, C. S.; Pearson, G. L. Journal of Applied Physics 1954 , 25 , 676- 677. 16. Mandelkorn, J.; McAfee, C.; Kesperis, J.; Schwartz, L.; Pharo, W. Journal of the Electrochemical Society 1962 , 109 , 313-318. 17. Green, M. A. Semiconductor Science and Technology 1993 , 8, 1-12. 18. Dimroth, F. New world record for solar efficiency at 46%. https://www.ise.fraunhofer.de/en/press-and-media/press-releases/press-releases-

2014/new-world-record-for-solar-cell-efficiency-at-46-percent (accessed December 30, 2015). 19. Conibeer, G. Materials Today 2007 , 10 , 42-50. 20. Pagliaro, M.; Palmisano, G.; Ciriminna, R. In Flexible Solar Cells ; Wiley-VCH Verlag GmbH & Co. KGaA: 2008, p 1-30. 21. Pagliaro, M.; Palmisano, G.; Ciriminna, R. In Flexible Solar Cells ; Wiley-VCH Verlag GmbH & Co. KGaA: 2008, p 85-106. 22. Hains, A. W.; Liang, Z. Q.; Woodhouse, M. A.; Gregg, B. A. Chemical Reviews 2010 , 110 , 6689-6735. 23. Prof. Yan's Research Group. Best Research Organic Solar Cell Efficiency. http:// http://blog.ust.hk/yanlab/ (accessed January 8, 2016). 24. Slade, A.; Garboushian, V. In Technical Digest, 15th International Photovoltaic Science and Engineering Conference Shanghai, 2005. 25. The Solar Trees. http://schmidhuber.de/en/project/german-pavilion-expo-milano-2015-0 (accessed December 20, 2015). 26. Spanggaard, H.; Krebs, F. C. Solar Energy Materials and Solar Cells 2004 , 83 , 125-146. 27. Gunes, S.; Neugebauer, H.; Sariciftci, N. S. Chemical Reviews 2007 , 107 , 1324-1338. 28. Tang, C. W. Applied Physics Letters 1986 , 48 , 183-185. 29. Hoppe, H.; Sariciftci, N. S. Journal of Materials Research 2004 , 19 , 1924-1945. 30. Krebs, F. C. Solar Energy Materials and Solar Cells 2009 , 93 , 394-412. 31. Aguirre, J. C.; Hawks, S. A.; Ferreira, A. S.; Yee, P.; Subramaniyan, S.; Jenekhe, S. A.; Tolbert, S. H.; Schwartz, B. J. Advanced Energy Materials 2015 , 5, 1402020. 32. Pagliaro, M.; Palmisano, G.; Ciriminna, R. In Flexible Solar Cells ; Wiley-VCH Verlag GmbH & Co. KGaA: 2008, p 31-54. 33. Potscavage, W. J.; Sharma, A.; Kippelen, B. Accounts of Chemical Research 2009 , 42 , 1758-1767. 34. Mishra, A.; Bauerle, P. Angewandte Chemie-International Edition 2012 , 51 , 2020-2067. 35. Campoy-Quiles, M.; Ferenczi, T.; Agostinelli, T.; Etchegoin, P. G.; Kim, Y.; Anthopoulos, T. D.; Stavrinou, P. N.; Bradley, D. D. C.; Nelson, J. Nature Materials 2008 , 7, 158-164.

36. Zeng, L. C.; Tang, C. W.; Chen, S. H. Applied Physics Letters 2010 , 97 , 053305. 37. Karak, S.; Homnick, P. J.; Della Pelle, A. M.; Bae, Y.; Duzhko, V. V.; Liu, F.; Russell, T. P.; Lahti, P. M.; Thayumanavan, S. ACS Applied Materials & Interfaces 2014 , 6, 11376- 11384. 38. Josey, D. S.; Castrucci, J. S.; Dang, J. D.; Lessard, B. H.; Bender, T. P. ChemPhysChem 2015 , 16 , 1245-1250. 39. Fulford, M. V.; Jaidka, D.; Paton, A. S.; Morse, G. E.; Brisson, E. R. L.; Lough, A. J.; Bender, T. P. Journal of Chemical & Engineering Data 2012 , 57 , 2756-2765. 40. Kumar, A.; Singh, S.; Mudahar, G. S.; Thind, K. S. Radiation Physics and Chemistry 2006 , 75 , 737-740. 41. Ghosh, A. K.; Morel, D. L.; Feng, T.; Shaw, R. F.; Rowe, C. A. Journal of Applied Physics 1974 , 45 , 230-236. 42. Tang, C. W.; Albrecht, A. C. Nature 1975 , 254 , 507-509. 43. Merritt, V. Y.; Hovel, H. J. Applied Physics Letters 1976 , 29 , 414-415. 44. Williams, E. W. IEE Journal on Solid-State and Electron Devices 1977 , 1, 185-192. 45. Cheng, Y. J.; Yang, S. H.; Hsu, C. S. Chemical Reviews 2009 , 109 , 5868-5923. 46. Kietzke, T.; Egbe, D. A. M.; Horhold, H. H.; Neher, D. Macromolecules 2006 , 39 , 4018- 4022. 47. Rand, B. P.; Burk, D. P.; Forrest, S. R. Physical Review B 2007 , 75 , 115327. 48. D'Andrade, B. W.; Datta, S.; Forrest, S. R.; Djurovich, P.; Polikarpov, E.; Thompson, M. E. Organic Electronics 2005 , 6, 11-20. 49. de la Torre, G.; Claessens, C. G.; Torres, T. Chemical Communications 2007 , 2000-2015. 50. Schumann, S.; Hatton, R. A.; Jones, T. S. The Journal of Physical Chemistry C 2011 , 115 , 4916-4921. 51. Lessard, B. H.; Dang, J. D.; Grant, T. M.; Gao, D.; Seferos, D. S.; Bender, T. P. ACS Applied Materials & Interfaces 2014 , 6, 15040-15051. 52. Lessard, B. H.; Al-Amar, M.; Grant, T. M.; White, R.; Lu, Z. H.; Bender, T. P. Journal of Materials Chemistry A 2015 , 3, 5047-5053. 53. Ren, B. Y.; Zhu, L.; Cui, G. R.; Sun, Y. G.; Zhang, X. T.; Liang, F. S.; Liu, Y. C. Tetrahedron Letters 2013 , 54 , 5953-5955.

54. Raboui, H.; Al-Amar, M.; Abdelrahman, A. I.; Bender, T. P. RSC Advances 2015 , 5, 45731-45739. 55. Kobayashi, N.; Ishizaki, T.; Ishii, K.; Konami, H. Journal of the American Chemical Society 1999 , 121 , 9096-9110. 56. Fleetham, T. B.; Mudrick, J. P.; Cao, W.; Klimes, K.; Xue, J.; Li, J. ACS Applied Materials & Interfaces 2014 , 6, 7254-7259. 57. Raïssi, M.; Vignau, L.; Cloutet, E.; Ratier, B. Organic Electronics 2015 , 21 , 86-91. 58. Blochwitz, J.; Pfeiffer, M.; Fritz, T.; Leo, K. Applied Physics Letters 1998 , 73 , 729-731. 59. Deng, Z. B.; Lu, Z. Y.; Chen, Y. L.; Yin, Y. H.; Zou, Y.; Xiao, J.; Wang, Y. S. Solid- State Electronics 2013 , 89 , 22-25. 60. Lu, Z. Y.; Li, X.; Wang, Y.; Xiao, J.; Xu, P. L. Current Applied Physics 2014 , 14 , 1465- 1469. 61. Bao, Z.; Lovinger, A. J.; Dodabalapur, A. Applied Physics Letters 1996 , 69 , 3066-3068. 62. Huang, W. G.; Sinha, J.; Yeh, M. L.; Hardigree, J. F. M.; LeCover, R.; Besar, K.; Rule, A. M.; Breysse, P. N.; Katz, H. E. Advanced Functional Materials 2013 , 23 , 4094-4104. 63. Kvitschal, A.; Cruz-Cruz, I.; Hummelgen, I. A. Organic Electronics 2015 , 27 , 155-159. 64. Morse, G. E.; Bender, T. P. ACS Applied Materials & Interfaces 2012 , 4, 5055-5068. 65. Claessens, C. G.; Gonzalez-Rodriguez, D.; McCallum, C. M.; Nohr, R. S.; Schuchmann, H. P.; Torres, T. Journal of Porphyrins and Phthalocyanines 2007 , 11 , 181-188. 66. Brisson, E. R. L.; Paton, A. S.; Morse, G. E.; Bender, T. P. Industrial & Engineering Chemistry Research 2011 , 50 , 10910-10917. 67. Morse, G. E.; Castrucci, J. S.; Helander, M. G.; Lu, Z. H.; Bender, T. P. ACS Applied Materials & Interfaces 2011 , 3, 3538-3544. 68. Paton, A. S.; Morse, G. E.; Castelino, D.; Bender, T. P. Journal of Organic Chemistry 2012 , 77 , 2531-2536. 69. Kamino, B. A.; Bender, T. P. Dalton Transactions 2013 , 42 , 13145-13150. 70. Morse, G. E.; Helander, M. G.; Stanwick, J.; Sauks, J. M.; Paton, A. S.; Lu, Z. H.; Bender, T. P. Journal of Physical Chemistry C 2011 , 115 , 11709-11718. 71. Geyer, M.; Plenzig, F.; Rauschnabel, J.; Hanack, M.; delRey, B.; Sastre, A.; Torres, T. Synthesis-Stuttgart 1996 , 1139-1151.

72. Morse, G. E.; Paton, A. S.; Lough, A.; Bender, T. P. Dalton Transactions 2010 , 39 , 3915- 3922. 73. Kipp, R. A.; Simon, J. A.; Beggs, M.; Ensley, H. E.; Schmehl, R. H. Journal of Physical Chemistry A 1998 , 102, 5659-5664. 74. Gonzalez- Rodriguez, D.; Torres, T.; Guldi, D. M.; Rivera, J.; Herranz, M. A.; Echegoyen, L. Journal of the American Chemical Society 2004 , 126 , 6301-6313. 75. Morse, G. E.; Helander, M. G.; Maka, J. F.; Lu, Z. H.; Bender, T. P. ACS Applied Materials & Interfaces 2010 , 2, 1934-1944. 76. Gonzalez-Rodriguez, D.; Torres, T.; Olmstead, M. M.; Rivera, J.; Herranz, M. A.; Echegoyen, L.; Castellanos, C. A.; Guldi, D. M. Journal of the American Chemical Society 2006 , 128 , 10680-10681. 77. Mutolo, K. L.; Mayo, E. I.; Rand, B. P.; Forrest, S. R.; Thompson, M. E. Journal of the American Chemical Society 2006 , 128 , 8108-8109. 78. Gommans, H.; Cheyns, D.; Aernouts, T.; Girotto, C.; Poortmans, J.; Heremans, P. Advanced Functional Materials 2007 , 17 , 2653-2658. 79. Pandey, R.; Zou, Y. L.; Holmes, R. J. Applied Physics Letters 2012 , 101 , 033308. 80. Gommans, H.; Aernouts, T.; Verreet, B.; Heremans, P.; Medina, A.; Claessens, C. G.; Torres, T. Advanced Functional Materials 2009 , 19 , 3435-3439. 81. Verreet, B.; Rand, B. P.; Cheyns, D.; Hadipour, A.; Aernouts, T.; Heremans, P.; Medina, A.; Claessens, C. G.; Torres, T. Advanced Energy Materials 2011 , 1, 565-568. 82. Ebenhoch, B.; Prasetya, N. B. A.; Rotello, V. M.; Cooke, G.; Samuel, I. D. W. Journal of Materials Chemistry A 2015 , 3, 7345-7352. 83. Cnops, K.; Rand, B. P.; Cheyns, D.; Verreet, B.; Empl, M. A.; Heremans, P. Nature Communications 2014 , 5, 3406. 84. Hancox, I.; Sullivan, P.; Chauhan, K. V.; Beaumont, N.; Rochford, L. A.; Hatton, R. A.; Jones, T. S. Organic Electronics 2010 , 11 , 2019-2025. 85. Díaz, D. D.; Bolink, H. J.; Cappelli, L.; Claessens, C. G.; Coronado, E.; Torres, T. Tetrahedron Letters 2007 , 48 , 4657-4660. 86. Helander, M. G.; Morse, G. E.; Qiu, J.; Castrucci, J. S.; Bender, T. P.; Lu, Z. H. ACS Applied Materials & Interfaces 2010 , 2, 3147-3152.

87. Yasuda, T.; Tsutsui, T. Molecular Crystals and Liquid Crystals 2007 , 462 , 3-9. 88. Renshaw, C. K.; Xu, X.; Forrest, S. R. Organic Electronics 2010 , 11 , 175-178. 89. Claessens, C. G.; González-Rodríguez, D.; Torres, T.; Martín, G.; Agulló-López, F.; Ledoux, I.; Zyss, J.; Ferro, V. R.; De La García Vega, J. M. Journal of Physical Chemistry B 2005 , 109 , 3800-3806. 90. Meller, A.; Ossko, A. Monatshefte fuer Chemie 1972 , 103 , 150-155. 91. Kasuga, K.; Idehara, T.; Handa, M. Inorganica Chimica Acta 1992 , 196 , 127-128. 92. Musluoglu, E.; Gürek, A.; Ahsen, V.; Gül, A.; Bekaroğlu, Ö. Chemische Berichte 1992 , 125 , 2337-2339. 93. Kobayashi, N.; Kondo, R.; Nakajima, S.; Osa, T. Journal of the American Chemical Society 1990 , 112 , 9640-9641. 94. Weitemeyer, A.; Kliesch, H.; Wӧhrle, D. Journal of Organic Chemistry 1995 , 60 , 4900- 4904. 95. Zyskowski, C. D.; Kennedy, V. O. Journal of Porphyrins and Phthalocyanines 2000 , 4, 707-712. 96. Claessens, C. G.; González-Rodríguez, D.; del Rey, B.; Torres, T.; Mark, G.; Schuchmann, H.-P. von Sonntag, C.; MacDonald, J. G.; Nohr, R. S. European Journal of Organic Chemistry 2003 , 2547-2551. 97. Paton, A. S.; Morse, G. E.; Lough, A. J.; Bender, T. P. CrystEngComm 2011 , 13 , 914- 919. 98. Paton, A. S.; Lough, A. J.; Bender, T. P. CrystEngComm 2011 , 13 , 3653-3656. 99. Bonnier, C.; Josey, D. S.; Bender, T. P. Australian Journal of Chemistry 2015 , 68 , 1750- 1758. 100. Paton, A. S.; Bender, T. P. Journal of Porphyrins and Phthalocyanines 2014 , 18 , 1051- 1056. 101. Potz, R.; Goldner, M.; Huckstadt, H.; Cornelissen, U.; Tutass, A.; Homborg, H. Zeitschrift fuer Anorganische and Allgemeine Chemie. 2000 , 626 , 588-596. 102. Kato, T.; Tham, F. S.; Boyd, P. D. W.; Reed, C. A. Heteroatom Chemistry 2006 , 17 , 209- 216.

103. Guilleme, J.; González-Rodríguez, D.; Torres, T. Angewandte Chemie International Edition 2011 , 50 , 3506-3509. 104. Nonell, S.; Rubio, N.; del Rey, B.; Torres, T. Journal of the Chemical Society, Perkin Transactions 2 2000 , 1091-1094. 105. Verreet, B.; Schols, S.; Cheyns, D.; Rand, B. P.; Gommans, H.; Aernouts, T.; Heremans, P.; Genoe, J. Journal of Materials Chemistry 2009 , 19 , 5295-5297. 106. Ma, B.; Woo, C. H.; Miyamoto, Y.; Frechet, J. M. J. Chemistry of Materials 2009 , 21 , 1413-1417. 107. Cheyns, D.; Rand, B. P.; Heremans, P. Applied Physics Letters 2010 , 97 , 033301/033301-033301/033303. 108. Kulshreshtha, C.; Kim, G. W.; Lampande, R.; Huh, D. H.; Chae, M.; Kwon, J. H. Journal of Materials Chemistry A 2013 , 1, 4077-4082. 109. Chen, G.; Sasabe, H.; Sano, T.; Wang, X.-F.; Hong, Z.; Kido, J.; Yang, Y. Nanotechnology 2013 , 24 , 484007/484001-484007/484009. 110. Verreet, B.; Cnops, K.; Cheyns, D.; Heremans, P.; Stesmans, A.; Zango, G.; Claessens, C. G.; Torres, T.; Rand, B. P. Advanced Energy Materials 2014 , 4, 1301413/1301411- 1301413/1301418. 111. Menke, S. M.; Holmes, R. J. ACS Applied Materials & Interfaces 2015 , 7, 2912-2918. 112. Matlaba, P.; Nyokong, T. Polyhedron 2002 , 21 , 2463-2472. 113. Rauschnabel, J.; Hanack, M. Tetrahedron Letters 1995 , 36 , 1629-1632. 114. Giribabu, L.; Kumar, C. V.; Surendar, A.; Reddy, V. G.; Chandrasekharam, M.; Reddy, P. Y. Synthetic Communications 2007 , 37 , 4141-4147. 115. Takao, Y.; Masuoka, T.; Yamamoto, K.; Mizutani, T.; Matsumoto, F.; Moriwaki, K.; Hida, K.; Iwai, T.; Ito, T.; Mizuno, T.; Ohno, T. Tetrahedron Letters 2014 , 55 , 4564- 4567. 116. Yamamoto, K.; Takagi, A.; Hada, M.; Taniwaki, R.; Mizutani, T.; Kumura, Y.; Takao, Y.; Moriwaki, K.; Matsumoto, F.; Ito, T.; Iwai, T.; Hida, K.; Mizuno, T.; Ohno, T. Tetrahedron 2016 , 72, 4918-4924. 117. Stork, J. R.; Potucek, R. J.; Durfee, W. S.; Noll, B. C. Tetrahedron Letters 1999 , 40 , 8055-8058.

118. Martin, G.; Rojo, G.; Agullo-Lopez, F.; Ferro, V. R.; Garcia de la Vega, J. M.; Martinez- Diaz, M. V.; Torres, T.; Ledoux, I.; Zyss, J. J. Phys. Chem. B 2002 , 106 , 13139-13145. 119. Pandey, R.; Kerner, R. A.; Menke, S. M.; Holst, J.; Josyula, K. V. B.; Holmes, R. J. Organic Electronics 2013 , 14 , 804-808. 120. Matsuo, Y.; Sato, Y.; Niinomi, T.; Soga, I.; Tanaka, H.; Nakamura, E. Journal of the American Chemical Society 2009 , 131 , 16048-16050. 121. Guide, M.; Lin, J. D. A.; Proctor, C. M.; Chen, J. R.; Garcia-Cervera, C.; Nguyen, T. Q. Journal of Materials Chemistry A 2014 , 2, 7890-7896. 122. Mitsumori, T.; Bendikov, M.; Dautel, O.; Wudl, F.; Shioya, T.; Sato, H.; Sato, Y. Journal of the American Chemical Society 2004 , 126 , 16793-16803. 123. Mi, B. X.; Wang, P. F.; Liu, M. W.; Kwong, H. L.; Wong, N. B.; Lee, C. S.; Lee, S. T. Chemistry of Materials 2003 , 15 , 3148-3151.

Chapter 2 A boron subphthalocyanine polymer: Poly(4-methylstyrene)-co - poly(phenoxy-boron-subphthalocyanine)

Adapted with permission from: Jeremy D. Dang, Jessica D. Virdo, Benoit H. Lessard, Elijah Bultz, Andrew S. Paton, and Timothy P. Bender. “A Boron Subphthalocyanine Polymer: Poly(4-methylstyrene)-co - poly(phenoxy boron subphthalocyanine).” Macromolecules 2012 , 45 , 7791-7798. DOI: 10.1021/ma301247p.

JDV and I contributed equally to the preparation of this paper.

ASP initiated this project by attempting to synthesize the isoeugenol-substituted BsubPc monomer 2b (Scheme 2.1). BHL determined that Br-BsubPc was a radical scavenger. EB developed the synthetic process for the homopolymers 3 and 4 (Scheme 2.2). JDV determined that the BsubPc polymer 5 could not be synthesized from a homopolymer approach, proposed a random copolymer approach, and carried out the syntheses of the copolymers 6-11 (Scheme 2.2). I attempted to synthesize the 4-vinyl phenoxy-BsubPc monomer 2a (Scheme 2.1), but found that either autopolymerization of 4-vinyl phenol occurred or a mixture of a small molecule BsubPc derivative and a BsubPc polymer was formed that could not be purified. I scaled up the synthesis of the copolymers 10a and 10b and optimized the synthesis of the BsubPc polymers 11a and 11b . I also characterized the intermediate copolymers 10a and 10b and the BsubPc polymers 11a and 11b via NMR, UV-vis absorption, and photoluminescence spectroscopy. The work was supervised by TPB. All authors approved the manuscript.

Supplementary information for this chapter is found in Appendix A.

2.1 Abstract BsubPcs have been explored as functional materials in a wide range of organic electronic applications, but only as a small molecule species. Herein, the synthesis of the first set of BsubPc-containing polymers, poly(4-methyl styrene)-co -poly(phenoxy-boron- subphthalocyanine), is reported along with its basic physical properties. The synthesis was only possible via a post-polymerization functionalization approach, where the BsubPc moiety is appended to the side chain of a styrene-based pre-polymer. The three-step transformation begins with the nitroxide mediated copolymerization of 4-methyl styrene and 4-acetoxystyrene to form poly(4-methyl styrene)-co -poly(4-acetoxystyrene). This is followed by deacetylation of the acetoxy groups to afford hydroxyl groups prior to phenoxylation with Br-BsubPc.

2.2 Introduction BsubPcs ( 1a /1b , Scheme 2.1) have demonstrated some sign of promise as organic photoactive materials not only in OPVs 1-4, but in other areas of OEs like NLOs, 5 OFETs, 6,7 and OLEDs 8,9 . Despite the widespread use of both small molecules and polymer materials as functional components in OEs, all investigations to date have been limited to the study of BsubPcs as small molecule species; there has yet to be a report on the synthesis and physical characterization of a BsubPc-containing polymer. In contrast, there are a number of reports on normal Pc containing polymers, which differ in the location of the Pc fragments within the macromolecule: either within the main chain, 10-14 as a side chain (pendant), 15-18 or as the main fragment making up a network polymer 19-24 .

Many semiconducting polymers of interest in the field of OEs contain the functional conjugated component within the main chain of the polymer. Examples include poly(paraphenylene)s, poly(thiophene)s, poly(pyrrole)s, poly(ethylene dioxythiophene)s, poly(fluorene)s, poly(carbazole)s, and various donor-acceptor copolymers. 25 In each case, the polymer is produced by the polycondensation of two difunctional monomers or a single bifunctional monomer. We can surmise that the lack of precedent for a BsubPc-containing polymer is attributed to the practical improbability of producing a difunctional or bifunctional monomer of the C 3V symmetric BsubPc.

It has been stated that polymeric OE materials benefit from cost effective fabrication techniques including roll-to-roll solution coating 26 and ink-jet printing 27 and it is thus desirable to study a polymeric version of BsubPc for direct comparison with its small molecule analogs. Herein, we describe the synthesis of the first polymeric BsubPc, poly(4-methylstyrene)-co -poly(phenoxy- boron-subphthalocyanine), which has been achieved using a post-polymerization transformation strategy. We emphasize the practical importance of this approach and outline the basic physical properties of the novel BsubPc-containing polymer(s).

2.3 Results and Discussion As mentioned above, incorporation of a BsubPc unit into a polymer main-chain by a polycondensation strategy would be synthetically challenging due to the C 3v symmetric nature of BsubPc. We therefore aimed to attach the BsubPc as a pendant group onto an inert polymer backbone. The obvious first step was the synthesis of a BsubPc molecule containing a vinyl functional group, which could then be polymerized using either conventional or reversible- deactivation radical polymeration 28 methods like nitroxide mediated polymerization (NMP) 29 . This pre-polymerization functionalization approach has been successfully employed for the synthesis of a range of monomers bearing functional pendants, including acrylate monomers bearing Disperse Red 1 30,31 and various other azobenzene pendants, 32 a vinylbenzene monomer bearing carbazole, 33 and a p-cyanomethylstyrene modified with a number of benzaldehyde pendants 34 .

In our case, we targeted a styrene-based polymer as styrene is well known to not interfere with charge transport while other polymers and their associated dipoles might hinder charge transport by the so-called pinning effect ( e.g. acrylates). 35 It has been shown many times that chloro boron subphthalocyanine (Cl-BsubPc, 1a , Scheme 2.1) can be axially substituted with a phenol or phenol derivative in refluxing toluene, chlorobenzene or other aromatic solvents. 36-40 Thus, our first attempt to synthesize a monomer of BsubPc utilized 4-vinyl phenol. Despite its structural simplicity, 4-vinyl phenol is not commercially available and was obtained through the hydrolysis of 4-acetoxy styrene (4-AS) with aqueous potassium hydroxide at 0 °C and either stored in the refrigerator prior to use or used immediately after preparation. 41 Reaction of 4-vinyl phenol with

1a resulted in autopolymerization of the 4-vinyl phenol at a temperature of 80 °C, which is lower than what would be required to facilitate the production of monomer 2a (Scheme 2.1). Similar observations were seen for the reaction of 1a with 2-methoxy-4-propenylphenol or isoeugenol in an attempt to make monomer 2b (Scheme 2.1).

Scheme 2.1 . Attempted synthesis of BsubPc-containing styrenic monomers via the pre- polymerization functionalization approach. Reagents and conditions: (i) X = Cl, R 1 = H, R 2 = H, toluene, 80 °C, 24 hours; (ii) X = Cl, R 1 = OCH 3, R 2 = CH 3, toluene 80 °C, 24 hours; (iii) X =

Br, R 1 = H, R 2 = H, toluene, 40 °C, 48 hours.

From these experiments, it was apparent that lower temperatures were required for the phenoxylation of 1a to overcome the issue of premature auto-polymerization. However, using a low reaction temperature will not permit phenoxylation of 1a to proceed. We thus turned to the use of the more reactive Br-BsubPc ( 1b , Scheme 2.1). Reaction of 1b with 4-vinyl phenol in toluene at 40 °C formed several products that included a compound with a different HPLC retention time (R t) than 1b and a UV-vis absorption profile characteristic of a BsubPc derivative, as well as a polymer (detected by GPC) with a molecular weight (M w) of ~20,000 Da. Attempts to isolate pure compounds from this mixture via silica gel column chromatography were unsuccessful due to the elution of the side products and what seemed to be the result of further reaction promoted by the silica gel. We surmised that the polymer was likely formed by a polymerization catalyzed by HBr acid, a side product liberated from the phenoxylation of 1b . In a second attempt, the experiment was repeated in the presence of pyridine so as to neutralize any HBr formed during the reaction. Premature polymerization was indeed inhibited and a compound

with the same HPLC R t and absorption profile as the previous base-free experiment was formed. However, the compound could not be isolated even after successive purification via silica gel chromatography for what we assume are the same reasons outlined above.

O O N O n X O O O N O N N N O O N n n B N N 5 N O (i) x N N N 3O O 4 OH (ii) N B N

O O N O O N O O n m n m X N N N (iii) NB N N O O OH 6a 7a (iv) 9a, 9b 10a, 10b O O N O n m

N N N 8a N B N 11a, 11b

Scheme 2.2 . The synthesis of BsubPc-containing copolymers via the post-polymerization functionalization approach. Conditions: (i) aqueous ammonium hydroxide, isopropanol, reflux, 18 hours; (ii) 1a (X = Cl) / 1b (X = Br), 1:1 ( v/v ) N,N-dimethylacetamide:1,2-dichlorobenzene, reflux, 24 hours; (iii) Method 1 (6a ): aqueous ammonium hydroxide, isopropanol, reflux, 70 hours; Method 2 (9a ): aqueous potassium hydroxide, isopropanol, reflux, 12 hours; Method 3 (9a and 9b ): sulfuric acid, toluene/ethanol, reflux, 12 hours; (iv) Method 1 (7a ): 1b , toluene/chlorobenzene, reflux, 2 hours; Method 2 (10a and 10b ): 1b , chlorobenzene, 120 °C, ~24-48 hours.

Before additional effort was put into obtaining a BsubPc vinyl monomer we first wanted to determine whether or not the BsubPc group is a radical scavenger. Therefore, two 4- methylstyrene (4-MS) homopolymerizations were carried out under NMP conditions, where one of the homopolymerizations was performed with the addition of 0.04 wt % 1b (as an additive, not as a monomer). Even with a minimal amount of 1b in the homopolymerization mixture significant termination was observed, as evident by the loss of linearity in the scaled conversion (ln((1-X)-1)) versus time plot (Figure S2.1). This termination therefore indicates that 1b and perhaps more generally BsubPcs are radical scavengers.

Due to the challenges associated with obtaining a styrenic monomer of BsubPc and its apparent ability to scavenge radicals, we turned to a post-polymerization functionalization approach (Scheme 2.2) whereby 1b would be reacted with a pre-formed polymer (pre-polymer) to produce the final BsubPc-containing polymer/macromolecule. Using this approach we first tried to functionalize the pre-polymer poly(vinyl phenol) with BsubPc. Rather than using commercially available poly(vinyl phenol), which has a very large polydispersity index (PDI ~6), we synthesized poly(vinyl phenol) in house using NMP. Specifically we produced the homopolymer by homopolymerization of 4-AS using benzoyl peroxide (BPO) as the initiator and 2,2,6,6- tetramethyl-1-piperidinyloxy (TEMPO) as the mediator at a BPO:TEMPO ratio of 1:1.5.

Intermediate poly(4-AS) ( 3, Scheme 2.2) had the following characteristics: M w = 21,600; PDI = 1.32. Subsequent hydrolysis of 3 to poly(vinyl phenol) ( 4, Scheme 2.2) was accomplished using aqueous ammonium hydroxide in an approach analogous to that of Barclay et al .42 Subsequent reaction of 4 with 1a or 1b turned out to be problematic as polymer 4 is not soluble in aromatic solvents such as toluene or chlorobenzene, solvents commonly used for phenoxylation of halo- BsubPcs. 20 Instead, a 1:1 ( v/v ) mixture of 1,2-dichlorobenzene and a polar aprotic solvent such as N,N -dimethylformamide (DMF) or N,N -dimethylacetamide (DMAc) was needed to solubilize polymer 4. Addition of either 1a or 1b with heating under a positive pressure of argon caused some initial conversion/partial substitution to the desired BsubPc-containing polymer 5 (as could be seen in the GPC chromatogram that the polymer obtained the characteristics UV-vis absorption profile of a BsubPc derivative), but significant decomposition of the BsubPc chromophore occurred well before 50% substitution. The unidentified decomposition product(s)

were green in color and it was suspected that the decomposition was a result of the known disproportionation and ring expansion of BsubPc derivatives, 43 although we could not isolate/purify the decomposition products for detailed characterization. We can, however, surmise that the decomposition is a result of the presence of the polar aprotic solvent or the combination of the aprotic solvent and the liberated HCl or HBr, as 1a is stable under refluxing dichlorobenzene for periods of time exceeding 100 hours 44 and 1b is stable in refluxing toluene for periods of at least 5 hours. 45 We also verified that decomposition of the BsubPc occurred regardless of the source or batch of pre-polymer used including if the pre-polymer is simply commercially poly(vinyl phenol). We can therefore eliminate the presence of nitroxide end- groups as another possible source of the decomposition. As a final confirmation of the effect of polar aprotic solvents, 1b was reacted with 5 equivalents of phenol under standard conditions except using 1:1 ( v/v ) of toluene:DMAc. In this case decomposition of the BsubPc was also observed definitely proving that it is the presence of the polar aprotic solvent possibly combined with the presence of the liberated HBr that causes the decomposition of the BsubPc chromophore.

Thereafter, we shifted our investigation to the use of copolymers as pre-polymers rather than poly(vinyl phenol) homopolymers. The aim was the inclusion of a selected comonomer that would produce a 4-AS-containing pre-polymer, which upon hydrolysis would produce a phenol- containing pre-polymer that was soluble in solvents such as toluene and/or chlorobenzene – solvents we knew that would not promote the decomposition of the BsubPc. As measured reactivity ratios were not widely available for 4-AS, we used the ‘Patterns of Reactivity’ method to examine a variety of potential comonomers (Table S2.1). Parameter values (r i,s , p i, ui and vi) for most monomers needed for this method were obtained from the Polymer Handbook 46 and used in the Patterns of Reactivity calculations directly (Eq S2.1-S2.3). In considering candidate monomers for copolymerization with 4-AS, several selection criteria were used: (1) The monomer, as stated above, should produce a copolymer with 4-vinyl phenol that is soluble in solvents such as toluene, chlorobenzene, or dichlorobenzene; (2) the monomer should produce a polymer that is insoluble in methanol and/or hexanes to facilitate workup via precipitation and Soxhlet extraction; (3) The monomer should not be an ester since hydrolysis is necessary to

transform the acetoxy groups of 4-AS to hydroxyl group of 4-vinyl phenol and such conditions would obviously also hydrolyze any other ester present.

Thus, on consideration, only 4-methyl styrene (4-MS) fit these criteria and was selected for copolymerization with 4-AS. The predicted reactivity ratios for 4-AS and 4-MS are closest to unity, and therefore this monomer pair is likely to produce a copolymer with a random tendency.

We initially set an arbitrary target of M w = 40,000 for a 4-AS/4-MS copolymer. The arbitrary target was thought to be high enough to ensure good film-forming properties of the resulting polymer. Using NMP, 4-MS and 4-AS were copolymerized at a charged ratio of 2:1 (mol/mol,

Scheme 2.2). This successfully yielded copolymer 6a with M w = 41,000 and PDI = 1.44. Gas chromatography (GC) analysis confirmed an equal rate of consumption for 4-AS and 4-MS, verifying that 6a is a random copolymer. Hydrolysis of 6a to 7a was accomplished by suspending 6a in isopropanol and heating to a reflux with excess ammonia hydroxide. As hydrolysis proceeded, the copolymer became increasingly soluble eventually forming a clear and homogeneous solution. The overall reaction time was quite long requiring approximately three days for complete hydrolysis.

Next the reaction of 7a with 1b was explored. The desired reaction solvent was toluene or chlorobenzene. Using either only formed a solution at 2 - 4 wt % of polymer 7a even at high temperatures. This was significantly more dilute than the typical phenoxylation reaction conditions used. Given the S N1 nature of the reaction, the low concentration would severely limit the reaction rate and thus, the reaction of 7a with 1b was only attempted due to its known higher reaction rate. 27 The reaction was monitored by GPC and after 90 min the reaction had stalled as indicated by the presence of unreacted and unchanging quantities of 1b . Removal of the mother liquor and isolation of the solubilized polymer showed that only ~5% of the polymer had stayed in solution. GPC analysis of the isolated polymer revealed that it did contain BsubPc (based on the characteristic UV-vis absorption spectrum) with a M w of 20,000 Da. Based on this observation, fractionation had occurred and this indicates that the target M w of 40,000 Da was too high when limited to the use of toluene or chlorobenzene as solvents. The synthesis was

repeated with a target M w of 15,000 - 18,000 in an attempt to ensure complete dissolution during the post-polymerization reaction process in either toluene or chlorobenzene.

Again using NMP, 4-MS and 4-AS were copolymerized in a 2:1 (mol/mol) ratio this time with a targeted M w of 15,000 - 18,000. The evolution of M w and PDI was monitored over time (Figure S2.2). Monomer consumption was also measured over time using GC against an internal standard of 1,2-dichlorobenzene (Figure S2.3). In this case, 4-AS was consumed at a slightly higher but comparable rate to 4-MS indicating approximate equivalence in their reactivity and the production of a random copolymer. Noting that copolymer 9a still had somewhat low solubility in toluene and chlorobenzene, we also synthesized a 4-MS/4-AS copolymer in a 4:1 (mol/mol) ratio to produce 9b . In the synthesis of both 9a and 9b , the process produced copolymers with M w that varied from batch-to-batch and, in some cases, the target M w was overshot especially at slightly larger scale. Despite this, copolymer 9a (M w = 30,800, PDI =

1.85) and 9b (M w = 25,200, PDI = 1.75) were both carried forward to the subsequent deacetylation reaction.

Deacetylation was carried out via base hydrolysis of 9a to give 2:1 poly(4-methylstyrene)-co - poly(4-vinyl phenol) ( 10a , Mw = 27,600, PDI = 2.29) and via acid-catalyzed transesterification of 9b to give 4:1 poly(4-methylstyrene)-co -poly(4-vinyl phenol) ( 10b , Mw = 24,200, PDI = 1.88). Different reaction conditions were employed for the deacetylation reaction due the poor solubility of 9b in aqueous basic solution. Copolymer 9a can also be converted to 10a using the same acid-catalyzed transesterification reaction used to deacetylate 9b . The deacetylated copolymers 10a and 10b were subsequently reacted with 1.2 equivalents of 1b in chlorobenzene at 120 °C and the reaction progress was monitored by GPC (Figure S2.4 & S2.5) until the ratio of BsubPc-containing copolymer (R t ~ 13.0 - 13.2 min) and unreacted 1b (R t ~ 16.7 - 17.0 min) had reached a steady value. Although the GPC retention time for the polymer peak did not shift with reaction progress, phenoxylation was known to occur based on a change in the UV-vis absorption profile for the polymer peak from that of a standard styrenic polymer to that of a characteristic BsubPc compound (Figure 2.1). Each of the 2:1 ( 11a , final M w = 8,700, PDI =

5.65) and 4:1 ( 11b, final M w = 17,600, PDI = 8.34) poly(4-methylstyrene)-co -poly(phenoxy- boron-subphthalocyanine) were isolated and purified via precipitation and a Soxhlet extraction.

(a)

(b) Figure 2.1 . GPC chromatograms and extracted UV-vis absorption spectra (inset) of BsubPc polymers 11a (a) and 11b (b).

The GPC spectrum of 11a and 11b (Figure 2.1) both showed a tailing effect, signifying a potential solubility issue with these copolymers. As a consequence of this, high PDIs were obtained, which were assumed to be inaccurate based on the fact that the PDIs for 9a and 9b were significantly lower (< 2). Furthermore, the reported M w of 11a and 11b were low compared to the M w of the pre-polymers again indicating a potential solubility issue resulting in a decreased hydrodynamic volume upon reaction with BsubPc.

The solubilities of 11a and 11b were confirmed to be generally poor in common organic solvents with 11b displaying a slightly better solubility than 11a . THF was found to be the best solvent for dissolving these BsubPc-containing copolymers. However, for NMR purposes THF was not an ideal solvent since one of its residual peaks overlapped with one of the aliphatic peaks of the copolymers. Chloroform-d was chosen instead as it demonstrated the best compromise between solubility and minimal residual peak overlap with the sample (Figure S2.6).

The UV-vis absorption spectra were acquired for 11a and 11b in non-deaerated toluene (Figure

2.2). Both polymers have a λ max of absorption at ~564 nm. This is similar to the λ max of absorption for small molecule phenoxy-BsubPc derivatives, which typically occurs over the range of 560-565 nm. A representative example spectrum of 3,4-dimethylphenoxy-BsubPc 47 is shown in Figure 2.2. When comparing the absorption spectra of polymers 11a and 11b each are essentially identical, (small differences in the UV) however, the respective long wavelength absorptions are slightly broadened when compared to the small molecule analog, indicating that the BsubPc chromophores are partially associated in solution. Photoluminescence (PL) emission spectra were also acquired for the polymers 11a and 11b (Figure 2.2) in non-deaerated toluene using a λ excitation of 564 nm. The PL spectra show that 11a and 11b have λ max of emission at 573 and 575 nm, respectively, which is again in the range typically observed for small molecule phenoxy-BsubPcs. The PL quantum yields (φ) for 11a and 11b were found to be relatively low at

0.12 and 0.06, respectively, relative to a standard of phenoxy-dodecafluoro-BsubPc (F 12BsubPc; Eq S2.4). This is significantly lower than the ~0.50 that is commonly observed for phenoxy- BsubPc derivatives. 9 While the Stokes shift is small, the low quantum yield clearly indicates that other processes are present within the polymer, except for the degeneration of the singlet to the ground state perhaps resulting from the aforementioned partial association of the BsubPc chromophores in solution.

Figure 2.2 . UV-vis absorption (blue) and photoluminescence (PL, red) spectra of 3,4- dimethylphenoxy-BsubPc and polymers 11a and 11b (as indicated).

In addition, PL excitation spectra were obtained for polymers 11a and 11b and the model compound 3,4-dimethylphenoxy-BsubPc in non-deaerated toluene using an observational λ emission of 573 and 575 nm for 11a and 11b , respectively (Figure S2.7). The excitation spectrum for 3,4- dimethylphenoxy-BsubPc is practically identical to that of the absorption spectrum (Figure S2.7a). The excitation spectrum for polymers 11a and 11b closely matches the respective absorption spectrum, however, each are slightly red-shifted relative to the absorption spectrum (Figure S2.7b & S2.7c) indicating only a fraction of the absorbing chromophores are responsible for the observed photoluminescence. These observations coupled with the broadened absorption spectrum and low quantum yield of photoluminescence suggest further study of the photophysics of the BsubPc-containing polymers is warranted.

Cyclic voltammetry (CV) of 11a and 11b were carried out to determine the energy levels of the FMO ( i.e. HOMO and LUMO). Initial experiments using Ag/AgCl as the reference electrode, Pt disk as the working electrode, Pt wire as the counter electrode, and tetrabutylammonium perchlorate in either DCM or THF as the electrolyte did not show any reduction or oxidation process for either 11a or 11b . It was suspected that the concentration of these copolymers in

DCM or THF was too low to be detected due to their generally poor solubilities. To address this issue, the Pt disk was coated by drop casting with a solution of 11a or 11b onto the electrode to ensure that their concentration at the working electrode was always high during the electrochemical study. Acetonitrile was used as the solvent as it does not solubilize the copolymers. However, once again, there was an absence of a reduction or oxidation process. To potentially address this, a high-surface-area working electrode, indium tin oxide (ITO) on glass substrate, was used instead to ensure that there would be a high amount of 11a or 11b at the working electrode. Coating onto the ITO plate was done via a drip-coat method, where the ITO plate was dipped into a THF solution of the copolymer and allowed to dry. Numerous CV studies using this setup were conducted and the results were inconsistent. In most cases no current was observed across the entire voltage scanned. Thus, the basic electrochemical properties of the title polymer remain unknown.

An attempt to integrate 11a or 11b as an organic photoactive material in OPV devices was not performed due to a chemical instability issue with the polymers. Some small molecule BsubPcs were found in both samples of 11a and 11b within two years of storage under ambient conditions as determined by GPC (Figure 2.3). Considering that both samples were exclusively the title polymer ( i.e. pure) prior to their storage, the source of the small molecule BsubPcs had to be derived from them decoupling from the poly(4-methylstyrene)-co -poly(4-vinyl phenol) scaffold. As a consequence of some loss of the photoactive component, 11a and 11b were not tested in OPV devices.

(a)

(b)

Figure 2.3 . GPC chromatograms and extracted UV-vis absorption spectra ( left : lower R T; right : higher R t) of BsubPc polymers 11a (a) and 11b (b) within two years of storage.

Although this was very discouraging, the work presented herein did demonstrate the feasibility of a BsubPc polymer, a mission that was not possible previously. The post-polymerization functionalization approach was later adopted by our group to couple Br-BsubPc to poly(styrene)- co -poly(acrylic acid) 48 (12 ) and poly(styrene)-co -poly(methacrylic acid) 49 (14 ) via the carboxylic acid moiety to form the corresponding BsubPc polymers 13 and 15 (Scheme 2.3). These BsubPc polymers, like 11a and 11b , had similar photophysical properties as their small molecule counterparts and were found to be stable even after ~1.5 year of storage. Polymer 15 , which was the easier one of the two to synthesize, was explored as a hole-transport layer in OLEDs, whereby orange emission was produced. This marked the first time a BsubPc polymer was studied in an OE device. 49 The characterization of 15 within OPV devices is the matter of future investigation.

Scheme 2.3 . The synthesis of BsubPc-containing copolymers 13 and 15 via the post- polymerization functionalization approach. Conditions: (i) BPO, TEMPO, 125 ˚C; (ii) toluene, reflux; (iii) V59, SG1, 80 ˚C.

2.4 Conclusions After outlining why the synthesis of a BsubPc-containing monomer was not successful and perhaps pointless since BsubPc may be a radical scavenger, we demonstrated the synthesis of BsubPc-containing polymers for the first time by employing a post-polymerization functionalization approach. Two polymers were synthesized: a 2:1 ( 11a ) and 4:1 ( 11b ) poly(4- methylstyrene)-co -poly(phenoxy-boron-subphthalocyanine). The three-step process begins with the nitroxide mediated copolymerization of 4-MS and 4-AS in a 2:1 and 4:1 molar ratio to give the 2:1 ( 9a ) and 4:1 ( 9b ) poly(4-methylstyrene)-co -poly(4-acetoxystyrene), respectively. Following deacetylation, 9a and 9b were converted to their respective 2:1 ( 10a ) and 4:1 ( 10b ) poly(4-methylstyrene)-co -poly(4-vinyl phenol), which was subsequently reacted with 1b via phenoxylation to afford the target 2:1 ( 11a ) and 4:1 ( 11b ) poly(4-methylstyrene)-co - poly(phenoxy-boron-subphthalocyanine). UV-vis absorption spectroscopy revealed that both 11a and 11b had similar absorption profiles as other BsubPcs derivatives. Despite this similarity, 11a and 11b were found to have low PL quantum yields. Cyclic voltammograms could not be obtained likely due to solubility issues of the polymers. The integration of these BsubPc polymers into OPV devices were not performed due to a chemical instability issue. However, the work herein led to the next generation of stable BsubPc polymers, whereby one of them was successfully incorporated into OLED devices.

2.5 References 1. Mutolo, K. L.; Mayo, E. I.; Rand, B. P.; Forrest, S. R.; Thompson, M. E. Journal of the American Chemical Society 2006 , 128 , 8108-8109. 2. Gommans, H.; Aernouts, T.; Verreet, B.; Heremans, P.; Medina, A.; Claessens, C. G.; Torres, T. Advanced Functional Materials 2009 , 19 , 3435-3439. 3. Verreet, B.; Rand, B. P.; Cheyns, D.; Hadipour, A.; Aernouts, T.; Heremans, P.; Medina, A.; Claessens, C. G.; Torres, T. Advanced Energy Materials 2011 , 1, 565-568. 4. Pandey, R.; Zou, Y. L.; Holmes, R. J. Applied Physics Letters 2012 , 101 , 033308. 5. Claessens, C. G.; González-Rodríguez, D.; Torres, T.; Martín, G.; Agulló-López, F.; Ledoux, I.; Zyss, J.; Ferro, V. R.; De La García Vega, J. M. Journal of Physical Chemistry B 2005 , 109 , 3800-3806. 6. Yasuda, T.; Tsutsui, T. Molecular Crystals and Liquid Crystals 2007 , 462 , 3-9. 7. Renshaw, C. K.; Xu, X.; Forrest, S. R. Organic Electronics 2010 , 11 , 175-178. 8. Diaz, D. D.; Bolink, H. J.; Cappelli, L.; Claessens, C. G.; Coronado, E.; Torres, T. Tetrahedron Letters 2007 , 48 , 4657-4660. 9. Morse, G. E.; Helander, M. G.; Maka, J. F.; Lu, Z.-H.; Bender, T. P. ACS Applied Materials & Interfaces 2010 , 2, 1934-1944. 10. Marks, T. J. Science 1985 , 227 , 881-889. 11. Bryant, G. C.; Cook, M. J.; Haslam, S. D.; Richardson, R. M.; Ryan, T. G.; Thorne, A. J. Journal of Materials Chemistry 1994 , 4, 209-216. 12. Bryant, G. C.; Cook, M. J.; Ryan, T. G.; Thorne, A. J. Journal of the Chemical Society- Chemical Communications 1995 , 467-468. 13. Shi, M. M.; Chen, H. Z.; Wang, M. Journal of Applied Polymer Science 1997 , 64 , 1769- 1774. 14. Kimura, M.; Wada, K.; Ohta, K.; Hanabusa, K.; Shirai, H.; Kobayashi, N. Macromolecules 2001 , 34 , 4706-4711. 15. Allcock, H. R.; Neenan, T. X. Macromolecules 1986 , 19 , 1495-1501. 16. Yang, H. X.; Sargent, J. R.; Hay, A. S. Journal of Polymer Science Part a-Polymer Chemistry 1995 , 33 , 989-997.

17. Chen, Y.; Hanack, M.; O'Flaherty, S.; Bernd, G.; Zeug, A.; Roeder, B.; Blau, W. J. Macromolecules 2003 , 36 , 3786-3788. 18. Zamora, F.; Gonzalez, C. Journal of Macromolecular Science-Physics 1996 , B35 , 709- 729. 19. Achar, B. N.; Fohlen, G. M.; Parker, J. A. Journal of Polymer Science Part a-Polymer Chemistry 1982 , 20 , 2773-2780. 20. Ahsen, V.; Yilmazer, E.; Bekaroglu, O.; Gul, A.; Bekaroglu, O. Makromolekulare Chemie-Rapid Communications 1987 , 8, 243-246. 21. Wohrle, D.; Schulte, B. Makromolekulare Chemie-Macromolecular Chemistry and Physics 1988 , 189 , 1167-1187. 22. Wohrle, D.; Schulte, B. Makromolekulare Chemie-Macromolecular Chemistry and Physics 1988 , 189 , 1229-1238. 23. Ahsen, V.; Yilmazer, E.; Bekaroglu, O. Makromolekulare Chemie-Macromolecular Chemistry and Physics 1988 , 189 , 2533-2543. 24. Ozdemir, M.; Agar, E. Spectroscopy Letters 1991 , 24 , 741-748. 25. Heeger, A. J. Chemical Society Reviews 2010 , 39 , 2354-2371. 26. Krebs, F. C.; Gevorgyan, S. A.; Alstrup, J. Journal of Materials Chemistry 2009 , 19 , 5442-5451. 27. Zschieschang, U.; Klauk, H.; Halik, M.; Schmid, G.; Dehm, C. Advanced Materials 2003 , 15 , 1147-1151. 28. Jenkins, A. D.; Jones, R. G.; Moad, G. Pure and Applied Chemistry 2010 , 82 , 483-491. 29. Grubbs, R. B. Polymer Reviews 2011 , 51 , 104-137. 30. Natansohn, A.; Rochon, P.; Gosselin, J.; Xie, S. Macromolecules 1992 , 25 , 2268-2273. 31. Barrett, C.; Natansohn, A.; Rochon, P. Macromolecules 1994 , 27 , 4781-4786. 32. Lim, S. L.; Li, N. J.; Lu, J. M.; Ling, Q. D.; Zhu, C. X.; Kang, E. T.; Neoh, K. G. ACS Applied Materials & Interfaces 2009 , 1, 60-71. 33. Lessard, B.; Ling, E. J. Y.; Morin, M. S. T.; Maric, M. Journal of Polymer Science Part a-Polymer Chemistry 2011 , 49 , 1033-1045. 34. Gupta, A.; Watkins, S. E.; Scully, A. D.; Singh, T. B.; Wilson, G. J.; Rozanski, L. J.; Evans, R. A. Synthetic Metals 2011 , 161 , 856-863.

35. Borsenberger, P. M.; Fitzgerald, J. J. Journal of Physical Chemistry 1993 , 97 , 4815-4819. 36. Morse, G. E.; Paton, A. S.; Lough, A.; Bender, T. P. Dalton Transactions 2010 , 39 , 3915- 3922. 37. Paton, A. S.; Lough, A. J.; Bender, T. P. CrystEngComm 2011 , 13 , 3653-3656. 38. Paton, A. S.; Morse, G. E.; Lough, A. J.; Bender, T. P. CrystEngComm 2011 , 13 , 914- 919. 39. Brisson, E. R. L.; Paton, A. S.; Morse, G. E.; Bender, T. P. Industrial & Engineering Chemistry Research 2011 , 50 , 10910-10917. 40. Morse, G. E.; Helander, M. G.; Stanwick, J.; Sauks, J. M.; Paton, A. S.; Lu, Z. H.; Bender, T. P. Journal of Physical Chemistry C 2011 , 115 , 11709-11718. 41. Sumner, M. J.; Weyers, R. Y.; Rosario, A. C.; Riffle, J. S.; Sorathia, U. Polymer 2004 , 45 , 5199-5206. 42. Barclay, G. G.; Hawker, C. J.; Ito, H.; Orellana, A.; Malenfant, P. R. L.; Sinta, R. F. Macromolecules 1998 , 31 , 1024-1031. 43. Kobayashi, N.; Ishizaki, T.; Ishii, K.; Konami, H. Journal of the Americal Chemical Society 1999 , 121 , 9096-9110. 44. Morse, G. E.; Castrucci, J. S.; Helander, M. G.; Lu, Z. H.; Bender, T. P. ACS Applied Materials & Interfaces 2011 , 3, 3538-3544. 45. Paton, A. S.; Morse, G. E.; Castelino, D.; Bender, T. P. Journal of Organic Chemistry 2012 , 77 , 2531-2536. 46. Brandrup, J. I., E. H.; Edmund, H.; Grulke, E. A.; Abe, A.; Bloch, D. R. In Polymer Handbook, 4th Ed. ; John Wiley & Sons: New York, 2005, p 181-288. 47. Paton, A. S.; Lough, A. J.; Bender, T. P. Industrial & Engineering Chemistry Research 2012 , 51 , 6290-6296. 48. Lessard, B. H.; Bender, T. P. Macromolecular Rapid Communications 2013 , 34 , 568- 573. 49. Lessard, B. H.; Sampson, K. L.; Plint, T.; Bender, T. P. Journal of Polymer Science Part a-Polymer Chemistry 2015 , 53 , 1996-2006.

Chapter 3 Process for the synthesis of symmetric and unsymmetric oxygen bridged dimers of boron subphthalocyanines (μ-oxo-(BsubPc) 2s)

Adapted with permission from: Jeremy D. Dang, Mabel V. Fulford, Brett A. Kamino, Andrew S. Paton, and Timothy P. Bender. "Process for the synthesis of symmetric and umsymmetric oxygen bridged dimers of boron subphthalocyanines (μ-oxo-(BsubPc) 2s)." Dalton Transactions 2015 , 44 , 4280-4288. DOI: 10.1039/C4DT02624A. and Jeffrey S. Castrucci, Richard K. Garner, Jeremy D. Dang, Emmanuel Thibau, Zheng-Hong Lu, and Timothy P. Bender. "Characterization of μ-oxo-(BsubPc) 2 in organic planar heterojunction photovoltaic devices." ACS Applied Materials & Interfaces . Paper under revision.

For the first paper, MVF carried out all of the experiments involved in the synthetic development of μ-oxo-(BsubPc) 2. BAK performed the CV experiment on μ-oxo-(BsubPc) 2 and initiated the work on the unsymmetric μ-oxo-(BsubPc) 2s by attempting to first synthesize F 12 BsubPc-O- BsubPc. I firmed up the synthesis of the unsymmetric derivatives, obtained their single crystals, and characterized all of the compounds within this chapter via NMR, optical absorption spectroscopy, photoluminescence spectroscopy, and cyclic voltammetry.

For the second paper, JSC and RKG performed all OPV device fabrication and testing and wrote the majority of the manuscript. I synthesized the organic materials, measured the film optical absorption spectra, and wrote the introduction section on μ-oxo-(BsubPc) 2. ET performed the UPS measurements.

The work was supervised by ZHL and TPB. All authors approved the manuscript.

Supplementary information for this chapter is found in Appendix B.

3.1 Abstract

A process for the gram scale synthesis of the oxygen bridged dimer of BsubPc, μ-oxo-(BsubPc) 2, is developed. During the development, it was found that a wide range of reaction pathways under diverse conditions lead to μ-oxo-(BsubPc) 2 formation. However, obtaining μ-oxo-(BsubPc) 2 as the main reaction product in appreciable yields and its subsequent isolation was extremely challenging. The best balance of purity, yield, and conversion was achieved with an equimolar reaction of HO-BsubPc with Br-BsubPc in the presence of K 3PO 4. This process was adapted towards the synthesis of three unsymmetric μ-oxo-(BsubPc) 2s. After synthesis, the solution-state properties of the unsymmetric dimers were investigated and compared to μ-oxo-(BsubPc) 2 and more broadly to other BsubPcs. The electronic properties of μ-oxo-(BsubPc) 2 were found to differ from its unsymmetric counterparts, but were found to be similar to halo-BsubPcs. Furthermore, the photophysical properties of these dimers differed greatly from all other known

BsubPcs. μ-Oxo-(BsubPc) 2 was tested and characterized in PHJ OPV devices, where it showed more promise as an electron donor than an electron acceptor and performed comparably with the prototypical Cl-BsubPc.

3.2 Introduction

The oxygen-bridged dimer of BsubPc, μ-oxo-(BsubPc) 2 (Figure 3.1), was serendipitously discovered when Cl-BsubPc was treated with NaOH in the presence of a phase transfer catalyst while attempting to synthesize HO-BsubPc. 1 It was immediately realized that this dimeric molecule had unique spectroscopic properties. For example, its optical absorption spectrum shows a significant hypsochromic/blue shift ( i.e. shift to higher energy) in the Q band compared to typical BsubPcs. Moreover the Q and B bands are similar in intensity, 1 whereas normal BsubPcs have Q bands that are more intense than the corresponding B band 2,3 . To further 1 illustrate its unusual spectroscopic behavior, the H NMR spectrum of μ-oxo-(BsubPc) 2 shows an up-field shift in the resonance signals for both terminal (2,3-position) and bay (1,4-position) hydrogens, where the effect is more pronounced on the latter. This effect is ascribed to the close proximity of these hydrogen atoms to the π-electron cloud of the second BsubPc macrocycle. 1

Figure 3.1. 2-D structure and ellipsoid plot (35% probability) of μ-oxo-(BsubPc) 2 (CCDC deposition number: 914863).7 Hydrogen atoms have been omitted for clarity. Colors: boron – yellow; nitrogen – blue; carbon – gray; oxygen – red.

Although μ-oxo-(BsubPc) 2 has been known for nearly twenty years now, its physical properties have been largely unexplored. This is likely attributed to the difficulties behind its synthesis as evident in the very few reports on its preparation. In 1996, Geyer et al . reported a synthetic method with a 10 mg (7.6%) yield starting from Cl-BsubPc and using NaOH as an oxygen source, dicyclohexano-18-crown-6 refluxing as a phase transfer catalyst and xylenes as the reaction solvent. The reaction pathway presumably involves the reaction of Cl-BsubPc with hydroxide (OH -) to yield HO-BsubPc in situ . HO-BsubPc either then self-condenses or reacts with another molecule of Cl-BsubPc yielding μ-oxo-(BsubPc) 2. This small sample of 10 mg was 1 enough to measure the characteristic absorption spectrum of μ-oxo-(BsubPc) 2. In 1999,

Kobayashi et al. described the synthesis of μ-oxo-(t-butyl -BsubPc) 2 and also described its absorption spectrum. The synthetic procedure involved the self-condensation of HO-t-butyl- BsubPc in nitrobenzene with some molecular sieves present and yielded 8.8 mg (4.7%). 4 No other synthesis was reported until a 2008 US patent application by Yamasaki, Mori, and Furuya which outlined two reaction pathways. The first is the self-condensation of HO-BsubPc to afford

the desired compound and water as a by-product. The second involves the reaction of a HO- BsubPc derivative with a halo-BsubPc derivative to produce a peripherally-substituted μ-oxo-

(BsubPc) 2 and a halo-acid as a byproduct. Each example was performed at a reasonable scale claiming to yield between 0.65 – 4.5 g of material.5 More recently and coincident with our interest in μ-oxo-(BsubPc) 2, Yamasaki & Mori published the material found in their patent application. Using the self-condensation reaction pathway, they reported the synthesis of μ-oxo-

(BsubPc) 2 in 34% yield and four peripherally substituted μ-oxo-(BsubPc) 2 compounds in yields that range from 2.3 to 28.4%. 6

Figure 3.2 . Absorption spectrum (pink) of μ-oxo-(BsubPc) 2 overlaid with the solar irradiance (black, sea level) spectrum. The red shaded area represents the area of highest solar irradiance (450-700 nm).

With respect to its properties, μ-oxo-(BsubPc) 2 has been found to have a higher solubility than its precursor HO-BsubPc or any halo-BsubPc in a broad range of organic solvents 6,7 and also found to be mobile under sublimation conditions, 8 despite its higher molecular mass. Both of these features are appealing for device integration purposes since either solution processing or vacuum

deposition techniques would be feasible. The solid state arrangement of the non-solvated, solvated, and hemi-hydrated μ-oxo-(BsubPc) 2 were found to exhibit high symmetry, close intermolecular interactions, and remarkably high crystal densities by comparison to other 7 BsubPcs. As mentioned earlier, μ-oxo-(BsubPc) 2 is blue-shifted in absorption compared to typical BsubPcs 1 and this is particularly favourable to us as its spectrum aligns better with the solar irradiance spectrum (Figure 3.2). Overall these results provided us with the impetus for further investigation of effective synthetic routes to μ-oxo-(BsubPc) 2.

In this chapter, we conduct a survey of the synthetic routes to μ-oxo-(BsubPc) 2 including those of Geyer et al. 1 and Yamasaki & Mori 6. Our assessment of the various routes is that the reaction of equal molar amounts of HO-BsubPc and Br-BsubPc in 1,2-dichlorobenzene in the presence of

K3PO 4 is the most effective method to reliably and reproducibly produce μ-oxo-(BsubPc) 2 in reasonable quantities. We also adapted this synthetic methodology towards the synthesis of three unsymmetric µ-oxo-(BsubPc) 2 derivatives. Each unsymmetric derivative was characterized for its solution-state photo- and electro-physical properties. Following this work, the use of μ-oxo-

(BsubPc) 2 as an electron donor material and as an electron acceptor material in PHJ OPVs is explored. In the former role, higher power conversion efficiencies (PCEs) were generated and were comparable with the prototypically studied Cl-BsubPc, thus opening a new class of BsubPc derived materials for future use in OPVs.

3.3 Results and Discussion Our first goal was to develop a practical synthetic method that yields reasonable quantities of doubly-train-sublimed μ-oxo-(BsubPc) 2. The criteria for practicability include the potential to carry out the reaction and subsequent work-up procedure using commonly available equipment in a research group involved in the synthesis of Pcs or organic electronic materials. The criteria for reasonable quantity was defined as a quantity having sufficient purity after being “doubly- sublimed” to enable the characterization of μ-oxo-(BsubPc) 2 as an organic electronic material within a device fabricated by vacuum deposition. 8 “Doubly-sublimed” is a standard purity threshold applied by our laboratory prior to electrical characterization, and means that once a sample is obtained that is pure by common analytical means ( e.g. HPLC, 1H NMR) it must then

be further purified by train sublimation 3 twice to ensure any trace impurities are removed. While atom and energy efficiency are typical priorities in process chemistry development, the priority and selection of the ‘best’ process herein was focused on obtaining train-sublimed pure samples of μ-oxo-(BsubPc) 2 at ‘reasonable’ yields above other process chemistry considerations.

Scheme 3.1 . Synthesis of μ-oxo-(BsubPc) 2. Reagents and conditions: (i) X = Cl, sodium hydroxide, dicyclohexano-18-crown-6, p-xylene, reflux; (ii) X = Br, water, pyridine, dimethyl sulfoxide, 60 °C; (iii) 1,2-dichlorobenzene, reflux; (iv) X = Br, 1,2-dichlorobenzene, water (trace), reflux; (v) X = Br, p-xylene (105 °C), nitrobenzene (200 °C), diphenyl ether (200 °C), or 1,2-dichlorobenzene (reflux); and (vi) X = Br, tripotassium phosphate, 1,2-dichlorobenzene, reflux.

We began our survey by attempting to replicate the first published synthetic method by Geyer et al .1 (Scheme 3.1, conditions (i)). This reported method involved the reaction of Cl-BsubPc with NaOH in p-xylene in the presence of a phase transfer catalyst, dicyclohexano-18-crown-6, under high dilution conditions. In our hands this process yielded no μ-oxo-(BsubPc) 2 and in fact HPLC analysis showed that the starting Cl-BsubPc remained unreacted (Table S3.1, Method 1.1). We then replaced Cl-BsubPc with Br-BsubPc which is known to be more reactive and still we observed no reaction (Table S3.1, Method 1.2). We then moved out of high dilution conditions by using ¼ the amount of p-xylene. In this case, the reaction did occur yielding 36% of μ-oxo-

(BsubPc) 2 and 64% of an unknown BsubPc compound, which was neither Br-BsubPc or HO-

BsubPc (Table S3.1, Method 1.3). Finally, we reduced the molar equivalents of NaOH used in the reaction to 1.0 (relative to Br-BsubPc), which resulted in a final reaction composition of ~2%

μ-oxo-(BsubPc) 2, 15% HO-BsubPc, 60% Br-BsubPc and 23% of an unknown BsubPc compound

(Table S3.1, Method 1.4). None of these variations yielded more than 36% μ-oxo-(BsubPc) 2 and therefore in our opinion none fit the criteria outlined above.

Turning our attention to the processes described by Yamasaki & Mori, 6 we began with the self- condensation of HO-BsubPc in 1,2-dichlorobenzene (Scheme 3.1, conditions (iii)). In our hands the result was a mixture of compounds as follows: 35% of μ-oxo-(BsubPc) 2, 24% of an unknown

BsubPc compound with a HPLC retention time (R t) of 3.4 min, 35% of HO-BsubPc and 6.1% of additional unknown BsubPc compounds (Table S3.2, Method 2.1). Despite our best attempts, including the workup procedure described by Yamasaki & Mori, we could not separate μ-oxo-

(BsubPc) 2 from the other BsubPc compounds. Especially difficult was the removal of the unknown BsubPc compound with R t = 3.4 min.

In an attempt to generate HO-BsubPc in situ and promote its self-condensation we refluxed Br- BsubPc in 1,2-dichlorobenzene with the periodic addition of drops of water to keep the reaction ‘wet’ (Scheme 3.1, conditions (iv)). This process actually produced a final reaction mixture containing 61% μ-oxo-(BsubPc) 2, however another unknown BsubPc with R t = 2.0 min was formed (35%). Small amounts of HO-BsubPc and other unknown BsubPcs were also present (2.2% and 1.4%, respectively, Table S3.2, Method 2.2). However again, we could not separate μ- oxo-(BsubPc) 2 from the other components especially from the unknown BsubPc with R t = 2.0 min.

In the introduction of the patent application by Yamasaki, Mori, and Furuya, 5 the inventors alluded to a method for the formation of μ-oxo-(BsubPc) 2 involving the reaction of HO-BsubPc with X-BsubPc (X = F, Cl, Br, I; albeit I-BsubPc is unknown and F-BsubPc is known to be non- reactive) in the presence of a base in a variety of solvents. Bases mentioned include hydrides, hydroxides and carbonates. However specific examples were only given in the patent application for the hydride case. Given that BsubPcs are known to be sensitive to bases we initially tried the

reaction of HO-BsubPc with Br-BsubPc without base (Scheme 3.1, conditions (v)). We selected the solvents p-xylene, nitrobenzene, diphenyl ether and 1,2-dichlorobenzene (Table S3.3, Method 3.1 – 3.6). In some cases we achieved conversions greater than 80%, however the unknown BsubPc compound with R t = 2.0 min persisted and we were unable to remove even small amounts (~10%) of it from the crude products to give pure μ-oxo-(BsubPc) 2.

After achieving relatively high conversions by the equimolar reaction of HO-BsubPc with Br- BsubPc, we then considered the addition of base to the process with the hypothesis that the persistence of the produced haloacid might lead to the formation of the persistent BsubPc impurity with R t = 2.0 min. Regarding the suggested addition of base by Yamasaki, Mori, and Furuya 5 we ruled out the use of hydrides for safety reasons. We felt the sensitivity of BsubPcs to bases negated the possibility of using hydroxides. We also ruled out the use of carbonates or bicarbonates due to the production of carbon dioxide gas on reaction with acids which might/could lead to excessive bubbling of the reaction mixture. We therefore decided to explore the use of tripotassium phosphate (K 3PO 4) as a scavenger of HBr (Scheme 3.1, conditions (vi)).

While not commonly referred to as a base, K 3PO 4 does have the ability to react with acids for example: HBr + K 3PO 4 ‰ KBr + HK 2PO 4. Of course K 3PO 4 is not soluble in common organic solvents so under this scenario it will be the surface of the K 3PO 4 that would react with HBr.

Therefore, we began by dispersing as much K 3PO 4 as possible in 1,2-dichlorobenzene (80 wt %

K3PO 4, Table S3.4, Method 4.1) and conversion to μ-oxo-(BsubPc) 2 was >96% with trace amounts of HO-BsubPc remaining. However, the absolute yield was low and was attributed to the excessively purple-colored solid filtered from the reaction, which undoubtedly contained adsorbed/absorbed BsubPc compounds on its surface. We were unable to remove the adsorbed/absorbed BsubPc compounds from the surface of the solid. We then reduced the amount of K 3PO 4 and subsequently used dried K 3PO 4 and also found high conversion (~97% and

~100% respectively, Table S3.4, Method 4.2 and 4.3). Finally, we tried using ‘wet’ K 3PO 4 to promote the hydrolysis of Br-BsubPc in situ to HO-BsubPc but found no conversion to μ-oxo-

(BsubPc) 2 (Table S3.4, Method 4.4). During these three reactions (Method 4.1-4.3), we noted qualitatively that although the conversion appeared to be >97% the intensity of the color of the solution seemed to decrease with time indicating decomposition of the BsubPc chromophore.

This was confirmed as only 8% of μ-oxo-(BsubPc) 2 was isolated from reaction 4.2. We then repeated 4.3 and 4.4 except the reaction progress was monitored as a function of time (Table S3.4, Method 4.5 and 4.6). Samples were taken of approximately the same volume, diluted with approximately the same amount of eluent and injected into the HPLC. The intensity of the μ- oxo-(BsubPc) 2 peak was observed to decrease over time. Given the lack of precise quantification the results are not provided, however it was found that a reaction time of 1 hour gave the best balance between conversion and isolated yield. Complete conversions were not achieved when the reaction was stopped after 1 hour, rather conversions of 67% and 78% were observed. Oddly enough residual HO-BsubPc was detected, perhaps indicating our conditions were not absolutely anhydrous.

Throughout these trials a variety of purification strategies were attempted. For example, removal of the reaction solvent followed by a Soxhlet extraction with toluene was tried. The work-up steps reported by Yamasaki et al. ,6 which involved dispersing the crude solid in N,N - dimethylformamide (DMF) followed by filtration and washing sequentially with DMF and water, were also tried. We also tried Kauffman column chromatography 3 using alumina as the adsorbent and toluene as the eluent. After 2 days of elution, a purple band likely corresponding to μ-oxo-(BsubPc) 2 was visible but did not elude out of the column. Kauffman column chromatography was repeated with dichloromethane (DCM) as the eluent and this turned out to be successful. We settled on a combination of a Soxhlet extraction with toluene, followed by a Kauffman column with DCM and finally a train sublimation 3 at 450 °C (measured external temperature) to achieve a pure sample of μ-oxo-(BsubPc) 2 at high purity. Typical yields were 53- 59% after Kauffman column chromatography and 50% following sublimation, resulting in an overall yield of 27-30 %. A final optimized synthetic method is detailed in method 4.9 in the ESI (Table S3.4).

Scheme 3.2 . Synthesis of F 12 BsubPc-O-BsubPc, Cl 6BsubPc-O-BsubPc, and Cl 12 BsubPc-O- BsubPc. Reagents and conditions: (i) tripotassium phosphate, 1,2-dichlorobenzene, reflux.

The feasibility and success of the K 3PO 4-mediated bimolecular condensation reaction for the formation of µ-oxo-(BsubPc) 2 motivated us to adapt this methodology towards the synthesis of unsymmetric µ-oxo-(BsubPc) 2 compounds, where the two BsubPc moieties are not identical. To the best of our knowledge, there is no precedent of an unsymmetric µ-oxo-(BsubPc) 2 compound in the literature. Using our K 3PO 4-mediated procedure, we synthesized three unsymmetric µ- oxo-(BsubPc) 2 compounds - F12 BsubPc-O-BsubPc, Cl 6BsubPc-O-BsubPc, and Cl 12 BsubPc-O- BsubPc – in 36%, 40%, and 43% yield, respectively. Each was made by reacting equimolar amounts of HO-BsubPc with Br-F12 BsubPc, Cl-Cl 6BsubPc, or Cl-Cl 12 BsubPc, respectively, in the presence of K 3PO 4 in 1,2-dichlorobenzene at 180 ˚C (Scheme 3.2). The reaction progress was monitored by HPLC for the complete consumption of the starting BsubPc materials, which normally took between 17 and 19 hours. Unlike the µ-oxo-(BsubPc) 2 synthesis, the color intensity of each solution for the three reactions did not decrease over time; these reactions were not time-sensitive at least for the first 17-19 hours. It is worth noting that for all three reactions the halo-BsubPcs (Br-F12 BsubPc, Cl-Cl 6BsubPc, and Cl-Cl 12 BsubPc) were fully consumed while residual amount of HO-BsubPc were present at the end of each reaction. Moreover, traces of µ- oxo-(BsubPc) 2 were detected via HPLC at the end of each reaction. Single crystals of F 12 BsubPc- O-BsubPc solvated with heptane (Figure 3.3a, CCDC deposition number: 1018494) and of

Cl 6BsubPc-O-BsubPc (Figure 3.3b, CCDC deposition number: 1018458) were successfully

grown by vapor diffusion of heptane into a DCM solution and by slow evaporation from DCM solution, respectively. X-ray diffraction analysis confirmed the molecular structures of the two compounds (Figure 3.3).

(a) (b)

Figure 3.3 . Ellipsoid plot (50% probability) showing the structure and atom numbering scheme of (a) F 12 BsubPc-O-BsubPc ·C7H16 (CCDC deposition number: 1018494) and (b) Cl 6BsubPc-O- BsubPc (CCDC deposition number: 1018458). Hydrogen atoms and solvent inclusions have been omitted for clarity. Colors: boron - yellow; nitrogen - blue; carbon - white; oxygen - red; fluorine - magenta; chlorine - green.

UV-Vis absorption spectra of µ-oxo-(BsubPc) 2, F12 BsubPc-O-BsubPc, Cl 6BsubPc-O-BsubPc, and Cl 12 BsubPc-O-BsubPc were acquired in toluene solutions at room temperature (Figure 3.4). All compounds were purified sequentially by column chromatography and train sublimation prior to performing these measurements. As already reported, the photophysical properties of µ- 1 oxo-(BsubPc) 2 is unique among BsubPcs. For example, it has a λ max of absorption (λ max = 533 nm in toluene) 1 that is significantly blue-shifted by comparison against typical monomeric BsubPcs, which are generally above 560 nm (Table S3.5). Furthermore, its absorption spectrum is broader compared to typical monomeric BsubPcs and this is attributed to the π-π interactions 6 of the two BsubPc moieties of the m -oxo-dimer (Figure S3.11a). Like µ-oxo-(BsubPc) 2 but to a

lesser degree, the three unsymmetric µ-oxo-(BsubPc)2 compounds have λ max values that are blue- shifted relative to monomeric BsubPcs; F 12 BsubPc-O-BsubPc, Cl 6BsubPc-O-BsubPc, and

Cl 12 BsubPc-O-BsubPc have a λ max of absorption at 547, 541 and 558 nm, respectively (Table

3.1). In addition to sharing this blue-shifting feature with µ-oxo-(BsubPc) 2, the absorption spectra of the three unsymmetric µ-oxo-(BsubPc) 2 compounds are broader in comparison to monomeric BsubPcs (Figure S3.12a), indicating that these compounds also experience π-π interaction between the BsubPc macrocycles in a similar manner as µ-oxo-(BsubPc) 2.

Figure 3.4 . Absorption (blue) and photoluminescence (red) spectra of (a) μ-oxo-(BsubPc) 2, (b)

F12 BsubPc-O-BsubPc, (c) Cl 6BsubPc-O-BsubPc, and (d) Cl 12 BsubPc-O-BsubPc in toluene solutions at room temperature.

The photoluminescence (PL) spectra of μ-oxo-(BsubPc) 2, F 12 BsubPc-O-BsubPc, Cl 6BsubPc-O-

BsubPc, and Cl 12 BsubPc-O-BsubPc were also acquired in toluene solutions at room temperature

(Figure 3.4). Although the absorption properties of μ-oxo-(BsubPc) 2 is known, its fluorescence properties have not been explored. The PL spectrum of μ-oxo-(BsubPc) 2 at an excitation wavelength (λ ex ) of 532 nm showed a minor shoulder peak at 576 nm and a primary peak at 636 nm (Table 3.1). The latter emission peak is heavily red-shifted compared to the emission spectra of typical monomeric BsubPcs (568 to 572 nm for halo-BsubPcs 9, Figure S3.11b), demonstrating once again its uniqueness among BsubPcs. Similar to μ-oxo-(BsubPc) 2, the PL spectra of the three unsymmetric μ-oxo-(BsubPc) 2 compounds obtained at a λ ex of 530 nm each showed two emission peaks (Figure 3.4 and Figure S3.12b). F 12 BsubPc-O-BsubPc has a more intensive emission peak at 575 nm and a less intensive peak at 729 nm, Cl 6BsubPc-O-BsubPc has a less intensive peak at 575 nm and a more intensive peak at 672 nm, while Cl 12 BsubPc-O-BsubPc has a less intensive peak at 585 nm and a more intensive peak at 739 nm (Table 3.1).

The presence of two emission peaks is uncharacteristic of BsubPcs in the solution state. This dual emission behavior, however, has previously been observed for pentafluorophenoxy-BsubPc

(F 5PhO-BsubPc) in solvent mixtures consisting of DMF (a good solvent) and water (a poor solvent) in a 1:128 v/v ratio or lower. 11 Under these conditions, a secondary emission peak at 706 nm emerges and its intensity increases relative to the primary emission peak at the 576-593 nm range as the ratio of DMF/H 2O decreases. This phenomenon is caused by molecular aggregation, an effect that becomes more pronounced with a higher concentration of water. 11 An aggregation effect and thus a secondary emission peak was not anticipated for the three unsymmetric μ-oxo-

(BsubPc) 2 compounds considering that toluene, a normally good solvent for BsubPcs, is used to dissolve these compounds and also that the analyses were done under dilute conditions (solution- state absorbance < 0.05).

Table 3.1 . Photophysical and electronic properties of µ-oxo-(BsubPc) 2, F12 BsubPc-O-BsubPc,

Cl 6BsubPc-O-BsubPc, and Cl 12 BsubPc-O-BsubPc.

a a a,b a 1 2 c 1 2 c Compound λmax,abs λmax,ems ΦPL Stokes Shift E ox | E ox (V) E red | E red (V) (nm) (nm) (%) (nm) d d µ-oxo-(BsubPc) 2 532 576, 636 2.0 44, 104 +0.95 -1.03 d e e F12 BsubPc-O-BsubPc 547 575, 729 0.3 28, 182 +1.15 -0.60 | -0.89 d d d Cl 6BsubPc-O-BsubPc 541 575, 672 1.1 34, 131 +1.08 -0.55 | -0.84 d d e e Cl 12 BsubPc-O-BsubPc 558 585, 745 1.9 27, 187 +1.13 | +1.26 -0.60 | -0.93 Cl-BsubPc 565 f 571 f 73 f 6f +1.04 g -1.05 g Br-BsubPc 566 f 572 f 10 f 6f +1.03 h -1.06 h a b In toluene solution. Relative to a PhO-F12 BsubPc standard using an excitation wavelength of 533 nm for µ-oxo- c (BsubPc) 2 and 530 nm for the three unsymmetric µ-oxo-(BsubPc)2 compounds. In degassed DCM solution relative to Ag/AgCl. d Peak potential. e Half-wave potential. f Data taken from Fulford et al .6 g Data taken from del Rey et al .30 h Data taken from Kasuga et al .10

To examine whether the emission peak at the longer wavelength - 636, 729, 672, and 739 nm - for μ-oxo-(BsubPc) 2, F12 BsubPc-O-BsubPc, Cl 6BsubPc-O-BsubPc, and Cl 12 BsubPc-O-BsubPc, respectively, was actually due to an aggregation effect, PL spectra (concentration = 2.30 x 10 -6

M) were acquired in a series of DMF/H 2O mixtures ranging from pure DMF to 1:200 ( v/v )

DMF/H 2O (Figure 3.5). For μ-oxo-(BsubPc) 2 in pure DMF, two emission peaks at 572 and 647 nm of nearly equal intensities were observed. Strangely between 10:1 and 1:1 DMF/H 2O mixture, the shorter wavelength emission (574-576 nm) was only seen. The longer wavelength emission (661-662 nm) began to appear at 1:10 DMF/H2O and intensified relative to the shorter wavelength emission as the ratio dropped towards 1:200 DMF/H 2O. For all three unsymmetric μ- oxo-(BsubPc) 2 in pure DMF, a single emission peak at the shorter wavelength was observed. The longer wavelength emission peak began to emerge at 1:10, 1:1, and 1:1 DMF/H 2O for

F12 BsubPc-O-BsubPc, Cl 6BsubPc-O-BsubPc, and Cl 12 BsubPc-O-BsubPc, respectively, and increased in intensity relative to the shorter wavelength emission peak as the DMF/H 2O ratio dropped towards 1:200. Moreover, the longer wavelength emission peak became the strongest/primary peak for both Cl 6BsubPc-O-BsubPc (at 1:50 DMF/H 2O and below) and

Cl 12 BsubPc-O-BsubPc (at 1:10 DMF/H 2O and below). It is also interesting to note that for only

Cl 12 BsubPc-O-BsubPc, a third emission peak (635-640 nm) was formed in DMF/H 2O ratio of

1:10 and below. Overall, these results/trends are consistent with those reported for F 5PhO- BsubPc 11 as mentioned earlier, suggesting that the PL emission at the longer wavelength - 636,

729, 672, and 739 nm for μ-oxo-(BsubPc) 2, F12 BsubPc-O-BsubPc, Cl 6BsubPc-O-BsubPc, and

Cl 12 BsubPc-O-BsubPc, respectively - in toluene solutions at room temperature is likely attributable to an aggregation-caused emission. An examination of the rate at which the longer wavelength emission peak increased relative to the shorter wavelength emission peak as the

DMF/H 2O ratio approached 1:200, suggests that the order in the degree of aggregation is

Cl 12 BsubPc-O-BsubPc > Cl 6BsubPc-O-BsubPc > μ-oxo-(BsubPc) 2 > F 12 BsubPc-O-BsubPc.

It is also very apparent from the PL spectra acquired from the series of DMF/H 2O mixture that the emission intensities weakened with increasing volume fraction of H 2O (Figure 3.5). In addition to this, small shifts in the position of the PL spectra were generally observed with the exception of Cl 12 BsubPc-O-BsubPc. A bathochromic shift of 4-5 nm and 2-3 nm were observed for μ-oxo-(BsubPc) 2 and F 12 BsubPc-O-BsubPc, respectively, while a hypsochromic shift of 6-8 nm and 45-46 nm were observed for Cl 6BsubPc-O-BsubPc and Cl 12 BsubPc-O-BsubPc, respectively, from pure DMF to 1:200 DMF/H 2O. Based on the observed light emission quenching effect and the shifts, although not very pronounced, in the PL spectra with increasing volume fraction of H 2O, the four compounds under study especially Cl 12 BsubPc-O-BsubPc display solvatochromic fluorescence ( i.e. solvent-dependent emission).

UV-Vis absorption spectra were also acquired in the same series of DMF/H 2O mixtures as used above to determine if the four compounds under study are also solvatochromic in absorption (Figure 3.5). There were generally small changes to the position of the spectra going from pure

DMF to 1:200 DMF/H 2O. A bathochromic shift of 6-8 nm was observed for μ-oxo-(BsubPc) 2, while a hypsochromic shift of 13-14 nm and 4-6 nm were observed for F 12 BsubPc-O-BsubPc and

Cl 6BsubPc-O-BsubPc, respectively. For Cl 12 BsubPc-O-BsubPc, a trend was not very clear as the spectra blue shifted initially before it red shifted. However, by just looking at the λ max values in pure DMF (560 nm) and in 1:200 DMF/H 2O (552 nm), a general hypsochromic shift is observed.

The absorbance intensities were also found to decrease with increasing volume fraction of H 2O. Given this observed effect combined with the shifts, although not very pronounced, in the absorption spectra with increasing volume fraction of H 2O, the four compounds under study are solvatochromic in absorption ( i.e. solvent-dependent absorption).

It is interesting to note that in toluene the characteristics are different. For μ-oxo-(BsubPc) 2 and

Cl 6BsubPc-O-BsubPc in toluene at room temperature the primary/strongest emission peak is actually the aggregate-induced emission peak at 636 and 672 nm, respectively (Figure 3.4). For

F12 BsubPc-O-BsubPc and Cl 12 BsubPc-O-BsubPc the primary emission peak is the one at 575 and 585 nm (shorter wavelength emission), respectively. This suggests that μ-oxo-(BsubPc) 2 and

Cl 6BsubPc-O-BsubPc are more prone to aggregation compared to F 12 BsubPc-O-BsubPc and

Cl 12 BsubPc-O-BsubPc. Since the intensity of the shorter wavelength emission peak is nearly the same for Cl 6BsubPc-O-BsubPc (~16%) than μ-oxo-(BsubPc) 2 (~15%), this suggests that both of these compounds aggregate to the same degree. Likewise, since the intensity of the longer wavelength emission peak is higher for F 12 BsubPc-O-BsubPc (~60%) than Cl 12 BsubPc-O-

BsubPc (~55%), this suggests that F 12 BsubPc-O-BsubPc aggregates slightly more than

Cl 12 BsubPc-O-BsubPc. Overall, the trend in the degree of aggregation observed here is μ-oxo-

(BsubPc) 2 ~ Cl 6BsubPc-O-BsubPc > F 12 BsubPc-O-BsubPc > Cl 12 BsubPc-O-BsubPc. This contradicts the trend found from the DMF/H 2O-aggregation experiments (Cl 12 BsubPc-O-BsubPc

> Cl 6BsubPc-O-BsubPc > μ-oxo-(BsubPc) 2 > F 12 BsubPc-O-BsubPc), signifying that solvent effects play a major factor in the aggregation of these BsubPc dimers.

As mentioned earlier, the PL emission spectrum of μ-oxo-(BsubPc) 2 is considerably red-shifted relative to its absorption spectrum, leading to a Stokes shift of 44 nm for the shorter wavelength emission and 104 nm for the longer wavelength emission (Table 3.1). This is not characteristic of monomeric BsubPcs, which normally have small Stokes shifts ( e.g. 6 nm for halo-BsubPcs 9, 20- 29 nm range for fluorinated PhO-BsubPcs 3). The large Stokes shift indicates that μ-oxo-

(BsubPc) 2 undergoes significant geometric relaxation ( i.e. nonradiative) following photoexcitation; the ground state structure is significantly different from the excited state structure. For the unsymmetric μ-oxo-(BsubPc) 2 compounds, the smaller Stokes shift is in line with fluorinated phenoxy-BsubPcs (20-29 nm) 3 while the larger Stokes shift surpasses that of the

µ-oxo-(BsubPc) 2 (Table 3.1). These large Stokes shifts signify that these unsymmetric μ-oxo-

(BsubPc) 2 compounds undergo even greater structural relaxation in the excited state compared to

μ-oxo-(BsubPc) 2.

Figure 3.5 . UV-vis absorption (solid) and PL (λ exc = 530 nm, dash) spectra (concentration = 2.30 -6 x 10 M) of (a) µ-oxo-(BsubPc) 2, (b) F12 BsubPc-O-BsubPc, (c) Cl 6BsubPc-O-BsubPc, and (d)

Cl 12 BsubPc-O-BsubPc in DMF (violet), 10:1 (blue), 1:1 (green), 1:10 (orange), 1:50 (red), and

1:200 (black) DMF/H 2O (v/v ).

Fluorescence quantum yields (Φ) were measured for the four compounds under study and they were found to be in the range of 0.3-2.0% relative to a standard of phenoxy-dodecafluoro-

BsubPc (PhO-F12 BsubPc, Table 3.1, and Eq S3.1). These low Φ values are not characteristic of BsubPcs, which normally have much higher quantum yields (Φ > 40%). 2,3,12 The low Φs are likely related to the structure of the dimers, where the two BsubPc chromophores are in close proximity to one another. Therefore even under very dilute conditions (solution-state absorbance

< 0.05), interactions between the two BsubPc chromophores cannot be neglected and may lead to self-quenching of the fluorescence.

(a) (b)

Figure 3.6 . Cyclic voltammograms of (a) μ-oxo-(BsubPc) 2, (b) F12 BsubPc-O-BsubPc, (c)

Cl 6BsubPc-O-BsubPc, and (d) Cl 12 BsubPc-O-BsubPc in DCM with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference

(E 1/2,red = -0.012 V vs. Ag/AgCl) at room temperature.

The electrochemical properties of μ-oxo-(BsubPc) 2, F 12 BsubPc-O-BsubPc, Cl 6BsubPc-O-

BsubPc, and Cl 12 BsubPc-O-BsubPc were analyzed via cyclic voltammetry in degassed DCM solution containing 0.1 M tetrabutylammonium perchlorate at room temperature (Figure 3.6 and

Table 3.1). All potentials were corrected to the half-wave reduction potential (E 1/2,red ) of decamethylferrocene, which was previously reported to be -0.012 V vs. Ag/AgCl. 13 All compounds underwent an irreversible oxidation while Cl 12 BsubPc-O-BsubPc underwent a

second irreversible oxidation. The irreversibility in the oxidation regime is a property that is commonly observed for BsubPcs. 2,3 Therefore since all oxidation peaks were irreversible, oxidation potentials (E ox ) were reported based on peak potentials as opposed to half-wave potentials. For the reduction regime, μ-oxo-(BsubPc) 2 was found to undergo a single irreversible reductive process while the unsymmetric μ-oxo-(BsubPc) 2 compounds were found to undergo two reductive processes. Both reduction events were irreversible for Cl 6BsubPc-O-BsubPc and reversible for both F 12 BsubPc-O-BsubPc and Cl 12 BsubPc-O-BsubPc. Reduction potentials (E red ) were reported based on half-wave potentials for reversible reduction peaks and on peak potentials for irreversible reduction peaks.

1 1 The first oxidation (E ox ) and reduction (E red ) potentials of μ-oxo-(BsubPc) 2 were found to be similar to both Cl-BsubPc and Br-BsubPc, but were found to differ from the unsymmetric μ-oxo- 1 (BsubPc) 2 compounds (Table 3.1). When comparing the E ox across the four compounds under study, μ-oxo-(BsubPc) 2 (+0.95 V) was the easiest to oxidize, followed by Cl 6BsubPc-O-BsubPc

(+1.08 V), Cl 12 BsubPc-O-BsubPc (+1.13 V), and lastly F 12 BsubPc-O-BsubPc (+1.15 V). Based 1 on this trend, peripheral halogenation on the μ-oxo-(BsubPc) 2 scaffold raises the E ox (i.e. harder 1 to oxidize); having chlorine or fluorine in place of the peripheral hydrogen increases the E ox . 1 From this trend, it also appears that the more peripheral halogens present, the higher the E ox . 1 When comparing the E red across the four compounds, Cl 6BsubPc-O-BsubPc (-0.55 V) was the easiest to reduce, followed by both Cl 12 BsubPc-O-BsubPc and F 12 BsubPc-O-BsubPc (-0.60 V), and lastly μ-oxo-(BsubPc) 2 (-1.03 V). Based on this result, a clear trend is not observed other 1 than the fact that peripheral halogenation lowers the E red (i.e. easier to reduce); having chlorine 1 or fluorine in place of the pheripheral hydrogen decreases the E red . Overall, peripheral 1 1 halogenation on the μ-oxo-(BsubPc) 2 scaffold increases the E ox and decreases the E red but its 1 effect is more severe for the E red . A similar effect was reported for the peripherally iodinated Cl- BsubPc (+1.13 V and -0.92 V), where it was found that its reduction potential was lower ( i.e. less negative) and its oxidation potential was higher ( i.e. more positive) relative to Cl-BsubPc (+1.04 V and -1.05 V). 14

μ-Oxo-(BsubPc) 2 (Figure 3.7) was integrated into PHJ OPVs as the electron acceptor layer and as the electron donor layer to measure its device performance characteristics and assess its potential to serve each role (Table 3.2). For comparison purposes, the prototypically studied Cl- BsubPc (Figure 3.7) was also tested under the same condition. OPV devices of the structure indium tin oxide (ITO)/poly(3,4-ethylenedioxythiophene) poly(styrenesulfonate) (PEDOT:PSS)/sexithiophene (α-6T)/acceptor/2,2’,2’’-(1,3,5-benzinetriyl)-tris(1-phenyl-1-H- benzimidazole) (TPBi)/silver (Ag) and ITO/PEDOT:PSS/molybdenum oxide

(MoO x)/donor/C 60/70/TPBi/Ag were fabricated for acceptor-based and donor-based cells, respectively (Figure 3.8). These device configurations were selected for this study given their 15-18 literature precedent. The incorporation of a MoO x layer in the latter configuration serves as a hole extraction layer that have proven to be effective in improving the FF and V OC for devices using BsubPc or Pc as the donor. 19-22

Figure 3.7 . Chemical structure of donor and acceptor materials used in this study. From left to right: Cl-BsubPc, μ-oxo-(BsubPc) 2, α-6T, C 60 , and C 70 .

(a) (b)

Figure 3.8 . Schematic of the device structure for testing μ-oxo-(BsubPc) 2 and Cl-BsubPc as an (a) electron donor and (b) electron acceptor.

Table 3.2 . Average device parameters of PHJ OPVs of the following configuration:

ITO/PEDOT:PSS/α-6T/acceptor/TPBi/Ag or ITO/PEDOT:PSS/MoO x/donor/C 60/70 /TPBi/Ag. The standard deviation is indicated in brackets next to each parameter.

a b -2 c d Donor Acceptor VOC (V) JSC (mA·cm ) FF PCE (%) Number of Devices α-6T Cl-BsubPc 1.08 (0.025) 5.45 (0.540) 0.55 (0.063) 3.20 (0.437) 17 α-6T μ-oxo- 0.98 (0.179) 1.57 (0.192) 0.35 (0.051) 0.54 (0.137) 18 (BsubPc) 2 Cl-BsubPc C60 1.06 (0.012) 4.35 (0.467) 0.52 (0.047) 2.40 (0.204) 12 μ-oxo- C60 0.90 (0.003) 3.87 (0.129) 0.59 (0.013) 2.05 (0.071) 7 (BsubPc) 2 Cl-BsubPc C70 1.08 (0.007) 4.62 (0.114) 0.52 (0.012) 2.58 (0.087) 12 μ-oxo- C70 0.91 (0.005) 6.32 (0.136) 0.48 (0.019) 2.74 (0.149) 15 (BsubPc) 2 a Open-circuit voltage. b Short-circuit current density. c Fill factor. d Power conversion efficiency.

As an acceptor when paired with α-6T (Figure 3.7), μ-oxo-(BsubPc) 2 perform very poorly producing OPV cells with an average PCE of 0.54% versus the 3.20% observed for Cl-BsubPc- based cells. The majority of the performance loss is attributed to a ~70% reduction in J SC with a moderate loss in the FF. As a donor when paired with C 60 (Figure 3.7), μ-oxo-(BsubPc) 2 perform similarly with an average PCE of 2.05% compared to 2.40% for Cl-BsubPc. When paired with the more π-extended fullerene derivative C 70 (Figure 3.7), μ-oxo-(BsubPc) 2-based devices

actually perform better with an average PCE of 2.74% compared to 2.58%. Based on these results, it is clear that μ-oxo-(BsubPc) 2 shows more promise as a donor material than as an acceptor material in PHJ OPVs.

3.4 Conclusions In this chapter we presented an optimized procedure for the synthesis and isolation of μ-oxo-

(BsubPc) 2 in a sufficient yield and purity. It was discovered that, while μ-oxo-(BsubPc) 2 can be created through a broad range of reaction pathways and under diverse conditions, its formation as the major product and its subsequent isolation is extremely challenging. The best compromise between the competing goals of purity, yield and extent of conversion was found to be an equimolar addition reaction of HO-BsubPc and Br-BsubPc in the presence of K 3PO 4 in 1,2- dichlorobenzene at 180 °C for 1 hour. Following the development of this methodology, we adapted it towards the synthesis of three unsymmetric µ-oxo-(BsubPc) 2 compounds - F12 BsubPc-

O-BsubPc, Cl 6BsubPc-O-BsubPc, and Cl 12 BsubPc-O-BsubPc. UV-vis absorption and fluorescence spectroscopy revealed that the μ-oxo-(BsubPc) 2 and the three unsymmetric μ-oxo-

(BsubPc) 2 compounds had absorption and fluorescence properties that differ greatly from typical monomeric BsubPcs. Moreover, very low fluorescence quantum yields were measured for the four compounds under study and these values are uncharacteristic of typical monomeric BsubPcs. A self-quenching effect between the BsubPc chromophores is likely occurring following photoexcitation as a result of the close proximity of the BsubPc moieties within the dimeric environment. The electrochemical properties of µ-oxo-(BsubPc) 2 were found to be similar to those of the halo-BsubPcs, while the unsymmetric µ-oxo-(BsubPc) 2 compounds were found to have higher oxidation potentials and much lower reduction potentials compared to µ- oxo-(BsubPc) 2. The potential application of μ-oxo-(BsubPc) 2 as the electron donor and as the electron acceptor in PHJ OPVs was examined. As an acceptor, μ-oxo-(BsubPc) 2 performed very poorly when paired with α-6T. When employed as a donor paired with either C 60 or C 70 , PCEs were much higher and comparable to the analogous Cl-BsubPc devices. Given the promising results as an electron donor, this work overall demonstrates that a BsubPc derivative other than the prototypical Cl-BsubPc can be and is worthy of being studied in OPVs.

3.5 References 1. Geyer, M.; Plenzig, F.; Rauschnabel, J.; Hanack, M.; Del Rey, B.; Sastre, A.; Torres, T. Synthesis 1996 , 1139-1151. 2. Gonzalez-Rodriguez, D.; Torres, T.; Guldi, D. M.; Rivera, J.; Herranz, M. A.; Echegoyen, L. Journal of the American Chemical Society 2004 , 126 , 6301-6313. 3. Morse, G. E.; Helander, M. G.; Maka, J. F.; Lu, Z.-H.; Bender, T. P. ACS Applied Material & Interfaces 2010 , 2, 1934-1944. 4. Kobayashi, N.; Ishizaki, T.; Ishii, K.; Konami, H. Journal of the American Chemical Society 1999 , 121 , 9096-9110. 5. Mori, T.; Furuya, F.; Yamasaki, Y. Optical layer including mu-oxo-bridged boron- subphthalocyanine dimer. U.S. Patent Application US 2008/0210128, September 4, 2008. 6. Yamasaki, Y.; Mori, T. Bulletin of the Chemical Society of Japan 2011 , 84 , 1208-1214. 7. Fulford, M. V.; Lough, A. J.; Bender, T. P. Acta Crystallographica Section B-Structural Science 2012 , 68 , 636-645. 8. Castrucci, J. S.; Dang, J. D.; Kamino, B. A.; Campbell, A.; Pitts, D.; Lu, Z.-H.; Bender, T. P. Vacuum 2014 , 109 , 26-33. 9. Fulford, M. V.; Jaidka, D.; Paton, A. S.; Morse, G. E.; Brisson, E. R. L.; Lough, A. J.; Bender, T. P. Journal of Chemical & Engineering Data 2012 , 57 , 2756-2765. 10. Kasuga, K.; Idehara, T.; Handa, M.; Ueda, Y.; Fujiwara, T.; Isa, K. Bulletin of the Chemical Society of Japan 1996 , 69 , 2559-2563. 11. Helander, M. G.; Morse, G. E.; Qiu, J.; Castrucci, J. S.; Bender, T. P.; Lu, Z.-H. ACS Applied Materials & Interfaces 2010, 2, 3147-3152. 12. Morse, G. E.; Castrucci, J. S.; Helander, M. G.; Lu, Z. H.; Bender, T. P. ACS Applied Materials & Interfaces 2011 , 3, 3538-3544. 13. Noviandri, I.; Brown, K. N.; Fleming, D. S.; Gulyas, P. T.; Lay, P. A.; Masters, A. F.; Phillips, L. Journal of Physical Chemistry B 1999 , 103 , 6713-6722. 14. Del Rey, B.; Keller, U.; Torres, T.; Rojo, G.; Agullo-Lopez, F.; Nonell, S.; Marti, C.; Brasselet, S.; Ledoux, I.; Zyss, J. Journal of the American Chemical Society 1998 , 120 , 12808-12817.

15. Mutolo, K. L.; Mayo, E. I.; Rand, B. P.; Forrest, S. R.; Thompson, M. E. Journal of the American Chemical Society 2006 , 128 , 8108-8109. 16. Sullivan, P.; Duraud, A.; Hancox, I.; Beaumont, N.; Mirri, G.; Tucker, J. H. R.; Hatton, R. A.; Shipman, M.; Jones, T. S. Advanced Energy Materials 2011 , 1, 352-355. 17. Cnops, K.; Rand, B. P.; Cheyns, D.; Verreet, B.; Empl, M. A.; Heremans, P. Nature Communications 2014 , 5, 3406. 18. Josey, D. S.; Castrucci, J. S.; Dang, J. D.; Lessard, B. H.; Bender, T. P. ChemPhysChem 2015 , 16 , 1245-1250. 19. Hancox, I.; Sullivan, P.; Chauhan, K. V.; Beaumont, N.; Rochford, L. A.; Hatton, R. A.; Jones, T. S. Organic Electronics 2010 , 11 , 2019-2025. 20. Hancox, I.; Chauhan, K. V.; Sullivan, P.; Hatton, R. A.; Moshar, A.; Mulcahy, C. P. A.; Jones, T. S. Energy & Environmental Science 2010 , 3, 107-110. 21. Hancox, I.; Rochford, L. A.; Clare, D.; Sullivan, P.; Jones, T. S. Applied Physics Letters 2011 , 99 , 013304. 22. Cattin, L.; Dahou, F.; Lare, Y.; Morsli, M.; Tricot, R.; Houari, S.; Mokrani, A.; Jondo, K.; Khelil, A.; Napo, K.; Bernede, J. C. Journal of Applied Physics 2009 , 105 , 034507.

Chapter 4 Growth of μ-oxo-boron subphthalocyanine (μ-oxo-(BsubPc) 2) thin films via organic chemical vapor deposition (OCVD)

The author list is: Jeremy D. Dang, Anjuli M. Szawiola, David S. Josey, and Timothy P. Bender.

I carried out all experiments involved with the growth of μ-oxo-(BsubPc) 2 thin films and their characterizations. AMS and DSJ measured the thicknesses of the μ-oxo-(BsubPc) 2 thin films acquired via OCVD. The work was supervised by TPB.

Supplementary information for this chapter is found in Appendix C.

4.1 Abstract

An OCVD process was designed and developed for the growth of thin films of μ-oxo-(BsubPc) 2 from Br-BsubPc. Initially, micron-thick films were produced and they were a mixture of three

BsubPc compounds - μ-oxo-(BsubPc) 2, HO-BsubPc, and Br-BsubPc. It was shown that the relative composition can be altered by modifying the thermal profile of the process. Following some optimizations, nanometer-thick films containing μ-oxo-(BsubPc) 2 as high as 74% were afforded. Neat films have not been realized via this technique, but given the promising performance properties for a mixture of μ-oxo-(BsubPc) 2 and Cl-BsubPc as a single layer in PHJ OPVs, efforts will be made to examine the applicability of these OCVD films within devices.

4.2 Introduction

The oxygen-bridged dimer of BsubPc, μ-oxo-(BsubPc) 2 (Figure 4.1), was first described in the literature in 1996, 1 but its physical properties were greatly underexplored for nearly twenty years; something attributable to the difficulties behind its synthesis. In response to this and as described in Chapter 3, we have recently developed a synthetic method to produce gram quantities of highly pure μ-oxo-(BsubPc) 2 via a reaction of HO-BsubPc with Br-BsubPc in the 2,3 presence of tripotassium phosphate (K 3PO 4) in refluxing 1,2-dichlorobenzene (Scheme 4.1 ). 2 This work permitted us to examine μ-oxo-(BsubPc) 2's crystal packing arrangement (Figure 4.1), its solution-state properties (described in Chapter 3) such as its absorption, fluorescence and electrochemical behaviors, 3 and its performance characteristics (described in Chapter 3) in OPV devices, where PCEs as high as 2.74% were achieved 4.

While reasonable amounts of μ-oxo-(BsubPc) 2 with a purity level suitable for OPV devices can be attained in our laboratory, the process is overall fairly labor-intensive and first requires the synthesis of two BsubPc precursors (Scheme 4.1). The precursors, Br-BsubPc and HO-BsubPc, are prepared from the cyclotrimerization reaction of phthalonitrile with boron tribromide and from the hydrolysis of Br-BsubPc, respectively. Both BsubPc starting materials are then purified prior to their bimolecular condensation reaction to ultimately afford crude μ-oxo-(BsubPc) 2. μ-

Oxo-(BsubPc) 2 is subsequently purified in a four-step process consisting of a Soxhlet extraction, Kaufmann column chromatography, and two rounds of train sublimation. 2,3

Figure 4.1 . Chemical and crystal packing structure of μ-oxo-(BsubPc) 2 (CCDC deposition number: 914863).

CN Br (i) N N N CN N B N N N N B N O N N N (ii) (iii) N N N N B N

OH N N N N N B N N

Scheme 4.1 . Synthetic route to μ-oxo-(BsubPc) 2. Reagents and conditions: (i) boron tribromide, toluene, bromobenzene; (ii) water, pyridine, dimethyl sulfoxide, 60 °C; and (iii) tripotassium phosphate, 1,2-dichlorobenzene, reflux.

μ-Oxo-(BsubPc) 2 is known to form as a side product during the reaction of X-BsubPcs. Presumably this occurs through the reaction with trace amounts of water yielding HO-BsubPc, which then condenses with X-BsubPc within the process yielding μ-oxo-(BsubPc) 2. Leveraging

this process, a potential way to address the intensive synthetic procedure described above for μ- oxo-(BsubPc) 2 is to adopt a chemical vapor deposition (CVD) approach. CVD is a well-known technology that dates back as early as 1855, 5 and is used to coat a substrate with a material that is made in situ from the chemical reaction of vapor/gaseous precursors. The chemical reactions found in this technique are normally either of the decomposition- or displacement-type. CVD have predominantly been used for inorganic and metalorganic precursors to yield highly pure elemental metals or metal alloys. 6-12 Its use with organic precursors are limited. Earliest examples of this involve the deposition of graphite on a number of different substrates from the pyrolysis of hydrocarbons. 13,14 This technique was also extended towards the growth of graphene, 15-17 carbon nanotubes, 18,19 and carbon fibers 19,20 . Examples outside of these carbon allotropes have been demonstrated in several cases by Gleason et al . for the deposition of various polymers like poly(3,4-ethylenedioxythiophene) (PEDOT), 21-24 poly(tetrafluoroethylene), 25 poly(dimethylaminomethylstyrene), 26 and poly(cyclohexyl methacrylate-co -ethylene glycol dimethacrylate) 27 from their respective monomers onto various substrates. Another report is that by Burrows et al ., where 4’-dimethylamino-N-methyl-4-stilbazolium tosylate was formed from the reaction of 4’-dimethylamino-4-stilbazole with methyl p-toluenesulfonate. It is worth noting that the authors referred to this process as an organic vapor phase deposition (OVPD) despite the fact that it is more suited as a CVD. 28

With the limited findings on CVD of organic precursors, it is not unexpected that the use of this approach for the growth of thin films within OPV devices is further rare. Even considering the reports cited above with regards to EDOT (the monomer of PEDOT) deposition, we are only able to cite four publications whereby CVD was merged/integrated into a functional organic electronic device. In one report, methane-derived graphene films were made and deposited as the 29 anode electrode within an OPV. In a second report, PEDOT was synthesized via FeCl3- mediated polymerization of 3,4-ethylenedioxythiophene and used as the anode electrode. 30 By means of the same chemistry, polythiophene was deposited as the donor layer on indium tin 31 oxide (ITO) from thiophene monomers and FeCl 3 in the third report. In the fourth report, CVD- derived poly( p-phenylenevinylene) was paired with a fullerene acceptor in an OPV cell. 32 In

contrast to these reports, there are a number of papers on inorganic photovoltaic devices fabricated by CVD. 33-38

Herein, we describe our efforts in developing a CVD process for yielding thin films of μ-oxo-

(BsubPc) 2 at low pressure (vacuum) using Br-BsubPc as the sole reactant. The advantages of this approach are twofold: (1) only a single BsubPc reactant, Br-BsubPc, is used as opposed to the combination of Br-BsubPc and HO-BsubPc and (2) the laborious steps associated with purifying

μ-oxo-(BsubPc) 2 are avoided. Another unique feature about our process is that the chemical reaction under study is synthetic in nature, where a larger, more complex organic small molecule is formed with minimal mass loss. This is opposed to the decomposition or displacement reactions often seen in CVD, whereby mass loss is significant. For the reason that the BsubPc precursors are organic in nature and that CVDs of metalorganic/organometallic compounds are classified as metalorganic CVD (MOCVD), we have therefore appropriately termed our technique “organic chemical vapor deposition (OCVD)”, which is not to be confused with oxidative chemical vapor deposition (oCVD) 39 .

4.3 Results and Discussion

Recall that μ-oxo-(BsubPc) 2 is often formed as a side product of X-BsubPc reactions. We hypothesize that it is formed in a two-step process: 1) hydrolysis of X-BsubPc with trace amounts of water to form HO-BsubPc and 2) bimolecular condensation reaction of X-BsubPc and HO-BsubPc. In an effort to develop an OCVD process for the growth of μ-oxo-(BsubPc) 2 thin films, Br-BsubPc was selected as the sole precursor due to its susceptibility to hydrolyze and form HO-BsubPc while being stored under ambient conditions.

The design and build of our OCVD apparatus (Figure 4.2a) was adapted from our train sublimation system (Figure S4.1). 40 A borosilicate glass tube is used as a reactor and equipped with a cold finger ( i.e. condenser) at the downstream end. The tube is held horizontally through the support of a frame and a copper-nickel pipe situated near the upstream end. A band heater, wrapped with flexible ceramic insulator, is mounted around the copper-nickel pipe and is powered by a temperature indicating controller with a K type thermocouple wire sandwiched

between the band heater and the copper-nickel pipe. Carrier gas is introduced into the reactor at the upstream end with its flow controlled by a ball valve and a needle valve. The option to heat the carrier gas prior to its flow into the reactor is facilitated by a coil wrapped with a layer of two heating tapes and multiple layers of aluminum foil. Each heating tape is connected to a temperature controller, whereby temperature is monitored through a K type thermocouple wire and temperature indicator. A vacuum pump is installed at the far downstream side of the apparatus to generate a pressure gradient and is equipped with a pressure indicator.

In a typical experiment, a mass of Br-BsubPc is added to a stainless steel boat (L = 2”) and positioned inside the OCVD reactor within the heating zone (defined by the band heater). The substrate, either plain glass slide or indium-tin oxide (ITO) on glass, is sandwiched in a vertical orientation between hollow glass inserts that are placed downstream from the boat. The apparatus is pumped down to a pressure below 100 mTorr before the carrier gas is introduced into the system at a flow rate equal to ~100 mTorr above the pumped down pressure. In experiments where the carrier gas is pre-heated, the power controllers for the coil are turned on and its power output is user-regulated via a K type thermocouple wire and temperature reader. Heating of the boat containing the Br-BsubPc is applied using a thermal profile programmed on the temperature indicating controller. Film thickness was measured via surface profilometry using Kapton tape to partially mask the substrate to create step edges. The OCVD process is schematically illustrated in Figure 4.2b.

In consideration of the cost of ITO on glass substrates, all initial OCVD experiments were carried out using glass slides until an optimized process was developed. As a starting point, the thermal profile from the train sublimation experiment of μ-oxo-(BsubPc) 2 was adapted to our OCVD process. 3 Deposits of this dimeric compound were found at a distance between 4" and 5” downstream from the tail end of the boat. Given this, all following experiments had the substrate positioned at ~4” from the boat.

UPSTREAM END DOWNSTREAM END

BOAT SUBSTRATE GLASS INSERT TI TEMPERATURE INDICATOR COIL WRAPPED OCVD REACTOR WITHIN OCVD WITH HEAT TAPE REACTOR CONDENSER PRESSURE INDICATOR

PI BAND COPPER-NICKEL PIPE NEEDLE TC HEATER VALVE CERAMIC TEMPERATURE VACUUM INSULATOR CONTROLLER TIC PUMP

BALL TEMPERATURE VALVE INDICATING CONTROLLER

(a) GAS INLET GAS OUTLET

(b) Figure 4.2 . Schematic diagram of (a) the OCVD apparatus and (b) the OCVD process within the reactor. The hollow glass inserts have been omitted for clarity.

In the first OCVD experiment, the substrate was found to contain a mixture of three BsubPc compounds as determined by a HPLC-UV/Vis analysis (Table S4.1, Method 1.1). The three

BsubPc compounds were identified, based on their known retention times (R t) and optical absorption spectra, as HO-BsubPc, Br-BsubPc, and μ-oxo-(BsubPc) 2 at a relative composition of 2, 98, and <1%, respectively. HPLC analysis was also performed on the residual material in the boat and on the deposited film on the glass inserts and the results are summarized in Table S4.1. A fourth unknown BsubPc compound was detected in the boat and in one of the inserts, but was only formed in minor amounts. The deposited film on the glass substrate was noticeably not uniform by visual inspection (Figure 4.3a). By profilometry, the film surface was rough and characterized by the presence of crystallites (Figure 4.3b). The film thickness was measured to be in the range of 0.9-1.7 microns.

(a)

(b) Figure 4.3 . (a) Picture of the glass substrate at three different locations and (b) a close-up picture (black and white camera) showing the crystallites following an OCVD (Method 1.1).

Given the near negligible composition of the target product on the substrate, the carrier gas was heated to 170-180 °C in a subsequent experiment to potentially facilitate the synthetic reaction

(Method 1.2). Nearly identical results were observed, indicating that perhaps the temperature of the gas was not hot enough. However, the 170-180 °C is the upper limit for pre-heating the nitrogen gas before a noticeable odour forms. For this reason, the thermal profile was next tuned towards a higher end temperature of 500 ˚C (Method 1.3). This resulted in a much greater amount of μ-oxo-(BsubPc) 2 on the substrate (18% vs. <1% composition). Like the first OCVD experiment (Method 1.1), the film lacked uniformity and its thickness was roughly on the micron scale. Given that the thickness of the organic photoactive layers in PHJ OPV devices are on the order of 10-40 nm, the next task was to decrease the film thickness by using less Br-BsubPc in the boat. Moreover, targeting thinner films could solve or reduce the uniformity issue due to less aggregation and crystallite formation. With less Br-BsubPc mass, the relative composition of μ- oxo-(BsubPc) 2 on the substrate following OCVD was found to be higher. The use of 10 mg and 5 mg of Br-BsubPc resulted in a 39% and 32% relative composition of μ-oxo-(BsubPc) 2, respectively (Method 1.4-1.5). The thickness of the former film was still on the micron scale while the thickness of the latter film was roughly in the ~240-500 nm range. Before targeting even thinner films, efforts were made to enhance the μ-oxo-(BsubPc) 2 composition by increasing the the end temperature to 550 °C (Method 1.6) in one experiment and to 600 °C in a second experiment (Method 1.7). The former experiment produced a film with the highest μ-oxo-

(BsubPc) 2 composition of 52% while the latter experiment produced a film with a low μ-oxo-

(BsubPc) 2 composition of 6%.

Despite using less Br-BsubPc precursor in the previous set of experiments, all films appeared rough under visual inspection and by profilometry. Considering this roughness, LRMS analysis was performed on three different areas of the film produced via Method 1.6 (highest μ-oxo-

(BsubPc) 2 composition) to determine if the relative compositions of the BsubPcs were uniform across the substrate (Figure S4.3 and Table S4.2). The results varied among the three sample sites. In all cases, the μ-oxo-(BsubPc) 2 m/z peak intensity was the highest. A Br-BsubPc m/z peak was not present and this was due to molecular fragmentation given its high sensitivity to axial substitution.

We then attributed the uniformity issue to the presence of surface contaminants (organic residue) on the glass substrate that were not removed via simple washing techniques. To determine that this is the cause and to potentially overcome this problem, a glass substrate was cleaned by successive sonications in detergent, acetone, and methanol followed by atmospheric

O2 plasma treatment. The “cleaned” substrate was immediately used in an OCVD experiment adopting the thermal profile of Method 1.6 (Method 1.8). The resulting film was much more uniform than those produced prior, verifying that surface contaminants are a major culprit

(Figure 4.4). Moreover, and perhaps just as important, the μ-oxo-(BsubPc) 2 composition was found to be higher at 73 % (vs. 52%)

We next looked into the effect of the carrier gas. A substitution for a less wet nitrogen gas (<5 ppm vs. 20 ppm of H 2O) was done and this produced very similar results (Method 1.9). This was also the case with the use of CO 2 carrier gas (<5 ppm of H 2O, Method 1.10).

(a) (b) Figure 4.4 . (a) Picture of the glass substrate and (b) a close-up picture (black and white camera) following an OCVD (Method 1.8).

With a semi-optimized OCVD process, ITO on glass substrate was next incorporated. Once again, via the thermal profile adopted in Method 1.6, the OCVD experiment yielded a 59% μ- oxo-(BsubPc) 2 film (Method 1.11). This composition is lower compared to that of the regular glass substrate (73%). Integrating a thin layer of PEDOT:PSS (~35 nm) on top of the ITO substrate produced a 52% μ-oxo-(BsubPc) 2 film after an OCVD process (Method 1.12).

To determine if a self-condensation of HO-BsubPc is a better approach for affording μ-oxo-

(BsubPc) 2, an OCVD experiment using HO-BsubPc was next carried out (Method 1.13). Via the thermal profile of Method 1.6, a film was produced with a much lower composition of the target product (28%), indicating that this route is evidently less beneficial. Secondly, this result demonstrates that the high μ-oxo-(BsubPc) 2 composition from the previous experiments are likely due to a bimolecular condensation reaction between Br-BsubPc and HO-BsubPc, made in situ from the hydrolysis of Br-BsubPc as we have hypothesized.

Seeing that it may not be possible to achieve a neat film of μ-oxo-(BsubPc) 2, the next set of efforts should focus on determining if the aforementioned BsubPc mixture is actually problematic to OPV device performance. We have results within the group showing that mixtures consisting of μ-oxo-(BsubPc) 2 and Cl-BsubPc actually perform well as a single layer within PHJ OPV devices. 4

Before this motion can be put into effect, thinner films on the order of 10-40 nm will need to be targeted. This will require using even less Br-BsubPc (<5 mg) in the OCVD process. Once the appropriate film thickness has been established, the film should be applied into a PHJ OPV device of the following architecture: ITO/PEDOT:PSS/μ-oxo-(BsubPc) 2 + X-BsubPcs/C 60/70 /Ag. If the OPV cell generates a very low PCE or does not work, then it can be concluded that this particular mixture format is detrimental to the performance of the cell. Further optimizations in the OCVD process can then be made in an endeavor to get a neat film of μ-oxo-(BsubPc) 2. Variables such as gas pressure, gas temperature, and temperature profile ( i.e. rate of temperature increase) can all be tuned. It may also be worthwhile to wrap the reactor tube with flexible ceramic insulator to create a longer temperature gradient or smaller drop in temperature as a function of distance downstream from the boat/heating band. The motive behind this is to promote the continual sublimation of the smaller and lighter BsubPcs, Br-BsubPc and HO-

BsubPc, near and on the substrate. This should lead to a higher composition of μ-oxo-(BsubPc) 2 in the film. If the OPV cell is shown to operate, then our OCVD process would have been successfully merged into a functional device.

4.4 Conclusions

An OCVD process for the formation of μ-oxo-(BsubPc)2 from Br-BsubPc was proposed, designed, and developed. In its earliest attempts, thick films on the micron scale were afforded.

These films comprised three BsubPc compounds - μ-oxo-(BsubPc) 2, HO-BsubPc, and Br- BsubPc. It was demonstrated herein that the relative composition can be optimized by tuning the thermal profile of the process, yielding films with μ-oxo-(BsubPc) 2 as high as 74% on plain glass substrate. Some issues with film uniformity and roughness were encountered, but was later resolved with the use of a more intensive cleaning method to remove surface contaminants from the substrate. It was also determined that the carrier gas is not a factor in the process. Thus far, we are unable to produce neat films of μ-oxo-(BsubPc) 2 via this OCVD technique. However, before additional effort is put into fully optimizing this process, an endeavour will be made to determine if this particular mixture format is actually detrimental to OPV device performance. Following this route, it was shown that the switch to ITO glass substrate, with and without a layer of PEDOT:PSS, is adaptable and resulted in a small decrease in the μ-oxo-(BsubPc) 2 composition relative to plain glass. The next step is to fully integrate/merge this process towards an OPV device.

4.5 References 1. Geyer, M.; Plenzig, F.; Rauschnabel, J.; Hanack, M.; Del Rey, B.; Sastre, A.; Torres, T. Synthesis 1996 , 1139-1151. 2. Fulford, M. V.; Lough, A. J.; Bender, T. P. Acta Crystallographica Section B-Structural Science 2012 , 68 , 636-645. 3. Dang, J. D.; Fulford, M. V.; Kamino, B. A.; Paton, A. S.; Bender, T. P. Dalton Transactions 2015 , 44 , 4280-4288. 4. Castrucci, J. S.; Garner, R. K.; Dang, J. D.; Thibau, E.; Lu, Z. H.; Bender, T. P. ACS Applied Materials & Interfaces . Paper under revision. 5. Wohler, F.; Uslar, L. Justus Liebigs Annalen der Chemie 1855 , 94 , 255-256. 6. Mond, L.; Langer, C.; Quincke, F. Journal of the Chemical Society, Transactions 1890 , 57 , 749-753. 7. Pring, J. N.; Fielding, W. Journal of the Chemical Society 1909 , 95 , 1497-1506. 8. Didchenko, R.; Alix, J. E.; Toeniskoetter, R. H. Journal of Inorganic & Nuclear Chemistry 1960 , 14 , 35-37. 9. Harrison, B. C.; Tompkins, E. H. Inorganic Chemistry 1962 , 1, 951-953. 10. Lilley, P.; Kay, P. M. R.; Litting, C. N. W. Journal of Materials Science 1975 , 10 , 1317- 1322. 11. Cooke, M. J.; Heinecke, R. A.; Stern, R. C.; Maes, J. W. C. Solid State Technology 1982 , 25 , 62-65. 12. Dormans, G. J. M.; Vanveldhoven, P. J.; Dekeijser, M. Journal of Crystal Growth 1992 , 123 , 537-544. 13. Karu, A. E.; Beer, M. Journal of Applied Physics 1966 , 37 , 2179-2181. 14. Robertson, S. D. Nature 1969 , 221 , 1044-1046. 15. Sutter, P. W.; Flege, J. I.; Sutter, E. A. Nature Materials 2008 , 7, 406-411. 16. Reina, A.; Jia, X. T.; Ho, J.; Nezich, D.; Son, H. B.; Bulovic, V.; Dresselhaus, M. S.; Kong, J. Nano Letters 2009 , 9, 30-35. 17 Li, X. S.; Magnuson, C. W.; Venugopal, A.; An, J. H.; Suk, J. W.; Han, B. Y.; Borysiak, M.; Cai, W. W.; Velamakanni, A.; Zhu, Y. W.; Fu, L. F.; Vogel, E. M.; Voelkl, E.; Colombo, L.; Ruoff, R. S. Nano Letters 2010 , 10 , 4328-4334.

18. Kong, J.; Cassell, A. M.; Dai, H. J. Chemical Physics Letters 1998 , 292 , 567-574. 19. Che, G.; Lakshmi, B. B.; Martin, C. R.; Fisher, E. R.; Ruoff, R. S. Chemistry of Materials 1998 , 10 , 260-267. 20. Tibbetts, G. G. Carbon 1989 , 27 , 745-747. 21. Lock, J. P.; Im, S. G.; Gleason, K. K. Macromolecules 2006 , 39 , 5326-5329. 22. Im, S. G.; Yoo, P. J.; Hammond, P. T.; Gleason, K. K. Advanced Materials 2007 , 19 , 2863-2867. 23. Im, S. G.; Gleason, K. K. Macromolecules 2007 , 40 , 6552-6556. 24. Im, S. G.; Kusters, D.; Choi, W.; Baxamusa, S. H.; de Sanden, M.; Gleason, K. K. ACS Nano 2008 , 2, 1959-1967. 25. Cruden, B. A.; Gleason, K. K.; Sawin, H. H. Journal of Vacuum Science & Technology B 2002 , 20 , 690-695. 26. Martin, T. P.; Gleason, K. K. Chemical Vapor Deposition 2006 , 12 , 685-691. 27. Chan, K.; Gleason, K. K. Journal of the Electrochemical Society 2006 , 153 , C223-C228. 28. Burrows, P. E.; Forrest, S. R.; Sapochak, L. S.; Schwartz, J.; Fenter, P.; Buma, T.; Ban, V. S.; Forrest, J. L. Journal of Crystal Growth 1995 , 156 , 91-98. 29. De Arco, L. G.; Zhang, Y.; Schlenker, C. W.; Ryu, K.; Thompson, M. E.; Zhou, C. W. ACS Nano 2010 , 4, 2865-2873. 30. Barr, M. C.; Rowehl, J. A.; Lunt, R. R.; Xu, J. J.; Wang, A. N.; Boyce, C. M.; Im, S. G.; Bulovic, V.; Gleason, K. K. Advanced Materials 2011 , 23 , 3500-3505. 31. Borrelli, D. C.; Barr, M. C.; Bulovic, V.; Gleason, K. K. Solar Energy Materials and Solar Cells 2012 , 99 , 190-196. 32. Supangat, A.; Zhou, X. J.; Belcher, W.; Dastoor, P. Materials Research Innovations 2011 , 15 , 18-20. 33. Dupuis, R. D.; Dapkus, P. D.; Yingling, R. D.; Moudy, L. A. Applied Physics Letters 1977 , 31 , 201-203. 34. Chen, J. C.; Ristow, M. L.; Cubbage, J. I.; Werthen, J. G. Journal of Electronic Materials 1992 , 21 , 347-353. 35. Chen, X. L.; Xu, B. H.; Xue, J. M.; Zhao, Y.; Wei, C. C.; Sun, J.; Wang, Y.; Zhang, X. D.; Geng, X. H. Thin Solid Films 2007 , 515 , 3753-3759.

36. Clayton, A. J.; Irvine, S. J. C.; Barrioz, V.; Brooks, W. S. M.; Zoppi, G.; Forbes, I.; Rogers, K. D.; Lane, D. W.; Hutchings, K.; Roncallo, S. Thin Solid Films 2011 , 519 , 7360-7363. 37. Ou, K. L.; Tadytin, D.; Steirer, K. X.; Placencia, D.; Nguyen, M.; Lee, P.; Armstrong, N. R. Journal of Materials Chemistry A 2013 , 1, 6794-6803. 38. Kartopu, G.; Barrioz, V.; Irvine, S. J. C.; Clayton, A. J.; Monir, S.; Lamb, D. A. Thin Solid Films 2014 , 558 , 374-377. 39. Tenhaeff, W. E.; Gleason, K. K. Advanced Functional Materials 2008 , 18 , 979-992. 40. Morse, G. E.; Helander, M. G.; Maka, J. F.; Lu, Z. H.; Bender, T. P. ACS Applied Materials & Interfaces 2010 , 2, 1934-1944.

Chapter 5 The mixed alloyed chemical composition of chloro-(chloro) n-boron subnaphthalocyanines dictate their physical properties and performance in organic photovoltaic devices

Adapted with permission from: Jeremy D. Dang, David S. Josey, Alan J. Lough, Yiying Li, Alaa Sifate, Zheng-Hong Lu, and

Timothy P. Bender. "The mixed alloyed chemical composition of chloro-(chloro) n-boron subnaphthalocyanines dictate their physical properties and performance in organic photovoltaic devices." Journal of Materials Chemistry A 2016 . DOI: 10.1039/C6TA02457B.

I carried out all synthetic experiments (with some aid from AS, summer student) and characterized the products via NMR, optical absorption, photoluminescence, and voltammetry. DSJ performed all OPV device fabrication and testing. AJL diffracted and determined all single crystals. YL performed the UPS and XPS measurements. The work was supervised by ZHL and TPB. All authors approved the manuscript.

Supplementary information for this chapter is found in Appendix D.

5.1 Abstract Chloro-boron subnaphthalocyanine (Cl-BsubNc) has recently attracted significant interest as a light-harvesting and charge transporting material in organic photovoltaics (OPVs) by enabling an 8.4% efficient PHJ OPV cell. We present herein a variety of experimental data that supports the conclusion that Cl-BsubNc, whether synthesized via literature methods or our in-house methods or purchased commercially, is actually a mix alloyed composition of Cl-BsubNcs with random amounts of chlorination at the bay position(s) of the BsubNc macrocyclic structure (Cl-

Cl nBsubNc(s)). We outline our efforts to develop alternative chemical processes, whereby we did obtain samples with lower and higher amounts of bay position chlorination. However, we were unable to obtain a pure, non-chlorinated sample of Cl-BsubNc. The positions and frequencies of the peripheral chlorine atoms were determined via single crystal X-ray crystallography, mass spectrometry, and X-ray photoelectron spectroscopy. The photo- and electro-physical properties were found to differ among the Cl-Cl nBsubNc samples. These differences also translated into varying performance within PHJ OPVs, whereby a mixture of Cl-Cl nBsubNcs with lower amounts of chlorination produced less efficient OPVs compared to a mixture with higher amounts of chlorination. Additionally, an in-house made sample of Cl-Cl nBsubNc with the highest level of bay position chlorination, yielded the best performing OPVs through an improved fill factor. A commercial sample of Cl-Cl nBsubNc also yielded OPVs with efficiencies equivalent to a Cl-Cl nBsubNc sample made in our laboratory. This mixture of Cl-Cl nBsubNcs are therefore likely present in the reported 8.4% efficient OPV. Our results therefore offer a cautionary note that the Cl-BsubNc samples used within the existing literature are likely not a pure chemical composition but are rather mixtures of Cl-Cl nBsubNcs with bay position chlorination. Our findings clarify the previous literature on the chemistry of Cl-BsubNc, firms up the photo- and electro-physical properties of these materials, and offer additional insight into their application as functional materials in efficient OPVs.

5.2 Introduction Boron subnaphthalocyanines (BsubNcs, Figure 5.1) are but minor constituents within a broader class of well-known and highly studied compounds known as phthalocyanines (Pcs). 1-3 BsubNcs are more structurally similar to another minor constituent of the Pc

family, boron subphthalocyanines (BsubPcs, Figure 5.1). BsubPcs are comprised of three nitrogen-bridged isoindoline subunits that chelate a single boron atom within their internal cavity. 4 BsubPcs are known to be aromatic, non-planar, and cone-shaped molecules. 5-8 Despite being first synthesized and characterized in the early 1970s, the majority of the synthesis and application work on BsubPcs has occurred within the last 10 to 15 years. 9-12

Figure 5.1. Generic chemical structures of boron subphthalocyanine (BsubPc) and boron subnaphthalocyanine (BsubNc) and the generic synthesis of BsubPcs and BsubNcs where X = Cl, Br, Ph, or other. R group designations omitted in places for clarity.

Halo-BsubPcs are relatively easy to prepare by the reaction of phthalonitrile (an ortho -dinitrile) 10,13 14,15 with a boron trihalide templating agent such as BCl3 or BBr 3 in an aromatic solvent resulting in the formation of Cl-BsubPc and Br-BsubPc, respectively, in high yield and high

purity (Figure 5.1). The halide of no other metal from the periodic table is known to template the formation of BsubPcs, whereas all other elements template the formation of Pcs.

Like BsubPcs, BsubNcs have been reportedly prepared by heating 2,3- dicyanonaphthalene (also an ortho -dinitrile) with a boron templating agent ( e.g. PhBCl 2,

BBr 3, BCl 3) in an aromatic solvent (Figure 5.1). A relatively limited amount of work has been reported in the literature on the synthesis of BsubNcs compared to BsubPcs. In summary, the first report of the synthesis of a BsubNc was by Hanack and Rauschnabel in 1995 (Table S5.1, Entry 1), where they treated 2,3-dicyanonaphthalene and an alkylated

2,3-dicyanonaphthalene derivative with PhBCl 2 in boiling naphthalene to produce the respective Ph-BsubNcs, albeit in only analytical amounts. 16,17 In 1999, Kobayashi et al. reported the use of BBr 3 as the boron templating agent to synthesize Br-BsubNc in a 34.6% yield (Table S5.1, Entry 2), 9 although Torres et al. 18 had questioned the purity of this sample. In 2000, both Kennedy 13 and Torres 18 reported the synthesis of Cl-BsubNc in

53% and 35%, respectively, using BCl 3 as the boron template and 2,3- dicyanonaphthalene as the starting material (Table S5.1, Entry 3 & 4). In 2007 Giribabu et al. reportedly employed microwave irradiation to afford a tert -butylated Cl-BsubNc in 82% yield (Table S5.1, Entry 5). 19 In 2014, Takao et al . made two peripherally fluorinated Cl-BsubNc derivatives, Cl-F6BsubNc (26% yield) and Cl-F12 BsubNc (44% yield), from their respective fluorinated 2,3-dicyanonaphthalene via a multistep synthesis

(Table S5.1, Entry 6). Takao et al . also treated the two products with AgBF 4 to substitute the axial chloride with a fluoride. 20 Very recently, Mizutani et al . made two other peripherally halogenated Cl-BsubNc derivatives (Table S5.1, Entry 7), Cl-Cl 6BsubNc

(22% yield) and Cl-I6BsubNc (73% yield), and also carried out the axial chloride-to- chloride transformation. 21 In addition to the synthesis of ‘standard’ BsubNcs from 2,3- dicyanonaphthalene, Kobayashi et al. also made chiral analogs of BsubNcs via the reaction of 1,2-dicyanonaphthalene with BCl 3 to form chiral Cl-1,2-BsubNc and subsequent phenoxy derivatives. 22

A review of the above mentioned studies (see Table S5.1) reveals that detailed characterization of the BsubNcs is generally limited. 1H NMR data has been numerically reported (Entry 1, 3, 4, 6, and 7) but spectra were only published for the latter two entries to graphically illustrate the purity of the sample(s). High resolution mass spectrometry (HRMS) data has only been reported for the BsubNcs in Entry 6 and 7; low resolution mass spectrometry (LRMS) data was reported for Entry 3 and 4. BsubNcs from Entry 2 and 4 have reported elemental analysis (EA) results, however, the EA data was only supplied by the former entry and the analysis was not within an acceptable range from the calculated values of the respective proposed molecular formula (C 36 H18 N6BBr) for publication ( i.e. +/- 0.4%). Overall, adequate characterizations are only reported for the BsubNcs in Entry 6 and 7.20,21

BsubPcs have recently been applied as functional materials in organic electronic devices most notably in organic photovoltaic devices (OPVs). 23-32 Although the recent efforts of our group 30,31 and others 26-29,32 have shown a variety of BsubPcs can have application in OPVs, the majority of the applications have focused on the use of the prototypical derivative Cl-BsubPc. 33 Through its use, power conversion efficiencies as high as 4.69% have been achieved. 32

At the same time, Cl-BsubNc has also been applied in OPVs. 28,32,34-40 As a result of the more-extended π-conjugation system compared to Cl-BsubPc, Cl-BsubNc has a λ max absorption that is significantly red-shifted to the mid 600 nm region in solution 13 and at 686 nm in the solid-state, 34 making Cl-BsubNc able to capture regions of red light for OPV applications. The first examples of Cl-BsubNc in OPVs showed that its substitutions for Cl-BsubPc as an electron donating and charge generating material paired with 34-36,38 fullerene (C 60 ) resulted in small improvements in photocurrent generation. This substitution reduced the spectral overlap between the donor and acceptor, but lost the photocurrent contributions in the 500-600 nm region from the Cl-BsubPc. Gommans et al. were the first to pair a BsubPc with Cl-BsubNc, reclaiming the BsubPc region of the 28 spectrum but losing the blue region contributions from C 60 .

Recently, Cheyns and Cnops et al. have demonstrated that the inclusion of Cl-BsubNc into a planar heterojunction (PHJ) OPV enables power conversion efficiencies (PCEs) as high as 6.4% 40 and 8.4% 32. These measured efficiencies are amongst the highest achieved for PHJ OPVs. Cheyns, Cnops et al. show evidence that the high PCEs are attributable to the complementary absorption profiles of BsubPc and BsubNc, broadening the overall absorption range of the OPV. To achieve the 8.4% efficiency, an energy-relay cascade was used to transfer excitons to a single dissociating interface, avoiding the drop in V OC typically seen in other 3-layer OPVs where multiple dissociating interfaces are incorporated. 32

Now considering the chemical nature of the Cl-BsubNc used in the above mentioned OPVs, in three cases, Cl-BsubNc was obtained from a commercial source with an advertised “75% dye content” ( i.e. 75% purity). 28,34,35 In two of these three studies, 28,35 the Cl-BsubNc was used “as is” while in the third, 34 Cl-BsubNc was purified via thermal gradient sublimation prior to use. This commercial supplier no longer sells Cl-BsubNc. In four other studies, Cl-BsubNc was obtained from a second (and current) commercial supplier who does not advertise its purity. 37-39,41 In two other papers, the source of the Cl- BsubNc was not reported but is assumed to be synthesized in the respective laboratories. 32,40

Concurrent with the communications from Verreet et al. 40 and Cnops et al. 32 our group was investigating the process chemistry surrounding Cl-BsubNc. What we found was that the literature procedure for producing Cl-BsubNc in fact produces a product mixture with a considerable amount of chlorination of the BsubNc chromophore. We also determined that the Cl-BsubNc offered by a commercial supplier has equivalent amounts of chlorination. We have identified that the chlorination is present exclusively at the bay position of the BsubNc chromophore. This is in marked contrast to our past experience with the synthesis of Cl-BsubPc 13 using a similar synthetic protocol whereby no peripheral chlorination is present or has ever been observed. This is of general concern given that peripheral chlorination of Pcs is known to affect the resulting OPV

100

characteristics. 42,43 Very recently, Kahn et al. published a paper outlining the use of XPS to determine the presence of peripheral chlorination within a sample of Cl-BsubNc also from a commercial supplier. 44

Herein, we outline our multi-year synthetic and device integration study, how we arrived at these conclusions and how our conclusions further support the very recent observations of Kahn et al .44 Included herein is extensive documentation of our conclusions and our efforts in modifying the synthetic and work-up procedures to obtain a highly pure (or purer) version of Cl-BsubNc with little chlorination and alternately a highly chlorinated sample(s) of Cl-BsubNc to provide points of contrast and comparison. We also present the subsequent detailed physical characterization of chlorinated Cl-BsubNcs obtained using our new processes including their incorporation into PHJ OPVs and their direct comparison with the “literature procedure” Cl- BsubNc and “commercially available” Cl-BsubNc samples. Points of comparison also include the changes in the fundamental photo- and electro-physical properties as a function of the amount of bay position chlorination.

5.3 Results and Discussion Cl-BsubNc Synthesis Our initial investigation began with the synthesis of Cl-BsubNc via the literature method of 18 Torres et al. , where 2,3-dicyanonaphthalene was reacted with BCl3 in a mixture of chlorobenzene and toluene (1:1 v/v ) at 130 °C (Table S5.2 - Method 1.1). This synthetic procedure is directly analogous to the one used reproducibly for the synthesis of Cl-BsubPc. 13 In our hands, this procedure resulted in the formation of five Cl-BsubNc compounds as determined by HPLC-UV/Vis analysis of the reaction mixture with R t ~ 3.6, 4.4, 5.4, 6.9, and 7.1 min, each with an absorption profile characteristic of a BsubNc chromophore (Figure S5.1). The reported purification method, a Soxhlet extraction using hexane followed by chromatography on silica gel using toluene, did not produce a chromatographically pure sample of Cl-BsubNc rather it still gave a mixture of five BsubNc products with some other non-BsubNc impurities remaining. Two additional reactions based on a modification of this method were carried out; an increase in the molar equivalencies of BCl 3 from 1.0 to 2.5 (Table S5.2 - Method 1.2) and substitution of

101

toluene for para -xylene (Table S5.2 - Method 1.3). Each led to the formation of the same five BsubNc compounds.

To determine if the formation of the five BsubNc compounds was exclusive under these conditions, we moved to the method described by Kennedy 13 whereby 2,3- dicyanonaphthalene was reacted with BCl 3 in dried 1,2-dichlorobenzene at 180 °C (Table S5.3 - Method 2.1). This reaction also produced the same five Cl-BsubNc compounds (Figure S5.2) albeit with a higher overall conversion. Their reported purification method - chromatography on Al 2O3 (IV) using toluene - also did not produce a chromatographically pure sample of Cl-BsubNc.

Based on the fact that Cl-BsubNc is vacuum sublimable, 28,32,34,37-40 we turned to the use of train sublimation 24 in an attempt to purify these sample of Cl-BsubNc. Train subliming crude Cl-BsubNc made via an adaptation of the Kennedy method (Table S5.3 – Method 2.3) was successful in removing most of the non-Cl-BsubNc impurities from the crude mixture after two successive train sublimation runs. While this technique was not successful in separating each of the five Cl-BsubNc compounds, the sample became suitable for analysis by low resolution mass spectroscopy (LRMS) and X-ray crystallography.

LRMS results from the sublimed Cl-BsubNc mixture showed five clusters of peaks of interest (Figure 5.2). The first cluster is consistent with the +BsubNc fragment of Cl-BsubNc ( m/z 545.2). The presence of this peak indicates that simple fragmentation of the axial chloride from Cl-BsubNc occurs using our MS method. The second cluster is consistent with the mass of Cl- BsubNc ( m/z 580.1) however the ratio of the isotopic peaks does not. The presence of the remaining higher m/z peaks is explainable with that observation. The third, fourth, and fifth clusters at progressively higher m/z ratios are consistent with mono-chlorinated ( m/z 614.1), dichlorinated (m/z 648.1), and tri-chlorinated ( m/z 682.0) Cl-BsubNc derivatives, which we will from here on refer to as Cl-Cl 1BsubPc, Cl-Cl 2BsubNc, and Cl-Cl 3BsubNc, respectively. Assuming that each chlorinated BsubNc also fragments, the presence of the respective fragment

102

one cluster lower explains the inconsistency in the isotopic ratios mentioned above. The ratio of the isotopic peaks for Cl-Cl 3BsubNc is consistent with the structure indicating that the fragment of a higher chlorinated species is not present. Due to the relative impurity of the sample ( i.e. mixture of BsubNc compounds), a HRMS analysis could not be done. It should be noted that the number of Cl-Cl nBsubNc products via LRMS ( i.e. 4) is one lower than that by HPLC ( i.e. 5).

This could likely be explained by a set of structural isomers having slightly different R ts (6.9 and 7.1 min).

Another possible explanation is that a MS peak for a fifth Cl-Cl nBsubNc was not observed due to its intensity being below the detection limit of our LRMS instrument. Regardless of the analysis, the sample is a mixture consisting of differing amounts of peripheral chlorinates.

The train sublimation process also produced crystals suitable for X-ray diffraction. While this does mark the first time a diffractable crystal of Cl-BsubNc has been analyzed and reported (Figure 5.3a), what was of more interest was the fact that the crystal was also shown to contained mixtures of Cl-BsubNc and chlorinated Cl-BsubNcs. We will therefore from here on refer to this crystal as literature-Cl-Cl nBsubNc (Cl-Cl nBsubNc produced from the literature procedure). The X-ray diffraction analysis indicated that the crystal was a mixture of compounds with differing levels of peripheral chlorination, whereby the chlorine atoms were only found to occupy the bay positions (C4, C4A, C9, C9A, C16, C16A) of the BsubNc molecular fragment. The structure was also found to be symmetric and hence, there are three different sites for potential chlorine occupancies (C4, C9, and C16 and their symmetric sites). The probability of a chlorine atom at C4, C9, and C16 was found to be 23.6%, 16.9%, and 16.1%, respectively, resulting in a total occupancy of 1.13 peripheral chlorines per molecule and an average molecular formula of

C36 H16.87 BCl 2.13 N6.

103

Figure 5.2. Mass spectrum of a sublimed sample of Cl-Cl nBsubNc made via an adaptation of the Kennedy method indicating the presence of Cl-Cl 1BsubPc, Cl-Cl 2BsubNc, and Cl-Cl 3BsubNc.

104

Thereafter we set out to develop synthetic conditions that might limit the amount of bay position chlorination. It has been reported that peripheral chlorination (or halogenation) can be reduced by using an electron-rich aromatic co-solvent to scavenge any chlorinating species. 45,46 For this role, we chose 1-methylnaphthalene (1-MNAP) as a co-solvent 46 and

2,3-dicyanonaphthalene was treated with BCl 3 in a 1,2,4-trichlorobenzene:1-MNAP mixture (4:1 v/v ) at 130 °C (Table S5.4 - Method 3.6). The introduction of 1-MNAP into the process completely suppressed the formation of three of the five Cl-BsubNc peaks leading to just two Cl-BsubNcs being present in the HPLC analysis (Figure S5.3a). Moreover, the integration of the peak at 3.6 min is much greater than the peak at 4.4 min which is opposite to what is observed above. Attempts to isolate a pure sample of Cl- BsubNc were still however unsuccessful. We tried train sublimation to purify the mixture. Like the previous example (Method 2.3), the technique was not successful in separating the two Cl-BsubNc compounds. Moreover, the sublimation yield was found to be extremely low (highest obtained yield of 3.7%) making it very challenging and laborious to obtain adequate amounts of sublimed material. However, a LRMS analysis (Figure S5.3b) on the sublimed mixture was consistent with the target Cl-BsubNc ( m/z 580.1) and a trace of Cl-Cl 1BsubNc ( m/z 614.1) This result combined with the greater HPLC peak intensity at 3.6 min versus 4.4 min allowed for the identification of the Cl-BsubNcs at 3.6 and 4.4 min to be Cl-BsubNc (non-chlorinated) and Cl-Cl 1BsubNc (mono-chlorinated), respectively.

The low yield of this process is not entirely attributable to the sublimation step. While the conversion by HPLC was measured to be ~58% this does not accurately reflect the actual conversion. A considerable amount of black, insoluble mass is formed during the reaction, making it necessary to do a filtration and wash with hot toluene to separate the black mass from the extracted Cl-BsubNc products.

105

Figure 5.3 . Ellipsoid plot (50 % probability) showing the structure and atom numbering scheme of Cl-Cl nBsubNc crystals obtained from (a) the literature method (CCDC deposition number: 1452383) and (b) the nitrobenzene method (CCDC deposition number: 1452384). Hydrogen atoms have been omitted for clarity. Colors: boron - orange; nitrogen - blue; carbon - black; chlorine - green.

We therefore next moved to reduce the amount of black mass produced during the process. Our hypothesis was that the Lewis acidic BCl 3 was promoting the polymerization of 2,3-dicyanonaphthalene and producing the black mass. We scoped the reaction conditions whereby the molar concentration of 2,3-dicyanonaphthalene, the molar equivalencies of the boron template, and the solvent system were changed (Table S5.5 - Methods 4.1-4.15). Unfortunately, none of the attempts were successful in producing a non-chlorinated Cl-BsubNc; either a mixture of Cl-Cl nBsubNcs or no Cl-BsubNc was formed and in all cases a considerable amount of black mass was still produced. We also added ethylene glycol to the reaction mixture in an attempt to lower the reactivity/Lewis acidity of BCl 3 through a chloride-oxygen ligand displacement reaction(s) (Table S5.6 – Methods 5.1-5.10). In short, none of the reactions were successful in producing

106

exclusively the desired, non-peripherally chlorinated Cl-BsubNc and nor did it they reduce the amount of black mass produced.

Considering that the addition of 1-MNAP as a chlorinating species scavenger was the most promising by producing just two Cl-Cl nBsubNcs, we turned to the exploration of other potentially more effective chlorinating species scavengers. Potential scavengers were selected to be liquids with a boiling point greater than 180 ˚C (the maximum temperature that the reaction was heated to) and since the exact mechanism of the chlorination is not known, scavengers were selected to specifically target Cl + (cationic) and Cl • (radical) species. Four scavengers were selected: dipentene (Cl +), (-)-β-pinene (Cl +), 1-octadecene (Cl +), and p-cymene (Cl + and Cl •) (Table S5.7 – Methods 6.1-6.4). The addition of 1-octadecence and dipentene resulted in complete stalling of the reaction; no Cl-BsubNc was produced. The reactions containing either (-)-β-pinene or p-cymene resulted in the formation of the same two Cl-BsubNc compounds (R t = 3.6 and 4.4 min,

Cl-BsubNc and Cl-Cl 1BsubNc) that were observed in the 1-MNAP experiment. The apparent conversion by HPLC was poor for the (-)-β-pinene experiment (17%) and higher for the p-cymene experiment (77%), however there was still a considerable amount of black mass produced. A comparison of the HPLC analysis for the 1-MNAP (Figure S5.3a) and p-cymene process (Figure S5.4a) shows that the intensity of the peak at 4.4 min (Cl-Cl 1BsubNc) relative to the peak at 3.6 min (Cl-BsubNc) is lower for the p- cymene reaction. It therefore appears that p-cymene is slightly more effective at scavenging chlorine species than 1-MNAP. For this reason, p-cymene was investigated furthermore.

Sublimation of the crude Cl-BsubNc material made via the p-cymene process was still very low yielding (highest yield obtained was 6.1%), but not as low as the 1-MNAP case (highest yield obtained was 3.7%). Furthermore, the technique was again not successful in separating the two BsubNc compounds (as indicated by HPLC analysis). LRMS analysis (Figure S5.4b) on the sublimed mixture was consistent with Cl-BsubNc ( m/z 580.1). A

107

MS peak for Cl-Cl 1BsubNc ( m/z 614.1) species was not observed but this was likely due to its intensity being below the detection limit of our LRMS instrument.

We then set out to synthesize a comparative Cl-Cl nBsubNc sample with a maximized level of peripheral chlorination. For this purpose, we chose an aromatic solvent with limited or no chlorine scavenging properties. Three different electron-poor solvents - nitrobenzene (NB), 1,2,4-trichlorobenzene, and 2,4-dinitro-1-fluorobenzene - were chosen (Table S5.8 - Methods 7.1-7.4). The experiment done in NB was the most interesting as it produced a significantly different HPLC chromatogram than any of the aforementioned processes (Figure S5.5a). Five BsubNc compounds were detected (R t = 4.4, 5.4, 6.7, 6.9, and 8.7 min) like the literature process, but the peak at 3.6 min is absent and a new peak at 8.7 min is formed. Moreover, the peaks at the higher R ts are more intensive than those at the lower R ts, indicating that this mixture is composed of a higher composition of chlorinated species. Like all cases above, train sublimation was not successful in separating any of the BsubNc compounds. Nevertheless, sublimation of the crude product was found to be higher-yielding (average: 10 %) than the p-cymene or 1- MNAP process. LRMS analysis on the sublimed mixture was consistent with the presence of Cl-Cl 1BsubNc ( m/z 614.1), Cl-Cl 2BsubNc ( m/z 648.1), Cl-Cl 3BsubNc ( m/z 682.0), Cl-

Cl 4BsubNc ( m/z 716.0), and Cl-Cl 5BsubNc ( m/z 749.9) (Figure S5.5b). It should be emphasized that the non-chlorinated Cl-BsubNc was not formed in the NB process as evident in the absence of its HPLC peak at 3.6 minutes (Figure S5.5a) and the m/z 580.1 peak (Figure S5.5b).

The sublimation process also produced single crystals suitable for X-ray diffraction (Figure 5.3b). Like the previous analysis outlined above, the X-ray analysis of the crystals determined they were a mixture of BsubNcs with differing levels of peripheral chlorination, where the chlorine atoms were present exclusively at the bay position. Unlike the previous sample for literature-Cl-BsubNc, the level of peripheral chlorination is much higher for the nitrobenzene-Cl-BsubNc (2.96 vs. 1.13 peripheral chlorines per molecule). This results in an average molecular formula of C 36 H15.04 BCl 3.96N6 and is

108

consistent with the LRMS results presented above. Beyond this analysis, both were orthorhombic crystals of the Pnma space group. Unit cell dimensions were found to only be different by a marginal amount (0.28 Å, 0.09 Å, and 0.38 Å for a, b, and c respectively, Table S5.9). The crystals from the nitrobenzene process were of 0.07 Mg cm -3 higher calculated density owing to the higher chlorine content. Additionally, having two crystal structures provides a point of comparison for the respective solid-state arrangements as well. It was determined that each crystal had the same solid-state arrangement (Figure S5.6 & S5.7). Again, it is important to note, these are diffractable crystals consisting of a mixture of compounds.

To summarize the synthetic process development, we were able to produce four samples of Cl-(Cl x)BsubNc with varying degrees of peripheral chlorination; a sample of Cl- BsubNcs made via the Kennedy’s method is a mixture of three compounds: Cl-BsubNc,

Cl-Cl 1BsubNc, and Cl-Cl 2BsubNc; a sample using the 1-MNAP procedure is a similar mixture of two compounds: Cl-BsubNc and Cl-Cl 1BsubNc; a sample using the p-cymene method which is also a mixture of two compounds: Cl-BsubNc and Cl-Cl 1BsubNc; and a sample using the nitrobenzene (NB) procedure is a mixture of five compounds: Cl-

Cl 1BsubNc, Cl-Cl 2BsubNc, Cl-Cl 3BsubNc, Cl-Cl 4BsubNc, and Cl-Cl 5BsubNc. We will term each of these samples “literature-Cl-Cl nBsubNc”, “1-MNAP-Cl-Cl nBsubNc”, “ p- cymene-Cl-Cl nBsubNc”, and “nitrobenzene-Cl-Cl nBsubNc, respectively. In the 1-MNAP- and p-cymene-Cl-Cl nBsubNc samples, the relative amount of Cl-BsubNc and Cl-

Cl 1BsubNc are slightly different. Finally, we were only able to produce adequate amounts of train sublimed samples of “literature-Cl-Cl nBsubNc”, “p-cymene-Cl-Cl nBsubNc”, and

"nitrobenzene-Cl-Cl nBsubNc" for further investigation. "1-MNAP-Cl-Cl nBsubNc" could not be obtained due to its very poor sublimation yield (highest obtained yield = 3.7%).

Numerous attempts were made to optimize the yield of 1-MNAP-Cl-Cl nBsubNc, but none of them could produce a yield greater than 3.7%. We surmise that the low sublimation yields are attributed to a narrow temperature window for the desired sublimation process to occur and the undesired decomposition process. The latter is based on the fact that the residual material in the boat following a sublimation process is black, insoluble char.

109

Finally and more generally, we believe based on our results there is little prospect for the development of a synthetic or purification process that will yield pure Cl-BsubNc.

Analysis of Commercial Cl-BsubNc Sample We also acquired a sample of commercial Cl-BsubNc (termed “commercial-Cl-

Cl nBsubNc”) and analyzed it by HPLC and LRMS. HPLC (Figure S5.8a) revealed the presence of the same five BsubNc compounds produced from the literature-Cl-Cl nBsubNc process. Also, the LRMS analysis results of commercial-Cl-Cl nBsubNc matched that of the literature-Cl-Cl nBsubNc (Figure S5.8b). Combined, these analyses demonstrate that the commercial source of Cl-Cl nBsubNc was not a pure sample, rather a mixture of peripherally chlorinated BsubNc products.

X-ray Photoelectron Spectroscopy (XPS) Analysis Core level XPS (Cl 2p) analysis was also performed on thermally evaporated, vacuum- deposited thin films of the Cl-Cl nBsubNcs (Figure 5.4, Table S5.10, and Figure S5.9). For all Cl-Cl nBsubNcs, two distinct chemical states are found for chlorine, chlorine bonded to boron (Cl-B, binding energy = 199.5-199.6 eV) and chlorine bonded to carbon (Cl-C, binding energy = 200.5-200.8 eV). These results are consistent with those of Kahn et al . who found two distinct chlorine peaks at 199.3 eV and 200.5 eV. 44 XPS analysis for boron and nitrogen (Figure S5.10, S5.11, S5.12) are also consistent with the chemical structure of BsubNc.

Bay position chlorination was further quantified by examining the deconvoluted XPS Cl peak areas (Table S5.11) and their relative ratios (Cl-C/Cl-B). The ratios are found to be 1.21, 1.61, 0.18, and 4.17 for literature-, commercial-, p-cymene-, and nitrobenzene-Cl- 44 Cl nBsubNc, respectively, while Kahn et al . found a ratio of ~1.5 for commercial Cl- BsubNc. Alternatively, the number of bay position chlorine atoms can also be calculated by considering the Cl total /N total ratio by multiplying by 6 ( i.e. the total number of nitrogens per molecule) and subtracting by 1 ( i.e. the number of chlorine bonded to boron). This approach produces similar numbers of 1.40, 1.76, 0.32, and 4.40 for literature-,

110

commercial-, p-cymene-, and nitrobenzene-Cl-ClnBsubNc, respectively (Table S5.11). Overall, these measurements are consistent with the differing degree of bay position chlorination among the samples.

Figure 5.4. Overlay of the literature-Cl-Cl nBsubNc (blue), commercial-Cl-Cl nBsubNc (red), p- cymene-Cl-Cl nBsubNc (green), and nitrobenzene-Cl-Cl nBsubNc (purple) normalized Cl 2p core- level XPS spectra.

1H NMR analysis 1 H NMR analysis was carried out for the four synthetic Cl-Cl nBsubNc samples, each doubly sublimed along with the commercial sample (Figure S5.13). For literature-, 1- 1 MNAP-, and p-cymene-Cl-Cl nBsubNc, the H NMR spectra are very similar in terms of the chemical shift and multiplicity. Integration of the bay position 1H resonances found at ~ 9.45 ppm indicates an increased presence of the bay position protons for the 1-MNAP-

Cl-Cl nBsubNc and p-cymene-Cl-Cl nBsubNc, an observation consistent with the preceding data. A noticeable observation is the presence of additional peaks ( i.e. impurities) in the 1 spectrum of commercial-Cl-Cl nBsubNc. The H NMR spectrum of nitrobenzene-Cl-

Cl nBsubNc, on the other hand, is very different compared to the other three spectra for the synthetic samples. There are additional peaks in the aromatic region consistent with a

111

large number of chlorinated BsubNc isomers with asymmetry. Assignment of the resonances and quantification is problematic. Kahn et al . have briefly reported similar results on the purity of a commercial source of Cl-BsubNc via 1H NMR. They had concluded the same and that the commercial Cl-BsubNc sample is a mixture of chlorinated species and outlined that full detailed NMR investigation is currently underway within their group. 44

Photophysical Analyses of Cl-Cl nBsubNcs Given there is a discrepancy in the literature regarding the photophysical properties of 9,18 BsubNcs, UV-vis absorption spectra of literature-Cl-Cl nBsubNc, commercial-Cl-

Cl nBsubNc, p-cymene-Cl-Cl nBsubNc, and nitrobenzene-Cl-Cl nBsubNc were acquired in toluene solutions at room temperature (Table 5.1 and Figure S5.14a). Due to the challenges associated with subliming 1-MNAP-Cl-Cl nBsubNc, this compound was excluded from this photophysical and subsequent study. The absorption profiles of literature-Cl-Cl nBsubNc and commercial-Cl-Cl nBsubNc are found to be nearly identical, both with a λ max value of 656 nm. In contrast to this, p-cymene-Cl-Cl nBsubNc has a blue- shifted λ max value of 651 nm while nitrobenzene-Cl-Cl nBsubNc has a red-shifted λ max value of 664 nm. The shift in the λ max,abs values is not surprising to us as we have observed a similar effect for the analogous Cl-BsubPc, whereby peripheral chlorination red-shifts the λ max,Abs : Cl-BsubPc, Cl-Cl 6BsubPc, and Cl-Cl 12 BsubPc at 565, 574, and 593 nm, respectively (Table S5.12). Torres et al . have reported a higher λ max,Abs value of 661 18 nm for their Cl-BsubNc. The extinction coefficients (ε) for the four Cl-ClnBsubNcs were found to be about the same and also consistent with the value reported by Torres et al 18 .

The photoluminescence (PL) spectra of the four Cl-Cl nBsubNc samples were also acquired in toluene solutions at room temperature (Table 5.1, Figure S5.14b). Like the results from the absorption study, the PL spectra are nearly identical for both literature-

Cl-Cl nBsubNc and commercial-Cl-Cl nBsubNc (λ max,PL = 666 nm) while a blue-shift in the spectrum of p-cymene- (λ max = 662 nm) and a red-shift in the spectrum of nitrobenzene-

112

Cl-Cl nBsubNc (λ max = 674 nm) is observed. Despite the differences in the λ max of absorption and emission spectra, all the Stokes shifts are found to be similar (10-11 nm).

Torres et al . has reported a slightly more red-shifted λ max,PL of 677 nm for their Cl- BsubNc, resulting in a marginally larger measured Stokes shift of 16 nm. 37 Fluorescence quantum yields (Φ PL ) were also measured and they were found to be between 24 and 29 % relative to a standard of oxazine 170 (Table 5.1, Eq S5.1). These values are in line with 18 that reported by Torres et al . (ΦPL = 0.22) .

Table 5.1 . Photophysical properties of Cl-Cl nBsubNcs.

a a,c a a,d Compound λmax,abs | Extinction Coefficient λmax,PL | Stokes Shift | ΦPL b -1 -1 b,c b λmax,abs (nm) (e , M cm ) λmax,PL (nm) Stokes Shift (%) (nm) Literature-Cl-Cl nBsubNc 656 | 688 78,400 666 | 746 10 | 58 27 Commercial-Cl-Cl nBsubNc 656 | 691 78,400 666 | 738 10 | 47 24 p-Cymene-Cl-Cl nBsubNc 651 | 685 79,200 662 | 746 11 | 61 29 Nitrobenzene-Cl-Cl nBsubNc 664 | 698 77,800 674 | 750 10 | 52 21 Torres' Cl-BsubNc 661 e 79,400 e 677 e 16 22 e a b c d In toluene solution. Solid state (film). λexc = 630 nm. Relative to a oxazine 170 standard using an excitation wavelength of 630 nm. e Data taken from Torres et al .18

Solid state (film) UV-vis absorption and PL spectra were also acquired and compared to the solution state (Table 5.1 and Figure S5.15). As expected, the spectra are broader but still very similar in shape and form. A moderate bathochromic shift of 32-35 nm in the absorption spectra and an even greater bathochromic shift of 72-84 nm in the PL spectra are observed relative to their respective solution state spectra. Large Stokes shifts of 47- 61 nm are seen in the solid state and are indicative of intermolecular organization and aggregation.

Electrochemical Analyses of Cl-Cl nBsubNcs

The electrochemical properties of the four Cl-Cl nBsubNc samples were then analysed via cyclic voltammetry (CV) in degassed DCM solution containing 0.1 M tetrabutylammonium perchlorate (TBAPC) (Table 5.2, Figure S5.16a-S5.18a). All potentials (E) were corrected to the half-wave reduction potential (E 1/2,red ) of

113

decamethylferrocene (-0.012 V vs. Ag/AgCl 47). Literature-, commercial-, and p-cymene- 1 Cl-Cl nBsubNc have a nearly identical reversible first (E ox ) oxidation and an irreversible 2 second (E ox ) oxidation. On the reduction side, a single pseudo-reversible reductive process is also observed. Unlike what is observed in the oxidation regime, the reduction potentials are found to be different depending on the nature of the BsubNc mixture; literature- (-0.87 V) and commercial-Cl-Cl nBsubNc (-0.89 V) shared a similar reduction 1 potential (E red ) while p-cymene Cl-Cl nBsubNc has a much higher potential (-0.99 V). A cyclic voltammogram of nitrobenzene-Cl-Cl nBsubNc could not be obtained due to poor signal-to-noise ratio likely caused by its observed poor solubility in DCM. While chlorination did not impact the oxidative potential and behavior of Cl-Cl nBsubNcs, chlorination did lower the reduction potential.

Table 5.2 . Electrochemical properties of Cl-Cl nBsubNcs.

1 2 a 1 a 1 2 b 1 2 3 b Compound E ox | E ox (V) E red (V) E ox | E ox (V) E red | E red | E red (V) d c c c c ce ce Literature-Cl-Cl nBsubNc +0.83 | +1.28 -0.87 +0.84 -0.92 | -1.06 | -1.36 d c c c c ce ce Commercial-Cl-Cl nBsubNc +0.84 | +1.29 -0.89 +0.85 -0.91 | -1.07 | -1.37 d c c c c c ce ce p-Cymene-Cl-Cl nBsubNc +0.82 | +1.28 -0.99 +0.75 | +1.11 -0.94 | -1.06 | -1.38 c c c ce Nitrobenzene-Cl-Cl nBsubNc - - +0.86 | +1.11 -0.81 | -1.28 Nonell et al. - - +0.68 f -0.85 f E = redox potential from CV. a In degassed DCM solution relative to Ag/AgCl. b In degassed DMF solution relative to Ag/AgCl. c Peak potential. d Half-wave potential. e Minor peak. f Data taken from Nonell et al .51

Due to the poor solubility of nitrobenzene-Cl-Cl nBsubNc in DCM solution, CV analyses of the four samples were thereafter redone in degassed N,N -dimethylformamide (DMF) solution (Table 5.2 and Figure S5.19a-S5.22a). Within DMF, all redox events were found to be irreversible or pseudo-reversible at best. Literature- and commercial-Cl-Cl nBsubNc 1 1 2 3 have nearly the same voltammograms with similar E ox , E red , E red , and E red values while noticeable differences are observed for p-cymene- and nitrobenzene-Cl-Cl nBsubNc. For 1 2 p-cymene-Cl-Cl nBsubNc, it has a lower E ox (+0.75 V) and also has a E ox (+1.11 V). In the reduction regime, there are three reductive events (-0.94, -1.06, -1.38 V). For 1 nitrobenzene-Cl-Cl nBsubNc, there are two oxidative events where E oxid (+0.86 V) and 2 E oxid (+1.11 V) are in line with the values seen for literature-/commercial-Cl-Cl nBsubNc and p-cymene-Cl-Cl nBsubNc, respectively. On the reduction side, two peaks (-0.81, -1.28

114

1 V) are noted with E red being lower than that of literature-, commercial-, or p-cymene-Cl-

Cl nBsubNc. Overall, these results do not totally match the CV results obtained in DCM suggesting that the electrochemical properties of Cl-Cl nBsubNc is solvent-dependent as expected. 48-50 Additionally, these results are not consistent with the electrochemical properties of Cl-BsubNc previously reported by Nonell et al . measured in degassed DMF with 0.1 M tetrabutylammonium hexafluorophosphate (TBAHF) 51 using Cl-BsubNc prepared according to the procedure published of Torres et al. 18 This sample likely has the same level of peripheral chlorination as our literature- and commercial-Cl-Cl nBsubNc samples. A single irreversible oxidation at +0.68 V and a pseudo-reversible reduction at - 1 1 0.85 V were reported, both lower than the E ox and E red of literature- or commercial-Cl-

Cl nBsubNc measured herein.

Table 5.3 . Electrochemical properties (DPV) and ultraviolet photoelectron spectroscopy (UPS) characteristics of Cl-Cl nBsubNcs.

1' 2' a 1' 2' 3' a d Compound E ox | E ox (V) E red | E red | E red (V) IE (HOMO) (eV) b bc b b b Literature-Cl-Cl nBsubNc +0.75 | +1.00 -0.87 | -1.02 | -1.33 5.32 b bc b b b Commercial-Cl-Cl nBsubNc +0.76 | +1.02 -0.86 | -1.01 | -1.29 5.31 b b b b b p-Cymene-Cl-Cl nBsubNc +0.69 | +1.01 -0.93 | -1.08 | -1.35 5.29 b b b bc bc Nitrobenzene-Cl-Cl nBsubNc +0.78 | +1.02 -0.79 | -0.98 | -1.26 5.42 Kahn et al . - - 5.35 e E' = redox potential from DPV. a In degassed DMF solution relative to Ag/AgCl. b Peak potential. c Minor peak. d Measured by UPS. e Data taken from Kahn et al .44

Differential pulse voltammetry (DPV, 0.1 M TBAPC) was also performed (Table 5.3, Table S5.13, and Figure S5.16b-S5.22b). DPV redox potentials (E') were found to be generally lower than CV redox potentials (E) in either DCM or DMF. In DCM, additional peaks were detected via DPV in the reduction region for p-cymene- and commercial-Cl-

Cl nBsubNc (Figure S5.19b and S5.20b, respectively). In DMF, additional peaks were detected via DPV in the oxidation region for literature- and commercial-Cl-Cl nBsubNc and in the reduction region for nitrobenzene-Cl-Cl nBsubNc (Figure S5.21b, S5.24b, and S5.23b, respectively). Despite these differences, the findings within the DPV dataset are in line with those within the CV dataset of the same solvent system.

115

Considering that the four samples under study are mixtures ( i.e. not a single, pure compound) it should be noted that this would or could affect the electrochemical signals' dimensions (broadness, height, resolution, etc.) and thus, the apparent peak potential values. For this reason, three other processing methods were performed on the DPV dataset to determine if the results acquired from the apex peak method are plausible (Table S5.14) and in short, these further studies supported the overall trends. Further details about these processing methods can be found in the appendix.

Ultraviolet Photoelectron Spectroscopy (UPS) Analysis

Given the application of interest for Cl-Cl nBsubNcs is within thin solid film of an organic photovoltaic device (OPV), we then examined each of the four mixtures using ultraviolet photoelectron spectroscopy (UPS). UPS was performed on thermally evaporated, vacuum-deposited thin films to determine the HOMO energy level ( i.e. ionization energy) of each sample and to further demonstrate their electronic property differences (Figure 5.5, Table 5.3, and Figure S5.23). The HOMO level is found to be the shallowest for p- cymene-Cl-Cl nBsubNc (5.29 eV), about the same for literature- (5.32 eV) and commercial-Cl-Cl nBsubNc (5.31 eV), and the deepest for nitrobenzene-Cl-Cl nBsubNc (5.42 eV). Kahn et al . have also reported a very similar value of 5.35 eV for a commercial sample of Cl-BsubNc. 44 A trend where the HOMO level deepens with increasing chlorination is evident here. A similar effect was reported for BsubPc, where the UPS- determined HOMO level shifted away from the vacuum line with increasing fluorination. 24 Also, the peak at a binding energy of ~4.5 eV is more pronounced in the nitrobenzene-Cl-Cl nBsubNc sample, an observation that is likely attributed to the higher chlorinated species. Further UPS details can be found in the appendix.

116

Figure 5.5. Overlay of the literature-Cl-Cl nBsubNc (blue), commercial-Cl-Cl nBsubNc (red), p- cymene-Cl-Cl nBsubNc (green) and nitrobenzene-Cl-Cl nBsubNc (purple) normalized UPS valence band spectra.

OPV Device Integration of Cl-Cl nBsubNcs Although we could not purify any of our synthetic samples to an absolutely pure single Cl-BsubNc compound, we did examine the performance of each within a PHJ OPV to determine if there are any performance differences attributable to the differing amounts of peripheral chlorination. Given that diffractable crystals were obtained via sublimation and XPS results obtained from sublimed films were consistent, we were not concerned with any alteration of the chemical composition on fabrication of PHJ OPVs. We also incorporated the “commercial-Cl-Cl nBsubNc” into OPVs as an additional point of comparison.

Four sets of OPV devices were therefore fabricated, pairing each sample of Cl-Cl nBsubNc as the electron acceptor with sexithiophene (α-6T) as an electron donor, with the following device configuration: indium tin oxide (ITO)/poly(3,4- ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS)/sexithiophene (α-6T, 55 nm)/Cl-Cl nBsubNc(25 nm)/bathocuprione (BCP, 10 nm)/silver (Ag, 80 nm). These

117

devices mimic our previous work using Cl-BsubPc paired with α-6T 52 and are consistent with and comparable to the PHJ OPVs fabricated by Cnops et al .32 Current density- voltage (J-V) characteristics (Figure 5.6a) were measured under 100 mW·cm -2 of simulated solar illumination (AM1.5). The measured external quantum efficiency (EQE) spectra are shown in Figure 5.6b. Data is derived from the measurement of at least 6 PHJ OPV devices and therefore the standard deviation for each device metric is also presented (Table 5.4).

Figure 5.6 . (a) J/V characteristics and (b) external quantum efficiency (EQE) spectra of p- cymene-Cl-Cl nBsubNc, literature-Cl-Cl nBsubNc, and commercial-Cl-Cl nBsubNc. PHJ OPVs of the following configuration: ITO/PEDOT:PSS/α-6T(55 nm)/Cl-Cl nBsubNc(25 nm)/BCP(10 nm)/Ag(80 nm) where the Cl-Cl nBsubNc layer is made of nitrobenzene-Cl-Cl nBsubNc p- cymene-Cl-Cl nBsubNc, literature-Cl-Cl nBsubNc, or commercial-Cl-Cl nBsubNc.

The OPVs made with literature-Cl-Cl nBsubNc and commercial-Cl-Cl nBsubNc samples perform nearly identically. The performance of these devices fabricated in our laboratory is not quite as high as for those made by Cnops et al. 32, but the losses can be attributed to a less extensive optimization of the layer thicknesses undertaken in our fabrication system. A much lower performance is achieved for devices made with the p-cymene-Cl-

Cl nBsubNc sample, with less bay position chlorination, where both the short-circuit

118

current (J SC ) and the fill factor (FF) fall significantly short of the OPVs derived from Cl-

Cl nBsubNcs with higher levels of chlorination. Conversely a small yet statistically 44 relevant increase in V OC was observed, as was suggested by Khan et al . Now

considering nitrobenzene-Cl-Cl nBsubNc, while we saw an equally small reduction in the

VOC of the PHJ OPV and a small reduction in the J SC , a notable increase in the fill factor

therefore resulted in the highest PCE. The observed small changes in V OC are in line and proportional with the measured HOMO by UPS.

Table 5.4 . Average device parameters of PHJ OPVs of the following configuration:

ITO/PEDOT:PSS/α-6T(55 nm)/Cl-Cl nBsubNc(25 nm)/BCP(10 nm)/Ag(80 nm) whereby the Cl-

Cl nBsubNc layer is made of nitrobenzene-Cl-Cl nBsubNc, p-cymene-Cl-Cl nBsubNc, literature-Cl-

Cl nBsubNc, or commercial-Cl-Cl nBsubNc. The standard deviation is indicated in brackets next to each parameter. Each parameter is a result of the measurement of at least 6 devices.

a b -2 c d Acceptor VOC (V) JSC (mA cm ) FF PCE (%) p-Cymene-Cl-Cl nBsubPc 1.004 (0.009) 6.05 (0.07) 0.40 (0.01) 2.44 (0.13)

Literature-Cl-Cl nBsubPc 0.972 (0.018) 8.92 (0.37) 0.45 (0.01) 3.90 (0.12)

Commercial-Cl-Cl nBsubPc 0.976 (0.005) 8.91 (0.13) 0.46 (0.01) 3.96 (0.07)

Nitrobenzene-Cl-Cl nBsubPc 0.930 (0.002) 8.80 (0.19) 0.53 (0.01) 4.32 (0.11) aOpen-circuit voltage, bShort-circuit current, cFill factor, dPower conversion efficiency

These results suggest that inclusion of the bay position chlorinated Cl-Cl nBsubNc derivatives in a mixture with Cl-BsubNc improves the overall performance of the layer in OPVs. This is conceptually similar to the work of Fleetham et al . who showed an enhancement in OPV performance using a mixture of peripherally chlorinated zinc phthalocyanine (ZnPc) and pure ZnPc. 42,43

Traditionally it has been believed that impurities in organic semiconductors will always act as traps, hindering device performance. 53,54 Street et al . have recently challenged that notion, demonstrating that different materials, specifically fullerene derivatives, can be blended together to achieve an alloying effect of their electronic properties, so long as their size allows them to easily intermix without disrupting the local order of one another. 55,56 The crystal structure of

119

literature-Cl-Cl nBsubNc shows co-crystallization of the different bay position chlorinated Cl- BsubNc compounds, indicating that the local order is not disrupted by the presence of bay position chlorination. Therefore, an alloying effect similar to what Street et al. 55,56 observed for fullerenes in bulk heterojunction OPVs could be occurring. This could explain why the performance of a Cl-Cl nBsubNc mixture in OPVs is not crippled by impurity trap states rather it is improved. While a comparison against absolutely pure Cl-BsubNc was not possible, and may not be possible, the strong performance in OPVs utilizing “Cl-BsubNc” reported in literature may in fact be caused by the alloying together of slightly different electronic states.

5.4 Conclusions In this paper, we outline proof that synthetic and commercial samples of Cl-BsubNc are actually a mixed alloyed chemical composition with random and significant amounts of chlorination in the bay position of the BsubNc chromophore. We also outline our efforts to prepare a single, pure Cl-BsubNc compound instead of a mixture of Cl-Cl nBsubNc compounds with differing levels of bay position chlorination. After realizing that this was unfeasible, we shifted our focus into the synthesis of Cl-Cl nBsubNc mixtures with less and more chlorination to ultimately determine any differences in the basic properties and performance within OPV devices. Single crystals of literature-Cl-Cl nBsubNc and nitrobenzene-Cl-Cl nBsubNc were obtained, analyzed, and were shown to have chlorines at the bay positions of the BsubNc molecular fragment with differing amounts (1.13 vs. 2.96 average chlorines per molecule respectively). Among three Cl-

Cl nBsubNcs and a commercial sample, the photo- and electro-physical change in properties were found to translated to differing performances in planar heterojunction organic photovoltaics (PHJ OPVs) when applied as an electron accepting material paired with α-sexithiophene (a-6T, as an electron donor material). PHJ OPVs made from Cl-Cl nBsubNc with lower amounts of chlorine

(p-cymene-Cl-Cl nBsubNc) was found to perform poorly when compared against PHJ OPVs made from Cl-Cl nBsubNc with higher/highest amounts of chlorination (nitrobenzene-Cl-

Cl nBsubNc). Ultimately nitrobenzene-Cl-Cl nBsubPc was found to perform moderately better than Cl-Cl nBsubNc made via the literature procedure (literature-Cl-Cl nBsubNc) or a commercially-available Cl-Cl nBsubNc sample (commercial-Cl-Cl nBsubNc). However, we feel these rather small differences do not call into question the 8.4% efficient PHJ OPV results of

120

Cnops et al. 32, rather it offers a cautionary note as to the actual chemical composition of the “Cl- BsubNc” used previously in their study, other studies presented in the literature, and any moving forward. Our current work is focused on obtaining Cl-BsubNc with the bay positions chemically blocked so as to avoid random bay position chlorination.

121

5.5 References 1. Edwards, L.; Gouterman, M. Journal of Molecular Spectroscopy 1970 , 33 , 292-310. 2. Tang, C. W. Applied Physics Letters 1986 , 48 , 183-185. 3. Peumans, P.; Forrest, S. R. Applied Physics Letters 2001 , 79 , 126-128. 4. Meller, A.; Ossko, A. Monatshefte fuer Chemie 1972 , 103 , 150-155. 5. Kietaibl, H. Monatshefte fuer Chemie 1974 , 105 , 405-418. 6. Paton, A. S.; Morse, G. E.; Lough, A. J.; Bender, T. P. CrystEngComm 2011 , 13 , 914- 919. 7. Virdo, J. D.; Kawar, Y. H.; Lough, A. J.; Bender, T. P. CrystEngComm 2013 , 15 , 3187- 3199. 8. Morse, G. E.; Gong, I.; Kawar, Y.; Lough, A. J.; Bender, T. P. Crystal Growth & Design 2014 , 14 , 2138-2147. 9. Kobayashi, N.; Ishizaki, T.; Ishii, K.; Konami, H. Journal of the American Chemical Society 1999 , 121 , 9096-9110. 10. Claessens, C. G.; Gonzalez-Rodriguez, D.; del Rey, B.; Torres, T.; Mark, G.; Schuchmann, H.-P.; von Sonntag, C.; MacDonald, J. G.; Nohr, R. S. European Journal of Organic Chemistry 2003 , 2547-2551. 11. Gonzalez-Rodriguez, D.; Torres, T.; Guldi, D. M.; Rivera, J.; Herranz, M. A.; Echegoyen, L. Journal of the American Chemical Society 2004, 126 , 6301-6313. 12. Gonzalez-Rodriguez, D.; Torres, T.; Olmstead, M. M.; Rivera, J.; Angeles Herranz, M.; Echegoyen, L.; Atienza Castellanos, C.; Guldi, D. M. Journal of the American Chemical Society 2006 , 128 , 10680-10681. 13. Zyskowski, C. D.; Kennedy, V. O. Journal of Porphyrins and Phthalocyanines 2000 , 4, 707-712. 14. Potz, R.; Goldner, M.; Huckstadt, H.; Cornelissen, U.; Tutass, A.; Homborg, H. Zeitschrift fuer Anorganische und Allgemeine Chemie 2000 , 626 , 588-596. 15. Dang, J. D.; Virdo, J. D.; Lessard, B. H.; Bultz, E.; Paton, A. S.; Bender, T. P. Macromolecules 2012 , 45 , 7791-7798. 16. Rauschnabel, J.; Hanack, M. Tetrahedron Letters 1995 , 36 , 1629-1632.

122

17. Geyer, M.; Plenzig, F.; Rauschnabel, J.; Hanack, M.; Del Rey, B.; Sastre, A.; Torres, T. Synthesis 1996 , 1139-1151. 18. Nonell, S.; Rubio, N.; del Rey, B.; Torres, T. Journal of the Chemical Society, Perkin Transactions 2 2000 , 1091-1094. 19. Giribabu, L.; Kumar, C. V.; Surendar, A.; Reddy, V. G.; Chandrasekharam, M.; Reddy, P. Y. Synthetic Communications 2007 , 37 , 4141-4147. 20. Takao, Y.; Masuoka, T.; Yamamoto, K.; Mizutani, T.; Matsumoto, F.; Moriwaki, K.; Hida, K.; Iwai, T.; Ito, T.; Mizuno, T.; Ohno, T. Tetrahedron Letters 2014 , 55 , 4564- 4567. 21. Yamamoto, K.; Takagi, A.; Hada, M.; Taniwaki, R.; Mizutani, T.; Kumura, Y.; Takao, Y.; Moriwaki, K.; Matsumoto, F.; Ito, T.; Iwai, T.; Hida, K.; Mizuno, T.; Ohno, T. Tetrahedron 2016 , 72, 4918-4924. 22. Shimizu, S.; Miura, A.; Khene, S.; Nyokong, T.; Kobayashi, N. Journal of the American Chemical Society 2011 , 133 , 17322-17328. 23. Diaz, D. D.; Bolink, H. J.; Cappelli, L.; Claessens, C. G.; Coronado, E.; Torres, T. Tetrahedron Letters 2007 , 48 , 4657-4660. 24. Morse, G. E.; Helander, M. G.; Maka, J. F.; Lu, Z.-H.; Bender, T. P. ACS Applied Materials & Interfaces 2010 , 2, 1934-1944. 25. Helander, M. G.; Morse, G. E.; Qiu, J.; Castrucci, J. S.; Bender, T. P.; Lu, Z.-H. ACS Applied Materials & Interfaces 2010 , 2, 3147-3152. 26. Mutolo, K. L.; Mayo, E. I.; Rand, B. P.; Forrest, S. R.; Thompson, M. E. Journal of the American Chemical Society 2006 , 128 , 8108-8109. 27. Gommans, H.; Cheyns, D.; Aernouts, T.; Girotto, C.; Poortmans, J.; Heremans, P. Advanced Functional Materials 2007 , 17 , 2653-2658. 28. Gommans, H.; Aernouts, T.; Verreet, B.; Heremans, P.; Medina, A.; Claessens, C. G.; Torres, T. Advanced Functional Materials 2009 , 19 , 3435-3439. 29. Sullivan, P.; Duraud, A.; Hancox, I.; Beaumont, N.; Mirri, G.; Tucker, J. H. R.; Hatton, R. A.; Shipman, M.; Jones, T. S. Advanced Energy Materials 2011 , 1, 352-355. 30. Morse, G. E.; Gantz, J. L.; Steirer, K. X.; Armstrong, N. R.; Bender, T. P. ACS Applied Materials & Interfaces 2014 , 6, 1515-1524.

123

31. Beaumont, N.; Castrucci, J. S.; Sullivan, P.; Morse, G. E.; Paton, A. S.; Lu, Z.-H.; Bender, T. P.; Jones, T. S. Journal of Physical Chemistry C 2014 , 118 , 14813-14823. 32. Cnops, K.; Empl, M. A.; Heremans, P.; Rand, B. P.; Cheyns, D.; Verreet, B. Nature Communications 2014 , 5, 3406. 33. Morse, G. E.; Bender, T. P. ACS Applied Materials & Interfaces 2012 , 4, 5055-5068. 34. Verreet, B.; Schols, S.; Cheyns, D.; Rand, B. P.; Gommans, H.; Aernouts, T.; Heremans, P.; Genoe, J. Journal of Materials Chemistry 2009 , 19 , 5295-5297. 35. Ma, B.; Woo, C. H.; Miyamoto, Y.; Frechet, J. M. J. Chemistry of Materials 2009 , 21 , 1413-1417. 36. Heremans, P.; Cheyns, D.; Rand, B. P. Accounts of Chemical Research 2009 , 42 , 1740- 1747. 37. Cheyns, D.; Rand, B. P.; Heremans, P. Applied Physics Letters 2010 , 97 , 033301/033301-033301/033303. 38. Kulshreshtha, C.; Kim, G. W.; Lampande, R.; Huh, D. H.; Chae, M.; Kwon, J. H. Journal of Materials Chemistry A 2013 , 1, 4077-4082. 39. Chen, G.; Sasabe, H.; Sano, T.; Wang, X.-F.; Hong, Z.; Kido, J.; Yang, Y. Nanotechnology 2013 , 24 , 484007/484001-484007/484009. 40. Verreet, B.; Cnops, K.; Cheyns, D.; Heremans, P.; Stesmans, A.; Zango, G.; Claessens, C. G.; Torres, T.; Rand, B. P. Advanced Energy Materials 2014 , 4, 1301413/1301411- 1301413/1301418. 41. Menke, S. M.; Holmes, R. J. ACS Applied Materials & Interfaces 2015 , 7, 2912-2918. 42. Fleetham, T. B.; Bakkan, N.; Mudrick, J. P.; Myers, J. D.; Cassidy, V. D.; Cui, J.; Xue, J.; Li, J. Journal of Materials Science 2013 , 48 , 7104-7114. 43. Fleetham, T. B.; Mudrick, J. P.; Cao, W.; Klimes, K.; Xue, J.; Li, J. ACS Applied Materials & Interfaces 2014 , 6, 7254-7259. 44. Endres, J.; Pelczer, I.; Rand, B. P.; Kahn, A. Chemistry of Materials 2016 , 28 , 794-801. 45. Stork, J. R.; Potucek, R. J.; Durfee, W. S.; Noll, B. C. Tetrahedron Letters 1999 , 40 , 8055-8058.

124

46. Martin, G.; Rojo, G.; Agullo-Lopez, F.; Ferro, V. R.; Garcia de la Vega, J. M.; Martinez- Diaz, M. V.; Torres, T.; Ledoux, I.; Zyss, J. Journal of Physical Chemistry B 2002 , 106 , 13139-13145. 47. Noviandri, I.; Brown, K. N.; Fleming, D. S.; Gulyas, P. T.; Lay, P. A.; Masters, A. F.; Phillips, L. Journal of Physical Chemistry B 1999 , 103 , 6713-6722. 48. Kadish, K. M.; Morrison, M. M. Journal of the American Chemical Society 1976 , 98 , 3326-3328. 49. Tsierkezos, N. G. Journal of Solution Chemistry 2007 , 36 , 289-302. 50. Ajloo, D.; Yoonesi, B.; Soleymanpour, A. International Journal of Electrochemical Science 2010 , 5, 459-477. 51. Rubio, N.; Jimenez-Banzo, A.; Torres, T.; Nonell, S. Journal of Photochemistry and Photobiolology, A: Chemistry 2007 , 185 , 214-219. 52. Josey, D. S.; Castrucci, J. S.; Dang, J. D.; Lessard, B. H.; Bender, T. P. ChemPhysChem 2015 , 16 , 1245-1250. 53. Salzman, R. F.; Xue, J.; Rand, B. P.; Alexander, A.; Thompson, M. E.; Forrest, S. R. Organic Electronics 2005 , 6, 242-246. 54. Pai, D. M.; Yanus, J. F.; Stolka, M. Journal of Physical Chemistry 1984 , 88 , 4714-4717. 55. Street, R. A.; Davies, D.; Khlyabich, P. P.; Burkhart, B.; Thompson, B. C. Journal of the American Chemical Society 2013 , 135 , 986-989. 56. Street, R. A.; Khlyabich, P. P.; Rudenko, A. E.; Thompson, B. C. Journal of Physical Chemistry C 2014 , 118 , 26569-26576.

125

Chapter 6 Phenoxy boron subnaphthalocyanines: synthesis, properties, and their applications in organic planar heterojunction photovoltaics

The work in this chapter is currently being prepared for submission as a manuscript to RSC Advances . The author list will be: Jeremy D. Dang, David S. Josey, and Timothy P. Bender.

I carried out all synthetic experiments and characterized the products via NMR, optical absorption, photoluminesence, and voltammetry. DSJ performed all OPV device fabrication and testing. The work was supervised by TPB. All authors approved the manuscript.

Supplementary information for this chapter is found in Appendix E.

126

6.1 Abstract

The first set of phenoxy BsubNc compounds – PhO-Cl nBsubNc and F 5-Cl nBsubNc – are synthesized through an axial displacement reaction of Cl-Cl nBsubNc with phenol or pentafluorophenol. Like their precursor, the products are composed of a mixture of phenoxylated

Cl nBsubNcs with differing level of peripheral chlorinates. Single crystals of PhO-Cl nBsubNc were obtained, analyzed, and determined to have an average bay position chlorine occupancy of

0.99. While the phenoxy Cl nBsubNcs share similar photophysical behaviours to Cl-Cl nBsubNc, their electrochemical properties were found to be modified following phenoxylation. These differences are reflected in their poorer performance in PHJ OPV devices. However, their inclusion resulted in devices with higher open-circuit voltages, suggesting that higher efficiencies could be afforded with a better paired electron donating material. The work herein marks the first examples of the use of non-Cl-BsubNcs in OPVs. Contrary to BsubPcs, the axial derivatization of Cl nBsubNcs has a pronounced effect on the electrochemical properties and PHJ OPV performance.

6.2 Introduction Boron subnaphthalocyanines (BsubNcs) are lower homologues of naphthalocyanines (Ncs) composed of three N-fused 1,3-diiminobenzoisoindoline subunits around a single, central boron atom (Figure 6.1). They are formed from a cyclotrimerization reaction of 2,3- dicyanonaphthalene with a boron template ( e.g. PhBCl 2, BBr 3, BCl 3) in an aromatic solvent or solvent mixtures under an inert atmosphere (Scheme 6.1). Like the boron subphthalocyanines (BsubPcs, Figure 6.1), which are instead composed of 1,3-diiminoisoindoline subunits, BsubNcs are non-planar, cone-shaped, and aromatic in structure. 1-6 Unlike the BsubPcs, BsubNcs have a longer π-conjugation system, causing their absorption spectra to be heavily red-shifted; the Q band is found in the mid and high 600 nm region in solution 2-6 and film 7, respectively. This feature of capturing light in the red portion of the visible spectrum, combined with their high molar extinction coefficients, have made them desirable targets as a photosensitizing agent in photodynamic therapy. 5 Additionally, and more relevant to our research group, Cl-BsubNc (Figure 6.1, X = Cl) has seen increasing interest as a photo- and electro-active material in organic photovoltaics (OPVs) since its debut in 2009. 7-9 No other BsubNc derivative has been explored

127

in this application. At the present time, there are ten total reports on the use of Cl-BsubNc in OPVs. 7-16 In one of these studies, its inclusion resulted in an 8.4% efficient planar heterojunction (PHJ) OPV cell, a power conversion efficiency (PCE) value that is one of the highest within the space of PHJ OPVs. 14

Figure 6.1 . Chemical structure of boron subphthalocyanine (BsubPc) and boron subnaphthalocyanine (BsubNc).

In the last chapter, we had demonstrated that Cl-BsubNc, whether synthesized via literature methods or our in-house developed processes or purchased from a commercial vendor, is actually a mixture of Cl-BsubNc products with random amounts of chlorination at the bay positions of the BsubNc structure. As a result of this mixture and chlorination format, we appropriately refer to the compound as Cl-Cl nBsubNcs. We were unable to develop a chemical process to produce a pure, non-chlorinated sample of Cl-BsubNc. However, we were able to develop a chemical process to produce Cl-Cl nBsubNcs with either lower or higher amounts of peripheral chlorination relative to the literature or commercial Cl-Cl nBsubNc sample. The bay position chlorination and its extent were determined by single crystal X-ray crystallography, X- ray photoelectron spectroscopy, and mass spectrometry. Following these structural analyses, we showed that the photo- and electro-physical properties as well as the PHJ OPV device characteristics were found to be different among the Cl-Cl nBsubNc samples with varying amounts of chlorination. Our findings strongly suggest that the Cl-BsubNc samples used in all previous reported OPV studies are likely not pure, consisting of a mixture of Cl-Cl nBsubNcs with bay position chlorination. 16

128

Scheme 6.1 . Synthesis of BsubNc from 2,3-dicyanonaphthalene.

Within the literature, there are only six derivatives beyond Cl-BsubNc. Three of these compounds are derived from using substituted 2,3-dicyanonaphthalene, 1,2,6 one compound is 3 derived from using BBr 3 as the Lewis acid and templating agent, and two compounds are derived from an axial displacement reaction 6. Despite the limited examples, the BsubNc scaffold is derivatizable at the axial position and the periphery much like their BsubPc counterparts. This chemical flexibility has proven to be a powerful tool for tailoring the physicochemical properties of BsubPcs 17,18 such as their solubilities, 19 optical absorption properties, 20 solid state arrangements, 21,22 and electrochemical properties 23,24 . From the collection of known BsubNcs, albeit small in quantity, it appears that the BsubNcs also share this tunability feature. For example, peripheral fluorination blue-shifts the absorption spectra and lowers the HOMO energy level of the resulting BsubNc compound relative to Cl-BsubNc. 6

As mentioned above, a pure, non-peripherally chlorinated Cl-BsubNc cannot be obtained, but its mixture of peripherally chlorinated species collectively shows great promise as a light harvesting and charge transporting material in OPVs. 16 Given this performance appeal and the fact that no other BsubNcs outside of Cl-BsubNc have been studied in OPVs, we were motivated to examine other derivatives of this class. In particular, we were interested in adopting axial phenoxylation chemistry due to our familiarity and our established protocols for introducing phenoxy substituents at the axial position(s) of BsubPcs and phthalocyanines (Pcs). We have shown that

129

displacing the axial halide with a phenoxy group(s) leads to improvements or benefits on several different levels: (1) hydrolytic stability (B-O > B-Cl) for BsubPcs, 25 (2) open-circuit voltage 26,27 (V OC ) output stemming from a better solid-state arrangement for Pcs, (3) charge transport properties for BsubPcs, 28 (4) solubility in organic solvents for BsubPcs 19,29 and Pcs 30 , and (5) ambipolar ( i.e. n- and p-type) characteristics for BsubPcs 31 .

Herein, we first describe the synthesis of two phenoxylated derivatives of BsubNc, phenoxy-

BsubNc (PhO-Cl nBsubNc) and pentafluorophenoxy-BsubNc (F 5-Cl nBsubNc), from an axial displacement reaction of Cl-Cl nBsubNc obtained from a literature procedure. Reacting this mixture intermediate with a phenol derivative also resulted in a product mixture, whereby bay position chlorination was further confirmed via X-ray crystallography. This work is followed by an examination of the phenoxy Cl nBsubNcs’ basic photo- and electro-physical properties and their PHJ OPV device characteristics. Comparison of these properties is made to those of Cl-

Cl nBsubNc. Again, we emphasize that this work not only augments the very small collection of known BsubNc derivatives, but it marks the first set of examples of BsubNc other than the prototypical Cl-BsubNc as the organic photoactive material in OPVs.

6.3 Results and Discussion From our previous work, a number of processes were developed for the synthesis of Cl-

Cl nBsubNcs. Each distinct route resulted in a mixture of Cl-Cl nBsubNc products that differed among each other in the level of peripheral chlorinates. 16 The best route with respect to mass yield was the one adapted using the literature process published by Kennedy et al .4,16 The resulting product, which we had termed literature-Cl-Cl nBsubNc, comprised Cl-BsubNc, Cl-

Cl 1BsubNc, Cl-Cl 2BsubNc, and Cl-Cl 3BsubNc. Through the same axial phenoxylation chemistry 21,25 that has been developed for BsubPcs, literature-Cl-Cl nBsubNc was treated with phenol and pentafluorophenol in refluxing toluene solution to produce PhO-Cl nBsubNc and F 5-Cl nBsubNc, respectively (Scheme 6.2). HPLC-UV/Vis analysis revealed a mixture of five BsubNc compounds, based on the absorption profile characteristic of a BsubNc chromophore, with R t ~

3.4, 3.9, 4.6, 5.4, and 5.6 min and R t ~ 4.1, 4.8, 5.6, 6.6, and 6.9 min for PhO-Cl nBsubNc and F 5-

130

Cl nBsubNc, respectively (Figure S6.1a & S6.2a). The five products in both samples matches 16 those of literature-Cl-Cl nBsubNc (R t ~ 3.6, 4.4, 5.4, 6.9, and 7.1 min) .

Scheme 6.2 . Synthesis of PhO-Cl nBsubNc and F 5-Cl nBsubNc.

Following their synthesis, the phenoxy BsubNcs were purified via train sublimation. Like our 16 previous findings on the processes leading to any of the Cl-Cl nBsubNcs, the technique was not successful in separating any of the BsubNc compounds. This inability to purify the mixture to a single compound was not a concern given that mixtures of Cl-Cl nBsubNcs perform well in PHJ 16 OPVs. The sublimed sample of each phenoxy Cl nBsubNc was suitable for analysis by LRMS

(Figure S6.1b & S6.2b); LRMS analysis of PhO-Cl nBsubNc and F 5-Cl nBsubNc is consistent with a mixture of four compounds – PhO-BsubNc ( m/z 638.2), PhO-ClBsubNc ( m/z 672.2), PhO-

Cl 2BsubNc ( m/z 706.1) and PhO-Cl 3BsubNc ( m/z 740.1) – and a mixture of three compounds, –

F5-BsubNc ( m/z 728.2), F 5-ClBsubNc ( m/z 762.1) and F 5-Cl 2BsubNc ( m/z 796.1) – respectively.

It is worth noting two observations. The first is that the F 5-Cl nBsubNc sample has one less peripherally chlorinated species (3 vs. 4) compared to the PhO-Cl nBsubNc sample. Second, the number of compounds as detected by LRMS ( i.e. 3/4) is lower than that by HPLC ( i.e. 5). This is 16 consistent with the previous findings for literature-Cl-Cl nBsubNc and for a commercial sample. We attribute this difference to a structural isomeric effect, whereby a pair of structural isomers may have different R ts on HPLC. It is also possible that this difference could be attributed to the sensitivity of the LRMS instrument, whereby the intensity of a fourth and fifth BsubNc species is

131

below the detection limit. To the best of our knowledge, these two BsubNc compounds are the first set of phenoxy BsubNc derivatives and this is the second known example of the use of axial functionalization of BsubNc. In contrast to this, there are several known phenoxy derivatives of BsubPcs. 20,21,25,32,33

For PhO-Cl nBsubNc, crystals suitable for X-ray diffraction were grown by slow evaporation from a DCM solution (Figure 6.1, Table 6.1, Figure S6.3, and Table S6.1-S6.6). As expected, the crystal analysis confirmed a mixture of chlorinated species and that phenoxylation was exclusive 16 to the axial position. Like the literature-Cl nBsubNc’s X-ray determined structure, the chlorines were found situated only on the bay position. The average peripheral chlorine occupancy was calculated to be 0.99 per molecule, resulting in an average molecular formula of

C42 H22.01 BCl 0.99 N6O. This value is slightly lower than that found for literature-Cl-Cl nBsubNc (1.13). 16

Figure 6.2 . Ellipsoid plot (50% probability) showing the structure and atom numbering scheme of PhO-Cl nBsubNc. Hydrogen atoms have been omitted for clarity. Colors: boron - orange; nitrogen - blue; carbon - white; oxygen - red.

132

Table 6.1 . X-ray diffraction results. BsubNc Compound LRMS Analysis Average Peripheral Cl Average Molecular Occupancy (/molecule) Formula PhO-Cl nBsubNc PhO-BsubNc, 0.99 C42 H22.01 BCl 0.99 N6O PhO-Cl 1BsubNc, PhO-Cl 2BsubNc, PhO-Cl 3BsubNc a Literature-Cl-Cl nBsubNc Cl-BsubNc, 1.13 C36 H16.87 BCl 2.13 N6 Cl-Cl 1BsubNc, Cl-Cl 2BsubNc, Cl-Cl 3BsubNc a Data taken from Dang et al .16

UV-vis absorption spectra of the two phenoxy Cl nBsubNcs were acquired in toluene solutions at room temperature (Figure 6.3a and Table 6.2). Both compounds had nearly identical profiles to each other and are characteristic of a BsubNc chromophore. The main Q band for PhO-

Cl nBsubNc and F5-Cl nBsubNc were found at a λ max of 650 and 654 nm with a distinct shoulder 16 peak at 587 and 591 nm, respectively. In comparison to literature-Cl-Cl nBsubNc, the phenoxy derivatives are slightly blue-shifted ( i.e. to higher energy) in absorption with slightly higher extinction coefficients.

The photoluminescence spectra (λ ex = 630 nm) of the two phenoxy Cl nBsubNc compounds were also acquired in toluene solutions at room temperature (Figure 6.3b and Table 6.2). Like the results from the absorption experiments, the PL spectra of PhO-Cl nBsubNc and F 5-Cl nBsubNc were very similar with λ max of emission at 662 and 666 nm, respectively. While the latter 16 spectrum is essentially identical to that of the literature-Cl-Cl nBsubNc, the former spectrum is slightly blue-shifted. With similarities in absorption and photoluminescence to literature-Cl-

Cl nBsubNc, it is not surprising that their Stokes shifts are also nearly identical. In contrast to their phenoxy BsubPc counterparts, 25 these Stokes shifts are noticeably smaller (12 nm vs. 20-29 nm). This suggests that their ground state and excited state are closer in structure than those of the phenoxy BsubPcs, a feature that is likely traced back to the more rigid nature of the benzoisoindoline subunits. Fluorescence quantum yields ( ΦPL ) of 19% were calculated relative to a standard of oxazine-170 (Table 6.2 and Eq S6.1) and they were somewhat lower than the value for literature-Cl-Cl nBsubNc.

133

Figure 6.3 . Overlay of the PhO-Cl nBsubNc (blue) and F 5-Cl nBsubNc (red) (a) normalized absorption spectra and (b) normalized photoluminescence (PL, λ ex = 630 nm) spectra. Cl-

Cl nBsubNc’s data is also shown for reference.

Table 6.2. Photophysical and electronic properties of phenoxy Cl nBsubNcs.

a a a,b 1 2 a 1 a Compound λmax,abs Extinction λmax,PL Stokes ΦPL E ox | E ox (V) E red (V) (nm) Coefficient (nm) Shift a (%) (ɛ, M -1cm -1) (nm) e d e PhO-Cl nBsubNc 650 90,000 662 12 19 +0.80 | +1.32 -0.99 e d e F5PhO-Cl nBsubNc 654 86,000 666 12 19 +0.85 | +1.34 -0.94 Literature-Cl- 656 78,400 666 10 27 +0.83 e | +1.28 d -0.87 d f Cl nBsubNc a In toluene solution. b Relative to a oxazine-170 standard using an excitation wavelength of 630 nm. c In degassed DCM solution relative to Ag/AgCl. d Peak potential. e Half-wave potential. f Data taken from Dang et al .16

The electrochemical properties of PhO-Cl nBsubNc and F 5-Cl nBsubNc were then analyzed via cyclic voltammetry (CV) in degassed DCM solution containing 0.1 M tetrabutylammonium perchlorate (Figure 6.4 and Table 6.2). All potentials (E) were corrected to the half-wave 34 reduction potential (E 1/2,red ) of decamethylferrocene (-0.012 V vs . Ag/AgCl ). Each compound 1 2 underwent a reversible first oxidation (E ox ), an irreversible second oxidation (E ox ), and a 1 1 reversible reduction (E red ). Amongst the two, PhO-Cl nBsubNc is easier to oxidize (E ox = 0.80 1 vs. 0.85 V) while F 5-Cl nBsubNc is easier to reduce (E red = -0.94 vs. -0.99 V). The same trend is 35 also observed for the related BsubPcs, whereby PhO-BsubPc is easier to oxidize while F 5-

134

25,36 1 1 BsubPc is easier to reduce. It is desirable to note that both the E ox and E red were reversible 16 20 processes, a feature that is absent for Cl-Cl nBsubNc and one that is not common for BsubPcs . Based on this, adding a phenol substituent in the axial position of BsubNc grants the molecule electrochemical reversibility in the reduction regime. This effect is also true for some phenoxylated BsubPcs. 25,36 The reversibility in both the oxidation and reduction region for the phenoxy Cl nBsubNcs is not overly surprising as the aromatic nature of the axial substituent combined with the more π-extended benzoisoindoline subunits are anticipated to stabilize the BsubNc in its radical anion or radical cation form. Differential pulse voltammetry (DPV, degassed DMF, 0.1 M TBAPC) was also carried out and it verified the CV results (Figure S6.4 and Table S6.7).

Figure 6.4 . Cyclic voltammograms of PhO-Cl nBsubNc and F 5-Cl nBsubNc in DCM with 0.1M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference (E 1/2,red = -0.012 V vs. Ag/AgCl) at room temperature. Cl-ClnBsubNc’s data is also shown for reference.

Given the similar photophysical properties to literature-Cl-Cl nBsubNc and the reversible electrochemical nature of phenoxy Cl nBsubNcs, the performance of PhO-Cl nBsubNc and F 5-

Cl nBsubNc were examined as acceptors within PHJ OPV devices. Each sample was paired with

135

sexithiophene (α-6T) using the following layer structure: indium tin oxide (ITO)/poly(3,4- ethylenedioxythiophene) poly(styrenesulfonate) (PEDOT:PSS)/α-6T (55 nm)/Cl-Cl nBsubNc (25 nm)/bathocuprione (BCP, 10 nm)/silver (Ag, 80 nm). This configuration replicates the one 16 adopted for the Cl-Cl nBsubNc-based PHJ OPV cells in our previous study to allow for direct comparison to be made herein. Current density-voltage (J-V) characteristics (Figure 6.5a) were measured under 100 mW·cm -2 of simulated solar illumination. The measured external quantum efficiency (EQE) spectra are shown in Figure 6.5b. The performance characteristics (Table 6.3) are derived from the measurement of at least 16 PHJ OPV devices.

Figure 6.5 . (a) J/V characteristics and (b) external quantum efficiency (EQE) spectra of PhO-

Cl nBsubNc, F 5-Cl nBsubNc, and Cl-Cl nBsubNc. PHJ OPVs of the following configuration: ITO/PEDOT:PSS/α-6T(55 nm)/BsubNc(25 nm)/BCP(10 nm)/Ag(80 nm) whereby the BsubNc layer is either PhO-Cl nBsubNc, F 5-Cl nBsubNc, or Cl-Cl nBsubNc.

OPV cells made with PhO-Cl nBsubNc outperform the cells made with F 5-Cl nBsubNc across the board ( i.e. VOC , J SC , FF), producing an average PCE of 1.87% for the former and 1.57% for the latter. The V OC s are higher than those based on literature-Cl-Cl nBsubNc and are consistent with a steric shielding effect, 37 observed with sterically bulky groups ( i.e. phenoxy groups) in small molecule-38-40 and polymer 41 -based OPVs. This effect is explained by an increase in the distance between the donor and acceptor materials induced by steric hindrance. This causes the potential

136

energy of electron-hole pairs at the donor-acceptor interface to increase, leading to an 37 enhancement in the V OC.

Table 6.3 . Average device parameters of PHJ OPVs of the following configuration: ITO/PEDOT:PSS/α-6T(55 nm)/BsubNc(25 nm)/BCP(10 nm)/Ag(80 nm) whereby the BsubNc layer is either PhO-Cl nBsubNc or F 5-Cl nBsubNc. The standard deviation is indicated in brackets next to each parameter. Each parameter is a result of the measurement of at least 16 devices.

a b -2 c d Acceptor VOC (V) JSC (mA·cm ) FF PCE (%) PhO-Cl nBsubNc 1.07 (0.024) 3.76 (0.39) 0.46 (0.01) 1.87 (0.24) F5PhO-Cl nBsubNc 1.02 (0.006) 3.67 (0.28) 0.42 (0.02) 1.57 (0.11) e Literature-Cl-Cl nBsubNc 0.96 (0.026) 8.96 (0.72) 0.45 (0.02) 3.88 (0.21) a Open-circuit voltage. b Short-circuit current density. c fill factor. d Power conversion efficiency. e Data taken from Dang et al .16

While these cells had higher V OC s than those based on literature-Cl-Cl nBsubNc, their PCEs were more than two times lower. The major loss for both set of devices was the result of a much lower extracted J SC s. Given that the photophysical properties of the phenoxy Cl nBsubNcs are similar to literature-Cl-Cl nBsubNc, it is unlikely that the origin of this loss is derived from the efficiency behind the photon absorption and exciton generation process. It is more likely that the lower J SC s are attributed to a decrease in the efficiency behind the exciton dissociation step, whereby the barrier for electron transfer from the α-6T to the BsubNc layer is higher as evident in the higher reduction potentials (i.e. weaker electron accepting ability). Moreover, the drop in the J SC s could be the result of a steric shielding effect (as described above) due to the higher potential energy of electron-hole pairs at the donor-acceptor interface, making it more difficult to separate and collect the individual charge carriers. 37 Overall, these results imply that contrary to BsubPcs, 31 adding phenoxy groups at the axial position of Cl nBsubNc mixtures is detrimental to OPV device performance. These results also imply that there could be a better electron donating material than a -6T for pairing with the phenoxy Cl nBsubNcs.

137

6.4 Conclusions

The first set of phenoxy BsubNc compounds – PhO-Cl nBsubNc and F 5-Cl nBsubNc – are synthesized through an axial displacement reaction of literature-Cl-Cl nBsubNc with phenol or pentafluorophenol. Like their precursor, the products are composed of a mixture of phenoxylated

Cl nBsubNcs with differing level of peripheral chlorinates. Single crystals of PhO-Cl nBsubNc were obtained, analyzed, and determined to have an average peripheral chlorine occupancy of

0.99. While the phenoxy Cl nBsubNcs share similar photophysical behaviours to literature-Cl-

Cl nBsubNc, their electrochemical properties were found to be modified following phenoxylation; they display reversible oxidative and reductive processes with the reduction potential being significantly higher compared with Cl-Cl nBsubNc. This change is also reflected in their greater than two times poorer performance within PHJ OPV devices compared to the precursor. However, their inclusion resulted in devices with higher open-circuit voltages, suggesting that a steric shielding effect between the donor and acceptor is likely in play and that higher efficiencies could be afforded with a better electron donating material than α-6T for pairing with the phenoxy Cl nBsubNcs. In conclusion, the work herein marks the first examples of the use of non-Cl-BsubNcs in OPVs. Contrary to BsubPcs, the axial derivatization of Cl nBsubNcs seems to have a pronounced effect on the electrochemical properties and PHJ OPV performance. More work is ongoing to deconvolute these observations and determine if other electron donating materials would work better when paired with phenoxy Cl nBsubNcs.

138

6.5 References 1. Rauschnabel, J.; Hanack, M. Tetrahedron Letters 1995 , 36 , 1629-1632. 2. Geyer, M.; Plenzig, F.; Rauschnabel, J.; Hanack, M.; Del Rey, B.; Sastre, A.; Torres, T. Synthesis 1996 , 1139-1151. 3. Kobayashi, N.; Ishizaki, T.; Ishii, K.; Konami, H. Journal of the American Chemical Society 1999 , 121 , 9096-9110. 4. Zyskowski, C. D.; Kennedy, V. O. Journal of Porphyrins and Phthalocyanines 2000 , 4, 707-712. 5. Nonell, S.; Rubio, N.; del Rey, B.; Torres, T. Journal of the Chemical Society, Perkin Transactions 2 2000 , 1091-1094. 6. Takao, Y.; Masuoka, T.; Yamamoto, K.; Mizutani, T.; Matsumoto, F.; Moriwaki, K.; Hida, K.; Iwai, T.; Ito, T.; Mizuno, T.; Ohno, T. Tetrahedron Letters 2014 , 55 , 4564- 4567. 7. Verreet, B.; Schols, S.; Cheyns, D.; Rand, B. P.; Gommans, H.; Aernouts, T.; Heremans, P.; Genoe, J. Journal of Materials Chemistry 2009 , 19 , 5295-5297. 8. Ma, B.; Woo, C. H.; Miyamoto, Y.; Frechet, J. M. J. Chemistry of Materials 2009 , 21 , 1413-1417. 9. Gommans, H.; Aernouts, T.; Verreet, B.; Heremans, P.; Medina, A.; Claessens, C. G.; Torres, T. Advanced Functional Materials 2009 , 19 , 3435-3439. 10. Cheyns, D.; Rand, B. P.; Heremans, P. Applied Physics Letters 2010 , 97 , 033301/033301-033301/033303. 11. Kulshreshtha, C.; Kim, G. W.; Lampande, R.; Huh, D. H.; Chae, M.; Kwon, J. H. Journal of Materials Chemistry A 2013 , 1, 4077-4082. 12. Chen, G.; Sasabe, H.; Sano, T.; Wang, X.-F.; Hong, Z.; Kido, J.; Yang, Y. Nanotechnology 2013 , 24 , 484007/484001-484007/484009. 13. Verreet, B.; Cnops, K.; Cheyns, D.; Heremans, P.; Stesmans, A.; Zango, G.; Claessens, C. G.; Torres, T.; Rand, B. P. Advanced Energy Materials 2014 , 4, 1301413/1301411- 1301413/1301418. 14. Cnops, K.; Empl, M. A.; Heremans, P.; Rand, B. P.; Cheyns, D.; Verreet, B. Nature Communications 2014 , 5, 3406.

139

15. Menke, S. M.; Holmes, R. J. ACS Applied Materials & Interfaces 2015 , 7, 2912-2918. 16. Dang, J. D.; Josey, D.; Lough, A.; Li, Y.; Sifate, A.; Lu, Z.; Bender, T. P. Journal of Materials Chemistry A 2016 . 17. Claessens, C. G.; Gonzalez-Rodriguez, D.; Torres, T. Chemical Reviews 2002 , 102 , 835- 853. 18. Morse, G. E.; Bender, T. P. ACS Applied Materials & Interfaces 2012 , 4, 5055-5068. 19. Brisson, E. R. L.; Paton, A. S.; Morse, G. E.; Bender, T. P. Industrial & Engineering Chemistry Research 2011 , 50 , 10910-10917. 20. Gonzalez-Rodriguez, D.; Torres, T.; Guldi, D. M.; Rivera, J.; Herranz, M. A.; Echegoyen, L. Journal of the American Chemical Society 2004 , 126 , 6301-6313. 21. Paton, A. S.; Lough, A. J.; Bender, T. P. CrystEngComm 2011 , 13 , 3653-3656. 22. Virdo, J. D.; Kawar, Y. H.; Lough, A. J.; Bender, T. P. CrystEngComm 2013 , 15 , 3187- 3199. 23. Morse, G. E.; Castrucci, J. S.; Helander, M. G.; Lu, Z. H.; Bender, T. P. ACS Applied Materials & Interfaces 2011 , 3, 3538-3544. 24. Kamino, B. A.; Bender, T. P. Dalton Transactions 2013 , 42 , 13145-13150. 25. Morse, G. E.; Helander, M. G.; Maka, J. F.; Lu, Z.-H.; Bender, T. P. ACS Applied Materials & Interfaces 2010 , 2, 1934-1944. 26. Lessard, B. H.; Grant, T. M.; White, R.; Thibau, E.; Lu, Z. H.; Bender, T. P. Journal of Materials Chemistry A 2015 , 3, 24512-24524. 27. Lessard, B. H.; White, R. T.; Al-Amar, M.; Plint, T.; Castrucci, J. S.; Josey, D. S.; Lu, Z. H.; Bender, T. P. ACS Applied Materials & Interfaces 2015 , 7, 5076-5088. 28. Castrucci, J. S.; Helander, M. G.; Morse, G. E.; Lu, Z. H.; Yip, C. M.; Bender, T. P. Crystal Growth & Design 2012 , 12 , 1095-1100. 29. Paton, A. S.; Lough, A. J.; Bender, T. P. Industrial & Engineering Chemistry Research 2012 , 51 , 6290-6296. 30. Lessard, B. H.; Dang, J. D.; Grant, T. M.; Gao, D.; Seferos, D. S.; Bender, T. P. ACS Applied Materials & Interfaces 2014 , 6, 15040-15051. 31. Morse, G. E.; Gantz, J. L.; Steirer, K. X.; Armstrong, N. R.; Bender, T. P. ACS Applied Materials & Interfaces 2014 , 6, 1515-1524.

140

32. Paton, A. S.; Morse, G. E.; Lough, A. J.; Bender, T. P. CrystEngComm 2011 , 13 , 914- 919. 33. Morse, G. E.; Paton, A. S.; Lough, A.; Bender, T. P. Dalton Transactions 2010 , 39 , 3915- 3922. 34. Noviandri, I.; Brown, K. N.; Fleming, D. S.; Gulyas, P. T.; Lay, P. A.; Masters, A. F.; Phillips, L. Journal of Physical Chemistry B 1999 , 103 , 6713-6722. 35. Morse, G. E.; Helander, M. G.; Stanwick, J.; Sauks, J. M.; Paton, A. S.; Lu, Z. H.; Bender, T. P. Journal of Physical Chemistry C 2011 , 115 , 11709-11718. 36. Sampson, K. L.; Josey, D. S.; Li, Y.; Lu, Z. H.; Bender, T. P. Under Submission. 37. Graham, K. R.; Erwin, P.; Nordlund, D.; Vandewal, K.; Li, R.; Ngongang Ndjawa, G. O.; Hoke, E. T.; Salleo, A.; Thompson, M. E.; McGehee, M. D.; Amassian, A. Advanced Materials 2013 , 25 , 6076-6082. 38. Erwin, P.; Thompson, M. E. Applied Physics Letters 2011 , 98 , 223305. 39. Lam, S. L.; Liu, X.; Zhao, F.; Lee, C.-L. K.; Kwan, W. L. Chemical Communications 2013 , 49 , 4543-4545. 40. Raboui, H.; Al-Amar, M.; Abdelrahman, A. I.; Bender, T. P. RSC Advances 2015 , 5, 45731-45739. 41. Yang, L.; Zhou, H.; You, W. Journal of Physical Chemistry C 2010 , 114 , 16793-16800.

141

Chapter 7 Boron, aluminum, gallium, and indium complexes of 1,3-bis(2- pyridylimino)isoindoline (BPI)

Adapted with permission from: Jeremy D. Dang and Timothy P. Bender. “Boron, aluminum, gallium, and indium complexes of 1,3-bis(2-pyridylimino)isoindoline (BPI).” Inorganic Chemistry Communications 2013 , 30 , 147- 151. DOI: 10.1016/j.inoche.2012.11.020.

All of the experiments in this chapter were performed by me. This work was supervised by TPB. All authors approved the manuscript.

Supplementary information for this chapter is found in Appendix F.

142

7.1 Abstract Numerous metal complexes of 1,3-bis(2-pyridylimino)isoindoline (BPI) are known, but coordination to this ligand has been limited to transition metals. Herein, we report on the synthesis of group 13 (boron, aluminum, gallium, and indium) complexes of BPI and their basic spectroscopic properties. In its deprotonated monoanionic form, the BPI ligand chelates to boron via a bidentate N^N interaction, forming a non-symmetric molecule. Coordination to aluminum, gallium, and indium occurs via a tridendate N^N^N interaction, forming a symmetric molecule. Absorption spectroscopy shows multiple absorption bands in the UV-visible region for the boron, gallium, and indium complex. A single broad absorption in the UV region was noted for the aluminum complex. Fluorescence spectroscopy showed small differences in emission behavior among the different complexes other than the aluminum complex, which was also found to be much more emissive.

7.2 Introduction 1,3-Bis(2-pyridylimino)isoindoline (BPI, Figure 7.1) is a pincer ligand consisting of two pyridyl groups attached to an isoindoline core via an imine bridging system. 1 In its deprotonated monoanionic form, BPI can serve as a tridendate N^N^N ligand 2 and has been shown to complex with many first and second row transition metals such as manganese, 3 iron, 4 cobalt, 5 nickel, 6 copper, 7 zinc, 8 and palladium 9 to form a 1:1 or 2:1 ligand:metal complex. Metal complexes of BPI have previously been employed as catalysts for several organic transformation processes, 10- 12 as model compounds for probing metal coordination environments of biological systems, 5,13 and as a mediator for controlled radical polymerization of acrylates. 14 BPI is generally prepared from the amination of o-phthalonitrile with two molecules of 2-aminopyridine or from the conversion of o-phthalonitrile to the 1,3-diiminoisoindoline intermediate using ammonia gas, followed by their reaction with two molecules of 2-aminopyridine (Scheme 7.1). 1,2

143

Figure 7.1 . From left to right: structure of 1,3-bis(2-pyridylimino)isoindoline (BPI), phthalocyanine (Pc), and boron subphthalocyanine (BsubPc).

Our interest in BPI lies in its structural similarity to the Pcs (Figure 7.1), which have demonstrated applications in OPVs 15,16 as well as other OEs 17,18 . Metal-free BPI has absorption properties that are not well suited as an organic photoactive material for OPVs (Figure 7.2). Its UV-vis absorption spectrum shows multiple bands, most of which are below 400 nm and are clearly far from the region of maximum solar irradiance (450-700 nm). However, complexation of BPI with transition metals like platinum 19,20 or iron 21 has been reported to create low-energy bands that appear within the target region as a result of metal-to-ligand charge transfer transitions. For example in the platination of BPI, a broad absorption band between 425 and 550 nm is present. 20

N NH 2 N N N N N N CN 2 ML 3 X NH N ML 2 N MX 2 CN N N N N N N

NH 3

Scheme 7.1 . Synthesis of BPI and the proposed synthesis of its group XIII complexes and its post-metallated derivatives. M is a group XIII element, L is a monodentate ligand, and X is a new functional group.

144

Figure 7.2 . UV-vis absorption spectrum of metal-free BPI.

Despite BPI’s rich coordination chemistry, to the best of our knowledge there is no precedent for their complexation with group XIII metals/metalloids. Besides this novelty aspect, group XIII metals are anticipated to form complexes of BPI with two ligands in the axial positions owing to their +3 oxidation state. This offers the opportunity to functionalize the metal center via ligand- exchange chemistry to gain access to more derivatives with the aim of altering the physical properties and understanding the structure-property relationships (Scheme 7.1). Moreover, these group XIII complexes are intriguing to us due to their resemblance to the BsubPcs (Figure 7.1), a class of materials that is of major interest in our group with signs of promise for OPV applications. 22,23 In this chapter, we report on the synthesis and basic spectroscopic properties of the boron (BPI·BF 2), aluminum (BPI·AlCl 2), gallium (BPI·GaCl 2), and indium (BPI·InCl 2) complexes of BPI.

7.3 Results and Discussion

The BPI·BF 2, BPI·AlCl2, BPI·GaCl 2, and BPI·InCl 2 complexes were synthesized by heating the BPI ligand, formed from the condensation of o-phthalonitrile and 2-aminopyridine, with the appropriate group XIII metal halide in the presence of triethylamine base (Scheme 7.2). Again, these metal halides were purposefully chosen with the intention to displace the halide ligands with other functional groups such as phenol derivatives to tune and improve their properties. This is a strategy that has been established for Pcs 24,25 and BsubPcs 26,27 . The group XIII

145

complexes were isolated in low yields and this also varied from batch-to-batch. The highest isolated yield obtained for BPI·BF 2, BPI·AlCl 2, BPI·GaCl 2, and BPI·InCl 2 was 19, 25, 17, and 27%, respectively. The poor yields were largely caused by the need to remove the triethylamine salt and unreacted group XIII metal halide from the crude product using water. The aqueous wash not only dissolved and removed the unwanted side products, but it also partially dissolved and removed the desired compound. Efforts to recover the product via liquid-liquid extraction with a variety of organic solvents were ineffective; the distribution coefficient of BPI·BF 2,

BPI·AlCl 2, BPI·GaCl 2, and BPI·InCl 2 clearly favors the aqueous phase over the organic phase. Purification via silica gel column chromatography was also ineffective as the salt product and the desired product stayed fixed on the column and thus, could not be eluted. Additional efforts were made to optimize the work-up procedure, where triethylamine was replaced with tri-n- butylamine. The aim behind the substitution for this fatty organic base was to avoid an aqueous wash work-up and to facilitate the removal of the side products via washing of the crude product with an organic solvent that would not solubilize the target compound ( e.g. hexane). The approach was unsuccessful as rinsing with water was still required to produce a pure product.

Scheme 7.2 . Synthesis of BPI·BF 2, BPI·AlCl 2, BPI·GaCl 2, and BPI·InCl 2. Reagents and conditions: (i) calcium chloride, n-hexanol, reflux, 12 hours; ( ii ) boron trifluoride diethyl etherate, toluene, 100 °C, 24 hours; ( iii ) MCl 3 (M = aluminum, gallium(III), indium(III)), toluene, 100 °C, 24 hours; ( iv ) gallium(III) chloride, n-hexanol, reflux, 2 days.

146

Since the yield of BPI·GaCl 2 was the lowest among the complexes, an attempt was made to determine whether BPI·GaCl 2, and perhaps the other group XIII complexes, could be prepared more readily through a direct approach. This was carried out by adding GaCl 3 into a reaction mixture of o-phthalonitrile and 2-aminopyridine, starting materials used for the synthesis of BPI

(Scheme 7.2). This direct approach produced BPI·GaCl 2 in 9% yield after two days of refluxing in n-hexanol. Although the yield is lower than the aforementioned synthetic method (9% vs. 17%), this approach bypassed the formation of the BPI intermediate.

(a)

(b)

Figure 7.3 . Ellipsoid plot (50% probability) showing the structure and atom numbering scheme for the (a) BPI·GaCl 2 complex (CCDC deposition number: 904845) and the (b) BPI·BF 2 complex (CCDC deposition number: 904844). Hydrogen atoms have been omitted for clarity.

1 The BPI·BF 2, BPI·AlCl 2, BPI·GaCl 2, and BPI·InCl 2 compounds were characterized by H NMR, 13 1 13 C NMR, and HRMS. In each case other than BPI·BF 2, H and C NMR spectra (Figure S7.1- S7.8) confirmed the symmetry of each complex by exhibiting six 1H and nine 13 C resonances.

147

This was supported by a single crystal X-ray diffraction analysis of BPI·GaCl 2 (Figure 7.3a), which showed the gallium metal binding to the nitrogen atom of each pyridyl ring and to the 1 13 nitrogen atom of the isoindoline ring ( i.e. N^N^N interaction). For BPI·BF 2, the H and C NMR spectra showed a unique chemical shift for each proton (12 resonances) and carbon (18 resonances) atom, suggesting that the molecule is not symmetric. This was confirmed by the single crystal structure of BPI·BF 2, where the boron atom was found to bind to a nitrogen atom of one pyridyl ring and to the nitrogen atom of the isoindoline ring (N^N interaction, Figure 7.3b). The B1-N4 bond distance (2.961 Å) was found to be nearly twice as long as the B1-N1

(1.517 Å) and B1-N3 (1.583 Å) bond lengths. Suitable crystals of BPI·AlCl 2 and BPI·InCl 2 for X-ray diffraction analysis could not be grown by slow evaporation from dichloromethane, methanol, toluene, or benzene, by slow cooling from hot toluene or n-hexanol, by vapour diffusion of pentane into dichloromethane or heptane into benzene, or by layering hexane onto dichloromethane.

It is worth noting that the coordination mode of BPI adopted in the BPI·BF 2 compound has only been observed in one other compound – a dimolybdenum complex of BPI. In this compound, one of the two molybdenum centers is bound to an isoindoline nitrogen and to a pyridyl nitrogen while the second molybdenum center is bound to a bridging imino nitrogen. 28 Moreover, the coordination motif in BPI·BF 2 strongly resembles that found in the highly fluorescent 4,4- difluoro-4-bora-3a,4a-diaza-s-indacene (BODIPY) dye, where the boron atom is bound to two fluorine atoms and to two pyrrole nitrogen atoms. 29,30

148

Figure 7.4 . UV-vis absorption spectra of BPI, BPI·BF 2, BPI·AlCl 2, BPI·GaCl 2, and BPI·InCl 2 in degassed dichloromethane solutions at room temperature.

The UV-vis absorption spectra for BPI, BPI·BF 2, BPI·AlCl 2, BPI·GaCl 2, and BPI·InCl 2 were acquired in degassed dichloromethane solutions at room temperature (Figure 7.4). With the exception of BPI·AlCl 2, which showed a single broad absorbing band at a λmax of 301 nm, the spectra for the complexes were broad and consisted of multiple absorption bands in the UV and short wavelength visible region. Similar findings were reported for the nickel(II) and 6,19,20 platinum(II) complexes of BPI. In comparison with the absorption spectrum of BPI (λ max =

386 nm), the spectra of BPI·GaCl 2 (λ max = 408 nm) and BPI·InCl 2 (λ max = 411 nm) were red shifted while the spectra of BPI·BF 2 (λ max = 370 nm) and BPI·AlCl 2 (λ max = 301 nm) were blue shifted. A trend in the absorption properties for the group XIII complexes was not clear.

149

Fluorescence emission spectra were also acquired for the group XIII complexes in degassed dichloromethane solutions at room temperature at three excitation wavelengths (λ ex ): (1) a common wavelength of 352 nm, (2) the global λ max of absorption, and (3) the longest λ max of absorption (Figure 7.5 and Table S7.1). When excited at 352 nm, a small red shift was observed in the spectra of BPI·BF 2 (λ max = 399 nm) and BPI·AlCl 2 (λ max = 399 nm) while a small blue shift was observed in the spectra of BPI·GaCl 2 (λ max = 396 nm) and BPI·InCl 2 (λ max = 398 nm) relative to the emission spectrum of BPI (λ max = 398 nm). Since the global and the longest λ max of absorption were different for each complex, comparisons of the emission spectra at these two λ ex could not be made. However, a trend is noted whereby the maximum emission intensity decreases with higher λ ex . For example, the maximum emission intensity is 13.9, 10.8, and 5.4

(arbitrary units) at a λ ex of 352, 370, and 388 nm, respectively, for BPI·BF 2. Although the concentration of each complex was not the same, the maximum absorbance value for each sample was kept in the range of 0.04 and 0.05. This allowed for qualitative comparisons to be made between the different compounds. At the common λ ex of 352 nm, the maximum emission intensity for all complexes was on the same order of magnitude. At the global and longest λ max of excitation, BPI·AlCl 2 was found to be orders of magnitude more emissive than BPI and the other group XIII complexes, which were all found to have similar emission intensities at both λ ex . Moreover, the shape of the emission curves was all similar to one another with the exception of

BPI·AlCl 2. From the absorption and fluorescence spectra, it is apparent that BPI·AlCl 2 differs significantly from the rest of the other complexes. Its dissimilar spectroscopic properties warrant further investigation.

150

Figure 7.5 . Fluorescence emission spectra (λ ex = 352 nm, highest and longest λ max of absorption as indicated) for BPI, BPI·BF 2, BPI·AlCl 2, BPI·GaCl 2, and BPI·InCl 2 in degassed dichloromethane solutions at room temperature.

The fluorescence quantum yields (φ) for all compounds other than BPI·AlCl 2 were measured to be in the range of 0.7-2.7% relative to a standard of 9,10-diphenylanthracene (Table S7.2 and Eq

151

S7.1). The very low values are in accordance with reported quantum yields for a platinum(II) complex of BPI and its post-metallated derivatives, which are all in the range of 0.03-3.79% in degassed dichloromethane solutions at ambient temperature. 19 It is interesting to note that

BPI·BF2 has a low quantum yield although it shares a similar coordination motif to that found in 29,30 the highly fluorescence BODIPYs. A quantum yield could not be measured for BPI·AlCl2 using this relative approach due to our inability to find a well characterized standard with absorption and emission bands that closely match BPI·AlCl 2. We can however surmise that given its higher emission intensity (Figure 7.5), its quantum yield is likely to be significantly higher.

None of the BPI·BF 2, BPI·AlCl 2, BPI·GaCl 2, and BPI·InCl 2 complexes were incorporated into OPVs as the organic photoactive material primarily due to their ill-suited absorption properties. These compounds have absorption bands outside of the target region of 450-700 nm, where the solar irradiance is the highest. Though it is a possibility to further red-shift the absorption of

BPI·GaCl 2 and BPI·InCl 2 via displacement of the chloride ligands with aromatic functional groups, their syntheses had proven to be very challenging as evident by the low yields and batch- to-batch variations. For these reasons, it is unworthy to undergo further investigation into group XIII complexes of BPI as photoactive materials for OPV applications.

7.4 Conclusions

We have reported on the synthesis of boron (BPI·BF 2), aluminum (BPI·AlCl 2), gallium

(BPI·GaCl 2), and indium (BPI·GaCl 2) complexes of 1,3-bis(2-pyridylimino)isoindoline (BPI). Their two-step synthesis began with the condensation reaction of o-phthalonitrile and 2- aminopyridine to form the BPI ligand, followed by complexation with the appropriate group XIII halide in the presence of an organic base. For BPI·AlCl 2, BPI·GaCl 2, and BPI·InCl 2, the group XIII centers were found to coordinate with BPI in a tridentate N^N^N manner, preserving the symmetric nature of the BPI ligand. In the case of BPI·BF 2, the boron atom was bound to BPI in a N^N manner, breaking the symmetry of the BPI ligand. The UV-Visible absorption spectra for

BPI·BF 2, BPI·GaCl 2, and BPI·InCl 2 were broad and consisted of multiple absorption bands in the UV-visible (short wavelength) region, while a single broad peak in the UV region was

152

observed for BPI·AlCl 2. Fluorescence emission spectra were acquired at three wavelengths of excitation. Small differences in the emission spectra of BPI and its group XIII complexes were observed when excited at 352 nm. At the λ max and longest λ of absorption, BPI·AlCl 2 was found to be much more emissive than the other complexes, which all had fluorescence emission intensities on the same order of magnitude. Very low fluorescence quantum yields in the range of

0.7-2.7% were measured for BPI, BPI·BF 2, BPI·GaCl 2, and BPI·InCl 2. These values are in line with reported quantum yields of platinated BPIs. As a result of a poor coverage of the solar irradiance spectrum, none of the group XIII complexes of BPI were studied as photoactive materials in OPV devices. An attempt to shift their absorption profiles to the desired spectral region via tuning of their photophysical properties by axial ligand displacement reactions were not performed due to the difficulties in their syntheses.

153

7.5 References 1. Elvidge, J. A.; Linstead, R. P. Journal of the Chemical Society 1952 , 5000-5007. 2. Siegl, W. O. The Journal of Organic Chemistry 1977 , 42 , 1872-1878. 3. Kaizer, J.; Baráth, G.; Speier, G.; Réglier, M.; Giorgi, M. Inorganic Chemistry Communications 2007 , 10 , 292-294. 4. Balogh-Hergovich, E.; Speier, G.; Réglier, M.; Giorgi, M.; Kuzmann, E.; Vértes, A. Inorganic Chemistry Communications 2005 , 8, 457-459. 5. Selvi, P. T.; Stoeckli-Evans, H.; Palaniandavar, M. Journal of Inorganic Biochemistry 2005 , 99 , 2110-2118. 6. Letcher, R. J.; Zhang, W.; Bensimon, C.; Crutchley, R. J. Inorganica Chimica Acta 1993 , 210 , 183-191. 7. Anderson, O. P.; la Cour, A.; Dodd, A.; Garrett, A. D.; Wicholas, M. Inorganic Chemistry 2003 , 42 , 122-127. 8. Siegl, W. O. Inorganic and Nuclear Chemistry Letters 1974 , 10 , 825-829. 9. Bröring, M.; Kleeberg, C.; Cónsul Tejero, E. European Journal of Inorganic Chemistry 2007 , 2007 , 3208-3216. 10. Saussine, L.; Brazi, E.; Robine, A.; Mimoun, H.; Fischer, J.; Weiss, R. Journal of the American Chemical Society 1985 , 107 , 3534-3540. 11. Meder, M. B.; Gade, L. H. European Journal of Inorganic Chemistry 2004 , 2716-2722. 12. Siggelkow, B.; Meder, M. B.; Galka, C. H.; Gade, L. H. European Journal of Inorganic Chemistry 2004 , 3424-3435. 13. Balogh-Hergovich, É.; Kaizer, J.; Speier, G.; Huttner, G.; Jacobi, A. Inorganic Chemistry 2000 , 39 , 4224-4229. 14. Langlotz, B. K.; Fillol, J. L.; Gross, J. H.; Wadepohl, H.; Gade, L. H. Chemistry - A European Journal 2008 , 14 , 10267-10279. 15. Fleetham, T. B.; Mudrick, J. P.; Cao, W.; Klimes, K.; Xue, J.; Li, J. ACS Applied Materials & Interfaces 2014 , 6, 7254-7259. 16. Raïssi, M.; Vignau, L.; Cloutet, E.; Ratier, B. Organic Electronics 2015 , 21 , 86-91. 17. Kvitschal, A.; Cruz-Cruz, I.; Hummelgen, I. A. Organic Electronics 2015 , 27 , 155-159.

154

18. Lu, Z. Y.; Li, X.; Wang, Y.; Xiao, J.; Xu, P. L. Current Applied Physics 2014 , 14 , 1465- 1469. 19. Wen, H. M.; Wu, Y. H.; Fan, Y.; Zhang, L. Y.; Chen, C. N.; Chen, Z. N. Inorganic Chemistry 2010 , 49 , 2210-2221. 20. Hanson, K.; Roskop, L.; Djurovich, P. I.; Zahariev, F.; Gordon, M. S.; Thompson, M. E. Journal of the American Chemical Society 2010 , 132 , 16247-16255. 21. Kripli, B.; Barath, G.; Balogh-Hergovich, E.; Giorgi, M.; Simaan, A. J.; Parkanyi, L.; Pap, J. S.; Kaizer, J.; Speier, G. Inorganic Chemistry Communications 2011 , 14 , 205-209. 22. Mutolo, K. L.; Mayo, E. I.; Rand, B. P.; Forrest, S. R.; Thompson, M. E. Journal of the American Chemical Society. 2006 , 128 , 8108-8109. 23. Gommans, H.; Aernouts, T.; Verreet, B.; Heremans, P.; Medina, A.; Claessens, C. G.; Torres, T. Advanced Functional Materials 2009 , 19 , 3435-3439. 24. Schumann, S.; Hatton, R. A.; Jones, T. S. Journal of Physical Chemistry C 2011 , 115 , 4916-4921. 25. Ren, B. Y.; Zhu, L.; Cui, G. R.; Sun, Y. G.; Zhang, X. T.; Liang, F. S.; Liu, Y. C. Tetrahedron Letters 2013 , 54 , 5953-5955. 26. Brisson, E. R. L.; Paton, A. S.; Morse, G. E.; Bender, T. P. Industrial & Engineering Chemistry Research 2011 , 50 , 10910-10917. 27. Morse, G. E.; Castrucci, J. S.; Helander, M. G.; Lu, Z. H.; Bender, T. P. ACS Applied Materials & Interfaces 2011 , 3, 3538-3544. 28. Baird, D. M.; Shih, K. Y.; Welch, J. H.; Bereman, R. D. Polyhedron 1989 , 8, 2359-2365. 29. Treibs, A.; Kreuzer, F.-H. Justus Liebigs Annalen der Chemie 1968 , 718 , 208-223. 30. Karolin, J.; Johansson, L. B. A.; Strandberg, L.; Ny, T. Journal of the American Chemical Society 1994 , 116 , 7801-7806.

155

Chapter 8 A synthetic and engineering process to boron tribenzosubporphyrins

The work in this chapter is currently being prepared for submission as a manuscript to Industrial & Engineering Chemistry Research or Organic Process Research & Development. The author list will be: Jeremy D. Dang, Alexander J. Peltekoff, Anjuli M. Szawiola, David S. Josey, and Timothy P. Bender.

AJP and I carried out all experiments and wrote the chapter together. AMS measured the thicknesses of the films acquired via OVPD. DSJ performed all OPV device fabrication and testing. The work was supervised by TPB. All authors approved the manuscript.

Supplementary information for this chapter is found in Appendix G.

156

8.1 Abstract An alternative process for synthesizing boron tribenzosubporphyrins (BsubPys) was developed using a solvent-assisted approach. Sulfolane was found to be the most promising solvent given its high boiling point, water-miscibility, and its ability to solubilize both organic and inorganic matter. A work-up procedure consisting of an aqueous wash, Soxhlet extraction, and a hybrid Soxhlet-column extraction produces BsubPy with purity up to 39 wt %. Attempts to train sublime the product to electronic grade purity were not feasible, leading us to develop an OVPD process. This process grew non-uniform films that performed very poorly in when merged into PHJ OPV devices.

8.2 Introduction Boron tribenzosubporphyrins (BsubPys) are a class of aromatic macrocycles that strongly resembles the BsubPcs (Figure 8.1). Both classes are composed of three isoindoline subunits that chelate a single boron atom within their internal cavity and are bowl-shaped in structure with a highly conjugated, 14 π-electron configuration. They differ in their bridging system, where BsubPys contain bridging methines (=CH-) while BsubPcs contain bridging imines (=N-). 1,2 As a result of the tetravalent nature of carbon, the methines could be substituted with aryl groups and this subclass is referred to as meso -aryl boron tribenzosubporphyrins ( mAr-BsubPys, Figure 8.1). 3 However, only the meso -phenyl derivatives are known in the literature. The non-meso - substituted class will be referred to as normal or n-BsubPy.

Figure 8.1. From left to right: chemical structure of a BsubPc, n-BsubPy, and mAr-BsubPy.

157

Despite the structural similarities, BsubPys have not received the same level of attention compared to the BsubPcs. A literature search produces only a few reports on BsubPys. The dearth of research on BsubPys is primarily attributed to their undeveloped chemistry. For example, BsubPcs can be made on the gram scale (>20 g) in yields up to 63%, 4,5 while BsubPys can be made on the milligram scale (<100 mg) in yields ranging from 1.4% up to 7.8% 1-3 (Scheme 8.1). The established synthetic protocols for BsubPcs have ultimately resulted in their employment as functional materials in OPVs 6-8 and other OE applications 9-11 . In contrast to this, BsubPys have never been explored in any OE application.

Although the BsubPys are structurally very similar to the BsubPcs, their optical absorption properties differ quite considerably (Figure 8.2). Their Q band is blue-shifted while their B band is red-shifted compared to typical BsubPcs. 1-3 Moreover, the B band is more intensive than the Q band, whereas the opposite is observed for BsubPcs9,12 and a near-equivalent intensity is 13 observed for μ-oxo-(BsubPc) 2 . Even though the Q band is much less intensive than the B band, its molar extinction coefficient is still reported to be very high at 9.0x10 4 and 6.6x10 4 L·mol - 1·cm -1 for n-BsubPy 1 and mPh-BsubPy 3, respectively. These numbers are comparable to the Pcs 14 and BsubPcs 13,15 . It is also worth noting that there are subtle differences in the solution-state absorption behaviour of n-BsubPys and mPh-BsubPy. There is a difference of 3 nm in the Q 1,3 peak and 17 nm in the B peak. Like the μ-oxo-(BsubPc) 2, the absorption spectra of either group of BsubPys align very closely to the maximum solar irradiance (Figure 8.2). Thus from an optical perspective, BsubPys are highly desirable synthetic targets to us for applications in OPVs.

Another attractive feature of the BsubPys is that they possess a high degree of chemical flexibility. Axial functionalization has been demonstrated for both n-BsubPy 1,2 and mPh- BsubPy 3, but fundamentally derivatization could be feasible at the periphery and at the meso position. Thus from a chemical design point of view, there are more possible avenues available compared to the Pcs or BsubPcs for constructing new derivatives with modified physicochemical properties.

158

Scheme 8.1 . Synthesis of HO-n-BsubPy 1,2 (left route) and HO-mAr-BsubPy 3 (right routes) and the ellipsoid plot (50% probability) of HO-n-BsubPy (CCDC deposition number: 281754) and

Ph 3SiO-mPh-BsubPy (CCDC deposition number: 676629).

In this chapter, a survey of synthetic processes is described in an effort to produce n- and mPh- BsubPys in a more facile manner than the literature procedures. The overall objectives are threefold: (1) to develop their synthetic methodologies and augment the very small collection of known BsubPy compounds, (2) to study their photo- and electro-physical properties, and (3) to assess their performance worth in OPV devices.

159

Figure 8.2 . Absorption spectrum of n-BsubPy (green) and mPh-BsubPy (orange) overlaid with the solar irradiance (black, sea level) spectrum. The red shaded area represents the area of highest solar irradiance (450-700 nm). The B and Q bands are indicated.

8.3 Results and Discussion n-BsubPys Our study began with the desire to synthesize n-BsubPys given that they differ chemically from the BsubPcs by a single degree ( i.e. bridging system) versus two degrees as seen in their m- phenyl counterparts ( i.e. bridging system and meso substituent). This will enable us to draw a more direct comparison between the two classes. In the reported synthesis of HO-n-BsubPy, three molecules of isoindolinone-3-acetic acid is reacted with boric acid at 350 ˚C under an atmosphere of nitrogen (Scheme 8.1). 1,2 We surmised that this reaction proceeds through an aldol-like condensation mechanism and that the low reported yield of 1.4% could possibly be improved by promoting this process via the use of a base such as sodium ethoxide or tert - butoxide in an anhydrous solvent. Moreover, an accompanying benefit with the use of a solvent

160

is the better mixing of the reacting species. Thus, the employment of both base and solvent are anticipated to make this reaction more feasible.

The first attempt at synthesizing HO-n-BsubPy involved treating isoindolinone-3-acetic acid with triethyl borate and sodium ethoxide in anhydrous THF at room temperature (Table S8.1, Method 1.1). Triethyl borate was used instead of boric acid to potentially reduce any side reaction that could occur between the boric acid and the sodium ethoxide base. No reaction occurred as determined by HPLC ( i.e. absence of BsubPy chromophore), suggesting that perhaps a higher reaction temperature or a stronger base was required for the reaction to proceed. The reaction was repeated, but heated to 66 ˚C (Table S8.1, Method 1.2) and it also showed no progress. A final attempt in this set of reactions was carried out using a stronger base, sodium tert -butoxide (Table S8.1, Entry 1.3). Again, a chromophore for the target n-BsubPy was not formed under this condition.

Taking into consideration the very high temperature of 350 ˚C that was used in the reported synthesis of HO-n-BsubPy, 1,2 we turned to the use of higher boiling point solvents (150 ˚C and higher). However, since the boiling point of triethyl borate is 117 ˚C, we opted to use solid boric acid instead. The reaction condition of Method 1.1 was adapted (Table S8.1, Method 1.4) and HPLC analysis revealed no reaction progress. This was repeated with the use of DMF (b.p. = 152 ˚C) (Table S8.1, Method 1.5) and it also showed no difference in the results. The stronger base, sodium tert -butoxide, was next employed in anhydrous THF and diglyme (b.p. = 162 ºC) (Table S8.1, Method 1.6 & 1.7), showing the same results. A final attempt in this second set of reactions was carried out using a very strong base, n-butyllithium (Table S8.1, Method 1.8). Unlike all previous experiments, several products were formed in this reaction. However, none of them had the characteristic optical absorption spectrum of a BsubPy compound. Seeing that the use of a base in combination with a high reaction temperature did not facilitate the desired reaction, we abandoned the use of the base and subsequently proposed the use of a more suitable solvent to assist in the reaction. In considering candidate solvents, three selection criteria were used: (1) ability to solubilize both inorganic and organic materials, (2) high boiling

161

point, and (3) good miscibility with water. The first criterion ensures that the boric acid and isoindolinone-3-acetic acid and their reacting intermediates are fully dissolved and well-mixed in the reaction mixture. The second criterion is to access even higher reaction temperatures than the ones previously employed to potentially drive the reaction to proceed. The third criterion is to permit the practical removal of the high boiling solvent via organic-aqueous extraction in the work-up step following a successful synthesis. Though solvents were employed in the prior experiments, none of them satisfied all three criteria. Sulfolane (b.p. = 285 ˚C) and dimethyl sulfone (b.p. = 238 ˚C) were two proposed solvents that fit these criteria.

The reaction of boric acid and isoindolinone-3-acetic acid was first carried out in dimethyl sulfone (Table S8.2, Method 2.1). Within a day of heating at 230 °C, HPLC revealed the formation of multiple n-BsubPy peaks based on their characteristic absorption profile. The relative integration of the main peak with R t ~ 1.9 min was found to increase over time, indicating that its formation was evolving. However, the use of dimethyl sulfone was challenging as it has a tendency to condense at the top of the reactor due to its high melting point (m.p. = 107-109 °C). As a result, external heating from a heat gun was regularly applied to melt the solidified dimethyl sulfone and return it back to the reaction medium. This issue was slightly improved with the use of a heating tape, heated up to 200 °C, to insulate the top portion of the reactor.

We then turned to the use of sulfolane, 16 which has a much lower melting point range of 20-26 °C and did not crystallize at the top of the reactor (Table S8.2, Method 2.2). Within a day of heating, HPLC analysis of the reaction mixture revealed the presence of multiple peaks with the characteristic n-BsubPy chromophore. The main peak with R t ~ 1.9 min increased in intensity over time up to 7 days. It should be noted that HPLC is not a reliable technique for quantitatively tracking the reaction progress for the reason that the relative integration value of the main peak did not persistently increase with time; this value was observed to decrease at certain time points, but there was a general increasing trend. As a result of this, it was difficult to determine an end

162

time for the reaction and 7 days was chosen to ensure good conversion to the HO-n-BsubPy product.

At the end of the reaction following a cool down to room temperature, a black solid mass was resulted. To remove the sulfolane and any water-soluble material, we found that the most practical method was to add water to the reaction flask and to stir and heat the mixture at 50 °C overnight. A gravity filtration with several rounds of aqueous washing was then performed to collect the aqueous-insoluble solid. A Soxhlet extraction using toluene was then carried out to extract out a black solid. A thin layer chromatography (TLC) analysis of this black-coloured product using a 1:2:2 ( v/v/v ) mixture of diethyl ether/dichloromethane/hexane as the developing solvent showed a bright green fluorescent spot with a retention factor (R f) of ~0.3 under UV 1,2 light. This R f and fluorescence feature matches those reported in the literature. A LRMS analysis was also consistent with a n-BsubPy compound minus the axial hydroxyl fragment.

Purity was evaluated optically using Beer-Lambert (B-L) law to calculate the weight percentage (wt %) of HO-n-BsubPy within the sample (Eq S8.1-S8.3). Unlike the HPLC method, this method is more accurate since it accounts for materials that do not absorb UV-visible light ( i.e. non-chromophores). Via the B-L analysis, the crude sample had a wt % of 8.5.

Before moving onto purification, some attempts were made to optimize the wt % yield of the reaction and thereby, determine the best process. A reaction that was half as concentrated and one that was twice as concentrated (Table S8.2, Method 2.3 & 2.4) was carried out. In the former case, there was not any appreciate mass after an aqueous work-up while in the latter case, the reaction mixture solidified within two days. A neat reaction (Method 2.6) was also performed and it also solidified, but did so within three hours. Other high boiling point solvents, phenyl ether (b.p. = 259 °C), 1,2-dichlorobenzene (b.p. = 180 °C), and 1,2,4-trichlorobenzene (b.p. = 214 °C), that did not fulfill the three criteria of a suitable solvent described above were also employed for comparison purposes. In all of these cases (Table S8.2, Method 2.7-2.9), a n- BsubPy chromophore was not formed, suggesting that the solvent's ability to solubilize both

163

inorganic and organic materials does play a key role. Additional efforts to optimize this reaction were not carried out because the precursor is not widely commercially available and had to be prepared in-house. Isoindolinone-3-acetic acid was synthesized in a two-step process with an overall yield of 64% beginning with the conversion of 1-nitroso-2-naphthol to the o- cyanoallocinnamic acid intermediate, followed by its transformation to the desired precursor. Based on these experimental results, the best process is the initial sulfolane process (Table S8.2, Method 2.2). This process was scaled up by a factor of three (Method 2.5), producing a crude product with a similar purity level (8.8 wt %).

In the reported synthesis of HO-n-BsubPy, the compound was purified through at least three rounds of silica gel column chromatography. 1,2 We do not find this to be a practical route and thus, attempted to develop our own purification process. Initially, an attempt was made to purify the HO-n-BsubPy via silica gel column chromatography eluting with dichloromethane. It was found that the compound adsorbed very strongly to the stationary phase and so, a transition to pure methanol was necessary for its elution. Even with the use of this very polar solvent, a very large volume was required. As a result of the solvent-wasteful nature of this process, we turned to the use of a hybrid Soxhlet-column extraction. In this technique, dry-loaded HO-n-BsubPy on silica gel was added on top of fresh silica gel in a glass thimble with a fritted disk (medium pore size) equipped at the bottom. The thimble was placed inside of a Soxhlet extraction apparatus and was initially extracted with dichloromethane to remove a side product with a purple fluorescence when irradiated with a UV lamp. Once this fluorescence was no longer observed and the filtrate coming out of the thimble was clear, a switch to methanol was made to elute out the desired product. This process was again monitored by a UV lamp for its characteristic green fluorescence. After the n-BsubPy was eluted, the filtrate was concentrated to produce green- black metallic-like solids. 1H NMR analysis of this product was consistent with a methoxy- substituted n-BsubPy (MeO-n-BsubPy) (Figure S8.3). This axial substitution was not surprising as Osuka et al. have reported that HO-n-BsubPy can be converted quantitatively to the methoxy derivative by heating it to reflux in methanol for 30 minutes. 1,2

164

Overall, the process up to this point is a long one consisting of ~7 days for the synthesis, ~5 days combined for the aqueous work-up and the Soxhlet extraction with toluene, and ~8 days for the hybrid Soxhlet-column extraction. The purity of the product following this process via B-L law was measured to be 28-31 wt %, a value that is more than three times greater than that of the pre- purified sample (8.5-8.8 wt %). This purity is obviously too low for testing in OPV devices and so we turned to the use of sublimation. In consideration of the long synthetic route to semi-pure MeO-n-BsubPy, train sublimation was initially performed on the analogous MeO-BsubPc (Table S8.4, Method 4.1-4.4) to approximate an appropriate thermal profile for subliming this material. With the optimized thermal profile for MeO-BsubPc, train sublimation of MeO-n-BsubPy was next performed (Table S8.5, Method 5.1). A maroon red, thin band was observed at a distance of 1.6-2.5” downstream from the tail end of the boat (Figure 8.3). There was not enough material in the red band to collect, but there was just enough to analyze it via LRMS, optical absorption and photoluminescence spectroscopy (Figure S8.6), and 1H NMR spectroscopy. The results from these analyses were consistent with a n-BsubPy molecule. Given the considerable residual mass in the boat, the sublimation process was repeated with a higher end temperature of 500 °C and 530 °C (Table S8.5, Method 5.2 & 5.3). Both experiments resulted in nearly identical amounts of residual mass in the boat ( i.e. same overall mass yield).

Figure 8.3 . A picture of the train sublimation experiment (Method 5.1) with a close-up image of the maroon red band ( n-BsubPy).

Given the very low sublimation yield and the lengthy process for producing semi-pure MeO-n- BsubPy, combined with the observation that its film can be formed under sublimation condition, we focused our efforts into developing an organic vapor phase deposition (OVPD) process for merging with OPV devices. OVPD is the coating of vapor-state organic compounds onto a

165

substrate ( i.e. growth of organic thin films). 17 Unlike the CVD process seen in Chapter 4, this process does not involve any chemical reaction. The intention here is to gain some insights into their value as a light-harvesting and charge transporting material. If n-BsubPy performs well in these OVPD-derived OPV devices, additional efforts will be put into optimizing their synthetic process to obtain appreciable quantities (>100 mg) of suitable purity for their subsequent integration in OPV devices via the conventional vacuum thermal evaporation technique. If n- BsubPy performs very poorly, we will then know that despite their favourable photophysical properties, they are not promising candidates in OPVs. Our pursuit into BsubPys will justifiably come to a halt.

Figure 8.4 . Pictures of the OVPD experiment (Method 5.5) showing the seven ITO on glass substrates, each with a pre-deposited layer of PEDOT:PSS.

An OVPD experiment was carried out using the OCVD reactor described in Chapter 4. An ITO on glass substrate was placed in a vertical orientation at a distance of ~4" downstream from the boat. Via the same thermal profile used in Method 5.3 (Table S8.5, Method 5.4), a thin film with was formed on the substrate and it was consistent with a n-BsubPy compound via UV-vis absorption spectroscopy (Figure S8.7). After determining the feasibility of n-BsubPy for OVPD, the next experiment involved using seven ITO substrates, each with a pre-deposited layer of PEDOT:PSS (~35 nm thickness), spaced out within the reactor (Figure 8.4). The intention

166

behind using multiple substrates was to improve the probability of a functional OPV device once the other layers ( i.e. electron acceptor and cathode) have been added.

5 114 99 200 Top 116 51 100 181 212 4 24 25 107 51 31 3

Cell Cell # 18 86 80 50

2 50 50 87 40 20

24 21 10 1 Bottom

2 3 4 5 6

Substrate #

Figure 8.5 . Cartoon image showing the film thickness at various locations across Substrate #2-6. The numbers within the boxes are in units of nanometers.

The films from the OVPD experiment were visually non-uniform. This was confirmed via profilometry (Figure 8.5) on five of the seven substrates (Substrate #2-6). The substrate number corresponds to its location within the OVPD reactor, whereby Substrate #1 is closest to the boat while Substrate #7 is furthest away from the boat. Profilometry also showed that the films were the thickest at the top end (end nearest to the center of the reactor) and the thinnest at the bottom end (end sitting at the bottom of the reactor). Moreover, the thickness range differed between the substrates, suggesting that their position within the reactor is a factor in the amount of material deposition. At first sight, this non-uniformity was an issue from a device characterization perspective. However, the variable film thickness across each substrate presents the advantage of enabling us to study the device characteristics as a function of film thickness.

Of the seven substrates, Substrate #2-5 were selected and carried forward to device integration based on their film absorption spectra (Figure S8.10). To determine a suitable acceptor material for pairing, the electrochemical properties of the BsubPy films were examined via CV and DPV

167

in DCM solution containing 0.1 M tetrabutylammonium perchlorate (Figure 8.6 and Table S8.7). 1 CV analysis revealed an irreversible first oxidation (E ox ) at +0.66 V, a reversible second 2 3 oxidation (E ox ) at +0.81 V, and an irreversible third oxidation (E ox ) at +1.44 V. A reduction process was not observed. Based on this, MeO-n-BsubPy shows more promise as an electron donor material than an acceptor material. Differential pulse voltammetry (DPV, degassed DMF, 0.1 M TBAPC) confirmed the CV results and also showed a minor reduction peak at -0.79 V 1 2 (Figure 8.6 and Table S8.7). Using either the E ox , which is a minor peak, or the E ox equates to a HOMO energy level of 5.4 or 5.6 eV, respectively, via the Thompson-Forrest equation 18 (Eq S8.4). The LUMO value is calculated to be 3.3 and 3.5 eV, respectively, using the energy band gap value obtained from the onset of absorption (Eq S8.5 & S8.6). These HOMO and LUMO values suggest that Cl-Cl 12 BsubPc, C 60 , or C 70 would be suitable acceptors to produce a donor- acceptor heterojunction with a staggered energy gap (Figure 8.7). This would enable electrons to flow “downhill” via the LUMO state towards the cathode and for holes to flow “uphill” via the HOMO state towards the anode, allowing charge carriers to be separated and collected.

Figure 8.6 . Cyclic (left) and differential pulse (right) voltammetry traces of MeO-n-BsubPy in DCM with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference at room temperature.

168

3.3 / 3.5

3.9 4.1 4.4 BsubPy BsubPc 60 12 C 70

5.4 / 5.6 C Cl-Cl Energy Energy Level(eV) 6.3 6.3 6.4

19 20 Figure 8.7 . Energy level (eV) band diagram of MeO-n-BsubPy, Cl-Cl 12 BsubPc, C60 , and 21 C70 .

Fresh films of MeO-n-BsubPy were made and integrated into the following PHJ OPV device structure: indium tin oxide (ITO)/poly(3,4-ethylenedioxythiophene) poly(styrenesulfonate)

(PEDOT:PSS)/MeO-n-BsubPy/Cl-Cl 12 BsubPc (20 nm)/bathocuprione (BCP, 10 nm)/silver (Ag, 80 nm). Current density-voltage (J-V) characteristics (Figure 8.8a) were measured under 100 mW·cm -2 of simulated solar illumination. The measured external quantum efficiency (EQE) spectra are shown in Figure 8.8b. Five cells were measured per substrate for a total of 20 cells and were labeled #1-5 based on their location, where #1 is at the bottom and #5 is at the top of the substrate (Figure 8.5).

Of the 20 cells, 15 of them short-circuited while the remaining 5 showed very poor to null device characteristics (Figure 8.8 and Table 8.1), with PCE values that barely registered. EQE spectra of the five cells show that the slight photocurrent can be attributed to Cl-Cl 12 BsubPc contributions, with no observable contribution from BsubPy. Although these results are disappointing, this work marks the closest that a BsubPy compound has ever come to being studied in OPV devices. These results are inconclusive in determining if the poor device data is directly attributed to the nature of the BsubPy material or the OVPD process. The next step forward is to carry out a control experiment using a well-known functional material. Films of Cl-BsubPc should be grown

169

via the same OVPD process and merged into PHJ OPV devices. If the resulting device characteristics are similar to cells fabricated using a standard deposition method, then it is likely that the poor performance of the BsubPy cells is due to the properties of the BsubPy. If the device characteristics are poor with the Cl-BsubPc control, then the poor performance of the BsubPy cells is likely related to the OVPD process.

(a)

(b) Figure 8.8 . (a) J/V characteristics and (b) external quantum efficiency spectra. PHJ OPVs of the following configuration: ITO/PEDOT:PSS/MeO-n-BsubPy/Cl-Cl 12 BsubPc/BCP/Ag.

170

Table 8.1 . Device parameters of PHJ OPVs of the following configuration:

ITO/PEDOT:PSS/MeO-n-BsubPy/Cl-Cl 12 BsubPc/BCP/Ag.

a b c 2 d e Substrate No. Cell No. VOC (V) JSC (ma/cm ) FF PCE (%) 2 1 0 0 0 0 2 0 0 0 0 3 0 0 0 0 4 0.09 0.08 0.26 0.002 5 0.12 0.07 0.25 0.002 3 1 0 0 0 0 2 0 0 0 0 3 0.10 0.12 0.24 0.003 4 0.14 0.18 0.25 0.006 5 0.21 0.09 0.24 0.004 4 1 0 0 0 0 2 0 0 0 0 3 0 0 0 0 4 0 0 0 0 5 0 0 0 0 5 1 0 0 0 0 2 0 0 0 0 3 0 0 0 0 4 0 0 0 0 5 0 0 0 0 a Cell number is based on location on the substrate (1 = bottom, 5 = top). b Open-circuit voltage. c Short-circuit current. d Fill factor. e Power conversion efficiency. mPh-BsubPys We also scoped the synthesis of the m-phenyl class. For HO-mPh-BsubPy, the reported synthesis follows a similar route as HO-n-BsubPy whereby 3-benzalphthalimidine is used as the starting material along with a slightly higher temperature of 360 ˚C (Scheme 8.1). The HO-mPh-BsubPy can also be formed from phthalimide and phenylacetic acid (Scheme 8.1). 3 Although the HO- mPh-BsubPy is more structurally complex, their formation is higher yielding (7.8% and 3.6%) 3 than the normal derivative (1.4%) 1,2 . Considering the two reported routes to HO-mPh-BsubPy, we were actually more interested in the lower yielding process for the reasons that the precursors are commercially available and inexpensive and that the process is a single-step. On top of this, the former route requires the use of ammonia gas, posing additional safety considerations.

171

Via the solvent-assisted approach outlined earlier for the n-BsubPy, the reaction of phthalimide, phenylacetic acid, and boric acid was initially carried out in sulfolane and dimethyl sulfone (Table S8.3, Method 3.1-3.12). Unlike the n-BsubPy case, HPLC was much more reliable here for quantifying the reaction progress as the main peak with R t ~3.6 min persistently evolved in formation over time. The reaction times for HO-mPh-BsubPy synthesis were long; ~7 days of heating was needed to reach steady state. To semi-quantitatively determine if the formation of the HO-mPh-BsubPy was actually increasing over time, a plot of the detector intensity at 510 nm

(i.e. ~λ max of absorption of the Q band) over time was made (Figure S8.1). The plot showed an increase in the detector intensity over time, confirming that the long reaction time is actually required.

In comparing the two solvents, dimethyl sulfone (Table S8.3, Method 3.1-3.5) and sulfolane (Table S8.3, Method 3.6-3.11), the synthesis of HO-mPh-BsubPy was found to be more reproducible in the latter solvent. This was especially true for scale-up experiments, whereby there was not a successful scale-up process using dimethyl sulfone (Table S8.3, Method 3.5). Other high boiling point solvents, diphenyl sufone (b.p. = 379 ˚C) and tetraethylene glycol (b.p. = 314 °C), were also attempted for comparison purposes and neither showed any promise (Table S8.3, Method 3.12 & 3.13). Following the same aqueous wash and Soxhlet extraction with toluene used in the n-BsubPy work-up, the purity of the HO-mPh-BsubPy product from only the promising experiments via B-L law was determined to be in the range of 8.3-13.5 wt % (Table S8.3). Although the wt % was found to be slightly higher for the dimethyl sulfone process, the sulfolane process was preferred and carried forward given its better reproducibility and also due to its tendency to not condense at the top of the reactor. Of the sulfolane-based experiments, the best process is the one of Method 3.7.

In the reported synthesis of HO-mPh-BsubPy, the compound was purified through two rounds of column chromatography using neutral alumina as the adsorbent followed by a gel permeation column. 3 Again like the n-BsubPy, we find this process to be impractical and attempted to develop our own purification process. Seeing that alumina adsorbent was used in the reported

172

literature process, we initially carried out an alumina column chromatography using dichloromethane as the eluent. From a 3 g crude sample of HO-mPh-BsubPy, ~200 mg of a shiny green-black product with a HPLC purity level >85% was obtained (Figure S8.2a). Four attempts to purify this product via train sublimation up to a final temperature (measured externally) of 450 ˚C, 500 ˚C, and 530 ˚C were all unsuccessful despite the fairly high purity (Table S8.6, Method 6.1-6.4). We attributed the sublimation issue to either the presence of the black-coloured impurities or to the nature of the axial hydroxyl functional group. Our reasoning for the latter is due to our experience in subliming >99% pure HO-BsubPc, which we had found to be very low- yielding and unstable. In an attempt to remove the black impurities, a silica gel column chromatography using dichloromethane as the eluent was carried out. Like the n-BsubPy case, the HO-mPh-BsubPy was found to adsorb very strongly to the silica gel stationary phase and could not be eluted using dichloromethane. A solvent switch to methanol was required for its elution and the resulting product was a maroon red solid with a 90% purity as measured by HPLC (Figure S8.2b). A 1H NMR analysis of the product was consistent with a MeO-mPh- BsubPy compound, indicating than an axial exchange reaction with methanol had occurred in the column. Kobayashi et al . had also observed a similar substitution phenomenon in their purification process but with the use of ethanol. 3 The maroon red product was train sublimed to produce a bright orange compound in a 33% mass yield (Figure S8.2c and Table S8.6, Method 6.5).

Characterizations via NMR (Figure S8.4), optical absorption and photoluminescence spectroscopy (Figure S8.8), and HRMS were all consistent with the compound. The electrochemical properties were also examined via cyclic voltammetry (CV) in degassed DCM solution containing 0.1 M tetrabutylammonium perchlorate (Figure 8.9 and Table S8.7). The 1 analysis revealed a reversible first oxidation (E ox ) at +0.83 V, an irreversible second oxidation 2 1 (E ox ) at +1.38 V, and an irreversible reduction (E red ) at -1.08 V. Based on these redox 1 potentials, the compound shows more promise as an electron donor due to the E ox being lower 1 1 than the E red , implying that it is easier to be oxidized than reduced. Also, the E ox is reversible

173

1 while the E red is not reversible. Differential pulse voltammetry (DPV, degassed DMF, 0.1 M TBAPC) was also carried out and it verified the CV results (Figure 8.9 and Table S8.7).

Figure 8.9 . Cyclic (left) and differential pulse (right) voltammetry traces of MeO-mPh-BsubPy in DCM with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference at room temperature.

The ability to obtain a somewhat appreciable mass of sublimed MeO-mPh-BsubPy (21 mg) was obviously very encouraging as we were one step closer to characterizing the compound within devices. Unfortunately, our supply of alumina (Fisher Scientific) was consumed and was no longer commercially available in Canada. Other alumina adsorbents were purchased and used, but none of them were successful in replicating the chromatographic process. Although the 21 mg of material is insufficient for integration into OPV devices with our current vacuum deposition system, a new design to our deposition system is underway that can accommodate very small quantities. This work is to be undertaken in the future.

We then moved onto adapting the hybrid Soxhlet-column extraction process employed in the n- BsubPy case for the mPh-BsubPy. Like the HO-n-BsubPy compound, the HO-mPh-BsubPy adsorbed very strongly to the silica gel when using dichloromethane as the eluent. Unlike the n- BsubPy case, whereby there was a clear sign to make a solvent switch from dichloromethane to methanol to elute the strongly-adsorbed desired product once the filtrate had become transparent,

174

there was no visible indication to do so here. Continual elution with dichloromethane for almost two weeks eventually led to the complete extraction of the n-BsubPy. To reduce the lengthy chromatographic process, a solvent switch to methanol was made after a couple days of dichloromethane elution. From one chromatographic trial, the resulting product was a red-brown solid and from a second trial, it was a dark brown solid. This colour difference is a clear sign of a reproducibility issue, especially given that both products were derived from the same synthetic batch.

Given the observation that it took nearly two weeks to elute HO-mPh-BsubPy using dichloromethane via the hybrid Soxhlet-column extraction, we became interested in displacing the axial polar hydroxyl group with the less polar phenol group. The motivation behind this is to weaken its interaction with the silica gel adsorbent and thereby, enable the faster elution of the PhO-mPh-BsubPy using dichloromethane. PhO-mPh-BsubPy was prepared by treating crude HO-mPh-BsubPy, obtained after the aqueous work-up step, with excess phenol in toluene solution under refluxing condition overnight. Contrary to our expectation, the PhO-mPh-BsubPy was found to adsorb even more strongly than HO-mPh-BsubPy to the silica gel adsorbent. At first this was discouraging, however this route permitted the extraction of undesired compounds using dichloromethane without extracting the target product and after three days of continual elution, the filtrate coming out of the thimble became clear. This provided us the visual cue to make a solvent switch to methanol to elute the green-fluorescing BsubPy compound. The resulting dark brown product was no longer PhO-mPh-BsubPy, but MeO-mPh-BsubPy. An axial displacement reaction had assumingly occurred in the column and/or while the eluted PhO-mPh- BsubPy was being heated in methanol in the pot still. This substitution reaction was verified by a change in HPLC R t for the reaction of pre-chromatographic PhO-mPh-BsubPy in methanol heated at reflux. Despite the added step of phenoxylation, this approach addresses the chromatographic reproducibility concern observed in the HO-mPh-BsubPy case. The resulting MeO-mPh-BsubPy product from this hybrid Soxhlet-column extraction had a purity of 29.1 wt %, a value that is 2-3 times greater than the unpurified product (8.3-13.5 wt %).

175

We also observed during the methanol elution of PhO-mPh-BsubPy that the initial fraction coming out of the thimble was dark brown-black in colour. This is a colour that is not related to the BsubPy chromophore. We then modified the work-up procedure the second time around, whereby this dark brown-black was collected and discarded. The resulting product was a brown- red solid with a 39 wt % purity, a purity that is nearly 10% higher than the product obtained via the previous work-up procedure. Characterizations via 1H NMR (Figure S8.5) and optical absorption and photoluminescence spectroscopy (Figure S8.9) were all consistent with a MeO- mPh-BsubPy compound.

Figure 8.10 . Pictures of the OVPD experiment (Method 6.12) showing the seven ITO on glass substrates, each with a pre-deposited layer of PEDOT:PSS.

This semi-pure product was next train sublimed (Table S8.6, Method 6.6) via the same thermal profile adopted previously (Method 6.5) and this resulted in a very low ~2% mass yield. Taking into consideration that some of the material in the boat charred, a lower end temperature of 500 °C was next used (Method 6.7) and it resulted in a similar mass yield. Increasing the ramp time to an end temperature of 500 °C (Method 6.8) or 530 °C (Method 6.9) or increasing the mass input (Method 6.10 & 6.11) all gave similar results. The inability to sublime this 39 wt % MeO- mPh-BsubPy using a previously proven thermal profile indicates that the impurities are likely impeding the sublimation process. Given that some of this material can be sublimed, albeit at very low yields, a similar OVPD process as described for the n-BsubPy was adopted. Like the n-

176

BsubPy, films of MeO-mPh-BsubPy were not uniform (Figure 8.10) with the formation of crystallites near the top end of the substrate. Unlike the n-BsubPy case, the crystallites are concerning as their presence will likely act as charge traps. For this reason, the films were not carried forward and merged into OPV devices.

8.4 Conclusions An alternative process for synthesizing n-BsubPy and mPh-BsubPy was proposed and developed using a solvent-assisted approach. Sulfolane was found to be the most promising solvent in this approach for both classes of compounds given its high boiling point, water-miscibility, and its ability to solubilize both organic and inorganic matter. However, with its use along with other attempted solvents that worked, the reaction times were found to be very long on the order of ~7 days. A work-up procedure was subsequently developed, consisting of an aqueous wash, Soxhlet extraction, and a hybrid Soxhlet-column extraction. This is a two-week process that produces MeO-n-BsubPy or MeO-mPh-BsubPy in yields up to 31 or 39 wt %, respectively. Despite the lengthy operation, the process only involves a single chromatography step as opposed to the many rounds of chromatography as seen in the literature processes. Attempts to train sublime either material to electronic grade purity were not feasible, which led us to develop an OVPD process. For MeO-n-BsubPy, non-uniform films were afforded and were merged into PHJ OPV devices using Cl-Cl 12 BsubPc as the acceptor to examine the effect of film thickness on device characteristics. Unfortunately, all working cells showed very poor to null device characteristics with PCEs that barely registered. This marked the first time a BsubPy compound was studied in devices. For MeO-mPh-BsubPy, formation of crystallites was observed following OVPD, deterring us from merging the films into devices for charge-trapping concerns.

177

8.5 References 1. Inokuma, Y.; Kwon, J. H.; Ahn, T. K.; Yoo, M. C.; Kim, D.; Osuka, A. Angewandte Chemie-International Edition 2006 , 45 , 961-964. 2. Inokuma, Y.; Osuka, A. Dalton Transactions 2008 , 2517-2526. 3. Makarova, E. A.; Shimizu, S.; Matsuda, A.; Luk'yanets, E. A.; Kobayashi, N. Chemical Communications 2008 , 2109-2111. 4. Morse, G. E.; Paton, A. S.; Lough, A.; Bender, T. P. Dalton Transactions 2010 , 39 , 3915- 3922. 5. Dang, J. D.; Virdo, J. D.; Lessard, B. H.; Bultz, E.; Paton, A. S.; Bender, T. P. Macromolecules 2012 , 45 , 7791-7798. 6. Mutolo, K. L.; Mayo, E. I.; Rand, B. P.; Forrest, S. R.; Thompson, M. E. Journal of the American Chemical Society 2006 , 128 , 8108-8109. 7. Verreet, B.; Rand, B. P.; Cheyns, D.; Hadipour, A.; Aernouts, T.; Heremans, P.; Medina, A.; Claessens, C. G.; Torres, T. Advanced Energy Materials 2011 , 1, 565-568. 8. Pandey, R.; Zou, Y. L.; Holmes, R. J. Applied Physics Letters 2012 , 101 , 033308. 9. Morse, G. E.; Helander, M. G.; Maka, J. F.; Lu, Z. H.; Bender, T. P. ACS Applied Materials & Interfaces 2010 , 2, 1934-1944. 10. Yasuda, T.; Tsutsui, T. Molecular Crystals and Liquid Crystals 2007 , 462 , 3-9. 11. Claessens, C. G.; Gonzalez-Rodriguez, D.; Torres, T.; Martin, G.; Agullo-Lopez, F.; Ledoux, I.; Zyss, J.; Ferro, V. R.; de la Vega, J. M. G. Journal of Physical Chemistry B 2005 , 109 , 3800-3806. 12. Gonzalez- Rodriguez, D.; Torres, T.; Guldi, D. M.; Rivera, J.; Herranz, M. A.; Echegoyen, L. Journal of the American Chemical Society 2004 , 126 , 6301-6313. 13. Geyer, M.; Plenzig, F.; Rauschnabel, J.; Hanack, M.; Del Rey, B.; Sastre, A.; Torres, T. Synthesis 1996 , 1139-1151. 14. Kobayashi, N.; Ishizaki, T.; Ishii, K.; Konami, H. Journal of the Americal Chemical Society 1999 , 121 , 9096-9110. 15. Fulford, M. V.; Jaidka, D.; Paton, A. S.; Morse, G. E.; Brisson, E. R. L.; Lough, A. J.; Bender, T. P. Journal of Chemical & Engineering Data 2012 , 57 , 2756-2765.

178

16. Tilstam, U. Organic Process Research & Development 2012 , 16 , 1273-1278. 17. Baldo, M.; Deutsch, M.; Burrows, P.; Gossenberger, H.; Gerstenberg, M.; Ban, V.; Forrest, S. Advanced Materials 1998 , 10 , 1505-1514. 18. D'Andrade, B. W.; Datta, S.; Forrest, S. R.; Djurovich, P.; Polikarpov, E.; Thompson, M. E. Organic Electronics 2005 , 6, 11-20. 19. Castrucci, J. S.; Josey, D. S.; Thibau, E.; Lu, Z.-H.; Bender, T. P. Journal of Physical Chemistry Letters 2015 , 6, 3121-3125. 20. Morse, G. E.; Gantz, J. L.; Steirer, K. X.; Armstrong, N. R.; Bender, T. P. ACS Applied Materials & Interfaces 2014 , 6, 1515-1524. 21. Benning, P. J.; Poirier, D. M.; Ohno, T. R.; Chen, Y.; Jost, M. B.; Stepniak, F.; Kroll, G. H.; Weaver, J. H.; Fure, J.; Smalley, R. E. Physical Review B 1992 , 45 , 6899-6913.

179

Chapter 9 Summary and Future Work

9.1 Summary Isoindole derivatives comprise a very small space within the domain of organic photoactive materials for OPV applications. Despite their unexplored nature, well known isoindole-based compounds like the Pcs 1 and the less known BsubPcs 2 show signs of promise. This thesis aims to primarily develop and strengthen the fundamental understanding of the structure-property- performance relationships for these general classes of compounds for OPVs. To enable this advancement, five unprecedented, unexplored, or underexplored classes of isoindole-based compounds were studied and they were the: 1) BsubPc polymers (Chapter 2);

2) Oxygen-bridged dimers of BsubPc (μ-oxo-(BsubPc) 2 (Chapter 3 & 4); 3) Boron subnaphthalocyanines (BsubNcs) (Chapter 5 & 6); 4) Group XIII complexes of 1,3-bis(2-pyridylimino)isoindoline (BPI) (Chapter 7); and 5) Boron tribenzosubporphyrins (BsubPys) (Chapter 8). The BsubPc polymers and group XIII complexes of BPI were unprecedented in the literature while the μ-oxo-(BsubPc) 2 dimers, BsubNcs, and BsubPys were known but were unexplored or underexplored in OPV devices. As a consequence of their undeveloped nature, the syntheses of these compounds were found to be challenging. Thus, leading to the genesis of my thesis statement:

To recall, the general course of action that is taken in studying any new material in our group is to first synthesize and purify the compound(s). This is followed by an assessment of its basic physical properties. If these properties are not appealing for OPVs, then its continual

180

investigation comes to a halt. If they are found suitable and promising, then its chemical process is optimized in an effort to obtain sufficient amounts of the compound(s) (ideally >250 mg). Lastly, the compound(s) is integrated within OPV devices and its performance is assessed.

In Chapter 2, a set of BsubPc-containing polymers were synthesized for the first time. This was only feasible through a post-polymerization functionalization approach, whereby the BsubPc moiety was appended onto the side chain of a styrene-based pre-polymer using axial phenoxylation chemistry. A pre-polymerization functionalization approach was not feasible due to premature auto-polymerization of the monomeric BsubPc coupling partner, purification issue of the resulting BsubPc-containing monomer, and to the potential radical-scavenging effect of BsubPc. The unique and desirable photophysical properties of BsubPcs were found preserved in these polymers. 3 Unfortunately, the BsubPc polymers were not stable as some of the appended BsubPcs were found decoupled from the polymer chain during its storage under ambient conditions. On the positive note, this work led to the formation of second generation BsubPc polymers, which were found stable after ~1.5 year of storage, and to their subsequent application in OLEDs as an orange-emitting material. This marked the first time a BsubPc polymer was studied in an OE device. 4,5

In Chapter 3, a process was developed for the synthesis of μ-oxo-(BsubPc) 2 at the gram quantity scale and with sufficiently high purity. The key to this synthesis was the addition of an acid/HBr scavenger, tripotassium phosphate, in the reaction of Br-BsubPc and HO-BsubPc. This work permitted μ-oxo-(BsubPc) 2’s photo- and electro-physical properties to be examined as these physical characterizations, other than the optical absorption behaviours, were not previously known. The process was also adaptable towards the synthesis of three unsymmetric μ-oxo-

(BsubPc) 2 derivatives, whereby the two BsubPc moieties are chemically distinct. This is a subclass of oxygen-bridged BsubPc dimers that was unprecedented in the literature. 6 μ-Oxo-

(BsubPc) 2 was finally tested and characterized in PHJ OPV devices after being first reported in the literature about twenty years ago. The compound showed more promise as an electron donor than an electron acceptor, performing comparable with the prototypical Cl-BsubPc. This

181

investigation resulted in the addition of another member in the presently small collection of OPV-functional BsubPc compounds. 7

In Chapter 4, an OCVD process was proposed, designed, and developed for the growth of thin films of μ-oxo-(BsubPc) 2 from Br-BsubPc given its performance value. This work aimed to address the labor-intensive nature behind the synthesis of μ-oxo-(BsubPc) 2 and to more directly produce μ-oxo-(BsubPc) 2-based OPV devices. Initially, micron-thick films were produced and they were a mixture of three BsubPc compounds - μ-oxo-(BsubPc) 2, HO-BsubPc, and Br- BsubPc. It was shown that the relative composition can be altered by modifying the thermal profile of the process. Following some optimizations, nanometer-thick films containing μ-oxo-

(BsubPc) 2 as high as 74% were afforded. Unfortunately, neat films of μ-oxo-(BsubPc) 2 have not been realized via this technique. However, given our recent interest in mixtures as functional materials for OPVs as highlighted in Chapter 5 & 6, coupled with the good device performance 7 properties of a mixture of μ-oxo-(BsubPc) 2 and Cl-BsubPc as a single layer in PHJ OPVs, efforts will be made to examine the applicability of these OCVD films within devices. Additional process optimization will also be made in an aim to obtain a neat film of μ-oxo-

(BsubPc) 2.

In Chapter 5, a recently emerging OPV-functional material, Cl-BsubNc, was studied in great lengths largely due to its lack of reported analytical characterizations. At the start of this study, there were only four manuscripts in the literature regarding Cl-BsubNc's utility in OPVs and since then, five more papers have been published including a Nature Communications 8. It is clear that there is a rising interest in this particular molecule. In our investigation, Cl-BsubNc, whether synthesized via literature methods or our in-house developed methods or purchased commercially, was determined to be a mixture of Cl-BsubNcs with differing number of chlorine atoms at the periphery. Attempts to produce the non-peripherally chlorinated Cl-BsubNc were unsuccessful. However, a sample of Cl-BsubNc with lower levels of chlorination was realized along with a sample of higher levels of chlorination. The photo- and electro-physical properties were found to differ among the different samples and this also translated to their PHJ OPV

182

device performance characteristics. It was observed that the sample of Cl-BsubNc with the higher levels of chlorination produced more efficient OPV devices than those based on the literature-prepared or commercially-bought Cl-BsubNc. 9 Peripheral halogenation of Cl-BsubNc has been previously speculated, 10 however, no evidence have been shown to support this. The work in this chapter presented several pieces of evidence to truly support and confirm this side reaction. Moreover, peripheral chlorination was found to occur exclusively at the bay positions and not in a random manner at both the bay and terminal positions of the BsubNc scaffold.

In Chapter 6, the first set of phenoxy BsubNc compounds – PhO-Cl nBsubNc and F 5-Cl nBsubNc

– were synthesized via an axial phenoxylation reaction of Cl-Cl nBsubNc. Like their precursor, the products are composed of a mixture of phenoxylated Cl nBsubNcs with differing level of peripheral chlorinates. Single crystals of PhO-Cl nBsubNc were obtained and it further confirmed that peripheral chlorination is exclusive to the bay position. While the phenoxy Cl nBsubNcs share similar photophysical behaviours to Cl-Cl nBsubNc, their electrochemical properties were found to be modified following phenoxylation; they display reversible oxidative and reductive processes with the reduction potential being significantly higher compared with Cl-Cl nBsubNc. This change is also reflected in their more than two times poorer performance within PHJ OPV devices compared to the precursor. However, their inclusion resulted in devices with higher open-circuit voltages, suggesting that a steric shielding effect between the donor and acceptor is likely in play and that better performance could be attained with a better electron donating material than α-6T for pairing with the phenoxy Cl nBsubNcs. In conclusion, the work herein marks the first examples of the use of non-Cl-BsubNcs in OPVs.

In Chapter 7, the boron, aluminum, gallium, and indium complexes of BPI were synthesized via a complexation reaction between BPI and the respective group XIII metal halide. These novel compounds were challenging to prepare as evident in their low isolated yields and their batch-to- batch variations. 11 Their optical absorption spectra were found to be ill-suited for light- harvesting purposes as their absorption bands do not overlap with the region (450-700 nm) of the

183

highest solar irradiance. For these reasons, the further study of group XIII complexes of BPI as organic photoactive materials for OPV applications was justifiably halted. In Chapter 8, an alternative process for synthesizing BsubPys, both normal and meso -phenyl, was proposed and developed using a solvent-assisted approach. Sulfolane was found to be the most promising solvent in this approach for both classes of compounds given its high boiling point, water-miscibility, and its ability to solubilize both organic and inorganic matter. However, with its use along with other attempted solvents that worked, the reaction times were found to be very long on the order of ~7 days. A work-up procedure was subsequently developed, consisting of an aqueous wash, Soxhlet extraction, and a hybrid Soxhlet-column extraction. This entire process takes nearly two weeks to carry out, resulting in MeO-n-BsubPy or MeO-mPh-BsubPy with purity up to 31 or 39 wt %, respectively. Despite the lengthy operation, the process employs a single chromatography step as opposed to the many rounds of column chromatography reported in the literature processes.12-14 Attempts to train sublime the material to electronic grade purity was not feasible and for this reason, an OVPD process was developed. This process grew non-uniform films for both materials. In the case of MeO-n-BsubPy, we were interested in taking advantage of the variable film thicknesses to study their effect on the device characteristics.

Films of it were merged into PHJ OPV devices using Cl-Cl 12 BsubPc as the acceptor, resulting in very poor performance. For MeO-mPh-BsubPy, formation of crystallites was observed following the OVPD process, deterring us from merging its films into devices due to charge trap concerns.

The progress/advancement of each of the five isoindole classes is summarized in Figure 9.1 with the size of the circle indicating the amount of effort spent for that specific step. For example, a considerate amount of effort was put into the chemical process optimization of BsubNcs. For the BsubPc polymers and group XIII complexes of BPI, advancements were not made beyond the assessment of their basic properties for reasons described above. The other three streams saw progress to the final step of OPV device integration and performance assessment. A process engineering and film characterization step were incorporated for the regular μ-oxo-(BsubPc) 2 and for both normal and meso -phenyl BsubPys to bypass the major issues with their chemical process

184

optimization step. The future direction for the BsubPc polymers, μ-oxo-(BsubPc) 2s, and BsubPys are also displayed (as dashed lines) in Figure 9.1 and will be discussed in the next section.

BsubPc Polymers BsubPc Dimers BsubNcs BPI Complexes BsubPys

N N N N B N

R = H Rn = F 12 , X = OPh, X = Cl Cl 6, F5Ph Cl 12 Initial Synthesis

Assessment of Basic Properties

Chemical Process X Optimization

Process Engineering

Film Characterization

Assessment of OPV Device Performance Accomplished Future Work

Figure 9.1 . The five classes of isoindole-based compounds studied in this dissertation showing their progress and future work. The isoindole subunit is emphasized in blue for each class of compounds.

185

9.2 Future Work 9.2.1 BsubPc Polymers in OPVs The optical properties of the first and second generation BsubPc polymers were shown to be preserved from their small molecule counterparts. 3-5 Also, a second generation BsubPc polymer has recently been successfully applied within OLED devices as an orange-emitting material. 5 Given these results, the next appropriate step is to determine whether they can be employed as an organic photoactive material in OPVs. This is after all, the ultimate objective of this investigation. Due to their polymeric nature, BsubPc polymers can only be integrated into OPV devices of the BHJ architecture. They should be explored as either an electron donor material blended with an established fullerene acceptor or as an electron acceptor material blended with a known donor like α-sexithiophene. Following this work, some insights into their value in either a donor or an acceptor role can be made and compared with small molecule BsubPcs. To emphasize again, this task has not been previously realized due to the inability to synthesize BsubPc polymers until now.

9.2.2 μ-Oxo-(BsubPc) 2 Films via OCVD

As described in Chapter 4, a film consisting of up to 74% μ-oxo-(BsubPc) 2 was grown via an OCVD process using Br-BsubPc as the starting material. Given our recent interest in mixtures as functional materials for OPVs, combined with the good device performance properties of a 7 mixture of μ-oxo-(BsubPc) 2 and Cl-BsubPc as a single layer in PHJ OPVs, this OCVD-derived film should next be tested in PHJ OPV devices with the following configuration:

ITO/PEDOT:PSS/μ-oxo-(BsubPc) 2 + X-BsubPcs/C 60/70 /Ag. If the devices work, then comparison of the performance characteristics can be made to the analogous cells based on pure μ-oxo-

(BsubPc) 2 (Chapter 3). If the cells do not work, then it is clear that this particular mixture format is detrimental to device performance and efforts will be put into further optimizing the OCVD process in an aim to obtain a neat film of μ-oxo-(BsubPc) 2. Variables such as carrier gas pressure, gas temperature, and temperature profile ( i.e. rate of temperature increase) can all be tuned. It is also worthwhile to wrap the reactor tube with flexible ceramic insulator to create a longer temperature gradient such that there is a smaller drop in temperature as a function of

186

distance downstream from the boat/heating band. The intention behind this is to promote further sublimation of the smaller and lighter BsubPcs ( i.e. Br-BsubPc and HO-BsubPc) near, on, and around the substrate, leaving behind a higher composition and possibly a pure film of μ-oxo-

(BsubPc) 2. Also, a highly uniform film is ideally desired for device fabrication and testing. Given that a carrier gas is used to transport the gaseous BsubPc materials within the reactor, there is a concern of turbulent fluid flow near and around the substrate. This would have an undesired effect on the quality of the resulting film ( i.e. uniformity). On the other hand, this would present us the opportunity to scope the effect of film thickness on device characteristics.

9.2.3 Unsymmetric μ-Oxo-(BsubPc) 2s in OPVs

μ-Oxo-(BsubPc) 2 has recently been shown to be a promising electron donor material in PHJ OPV devices with PCEs in line with the prototypically studied Cl-BsubPc-based cells. 7 The unsymmetric variants are a newly discovered subclass that displays desirable photo- and electro- physical behaviours for OPVs. From an optical absorption perspective, they are strong light- absorbers with Q bands within the region of maximum solar irradiance (450-700 nm). Furthermore, their Q bands are broader than those of the monomeric BsubPc types and as such, have a better spectral coverage. From an electrochemical perspective, the unsymmetric μ-oxo- 6 (BsubPc) 2s are unique and that they are easier to reduce ( i.e. low reduction potentials), demonstrating that they would be good candidates as electron acceptors. Integration of the unsymmetric dimers as acceptors within PHJ OPVs should next be undertaken and this study would also fit well with the increasing interest in alternative acceptors outside of the fullerene family. 15-17

187

Table 9.1 . Electronic properties of μ-oxo-(BsubPc) 2 and the unsymmetric variants.

1 a,b 1 a,b c d Compound E ox E red HOMO LUMO (V) (V) (eV) (eV) e e µ-oxo-(BsubPc) 2 +0.95 -1.03 5.9 3.7 F12 BsubPc-O-BsubPc +1.15 -0.60 6.0 4.1 Cl 6BsubPc-O-BsubPc +1.08 -0.55 5.9 4.1 Cl 12 BsubPc-O-BsubPc +1.13 -0.60 6.0 4.1 a Measured via CV in degassed DCM solution relative to Ag/AgCl. b Data taken from Dang et al .6 c Calculated using the equation developed by D’Andrade et al .18 d Calculated using the equation developed by Djurovich et al. 19 e Data taken from Castrucci et al .7

Before testing them as acceptors, it is necessary to accurately determine their FMO energy levels in order to pair it with an appropriate donor material (Figure 9.2). The HOMO energy level can be measured using UPS while the LUMO energy level can be calculated using the HOMO energy level and the energy band gap (from the onset of absorption). At the present moment, the HOMO 18 and LUMO 19 energy levels can be estimated based on CV-determined oxidation and reduction potentials, respectively, using models developed by Thompson and Forrest. Applying these models to the unsymmetric μ-oxo-(BsubPc) 2s produce HOMO values that are nearly the same as μ-oxo-(BsubPc) 2 and LUMO values that are lower lying (Table 9.1). From these estimations, Cl-BsubPc would be a suitable donor material for pairing with any of the three unsymmetric μ-oxo-(BsubPc) 2 dimers. However, the FMO energy levels should accurately be measured first.

2.4

3.5 3.7 3.7

α-6T 3.9

2 4.1 4.4 Cl-

4.9 BsubPc 6 μ-oxo- Cl-BsubPc 60 Cl- 70 BsubPc (BsubPc) Cl C 12 5.6 C 5.9 5.8 Cl 6.3 6.3 6.4

20 21 7 Figure 9.2 . Band diagram of energy levels (eV) for α-6T, Cl-BsubPc, μ-oxo-(BsubPc) 2, Cl- 22 23 21 24 Cl 6BsubPc, Cl-Cl 12 BsubPc, C60 , and C 70 .

188

Torres et al .25 and Jones et al .26 have each fabricated a set of "all-BsubPc" PHJ OPV devices, whereby BsubPc derivatives were integrated as both the donor and acceptor material (Table 9.2).

In the former case, Cl-BsubPc was paired with F-F12 BsubPc in one set of cells and with Cl-

F12 BsubPc in a second set of cells. In the latter case, Cl-BsubPc was paired with Cl-Cl 6BsubPc. In both cases, Cl-BsubPc was used as the donor material while the peripherally halogenated

BsubPcs were used as the acceptor material. For each of the three cells described, the J SC s were found lower than those based on the analogous C 60 -based cells. A possible cause for the lower

JSC s other than the lack of photocurrent generation in the high energy UV-visible region is a reduction in the exciton dissociation efficiency, whereby dissociation at the donor-acceptor interface is more difficult. To potentially remedy this issue and enhance the J SC output of these all-BsubPc devices, I would recommend the integration of an unsymmetric μ-oxo-(BsubPc) 2 interlayer. To be specific, F 12 BsubPc-O-BsubPc should be sandwiched between F/Cl-F12 BsubPc and Cl-BsubPc while Cl 6BsubPc-O-BsubPc should be sandwiched between Cl-Cl 6BsubPc and Cl-BsubPc. Their unsymmetric nature, combined with their structural similarities to both the donor and acceptor material, could create a favourable electronic bridge or surfactant for charge carriers to dissociate at the interface. To complete this study, the Cl 12 BsubPc-O-BsubPc derivative should also be integrated into an all-BsubPc device of the following configuration:

ITO/Cl-BsubPc/Cl 12 BsubPc-O-BsubPc/Cl-Cl 12 BsubPc/BCP/Al.

Table 9.2 . Device performance characteristics reported by Torres et al .25 and Jones et al .26

-2 Device Architecture VOC (V) JSC (mA·cm ) FF PCE (%) ITO/Cl-BsubPc/ 0.94 2.1 0.49 0.96 a F-F12 BsubPc/BCP/Al ITO/Cl-BsubPc/ 0.71 2.2 0.34 0.52 a Cl-F12 BsubPc/BCP/Al ITO/MoO x/Cl-BsubPc/ 1.32 3.28 0.63 2.70 b Cl-Cl 6BsubPc/BCP/Al ITO/Cl-BsubPc/ 0.92 5.4 0.61 3.0 a C60 /BCP/Al ITO/MoO x/Cl-BsubPc/ 1.10 4.66 0.65 3.29 b C60 /BCP/Al a Data taken from Torres et al .25 b Data taken from Jones et al .26

189

9.2.4 Other Symmetric μ-Oxo-(BsubPc) 2s

Following the proposed future work on the unsymmetric μ-oxo-(BsubPc) 2s in OPVs, it would also be worthwhile to synthesize a set of symmetric derivatives consisting of (F 12 BsubPc) 2O,

(Cl 6BsubPc) 2O, and (Cl 12 BsubPc) 2O. With these compounds, a systematic study into the effect of halogenation onto the periphery of one and both of the BsubPc subunits of the μ-oxo-

(BsubPc) 2 structure would be complete. To synthesize these compounds, I propose on adopting a similar approach for the unsymmetric μ-oxo-(BsubPc)2s (Scheme 9.1). Br-F12 BsubPc, Cl-

Cl6 BsubPc, and Cl-Cl 12 BsubPc are to be converted to their hydroxy-substituted form prior to their condensation reaction with their respective halo-substituted precursor to afford the target products. For the hydrolysis step, I would recommend using our group's recently developed and pyridine-free method. 27

Scheme 9.1 . Proposed synthesis of (F 12 BsubPc) 2O, (Cl 6BsubPc) 2O, and (Cl 12 BsubPc) 2O.

190

9.2.5 Cl-Cl 6BsubNc as a Synthetic Target

A mixture of Cl-Cl nBsubNcs with higher levels of peripheral chlorinates was shown to produce PHJ OPVs with higher PCEs than those with lower levels of peripheral chlorinates. To further confirm that having more chlorine atoms at the periphery/bay position is beneficial, I would recommend a pursuit into preparing Cl-Cl 6BsubNc. To make this molecule, I propose the following five pathways (Figure 9.3): 1) Chlorination of 2,3-dicyanonapthalene (2,3-DCNAP, red ); 2) Chlorination of α,α,α',α' -tetrabromo-o-xylene (TBX, blue );

3) Chlorination of literature-Cl-Cl nBsubNc ( green ); 4) Diels-Alder Adduct I ( pink ); and 5) Diels-Alder Adduct II ( purple ). The red, blue, and pink pathways all involve going through a 1,4-dichloro-2,3- dicyanonaphthalene intermediate before its cyclotrimerization with boron trichloride to give Cl-

Cl 6BsubNc. The red pathway involves chlorinating 2,3-DCNAP at both the 1- and 4-position (i.e. bay position). The blue pathway involves chlorinating TBX at both the benzylic positions prior to its reaction with fumaronitrile to form the desired intermediate. The pink and purple pathways begin with the Diels-Alder reaction of 2,3-DCNAP with hexachlorocyclopentadiene, 28 followed by chlorination of the adduct at the two bay positions to yield the dichlorinated adduct. The two pathways split apart at this point, whereby a Retro Diels-Alder reaction is proposed to give the desired intermediate via the pink pathway, while a cyclotrimerization and subsequent

Retro Diels-Alder reaction is proposed to afford Cl-Cl 6BsubNc via the purple pathway. The green pathway involves chlorinating all available bay positions of Cl-Cl nBsubNc.

191

Figure 9.3 . The five proposed synthetic routes to Cl-Cl 6BsubNc.

Some preliminary work has been done in each of the proposed routes. For the blue pathway, attempts to chlorinate TBX at the benzylic positions were made via adaptations from literature procedures 29,30 involving N-chlorosuccinimide (NCS) as the chlorinating agent in carbon tetrachloride (Scheme 9.2). The first reaction was heated to 75 °C, showing no progress via HPLC after a day. Given that the action of NCS proceeds via a radical mechanism, two radical initiators - benzoyl peroxide (BPO) and 2,2'-azodi(2-methylbutyronitrile) (V59) - were next used in separate experiments. Their use resulted in a colour change in the reaction mixture from beige to orange, signifying the release of bromine from TBX. The last reaction was performed under UV-irradiation at room temperature and it also produced the same beige-to-orange colour change. Based on these results and given the steric bulk of the bromine atoms, it is unlikely for chlorination to take place at the benzylic positions. Further efforts into this route are not warranted.

192

Scheme 9.2 . Attempted synthesis of α,α,α',α'-tetrabromo-α,α'-dichloro-o-xylene.

For the red pathway, a similar NCS chemistry was adopted in an effort to chlorinate 2,3-DCNAP and produce 1,4-dichloro-2,3-dicyanonaphthalene (Scheme 9.3). HPLC analysis of the reaction mixture showed the formation of a new peak, signifying that a new compound was being produced. However, a LRMS analysis of the reaction mixture was not consistent with the monochlorinated or dichlorinated product. The use of radical initiators like BPO or UV- irradiation should next be employed for this specific route.

Scheme 9.3 . Attempted synthesis of 1,4-dichloro-2,3-dicyanonaphthalene from 2,3- dicyanonaphthalene.

For the green pathway, a commercial sample of Cl-BsubNc was treated with BCl 3 in 1,2- dichlorobenzene at room temperature to "complete" the bay position chlorination with the aim of making Cl-Cl 6BsubNc (Scheme 9.4). The BsubNc chromophore was found absent within two

193

hours of reaction, demonstrating the sensitivity of the BsubNc structure to BCl 3 and perhaps to acids in general. For this reason, this specific route is not worth a further pursuit.

Scheme 9.4 . Attempted synthesis of Cl-Cl 6BsubNc from Cl-Cl nBsubNc.

For the pink and purple pathways, 2,3-DCNAP was treated with hexachlorocyclopentadiene in 1,2,4-trichlorobenzene at 160 °C. At the 20 hr mark, there was no sign of a reaction as determined by HPLC. It is very likely that a much higher reaction temperature is needed given that the aromaticity of 2,3-DCNAP would have to be broken to form the Diels-Alder adduct. Further efforts into this pathway should be considered.

9.2.6 Non-Peripherally Halogenated BsubNcs In Chapter 6, numerous attempts were made to prepare non-peripherally chlorinated Cl-BsubNc. All efforts were proven unsuccessful. To examine the photo- and electro-physical properties and OPV device characteristics of a non-peripherally halogenated BsubNc, an endeavour should be made to synthesize either F-BsubNc or Br-BsubNc. Br-BsubNc has already been prepared by Kobayashi et al., 31 whereby 2,3-DCNAP was first dissolved in 2,3-dimethyl-6-tert - o butylnaphthalene at 80 C under at an atmosphere of nitrogen and then slowly treated with BBr 3 before being heated to 180 oC. However, its purity has been questioned by Torres et al .32 The electron-rich, polysubstituted naphthalene solvent was made in-house prior to its use and is a key component in significantly reducing peripheral halogenation. However, the paper was not very clear on stating whether this side reaction was completely suppressed.

194

Some preliminary work has been done in this area. With the aim of making Br-BsubPc, Kobayashi’s method was first adapted with the use of 1-methylnaphthalene (1-MNAP) as the solvent (Table 9.3). Although 1-MNAP is not as electron rich as 2,3-dimethyl-6-tert - butylnaphthalene, it was chosen due to its commercial availability and thus, does not require additional synthetic work. With its use, the formation of a BsubNc chromophore was not observed (Method 1.1). Repeating this reaction, but going to a lower final temperature of 130 ˚C also did not produce a BsubNc compound (Method 1.2). The synthetic process to Br-BsubPc 3 was next adapted where 2,3-DCNAP was treated with BBr 3 in bromobenzene at room temperature (Method 1.3) and this also did not produce the desired product. Given the stronger

Lewis acidity of BBr 3 relative to BCl 3 and the sensitivity of Cl-BsubNc to BCl 3 (as described earlier), 1-MNAP was next incorporated as a cosolvent to mop up or lessen the reactivity of BBr 3 and to scavenge the bromine (Method 1.4). A single, faint peak with a UV-vis absorption profile characteristic of a BsubNc was observed after two hours following BBr 3 addition. This result was very encouraging of course, but this peak was found to decrease in intensity over time until it became nearly undetected at the 23-hour mark. This observation is a clear sign of a stability issue and for this reason, the reaction was subsequently carried out at 0 °C and allowed to slowly warm up to room temperature (Method 1.5). Again, a single and minor BsubNc peak was formed but its chromophore did not degrade over time unlike the previous trial. To potentially improve the very low conversion, the previous reaction was repeated in a closed system to reduce the chance of BBr 3 from leaving the reactor (Method 1.6). This process did not lead to any BsubNc formation, indicating that the reaction progress of interest is very sensitive to the amount of BBr 3 in the mixture. Further scoping of reaction conditions was performed where the reaction mixture was heated up to 130 °C (Method 1.7) and a substitution of bromobenzene for 1,2,4- trichlorobenzene was made (Method 1.8). In the former process, no BsubNc chromophore was observed. In the latter process, three BsubNc compounds were detected at the 15-minute mark and four compounds were detected at the 75-minute mark. Their intensities were also weak like those in Method 1.4 and Method 1.5. Based on these preliminary results, the most promising process is the use of a bromobenzene/1-MNAP solvent mixture (Method 1.4 & 1.5) since only a single BsubNc chromophore was formed. I would next recommend repeating these experiments

195

to determine that they are reproducible and more importantly, to positively identify whether the formed BsubNc compound is or is not halogenated at the periphery. Following this, other variables like the concentration of 2,3-DCNAP, amount of BBr 3 addition, and reaction temperature should all be explored in an aim to find the best conditions for making Br-BsubNc.

Table 9.3 . Attempted syntheses of Br-BsubNc. Method Boron Solvent System Temperature (°C) Number of BsubNc Template/Source Products 1.1 BBr 3 (0.5 eq) 1-MNAP 80→180 0 1.2 BBr 3 (0.5 eq) 1-MNAP 80→130 0 1.3 BBr 3 (0.33 eq) BrBZH rt 0 bc 1.4 BBr 3 (0.5 eq) BrBZH:1-MNAP (3:1) rt 1 b 1.5 BBr 3 (0.5 eq) BrBZH:1-MNAP (3:1) 0→rt 1 a 1.6 BBr 3 (0.5 eq) BrBZH:1-MNAP (3:1) 0→rt 0 1.7 BBr 3 (0.5 eq) BrBZH:1-MNAP (3:1) rt→130 0 b 1.8 BBr 3 (0.5 eq) 1,2,4-TCB:1-MNAP (3:1) rt→130 3 (15 min) 4b (1h 15 min) *BrBZH = bromobenzene; 1,2,4-TCB = 1,2,4-trichlorobenzene; 1-MNAP = 1-methylnaphthalene. a Closed system. b Faint detection. c Chromophore degraded over time.

F-BsubNc is not a known compound in the literature. My initial efforts to make this compound involved heating 2,3-DCNAP in the presence of BF 3·OEt 2 in four different solvents (Table 9.4, Method 2.1-2.4). This procedure was adapted from the synthesis of F-BsubPc, whereby Br- 33 BsubPc was treated with BF 3·OEt 2 and the mixture was heated to ~110 ˚C overnight. Three of the four solvents were ether-based (anisole, dibutyl ether, diethylene glycol dimethyl ether) to ensure good miscibility with the diethyl ether complex of BF 3 while the fourth solvent was chlorobenzene, a common aromatic solvent used in BsubPc reactions. A BsubNc chromophore was not detected in any of the four reactions, suggesting that a higher temperature may be needed for the cyclotrimerization reaction or perhaps less BF 3·OEt 2 should be added due to the likely sensitivity of the BsubNc chromophore to Lewis acid. Considering that BF 3·OEt 2 has a boiling point close to 130 °C, another solvated BF 3 with a higher boiling point such as BF 3·THF (b.p. = 180 °C) is recommended to access higher temperatures. It is worth noting that the use of

BF 3 with a coordinated basic ligand/solvent such as BF3·THF may not actually work due to a 34 strong association between the ligand/solvent and the BF 3. This complex needs to break in

196

order for the BF 3 to coordinate with 2,3-DCNAP and initiate the desired cyclotrimerization reaction.

Table 9.4 . Attempted syntheses of F-BsubNc. Method Boron Template/Source Solvent System Temperature (°C) Number of BsubNc Products 2.1 BF 3·OEt 2 (25 eq) ClBZH 130 0 2.2 BF 3·OEt 2 (25 eq) ANS 130 0 2.3 BF 3·OEt 2 (25 eq) DBE 130 0 2.4 BF 3·OEt 2 (25 eq) DIGLYME 130 0 *ClBZH = chlorobenzene; ANS = anisole; DBE = dibutyl ether; DIGLYME = diethylene glycol dimethyl ether.

9.2.7 Application of Cl-Cl nBsubNcs in OLEDs The adaptability of organic semiconductors among various OE technologies is not uncommon. For example, the Pcs, 35-38 BsubPcs, 39-42 and tetrabenzoporphyrins 43-46 have all been applied in

OPVs, OLEDs, OFETs, and NLOs. Given the recent success of Cl-Cl nBsubNcs as a functional material in OPVs, 8,47 coupled with its decent photoluminescence quantum yield, 9,32 it is worthwhile to explore its utility in OLEDs. Cl-Cl nBsubNcs would likely serve as a red-emitting material in OLED devices given its solid state photoluminescence in the 738-750 nm range.

9.2.8 BPI as an Electron Transport Layer in OPVs The absorption properties of the group XIII complexes of BPI were not desirable for the role as an organic photoactive material in OPVs. However, BPI could potentially serve as an electron transport layer (ETL) or electron injection layer for Pc- or BsubPc-based OPV devices because of its suitable electrochemical properties and its structural similarities to either isoindole-based macrocycle. The ETL, situated between the photoactive acceptor layer and the cathode, aids to extract and carry electrons from the former to the latter layer while blocking the flow of holes in this direction. This is facilitated by a low lying LUMO level for exergonic electron transfer and a very low lying HOMO level to block hole transfer. 48 A CV of BPI reveals n-type behaviour exclusively, a feature that is essential to the role of an ETL (Figure 9.4). Even just as important, BPI is anticipated to form a good interfacial contact with Pc or BsubPc as a result of structural similarities ("like interacts with like"). This could likely lead to an enhanced charge extraction

197

efficiency ( i.e. higher J SC ) and a higher overall PCE. In contrast, commonly used ETLs for OPVs 49 are based on alkali metals such as zinc oxide (ZnO) and titanium oxide (TiO x), whereby their inorganic nature do not form good contacts with organic surfaces.

Figure 9.4 . Cyclic voltammogram of BPI in DCM with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference (E 1/2,red = -0.012 V vs. Ag/AgCl) at room temperature. The dashed vertical line represents the reductive process.

9.2.9 BsubPys in OPVs via Vacuum Thermal Evaporation In Chapter 8, MeO-n-BsubPy films produced from an OVPD process resulted in PHJ OPV cells that performed very poorly with no observable photocurrent contribution from BsubPy. The work did not conclude whether the poor performance was attributed to the nature of the BsubPy material or to the OVPD process. To determine this, it is recommended that a control experiment using a well-known functional material such as Cl-BsubPc is carried out. This includes growing films of Cl-BsubPc via the same OVPD process and merging them into devices using the same architecture. If the device characteristics are similar to a standard cell based on Cl-BsubPc, then the poor performance of the MeO-n-BsubPy cells is likely connected to the material’s properties. If the device characteristics are poor with the Cl-BsubPc control, then the poor performance is likely related to the OVPD process.

198

For MeO-mPh-BsubPy, we were actually able to synthesize and purify a very small sample (21 mg), unfortunately not in a reproducible way. This quantity was also not sufficient for a study in OPV devices with our current vacuum deposition system. However, a new design to our deposition system is underway that can accommodate very small quantities. Once this system is operational, this material should be immediately tested and its device performance can be finally fully characterized.

199

9.3 References 1. de la Torre, G.; Claessens, C. G.; Torres, T. Chemical Communications 2007 , 2000-2015. 2. Morse, G. E.; Bender, T. P. ACS Applied Materials & Interfaces 2012 , 4, 5055-5068. 3. Dang, J. D.; Virdo, J. D.; Lessard, B. H.; Bultz, E.; Paton, A. S.; Bender, T. P. Macromolecules 2012 , 45 , 7791-7798. 4. Lessard, B. H.; Bender, T. P. Macromolecular Rapid Communications 2013 , 34 , 568- 573. 5. Lessard, B. H.; Sampson, K. L.; Plint, T.; Bender, T. P. Journal of Polymer Science Part a-Polymer Chemistry 2015 , 53 , 1996-2006. 6. Dang, J. D.; Fulford, M. V.; Kamino, B. A.; Paton, A. S.; Bender, T. P. Dalton Transactions 2015 , 44 , 4280-4288. 7. Castrucci, J. S.; Garner, R. K.; Dang, J. D.; Thibau, E.; Lu, Z. H.; Bender, T. P. ACS Applied Materials & Interfaces Paper under revision. 8. Cnops, K.; Empl, M. A.; Heremans, P.; Rand, B. P.; Cheyns, D.; Verreet, B. Nature Communications 2014 , 5, 3406. 9. Dang, J. D.; Josey, D.; Lough, A.; Li, Y.; Sifate, A.; Lu, Z.; Bender, T. P. Journal of Materials Chemistry A 2016 . 10. Martin, G.; Rojo, G.; Agullo-Lopez, F.; Ferro, V. R.; Garcia de la Vega, J. M.; Martinez- Diaz, M. V.; Torres, T.; Ledoux, I.; Zyss, J. J. Phys. Chem. B 2002 , 106 , 13139-13145. 11. Dang, J. D.; Bender, T. P. Inorganic Chemistry Communications 2013 , 30 , 147-151. 12. Inokuma, Y.; Kwon, J. H.; Ahn, T. K.; Yoo, M. C.; Kim, D.; Osuka, A. Angewandte Chemie-International Edition 2006 , 45 , 961-964. 13. Inokuma, Y.; Osuka, A. Dalton Transactions 2008 , 2517-2526. 14. Makarova, E. A.; Shimizu, S.; Matsuda, A.; Luk'yanets, E. A.; Kobayashi, N. Chemical Communications 2008 , 2109-2111. 15. Cheng, P.; Zhao, X. G.; Zhou, W. Y.; Hou, J. H.; Li, Y. F.; Zhan, X. W. Organic Electronics 2014 , 15 , 2270-2276. 16. Lin, Y. Z.; Zhan, X. W. Materials Horizons 2014 , 1, 470-488.

200

17. Sun, D.; Meng, D.; Cai, Y. H.; Fan, B. B.; Li, Y.; Jiang, W.; Huo, L. J.; Sun, Y. M.; Wang, Z. H. Journal of the American Chemical Society 2015 , 137 , 11156-11162. 18. D'Andrade, B. W.; Datta, S.; Forrest, S. R.; Djurovich, P.; Polikarpov, E.; Thompson, M. E. Organic Electronics 2005 , 6, 11-20. 19. Djurovich, P. I.; Mayo, E. I.; Forrest, S. R.; Thompson, M. E. Organic Electronics 2009 , 10 , 515-520. 20. Ge, Y.; Whitten, J. E. Chemical Physics Letters 2007 , 448 , 65-69. 21. Morse, G. E.; Gantz, J. L.; Steirer, K. X.; Armstrong, N. R.; Bender, T. P. ACS Applied Materials & Interfaces 2014 , 6, 1515-1524. 22. Sullivan, P.; Duraud, A.; Hancox, I.; Beaumont, N.; Mirri, G.; Tucker, J. H. R.; Hatton, R. A.; Shipman, M.; Jones, T. S. Advanced Energy Materials 2011 , 1, 352-355. 23. Castrucci, J. S.; Josey, D. S.; Thibau, E.; Lu, Z.-H.; Bender, T. P. Journal of Physical Chemistry Letters 2015 , 6, 3121-3125. 24. Benning, P. J.; Poirier, D. M.; Ohno, T. R.; Chen, Y.; Jost, M. B.; Stepniak, F.; Kroll, G. H.; Weaver, J. H.; Fure, J.; Smalley, R. E. Physical Review B 1992 , 45 , 6899-6913. 25. Gommans, H.; Aernouts, T.; Verreet, B.; Heremans, P.; Medina, A.; Claessens, C. G.; Torres, T. Advanced Functional Materials 2009 , 19 , 3435-3439. 26. Sullivan, P.; Schumann, S.; Da Campo, R.; Howells, T.; Duraud, A.; Shipman, M.; Hatton, R. A.; Jones, T. S. Advanced Energy Materials 2013 , 3, 239-244. 27. Paton, A. S.; Bender, T. P. Journal of Porphyrins and Phthalocyanines 2014 , 18 , 1051- 1056. 28. Danish, A. A.; Silverman, M.; Tajima, Y. A. Journal of the American Chemical Society 1954 , 76 , 6144-6150. 29. Tuleen, D. L.; Marcum, V. C. Journal of Organic Chemistry 1967 , 32 , 204-206. 30. Newkome, G. R.; Kiefer, G. E.; Xia, Y. J.; Gupta, V. K. Synthesis-Stuttgart 1984 , 676- 679. 31. Kobayashi, N.; Ishizaki, T.; Ishii, K.; Konami, H. Journal of the American Chemical Society 1999 , 121 , 9096-9110.

201

32. Nonell, S.; Rubio, N.; del Rey, B.; Torres, T. Journal of the Chemical Society, Perkin Transactions 2 2000 , 1091-1094. 33. Fulford, M. V.; Jaidka, D.; Paton, A. S.; Morse, G. E.; Brisson, E. R. L.; Lough, A. J.; Bender, T. P. Journal of Chemical & Engineering Data 2012 , 57 , 2756-2765. 34. Maria, P. C.; Gal, J. F.; Defranceschi, J.; Fargin, E. Journal of the American Chemical Society 1987 , 109 , 483-492. 35. De La Torre, G.; Vázquez, P.; Agulló-López, F.; Torres, T. Journal of Materials Chemistry 1998 , 8, 1671-1683. 36. Fleetham, T. B.; Mudrick, J. P.; Cao, W.; Klimes, K.; Xue, J.; Li, J. ACS Applied Materials & Interfaces 2014 , 6, 7254-7259. 37. Lu, Z. Y.; Li, X.; Wang, Y.; Xiao, J.; Xu, P. L. Current Applied Physics 2014 , 14 , 1465- 1469. 38. Kvitschal, A.; Cruz-Cruz, I.; Hummelgen, I. A. Organic Electronics 2015 , 27 , 155-159. 39. Claessens, C. G.; Gonzalez-Rodriguez, D.; Torres, T.; Martin, G.; Agullo-Lopez, F.; Ledoux, I.; Zyss, J.; Ferro, V. R.; de la Vega, J. M. G. Journal of Physical Chemistry B 2005 , 109 , 3800-3806. 40. Morse, G. E.; Helander, M. G.; Maka, J. F.; Lu, Z.-H.; Bender, T. P. ACS Applied Materials & Interfaces 2010 , 2, 1934-1944. 41. Renshaw, C. K.; Xu, X.; Forrest, S. R. Organic Electronics 2010 , 11 , 175-178. 42. Ebenhoch, B.; Prasetya, N. B. A.; Rotello, V. M.; Cooke, G.; Samuel, I. D. W. Journal of Materials Chemistry A 2015 , 3, 7345-7352. 43. Ono, N.; Ito, S.; Wu, C. H.; Chen, C. H.; Wen, T. C. Chemical Physics 2000 , 262 , 467- 473. 44. Shea, P. B.; Kanicki, J.; Ono, N. Journal of Applied Physics 2005 , 98 , 014503. 45. Matsuo, Y.; Sato, Y.; Niinomi, T.; Soga, I.; Tanaka, H.; Nakamura, E. Journal of the American Chemical Society 2009 , 131 , 16048-16050. 46. Graham, K. R.; Yang, Y.; Sommer, J. R.; Shelton, A. H.; Schanze, K. S.; Xue, J.; Reynolds, J. R. Chemistry of Materials 2011 , 23 , 5305-5312.

202

47. Verreet, B.; Cnops, K.; Cheyns, D.; Heremans, P.; Stesmans, A.; Zango, G.; Claessens, C. G.; Torres, T.; Rand, B. P. Advanced Energy Materials 2014 , 4, 1301413/1301411- 1301413/1301418. 48. Ma, H.; Yip, H. L.; Huang, F.; Jen, A. K. Y. Advanced Functional Materials 2010 , 20 , 1371-1388. 49. Lattante, S. Electronics 2014 , 3, 132-164.

203

Appendix General Experimental Section

Materials . 2,2,6,6-Tetramethyl-1-piperidinyloxy (TEMPO), 4-acetoxystyrene (4-AS), 4- methylstyrene (4-MS), benzoyl peroxide (BPO), boron tribromide, boron trichloride, boron trifluoride diethyl etherate, dicyclohexano-18-crown-6, aluminum chloride, aluminum chloride hexahydrate, iron(III) chloride hexahydrate, tin(IV) chloride, tin(IV) chloride pentahydrate, tripotassium phosphate, decamethylferrocene, tetrabutylammonium perchlorate, sexithiophene

(α-6T), molybdenum(VI) oxide (MoO x, 99.98% trace metals basis), 1-methylnaphthalene, 1,2,4- trichlorobenzene, chloronaphthalene, anisole, dibutyl ether, diethylene glycol dimethyl ether, dipentene, (-)-β-pinene, 1-octadecene, para -cymene, oxazine 170, bathocuprione (BCP, purity: 99.6%), decamethylferrocene, tetrabutylammonium perchlorate, neutral aluminum oxide

(Al 2O3), phenol, n-hexanol, aluminium chloride, tri-n-butylamine, calcium chloride, celite, deuterated dimethyl sulfoxide (DMSO-d6), ethanol, 1-nitroso-2-naphthol, p-toluenesulfonyl chloride, sodium ethoxide, sodium tert -butoxide, triethyl borate, boric acid, phthalimide, phenylacetic acid, sulfolane, dimethyl sulfone, n-butyllithium, and phenyl ether were purchased from Sigma-Aldrich Chemical Company (Mississauga, Ontario, Canada). Phthalonitrile, 4,5- dichlorophthalonitrile, tetrafluorophthalonitrile, fumaronitrile, and 2-aminopyridine were purchased from TCI America (Portland, Oregon, USA). Bromobenzene was purchased from Alfa Aesar (Ward Hill, Massachusetts, USA). Pentafluorophenol was purchased from Oakwood Products Inc (Estill, South Carolina, USA). Deuterated chloroform with 0.05% ( v/v ) tetramethylsilane (TMS) was purchased from Cambridge Isotope Laboratories (St. Leonard, Quebec, Canada). Tetrachlorophthalonitrile was purchased from Hangzhou Dayangchem Company Limited (Hangzhou, China). 2,2’,2’’-(1,3,5-Benzinetriyl)-tris(1-phenyl-1-H- benzimidazole) ( TPBi) and commercial chloro boron subnaphthalocyanine (Cl-BsubNc) were purchased from Luminescence Technology Corp (Hsin-Chu, Taiwan). α,α,α',α'-Tetrabromo-o- xylene was purchased from Leap Labchem Company Limited (Hangzhou, China). Gallium(III) chloride and indium(III) chloride were purchased from Strem Chemicals Inc. (Newburyport,

Massachusetts, USA). C60 (99.5%) and C 70 (99.5%) were purchased from SES Research

204

(Houston, Texas, USA). Poly(3,4-ethylenedioxythiophene) poly(styrenesulfonate) (PEDOT:PSS) was purchased from Heraeus Precious Metals GmbH & Co. (Leverkusen, Germany). Silver (Ag, purity: 99.999%) was purchased from R.D. Mathis Company (Long Beach, California, USA). Silver paint (Conductive Silver 187) was purchased from Ted Pella Inc. (Redding, California, USA). All other ACS and HPLC grade solvents, potassium hydroxide, sulfuric acid, sodium bicarbonate, basic alumina, potassium carbonate, sodium hydroxide, ethylene glycol, sodium iodide, hydrochloric acid, and charcoal were purchased from Caledon Labs (Caledon, Ontario,

Canada). α-6T, C 60 , and C 70 were purified once by train sublimation while all other materials were used without further purification. Extraction thimbles (cellulose, single thickness) and aluminum plates coated with silica (pore size of 60Å) and fluorescent indicator were purchased from Whatman International Ltd. (Kent, England). Glass extraction thimbles were purchased from Chemglass Life Sciences LLC (Vineland, New Jersey). Silica Gel P60 (mesh size 40-63 µm) was purchased from SiliCycle Inc. (Quebec City, Quebec, Canada). Glass substrates coated with indium-tin oxide (ITO) were purchased from Thin Film Devices Inc. (Anaheim, California, USA). Glass slides were purchased from Corning Incorporated (Corning, New York, USA). 'Wet' nitrogen (20 ppm of water) gas cylinder was purchased from Linde Canada Limited (Mississauga, Ontario, Canada). Band heater (L = 2", OD = 2"), frame structure, copper-nickel pipe (L = 12", 2 pipe size), and ceramic insulator were purchased from McMaster-Carr (Aurora, Ohio, USA). The temperature controller was purchased from Fisher Scientific (Pittsburgh, Pennsylvania, USA). The ball valve, needle valve, and stainless steel gas line were purchased from Weston Valve & Fitting Ltd (Mississauga, Ontario, Canada). The pressure gauge was purchased from Kurt J. Lesker Company (Concord, Ontario, Canada). The K type thermocouple wire was purchased from Omega Engineering Inc. (Laval, Quebec, Canada). The power controller and heating tape were purchased from HTS/Amptek Company (Stafford, Texas, USA). The temperature reader was purchased from Cole-Parmer Canada Company (Montreal, Quebec, Canada). The stainless steel sublimation boat was custom made by the Machine Shop (Department of Chemistry, University of Toronto). The glass reactor (L = 28", ID = 2"), glass inserts, and condenser for sublimation experiments were custom made by the Glass Blowing Shop (Department of Chemistry, University of Toronto).

205

Methods . Thin layer chromatography (TLC) was performed on aluminum plates coated with silica (pore size of 60Å) and fluorescent indicator and visualized under UV (254 nm) light. Column chromatography was performed using Silica Gel P60 (mesh size 40-63 µm) or alumina (various). Soxhlet extractions were performed using single thickness cellulose extraction thimbles (33 × 118 mm, 25 × 80mm). Nuclear magnetic resonance (NMR) spectra were recorded on a Varian Mercury spectrometer at 23 °C, operating at 400 MHz for 1H NMR. Chemical shifts (δ) are reported in parts per million (ppm) relative to tetramethylsilane or a residual solvent peak. Coupling constants ( J) are reported in hertz (Hz). Spin multiplicities are designated by the following abbreviations: s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet) and br (broad). Fourier transform infrared spectroscopy (FTIR) were performed on a PerkinElmer Spectrum 100 FT-IR spectrometer using pellets prepared with KBr. Ultraviolet-visible (UV-vis) absorption spectra were acquired on a PerkinElmer Lambda 25 UV/Vis spectrometer using a PerkinElmer quartz cuvette with a 10 mm path length. Photoluminescence (PL) spectra were recorded on a PerkinElmer LS55 fluorescence spectrometer using a PerkinElmer quartz cuvette with a 10mm path length. High pressure liquid chromatography (HPLC) analysis was carried out on a Waters 2695 separation module with a Waters 2998 photodiode array and a Waters Styragel® HR 2 THF 4.6 x 300 mm column. The mobile phase used was HPLC grade acetonitrile (80% by volume) and N,N -dimethylformamide (20% by volume). Gel permeation chromatography (GPC) were conducted on a Waters 2695 separation module with a Waters 2998 photodiode array, a Waters 2414 refractive index detector, and two Waters Styragel® 5 µm, HR 4E 7.8 x 300 mm column in series. The mobile phase used was HPLC grade THF, which was ran at a flow rate of 1.2 mL/min. Gas chromatography (GC) was conducted on a PerkinElmer AutoSystem with a Rtx®-35 (Fused Silica) column with dimensions of 15 m x 0.32 mm x 1.00 µm. Helium was used as the mobile phase and a flame ionization detector was used for detection. Electrochemical measurements were carried out using a Bioanalytical Systems C3 electrochemical workstation. The working electrode was a 1 mm platinum disk, the counter electrode was a platinum wire, and the reference electrode was Ag/AgCl 2 saturated salt solution. Spec-grade solvents were purged with argon gas at room temperature prior to their use. Three cycles from +1.6 to -1.6 V at a scan rate of 100 mV/s were measured for each sample.

206

Tetrabutylammonium perchlorate (0.1 M) was used as the supporting electrolyte. Decamethylferrocene was used as an internal reference. All half-wave potentials were corrected to the half-wave reduction potential (E 1/2,red ) of decamethylferrocene, which was previously reported to be -0.012 V vs. Ag/AgCl. 1 Photoelectron measurements were performed using a PHI 5500 Multi-Technique system attached to a Kurt J. Lesker multiaccess chamber ultra high vacuum cluster tool (base pressure of ~10 -10 Torr). Monochromated Al Kα photon source ( hν = 1486.7 eV) was used for X-ray photoelectron spectroscopy (XPS) while a non-monochromated He Iα photon source ( hν = 21.22 eV) was used for ultraviolet photoelectron spectroscopy (UPS). Work function and valence-band measurements were carried out using UPS with the sample tilted to a take-off angle of 89° and under an applied bias of -15 V. The various organic molecules were deposited in a dedicated organic chamber from alumina crucibles transfer-arm evaporator (TAE) cell onto freshly cleaved highly ordered pyrolytic graphite (HOPG) substrates. HOMO energy levels for each molecule were determined from 12 nm thick films as measured by a calibrated quartz crystal microbalance (QCM). Accurate mass determinations (HRMS) were carried out on an Agilent 6538 Q-TOF mass spectrometer equipped with an Agilent 1200 HPLC and an ESI ion source or on an AccuTOF mass spectrometer (JEOL USA Inc., Peabody, Massachusetts, USA) with a DART-SVP ion source (IonSense Inc., Saugus, Massachusetts, USA) using helium gas. Single crystal X-ray diffraction analyses were done by Dr. Alan Lough and details regarding the instrumentations and methods can be found online via the CCDC.

Device Fabrication . OPV devices were fabricated on 25 mm by 25 mm glass substrates coated with ITO having a sheet resistance of 15 Ω per square. The ITO was pre-patterned, leaving 8 mm from one side as uncoated glass. Substrates were cleaned by successive sonications in detergent and solvents, followed by 5 minutes of atmospheric plasma treatment. PEDOT:PSS was spin- coated onto the substrates, 500 rpm, 10 s; 4000 rpm, 30 s. Substrates were baked on a hot plate at

110 °C for 10 minutes, and then transferred into a nitrogen atmosphere glove box (O 2 < 10 ppm,

H2O < 10 ppm). Substrates were transferred to a custom-built thermal evaporation system attached to the nitrogen glove box without exposure to ambient conditions. All subsequent device layers were thermally evaporated at ~1.0 A/s and a working pressure of ~1 x 10 -7 Torr for

207

organic layers and ~1 x 10 -6 Torr for silver. Silver electrodes were evaporated through a shadow mask, defining 0.2 cm 2 as the active area for each device. A transfer back to the glove box was required between BCP or TPBi and silver layers to change the shadow masks.

Device Characterization . Layer thickness and deposition rates of evaporated films were monitored using a quartz crystal microbalance calibrated against films deposited on glass and measured with a KLA-Tencor P16+ surface profilometer. To enhance the electrical contact during testing, silver paint was applied to the ITO and metal electrode contact points and left to dry for 20 minutes. Devices were kept in the nitrogen-filled glove box throughout testing. Voltage sweeps of the devices were performed under full illumination by a 300W Xe arc lamp (Oriel) with an AM 1.5G filter, and the corresponding currents were measured with a Keithley 2401 Low Voltage SourceMeter. Light intensity was calibrated to 100 mW/cm 2 with reference to a calibrated silicon photodetector. Wavelengths scans at 10 nm intervals were performed using an in-line Cornerstone TM 260 1/4 m Monochromator and the corresponding currents were measured using a Newport Optical Power Meter 2936-R and converted to external quantum efficiencies using a reference wavelength scan of a calibrated silicon photodetector.

208

References 1. Noviandri, I.; Brown, K. N.; Fleming, D. S.; Gulyas, P. T.; Lay, P. A.; Masters, A. F.; Phillips, L. Journal of Physical Chemistry B 1999 , 103 , 6713-6722.

209

Appendix A: Supplementary Information for Chapter 2

Synthetic Procedures Bromo boron subphthalocyanine (Br-BsubPc, 1b). 1b was prepared according to a modified literature procedure. 1 To an oven-dried three-neck round bottom flask equipped with a condenser, an addition funnel, and a gas inlet was added phthalonitrile (42.3 g, 0.330 mol, 3.1 equiv), bromobenzene (136 mL), and toluene (362 mL) under argon. To the addition funnel was charged boron tribromide (10.0 mL, 26.5 g, 0.106 mol, 1 equiv), which was subsequently added to the reaction mixture drop wise. An immediate colour change to a dark brown was observed and the reaction was allowed to stir overnight. The following morning, the stirring was turned off and the reaction was left undisturbed for one hour. The reaction was gravity filtered and the solid was immediately washed with methanol (250 mL) and dried in an oven to give 1b (23.1 g, 46%)

1 as a brown solid. H NMR (400 MHz, CDCl 3) d 8.89-8.83 (m, 6H), 8.00-7.94 (m, 6H).

2:1 Poly(4-methylstyrene)-co -poly(4-acetoxystyrene) (9a ). A flame-dried three-neck round bottom flask equipped with a condenser, an internal temperature probe, and a gas inlet was purged with nitrogen. Two pipettes were packed with inhibitor remover replacement packing for removing tert-butylcatechol. 4-AS (11.31 g, 70 mmol) was uninhibited by passing it through one pipette and introduced into the reaction flask. 4-MS (16.49 g, 140 mmol) was uninhibited by passing it through the second pipette and introduced into the reaction flask. To the reaction mixture was added TEMPO (0.024 g, 0.15 mmol), followed by BPO (0.025 g, 0.10 mmol). A small amount of 1,2-dichlorobenzene (1.2 mL) was added as an internal standard to monitor the consumption of the monomer via GC. The reaction was purged with nitrogen for 15 minutes and allowed to heat to 125 °C. The reaction progress was monitored by GC and GPC, and the reaction was stopped when the polymer had reached a desirable molecular weight (Mw ~ 15,000 to 18,000 Da). The reaction was precipitated into methanol (1 L) and suction filtered to obtain 9a

1 as a white solid (16.2 g, 58%). H NMR (400 MHz, CDCl 3) d 7.05-6.17 (br d, 12H), 2.26 (br s, 9H), 1.78 (br s, 3H), 1.35 (br s, 6H).

210

4:1 Poly(4-methylstyrene)-co -poly(4-acetoxystyrene) (9b ). 9b was obtain as a white solid

1 (53%) via the same procedure used for the synthesis of 9a . H NMR (400 MHz, CDCl 3) d 7.05- 6.17 (br d, 20H), 2.26 (br s, 15H), 1.80 (br s, 5H), 1.35 (br s, 10H).

2:1 Poly(4-methylstyrene)-co -poly(4-vinyl phenol) (10a ). To a round bottom flask equipped with a condenser was added potassium hydroxide (1.12 g, 0.02 mol) and water (150 mL). To the solution was added 9a (2.00 g) and isopropanol (100 mL). The reaction mixture was heated to 75 °C for 12 hours and analyzed by FT-IR for the absence of the ester signal at 1768 cm -1. The reaction was allowed to cool to room temperature and the copolymer was precipitated out of solution by neutralizing the reaction mixture via a drop wise addition of 2 M HCl. The product was isolated by suction filtration and washed with water (3x 50 mL), followed by hexane (3 x 50 mL). The crude product was purified by dissolving it in minimum volume of THF with the assistance of a sonicator, precipitating it drop wise into ice cold stirring hexane (500 mL) via a cotton-plugged pipette, suction filtering the suspension, and washing the solids with water (3 x 50 mL) and hexane (3 x 50 mL) to obtain 10a (1.58 g, 88%) as a beige solid. 1H NMR (400

MHz, CDCl 3) d 7.10-6.20 (br d, 12H), 2.26 (br s, 6H), 1.86 (br s, 3H), 1.39 (br s, 6H).

4:1 Poly(4-methylstyrene)-co -poly(4-vinyl phenol) (10b ). To a round bottom flask equipped with a condenser was added 9b (2.50 g), toluene (60 mL), ethanol (20 mL), and concentrated sulfuric acid (5 drops). The reaction mixture was heated to 75 °C for 12 hours and analyzed by FT-IR for the absence of the ester signal at 1768 cm -1. The reaction was allowed to cool to room temperature where sodium bicarbonate (0.80 g) was added to the mixture and stirred for four hours to neutralize the sulfuric acid. The reaction mixture was filtered to remove the base and the filtrate was concentrated under reduced pressure. The crude product was purified by dissolving it in minimum volume of THF with the assistance of a sonicator, precipitating it drop wise into ice cold stirring hexane (600 mL) via a cotton-plugged pipette, suction filtering the suspension, and washing the solids with water (3 x 50 mL) and hexane (3 x 50 mL). The copolymer was re- precipitated a second time in hexane as described above and dried in a vacuum oven at 80°C

211

1 overnight to afford 10b (3.20 g, 85%) as a white solid. H NMR (400 MHz, CDCl 3) d 7.12-6.11 (br d, 20H), 2.27 (br s, 12H), 1.85 (br s, 5H), 1.38 (br s, 10H).

2:1 Poly(4-methylstyrene)-co -poly(phenoxy-boron-subphthalocyanine) (11a ). An oven-dried three-neck round bottom flask equipped with a condenser and a gas inlet was purged with argon for 15 minutes. To the reaction flask was added 10a (1.58 g) and anhydrous chlorobenzene (180 mL). The reaction was purged for 10 minutes before adding 1b (2.53 g, 5.33 mmol). The reaction mixture was heated to 120 °C under argon and the reaction progress was monitored via GPC. When the ratio of BsubPc-containing copolymer and unreacted 1b had reached a steady value, the reaction was stopped and was allowed to cool to room temperature. The reaction mixture was concentrated under reduced pressure to a volume of approximately 30 mL, precipitated drop wise into ice cold stirring methanol (1 L) via a cotton-plugged pipette, suction filtered, and washed with hexane (3 x 50 mL) and methanol (3 x 50 mL). The crude product was purified by adding it to a cellulose thimble and extracting with methanol in a Soxhlet extraction apparatus setup until the extract was clear. Three additional re-precipitation into ice cold methanol followed by a single re-precipitation into ice cold hexane was done as described above and the solid was dried in a vacuum oven at 80 °C overnight to afford 11a (1.85 g, 56%) as a

1 dark purple solid. H NMR (400 MHz, CDCl 3) d 8.74 (br s, 6H), 7.69 (br s, 6H), 7.05-6.12 (br d, 8H), 5.72 (br s, 2H), 5.19 (br s, 2H), 2.21 (br s, 6H), 1.85 (br s, 3H), 1.29 (br s, 6H).

4:1 Poly(4-methylstyrene)-co -poly(phenoxy-boron-subphthalocyanine) (11b ). An oven-dried three-neck round bottom flask equipped with a condenser and a gas inlet was purged with argon for 15 minutes. To the reaction flask was added 10b (2.00 g) and anhydrous chlorobenzene (250 mL). The reaction was purged for 10 minutes before adding 1b (1.90 g, 4.00 mmol). The reaction mixture was heated to 120 °C under argon and the reaction progress was monitored via GPC. When the ratio of BsubPc-containing copolymer and unreacted 1b had reached a steady value, the reaction was stopped and was allowed to cool to room temperature. The reaction mixture was concentrated under reduced pressure to a volume of approximately 30 mL, precipitated drop wise into ice cold stirring methanol (1 L) via a cotton-plugged pipette, suction

212

filtered, and washed with hexane (3 x 50 mL) and methanol (3 x 50 mL). The crude product was purified by adding it to a cellulose thimble and extracting with methanol in a Soxhlet extraction apparatus setup until the extract was clear. Three additional re-precipitation into ice cold methanol followed by a single re-precipitation into ice cold hexane was done as described above and the solid was dried in a vacuum oven at 80 °C overnight to afford 11b (1.98 g, 60%) as a

1 bright purple solid. H NMR (400 MHz, CDCl 3) d 8.75 (br s, 6H), 7.71 (br s, 6H), 7.05-6.10 (br d, 16H), 5.73 (br s, 2H), 5.11 (br s, 2H), 2.23 (br s, 12H), 1.81 (br s, 5H), 1.33 (br s, 10H).

213

Determination of Free-Radical Scavenging Capabilities of Br-BsubPc As mentioned in Chapter 2, 4-methylstyrene (4-MS) was homopolymerized using a 1:2 molar mixture of BPO:TEMPO at 130 °C in bulk (target molecular weight of 15 kg·mol -1). In one case, the addition of 0.04 wt % of Br-BsubPc was added to the mixture prior to the homopolymerization (2 mg of Br-BsubPc for 4.93 g of 4-MS). The homopolymerization conversion was determined by gravimetric analysis and the molecular weight distribution was determined using GPC. The resulting scaled conversion (ln(1-X)-1) versus time and number average molecular weight (M n) versus conversion plots are found in Figure S2.1. The 4-MS homopolymerization (with no additional Br-BsubPc) resulted in a linear increase in ln(1-X)-1 versus time and M n versus X. It is clear that the inclusion of a minimal amount of Br-BsubPc had a pronounced effect on the homopolymerization of 4-MS. Very little 4-MS homopolymer was produced until the ~4h mark, where the 4-MS homopolymerization apparently began (Figure S2.1). These results coincided with the significant change in the reaction solution colour going from a bright magenta at the beginning of the homopolymerization to a light brown-green after 4h (Figure S2.1C). These results indicated that while Br-BsubPc was present, the homopolymerization did not take place due to the scavenging of the propagating radicals. However, upon consumption or degradation of Br-BsubPc (as discussed in the chapter) the homopolymerization of 4-MS proceeded. These results therefore indicate that the inclusion of as little as 0.04 wt % of Br-BsubPc will result in no significant homopolymerization due to scavenging of the propagating radicals.

214

Figure S2.1 . A) Scaled conversion (ln(1-X)-1) versus time and B) number average molecular weight (M n) versus conversion plots for 4-methylstyrene (4-MS) homopolymerization using a 1:2 molar mixture of BPO:TEMPO at 130 °C in bulk (target molecular weight of 15 kg·mol -1) and a similar homopolymerization with the inclusion of 0.04 wt % of Br-BsubPc. C) Picture of the reaction mixture drawn from the homopolymerization of 4-MS with the inclusion of 0.04 wt % of Br-BsubPc

215

“Patterns of Reactivity” Method for Predicting Monomer Reactivity Ratios

Introduction to Reactivity Ratios

Classically, for a given pair of a monomers (M 1 and M 2) undergoing free-radical copolymerization, the reactivity ratios (r 1,2 and r 2,1 ) are defined as:

and (Eq S2.1)

The reactivity ratios for a given monomer pair are typically determined experimentally, since each pair of reactivity ratios is unique to that specific pair of monomers. There are countless possible monomer pairs, and experimentally-determined reactivity ratios have not been determined for all possible pairs directly. The Patterns of Reactivity method, 2 described briefly below, can be used to predict reactivity ratios in the absence of direct experimental data.

This method uses eight experimentally-determined input parameters to predict the reactivity ratios for a given pair of monomers, as shown in the pair of equations:

(Eq S2.2)

Here r 1,s and r 2,s are the experimentally-determined reactivity ratios of monomer 1 and monomer 2 copolymerized with styrene, respectively.

The polarity parameter, p i, for monomer i is calculated according to:

(Eq S2.3)

Here r i,a and r i,s are the experimentally-determined reactivity ratios of monomer i with acrylonitrile and styrene, respectively. Acrylonitrile and styrene are used for this calculation because they represent extremes in the range of radical polarity, with styrene having a very low value (zero) and acrylonitrile having one of the highest values known (0.701). 2

Finally, values for ui and vi are obtained by referring to data for the separate copolymerizations of monomer i with the members of a ‘basic monomer set.’ The basic monomer set includes five

216

monomers for which reliable data exists in the literature, namely: styrene, methyl methacrylate, methyl acrylate, methacrylonitrile and acrylonitrile. A plot is made of [log r B,i – log r B,s ] against p B. The subscript i refers to the monomer for which u and v values are to be determined; the subscript B refers to the members of the basic monomer set. The resulting plot yields a straight line, with the slope equal to –ui and the intercept on the ordinate axis equal to –vi, in conformance with equation (Eq S2.1).

This method can be used to calculate values for ui and vi even if copolymerization data is not available for monomer i with all five members of the basic monomer set. In that case, the extent of the data available can be used, as long as it includes data for acrylonitrile and styrene to ensure that the model is valid over a range of radical polarities.

217

Table S2.1 . Reactivity ratios of candidate monomers for copolymerization with 4-acetoxystyrene (predicted using the Patterns of Reactivity method 2). Actual values for acrylonitrile are given for comparison to calculated values. Monomer 1 Structure

4-acetoxystyrene Monomer 2 Structure r1,2 r2,1 Acrolein 0.27 0.34 O H2C H a Acrylonitrile 0.40 0.07

Crotonaldehyde 15.87 0.05 N-Vinylphthalimide 7.77 0.09

2-Vinylpyridine 0.56 1.06

4-Vinylpyridine 0.59 0.62

2,4-Dichlorostyrene 0.23 0.31

α-Methoxystyrene 4.01 0.07

α-Methylstyrene 1.35 0.47

p-Methylstyrene 1.03 0.83

N-Vinylsuccinimide 13.90 0.05

1-Vinyltetrazole 4.65 0.20

Vinyl methyl ketone 0.31 0.42

a Experimentally determined values, for comparison: r 1,2 =0.40 and r 2,1 =0.03

218

Figure S2.2 . Molecular weight (M w) and polydispersity index (PDI) over time for the low molecular weight copolymerization of 4-acetoxystyrene and 4-methylstyrene.

219

Monomer Consumption Rates for Low Molecular Weight Copolymerization

60%

50%

40% y = 0.0638x R² = 0.9876 30% y = 0.0536x R² = 0.9893 Consumption (%) Consumption 20%

10%

0% 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 Time (hours)

Methyl Styrene Consumption Acetoxy Styrene Consumption

Figure S2.3 . Percentage consumption of the 4-methylstyrene and 4-acetoxystyrene monomers over the course of the copolymerization. The slope of the lines indicates a consumption rate, which is also the rate at which the monomers were incorporated into the polymer product.

220

GPC Chromatograms

Figure S2.4 . GPC spectrum at 1.5 (top), 4.5 (middle), and 19 hours (bottom) for the reaction of 1b with 10a .

221

Figure S2.5 . GPC spectrum at 2 (top), 18 (middle), and 42 hours (bottom) for the reaction of 1b with 10b .

222

Figure S2.6 . NMR spectrum (400 MHz, CDCl 3) of 11a (top) and 11b (bottom).

223

Determination of Fluorescence Quantum Yields The fluorescence quantum yields (φ) for 11a and 11b were calculated using the formula below: 2 2 φ = φ R (I / I R)(OD R / OD)(n / n R ) (Eq S2.4) where I is the integrated fluorescence intensity, OD is the optical density or absorbance, and n is the refractive index of the solvent. The subscript R is F 12 BsubPc, a reference fluorophore, which has previously been reported to have a φ = 0.40. 3 Integrated fluorescence intensity values were acquired from PerkinElmer FL WinLab (version 4.00.03) while the optical density values were acquired from PerkinElmer UV WinLab (version 6.02.0723). All analyses were acquired in toluene and thus the calculation of the refractive index term of Eq S2.4 is irrelevant.

224

UV-Vis Absorption and Photoluminescence Plots

(a)

(b)

(c)

Figure S2.7 . UV-vis absorption (blue), photoluminescence (red), and photoexcitation (green) spectra of (a) 3,4-dimethylphenoxy-BsubPc, (b) polymer 11a and (c) polymer 11b .

225

References 1. Potz, R.; Goldner, M.; Huckstadt, H.; Cornelissen, U.; Tutass, A.; Homborg, H. Zeitschrift fuer Anorganische and Allgemeine Chemie. 2000 , 626 , 588-596. 2. Brandrup, J. I., E. H.; Edmund, H.; Grulke, E. A.; Abe, A.; Bloch, D. R. In Polymer Handbook, 4th Ed. ; John Wiley & Sons: New York, 2005, p 181-288. 3. Gonzalez-Rodriguez, D.; Torres, T.; Guldi, D. M.; Rivera, J.; Herranz, M. A.; Echegoyen, L. Journal of the American Chemical Society 2004 , 126 , 6301-6313.

226

Appendix B: Supplementary Information for Chapter 3

Synthetic Procedures for Starting BsubPc Materials: Bromo boron subphthalocyanine (Br-BsubPc) . Br-BsubPc was prepared according to literature procedure. 1 Hydroxy boron subphthalocyanine (HO-BsubPc) . HO-BsubPc was prepared according to literature procedure. 2

Bromo dodecafluoro boron subphthalocyanine (Br-F12 BsubPc) . Br-F12 BsubPc was prepared according to literature procedure. 3

Chloro hexachloro boron subphthalocyanine (Cl-Cl 6BsubPc) . Cl-Cl 6BsubPc was prepared according to literature procedure. 4

Chloro dodecachloro boron subphthalocyanine (Cl-Cl 12 BsubPc) . Cl-Cl 12 BsubPc was prepared according to literature procedure. 5

Attempted Synthetic Procedures for µ-Oxo-(BsubPc) 2: Methods 1.1 to 1.4 (Table S3.1) . For 1.1, conditions were maintained as reported by Geyer et al .7 For 1.2 -1.4 : Br-BsubPc was used in the place of Cl-BsubPc. In 1.3 , the volume of solvent was reduced by a factor of four. In 1.4 , the amount of NaOH was reduced to a molar amount equivalent to the moles of Br-BsubPc while keeping its molar equivalence relative to dicyclohexano-18-crown-6 the same.

6 Table S3.1. µ-Oxo-(BsubPc) 2 syntheses adapted from Geyer et al . Halo-BsubPc p-Xylene NaOH Dicyclohexano-18-Crown-6 Method Mass Moles Volume Mass Moles Halo Mass (g) Moles (mmol) (g) (mmol) (mL) (g) (mmol) 1.1 7 Cl 0.14 0.325 200 0.08 2 0.28 0.75 1.2 Br 0.154 0.325 200 0.08 2 0.28 0.75 1.3 Br 0.154 0.325 50 0.08 2 0.28 0.75 1.4 Br 0.154 0.325 50 0.013 0.325 0.045 0.122

227

Method 2.1 - Self-Condensation of HO-BsubPc Method (Table S3.2) . μ-Oxo-(BsubPc) 2 was synthesized by the self-condensation of HO-BsubPc as adapted from the method of Yamasaki & Mori. 7 1,2-Dichlorobenzene (1.4 mL) was added to make up a 10 wt% solution of HO-BsubPc (0.200 g, 0.485 mmol) in a 4 Dram (15 mL) vial fitted with a condenser. The reaction mixture was heated to reflux (180 °C) under an atmosphere of argon and the reaction progress was monitored by HPLC. After 120 h (5 days), full conversion was not achieved and the reaction was cooled to room temperature. The solvent was removed by rotary evaporation and the solid was dried in a vacuum oven overnight at 40°C. The crude product was purified by Kauffman column chromatography (alumina, DCM). The first band was collected and it contained μ-oxo-(BsubPc) 2 and two unknown impurities (HPLC R T: 2.0 min and 3.7 min) and the second band was a mixture of the same in addition to traces of other BsubPc impurities. The collected fractions were concentrated via rotary evaporation and dried in a vacuum oven overnight at 40°C to afford μ- oxo-(BsubPc) 2 (45 mg, 23%).

Method 2.2 - '1 Pot' Reaction Method (Table S3.2) . μ-Oxo-(BsubPc) 2 was synthesized as reported in Fulford et al .8 by the in situ hydrolysis of Br-BsubPc (0.200 g, 0.421 mmol) in the presence of water.

Table S3.2. µ-Oxo-(BsubPc) 2 syntheses via the self-condensation of HO-BsubPc method. Method BsubPc Mass Moles Volume of 1,2- Weight Temperature Reaction Starting (g) (mmol) Dichlorobenzene Percent (°C) Time (h) Material (mL) (%) 2.1 8 HO-BsubPc 0.20 0.485 1.4 10 180 120 2.2 9 Br-BsubPc 0.20 0.421 1.4 10 180 24

Methods 3.1 to 3.6 - Equimolar Method (Table S3.3) . μ-Oxo-(BsubPc) 2 was synthesized by reacting equimolar quantities of HO-BsubPc and Br-BsubPc, a reaction pathway which is included in the patent application of Mori et al .9 For 3.1 to 3.5 , solvent was added to HO-BsubPc (0.100 g, 0.243 mmol) and Br-BsubPc (0.115 g, 0.243 mmol) in a scintillation (20 mL) or 4 Dram (15 mL) vial fitted with a condenser, to make up a ~2 wt% or ~10 wt% solution, respectively. Each reaction was heated to a temperature (as indicated in Table S3.3) under an

228

atmosphere of argon and the reaction progress was monitored by HPLC. After a certain reaction time (as indicated in Table S3.3), the concentration of μ-oxo-(BsubPc) 2 had reach steady state as determined by HPLC. The reaction was cooled to room temperature. For 3.1 to 3.3 , no further workup was done. The subsequent steps for 3.5 and 3.6 (5x scale of 3.5 ) are described in (3). For

3.4 , solvent was added to the reaction mixture in an attempt to precipitate out μ-oxo-(BsubPc) 2. A significant quantity of cyclohexane (200 mL) was added, however no precipitate was observed until after the solution was left covered in an ice bath overnight. The recovered solids were dried in a vacuum oven overnight at 40°C (123 mg; HPLC (545nm channel): 82% μ-oxo-(BsubPc) 2 by area). In a replicate of 3.4 , methanol was added to the reaction mixture and no precipitate was observed.

Table S3.3 . µ-Oxo-(BsubPc) 2 syntheses via the equimolar method. Method Solvent Volume Weight Percent (wt %) Temperature (°C) Reaction Time (h) (mL) 3.1 p-xylene 10 2.4 105 139.5 3.2 nitrobenzene 10 1.8 200 31 3.3 diphenyl ether 10 2.0 200 31 3.4 diphenyl ether 1.8 10 200 15.5 3.5 1,2-dichlorobenzene 1.5 10 181 20 3.6 1,2-dichlorobenzene 7.5 10 181 20

Methods 4.1 - 4.3, 4.5 - Equimolar Tripotassium Phosphate Method (Table S3.4) . µ-Oxo-

(BsubPc) 2 was synthesized based on 3.5/3.6 , but with the addition of tripotassium phosphate

(K 3PO 4). 1,2-Dichlorobenzene was added to make up a 10 wt% solution of equimolar HO- BsubPc and Br-BsubPc in a 4 Dram (15 mL) vial fitted with a condenser. In 4.3 , glassware were dried in an oven at 100 °C overnight and the solvent was filtered through alumina before use. A mass of K 3PO 4 equivalent to 20% or 80% of the solvent’s mass, ground in a mortar and pestle and activated in an oven at 200 °C was added. The reaction mixture was heated to reflux under an atmosphere of argon and the reaction progress was monitored by HPLC. Samples were filtered through glass wool prior to analysis via HPLC to remove suspended solids. After a certain reaction time (as indicated in Table S3.4), the reaction was cooled to room temperature. Various workup steps followed.

229

For 4.1 , the solvent was removed by rotary evaporation and the solid was dried in a vacuum oven overnight at 40 °C. The solids were transferred to a Soxhlet thimble and a Soxhlet extraction using toluene was carried out. The solids remaining in the thimble had a deep turquoise colour and were insoluble in a range of organic solvents. The solvent was removed from the filtrate by rotary evaporation and the solid was then dried in a vacuum oven overnight at 40 °C (60 mg,

31%). Pure μ-oxo-(BsubPc) 2 was not isolated. For 4.2 , the same sequence was followed as 4.1 with a Soxhlet extraction using toluene (130 mg, 66%). The solids were train sublimed at 220 °C to remove small molecule impurities (70 mg, 36%). The solids remaining in the crucible were subsequently train sublimed at 450 °C following the same procedure as 3.5 , and μ-oxo-(BsubPc) 2 was collected as a loosely, partially crystalline solid (16 mg, 8%). For 4.3 , the reaction products were not isolated. For 4.5 , the reaction mixture was transferred directly into a Soxhlet thimble and a Soxhlet extraction using toluene was carried out. Alumina (40 g) was added to the toluene retentate (~100 mL) and stirred. The liquid was decanted into a filter, and more toluene was added to the flask containing alumina and stirred. This was repeated several times and the remaining alumina and liquid was poured into the filter and rinsed with toluene. The toluene rinse runoff remained deeply magenta coloured. The solvent was removed from the filtrate by rotary evaporation and the solid was dried in a vacuum oven overnight at 40 °C (42 mg, 21%). HPLC R T: 2.8 min (>99.9%, 545 nm channel). A Kauffman column was run on the retained and dried alumina using DCM as the eluent. The solvent was removed from the filtrate by rotary evaporation and the solid was dried in a vacuum oven overnight at 40 °C (14 mg, 7%). HPLC R T: 2.8 min (>99.9%,

545 nm channel). (Total Yield: 56 mg, 29%). μ-Oxo-(BsubPc) 2 could not separated from the non-BsubPc impurities.

Methods 4.4, 4.6 - 4.8 – ‘1 pot’ Tripotassium Phosphate Method (Table S3.4) . µ-Oxo-

(BsubPc) 2 was synthesized based on 2.2 , but with the addition of K 3PO 4. For 4.4 and 4.6 , K 3PO 4 was ground in a mortar and pestle and activated in an oven at 200 °C. It was then left exposed to the atmosphere to absorb water overnight or until its mass had increased by approximately 15%,

230

at which point the dry mass equivalent of wet K 3PO 4 corresponding to 20% of the solvent’s mass was added to the reaction vial (wet mass calculated using the measured wt % of water). For 4.7 and 4.8 , a mass of dry K 3PO 4 equivalent to 20% of the solvent’s mass, ground in a mortar and pestle and activated in an oven at 200 °C, was added to the reaction vial. This was then left open to absorb water until its mass had increased by approximately 15%. 1,2-Dichlorobenzene was added to make up a 10 wt % solution of Br-BsubPc in the 4 Dram (15 mL) ( 4.4, 4.6 ) or scintillation (20 mL) ( 4.7 - 4.8 ) vial fitted with a condenser. The reaction was heated to reflux under an atmosphere of argon and the reaction progress was monitored by HPLC. Samples were filtered through glass wool prior to analysis via HPLC to remove suspended solids. After a certain reaction time (as indicated in Table S3.4), the reaction was cooled to room temperature. Various workup steps followed. For 4.4 , 4.6-4.8 the reaction products were not isolated.

Method 4.9 – Final Method (Table S3.4) . µ-Oxo-(BsubPc) 2 was synthesized based on an adaptation of 4.5. 1,2-Dichlorobenzene (6.9 mL) was added to make up a 10 wt% solution of HO-BsubPc (0.464 g, 1.13 mmol) and Br-BsubPc (0.535 g, 1.13 mmol) in a scintillation (20 mL) vial fitted with a condenser. A mass of K 3PO 4 equivalent to 20% of the solvent’s mass (1.801 g, 8.49 mmol), ground in a mortar and pestle and activated in an oven at 200 °C was added. The reaction was heated to reflux under an atmosphere of argon and the reaction progress was monitored by HPLC. After 1 hour, the reaction was cooled and immediately transferred to a Soxhlet thimble. A Soxhlet extraction using toluene was carried out for 2 days. The solvents were removed by rotary evaporation and the solids were fully redissolved in DCM. A Kauffman column chromatography (alumina, DCM) was carried out. The column was left to dry overnight, after which a fresh flask of DCM was placed under the column and reflux was established. The Kauffman column was ran for another day. The solvent was removed by rotary evaporation and the solids were dried in a vacuum oven overnight at 40 °C (537 mg, 59% overall mass yield,

HPLC R T: 2.8 min (>99.9%, at 545 nm)). The solids were train sublimed at a temperature of 450 °C (held for 5 hrs 45min, total heating time 10 hrs 15min), and single crystals as well as a band of μ-oxo-(BsubPc) 2 were collected (357 mg, overall 39% mass yield; sublimation step yield

231

70%). Removed white crystals were found to be phthalimide (~10 mg) and bands containing small quantities of HO-BsubPc were also discarded (mass below collection limit). The μ-oxo-

(BsubPc) 2 was train sublimed a second time under the same conditions. Single crystals and μ- oxo-(BsubPc) 2 band were collected (sublimation step yield 58%). Overall yield: 0.204 g, 22.5%.

HPLC R T: 2.8 min (>99.9%, MAXPlot).

Table S3.4. µ-Oxo-(BsubPc) 2 syntheses via the tripotassium phosphate method.

Method Volume of 1,2- HO-BsubPc Br-BsubPc Mass of K 3PO 4 Reaction Dichlorobenzene Mass Moles Mass Moles % of Dry Wet Time (h) (mL) (g) (mmol) (g) (mmol) Solvent Mass Mass Mass (g) (g) 4.1 1.5 0.100 0.243 0.115 0.243 80 1.553 - 24 4.2 1.5 0.100 0.243 0.115 0.243 20 0.388 - 24 4.3 1.5 0.100 0.243 0.115 0.243 20 0.388 - 24 4.4 1.6 0 0 0.231 0.485 20 0.388 0.437 24 4.5 1.5 0.100 0.243 0.115 0.243 20 0.388 - 1 4.6 1.6 0 0 0.231 0.485 20 0.388 0.464 1 4.7 4.0 0 0 0.576 1.213 20 1.039 1.193 1.25 4.8 4.0 0 0 0.576 1.213 20 1.039 1.236 1 4.9 6.9 0.464 1.13 0.535 1.13 20 1.801 - 1

Synthetic Procedures for Asymmetric µ-Oxo-(BsubPc) 2 Compounds:

F12 BsubPc-O-BsubPc . To a single neck round bottom flask equipped with a condenser was added 1,2-dichlorobenzene (14 mL, ~10 w/w % relative to HO-BsubPc and Br-F12 BsubPc), HO-

BsubPc (0.75 g, 1.82 mmol, 1 equiv), and Br-F12 BsubPc (1.26 g, 1.82 mmol, 1 equiv) under an argon atmosphere. To this reaction mixture was added tripotassium phosphate (3.66 g, ~20% mass equivalent relative to 1,2-DCB), which was pre-ground and activated in an oven at 200 ºC for one hour prior to use. The reaction mixture was heated to 180 ºC and the reaction progress was monitored by HPLC. Once the reaction was complete (~19 h), the reaction was cooled to room temperature before it was concentrated to dryness via rotary evaporation. The crude product was purified by silica gel column chromatography (100% DCM to 95:5 DCM/ethyl acetate) to afford F 12 BsubPc-O-BsubPc (670 mg, 36%) as a dark purple-gold solid. The compound was further purified using train sublimation. The apparatus was operated under a vacuum with a controlled flow of carbon dioxide gas generating an internal pressure of 100 mTorr. The temperature was increased from room temperature up to ~440 °C and was held

232

constant at that temperature overnight (~12 h). A metallic gold band was confirmed to be the title compound (71 mg, 41% yield relative to mass placed in the train sublimation apparatus). Vapour diffusion of heptane into DCM solution produced single crystals suitable for X-ray diffraction

1 13 analysis. H NMR (400 MHz, CDCl 3, Me 4Si) d 8.70-8.64 (m, 6H), 7.87-7.81 (m, 6H); C NMR

11 (100 MHz, CDCl 3) d 150.6, 147.4, 130.7, 129.8, 122.0; B NMR (400 MHz, CDCl 3, BF 3•OEt 2)

19 d -18.3 (s); F NMR (400 MHz, CDCl 3, BF 3•OEt 2) d 15.67-14.93 (m, 6F), 4.28-3.60 (m, 6F); + HRMS (ESI): m/z [M+H] calcd for 1021.1397, found 1021.1368; UV-vis toluene λmax (ε): 547 nm (89800 M -1cm -1).

Cl 6BsubPc-O-BsubPc. To a single neck round bottom flask equipped with a condenser was added 1,2-dichlorobenzene (13 mL, ~10 w/w % relative to HO-BsubPc and Cl-Cl 6BsubPc), HO-

BsubPc (0.75 g, 1.82 mmol, 1 equiv), and Cl-Cl 6BsubPc (1.16 g, 1.82 mmol, 1 equiv) under an argon atmosphere. To this reaction mixture was added tripotassium phosphate (3.40 g, ~20% mass equivalent relative to 1,2-DCB), which was pre-ground and activated in an oven at 200 ºC for one hour prior to use. The reaction mixture was heated to 180 ºC and the reaction progress was monitored by HPLC. Once the reaction was complete (~17 h), the reaction was cooled to room temperature before it was concentrated to dryness via rotary evaporation. The crude product was purified by silica gel column chromatography (100% DCM to 95:5 DCM/ethyl acetate) to afford Cl 6BsubPc-O-BsubPc (737 mg, 40%) as a gold solid. The compound was further purified using train sublimation. The apparatus was operated under a vacuum with a controlled flow of carbon dioxide gas generating an internal pressure of 90 mTorr. The temperature was increased from room temperature up to ~440 °C and was held constant at that temperature overnight (~12 h). A metallic gold band was confirmed to be the title compound (50 mg, 35% yield relative to mass placed in the train sublimation apparatus). Slow evaporation from DCM solution produced single crystals suitable for X-ray diffraction analysis. 1H NMR (400 13 MHz, CDCl 3, Me 4Si) d 8.63-8.59 (m, 12H), 7.83-7.77 (m, 6H); C NMR (100 MHz, CDCl 3) d 11 150.5, 149.3, 134.4, 130.6, 129.53, 129.5, 123.5, 121.9; B NMR (400 MHz, CDCl 3, BF 3•OEt 2)

+ d -18.2 (s); HRMS (ESI): m/z [M+H] calcd for 1009.0189, found 1009.0166; UV-vis toluene λmax (ε): 541 nm (89800 M -1cm -1).

233

Cl 12 BsubPc-O-BsubPc. To a single neck round bottom flask equipped with a condenser was added 1,2-dichlorobenzene (10 mL, ~10 w/w % relative to HO-BsubPc and Cl-Cl 12 BsubPc), HO-

BsubPc (0.50 g, 1.21 mmol, 1 equiv), and Cl-Cl 12 BsubPc (1.02 g, 1.21 mmol, 1 equiv) under an argon atmosphere. To this reaction mixture was added tripotassium phosphate (2.60 g, ~20% mass equivalent relative to 1,2-DCB), which was pre-ground and activated in an oven at 200 ºC for one hour prior to use. The reaction mixture was heated to 180 ºC and the reaction progress was monitored by HPLC. Once the reaction was complete (~18 h), the reaction was cooled to room temperature before it was concentrated to dryness via rotary evaporation. The crude product was purified by silica gel column chromatography (100% DCM) to afford Cl 12 BsubPc- O-BsubPc (630 mg, 43%) as a dark purple solid. The compound was further purified using train sublimation. The apparatus was operated under a vacuum with a controlled flow of carbon dioxide gas generating an internal pressure of 90 mTorr. The temperature was increased from room temperature up to ~480 °C and was held constant at that temperature overnight (~12 h). A metallic gold band was confirmed to be the title compound (73 mg, 55% yield relative to mass 1 placed in the train sublimation apparatus). H NMR (400 MHz, CDCl 3, Me 4Si) d 8.69-8.64 (m,

13 6H), 7.86-7.80 (m, 6H); C NMR (100 MHz, CDCl 3) d 150.7, 148.1, 135.0, 130.7, 129.7, 128.5,

11 + 127.3, 122.1; B NMR (400 MHz, CDCl 3, BF 3•OEt 2) d -18.7 (s); MS (ESI): m/z [M+H] calcd -1 -1 for 1219.8, found 1219.8; UV-vis toluene λmax (ε): 558 nm (89600 M cm ).

234

NMR Spectra

1 Figure S3.1 . H NMR (400 MHz, CDCl 3) spectrum obtained at 296 K for F 12 BsubPc-O-BsubPc.

235

13 Figure S3.2 . C NMR (100 MHz, CDCl 3) spectrum obtained at 296 K for F 12 BsubPc-O- BsubPc.

236

11 Figure S3.3 . B NMR (400 MHz, CDCl 3, referenced to BF 3•OEt 2) spectrum obtained at 296 K for F 12 BsubPc-O-BsubPc.

237

19 Figure S3.4 . F NMR (400 MHz, CDCl 3, referenced to BF 3•OEt 2) spectrum obtained at 296 K for F 12 BsubPc-O-BsubPc.

238

1 Figure S3.5 . H NMR (400 MHz, CDCl 3) spectrum obtained at 296 K for Cl 6BsubPc-O-BsubPc.

239

13 Figure S3.6 . C NMR (100 MHz, CDCl 3) spectrum obtained at 296 K for Cl 6BsubPc-O- BsubPc.

240

11 Figure S3.7 . B NMR (400 MHz, CDCl 3, referenced to BF 3•OEt 2) spectrum obtained at 296 K for Cl 6BsubPc-O-BsubPc.

241

1 Figure S3.8 . H NMR (400 MHz, CDCl 3) spectrum obtained at 296 K for Cl 12 BsubPc-O- BsubPc.

242

13 Figure S3.9 . C NMR (100 MHz, CDCl 3) spectrum obtained at 296 K for Cl 12 BsubPc-O- BsubPc.

243

11 Figure S3.10 . B NMR (400 MHz, CDCl 3, referenced to BF 3•OEt 2) spectrum obtained at 296 K for Cl 12 BsubPc-O-BsubPc.

244

Wavelength of Maximum Absorption for Some Monomeric BsubPcs

Table S3.5 . λmax values for some monomeric BsubPc compounds.

BsubPc Compound λmax (nm) Br-BsubPc 566 Cl-BsubPc 565 F-BsubPc 562 PhO-BsubPc 563 HO-BsubPc 561 Br-F12 BsubPc 578 Cl-Cl 6BsubPc 574 Cl-Cl 12 BsubPc 593 All measurements were done in toluene solutions at room temperature.

245

UV-vis Absorption and Photoluminescence Plots

(a)

(b) Figure S3.11 . Overlay of μ-oxo BsubPc’s (red) and the halo-BsubPc’s (blue; F- solid, Cl- dotted, Br- dashed) (a) normalized absorption spectra; and (b) photoluminescence (PL) spectra. 9

246

(a)

(b)

Figure S3.12 . Overlay of F 12 BsubPc-O-BsubPc's (solid red), Cl 6BsubPc-O-BsubPc's (dotted red), and Cl 12 BsubPc-O-BsubPc's (dashed red) and the halo-BsubPc’s (blue; F- solid, Cl- dotted, Br- dashed) (a) normalized absorption spectra; and (b) normalized photoluminescence (PL) spectra.

247

Determination of Fluorescence Quantum Yields The fluorescence quantum yields (φ) were calculated using the formula below:

2 2 φ = φ R (I / I R)(OD R / OD)(n / n R ) (Eq S3.1) where I is the integrated fluorescence intensity, OD is the optical density ( i.e. absorbance), and n is the refractive index of the solvent. The subscript R is PhO-F12 BsubPc, a reference fluorophore, which has previously been reported to have a φ = 0.40. 10 Integrated fluorescence intensity values were acquired from PerkinElmer FL WinLab (version 4.00.03) while the optical density values were acquired from PerkinElmer UV WinLab (version 6.02.0723).

248

References 1. Dang, J. D.; Virdo, J. D.; Lessard, B. H.; Bultz, E.; Paton, A. S.; Bender, T. P. Macromolecules 2012 , 45 , 7791-7798. 2. Fulford, M. V.; Lough, A. J.; Bender, T. P. Acta Crystallographica Section B-Structural Science 2012 , 68 , 636-645. 3. Morse, G. E.; Helander, M. G.; Maka, J. F.; Lu, Z.-H.; Bender, T. P. ACS Applied Materials & Interfaces 2010 , 2, 1934-1944. 4. Sullivan, P.; Duraud, A.; Hancox, I.; Beaumont, N.; Mirri, G.; Tucker, J. H. R.; Hatton, R. A.; Shipman, M.; Jones, T. S. Advanced Energy Materials 2011 , 1, 352-355. 5. Morse, G. E.; Bender, T. P. Inorganic Chemistry 2012 , 51 , 6460-6467. 6. Geyer, M.; Plenzig, F.; Rauschnabel, J.; Hanack, M.; Del Rey, B.; Sastre, A.; Torres, T. Synthesis 1996 , 1139-1151. 7. Yamasaki, Y.; Mori, T. Bulletin of the Chemical Society of Japan 2011 , 84 , 1208-1214. 8. Fulford, M. V.; Jaidka, D.; Paton, A. S.; Morse, G. E.; Brisson, E. R. L.; Lough, A. J.; Bender, T. P. Journal of Chemical & Engineering Data 2012 , 57 , 2756-2765. 9. Mori, T.; Furuya, F.; Yamasaki, Y. Optical layer including mu-oxo-bridged boron- subphthalocyanine dimer. U.S. Patent Application US 2008/0210128, September 4, 2008. 10. Gonzalez-Rodriguez, D.; Torres, T.; Guldi, D. M.; Rivera, J.; Herranz, M. A.; Echegoyen, L. Journal of the American Chemical Society 2004 , 126 , 6301-6313.

249

Appendix C: Supplementary Information for Chapter 4

Synthetic Procedures Br-BsubPc . Br-BsubPc was prepared according to literature procedure. 1

Train Sublimation Apparatus

BOAT SUBSTRATE GLASS INSERT

OCVD REACTOR WITHIN OCVD REACTOR CONDENSER PRESSURE INDICATOR

PI BAND COPPER-NICKEL PIPE NEEDLE HEATER VALVE CERAMIC VACUUM INSULATOR TIC PUMP

BALL TEMPERATURE VALVE INDICATING CONTROLLER

GAS OUTLET GAS INLET Figure S4.1 . Schematic of the train sublimation system.

250

Table S4.1 . OCVD Experimental Results. Method Mass of Thermal Profile Pressure a (mTorr) | BsubPc Composition (%) b Film h Br-BsubPc T Ramp Soak Temperature (°C) Compound BT cg I1 dg SB eg I2 fg Thickness (mg) (˚C) Time (hr) Time (hr) of N 2 gas 1.1 25 150 1 1 100 | rt HO-BsubPc 76 31 2 10 ~0.9-1.7 μm Br-BsubPc 5 28 98 83 450 1 8 μ-oxo-(BsubPc) 2 9 41 <1 1 Other BsubPc 10 0 0 6 1.2 25 150 1 1 100 | 170-180 HO-BsubPc 78 23 1 48 - Br-BsubPc 5 39 98 16 450 1 8 μ-oxo-(BsubPc) 2 8 38 1 36 Other BsubPc 9 0 0 0 1.3 25 150 1 1 100 | rt HO-BsubPc 99 30 27 23 ~3.1-3.6 μm Br-BsubPc 0 12 55 45 500 1 8 μ-oxo-(BsubPc) 2 1 56 18 32 Other BsubPc 0 2 0 0 1.4 10 150 1 1 100 | rt HO-BsubPc 100 48 53 54 ~1.4-2.0 μm Br-BsubPc 0 2 8 8 500 1 8 μ-oxo-(BsubPc) 2 0 50 39 38 Other BsubPc 0 0 0 0 1.5 5 150 1 1 100 | rt HO-BsubPc 100 27 27 26 ~240-550 nm Br-BsubPc 0 2 41 23 500 1 8 μ-oxo-(BsubPc) 2 0 69 32 51 Other BsubPc 0 2 0 0 1.6 5 150 1 1 100 | rt HO-BsubPc 100 53 41 38 - Br-BsubPc 0 0 7 16 550 1 8 μ-oxo-(BsubPc) 2 0 47 52 46 Other BsubPc 0 0 0 0 1.7 5 150 1 1 100 | rt HO-BsubPc 0 57 24 59 ~127-220 nm Br-BsubPc 0 0 62 0 600 1 8 μ-oxo-(BsubPc) 2 0 43 6 47 Other BsubPc 0 0 8 0 1.8 i 5 150 1 1 100 | rt HO-BsubPc 100 56 25 43 - Br-BsubPc 0 2 2 2

251

550 1 8 μ-oxo-(BsubPc) 2 0 41 73 55 Other BsubPc 0 0 0 0 1.9 i 5 150 1 1 100 l | rt HO-BsubPc 100 46 28 38 ~150-190 nm Br-BsubPc 0 0 1 2 550 1 8 μ-oxo-(BsubPc) 2 0 54 71 60 Other BsubPc 0 0 0 0 1.10 i 5 150 1 1 100 m | rt HO-BsubPc 100 30 26 30 - Br-BsubPc 0 0 0 4 550 1 8 μ-oxo-(BsubPc) 2 0 70 74 66 Other BsubPc 0 0 0 0 1.11 j 5 150 1 1 100 | rt HO-BsubPc 100 54 34 40 ~100-200 nm Br-BsubPc 0 0 7 1 550 1 8 μ-oxo-(BsubPc) 2 0 46 59 59 Other BsubPc 0 0 0 0 1.12 jk 5 150 1 1 100 | rt HO-BsubPc 100 77 48 45 - Br-BsubPc 0 0 0 2 550 1 8 μ-oxo-(BsubPc) 2 0 23 52 53 Other BsubPc 0 0 0 0 1.13 i 5 (HO- 150 1 1 100 | rt HO-BsubPc 94 88 58 35 - BsubPc) Br-BsubPc 0 0 0 0 550 1 8 μ-oxo-(BsubPc) 2 0 5 28 64 Other BsubPc 6 7 14 1 a Added on top of the pump down pressure. b As determined by HPLC-UV/Vis (Extracted at 560 nm). c BT = boat. d I1 = right edge of insert 1. e SB = substrate. f I2 = left edge of insert #2. g See Figure S5.2 for position of BT, I1, SB, and I2. h Measured via surface profilometry. i Pre-cleaned glass substrate. j Pre-cleaned

k l m ITO glass. Layer of PEDOT:PSS. N2 (<5 ppm H 2O) gas. CO 2 (<5 ppm H 2O) gas.

252

INSERT 1 (I1) INSERT 2 (I2)

UPSTREAM END DOWNSTREAM END

BOAT (BT) SUBSTRATE (SB) Figure S4.2 . Side view profile of the OCVD reactor showing the orientation of the boat (BT), insert 1 (I1), substrate (SB), and insert 2 (I2).

Kapton tape

Sample site #3 Sample site #1

Sample site #2

Figure S4.3 . Cartoon image (front view) of an OCVD-derived film (Method 1.6) showing the three sample sites used for LRMS analysis.

Table S4.2 . LRMS (TOF MS EI+) results for an OCVD-derived film (Method 1.6). BsubPc Compound % Composition at Site #1 a % Composition at Site #2 a % Composition at Site #3 a μ-oxo-(BsubPc) 2 53 47 69 HO-BsubPc 20 23 14 +BsubPc 27 30 17 a See Figure S4.3 for position on substrate.

253

References 1. Dang, J. D.; Virdo, J. D.; Lessard, B. H.; Bultz, E.; Paton, A. S.; Bender, T. P. Macromolecules 2012 , 45 , 7791-7798.

254

Appendix D: Supplementary Information for Chapter 5

Table S5.1 . Summary of reported BsubNc synthesis. Entry Starting Material Boron Solvent Temperature (°C) Yield (%) Reference Source 1 PhBCl 2 NAP 218 Anal. 1,2

2 BBr 3 2,3-DM-6- 180 34.6 3 tBuNAP

3 BCl 3 1,2-DCB 180 53 4

4 BCl 3 ClBZH: 130 35 5 TOL (1:1)

5 BCl 3 1-ClNAP microwave 82 6

6 BCl 3 p-XYL: 150-180 Cl-F6BsubNc: 26 7 1,2-DCB Cl-F12 BsubNc: 44 (1:1)

7 BCl 3 p-XYL: Cl-Cl 6BsubNc: 160 Cl-Cl 6BsubNc: 22 8 1,2-DCB Cl-I6BsubNc: 180 Cl-I6BsubNc: 73 (1:1)

*NAP = naphthalene; 2,3-DM-6-tBuNAP = 2,3-dimethyl-6-tert -butylnaphthalene; 1,2-DCB = 1,2-dichlorobenzene; ClBZH = chlorobenzene; TOL = toluene; 1-ClNAP = 1-chloronaphthalene; p-XYL = para -xylene.

255

Synthesis of Starting Material 2,3-Dicyanonaphthalene (DCNAP). DCNAP was prepared according to literature procedure. 9

Synthesis of Cl-BsubNc (Torres Method): Method 1.1 to 1.3 (Table S5.2 ): For 1.1 , to a three-neck round bottom flask equipped with a gas inlet and a short-path condenser was added DCNAP (0.250 g, 1.40 mmol, 1 equiv), chlorobenzene (2.5 mL), and toluene (2.5 ml) under an atmosphere of argon gas. The mixture was stirred for 10 minutes at room temperature before BCl 3 (1.0 M in heptane, 1.40 mL, 1 equiv) was added. The reaction mixture was heated to 130 °C for two hours before it was cooled to room temperature. The solvent mixture was removed via rotary evaporation and the resulting solid was extracted with hexane in a Soxhlet extraction apparatus for 24 hours. Column chromatography (Si gel) was then performed on the washed product using toluene as the eluent and blue fractions were collected, concentrated, and dried in a vacuum oven (60 °C). In 1.2 , 2.5 molar equivalence of BCl 3 was used. In 1.3 , the reaction of 1.2 was repeated but para -xylene was substituted for toluene.

Table S5.2 . Attempted syntheses adapted from Torres et al .5 Method Boron Template/Source Solvent System Temperature % Relative Number of (°C) Conversion BsubNc via HPLC Products 1.1 BCl 3 ClBZH:TOL 130 35 5 (1.0 M heptane, 1.0 eq) (1:1, 0.28 M) 1.2 BCl 3 ClBZH:TOL 130 24 5 (1.0 M heptane, 2.5 eq) (1:1, 0.28 M) 1.3 BCl 3 ClBZH:XYL 130 46 5 (1.0 M heptane, 2.5 eq) (1:1, 0.28 M) *ClBZH = chlorobenzene; TOL = toluene; XYL = para -xylene.

256

HPLC-UV/Vis Chromatogram of Cl-BsubNc (Torres Method)

Figure S5.1 . HPLC chromatogram (extracted at 650 nm) and the respective UV-vis absorption spectrum of each peak for Cl-BsubNc made via an adaptation of the Torres method (Method 1.1 ). Note that the peak at ~1.6 min is from the N,N - dimethylformamide solvent.

257

Synthesis of Cl-BsubNc (Kennedy Method): Method 2.1 to 2.5 (Table S5.3 ): For 2.1 , to a three-neck round bottom flask equipped with a gas inlet and a short-path condenser was added DCNAP (0.250 g, 1.40 mmol, 1 equiv) and dried 1,2- dichlorobenzene (10 mL) under an atmosphere of argon gas. The mixture was stirred for 10 minutes at room temperature before BCl 3 (1.0 M in heptane, 0.93 mL, 0.66 equiv) was added. The reaction mixture was heated to 150 °C for one hour to distill off the heptane before it was heated to 180 °C for two hours. The reaction was cooled to room temperature and concentrated via rotary evaporation. Attempts to purify the crude product via column chromatography (neutral

Al 2O3) were performed using toluene as the eluent. In 2.2 , ACS grade 1,2-dichlorobenzene was used. In 2.3 and 2.4 , 1.1 and 3.3 molar equivalence of BCl 3 was used, respectively. In 2.5 , a solvent mixture of 1,2-dichlorobenzene and 1-methylnaphthalene (4:1 v/v ) was used. For 2.3 , the reaction was scaled up by a factor of 7.6. The crude product (0.730 g) was purified by train sublimation. The apparatus was operated under a vacuum with a controlled flow of carbon dioxide gas generating an internal pressure of 100 mTorr. The temperature was increased from room temperature to ~580 °C (external temperature) and was held constant at that temperature overnight. A metallic purple band was collected (0.150 g, 21% yield relative to mass placed in the train sublimation apparatus). The sublimed product was subsequently purified a second time by train sublimation under the same condition (0.115 g, 76% yield relative to mass placed in the train sublimation apparatus), producing single crystals suitable for X-ray diffraction. 1H NMR

(400 MHz, CDCl 3) d 9.43 (s, 6H), 8.37-8.33 (m, 6H), 7.77-7.73 (m, 6H); MS (EI): m/z 580.1, -1 -1 614.1, 648.1, 684.0; UV-vis toluene λmax (ε): 656 nm (78400 M cm ).

258

Table S5.3 . Attempted syntheses adapted from Zyskowski and Kennedy. 4 Method Boron Template/Source Solvent System Temperature % Relative Number of (°C) Conversion BsubNc via HPLC Products 2.1 BCl 3 1,2-DCB 180 88 5 (1.0 M heptane, 0.66 eq) (dried, 0.14 M) 2.2 BCl 3 1,2-DCB 180 84 5 (1.0 M heptane, 0.66 eq) (0.14 M) 2.3 BCl 3 1,2-DCB 180 81 5 (1.0 M heptane, 1.1 eq) (0.14 M) 2.4 BCl 3 1,2-DCB 180 87 5 (1.0 M heptane, 3.3 eq) (0.14 M) 2.5 BCl 3 1,2-DCB:1-MNAP 180 49 2 (1.0 M heptane, 3.3 eq) (4:1, 0.14 M) *1,2-DCB = 1,2-dichlorobenzene; 1-MNAP = 1-methylnaphthalene.

259

HPLC-UV/Vis Chromatogram of Cl-BsubNc (Kennedy Method)

Figure S5.2 . HPLC chromatogram (extracted at 650 nm) and the respective UV-vis absorption spectrum of each peak for Cl-BsubNc made via an adaptation of the Kennedy method (Method 2.3 ). Note that the peak at ~1.6 min is from the N,N - dimethylformamide solvent.

260

Synthesis of Cl-BsubNc (1-MNAP Method): Method 3.1 to 3.6 (Table S5.4 ): For 3.1 , to a three-neck round bottom flask equipped with a gas inlet and a short-path condenser was added DCNAP (0.500 g, 2.81 mmol, 1 equiv), 1,2,4- trichlorobenzene (16 mL) and 1-methylnaphthalene (4 mL) under an atmosphere of argon gas.

The mixture was stirred for 10 minutes at room temperature before BCl 3 (1.0 M in heptane, 1.90 mL, 0.66 equiv) was added. The reaction mixture was heated to 130 °C for two hours before it was hot filtered under vacuum. The solid black mass was washed with hot toluene (~100 °C) until the filtrate was no longer blue in colour. The filtrate solution was concentrated to remove the toluene via rotary evaporation before it was precipitated into ice cold stirring hexane (200 mL), gravity filtered, washed with hexane (3 x 50 mL), and dried in a vacuum oven (60 °C). Attempts to purify the crude product via Soxhlet extraction and column chromatography on both

Si gel and neutral Al 2O3 were made. For 3.2 to 3.5 , the molar equivalence of BCl 3 was increased to 1.1, 2.5, 4.2, and 4.9, respectively. For 3.6 , the reaction was scaled up by a factor of 2.2. The crude product (0.376 g) was purified by train sublimation. The apparatus was operated under a vacuum with a controlled flow of carbon dioxide gas generating an internal pressure of 100 mTorr. The temperature was increased from room temperature to ~550 °C (external temperature) and was held constant at that temperature overnight. A metallic purple band was collected (14 mg, 3.7% yield relative to mass placed in the train sublimation apparatus). 1H NMR (400 MHz,

CDCl 3) d 9.43 (s, 6H), 8.37-8.33 (m, 6H), 7.77-7.73 (m, 6H); MS (EI): m/z 580.1, 614.1. Table S5.4 . Attempted syntheses adapted from Noll et al .10 Method Boron Template/Source Solvent System Temperature % Relative Number of (°C) Conversion BsubNc via HPLC Products 3.1 BCl 3 1,2,4-TCB:1-MNAP 130 70 2 (1.0 M heptane, 0.66 eq) (4:1, 0.14 M) 3.2 BCl 3 1,2,4-TCB:1-MNAP 130 68 2 (1.0 M heptane, 1.1 eq) (4:1, 0.14 M) 3.3 BCl 3 1,2,4-TCB:1-MNAP 130 68 2 (1.0 M heptane, 2.5 eq) (4:1, 0.14 M) 3.4 BCl 3 1,2,4-TCB:1-MNAP 130 26 2 (1.0 M heptane, 4.2 eq) (4:1, 0.14 M) 3.5 BCl 3 1,2,4-TCB:1-MNAP 130 22 2 (1.0 M heptane, 4.9 eq) (4:1, 0.14 M) 3.6 BCl 3 1,2,4-TCB:1-MNAP 130 58 2 (1.0 M heptane, 1.1 eq) (4:1, 0.14 M) *1,2,4-TCB = 1,2,4-trichlorobenzene; 1-MNAP = 1-methylnaphthalene.

261

HPLC-UV/Vis Chromatogram and Mass Spectrum of Cl-BsubNc (1-MNAP Method)

Figure S5.3a. HPLC chromatogram (extracted at 650 nm) and the respective UV-vis absorption spectrum of each peak for Cl-BsubNc made via the 1-MNAP method (Method 3.6 ). Note that the peak at ~1.6 min is from the N,N -dimethylformamide solvent.

262

Figure S5.3b. Mass spectrum of sublimed Cl-BsubNc sample made via the 1-MNAP method

(Method 3.6 ) indicating the additional presence of Cl-Cl 1BsubPc and absence of Cl-Cl 2BsubNc and Cl-Cl 3BsubNc.

263

Synthesis of Cl-BsubNc (Scope of Reactions Method): Method 4.1 to 4.15 (Table S5.5 ): All reactions were performed in a similar manner as those in Method 1-3.

Table S5.5 . Scope of reactions varying molar equivalence of BCl 3, solvent mixture composition, and concentration of 2,3-dicyanonaphthalene. Method Boron Template/Source Solvent System Temperature % Relative Number of (°C) Conversion BsubNc via HPLC Products 4.1 BCl 3 ClNAP 130 31 5 (1.0 M heptane, 1.0 eq) (0.47 M) 4.2 BCl 3 1-MNAP 130 8 2 (1.0 M heptane, 1.0 eq) (0.47 M) 4.3 BCl 3 ClBZH:1-MNAP 130 4 2 (1.0 M heptane, 1.0 eq) (1:1, 0.47 M) 4.4 BCl 3 1,2-DCB:TOL 130 40 5 (1.0 M heptane, 1.0 eq) (1:1, 0.47 M) 4.5 BCl 3 1,2-DCB:1-MNAP 130 12 2 (1.0 M heptane, 1.0 eq) (1:1, 0.47 M) 4.6 BCl 3 1,2-DCB:1-MNAP 130 22 2 (1.0 M heptane, 2.0 eq) (1:1, 0.47 M) 4.7 BCl 3 1,2-DCB:1-MNAP 130 23 2 (1.0 M heptane, 4.0 eq) (1:1, 0.47 M) 4.8 BCl 3 1,2-DCB 180 70 5 (1.0 M heptane, 2.5 eq) (0.17 M) 4.9 BCl 3 1,2-DCB:1-MNAP 180 7 2 (1.0 M heptane, 2.5 eq) (3:1, 0.17 M) 4.10 BCl 3 1,2-DCB:1-MNAP 180 0 0 (1.0 M heptane, 2.5 eq) (6:1, 0.17 M) 4.11 BCl 3 1,2-DCB 180 88 5 (1.0 M heptane, 1.1 eq) (0.28 M) 4.12 BCl 3 1,2-DCB:1-MNAP 180 38 2 (1.0 M heptane, 1.1 eq) (4:1, 0.28 M) 4.13 BCl 3 1,2,4-TCB:1-MNAP 180 55 2 (1.0 M heptane, 1.1 eq) (4:1, 0.28 M) 4.14 BCl 3 1,2-DCB:1-MNAP 180 1 2 (1.0 M heptane, 2.2 eq) (4:1, 0.28 M) 4.15 BCl 3 1,2,4-TCB:1-MNAP 180 20 2 (1.0 M heptane, 2.2 eq) (4:1, 0.28 M) *ClNAP = 1-chloronaphthalene; 1,2,4-TCB = 1,2,4-trichlorobenzene; 1-MNAP = 1-methylnaphthalene; 1,2-DCB = 1,2-dichlorobenzene; ClBZH = chlorobenzene; TOL = toluene.

264

Synthesis of Cl-BsubNc (Ethylene Glycol Method): Method 5.1 to 5.10 (Table S5.6 ): Method 5.1 was carried out in a similar manner as those in Method 2. For 5.2 to 5.10 , to a three-neck round bottom flask equipped with a gas inlet and a short-path condenser was added DCNAP (0.150 g, 0.84 mmol, 1 equiv), 1,2-dichlorobenzene, and ethylene glycol under an atmosphere of argon gas. The mixture was stirred for 10 minutes at room temperature before BCl 3 (1.0 M in heptane, 2.0 mL, 2.4 equiv) was added. The reaction mixture was heated to 150 °C for one hour to distill off the heptane before it was heated to 180 °C for two hours. The crude product was precipitated into ice cold stirring hexane (50 mL), gravity filtered, washed with hexane (3 x 10 mL), and dried in a vacuum oven (60 °C).

265

Table S5.6 . Attempted syntheses using ethylene glycol as an additive. Method Boron Solvent System Temperature % Relative Number of Template/Source (°C) Conversion BsubNc via HPLC Products 5.1 BCl 3 1,2-DCB 180 85 5 (1.0 M heptane, 2.4 eq) (5 mL, 0.17 M) 5.2 BCl 3 1,2-DCB (4.94 mL):EG 180 34 5 (1.0 M heptane, 2.4 eq) (0.056 mL) - 0.5 eq wrt BCl 3 (0.17 M) 5.3 BCl 3 1,2-DCB (4.92 mL):EG 180 72 5 (1.0 M heptane, 2.4 eq) (0.084 mL) - 0.75 eq wrt BCl 3 (0.17 M) 5.4 BCl 3 1,2-DCB (4.91 mL):EG 180 84 5 (1.0 M heptane, 2.4 eq) (0.095 mL) - 0.85 eq wrt BCl 3 (0.17 M) 5.5 BCl 3 1,2-DCB (4.90 mL):EG 180 75 5 (1.0 M heptane, 2.4 eq) (0.100 mL) - 0.9 eq wrt BCl 3 (0.17 M) 5.6 BCl 3 1,2-DCB (4.89 mL):EG 180 81 5 (1.0 M heptane, 2.4 eq) (0.106 mL) - 0.95 eq wrt BCl 3 (0.17 M) 5.7 BCl 3 1,2-DCB (4.89 mL):EG 180 54 5 (1.0 M heptane, 2.4 eq) (0.108 mL) - 0.97 eq wrt BCl 3 (0.17 M) 5.8 BCl 3 1,2-DCB (4.89 mL):EG 180 87 5 (1.0 M heptane, 2.4 eq) (0.109 mL) - 0.98 eq wrt BCl 3 (0.17 M) 5.9 BCl 3 1,2-DCB (4.89 mL):EG 180 0 0 (1.0 M heptane, 2.4 eq) (0.112 mL) - 1 eq wrt BCl 3 (0.17 M) 5.10 BCl 3 1,2-DCB (4.78 mL):EG 180 0 0 (1.0 M heptane, 2.4 eq) (0.223 mL) - 2 eq wrt BCl 3 (0.17 M) *1,2-DCB = 1,2-dichlorobenzene; EG = ethylene glycol.

Synthesis of Cl-BsubNc (Chlorinating Species Scavenging Method): Method 6.1 to 6.4 (Table S5.7 ): All reactions were set-up in a similar manner as those in Method 2. To a three-neck round bottom flask equipped with a gas inlet and a short-path condenser was added DCNAP (1.10 g, 6.18 mmol, 1 equiv), 1,2,4-trichlorobenzene (16 mL), and the scavenger cosolvent (4 mL) under an atmosphere of argon gas. The mixture was stirred for

10 minutes at room temperature before BCl 3 (1.0 M in heptane, 6.5 mL, 1.1 equiv) was added. The reaction mixture was heated to 150 °C for one hour to distill off the heptane before it was heated to 180 °C for two hours. The reaction mixture was hot filtered under vacuum and the solid black mass was washed with hot toluene (~100 °C) until the filtrate was no longer blue in

266

colour. The filtrate solution was concentrated to remove the toluene via rotary evaporation before it was precipitated into ice cold stirring hexane (200 mL), gravity filtered, washed with hexane (3 x 50 mL), and dried in a vacuum oven (60 °C). For 6.4 , the crude product (0.295 g) was purified by train sublimation. The apparatus was operated under a vacuum with a controlled flow of carbon dioxide gas generating an internal pressure of 100 mTorr. The temperature was increased from room temperature to ~550 °C (external temperature) and was held constant at that temperature overnight. A metallic purple band was collected (18 mg, 6.1% yield relative to mass placed in the train sublimation apparatus). The sublimed product was subsequently purified a second time by train sublimation under the same condition (13 mg, 73% yield relative to mass

1 placed in the train sublimation apparatus). H NMR (400 MHz, CDCl 3) d 9.43 (s, 6H), 8.38-8.33 -1 -1 (m, 6H), 7.78-7.73 (m, 6H); MS (EI): m/z 580.1; UV-vis toluene λmax (ε): 651 nm (79200 M cm ).

Table S5.7 . Attempted syntheses using potential chlorinating species scavenger as an additive. Method Boron Template/Source Solvent System Temperature % Relative Number of (°C) Conversion BsubNc via HPLC Products 6.1 BCl 3 1,2,4-TCB:DP 180 0 0 (1.0 M heptane, 1.1 eq) (4:1, 0.28 M) 6.2 BCl 3 1,2,4-TCB:βP 180 17 2 (1.0 M heptane, 1.1 eq) (4:1, 0.28 M) 6.3 BCl 3 1,2,4-TCB:1-OD 180 0 0 (1.0 M heptane, 1.1 eq) (4:1, 0.28 M) 6.4 BCl 3 1,2,4-TCB: p-CYM 180 77 2 (1.0 M heptane, 1.1 eq) (4:1, 0.28 M) *1,2,4-TCB = 1,2,4-trichlorobenzene; DP = dipentene; βP = (-)-β-pinene, 1-OD = 1-octadecene; p-CYM = para - cymene.

267

HPLC-UV/Vis Chromatogram and Mass Spectrum of Cl-BsubNc ( p-Cymene Method)

Figure S5.4a. HPLC chromatogram (extracted at 650 nm) and the respective UV-vis absorption spectrum of each peak for Cl-BsubNc made via the p-cymene method (Method 6.4 ). Note that the peak at ~1.6 min is from the N,N -dimethylformamide solvent.

268

Figure S5.4b. Mass spectrum of sublimed Cl-BsubNc made via the p-cymene method (Method

6.4 ) indicating the additional presence of Cl-Cl 1BsubPc and absence of Cl-Cl 2BsubNc and Cl-

Cl 3BsubNc.

269

Synthesis of Cl-BsubNc (Electron-Poor Solvent Method): Method 7.1 to 7.3 (Table S5.8 ): All reactions were set-up in a similar manner as those in Method 2. To a three-neck round bottom flask equipped with a gas inlet and a short-path condenser was added DCNAP (0.100 mg, 0.56 mmol, 1 equiv) and the solvent (2 mL) under an atmosphere of argon gas. The mixture was stirred for 10 minutes at room temperature before

BCl 3 (1.0 M in heptane, 0.62 mL, 1.1 equiv) was added. The reaction mixture was heated to 150 °C for one hour to distill off the heptane before it was heated to 180 °C for two hours. The reaction mixture was hot filtered under vacuum and the solid black mass was washed with hot toluene (~100 °C) until the filtrate was no longer blue in colour. The filtrate solution was concentrated to remove the toluene via rotary evaporation before it was precipitated into ice cold stirring hexane (20 mL), gravity filtered, washed with hexane (3 x 10 mL), and dried in a vacuum oven (60 °C). For 7.4 , the reaction was scaled up by a factor of 45. The crude product (0.75 g) was purified by train sublimation. The apparatus was operated under a vacuum with a controlled flow of carbon dioxide gas generating an internal pressure of 100 mTorr. The temperature was increased from room temperature to ~580 °C (external temperature) and was held constant at that temperature overnight. A metallic purple band was collected (79 mg, 10% yield relative to mass placed in the train sublimation apparatus). The sublimed product was subsequently purified a second time by train sublimation under the same condition (61 mg, 77% yield relative to mass placed in the train sublimation apparatus), producing single crystals 1 suitable for X-ray diffraction. H NMR (400 MHz, CDCl 3) d 9.49-9.37 (m, 3H), 8.92-8.82 (m, 3H), 8.42-8.34 (m, 3H), 7.97-7.86 (m, 3H), 7.87-7.75 (m, 3H); MS (EI) 580.1, 614.1, 648.1, -1 -1 684.0, 716.0, 749.9: m/z ; UV-vis toluene λmax (ε): 664 nm (M cm ).

270

Table S5.8 . Attempted syntheses using electron-poor solvents. Method Boron Template/Source Solvent System Temperature % Relative Number of (°C) Conversion BsubNc via HPLC Products 7.1 BCl 3 NB 180 - 5 (1.0 M heptane, 1.1 eq) (0.28 M) 7.2 BCl 3 1,2,4-TCB 180 42 5 (1.0 M heptane, 1.1 eq) (0.28 M) 7.3 BCl 3 2,4-DNFB 180 0 0 (1.0 M heptane, 1.1 eq) (0.28 M) 7.4 BCl 3 NB 180 - 5 (1.0 M heptane, 1.1 eq) (0.28 M) *NB = nitrobenzene; 1,2,4-TCB = 1,2,4-trichlorobenzene; 2,4-DNFB = 2,4-dinitro-1-fluorobenzene.

271

HPLC-UV/Vis Chromatogram and Mass Spectrum of Cl-BsubNc (Nitrobenzene Method)

Figure S5.5a . HPLC chromatogram (extracted at 650 nm) and the respective UV-vis absorption spectrum of each peak for Cl-BsubNc made via the nitrobenzene method (Method 7.4 ). Note that the peak at ~1.6 min is from the N,N - dimethylformamide solvent.

272

Figure S5.5b. Mass spectrum of sublimed Cl-BsubNc made via the nitrobenzene method

(Method 7.4 ) indicating the additional presence of Cl-Cl 1BsubPc, Cl-Cl 2BsubPc, Cl-Cl 3BsubPc,

Cl-Cl 4BsubPc, and Cl-Cl 5BsubPc.

273

Table S5.9. Comparison of selected crystallographic parameters for the single crystals obtained

for literature-Cl-Cl nBsubNc and nitrobenzene-Cl-Cl nBsubNc.

Literature-Cl- Number Object1 Object2 Length Unit cell a 14.623 Å Cl nBsubNc dimensions = 1 centroid: C3 C2 centroid: C6 C7 C5 C4 C9 C8 3.654 Å b 18.453 Å C11 C10 C9 C4 = 2 centroid: C4 C5 centroid: C4 C9 C10 C3 C2 C11 3.654 Å c 10.3923 Å C9 C8 C7 C6 = 3 centroid: C17 C18 centroid: N1 C12 N2 C12 N1 B1 3.705 Å Density 1.468 Mg/m3 C18 C17 C16 C16 (calculated)

Nitrobenzene- Number Object1 Object2 Length Unit cell a 14.9062 Å Cl-Cl nBsubNc dimensions = 1 centroid: C12 C13 centroid: C8 C9 C10 C15 C16 C17 3.664 Å b 18.3548 Å C14 C15 C10 C11 = 2 centroid: C16 C15 centroid: C10 C11 C12 C13 C14 3.664 Å c 10.7766 Å C10 C9 C8 C17 C15 = 4 centroid: C1 C1 centroid: N4 B1 N4 C18 N5 C18 3.754 Å Density 1.538 Mg/m3 C2 C2 C3 C3 (calculated)

Differences Benzoisoindoline/Benzoisoindoline -0.010 Å Unit cell a -0.2832 Å dimensions = Benzoisoindoline/Benzoisoindoline -0.010 Å b 0.0982 Å = Benzoisoindoline/B|N|N|N|C|C -0.049 Å c -0.3843 Å = Density -0.07 Mg/m3 (calculated)

274

Figure S5.6 . Solid-state arrangement of Cl-Cl nBsubNc within crystals obtained from the literature method and train sublimed.

Figure S5.7 . Solid-state arrangement of Cl-Cl nBsubNc within crystals obtained from the nitrobenzene method and train sublimed.

275

HPLC-UV/Vis Chromatogram and Mass Spectrum of Cl-BsubNc (Commercial)

Figure S5.8a . HPLC chromatogram (extracted at 650 nm) and the respective UV-vis absorption spectrum of each peak for the commercial-Cl-BsubNc. Note that the peak at ~1.6 min is from the N,N - dimethylformamide solvent.

276

Figure S5.8b. Mass spectrum of commercial Cl-BsubNc indicating the additional presence of

Cl-Cl 1BsubPc, Cl-Cl 2BsubNc, and Cl-Cl 3BsubNc.

277

X-ray Photoelectron Spectroscopy (XPS) B 1s and N 1s core level XPS spectra are shown in Figure S5.10. Binding energies of various chlorine, boron, and nitrogen bonds for each Cl-Cl nBsubNc are summarized in Table S5.10. Deconvolution ( i.e. peak fitting) was performed for each spectrum (Cl 2p, B 1s, N 1s) to determine the number of distinct chemical states for chlorine (Figure S5.9), boron (Figure S5.11), and nitrogen (Figure S5.12). Each deconvoluted curve represents a distinct chemical state for B 1s and N 1s. For Cl 2p, due to a spin orbital splitting effect, each distinct chemical state is represented by two deconvoluted curves (2p 1/2 and 2p 3/2 ); there are four deconvoluted curves in each Cl 2p XPS plot to indicate two distinct chemical states for chlorine. Integrated peak intensity ratios are summarized in Table S5.11.

(a) (b)

Figure S5.9 . Deconvolution of Cl 2p XPS data of (a) literature-Cl-Cl nBsubNc, (b) commercial-

Cl-Cl nBsubNc, (c) p-cymene-Cl-Cl nBsubNc, and (d) nitrobenzene-Cl-Cl nBsubNc. Colors: blue – experimental XPS curve; red – summed of the fitted curves; light blue – Cl-C, Cl 2p 1/2 curve; green – Cl-B, Cl 2p 1/2 curve; orange – Cl-C, Cl 2p 3/2 curve; purple – Cl-B, Cl 2p 3/2 curve.

278

Table S5.10 . Core level binding energies of Cl-Cl nBsubNcs. Compound Cl 2p B 1s N1s Cl(-B) Cl(-C) B(-Cl and -N) N(-B) N(-C) Literature-Cl- 199.59 200.72 191.98 399.33 400.1 Cl nBsubNc Commercial-Cl- 199.50 200.66 191.96 399.30 400.07 Binding Cl nBsubNc Energy p-Cymene-Cl- 199.45 200.47 191.84 399.20 399.93 (eV) Cl nBsubNc Nitrobenzene-Cl- 199.57 200.75 192.06 399.44 400.21 Cl nBsubNc Average 199.53 200.65 191.96 399.32 400.08 Kahn et al .11 199.3 200.5 - - - *Cl(-B) = chlorine bonded to boron; Cl(-C) = chlorine bonded to carbon; N(-B) = isoindoline nitrogen; N(-C) = bridging nitrogen.

279

(a) (b)

Figure S5.10 . Overlay of the literature-Cl-Cl nBsubNc (blue), commercial-Cl-Cl nBsubNc (red), p-cymene-Cl-Cl nBsubNc (green) and nitrobenzene-Cl-Cl nBsubNc (purple) (a) normalized B 1s core-level XPS spectra and (b) normalized N 1s core-level XPS spectra.

280

(a) (b)

Figure S5.11 . Deconvolution of B 1s XPS data of (a) literature-Cl-Cl nBsubNc, (b) commercial-

Cl-Cl nBsubNc, (c) p-cymene-Cl-Cl nBsubNc, and (d) nitrobenzene-Cl-Cl nBsubNc. Colors: blue – experimental XPS curve; red – summed of the fitted curves; green – fitted curve 1.

281

(a) (b)

Figure S5.12 . Deconvolution of N 1s XPS data of (a) literature-Cl-Cl nBsubNc, (b) commercial-

Cl-Cl nBsubNc, (c) p-cymene-Cl-Cl nBsubNc, and (d) nitrobenzene-Cl-Cl nBsubNc. Colors: blue – experimental XPS curve; red – summed of the fitted curves; green – fitted curve 1; purple – fitted curve 2.

282

Table S5.11 . Integrated peak intensity ratios of Cl-Cl nBsubNcs.

Compound Cl-B / B Cl-C / B Cl-C / Cl-B Cl total / N total 6 (Cl total / N total ) - 1 Literature-Cl- 0.93 1.06 1.21 0.40 1.40 Cl nBsubNc Commercial-Cl- 0.90 1.36 1.61 0.46 1.76 Cl nBsubNc p-Cymene-Cl- 0.93 0.10 0.18 0.22 0.32 Cl nBsubNc Nitrobenzene-Cl- 0.80 3.12 4.17 0.90 4.40 Cl nBsubNc Kahn et al .11 - - ~1.5 - -

283

1 Figure S5.13 . H NMR (400 MHz, CDCl 3, 296 K) spectrum of the aromatic region for literature-

Cl-Cl nBsubNc (blue), 1-MNAP-Cl-Cl nBsubNc (orange), p-cymene-Cl-Cl nBsubNc (green), nitrobenzene-Cl-Cl nBsubNc (purple), and commercial-Cl-Cl nBsubNc (red). Integration values are shown below the peaks.

284

UV-Vis Absorption and Photoluminescence Plots of Cl-Cl nBsubNc

(a)

(b)

Figure S5.14 . Overlay of the literature-Cl-Cl nBsubNc (blue), commercial-Cl-Cl nBsubNc (red), p-cymene-Cl-Cl nBsubNc (green), and nitrobenzene-Cl-Cl nBsubNc (purple) (a) normalized absorption spectra and (b) normalized photoluminescence (PL) spectra.

285

Wavelength of Maximum Absorption of Cl-Cl nBsubPcs

Table S5.12 . λmax of Absorption Values of Peripherally Chlorinated Cl-BsubPc.

BsubPc Compound λmax (nm) Cl-BsubPc 565 Cl-Cl 6BsubPc 574 Cl-Cl 12 BsubPc 593 All measurements were done in toluene solutions at room temperature.

Determination of Fluorescence Quantum Yields The fluorescence quantum yields (Φ) were calculated using the formula below:

2 2 Φ = Φ R (I / I R)(OD R / OD)(n / n R ) (Eq S5.1) where I is the integrated fluorescence intensity, OD is the optical density ( i.e. absorbance), and n is the refractive index of the solvent. The subscript R is oxazine 170, a reference fluorophore, which has previously been reported to have a Φ = 0.579 12 in ethanol at room temperature. Integrated fluorescence intensity values were acquired from PerkinElmer FL WinLab (version 4.00.03) while the optical density values were acquired from PerkinElmer UV WinLab (version 6.02.0723). Analyses of all BsubNc compounds were acquired in toluene at room temperature.

286

Solid State UV-vis Absorption and Photoluminescence Plots of Cl-Cl nBsubNc

(a)

(b) (c)

Figure S5.15 . Overlay of literature-Cl-Cl nBsubNc (blue), p-cymene-Cl-Cl nBsubNc (green), nitrobenzene-Cl-Cl nBsubNc (purple), and commercial-Cl-Cl nBsubNc (red) (a) normalized solid- state absorption spectra and (b,c) normalized solid-state photoluminescence (PL) spectra excited at 630 nm and 650 nm, respectively.

287

Cyclic and Differential Pulse Voltammograms

(a)

(b)

Figure S5.16 . (a) Cyclic voltammetry and (b) differential pulse voltammetry traces of literature-

Cl-Cl nBsubNc in DCM with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference at room temperature.

288

(a)

(b)

Figure S5.17 . (a) Cyclic voltammetry and (b) differential pulse voltammetry traces of p-cymene-

Cl-Cl nBsubNc in DCM with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference at room temperature.

289

(a)

(b)

Figure S5.18 . (a) Cyclic voltammetry and (b) differential pulse voltammetry traces of commercial-Cl-Cl nBsubNc in DCM with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference at room temperature.

290

(a)

(b)

Figure S5.19 . (a) Cyclic voltammetry and (b) differential pulse voltammetry traces of literature-

Cl-Cl nBsubNc in DMF with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference at room temperature.

291

(a)

(b)

Figure S5.20 . (a) Cyclic voltammetry and (b) differential pulse voltammetry traces of p-cymene-

Cl-Cl nBsubNc in DMF with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference at room temperature.

292

(a)

(b)

Figure S5.21 . (a) Cyclic voltammetry and (b) differential pulse voltammetry traces of nitrobenzene-Cl-Cl nBsubNc in DMF with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference at room temperature.

293

(a)

(b)

Figure S5.22 . (a) Cyclic voltammetry and (b) differential pulse voltammetry traces of commercial-Cl-Cl nBsubNc in DMF with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference at room temperature.

294

Table S5.13 . Electrochemical properties (DPV) of literature, commercial, and p-cymene Cl-

Cl nBsubNc.

Alternative Processing Methods for Analyzing DP Voltammograms Three additional processing methods were employed to determine if the results from the apex peak method (used in all prior voltammetry experiments) are in line with these alternative methods. The first method derives the potential values from the onset of a peak, the second method derives the potential values from the first derivative zero point, and the third method derives the potential values from the maximum slope point. The results are summarized below:

Table S5.14 . DPV potential values derived from three different processing methods for the four

Cl-Cl nBsubNc sources. Method 1 a Method 2 b Method 3 c 1' 2' 1' 2' 1' 2' 1' 2' 1' 2' 1' 2' E ox | E ox E red | E red | E ox | E ox E red | E red | E ox | E ox E red | E red | 3' 3' 3' Compound (V) E red (V) (V) E red (V) (V) E red (V) Literature 0.65, 0.92 -0.75, N/A, -1.21 0.52, 0.91 -0.60, N/A, -1.17 0.70, 0.95 -0.83, N/A, -1.25 Commercial 0.66, 0.94 -0.75, N/A, -1.18 0.50, 0.93 -0.60, N/A, -1.16 0.71, 0.97 -0.82, N/A, -1.24 p-Cymene 0.59, 0.86 -0.85, -1.03, -1.27 0.48, 0.82 0.68, -1.02, -1.19 0.65, 0.92 -0.89, -1.05, -1.32 Nitrobenzene 0.68, 0.91 -0.69, -0.94, -1.17 0.52, 0.88 -0.55,-0.95,-1.08 0.73, 0.96 -0.75, -0.96, -1.21 a E’ derived from peak onset. b E’ derived from the first derivative zero point. c E’ derived from the maximum slope point.

295

Ultraviolet Photoelectron Spectroscopy (UPS) The ionization energy ( i.e. HOMO energy level) is determined using the following equation: Ionization energy = ɸ + Δe, (Eq S5.2) where ɸ is the work function and Δe is the HOMO-Fermi gap. The Δ e is determined from the UPS HOMO-derived plot (Figure S23b) via the extrapolation of the linear portion of the tailing peak with the x-axis. The ɸ is calculated from the following equation.

ɸ = E UV – SEC + B, (Eq S5.3) where E UV is the energy of the incident UV light, SEC is the secondary electron cutoff, and B is the electrical bias applied during measurement. The SEC is determined from the UPS SEC plot (Figure S23a) via the extrapolation of the linear portion of the onset peak with the x-axis. The

EUV and B used in these measurements are 21.22 and -15 eV, respectively.

296

(a)

(b)

Figure S5.23 . Ultraviolet photoelectron spectroscopy (UPS) He Iα (hν = 21.22 eV) valence band spectra of literature-Cl-Cl nBsubNc (blue), commercial-Cl-Cl nBsubNc (red), p-cymene-Cl-

Cl nBsubNc (green) and nitrobenzene-Cl-Cl nBsubNc (purple) showing (a) the secondary electron cutoff and (b) the HOMO-derived peak.

297

References 1. Rauschnabel, J.; Hanack, M. Tetrahedron Letters 1995 , 36 , 1629-1632. 2. Geyer, M.; Plenzig, F.; Rauschnabel, J.; Hanack, M.; Del Rey, B.; Sastre, A.; Torres, T. Synthesis 1996 , 1139-1151. 3. Kobayashi, N.; Ishizaki, T.; Ishii, K.; Konami, H. Journal of the American Chemical Society 1999 , 121 , 9096-9110. 4. Zyskowski, C. D.; Kennedy, V. O. Journal of Porphyrins and Phthalocyanines 2000 , 4, 707-712. 5. Nonell, S.; Rubio, N.; del Rey, B.; Torres, T. Journal of the Chemical Society, Perkin Transactions 2 2000 , 1091-1094. 6. Giribabu, L.; Kumar, C. V.; Surendar, A.; Reddy, V. G.; Chandrasekharam, M.; Reddy, P. Y. Synthetic Communications 2007 , 37 , 4141-4147. 7. Takao, Y.; Masuoka, T.; Yamamoto, K.; Mizutani, T.; Matsumoto, F.; Moriwaki, K.; Hida, K.; Iwai, T.; Ito, T.; Mizuno, T.; Ohno, T. Tetrahedron Letters 2014 , 55 , 4564- 4567. 8. Yamamoto, K.; Takagi, A.; Hada, M.; Taniwaki, R.; Mizutani, T.; Kumura, Y.; Takao, Y.; Moriwaki, K.; Matsumoto, F.; Ito, T.; Iwai, T.; Hida, K.; Mizuno, T.; Ohno, T. Tetrahedron 2016 , 72, 4918-4924. 9. Bouvet, M.; Bassoul, P.; Simon, J. Molecular Crystals and Liquid Crystals Science and Technology. Section A. Molecular Crystals and Liquid Crystals 1994 , 252 , 31-38. 10. Stork, J. R.; Potucek, R. J.; Durfee, W. S.; Noll, B. C. Tetrahedron Letters 1999 , 40 , 8055-8058. 11. Endres, J.; Pelczer, I.; Rand, B. P.; Kahn, A. Chemistry of Materials 2016 , 28 , 794-801. 12. Rurack, K.; Spieles, M. Analytical Chemistry 2011 , 83 , 1232-1242.

298

Appendix E: Supplementary Information for Chapter 6

Synthetic Procedures 2,3-Dicyanonaphthalene (DCNAP). DCNAP was prepared according to literature procedure. 1

Chloro boron subnaphthalocyanine (Cl-Cl nBsubNc) . To a three-neck round bottom flask equipped with a gas inlet and a short-path condenser was added DCNAP (3.80 g, 0.021 mol, 1 equiv) and dried 1,2-dichlorobenzene (80 mL) under an atmosphere of argon gas. The mixture was stirred for 10 minutes at room temperature before BCl 3 (1.0 M in heptane, 22.5 mL, 1.1 equiv) was added. The reaction mixture was heated to 150 °C for one hour to distill off the heptane before it was heated to 180 °C for two hours. The reaction was cooled to room temperature and concentrated via rotary evaporation. The crude product (0.730 g) was purified by train sublimation. The apparatus was operated under a vacuum with a controlled flow of carbon dioxide gas generating an internal pressure of 100 mTorr. The temperature was increased from room temperature to ~580 °C (external temperature) and was held constant at that temperature overnight. A metallic purple band was collected (0.150 g, 21% yield relative to mass placed in the train sublimation apparatus). The sublimed product was subsequently purified a second time by train sublimation under the same condition (0.115 g, 76% yield relative to mass

1 placed in the train sublimation apparatus). H NMR (400 MHz, CDCl 3) d 9.43 (s, 6H), 8.37-8.33

(m, 6H), 7.77-7.73 (m, 6H); MS (EI): m/z 580.1, 614.1, 648.1, 684.0; UV-vis toluene λmax (ε): 656 nm (78400 M -1cm -1).

Phenoxy boron subnaphthalocyanine (PhO-Cl nBsubNc) . To an oven-dried round bottom flask equipped with a condenser and a gas inlet was added Cl-Cl nBsubNc (0.41 g, 0.71 mmol, 1 equiv), phenol (0.33 g, 3.51 mmol, 5 equiv), and toluene (35 mL) under an atmosphere of argon gas. The reaction mixture was heated to reflux (~115 °C) and the reaction progress was monitored via HPLC. After ~16 hours, the reaction was cooled to room temperature and the solvent was removed under reduced pressure to yield a dark purple solid (0.48 g). The crude

299

product was purified twice by train sublimation. The apparatus was operated under a vacuum with a controlled flow of carbon dioxide gas generating an internal pressure of 100 mTorr. The temperature was increased from room temperature to ~560 °C (external temperature) and was held constant at that temperature overnight. A metallic purple band was collected (126 mg, 28%). Single crystals suitable for X-ray diffraction analysis were grown by slow evaporation

1 from DCM solution. H NMR (400 MHz, CDCl 3) d 9.40-9.25 (m, 6H), 8.35-8.22 (m, 6H), 7.72- 7.61 (m, 6H), 6.83-6.74 (m, 2H), 6.69-6.61 (m, 1 H), 5.64-5.55 (m, 2H); MS (EI): m/z 638.2, -1 -1 672.2, 706.1, 740.1; UV-vis toluene λmax (ε): 650 nm (90000 M cm ).

Pentafluorophenoxy boron subnaphthalocyanine (F 5-Cl nBsubNc) . To an oven-dried round bottom flask equipped with a condenser and a gas inlet was added Cl-Cl nBsubNc (0.33 g, 0.57 mmol, 1 equiv), pentafluorophenol (0.52 g, 2.84 mmol, 5 equiv), and toluene (30 mL) under an atmosphere of argon gas. The reaction mixture was heated to reflux (~115 °C) and the reaction progress was monitored via HPLC. After ~16 hours, the reaction was cooled to room temperature and the solvent was removed under reduced pressure to yield a dark purple solid (0.43 g). The crude product was purified twice by train sublimation. The apparatus was operated under a vacuum with a controlled flow of carbon dioxide gas generating an internal pressure of 100 mTorr. The temperature was increased from room temperature to ~560 °C (external temperature) and was held constant at that temperature overnight. A metallic purple band was

1 collected (0.13 g, 31%). H NMR (400 MHz, CDCl 3) d 9.45-9.34 (m, 6H), 8.39-3.20 (m, 6H), - 7.81-7.70 (m, 6H); MS (EI): m/z 728.2, 762.1, 796.1; UV-vis toluene λmax (ε): 654 nm (86000 M 1cm -1).

300

HPLC-UV/Vis Chromatograms and Mass Spectra

(a)

Figure S6.1a . (a) HPLC chromatogram (extracted at 650 nm) and (b) the respective UV-vis absorption spectrum of each peak for PhO-Cl nBsubNc. Note that the peak at ~1.6 min is from the N,N - dimethylformamide solvent.

(b)

301

Figure S6.1b . Mass spectrum of sublimed PhO-Cl nBsubNc.

302

(a)

Figure S6.2a . (a) HPLC chromatogram (extracted at 650 nm) and (b) the respective UV-vis absorption spectrum of each peak for F 5-Cl nBsubNc. Note that the peak at ~1.6 min is from the N,N - dimethylformamide solvent.

(b)

303

Figure S6.2b . Mass spectrum of sublimed F 5-Cl nBsubNc.

304

Crystal Structure Details of PhO-Cl nBsubNc

Figure S6.3 . Ellipsoid plot (50% probability) showing the structure and atom numbering scheme of PhO-Cl nBsubNc. Hydrogen atoms have been omitted for clarity. Colors: boron - yellow; nitrogen - blue; carbon - white; chlorine - green.

305

Table S6.1 . Crystal data and structure refinement for PhO-Cl nBsubNc.

Identification code d1541_a Empirical formula C42.50 H23.01 B Cl1.99 N6 O Formula weight 714.86 Temperature 147(2) K Wavelength 1.54178 Å Crystal system Monoclinic Space group C2/c Unit cell dimensions a = 29.4349(7) Å α= 90°. b = 14.1483(3) Å β= 130.4770(10)°. c = 20.0640(5) Å γ = 90°. Volume 6355.9(3) Å 3 Z 8 Density (calculated) 1.494 Mg/m 3 Absorption coefficient 2.216 mm -1 F(000) 2934 Crystal size 0.200 x 0.100 x 0.030 mm 3 Theta range for data collection 3.695 to 67.084°. Index ranges -35<=h<=34, -15<=k<=16, -23<=l<=23 Reflections collected 74573 Independent reflections 5651 [R(int) = 0.0385] Completeness to theta = 67.084° 99.6 % Absorption correction Semi-empirical from equivalents Max. and min. transmission 0.7529 and 0.5776 Refinement method Full-matrix least-squares on F 2 Data / restraints / parameters 5651 / 12 / 546 Goodness-of-fit on F 2 1.046 Final R indices [I>2sigma(I)] R1 = 0.0591, wR2 = 0.1483 R indices (all data) R1 = 0.0639, wR2 = 0.1525 Extinction coefficient n/a Largest diff. peak and hole 1.068 and -0.651 e.Å -3

306

Table S6.2 . Atomic coordinates (x 10 4) and equivalent isotropic displacement parameters (Å 2 x

10 3) for PhO-Cl nBsubNc. U(eq) is defined as one third of the trace of the orthogonalized U ij tensor. ______x y z U(eq) ______B(1) 2599(1) 4635(2) 3735(2) 27(1) Cl(1) 4909(1) 6468(2) 6013(2) 49(1) Cl(2) 3894(5) 5072(8) 2522(7) 67(4) Cl(3) 1513(4) 5281(6) 133(6) 38(3) Cl(4) 237(1) 6289(2) 1316(2) 43(1) Cl(5) 1089(1) 6550(2) 4188(2) 42(1) Cl(6) 3839(1) 6636(2) 6960(2) 46(1) O(1) 2702(1) 3651(1) 3955(1) 31(1) N(1) 3126(1) 5013(1) 3867(1) 28(1) N(2) 2648(1) 4980(1) 2357(1) 34(1) N(3) 2077(1) 4912(1) 2814(1) 28(1) N(4) 1520(1) 5620(1) 3166(1) 31(1) N(5) 2551(1) 5282(1) 4278(1) 28(1) N(6) 3578(1) 5717(1) 5245(1) 31(1) C(1) 3597(1) 5436(2) 4622(2) 30(1) C(2) 4020(1) 5650(2) 4494(2) 34(1) C(3) 4578(1) 6046(2) 5029(2) 42(1) C(4) 4882(1) 6173(2) 4703(2) 50(1) C(2A) 4020(1) 5650(2) 4494(2) 34(1) C(3A) 4578(1) 6046(2) 5029(2) 42(1) C(4A) 4882(1) 6173(2) 4703(2) 50(1) C(5) 5473(1) 6560(2) 5250(3) 67(1) C(6) 5758(2) 6664(2) 4919(4) 87(2) C(7) 5477(2) 6392(3) 4051(4) 90(2) C(8) 4921(2) 6027(2) 3521(3) 70(1) C(9) 4602(1) 5906(2) 3826(2) 50(1) C(10) 4021(1) 5523(2) 3277(2) 44(1) C(11) 3737(1) 5394(2) 3607(2) 35(1) C(9A) 4602(1) 5906(2) 3826(2) 50(1)

307

C(10A) 4021(1) 5523(2) 3277(2) 44(1) C(11A) 3737(1) 5394(2) 3607(2) 35(1) C(12) 3142(1) 5054(2) 3202(2) 31(1) C(13) 2123(1) 4976(2) 2177(2) 31(1) C(14) 1538(1) 5272(2) 1389(2) 34(1) C(15) 1308(1) 5398(2) 541(2) 41(1) C(16) 724(1) 5755(2) -86(2) 44(1) C(14A) 1538(1) 5272(2) 1389(2) 34(1) C(15A) 1308(1) 5398(2) 541(2) 41(1) C(16A) 724(1) 5755(2) -86(2) 44(1) C(17) 459(2) 5873(2) -980(2) 58(1) C(18) -101(2) 6228(2) -1577(2) 64(1) C(19) -428(2) 6495(2) -1327(2) 62(1) C(20) -198(1) 6387(2) -478(2) 53(1) C(21) 385(1) 6009(2) 165(2) 44(1) C(22) 633(1) 5885(2) 1047(2) 38(1) C(23) 1193(1) 5508(2) 1644(2) 33(1) C(21A) 385(1) 6009(2) 165(2) 44(1) C(22A) 633(1) 5885(2) 1047(2) 38(1) C(23A) 1193(1) 5508(2) 1644(2) 33(1) C(24) 1568(1) 5319(2) 2578(2) 31(1) C(25) 2024(1) 5632(2) 4012(2) 29(1) C(26) 2175(1) 6133(2) 4766(2) 31(1) C(27) 1836(1) 6528(2) 4943(2) 37(1) C(28) 2122(1) 6922(2) 5777(2) 39(1) C(26A) 2175(1) 6133(2) 4766(2) 31(1) C(27A) 1836(1) 6528(2) 4943(2) 37(1) C(28A) 2122(1) 6922(2) 5777(2) 39(1) C(29) 1790(2) 7252(2) 6020(2) 49(1) C(30) 2077(2) 7580(2) 6847(2) 58(1) C(31) 2701(2) 7634(2) 7464(2) 58(1) C(32) 3034(2) 7344(2) 7254(2) 49(1) C(33) 2759(1) 6963(2) 6413(2) 40(1) C(34) 3102(1) 6592(2) 6207(2) 37(1) C(35) 2821(1) 6164(2) 5418(2) 31(1) C(33A) 2759(1) 6963(2) 6413(2) 40(1)

308

C(34A) 3102(1) 6592(2) 6207(2) 37(1) C(35A) 2821(1) 6164(2) 5418(2) 31(1) C(36) 3043(1) 5676(2) 5039(1) 30(1) C(37) 2334(1) 2995(2) 3887(1) 28(1) C(38) 2547(1) 2076(2) 4103(2) 34(1) C(39) 2210(1) 1360(2) 4058(2) 40(1) C(40) 1658(1) 1552(2) 3811(2) 38(1) C(41) 1446(1) 2471(2) 3601(2) 38(1) C(42) 1779(1) 3193(2) 3632(2) 34(1) Cl(7) 5016(8) 6835(10) 7918(8) 516(15) Cl(8) 4884(3) 8313(8) 6875(7) 349(7) C(1S) 4580(5) 7759(11) 7249(11) 145(6) Cl(9) 5201(6) 6381(14) 7295(11) 84(6) Cl(10) 4707(6) 7444(16) 7920(11) 84(6) C(2S) 5322(7) 7228(17) 8007(15) 12(6) ______

309

Table S6.3 . Bond lengths [Å] and angles [°] for PhO-Cl nBsubNc. ______B(1)-O(1) 1.433(3) B(1)-N(1) 1.493(3) B(1)-N(3) 1.498(3) B(1)-N(5) 1.498(3) Cl(1)-C(3) 1.646(4) Cl(2)-C(10) 1.451(11) Cl(3)-C(15) 1.306(9) Cl(3)-C(16) 2.169(9) Cl(4)-C(22) 1.670(4) Cl(5)-C(27) 1.677(4) Cl(6)-C(34) 1.656(4) O(1)-C(37) 1.364(3) N(1)-C(1) 1.359(3) N(1)-C(12) 1.367(3) N(2)-C(13) 1.341(3) N(2)-C(12) 1.341(3) N(3)-C(24) 1.370(3) N(3)-C(13) 1.371(3) N(4)-C(24) 1.346(3) N(4)-C(25) 1.346(3) N(5)-C(25) 1.365(3) N(5)-C(36) 1.366(3) N(6)-C(36) 1.344(3) N(6)-C(1) 1.346(3) C(1)-C(2) 1.457(3) C(2)-C(3) 1.370(3) C(2)-C(11) 1.438(4) C(3)-C(4) 1.421(4) C(4)-C(9) 1.426(5) C(4)-C(5) 1.432(4) C(2A)-C(3A) 1.370(3) C(2A)-C(11A) 1.438(4) C(3A)-C(4A) 1.421(4)

310

C(3A)-H(3A) 0.9500 C(4A)-C(9A) 1.426(5) C(5)-C(6) 1.374(6) C(5)-H(5) 0.9500 C(6)-C(7) 1.414(7) C(6)-H(6A) 0.9500 C(7)-C(8) 1.348(6) C(7)-H(7A) 0.9500 C(8)-C(9) 1.425(4) C(8)-H(8A) 0.9500 C(9)-C(10) 1.410(4) C(10)-C(11) 1.373(3) C(11)-C(12) 1.454(3) C(9A)-C(10A) 1.410(4) C(10A)-C(11A) 1.373(3) C(10A)-H(10A) 0.9500 C(13)-C(14) 1.451(3) C(14)-C(15) 1.375(3) C(14)-C(23) 1.439(3) C(15)-C(16) 1.409(4) C(16)-C(21) 1.426(4) C(16)-C(17) 1.427(4) C(14A)-C(15A) 1.375(3) C(14A)-C(23A) 1.439(3) C(15A)-C(16A) 1.409(4) C(15A)-H(15A) 0.9500 C(16A)-C(21A) 1.426(4) C(16A)-C(17) 1.427(4) C(17)-C(18) 1.357(5) C(17)-H(17A) 0.9500 C(18)-C(19) 1.395(5) C(18)-H(18A) 0.9500 C(19)-C(20) 1.374(5) C(19)-H(19A) 0.9500 C(20)-C(21A) 1.422(4) C(20)-C(21) 1.422(4)

311

C(20)-H(20A) 0.9500 C(21)-C(22) 1.421(4) C(22)-C(23) 1.370(3) C(23)-C(24) 1.455(3) C(21A)-C(22A) 1.421(4) C(22A)-C(23A) 1.370(3) C(22A)-H(22A) 0.9500 C(25)-C(26) 1.454(3) C(26)-C(27) 1.375(3) C(26)-C(35) 1.449(3) C(27)-C(28) 1.412(4) C(28)-C(29) 1.425(4) C(28)-C(33) 1.429(4) C(26A)-C(27A) 1.375(3) C(26A)-C(35A) 1.449(3) C(27A)-C(28A) 1.412(4) C(27A)-H(27A) 0.9500 C(28A)-C(29) 1.425(4) C(28A)-C(33A) 1.429(4) C(29)-C(30) 1.365(4) C(29)-H(29A) 0.9500 C(30)-C(31) 1.401(5) C(30)-H(30A) 0.9500 C(31)-C(32) 1.358(4) C(31)-H(31A) 0.9500 C(32)-C(33A) 1.421(4) C(32)-C(33) 1.421(4) C(32)-H(32A) 0.9500 C(33)-C(34) 1.415(4) C(34)-C(35) 1.363(3) C(35)-C(36) 1.458(3) C(33A)-C(34A) 1.415(4) C(34A)-C(35A) 1.363(3) C(34A)-H(34A) 0.9500 C(35A)-C(36) 1.458(3) C(37)-C(38) 1.386(3)

312

C(37)-C(42) 1.386(3) C(38)-C(39) 1.380(3) C(38)-H(38A) 0.9500 C(39)-C(40) 1.382(4) C(39)-H(39A) 0.9500 C(40)-C(41) 1.384(4) C(40)-H(40A) 0.9500 C(41)-C(42) 1.388(3) C(41)-H(41A) 0.9500 C(42)-H(42A) 0.9500 Cl(7)-C(1S) 1.709(9) Cl(8)-C(1S) 1.688(9) C(1S)-H(1S1) 0.9900 C(1S)-H(1S2) 0.9900 Cl(9)-C(2S) 1.717(10) Cl(10)-C(2S) 1.729(10) C(2S)-H(2S1) 0.9900 C(2S)-H(2S2) 0.9900

O(1)-B(1)-N(1) 107.82(18) O(1)-B(1)-N(3) 118.45(19) N(1)-B(1)-N(3) 104.00(18) O(1)-B(1)-N(5) 118.00(19) N(1)-B(1)-N(5) 103.29(18) N(3)-B(1)-N(5) 103.46(18) C(15)-Cl(3)-C(16) 38.6(3) C(37)-O(1)-B(1) 129.47(18) C(1)-N(1)-C(12) 114.42(19) C(1)-N(1)-B(1) 123.06(19) C(12)-N(1)-B(1) 122.12(19) C(13)-N(2)-C(12) 117.02(19) C(24)-N(3)-C(13) 113.35(19) C(24)-N(3)-B(1) 123.48(19) C(13)-N(3)-B(1) 121.61(19) C(24)-N(4)-C(25) 116.99(19) C(25)-N(5)-C(36) 113.58(19)

313

C(25)-N(5)-B(1) 123.85(19) C(36)-N(5)-B(1) 122.02(18) C(36)-N(6)-C(1) 116.62(19) N(6)-C(1)-N(1) 122.5(2) N(6)-C(1)-C(2) 131.1(2) N(1)-C(1)-C(2) 105.3(2) C(3)-C(2)-C(11) 120.5(2) C(3)-C(2)-C(1) 132.5(2) C(11)-C(2)-C(1) 107.0(2) C(2)-C(3)-C(4) 119.0(3) C(2)-C(3)-Cl(1) 124.2(2) C(4)-C(3)-Cl(1) 116.6(2) C(3)-C(4)-C(9) 120.1(2) C(3)-C(4)-C(5) 120.7(3) C(9)-C(4)-C(5) 119.3(3) C(3A)-C(2A)-C(11A) 120.5(2) C(2A)-C(3A)-C(4A) 119.0(3) C(2A)-C(3A)-H(3A) 120.5 C(4A)-C(3A)-H(3A) 120.5 C(3A)-C(4A)-C(9A) 120.1(2) C(6)-C(5)-C(4) 119.3(4) C(6)-C(5)-H(5) 120.4 C(4)-C(5)-H(5) 120.4 C(5)-C(6)-C(7) 121.0(3) C(5)-C(6)-H(6A) 119.5 C(7)-C(6)-H(6A) 119.5 C(8)-C(7)-C(6) 120.9(4) C(8)-C(7)-H(7A) 119.6 C(6)-C(7)-H(7A) 119.6 C(7)-C(8)-C(9) 120.7(4) C(7)-C(8)-H(8A) 119.7 C(9)-C(8)-H(8A) 119.7 C(10)-C(9)-C(8) 120.9(3) C(10)-C(9)-C(4) 120.2(2) C(8)-C(9)-C(4) 118.9(3) C(11)-C(10)-C(9) 118.9(3)

314

C(11)-C(10)-Cl(2) 128.3(5) C(9)-C(10)-Cl(2) 110.1(5) C(10)-C(11)-C(2) 121.3(2) C(10)-C(11)-C(12) 131.6(3) C(2)-C(11)-C(12) 107.1(2) C(10A)-C(9A)-C(4A) 120.2(2) C(11A)-C(10A)-C(9A) 118.9(3) C(11A)-C(10A)-H(10A) 120.6 C(9A)-C(10A)-H(10A) 120.6 C(10A)-C(11A)-C(2A) 121.3(2) N(2)-C(12)-N(1) 122.4(2) N(2)-C(12)-C(11) 131.0(2) N(1)-C(12)-C(11) 105.1(2) N(2)-C(13)-N(3) 122.9(2) N(2)-C(13)-C(14) 129.7(2) N(3)-C(13)-C(14) 105.6(2) C(15)-C(14)-C(23) 121.0(2) C(15)-C(14)-C(13) 131.9(2) C(23)-C(14)-C(13) 107.0(2) Cl(3)-C(15)-C(14) 134.8(5) Cl(3)-C(15)-C(16) 106.1(4) C(14)-C(15)-C(16) 119.2(3) C(15)-C(16)-C(21) 120.1(2) C(15)-C(16)-C(17) 121.3(3) C(21)-C(16)-C(17) 118.6(3) C(15)-C(16)-Cl(3) 35.3(3) C(21)-C(16)-Cl(3) 155.3(3) C(17)-C(16)-Cl(3) 86.0(3) C(15A)-C(14A)-C(23A) 121.0(2) C(14A)-C(15A)-C(16A) 119.2(3) C(14A)-C(15A)-H(15A) 120.4 C(16A)-C(15A)-H(15A) 120.4 C(15A)-C(16A)-C(21A) 120.1(2) C(15A)-C(16A)-C(17) 121.3(3) C(21A)-C(16A)-C(17) 118.6(3) C(18)-C(17)-C(16) 120.7(3)

315

C(18)-C(17)-C(16A) 120.7(3) C(18)-C(17)-H(17A) 119.6 C(16)-C(17)-H(17A) 119.6 C(17)-C(18)-C(19) 120.7(3) C(17)-C(18)-H(18A) 119.7 C(19)-C(18)-H(18A) 119.7 C(20)-C(19)-C(18) 121.1(3) C(20)-C(19)-H(19A) 119.4 C(18)-C(19)-H(19A) 119.4 C(19)-C(20)-C(21A) 119.9(3) C(19)-C(20)-C(21) 119.9(3) C(19)-C(20)-H(20A) 120.1 C(21)-C(20)-H(20A) 120.1 C(22)-C(21)-C(20) 121.2(3) C(22)-C(21)-C(16) 119.9(2) C(20)-C(21)-C(16) 119.0(3) C(23)-C(22)-C(21) 119.3(3) C(23)-C(22)-Cl(4) 123.3(2) C(21)-C(22)-Cl(4) 117.3(2) C(22)-C(23)-C(14) 120.4(2) C(22)-C(23)-C(24) 132.4(2) C(14)-C(23)-C(24) 107.1(2) C(22A)-C(21A)-C(20) 121.2(3) C(22A)-C(21A)-C(16A) 119.9(2) C(20)-C(21A)-C(16A) 119.0(3) C(23A)-C(22A)-C(21A) 119.3(3) C(23A)-C(22A)-H(22A) 120.3 C(21A)-C(22A)-H(22A) 120.3 C(22A)-C(23A)-C(14A) 120.4(2) N(4)-C(24)-N(3) 122.9(2) N(4)-C(24)-C(23) 130.4(2) N(3)-C(24)-C(23) 105.5(2) N(4)-C(25)-N(5) 122.0(2) N(4)-C(25)-C(26) 131.4(2) N(5)-C(25)-C(26) 105.53(19) C(27)-C(26)-C(35) 120.1(2)

316

C(27)-C(26)-C(25) 133.1(2) C(35)-C(26)-C(25) 106.84(19) C(26)-C(27)-C(28) 119.5(2) C(26)-C(27)-Cl(5) 121.4(2) C(28)-C(27)-Cl(5) 119.1(2) C(27)-C(28)-C(29) 121.5(3) C(27)-C(28)-C(33) 120.1(2) C(29)-C(28)-C(33) 118.4(2) C(27A)-C(26A)-C(35A) 120.1(2) C(26A)-C(27A)-C(28A) 119.5(2) C(26A)-C(27A)-H(27A) 120.2 C(28A)-C(27A)-H(27A) 120.2 C(27A)-C(28A)-C(29) 121.5(3) C(27A)-C(28A)-C(33A) 120.1(2) C(29)-C(28A)-C(33A) 118.4(2) C(30)-C(29)-C(28) 120.5(3) C(30)-C(29)-C(28A) 120.5(3) C(30)-C(29)-H(29A) 119.8 C(28)-C(29)-H(29A) 119.8 C(29)-C(30)-C(31) 120.9(3) C(29)-C(30)-H(30A) 119.5 C(31)-C(30)-H(30A) 119.5 C(32)-C(31)-C(30) 120.4(3) C(32)-C(31)-H(31A) 119.8 C(30)-C(31)-H(31A) 119.8 C(31)-C(32)-C(33A) 120.9(3) C(31)-C(32)-C(33) 120.9(3) C(31)-C(32)-H(32A) 119.5 C(33)-C(32)-H(32A) 119.5 C(34)-C(33)-C(32) 121.5(3) C(34)-C(33)-C(28) 119.7(2) C(32)-C(33)-C(28) 118.8(2) C(35)-C(34)-C(33) 119.7(2) C(35)-C(34)-Cl(6) 121.5(2) C(33)-C(34)-Cl(6) 118.7(2) C(34)-C(35)-C(26) 120.8(2)

317

C(34)-C(35)-C(36) 132.7(2) C(26)-C(35)-C(36) 106.51(19) C(34A)-C(33A)-C(32) 121.5(3) C(34A)-C(33A)-C(28A) 119.7(2) C(32)-C(33A)-C(28A) 118.8(2) C(35A)-C(34A)-C(33A) 119.7(2) C(35A)-C(34A)-H(34A) 120.2 C(33A)-C(34A)-H(34A) 120.2 C(34A)-C(35A)-C(26A) 120.8(2) C(34A)-C(35A)-C(36) 132.7(2) C(26A)-C(35A)-C(36) 106.51(19) N(6)-C(36)-N(5) 122.7(2) N(6)-C(36)-C(35) 130.6(2) N(5)-C(36)-C(35) 105.51(19) O(1)-C(37)-C(38) 116.0(2) O(1)-C(37)-C(42) 124.6(2) C(38)-C(37)-C(42) 119.4(2) C(39)-C(38)-C(37) 120.5(2) C(39)-C(38)-H(38A) 119.7 C(37)-C(38)-H(38A) 119.7 C(38)-C(39)-C(40) 120.5(2) C(38)-C(39)-H(39A) 119.8 C(40)-C(39)-H(39A) 119.8 C(39)-C(40)-C(41) 119.1(2) C(39)-C(40)-H(40A) 120.5 C(41)-C(40)-H(40A) 120.5 C(40)-C(41)-C(42) 120.8(2) C(40)-C(41)-H(41A) 119.6 C(42)-C(41)-H(41A) 119.6 C(37)-C(42)-C(41) 119.7(2) C(37)-C(42)-H(42A) 120.1 C(41)-C(42)-H(42A) 120.1 Cl(8)-C(1S)-Cl(7) 111.1(8) Cl(8)-C(1S)-H(1S1) 109.4 Cl(7)-C(1S)-H(1S1) 109.4 Cl(8)-C(1S)-H(1S2) 109.4

318

Cl(7)-C(1S)-H(1S2) 109.4 H(1S1)-C(1S)-H(1S2) 108.0 Cl(9)-C(2S)-Cl(10) 113.7(9) Cl(9)-C(2S)-H(2S1) 108.8 Cl(10)-C(2S)-H(2S1) 108.8 Cl(9)-C(2S)-H(2S2) 108.8 Cl(10)-C(2S)-H(2S2) 108.8 H(2S1)-C(2S)-H(2S2) 107.7 ______Symmetry transformations used to generate equivalent atoms:

319

Table S6.4 . Anisotropic displacement parameters (Å 2 x 10 3) for PhO-Cl nBsubNc. The anisotropic displacement factor exponent takes the form: -2p 2[ h 2 a* 2U11 + ... + 2 h k a* b* U12 ] ______U11 U22 U33 U23 U13 U12 ______B(1) 27(1) 26(1) 33(1) 1(1) 21(1) -2(1) Cl(1) 30(2) 47(2) 51(2) -12(1) 18(1) -5(1) Cl(4) 28(2) 48(2) 47(2) 0(1) 22(2) 1(1) Cl(5) 32(2) 50(2) 47(2) 7(1) 27(2) 10(1) Cl(6) 38(2) 55(2) 32(2) -9(1) 17(1) -6(1) O(1) 31(1) 26(1) 40(1) 3(1) 25(1) -2(1) N(1) 28(1) 26(1) 35(1) 2(1) 23(1) -1(1) N(2) 40(1) 29(1) 38(1) 1(1) 29(1) -2(1) N(3) 29(1) 26(1) 31(1) 0(1) 20(1) -3(1) N(4) 29(1) 28(1) 37(1) 2(1) 22(1) -1(1) N(5) 28(1) 28(1) 30(1) 2(1) 19(1) -1(1) N(6) 27(1) 29(1) 34(1) 2(1) 18(1) 0(1) C(1) 27(1) 24(1) 37(1) 4(1) 20(1) 1(1) C(2) 29(1) 23(1) 51(1) 8(1) 27(1) 3(1) C(3) 31(1) 27(1) 64(2) 7(1) 28(1) 4(1) C(4) 34(1) 24(1) 94(2) 20(1) 42(2) 8(1) C(2A) 29(1) 23(1) 51(1) 8(1) 27(1) 3(1) C(3A) 31(1) 27(1) 64(2) 7(1) 28(1) 4(1) C(4A) 34(1) 24(1) 94(2) 20(1) 42(2) 8(1) C(5) 35(2) 27(1) 127(3) 19(2) 48(2) 8(1) C(6) 44(2) 34(2) 184(5) 39(2) 75(3) 14(1) C(7) 75(3) 58(2) 183(5) 54(3) 104(3) 27(2) C(8) 66(2) 54(2) 127(3) 45(2) 80(2) 25(2) C(9) 48(2) 36(1) 86(2) 25(1) 52(2) 14(1) C(10) 49(2) 37(1) 68(2) 17(1) 48(2) 11(1) C(11) 35(1) 28(1) 50(1) 11(1) 32(1) 6(1) C(9A) 48(2) 36(1) 86(2) 25(1) 52(2) 14(1) C(10A) 49(2) 37(1) 68(2) 17(1) 48(2) 11(1)

320

C(11A) 35(1) 28(1) 50(1) 11(1) 32(1) 6(1) C(12) 38(1) 24(1) 42(1) 3(1) 31(1) 2(1) C(13) 39(1) 25(1) 33(1) -2(1) 25(1) -3(1) C(14) 38(1) 25(1) 32(1) -2(1) 21(1) -6(1) C(15) 50(2) 31(1) 35(1) -4(1) 24(1) -9(1) C(16) 52(2) 29(1) 34(1) -3(1) 20(1) -10(1) C(14A) 38(1) 25(1) 32(1) -2(1) 21(1) -6(1) C(15A) 50(2) 31(1) 35(1) -4(1) 24(1) -9(1) C(16A) 52(2) 29(1) 34(1) -3(1) 20(1) -10(1) C(17) 69(2) 45(2) 34(1) -1(1) 21(2) -10(2) C(18) 68(2) 49(2) 37(2) 3(1) 17(2) -6(2) C(19) 50(2) 35(2) 40(2) 5(1) 1(1) -11(1) C(20) 40(2) 31(1) 47(2) 1(1) 10(1) -10(1) C(21) 38(1) 24(1) 36(1) -1(1) 10(1) -10(1) C(22) 32(1) 28(1) 40(1) -2(1) 17(1) -7(1) C(23) 32(1) 25(1) 33(1) -2(1) 17(1) -7(1) C(21A) 38(1) 24(1) 36(1) -1(1) 10(1) -10(1) C(22A) 32(1) 28(1) 40(1) -2(1) 17(1) -7(1) C(23A) 32(1) 25(1) 33(1) -2(1) 17(1) -7(1) C(24) 29(1) 26(1) 35(1) 0(1) 20(1) -4(1) C(25) 28(1) 27(1) 36(1) 5(1) 22(1) 2(1) C(26) 37(1) 26(1) 35(1) 5(1) 26(1) 3(1) C(27) 44(1) 30(1) 47(1) 8(1) 34(1) 6(1) C(28) 62(2) 24(1) 51(2) 7(1) 46(1) 6(1) C(26A) 37(1) 26(1) 35(1) 5(1) 26(1) 3(1) C(27A) 44(1) 30(1) 47(1) 8(1) 34(1) 6(1) C(28A) 62(2) 24(1) 51(2) 7(1) 46(1) 6(1) C(29) 76(2) 31(1) 70(2) 10(1) 61(2) 13(1) C(30) 109(3) 32(2) 77(2) 6(1) 80(2) 11(2) C(31) 109(3) 34(2) 62(2) -5(1) 69(2) -6(2) C(32) 79(2) 32(1) 49(2) -6(1) 47(2) -11(1) C(33) 64(2) 23(1) 45(1) 1(1) 42(1) -4(1) C(34) 46(1) 31(1) 37(1) 2(1) 28(1) -3(1) C(35) 37(1) 25(1) 34(1) 3(1) 25(1) -1(1) C(33A) 64(2) 23(1) 45(1) 1(1) 42(1) -4(1) C(34A) 46(1) 31(1) 37(1) 2(1) 28(1) -3(1)

321

C(35A) 37(1) 25(1) 34(1) 3(1) 25(1) -1(1) C(36) 31(1) 26(1) 30(1) 3(1) 19(1) 0(1) C(37) 31(1) 28(1) 25(1) 0(1) 19(1) -5(1) C(38) 39(1) 32(1) 39(1) 5(1) 29(1) 2(1) C(39) 49(2) 29(1) 50(2) 6(1) 35(1) 1(1) C(40) 38(1) 36(1) 40(1) 2(1) 26(1) -8(1) C(41) 29(1) 42(1) 40(1) 7(1) 22(1) -2(1) C(42) 33(1) 31(1) 39(1) 6(1) 23(1) 2(1) Cl(7) 395(18) 700(30) 263(15) 318(19) 127(16) 30(20) Cl(8) 183(6) 473(14) 469(14) 331(12) 246(8) 136(7) C(1S) 70(7) 173(12) 158(12) 55(10) 59(8) 14(8) Cl(9) 32(6) 151(13) 56(7) -3(7) 22(5) -8(6) Cl(10) 32(6) 151(13) 56(7) -3(7) 22(5) -8(6) ______

322

Table S6.5 . Hydrogen coordinates (x 10 4) and isotropic displacement parameters (Å 2 x 10 3) for

PhO-Cl nBsubNc. ______x y z U(eq) ______

H(3A) 4759 6233 5608 51 H(5) 5666 6741 5836 80 H(6A) 6149 6924 5280 104 H(7A) 5684 6466 3837 108 H(8A) 4739 5850 2939 84 H(10A) 3829 5356 2689 53 H(15A) 1540 5246 381 49 H(17A) 678 5701 -1160 70 H(18A) -272 6295 -2171 77 H(19A) -815 6757 -1751 75 H(20A) -429 6564 -320 63 H(22A) 412 6062 1221 46 H(27A) 1413 6535 4506 44 H(29A) 1366 7243 5604 58 H(30A) 1849 7776 7006 70 H(31A) 2893 7875 8034 70 H(32A) 3457 7397 7674 59 H(34A) 3525 6642 6617 44 H(38A) 2929 1938 4284 41 H(39A) 2357 730 4198 48 H(40A) 1427 1060 3785 45 H(41A) 1069 2609 3435 45 H(42A) 1627 3819 3478 41 H(1S1) 4178 7523 6747 174 H(1S2) 4535 8215 7577 174 H(2S1) 5440 7824 7898 14 H(2S2) 5659 7025 8614 14 ______

323

Table S6.6 . Torsion angles [°] for PhO-Cl nBsubNc. ______N(1)-B(1)-O(1)-C(37) -175.47(19) N(3)-B(1)-O(1)-C(37) -57.9(3) N(5)-B(1)-O(1)-C(37) 68.2(3) O(1)-B(1)-N(1)-C(1) -94.3(2) N(3)-B(1)-N(1)-C(1) 139.2(2) N(5)-B(1)-N(1)-C(1) 31.4(3) O(1)-B(1)-N(1)-C(12) 93.3(2) N(3)-B(1)-N(1)-C(12) -33.3(3) N(5)-B(1)-N(1)-C(12) -141.1(2) O(1)-B(1)-N(3)-C(24) 107.2(2) N(1)-B(1)-N(3)-C(24) -133.2(2) N(5)-B(1)-N(3)-C(24) -25.6(3) O(1)-B(1)-N(3)-C(13) -88.0(3) N(1)-B(1)-N(3)-C(13) 31.6(3) N(5)-B(1)-N(3)-C(13) 139.21(19) O(1)-B(1)-N(5)-C(25) -103.4(2) N(1)-B(1)-N(5)-C(25) 137.8(2) N(3)-B(1)-N(5)-C(25) 29.6(3) O(1)-B(1)-N(5)-C(36) 85.8(3) N(1)-B(1)-N(5)-C(36) -33.0(3) N(3)-B(1)-N(5)-C(36) -141.23(19) C(36)-N(6)-C(1)-N(1) -8.7(3) C(36)-N(6)-C(1)-C(2) 157.9(2) C(12)-N(1)-C(1)-N(6) 159.6(2) B(1)-N(1)-C(1)-N(6) -13.3(3) C(12)-N(1)-C(1)-C(2) -10.0(2) B(1)-N(1)-C(1)-C(2) 177.09(19) N(6)-C(1)-C(2)-C(3) 13.9(4) N(1)-C(1)-C(2)-C(3) -177.7(2) N(6)-C(1)-C(2)-C(11) -163.7(2) N(1)-C(1)-C(2)-C(11) 4.6(2) C(11)-C(2)-C(3)-C(4) -2.0(3) C(1)-C(2)-C(3)-C(4) -179.3(2)

324

C(11)-C(2)-C(3)-Cl(1) 173.1(2) C(1)-C(2)-C(3)-Cl(1) -4.2(4) C(2)-C(3)-C(4)-C(9) 1.2(4) Cl(1)-C(3)-C(4)-C(9) -174.3(2) C(2)-C(3)-C(4)-C(5) -178.2(2) Cl(1)-C(3)-C(4)-C(5) 6.4(3) C(11A)-C(2A)-C(3A)-C(4A) -2.0(3) C(2A)-C(3A)-C(4A)-C(9A) 1.2(4) C(3)-C(4)-C(5)-C(6) 179.3(2) C(9)-C(4)-C(5)-C(6) 0.0(4) C(4)-C(5)-C(6)-C(7) -0.4(4) C(5)-C(6)-C(7)-C(8) 0.5(5) C(6)-C(7)-C(8)-C(9) -0.1(5) C(7)-C(8)-C(9)-C(10) -179.8(3) C(7)-C(8)-C(9)-C(4) -0.3(4) C(3)-C(4)-C(9)-C(10) 0.5(4) C(5)-C(4)-C(9)-C(10) 179.9(2) C(3)-C(4)-C(9)-C(8) -179.0(2) C(5)-C(4)-C(9)-C(8) 0.3(4) C(8)-C(9)-C(10)-C(11) 178.1(2) C(4)-C(9)-C(10)-C(11) -1.4(4) C(8)-C(9)-C(10)-Cl(2) 15.1(6) C(4)-C(9)-C(10)-Cl(2) -164.4(5) C(9)-C(10)-C(11)-C(2) 0.6(4) Cl(2)-C(10)-C(11)-C(2) 160.1(6) C(9)-C(10)-C(11)-C(12) 177.4(2) Cl(2)-C(10)-C(11)-C(12) -23.0(7) C(3)-C(2)-C(11)-C(10) 1.1(4) C(1)-C(2)-C(11)-C(10) 179.1(2) C(3)-C(2)-C(11)-C(12) -176.4(2) C(1)-C(2)-C(11)-C(12) 1.6(2) C(3A)-C(4A)-C(9A)-C(10A) 0.5(4) C(4A)-C(9A)-C(10A)-C(11A) -1.4(4) C(9A)-C(10A)-C(11A)-C(2A) 0.6(4) C(3A)-C(2A)-C(11A)-C(10A) 1.1(4) C(13)-N(2)-C(12)-N(1) 7.9(3)

325

C(13)-N(2)-C(12)-C(11) -156.1(2) C(1)-N(1)-C(12)-N(2) -156.6(2) B(1)-N(1)-C(12)-N(2) 16.4(3) C(1)-N(1)-C(12)-C(11) 10.9(3) B(1)-N(1)-C(12)-C(11) -176.04(19) C(10)-C(11)-C(12)-N(2) -18.3(4) C(2)-C(11)-C(12)-N(2) 158.9(2) C(10)-C(11)-C(12)-N(1) 175.7(2) C(2)-C(11)-C(12)-N(1) -7.2(2) C(12)-N(2)-C(13)-N(3) -9.4(3) C(12)-N(2)-C(13)-C(14) 152.9(2) C(24)-N(3)-C(13)-N(2) 153.1(2) B(1)-N(3)-C(13)-N(2) -13.2(3) C(24)-N(3)-C(13)-C(14) -12.9(2) B(1)-N(3)-C(13)-C(14) -179.09(19) N(2)-C(13)-C(14)-C(15) 19.8(4) N(3)-C(13)-C(14)-C(15) -175.6(2) N(2)-C(13)-C(14)-C(23) -156.0(2) N(3)-C(13)-C(14)-C(23) 8.7(2) C(16)-Cl(3)-C(15)-C(14) -178.2(7) C(23)-C(14)-C(15)-Cl(3) 177.4(6) C(13)-C(14)-C(15)-Cl(3) 2.1(7) C(23)-C(14)-C(15)-C(16) -0.6(4) C(13)-C(14)-C(15)-C(16) -175.9(2) Cl(3)-C(15)-C(16)-C(21) -176.6(5) C(14)-C(15)-C(16)-C(21) 2.0(4) Cl(3)-C(15)-C(16)-C(17) 3.3(5) C(14)-C(15)-C(16)-C(17) -178.2(2) C(14)-C(15)-C(16)-Cl(3) 178.5(5) C(23A)-C(14A)-C(15A)-C(16A) -0.6(4) C(14A)-C(15A)-C(16A)-C(21A) 2.0(4) C(14A)-C(15A)-C(16A)-C(17) -178.2(2) C(15)-C(16)-C(17)-C(18) -179.2(3) C(21)-C(16)-C(17)-C(18) 0.6(4) Cl(3)-C(16)-C(17)-C(18) -177.3(4) C(15A)-C(16A)-C(17)-C(18) -179.2(3)

326

C(21A)-C(16A)-C(17)-C(18) 0.6(4) C(16)-C(17)-C(18)-C(19) 0.7(5) C(16A)-C(17)-C(18)-C(19) 0.7(5) C(17)-C(18)-C(19)-C(20) -1.5(5) C(18)-C(19)-C(20)-C(21A) 0.9(4) C(18)-C(19)-C(20)-C(21) 0.9(4) C(19)-C(20)-C(21)-C(22) -179.7(2) C(19)-C(20)-C(21)-C(16) 0.4(4) C(15)-C(16)-C(21)-C(22) -1.2(4) C(17)-C(16)-C(21)-C(22) 179.0(2) Cl(3)-C(16)-C(21)-C(22) -6.0(8) C(15)-C(16)-C(21)-C(20) 178.7(2) C(17)-C(16)-C(21)-C(20) -1.2(4) Cl(3)-C(16)-C(21)-C(20) 173.9(7) C(20)-C(21)-C(22)-C(23) 179.2(2) C(16)-C(21)-C(22)-C(23) -0.9(3) C(20)-C(21)-C(22)-Cl(4) -5.4(3) C(16)-C(21)-C(22)-Cl(4) 174.5(2) C(21)-C(22)-C(23)-C(14) 2.3(3) Cl(4)-C(22)-C(23)-C(14) -172.9(2) C(21)-C(22)-C(23)-C(24) 178.2(2) Cl(4)-C(22)-C(23)-C(24) 3.0(4) C(15)-C(14)-C(23)-C(22) -1.5(3) C(13)-C(14)-C(23)-C(22) 174.8(2) C(15)-C(14)-C(23)-C(24) -178.4(2) C(13)-C(14)-C(23)-C(24) -2.1(2) C(19)-C(20)-C(21A)-C(22A) -179.7(2) C(19)-C(20)-C(21A)-C(16A) 0.4(4) C(15A)-C(16A)-C(21A)-C(22A) -1.2(4) C(17)-C(16A)-C(21A)-C(22A) 179.0(2) C(15A)-C(16A)-C(21A)-C(20) 178.7(2) C(17)-C(16A)-C(21A)-C(20) -1.2(4) C(20)-C(21A)-C(22A)-C(23A) 179.2(2) C(16A)-C(21A)-C(22A)-C(23A) -0.9(3) C(21A)-C(22A)-C(23A)-C(14A) 2.3(3) C(15A)-C(14A)-C(23A)-C(22A) -1.5(3)

327

C(25)-N(4)-C(24)-N(3) 9.3(3) C(25)-N(4)-C(24)-C(23) -156.3(2) C(13)-N(3)-C(24)-N(4) -157.1(2) B(1)-N(3)-C(24)-N(4) 8.8(3) C(13)-N(3)-C(24)-C(23) 11.6(2) B(1)-N(3)-C(24)-C(23) 177.49(19) C(22)-C(23)-C(24)-N(4) -14.2(4) C(14)-C(23)-C(24)-N(4) 162.2(2) C(22)-C(23)-C(24)-N(3) 178.4(2) C(14)-C(23)-C(24)-N(3) -5.3(2) C(24)-N(4)-C(25)-N(5) -5.4(3) C(24)-N(4)-C(25)-C(26) 160.8(2) C(36)-N(5)-C(25)-N(4) 154.7(2) B(1)-N(5)-C(25)-N(4) -16.8(3) C(36)-N(5)-C(25)-C(26) -14.6(2) B(1)-N(5)-C(25)-C(26) 173.88(19) N(4)-C(25)-C(26)-C(27) 21.5(4) N(5)-C(25)-C(26)-C(27) -170.6(2) N(4)-C(25)-C(26)-C(35) -159.5(2) N(5)-C(25)-C(26)-C(35) 8.4(2) C(35)-C(26)-C(27)-C(28) -3.0(3) C(25)-C(26)-C(27)-C(28) 175.9(2) C(35)-C(26)-C(27)-Cl(5) 177.1(2) C(25)-C(26)-C(27)-Cl(5) -4.0(4) C(26)-C(27)-C(28)-C(29) -173.6(2) Cl(5)-C(27)-C(28)-C(29) 6.3(3) C(26)-C(27)-C(28)-C(33) 3.9(3) Cl(5)-C(27)-C(28)-C(33) -176.2(2) C(35A)-C(26A)-C(27A)-C(28A) -3.0(3) C(26A)-C(27A)-C(28A)-C(29) -173.6(2) C(26A)-C(27A)-C(28A)-C(33A) 3.9(3) C(27)-C(28)-C(29)-C(30) 176.2(2) C(33)-C(28)-C(29)-C(30) -1.3(4) C(27A)-C(28A)-C(29)-C(30) 176.2(2) C(33A)-C(28A)-C(29)-C(30) -1.3(4) C(28)-C(29)-C(30)-C(31) 2.4(4)

328

C(28A)-C(29)-C(30)-C(31) 2.4(4) C(29)-C(30)-C(31)-C(32) -1.1(4) C(30)-C(31)-C(32)-C(33A) -1.3(4) C(30)-C(31)-C(32)-C(33) -1.3(4) C(31)-C(32)-C(33)-C(34) -175.1(2) C(31)-C(32)-C(33)-C(28) 2.4(4) C(27)-C(28)-C(33)-C(34) -1.0(3) C(29)-C(28)-C(33)-C(34) 176.5(2) C(27)-C(28)-C(33)-C(32) -178.6(2) C(29)-C(28)-C(33)-C(32) -1.1(3) C(32)-C(33)-C(34)-C(35) 174.8(2) C(28)-C(33)-C(34)-C(35) -2.7(3) C(32)-C(33)-C(34)-Cl(6) -1.7(3) C(28)-C(33)-C(34)-Cl(6) -179.2(2) C(33)-C(34)-C(35)-C(26) 3.7(3) Cl(6)-C(34)-C(35)-C(26) -180.0(2) C(33)-C(34)-C(35)-C(36) -176.0(2) Cl(6)-C(34)-C(35)-C(36) 0.4(4) C(27)-C(26)-C(35)-C(34) -0.8(3) C(25)-C(26)-C(35)-C(34) -179.9(2) C(27)-C(26)-C(35)-C(36) 179.0(2) C(25)-C(26)-C(35)-C(36) -0.2(2) C(31)-C(32)-C(33A)-C(34A) -175.1(2) C(31)-C(32)-C(33A)-C(28A) 2.4(4) C(27A)-C(28A)-C(33A)-C(34A) -1.0(3) C(29)-C(28A)-C(33A)-C(34A) 176.5(2) C(27A)-C(28A)-C(33A)-C(32) -178.6(2) C(29)-C(28A)-C(33A)-C(32) -1.1(3) C(32)-C(33A)-C(34A)-C(35A) 174.8(2) C(28A)-C(33A)-C(34A)-C(35A) -2.7(3) C(33A)-C(34A)-C(35A)-C(26A) 3.7(3) C(33A)-C(34A)-C(35A)-C(36) -176.0(2) C(27A)-C(26A)-C(35A)-C(34A) -0.8(3) C(27A)-C(26A)-C(35A)-C(36) 179.0(2) C(1)-N(6)-C(36)-N(5) 6.7(3) C(1)-N(6)-C(36)-C(35) -159.3(2)

329

C(25)-N(5)-C(36)-N(6) -154.5(2) B(1)-N(5)-C(36)-N(6) 17.2(3) C(25)-N(5)-C(36)-C(35) 14.5(2) B(1)-N(5)-C(36)-C(35) -173.82(19) C(34)-C(35)-C(36)-N(6) -20.7(4) C(26)-C(35)-C(36)-N(6) 159.6(2) C(34)-C(35)-C(36)-N(5) 171.6(2) C(26)-C(35)-C(36)-N(5) -8.1(2) B(1)-O(1)-C(37)-C(38) 176.0(2) B(1)-O(1)-C(37)-C(42) -4.7(4) O(1)-C(37)-C(38)-C(39) 179.9(2) C(42)-C(37)-C(38)-C(39) 0.6(4) C(37)-C(38)-C(39)-C(40) -1.1(4) C(38)-C(39)-C(40)-C(41) 0.6(4) C(39)-C(40)-C(41)-C(42) 0.4(4) O(1)-C(37)-C(42)-C(41) -178.8(2) C(38)-C(37)-C(42)-C(41) 0.5(3) C(40)-C(41)-C(42)-C(37) -1.0(4) ______Symmetry transformations used to generate equivalent atoms:

330

Differential Pulse Voltammetry

(a)

(b)

Figure S6.4 . Differential pulse voltammograms of (a) PhO-Cl nBsubNc and (b) F 5-Cl nBsubNc in DCM with 0.1 M tetrabutylammonium perchlorate as the supporting electrolyte and decamethylferrocene as the internal reference at room temperature.

Table S6.7 . DPV characteristics of phenoxy Cl nBsubNcs.

1' 2' a 1' a Compound E ox | E ox (V) E red (V) PhO-Cl nBsubNc +0.78 | +1.25 -1.01 F5PhO-Cl nBsubNc +0.82 | +1.26 -0.96 E' = redox potential from DPV. a In degassed DCM solution relative to Ag/AgCl.

331

Determination of Fluorescence Quantum Yields The fluorescence quantum yields (Φ) were calculated using the formula below:

2 2 Φ = Φ R (I / I R)(OD R / OD)(n / n R ) (Eq S6.1) where I is the integrated fluorescence intensity, OD is the optical density ( i.e. absorbance), and n is the refractive index of the solvent. The subscript R is oxazine 170, a reference fluorophore, which has previously been reported to have a Φ = 0.579 2 in ethanol at room temperature. Integrated fluorescence intensity values were acquired from PerkinElmer FL WinLab (version 4.00.03) while the optical density values were acquired from PerkinElmer UV WinLab (version 6.02.0723). Analyses of all BsubNc compounds were acquired in toluene at room temperature.

332

NMR Spectra

1 Figure S6.5 . H NMR (400 MHz, CDCl 3) spectrum obtained at 296 K for PhO-Cl nBsubNc.

333

1 Figure S6.6 . H NMR (400 MHz, CDCl 3) spectrum obtained at 296 K for F 5-Cl nBsubNc.

334

References 1. Bouvet, M.; Bassoul, P.; Simon, J. Molecular Crystals and Liquid Crystals Science and Technology. Section A. Molecular Crystals and Liquid Crystals 1994 , 252 , 31-38. 2. Rurack, K.; Spieles, M. Analytical Chemistry 2011 , 83 , 1232-1242.

335

Appendix F: Supplementary Information for Chapter 7

Synthetic Procedures 1,3-Bis(2-pyridylimino)isoindoline (BPI ). BPI is a known compound and was prepared according to literature procedure. 1 To a three-neck round bottom flask equipped with a condenser and a gas inlet was added o-phthalonitrile (5.00 g, 0.039 mol, 1 equiv), n-hexanol (120 mL), 2- aminopyridine (8.50 g, 0.090 mol, 2.3 equiv), and calcium chloride (2.15 g, 0.019 mol, 0.5 equiv) under argon. The reaction was heated to reflux (~155 °C) and monitored by TLC (silica, dichloromethane/ethyl acetate - 3:2) until all of the o-phthalonitrile had reacted (~12 hours). Cooling of the reaction mixture to room temperature allowed for the product to precipitate out of solution. The solid was isolated via vacuum filtration and washed with hexane (3 x 100 mL) and water (3 x 100 mL). The crude product was dissolved in a minimum volume of dichloromethane and filtered through a short plug of celite. The filtrate was concentrated to dryness via rotary evaporation and dried in a vacuum oven at 80 °C overnight to give BPI (7.71 g, 66%) as a bright

1 yellow/green solid. H NMR (400 MHz, DMSO-d6) d 8.74-8.70 (m, 2H), 8.04 (dd, J = 3.0 Hz, J = 5 Hz, 2H), 7.92 (td, J = 2 Hz, J = 7.7 Hz, 2H), 7.77 (dd, J = 3.0 Hz, J = 5.6 Hz, 2H), 7.48 (d, J 13 = 8.0 Hz, 2H), 7.31-7.26 (m, 2H); C NMR (100 MHz, DMSO-d6) δ 159.6, 152.6, 148.2, 138.8, 135.0, 132.2, 123.1, 122.5, 120.8.

BPI·BF 2. A three-neck round bottom flask equipped with a condenser and a gas inlet was purged with argon for 15 minutes. To the reaction flask was added BPI (0.50 g, 1.67 mmol, 1 equiv), anhydrous toluene (15 mL), triethylamine (0.19 g, 0.26 mL, 1.84 mmol, 1.1 equiv), and boron trifluoride diethyl etherate (0.26 g, 0.20 mL, 1.84 mmol, 1.1 equiv). The reaction was heated to 100 °C under argon for 24 hours before it was allowed to cool down to room temperature. The reaction mixture was suction filtered and the isolated solid was washed with warm toluene (50 °C, 5 x 30 mL), dichloromethane (3 x 10 mL), and water (3 x 5 mL). The solid was dried in a vacuum oven at 80 °C overnight to afford BPI·BF 2 as a bright yellow solid in yields up to 19%. Slow evaporation from dichloromethane produced single crystals suitable for X-ray diffraction

336

1 analysis. H NMR (400 MHz, DMSO-d6) d 8.65 (dd, J = 5.8 Hz 1H), 8.51-8.44 (m, 1H), 8.41- 8.34 (m, 1H), 8.06 (d, J = 7.7 Hz, 1H), 7.90 (td, J = 1.8 Hz, J = 7.7 Hz, 1H), 7.76 (d, J = 8.2 Hz, 1H), 7.72 (td, J = 7.6 Hz, 1H), 7.68-7.62 (m, 1H), 7.52 (t, J = 7.6 Hz, 1H), 7.33-7.26 (m, 1H), 13 7.10 (d, J = 8.0 Hz, 1H), 6.38 (d, J = 7.7 Hz, 1H); C NMR (100 MHz, DMSO-d6) δ 160.2, 159.9, 154.8, 153.4, 148.8, 144.9, 140.2, 138.6, 135.2, 133.2, 132.7, 128.6, 125.3, 124.4, 123.3, 121.6, 120.3, 115.1; HRMS (DART) m/z [M+H] + calcd for 348.12321, found 348.12326.

BPI·AlCl 2. Complex BPI·AlCl 2 was obtained as a bright yellow solid in yields up to 25% via the same procedure used for the synthesis of BPI·BF 2, but aluminum chloride was used as the metal

1 complex. H NMR (400 MHz, DMSO-d6) d 8.14 (dd, J = 1.7 Hz, J = 6.0 Hz, 2H), 8.09 (dd, J = 3.0 Hz, J = 5.5 Hz, 2H), 7.85-7.79 (m, 4H), 7.47 (d, J = 8.2 Hz, 2H), 7.44 (s, 1H), 6.84-6.76 (t, 13 2H); C NMR (100 MHz, DMSO-d6) δ 162.8, 156.8, 146.3, 141.5, 137.9, 132.0, 125.7, 122.5, 120.0; HRMS (DART) m/z [M+H] + calcd for 396.03634, found 396.03755.

BPI·GaCl 2. Complex BPI·GaCl 2 was obtained as a bright yellow solid in yields up to 17% via the same procedure used for the synthesis of BPI·BF2, but gallium(III) chloride was used as the metal complex. Slow evaporation from dichloromethane produced single crystals suitable for X- ray diffraction analysis. Alternative method: A three-neck round bottom flask equipped with a condenser and a gas inlet was purged with argon for 15 minutes. To the reaction flask was added o-phthalonitrile (0.15 g, 1.14 mmol, 1 equiv), n-hexanol (5 mL), 2-aminopyridine (.22 g, 2.39 mmol, 2.1 equiv), and a chlorobenzene solution (5 mL) of gallium(III) chloride (0.22 g, 1.25 mmol, 1.1 equiv). The reaction was heated to reflux (~155 °C) and monitored by TLC (silica, dichloromethane/ethyl acetate - 3:2) until all of the o-phthalonitrile had reacted (~ 2 days). Cooling of the reaction mixture to room temperature allowed for the product to precipitate out of solution. The solid was isolated via vacuum filtration and washed with hexane (3 x 100 mL), warm toluene (50 °C, 5 x 20 mL), dichloromethane (3 x 10 mL), and water (3 x 5 mL). The solid was dried in a vacuum oven at 80 °C overnight to give BPI·GaCl 2 (46 mg, 9%) as a bright yellow solid. Some crystals suitable for single crystal X-ray diffraction study were formed during cooling of the reaction

337

1 mixture to room temperature. H NMR (400 MHz, DMSO-d6) d 8.83 (d, J = 5.5 Hz, 2H), 8.20 (t, J = 7.7 Hz, 2H), 8.05 (dd, J = 3 Hz, J = 5.5 Hz, 2H), 7.81 (dd, J = 3.0 Hz, J = 5.6 Hz, 2H), 7.64 13 (d, J = 8.0 Hz, 2H), 7.58 (t, J = 6.6 Hz, 2H); C NMR (100 MHz, DMSO-d6) δ 159.6, 152.6, 148.2, 138.8, 135.0, 132.3, 123.1, 122.4, 120.8; HRMS (DART) m/z [M+H] + calcd for 437.98038, found 437.98091.

BPI·InCl 2. Complex BPI·InCl 2 was obtained as a bright yellow solid in yields up to 27% via the same procedure used for the synthesis of BPI·BF 2, but indium(III) chloride was used as the metal 1 complex. H NMR (400 MHz, DMSO-d6) d 8.72 (d, J = 4.5 Hz, 2H), 8.03 (dd, J = 3.1 Hz, J = 5.5 Hz, 2H), 7.92 (td, J = 1.9 Hz, J = 7.8 Hz, 2H), 7.76 (dd, J = 3.1 Hz, J = 5.6 Hz, 2H), 7.45 (d, 13 J = 8.0 Hz, 2H), 7.32-7.25 (m, 2H); C NMR (100 MHz, DMSO-d6) δ 159.6, 152.5, 148.1, 138.8, 135.0, 132.2, 123.0, 122.4, 120.8; HRMS (DART) m/z [M+H] + calcd for 483.95868, found 483.95882.

338

NMR Spectra

1 Figure S7.1 . H NMR (400 MHz, DMSO-d6) spectrum obtained at 296 K for compound

BPI·BF 2.

339

N Cl N Al Cl N

1 Figure S7.2 . H NMR (400 MHz, DMSO-d6) spectrum obtained at 296 K for compound

BPI·AlCl 2.

340

N Cl N Ga Cl N

1 Figure S7.3 . H NMR (400 MHz, DMSO-d6) spectrum obtained at 296 K for compound

BPI·GaCl 2.

341

N Cl N In Cl N

1 Figure S7.4 . H NMR (400 MHz, DMSO-d6) spectrum obtained at 296 K for compound

BPI·InCl 2.

342

13 Figure S7.5 . C NMR (100 MHz, DMSO-d6) spectrum obtained at 296 K for compound

BPI·BF 2.

343

N Cl N Al Cl N

13 Figure S7.6 . C NMR (100 MHz, DMSO-d6) spectrum obtained at 296 K for compound

BPI·AlCl 2.

344

13 Figure S7.7 . C NMR (100 MHz, DMSO-d6) spectrum obtained at 296 K for compound

BPI·GaCl 2.

345

13 Figure S7.8 . C NMR (100 MHz, DMSO-d6) spectrum obtained at 296 K for compound

BPI·InCl 2.

346

Fluorescence Emission Data

Table S7.1 . Observed λ max of emission and intensity at three excitation wavelengths (λ ex ) for

BPI, BPI·BF 2, BPI·AlCl 2, BPI·GaCl 2, and BPI·InCl 2. a Compound λex λmax of Emission Emission Intensity BPI 352 398 12.6 386 436 5.7 408 466 2.0 BPI·BF 2 352 399 13.9 370 418 10.8 388 439 5.4 BPI·AlCl 2 301 367 637.4 312 368 533.2 352 399 16.4 BPI·GaCl 2 352 396 13.7 408 467 2.5 434 503 2.1 BPI·InCl 2 352 398 14.2 411 470 2.2 437 506 2.2 a The maximum absorbance values of each analyte was kept in the range of 0.04 and 0.05; concentrations for each analyte were not the same. Emission intensity values are reported here for qualitative comparison purposes.

347

Determination of Fluorescence Quantum Yields The fluorescence quantum yields (φ) were calculated using the formula below:

2 2 φ = φ R (I / I R)(OD R / OD)(n / n R ) (Eq S7.1) where I is the integrated fluorescence intensity, OD is the optical density ( i.e. absorbance), and n is the refractive index of the solvent. The subscript R is 9,10-diphenylanthracence, a reference fluorophore, which has previously been reported to have a φ = 0.97 in degassed cyclohexane at room temperature. 2 Integrated fluorescence intensity values were acquired from PerkinElmer FL WinLab (version 4.00.03) while the optical density values were acquired from PerkinElmer UV WinLab (version 6.02.0723). Analyses of the BPI ligand and its group XIII complexes were acquired in degassed dichloromethane at room temperature.

Table S7.2 . Calculated fluorescence quantum yields (ɸ) for BPI and its group XIII complexes.

BPI BPI ·BF 2 BPI ·GaCl 2 BPI ·InCl 2 ɸ (%) 0.8 2.7 0.9 0.7

348

References 1. Siegl, W. O. The Journal of Organic Chemistry 1977 , 42 , 1872-1878. 2. Meech, S. R.; Phillips, D. Journal of Photochemistry 1983 , 23 , 193-217.

349

Appendix G: Supplementary Information for Chapter 8

Synthetic Procedures o-Cyanoallocinnamic acid . This compound was prepared according to a modified literature procedure. 1,2 In a round-bottom flask, 1-nitroso-2-naphthol (20.0 g, 0.115 mol, 1 equiv) was dissolved in tetrahydrofuran (100 mL). p-Toluenesulfonyl chloride (26.4 g, 0.139 mol, 1.2 equiv) was added and the reaction mixture was stirred for 20 minutes. An aqueous solution of NaOH (3M, 0.403 mol, 3.5 equiv) was added dropwise at 0 ˚C. The mixture was stirred at room temperature for 18 hours and then concentrated via rotary evaporation to remove the tetrahydrofuran. The resulting aqueous solution was treated with a spatula tip of charcoal before it was gravity filtered. The aqueous filtrate was acidified at 0 ˚C with HCl (conc) dropwise until litmus paper turned red. The precipitate was collected via suction filtration and washed with cold water (3 x 30 mL). The crude product was purified via recrystallization from hot water (~600 mL) using charcoal to give fluffy white solids (15.0 g, 75%).

Isoindolinone-3-acetic acid . This compound was prepared according to a modified literature procedure. 1 In a round-bottom flask equipped with a condenser, o-cyanoallocinnamic acid (15.0 g, 0.087 mol, 1 equiv) was dissolved in an aqueous solution of NaOH (3M, 0.217 mol, 2.5 equiv) and heated to reflux for 3 hours. The mixture was cooled to 0 ˚C before HCl (conc) was added dropwise until litmus paper turned red. The precipitate was collected via suction filtration and wash with cold water (3 x 30 mL). The crude product was purified via recrystallization from hot water (~300 mL) using charcoal to produce white powder (14.0 g, 85%).

Bromo boron subphthalocyanine (Br-BsubPc) . Br-BsubPc was prepared according to literature procedure. 3

Methoxy boron subphthalocyanine (MeO-BsubPc) . To a round-bottom flask equipped with a condenser was added Br-BsubPc (5.0 g, 10.5 mol) and methanol (250 mL). The mixture was

350

heated to reflux and its progress was monitored by HPLC. Once all of the Br-BsubPc had reacted (within 12 hours), the reaction mixture was cooled to room temperature and concentrated to dryness via rotary evaporation. The crude product was purified via train sublimation, whereby the apparatus was operated under a vacuum with a controlled flow of carbon dioxide gas (~100 mTorr above pump down pressure). The best thermal profile in terms of yield and purity is

1 summarized in Method 4.2 (Table S8.4). H NMR (400 MHz, CDCl 3) d 8.88-8.83 (m, 6H), 7.92-

13 7.87 (m, 6H), 1.50 (s, 3H); C NMR (100 MHz, CDCl 3) d 151.7, 131.1, 129.8, 122.2, 46.9.

Attempted Synthetic Procedures for n-BsubPy: Methods 1.1 to 1.8 (Table S8.1) . In a general reaction, isoindolinone-3-acetic acid (300 mg, 1.57 mmol, 3.1 equiv) and a solvent (15 mL) were added to a round-bottom flask under an atmosphere of nitrogen. The boron template (1 equiv) was next added, following by the dropwise addition of a base (10.2 equiv). The reaction mixture was stirred at room temperature for Method 1.1 and heated to reflux for all other reactions. The reaction progress was monitored by HPLC.

Table S8.1 . Attempted syntheses of n-BsubPy via the aldol-promotion approach. Method Starting Material Boron Base Solvent Temperature Formation Template (˚C) (%) a 1.1 B(OEt) 3 NaOEt THF (anh.) r.t. 0 1.2 B(OEt) 3 NaOEt THF (anh.) 66 0 t 1.3 B(OEt) 3 NaO Bu THF (anh.) 66 0 1.4 B(OH) 3 NaOEt THF (anh.) 66 0 1.5 B(OH) 3 NaOEt DMF (dist.) 152 0 t 1.6 B(OH) 3 NaO Bu THF (anh.) 66 0 t 1.7 B(OH) 3 NaO Bu diglyme 162 0 1.8 B(OH) 3 n-BuLi THF (anh.) 66 0 a As determined by HPLC-UV/Vis (MAXPLOT) for BsubPy. B(OEt) 3 = triethyl borate; NaOEt = sodium ethoxide; NaO tBu = sodium tert -butoxide; n-BuLi = n-butyllithium; THF = tetrahydrofuran; EtOH = ethanol; diglyme = diethylene glycol dimethyl ether.

Methods 2.1 to 2.10 (Table S8.2) . For 2.1 , isoindolinone-3-acetic acid (3.00 g, 15.7 mmol, 1 equiv), boric acid (0.97 g, 15.7 mmol, 1 equiv), and dimethyl sulfone (7 mL) were added to a 3- neck round-bottom flask equipped with a short path condenser under an atmosphere of argon. The reaction mixture was heated and the reaction progress was monitored by HPLC. For 2.2 ,

351

sulfolane was used instead. For 2.3 and 2.4 , the volume of sulfolane were halved and doubled, respectively. For 2.5 , the reaction in 2.2 was scaled up by a factor of 3. For 2.6 , a solvent was not used. For 2.7 -2.10 , diphenyl sulfone, phenyl ether, 1,2-dichlorobenzene, and 1,2,4- trichlorobenzene were used as the solvent, respectively.

Work-up Procedure . Once the reaction progress had stagnated as determined via HPLC, the reaction was allowed to cool to room temperature to produce a black solid mass. Water was added to nearly fill the flask and the mixture was stirred and heated at 50 °C overnight. The suspension was gravity filtered and the solid was rinsed with an excess volume of water. A Soxhlet extraction using toluene was carried out to extract out a black organic-soluble solid after rotary evaporation. The black solid was dry loaded onto silica gel using dichloromethane and then loaded onto fresh silica gel in a glass thimble with a fritted disk (medium pore size) at the bottom. The thimble was placed inside of a Soxhlet extraction apparatus and was initially extracted with dichloromethane to remove a side product with a purple fluorescence when irradiated with a UV lamp. Once this fluorescence was no longer observed and the filtrate coming out of the thimble was clear, a switch to methanol was made to elute out the desired product. This process was monitored by a UV light source for its characteristic green fluorescence. Once this was collected, the filtrate was concentrated to produce green metallic- like solids. The product was further purified via train sublimation (see Table S8.5).

352

Table S8.2 . Attempted syntheses of HO-n-BsubPy via the solvent-assisted approach. Method Concentration (mol/L) a | Solvent Temperature Reaction Time Wt % Relative Scale-Up b (˚C) (h) (Before | After) c 2.1 5.7 |1 DMS 230 168 - 2.2 5.7 | 1 SULF 280 168 8.5 | 28 2.3 2.9 | 1 SULF 280 168 - 2.4 11.4 | 1 SULF 280 48 - 2.5 5.7 | 3 SULF 280 168 8.8 | 31 2.6 Neat | 1 - 280 3 - 2.7 5.7 |1 PE 260 19 - 2.8 5.7 |1 DCB 180 46 - 2.9 5.7 |1 TCB 215 44 - a Calculated with consideration for the sum of the two starting materials. b Relative to the experiment in Method 2.1/2.2. c As determined optically using Beer-Lambert Law before and after a full work-up. DMS = dimethyl sulfone; SUFL = sulfolane; DPS = diphenyl sulfone; PE = phenyl ether; DCB = 1,2-dichlorobenzene; TCB = 1,2,4- trichlorobenzene.

Attempted Synthetic Procedures for mPh-BsubPy: Methods 3.1 to 3.14 (Table S8.3) . For 3.1 , phthalimide (3.00 g, 20.4 mmol, 1 equiv), phenylacetic acid (4.16 g, 30.8 mmol, 1.5 equiv), boric acid (1.26, 20.4 mmol, 1 equiv), and dimethyl sulfone (7 mL) were added to a 3-neck round bottom flask equipped with a short path condenser under an atmosphere of argon. The reaction mixture was heated and the reaction progress was monitored by HPLC. For 3.2 , the relative molar equivalent of phenylacetic acid was lowered to one. For 3.3 and 3.4 , the volume of dimethyl sulfone was halved and doubled, respectively. For 3.5 , the reaction in 3.1 was scaled up by a factor of 3. For 3.6 , the reaction in 3.1 was done in sulfolane. For 3.7 , the reaction in 3.6 was repeated at a higher temperature of 280 °C. For 3.8 , the reaction in 3.7 was scaled up by a factor of 3. For 3.9 , the reaction in 3.7 was repeated at twice the volume of sulfolane. For 3.10 , the reaction in 3.9 was scaled up by a factor of 3. For 3.11 , the reaction in 3.7 was stopped at 24 hours. For 3.12 and 3.13 , diphenyl sulfone and tetraethylene glycol were used as the solvent, respectively.

Work-up Procedure . For 3.7 , once the reaction progress had stagnated as determined via HPLC, the reaction was allowed to cool to room temperature to produce a black solid mass. Water was added to nearly fill the flask and the mixture was stirred and heated at 50 °C overnight. The suspension was gravity filtered and the solid was rinsed with an excess volume of

353

water. A Soxhlet extraction using toluene was carried out to extract out a black organic-soluble solid (3.77 g) after rotary evaporation. The crude product (3.0 g) was purified via alumina column chromatography (dichloromethane) to give a shiny green-black product (197 mg, HPLC purity > 85%). The product was purified a second time by silica gel column chromatography (dichloromethane to methanol) to give a maroon red product (63 mg). The product was further purified via train sublimation, whereby the apparatus was operated under a vacuum with a controlled flow of carbon dioxide gas (~100 mTorr above pump down pressure). The thermal profile used is shown in Table S8.6, Method 6.5. This afforded MeO-mPh-BsubPy as a bright orange band (21 mg, 33% yield relative to mass placed in the train sublimation apparatus). 1H

NMR (400 MHz, CDCl 3) d 8.56 (br d, J = 7.22 Hz, 3H), 7.94 (br t, J = 7.52 Hz, 3H), 7.83 (t, J = 7.62 Hz, 3H), 7.68-7.62 (m, 9H), 7.48-7.43 (m, 6H), 7.42 (br d, J = 7.19 Hz, 3H), 0.96 (s, 3H);

13 C NMR (100 MHz, CDCl 3) d 137.1, 134.3, 132.8, 132.6, 131.4, 129.04, 129.00, 128.9, 126.4, 123.3, 117.0, 47.4; HRMS (EI-TOF+) m/z [M] calcd for 651.2482, found 651.2488.

Table S8.3 . Attempted syntheses of HO-mPh-BsubPy. Method Relative Concentration Solvent Temperature Reaction Formation Weight Molar (mol/L) b | (˚C) Time (h) (%) d Percentage Ratio a Relative Scale-Up c (%) e 3.1 1 : 1.5 : 1 10.3 | 1 DMS 230 166 27 9.6 | - 3.2 1 : 1 : 1 8.7 | 1 DMS 230 170 4 - 3.3 1 : 1.5 : 1 20.7 | 1 DMS 230 186 24 13.5 | - 3.4 1 : 1.5 : 1 5.2 | 1 DMS 230 93 0 - 3.5 1 : 1.5 : 1 10.3 | 3 DMS 230 102 0 - 3.6 1 : 1.5 : 1 10.3 | 1 SUFL 260 189 20 8.3 | - 3.7 1 : 1.5 : 1 10.3 | 1 SUFL 280 187 39 11.1 | - 3.8 1 : 1.5 : 1 10.3 | 3 SUFL 280 187 24 - 3.9 1 : 1.5 : 1 5.2 | 1 SUFL 280 190 20 8.4 | - 3.10 1 : 1.5 : 1 5.2 | 3 SUFL 280 119 14 - 3.11 1 : 1.5 : 1 10.3 | 1 SUFL 280 24 <1 - 3.12 1 : 1.5 : 1 10.3 | 1 DPS 280 168 6 - 3.13 1 : 1.5 : 1 10.3 | 1 TEG 280 43 0 - a Phthalimide : phenylacetic acid : boric acid. b Calculated with consideration for the sum of all three starting materials. c Relative to the base experiment ( i.e. Entry 1). d As determined by HPLC-UV/Vis (MAXPLOT) for BsubPy. e As determined optically using Beer-Lambert Law before and after a full work-up. DMS = dimethyl sulfone; SUFL = sulfolane; DPS = diphenyl sulfone; TEG = tetra(ethylene glycol).

354

Phenoxylation Route: Phenoxy meso -phenyl tribenzo boron subporphyrin (PhO-mPh-BsubPy) . To a round-bottom flask equipped with a condenser was added pre-chromatographic HO-mPh-BsubPy (3.00 g, 4.71 mmol, 1 equiv), phenol (2.21 g, 23.5 mmol, 5 equiv), and toluene (75 mL). The mixture was heated to reflux and the reaction progress was monitored by HPLC. Once all of the HO-mPh- BsubPy had reacted (within 12 hours), the reaction mixture was cooled to room temperature and concentrated to dryness via rotary evaporation.

Work-up Procedure I (WPI) . The crude PhO-mPh-BsubPy was dry loaded onto silica gel using dichloromethane and then loaded onto fresh silica gel in a glass thimble with a fritted disk (medium pore size) at the bottom. The thimble was placed inside of a Soxhlet extraction apparatus and was initially extracted with dichloromethane. Once the filtrate coming out of the thimble became clear, a switch to methanol was made to elute out the desired product. This process was monitored by a UV light source for its characteristic green fluorescence. Once this was collected, the filtrate was concentrated to produce a dark brown solid (29.1 wt %).

Work-up Procedure II (WPII) . The procedure was carried out in a similar manner as WPI, but with an additional step. After elution with dichloromethane was done, the thimble was removed from the Soxhlet apparatus and manually eluted with methanol until a dark brown-black fraction was collected and discarded. The thimble was placed back into the Soxhlet apparatus and a green-fluorescing product was collected and concentrated to a brown-red solid (39 wt %).

355

Detector Intensity vs. Reaction Time

Figure S8.1 . Detector intensity (at 510 nm) versus time for the formation of HO-mPh-BsubPy via the reaction of phthalimide, phenylacetic acid, and boric acid in sulfolane at 280 ˚C.

356

Beer-Lambert Law to Calculate Weight Percentage The wt % is calculated via the following formulas:

capparent = m/M mV (Eq S8.1)

A = εbc actual (Eq S8.2)

wt % = 100 % x (c actual / c apparent ) (Eq S8.3)

where c actual and c apparent is the respective actual and apparent concentration, ε is the molar absorptivity, b is the path length, m is the mass of the analyte, M m is the molar mass of the analyte, and V is the volume of the analyte solution. The c apparent is calculated based on a known mass of the crude sample dissolved into a solution of a known volume (Eq S8.1). The c actual is determined optically using B-L law using the reported ε of BsubPy (Eq S8.2). The wt % is then calculated by taking the ratio of the c actual and c apparent (Eq S8.3).

357

Sublimation Data Table S8.4 . Train sublimation experiments on MeO-BsubPc. Method a Thermal Profile Amount in Boat Amount Purity (%) b T (°C) Ramp Soak (mg) Sublimed Time (hr) Time (hr) (mg | % yield) 4.1 150 1 1 500 95 | 19 95 350 1 2 400 1 4 420 2 4 4.2 150 1 1 500 150 | 30 97 350 1 2 400 1 4 450 2 4 4.3 150 1 1 500 170 | 34 73 350 1 2 400 1 4 500 2 4 4.4 150 1 1 500 110 | 22 91 350 1 2 400 1 4 460 2 4 a b 100 mTorr of CO 2 gas added on top of the pump down pressure. As determined via HPLC (MAXPLOT).

358

Table S8.5 . Train sublimation experiments on MeO-n-BsubPy. Method a Thermal Profile Amount in Boat Amount Purity (%) b T (°C) Ramp Soak | Residual Sublimed Time (hr) Time (hr) Amount (mg) (mg | % yield) 5.1 150 1 1 60 | 38 Thin film - 350 1 2 400 1 4 450 2 4 5.2 150 1 1 60 | 36 Thin film - 350 1 2 400 1 4 500 2 4 5.3 150 1 1 60 | 36 Thin film - 350 1 2 400 1 4 530 2 4 5.4 c 150 1 1 60 | 35 Thin Film - 350 1 2 400 1 4 530 2 4 5.5 c 150 1 1 30 | 19 Thin film - 350 1 2 400 1 4 530 2 4 a b c 100 mTorr of CO 2 gas added on top of the pump down pressure. As determined via HPLC (MAXPLOT). OVPD process using ITO on glass substrate.

359

Table S8.6 . Train sublimation on mPh-BsubPy. Method a Thermal Profile Amount in Boat (mg) | Amount Sublimed T (°C) Ramp Soak Time Compound (mg | % yield) Time (hr) (hr) 6.1 150 1 1 92 | HO-mPh-BsubPy 0 | 0 350 1 2 400 1 4 450 2 4 6.2 150 1 1 90 | HO-mPh-BsubPy 0 | 0 350 1 2 400 1 4 500 2 4 6.3 150 1 1 90 | HO-mPh-BsubPy 0 | 0 350 1 2 400 1 4 530 2 4 6.4 150 1 1 94 | HO-mPh-BsubPy 0 | 0 350 1 2 400 1 4 530 2 8 6.5 150 1 1 63 | MeO-mPh-BsubPy 21 | 33 350 1 2 400 1 4 530 2 4 6.6 150 1 1 50 | MeO-mPh-BsubPy ~1 | ~2 350 1 2 400 1 4 530 2 4 6.7 150 1 1 50 | MeO-mPh-BsubPy ~1 | ~2 350 1 2 400 1 4 500 2 4 6.8 150 1 1 50 | MeO-mPh-BsubPy ~1 | ~2 350 1 2 400 1 4 500 4 4 6.9 150 1 1 50 | MeO-mPh-BsubPy ~1 | ~2 350 1 2 400 1 4 530 4 4 6.10 150 1 1 200 | MeO-mPh-BsubPy ~1 | ~1 350 1 2 400 1 4 500 4 4 6.11 150 1 1 100 | MeO-mPh-BsubPy ~1 | ~1 350 1 2 400 1 4 500 1 4 6.12 b 150 1 1 30 | MeO-mPh-BsubPy Thin film 350 1 2

360

400 1 4 530 2 4 a b 100 mTorr of CO 2 gas added on top of the pump down pressure. OVPD process using ITO on glass substrate.

(a) (b) (c) Figure S8.2 . Picture of (a) HO-mPh-BsubPy post-alumina column, (b) MeO-mPh-BsubPy post- silica gel column, and (c) MeO-mPh-BsubPy post-train sublimation (Method 3.7).

361

NMR Spectra

1 Figure S8.3 . H NMR (400 MHz, CDCl 3) spectrum obtained at 296 K of MeO-n-BsubPy after the hybrid Soxhlet-column extraction (Method 2.2).

362

1 Figure S8.4a . H NMR (400 MHz, CDCl 3) spectrum obtained at 296 K of sublimed MeO-mPh- BsubPy (Method 6.5).

363

13 Figure S8.4b . C NMR (100 MHz, CDCl 3) spectrum obtained at 296 K of sublimed MeO-mPh- BsubPy (Method 6.5).

364

1 Figure S8.5 . H NMR (400 MHz, CDCl 3) spectrum obtained at 296 K of sublimed MeO-mPh- BsubPy obtained via WPII (Method 6.6).

365

UV-vis Absorption and Photoluminescence Spectra

Figure S8.6 . Normalized absorption (blue) and photoluminescence (red, λ ex = 505 nm) spectra of the maroon red band ( n-BsubPy) following sublimation (Method 5.1) in toluene solution.

Figure S8.7 . Absorption spectrum of the film on ITO glass substrate ( n-BsubPy) following an OVPD process (Method 5.4) in toluene solution.

366

Figure S8.8 . Normalized absorption (blue) and photoluminescence (red, λ ex = 505 nm) spectra of MeO-mPh-BsubPy following sublimation (Method 6.5) in toluene solution.

Figure S8.9 . Normalized absorption (blue) and photoluminescence (red, λ ex = 505 nm) spectra of MeO-mPh-BsubPy obtained via WPII following sublimation (Method 6.6) in toluene solution.

367

Figure S8.10 . Absorption (blue) spectra of MeO-n-BsubPy films (deposited on ITO on glass substrates with pre-deposited PEDOT:PSS) obtained via OVPD (Method 5.5).

368

Table S8.7 . Electrochemical properties (CV and DPV).

1 2 3 a 1 a 1’ 2’ 3’ a 1’ a Compound E ox | E ox | E ox (V) E red (V) E ox | E ox | E ox (V) E red (V) MeO-n-BsubPy d +0.66 c | +0.81 b | +1.44 c - +0.60 c | +0.79 c | +1.36 c -0.79 c MeO-mPh-BsubPy e +0.83 b | +1.38 c -1.08 c +0.82 c | +1.32 c -1.03 c E = redox potential from CV. E’ = redox potential from DPV. a In degassed DCM solution relative to Ag/AgCl. b Half-wave potential. c Peak potential. d Compound obtained from Method 5.5. e Compound obtained from Method 6.5.

369

Estimating HOMO and LUMO Energy Levels The HOMO energy level can be estimated using the Thompson-Forrest equation 4:

EHOMO = -1.2 qVCV – 4.6, (Eq S8.4)

where q is the electron charge and V CV is the oxidation potential as determined from CV (in DCM solution).

The LUMO energy level can be calculated by:

ELUMO = HOMO + E g, (Eq S8.5)

where E g is the energy band gap that is calculated using the following equation:

Eg = hc /λ (Eq S8.6) where h is Planck’s constant (4.136 x 10 -15 eV·s), c is the speed of light (3.0 x 10 8 m/s), and λ is the wavelength. The λ is determined from the onset of the absorption with the lowest energy transition (solid state).

370

References 1. Edwards, L.; Gouterman, M.; Rose, C. B. Journal of the American Chemical Society 1976 , 98 , 7638-7641. 2. Alabaster, R. J.; Cottrell, I. F.; Hands, D.; Humphrey, G. R.; Kennedy, D. J.; Wright, S. H. B. Synthesis 1989 , 1989 , 598-603. 3. Dang, J. D.; Virdo, J. D.; Lessard, B. H.; Bultz, E.; Paton, A. S.; Bender, T. P. Macromolecules 2012 , 45 , 7791-7798. 4. D'Andrade, B. W.; Datta, S.; Forrest, S. R.; Djurovich, P.; Polikarpov, E.; Thompson, M. E. Organic Electronics 2005 , 6, 11-20.