<%BANNER%>

Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2011-08-31.

Permanent Link: http://ufdc.ufl.edu/UFE0024898/00001

Material Information

Title: Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2011-08-31.
Physical Description: Book
Language: english
Creator: Beaujuge, Pierre
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Statement of Responsibility: by Pierre Beaujuge.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Reynolds, John R.
Electronic Access: INACCESSIBLE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024898:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024898/00001

Material Information

Title: Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2011-08-31.
Physical Description: Book
Language: english
Creator: Beaujuge, Pierre
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Statement of Responsibility: by Pierre Beaujuge.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Reynolds, John R.
Electronic Access: INACCESSIBLE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024898:00001


This item has the following downloads:


Full Text

PAGE 1

1 SPECTRAL ENGINEERING IN CONJUGATED ORGANIC POLYMERS By PIERRE M. BEAUJUGE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 Pierre M. Beaujuge

PAGE 3

3 To my partners in research

PAGE 4

4 ACKNOWLEDGMENTS My acknowledgments go to my graduate research advisor, Professor John R. Reynolds, for letting me work with him, not -for him -. Timeless c ommitment to each member in the group, respect and appreciation of the research effort cultivated, and continuous encouragements to developing personal initiative are only some of the most valuable directions fo llowed by John I am grateful to my father who supported me and kept me focused over the years. Many thanks to those who made my journey at UF especially enjoyable including my research teammates as well as Professor Franky So

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TABLES ................................................................................................................................ 9 LIST OF FIGURES ............................................................................................................................ 12 LIST OF ABBREVIATIONS ............................................................................................................ 22 ABSTRACT ........................................................................................................................................ 24 CHAPTER 1 THE DONOR -ACCEPTOR APPROACH IN CONJUGATED POLYMERS: ORIGIN, PRINCIPLES AND APPLICATIONS ..................................................................... 26 1.1 The Fundamentals of Bandgap Control in Conjugated Polymers .................................. 26 1.2 High Performance Conjugated Polymers for Device Applications ........................... 28 1.2.1 In Field Effect Transistors ..................................................................................... 29 1.2.2 In Light Emitti ng Diodes ...................................................................................... 3 3 1.2.3 In Chemical Sensors .............................................................................................. 36 1.2.4 In Photovoltaic Cells ............................................................................................. 38 1.2.5 In Electrochromic Devices (ECDs) ...................................................................... 48 1.2.5.1 Electrochromism .................................................................................... 48 1.2.5.2 Electrochromic polymers (ECPs) ......................................................... 48 1.2.5.3 Electrochromic devices (ECDs) ............................................................ 50 1.2.5.4 Absorption/Transmission ECDs ........................................................... 51 1.2.5.5 Reflective ECDs .................................................................................... 52 1.3 Thienylene Benzothiadiazole Based Polymers (PTBTDs) .......................................... 53 1.4 -Conjugated Polymers in this Disssertation .................................................................. 57 2 CHARACTERIZATION TECHNIQUES AND EXPERIMENTAL METHODS ................ 60 2.1 Introduction ....................................................................................................................... 60 2.2 General Synthetic Methods .............................................................................................. 60 2.2.1 Selected Palladium -Mediated Cross Couplings .................................................. 61 2.2.1.1 Stille cross -couplings ............................................................................. 61 2.2.1.2 Suzuki cross -couplings .......................................................................... 64 2.2.2 Oxidative Polycondensations ................................................................................ 67 2.3 Electrochemical Methods ................................................................................................. 69 2.4 Spectroscopic Methods suitable for ECP Characterization ............................................ 71 2.4.1 Spectroelectrochemistry vs. Luminance Measurements ..................................... 71 2.4.2 Colorimetry ............................................................................................................ 73 2.4.3 Coloration Effi ciency ............................................................................................ 75 2.5 Surface Morphology Characterization Tools .................................................................. 76

PAGE 6

6 2.6 Structural Analysis by 2D -WAXS ................................................................................... 80 2.7 Charge -Carrier Mobility Measurements in FETs ............................................................ 82 2.8 Photovoltaic Devices ......................................................................................................... 84 2.9 Space Charg e Limited Current (SCLC) Modeling .......................................................... 86 3 SPRAY PROCESSABLE GREEN TO TRANSMI SSIVE SWITCHING 3,4 DIOXYTHIOPHENE 2,1,3 -BENZOTHIADIAZOLE DONOR ACCEPTOR ELECTROCHROMIC POLYMERS ......................................................................................... 90 3.1 Context, Proposed Design and Rational .......................................................................... 90 3.2 Synthesis and Characterization of 3,4 Dioxithiophene 2,1,3 -Benzothiadiazole Polymers ............................................................................................................................ 95 3.2.1 Synthesis and Physical Characterization .............................................................. 95 3.2.2 Optical Characterization ........................................................................................ 97 3.2.3 Polymer Redox and Spectroelectrochemical Characterization ........................... 99 3.2.4 Polymer Colorimetric Analysis .......................................................................... 101 3.2. 5 Polymer Switching Study.................................................................................... 103 3.2.6 Synthetic Details .................................................................................................. 105 3.3 Conclusions and Outlook ................................................................................................ 109 4 SYNTHETIC CONTROL OF THE SPECTRAL ABSORPTION IN 3,4 DIOXYTHIOPHENE 2,1,3 -BENZOTHIADIAZOLE DONOR ACCEPTOR ELECTROCHROMIC POLYMERS: FROM GREEN TO BLACK .................................... 111 4.1 Co ntext and Motivations for Tailoring a Two Band Absorption in the Visible ......... 111 4.2 Proposed Design and Polymer Optical Characterization.............................................. 114 4.2.1 Design and Synthesis of 3,4 Dialkoxythiophene 2,1,3 Benzothiadiazole Polymers .............................................................................................................. 114 4.2.2 3,4 Dialkoxythiophene 2,1,3 -Benzothiadiazole Polymer Optical Characterization .................................................................................................. 116 4.2.3 Design and Synthesis of 3,4 -Propylenedioxythiophene 2,1,3 Benzothiadiazole Polymers ................................................................................ 119 4.2.4 3,4 Propylenedioxythiophene 2,1,3 -Benzothiadiazole Polymer Optical Characterization .................................................................................................. 122 4.3 Extension of the Proposed Approach to the Synthesis of Black to Transmissive Switching Polymer Electrochromes ............................................................................... 124 4.3.1 Design, Synthesis and Physical Characterization .............................................. 124 4.3.2 Polymer Redox and Spectroelectrochemical Characterization ......................... 126 4.3.3 Polymer Colorimetric Analysis .......................................................................... 127 4.3.4 Polymer Switching Study.................................................................................... 129 4.4 Electrochromic Devices .................................................................................................. 131 4.5 Further Insight into the Optical Changes Observed When Using the Donor Acceptor Approach to Tailor a Two -Band Absorption in the Visible ......................... 132 4.5.1 Determination of the Attenuation Coefficients of P1 -P8 .................................. 132 4.5.2 Donor -Acceptor Conjugated Polymers Seen with Distinct Chromophores along the Backbone (Model #1 and #1 ).......................................................... 136

PAGE 7

7 4.5.3 Donor -Acceptor Conjugated Polymers Seen as Following Frontier Molecular Orbital Theory Principles (Model #2) ............................................ 136 4.5.4 In Summary .......................................................................................................... 139 4.6 Synthetic Details ............................................................................................................. 142 4.6.1 Synthesis of P1 P3 ............................................................................................... 142 4.6.2 Synthesis of P5 P7 ............................................................................................... 147 4.6.3 Synthesis of P9 ..................................................................................................... 152 4.7 Conclusions and Outlook ................................................................................................ 152 5 FINE BANDGAP TUNING IN 3,4 DIOXYTHIOPHENE 2,1,3 BENZOTHIADIAZOLE DONOR ACCEPTOR ELECTROCHROMIC POLYMERS VIA THE USE OF UNSATURATED LINNKAGES: FROM BLUE TO GREEN ............ 154 5.1 Motivations for Simplifying the Access to Colors Difficult to Achieve and Proposed Design .............................................................................................................. 154 5.2 Synthesis and Characterization of 3,4 Dioxythiophe ne 2,1,3 Benzothiadiazole Polymers Containing Unsaturated Linkages ................................................................. 155 5.2.1 Synthesis and Physical Characterization ............................................................ 155 5.2.2 Polymer Optical Characterization ...................................................................... 160 5.2.3 Polymer Redox Properties .................................................................................. 162 5.2.4 Polymer Spectroelectrochemical Characterization ............................................ 166 5.2.5 Polymer Colorimetric Analysis .......................................................................... 171 5.2.6 Polymer Switching Studies ................................................................................. 175 5.2.7 Synthetic Details .................................................................................................. 180 5.3 Conclusions and Outlook ................................................................................................ 182 6 STRUCTURE PROPERTY RELATIONSHIPS AND DESIGN RULES I N 3,4 DIOXYTHIOPHENE 2,1,3 -BENZOTHIADIAZOLE DONOR ACCEPTOR GREEN POLYMERS FOR PHOTOVOLTAIC APPLICATIONS ...................................... 185 6.1 Motivations for a Structure Relationship Study of 3,4 Dioxythiophene 2,1,3Benzothiadiazole Polymers ............................................................................................ 185 6.2 Proposed Design and Rational ....................................................................................... 187 6.3 Synthesis and Characterization of the 3,4 Dioxythiophene 2,1,3 Benzothiadiazole Polymer Hybrids ................................................................................ 189 6.3.1 Design, Synthesis and Physical Characterization .............................................. 189 6.3.2 Polymer Optical Ch aracterization ...................................................................... 193 6.3.3 Polymer Electrochemistry ................................................................................... 195 6.3.4 Photvoltaic Device Performance ........................................................................ 199 6.3.5 Device Morphology Study .................................................................................. 207 6.3.6 Charge Transport in Devices .............................................................................. 209 6.3.7 Synthetic Details .................................................................................................. 211 6.4 Conclusions and Outlook ................................................................................................ 214 7 SYNTHETIC CONTROL OF THE SPECTRAL ABSORPTION AND CHARGE TRANSPORT PROPERTIES IN DITHIENOSIL OLE 2,1,3 BENZOTHIADIAZOLE DONOR ACCEPTOR POLYMERS FOR PHOTOVOLTAIC APPLICATIONS ........... 216

PAGE 8

8 7.1 Context and Motivations for the Design of Broadly Absorbing Polymers with High Charge -Carrier Mobil ities ..................................................................................... 216 7.2 Proposed Design and Rational ....................................................................................... 222 7.3 Synthesis and Characterization of Dithienosilole 2,1,3 Benzothiadiazole Pol ymers .......................................................................................................................... 225 7.3.1 Synthesis and Physical Characterization ............................................................ 225 7.3.2 Polymer Optical Characterization ...................................................................... 228 7.3.3 Polymer Electrochemical Characterization and Band Structure ...................... 230 7.3.4 Polymer Structural Analysis ............................................................................... 233 7.3.5 Field Effect Transistors ....................................................................................... 234 7.3.6 Photovoltaic Devices ........................................................................................... 237 7.3.7 Synthetic Details .................................................................................................. 239 7.4 Conclusions and Outlook ................................................................................................ 244 8 PERSPECTIVES AND OUTLOOK ....................................................................................... 245 8.1 -Conjugated Polymers in Context ................................................................................ 245 8.2 The Donor -Acceptor Approach in Conjugated Polymers with Perspectives in Electrochromic Applications .......................................................................................... 246 8.3 The Donor -Acceptor Approach in Conjugated Polymers with Perspectives in Photovoltaic Applications ............................................................................................... 249 LIST OF REFERENCES ................................................................................................................. 254 BIOGRAPHICAL SKETCH ........................................................................................................... 261

PAGE 9

9 LIST OF TABLES Table page 1 1 Electrochemically determined HOMO and LUMO energy levels by cyclic voltammetry (CV) an d by differential pulse voltammetry (DPV) for the fullerenes PC60BM and PC70BM, corresponding electrochemical bandgaps, and comparison with their optically estimated values (from thin films) ........................................................ 47 3 1 Number average molecular weight ( Mn, g mol1), weight average molecular weight (Mw, g mol1), polydispersity index (PDI), average number of repeat units, and average number of rings for the donor acceptor copolymers PA and PB .......................... 97 3 2 Local absorption maxima (solution in toluene and solid state) and film optical bandgap ( Eg, eV) for the donor acceptor copolymers PA and PB ..................................... 99 3 3 Square -wave potential stepping EC switching of spray-cast PA onto ITO in 0.1 M LiBF4 / propylene carbonate solution switching between 0.5 V and +1.05 V ( vs. Fc/Fc+) with a switch time of 2 s ......................................................................................... 104 4 1 Number average molecular weight ( Mn, g mol1), weight average molecular weight (Mw, g mol1), polydispersity index (PDI), average number of repeat units, and average number of rings for the donor acceptor copolymers P1 -P3 ................................ 116 4 2 Local absorption maxima (solution in toluene and solid state) and film optical bandgap ( Eg, eV) for the donor acceptor copolymers P1 -P3 ............................................ 118 4 3 Number average molecular weight (Mn, g mol 1), weight average molecular weight (Mw, g mol 1), polydispersity index (PDI), average number of repeat units, and average number of rings for the donor acceptor copolymers P5 -P7 ................................ 122 4 4 Local absorption maxima (solution in toluene and solid state) and film optical bandgap ( Eg, eV) for the donor acceptor copolymers P5 -P7 ............................................ 124 4 5 Number average molecular weight ( Mn, g mol1), weight average molecular weight (Mw, g mol1), polydispersity index (PDI), average number of repeat units, and average number of rings for the donor acceptor copolymer hybrid P9 ............................ 125 5 1 Weight average molecular weight ( Mw, g mol1), polydispersity index (PDI), average number of repeat units, average number of rings, and Onset of decomposition temperature (as measured by TGA under nitrogen atmosphere) for the donor acceptor copolymers P1 -P6 ................................................................................................. 160 5 2 Local absorption maxima (solution in toluene and solid state) and film optical bandgap ( Eg, eV) for the donor acceptor copolymers P1 -P6 ............................................ 162

PAGE 10

10 5 3 Electrochemically determined HOMO and LUMO energy levels (by CV and by DPV), Corresponding electrochemical bandgaps, and Comparison with their optically estimated values for the copolymers P1 P6 ........................................................ 166 6 1 Number average molecular weight ( Mn, g mol1), Weight average molecular weight (Mw, g mol1), Polydispersity index (PDI), Average number of repeat units, Average number of rings, and O nset of decomposition temperature (as measured by TGA under nitrogen atmosphere) for the donor acceptor copolymers P1 -P3 ........................... 193 6 2 Local absorption maxima (solution in toluene and solid state) and film optical bandgap ( Eg, eV) for the donor acceptor copolymers P1 -P3 ............................................ 195 6 3 Electrochemically determined HOMO and LUMO energy levels (by CV and by DPV), Corresponding electrochemical ban dgaps, and Comparison with their optically estimated values for the copolymers P1 P3 ........................................................ 197 6 4 Solar cell device performance for P1 (under AM 1.5 illumination at an irradiation intensity of 100 mW cm2) as a function of polymer:PC60BM blend composition .......... 200 6 5 Solar cell device performance for P2 (under AM 1.5 illumination at an irradiation intensity of 100 mW cm2) as a function of polymer:PC60BM blend composition .......... 201 6 6 Solar cell device performance for P3 (under AM 1.5 illumination at an irradiation intensity of 100 mW cm2) as a function of polymer:PC60BM blend compo sition .......... 203 6 7 Optimization of the solar cell device performance for P3 (under AM 1.5 illumination at an irradiation intensity 100 mW cm2) by varying the hole -transporting interface layer (PE DOT or MoO3) and the type of PCBM used (PC60BM or PC70BM) ................ 205 6 8 Zero -field hole mobility in the pristine copolymers P1 -P3 and in the polymer phase of the optimized blends, derived from fitti ng the J -V data to trap -free single -carrier SCLC model ......................................................................................................................... 211 7 1 Weight average molecular weight ( Mw, kDa), polydispersity (PDI), number average molecular weight ( Mn, kDa, calculated from Mw and PDI), film optical bandgap (Eg, eV), -stacking ( nm) and chain to -chain distances ( d nm), FET charge carrier mobilities at saturation (sat, cm2 V1 s1), and current on/off ratios ( Ion: Ioff) for the all donor copolymers developed by Marks et al. .................................................................... 223 7 2 Elemental Analysis (C, H, N) as calculated (left) and experimentally found (right) by Atlantic Microlab, Inc. for the donor acceptor copolymers P1 -P4 .................................. 228 7 3 Number average molecular weight ( Mn, g mol1), Weight average molecular weight (Mw, g mol1), Polydispersity index (PDI), Average number of repeat units, Average number of rings, and Onset of decomposition temperature (as measured by TGA under nitrogen atmosphere) for the donor acceptor copolymers P1 -P4 ........................... 228

PAGE 11

11 7 4 Local absorption maxima (solution in toluene and solid state) and film optical bandgap ( Eg, eV) for the donor acceptor copolymers P1 -P4 ............................................ 229 7 5 Electrochemically determined HOMO and LUMO energy levels (by CV and by DPV), Corresponding electrochemical bandgaps, and Comparison with their optic ally estimated values for the copolymers P1 P4 ........................................................ 230 7 6 -stacking ( nm) and chain -to chain distances ( d nm) for the donor acceptor copolymers P1 -P4 ................................................................................................................ 234 7 7 Number average molecular weight ( Mn, kDa), polydispersity (PDI), film optical bandgap (Eg, eV), -stacking ( nm) and chain to chain distances ( d nm), FET charge carrier mobilities at saturation (sat, cm2 V1 s1), and cur rent on/off ratios (Ion: Ioff) for the donor acceptor copolymers P1 -P4 ............................................................ 235

PAGE 12

12 LIST OF FIGURES Figure page 1 1 Schematic representation of the evolution o f the bandgap with the extent of conjugation in -conjugated polymers .................................................................................. 27 1 2 The two mesomeric forms of polythiophene are not equivalent in energy (nondegenerate) due to the bond length alternation that results from the presence of both si ngle and double covalent bonding between atoms ............................................................ 27 1 3 Structural factors influencing the extent of the bandgap in semiconducting polymers ..... 28 1 4 Representative architectures employed in the construction of organic fieldeffect transistors (OFETs) ................................................................................................................ 30 1 5 Chart of the representative semiconducting -conjugated polymers with state -of the art performance in solution-processed FET devices ............................................................ 32 1 6 Representative architectures employed in the construction of organic light emitting diodes (OLE Ds) ...................................................................................................................... 33 1 7 Schematic representations of a) single-layer OLED device energy band diagram, b) two layer OLED device energy band diagram including a p-type (i.e. hole transporting) emitter and an electron transport material (ETM) ......................................... 35 1 8 Chart of commonly employed emitting -conjugated polymers with well -behaved light emission in solution -processed OLED devices ........................................................... 35 1 9 K+ induced aggregation of 15-crown 5 -functionalized poly( p -phenylene ethynylene) via intermolec ular complexation as K+ produces 1:2 complexes with 15 -crown 5 ........... 37 1 10 K+ complexation induces a polythiophene backbone to twist out of planarity with the extent of this twist depending on the diameter of the metal ion and the size of the complexing macrocycle ......................................................................................................... 38 1 11 Comparative illustration of a) a bilayer single junction polymer:PCBM based device and b) a bulk -heterojunction p olymer:PCBM based device as first proposed by Heeger et al. in 1995 ......................................................................................................... 39 1 12 a) Solar spectral irradiance (under Air Mass 1.5 conditions) as a function of photon wavelength and energy, b) the solar spectral irradiance is converted into power density, c) the photon flux is calculated by simply dividing the power density by the energy of the corresponding photon ..................................................................................... 41 1 13 Overview me chanism illustrating the solar power -conversion taking place in organic solar cells ................................................................................................................................ 43

PAGE 13

13 1 14 Schematic representations of a) PV device energy band diagram in open -circuit conditions, b) PV devi ce energy band diagram in short -circuit conditions ........................ 44 1 15 Schematic illustration of the loss mechanism encountered in polymer solar cells, and leading to the device open-circuit voltage ( VO C) being ultimately controlled by the optical bandgap of the smallest gap component i n the donor acceptor blend ............. 45 1 16 Schematic illustration of the possible redox degradation mechanisms unde rgone by semiconducting polymers under environmental conditions ................................................ 46 1 17 a) [6,6]-phenyl C61-butyric acid methyl ester or PC60BM, b) [6,6] -phenyl C71-butyric acid methyl ester or PC70BM ................................................................................................. 47 1 18 a) Schematic illustration of a conventional dual polymer transmission ECD, and b) Schematic illustration of a reflectivetype PECD using a porous membrane electrode .... 52 1 19 Representative electropolymerized PTBTD type narrow bandgap polymer analogues initially developed by Yamashita et al. from aromatic -donors and quinonoidacceptors ................................................................................................................................. 54 1 20 Representative chemically polymerized PTBTD -type narrow bandgap polymer analogues developed by Krebs et al. from aromatic -donors and quinonoid acceptors ..... 55 1 21 Soluti on-processable CPDT BTD type narrow bandgap polymer analogues as developed by Brabec et al. and further investigated by various groups such as Bazan, Heeger, Mllen et al. .............................................................................................................. 56 1 22 Chemically p olymerized PTDPP type narrow bandgap polymer analogues developed by Janssen et al. and Winnewisser et al. have recently emerged as an alternative to the now more established PTBTDs ....................................................................................... 57 2 1 Cata lytic cycle for the cross -coupling of organotins with aryl halides (electrophiles) according to the Stille cross -coupling conditions ................................................................ 63 2 2 Step -growth mechanism commonly proposed to illustrat e Stille -type polycondensations .................................................................................................................. 64 2 3 Catalytic cycle for the cross -coupling of organoboranes with aryl halides (electrophiles) according to the Suzuki cross -coupling conditions ................................. 67 2 4 Proposed mechanism for the oxidative chemical polycondensation of easily oxidized monomers ............................................................................................................................... 69 2 5 Converting electrochemical potenti als measured as a function of a particular reference electrode to a different reference electrode .......................................................... 70 2 6 Spectroelectrochemical experiments monitoring the formation of ionic states, namely pol arons (radical -cations/anions) and bipolarons (dications/dianions), upon

PAGE 14

14 progressive application of an electrical bias, can be used to evaluate the transmittance changes undergone over a broad range of wavelengths ............................... 72 2 7 Relative luminance as a function of applied potential for various film thicknesses of a polymer solution-cast on ITO -coated glass (with the thickest film represented by the green curve) ...................................................................................................................... 73 2 8 Effect of growth rate and thermal annealing on the morphology of P3HT/PCBM composite films (PCBM concentration = 50 wt%) .............................................................. 78 2 9 a) Schematic of the cross -section of a P3HT/PCBM BHJ device (1:0.7) and photographs representing the focusedand defocused cross -sectional TEM images of the corresponding device, b) TEM of a 120 nm thick slice of active layer with the corresponding binary image showing the continuous interp enetrating network favoring the assumption of a direct percolation pathway across the device ...................... 79 2 10 a) Schematic illustration of a semiconducting fiber as extruded by Pisula et al. b) Charac teristic features of a 2D -WAXS pattern for a -conjugated polymer exhibiting relatively pronounced -stacking interactions ...................................................................... 81 2 11 Schematic illustration of p type and n type field effect tran sistors (FETs) which, in turn, can be used to determine the hole and electron mobility (respectively) of semiconducting polymers ...................................................................................................... 83 2 12 a) Output characteristic curves of the source drain cur rent ( ISD) as a function of the drain voltage ( VSD) for various gate ( VG) voltages applied. b) Transfer plot ( ISD as a function of VG) from which the charge -carrie r mobility at saturation can be determined for the semiconducting polymer according to Equati on (2) ............................ 84 2 13 Current density response of a PV cell in the dark (blue curve) and under solar illumination (red curve) from which the short -circuit current ( JSC), open-circuit voltage ( VOC), a nd the current and voltage at the maximum power point ( Jm and Vm) can be determined ................................................................................................................... 85 2 14 The highest device fill factors (0 < FF < 100%) are obtained when the shunt resistance ( Rshunt) in r everse bias is maximal ( JSC relatively constant with V ), whereas the series resistance must be as negligible as possib le at higher voltages ( JSC goes to infinity at VOC) ........................................................................................................................ 85 2 15 Schematic conf iguration of a hole only device used in the characterization of the hole transport taking place in the active layer of conventional vertically stacked devices such as BHJ solar cells ............................................................................................. 87 2 16 Current density as a function of the effective electric field in a polymer based hole only device .............................................................................................................................. 88 3 1 The first promising neutral green conjugated ECP by Wudl et al. a donor acceptor based material exhibiting two bands of absorption in the visible spectrum ....................... 91

PAGE 15

15 3 2 Strictly linear and alternating DA systems involving two distinct energy transitions have been described by Salzner e t al. ................................................................................... 92 3 3 An example of the existence of strictly linear and alternating DA polymers involving two distinct energy transitions can be found in work from Krebs et al. on low bandgap conjugat ed polymers based on thiophenes and 2,1, 3 -benzothiadiazole .............. 93 3 4 Engineering solution processable neutral state green donor acceptor -conjugated polymers exhibiting fast switching properties and highly transmissive oxidized states ... 94 3 5 Synthetic route to ProDOT BTD based donor acceptor -conjugated copolymers PA and PB .............................................................................................................................. 96 3 6 a) Solution optical absorption spectra (in toluene) of Pentamer A and Pentamer B (the oligomer precursors for polymer PA and PB ) superimposed to those of their Trimer precursor (Trimer A and Trimer B), b) Solution optic al absorption spectra (in toluene) of DA -copolymers PA and PB superimposed to those of their Pentamer precursor (Pentamer A and Pentamer B) .............................................................................. 98 3 7 Spectroelectrochemistry of a) PA and b) PB ..................................................................... 100 3 8 Cyclic voltammetry (CV) of a thin film of a) PA and c) PB drop -cast on a Pt button electrode (9 scans), Differential Pulse Voltammetry (DPV) on a thin film of b) PA and d) PB drop cast on a Pt button electrode ..................................................................... 101 3 9 Relative Luminance (%) as a function of applied potential for both spray -coated PA and PB ................................................................................................................................... 102 3 10 a) Square-wave potential step absorptometry of spray -coated PA (monitored at 672 nm, from 0.5 V to +1.05 V vs. Fc/Fc+) and b) PB (monitored at 710 nm, from 0.7 V to +1 V vs. Fc/Fc+) onto ITO in 0.1 M LiBF4/ACN soluti on ............................................ 103 4 1 a) Switchable windows (courtesy of Sage Electrochromics Inc.) and b) Electronic paper (courtesy of Plastic Logic) ........................................................................................ 111 4 2 a) Schematic representation of the optical spectrum of a two -band absorbing chromophore reflecting the color green (green curve), b) Schematic representation of the optical spectrum of an ink like blue (blue curve) and black chromophores (black curve) .................................................................................................................................... 113 4 3 Designing 3,4 -dialkoxythiophene (DalkOT) and 2,1,3-benzothiadiazole (BTD) based polymers ( P1 -P3 ) whereby the BTDs are increasingly spaced by the DalkOTs along the backbone ............................................................................................................... 114 4 4 Synthetic route to DalkOT BTD based donor acceptor -conjugated copolymers P1 -P3 .................................................................................................................................... 115

PAGE 16

16 4 5 a) Solution optical absorption spectra (in toluene) of M1 -M3 (oligomer precursors for polymer P1 -P3 ), b) Solution optical absorption spectra (in toluene) of DA copolymers P1 -P3 along with that of the control polymer P4 The spectrum of each system is normalized at the longer wavelength absorption maximum ............................. 117 4 6 Designing 3,4 -propylenedioxythiophene (ProDOT) and 2,1,3 benzothiadiazole (BTD) based polymers ( P5 -P7 ) whereby the BTDs are increasingly spaced by the ProDOTs along the backbone .............................................................................................. 119 4 7 Synthetic route to ProDOT BTD based donor acceptor -conjugated copolymers P5 -P7 .................................................................................................................................... 120 4 8 a) Solution optical absorption spectra (in toluene) of M5 -M7 (oligomer precursors for polymer P5 -P7 ), b) Solution optical absorption spectra (in toluene) of DA copolymers P5 -P7 along with that of the control polymer P8 ......................................... 123 4 9 a) Schematic representation of the optical spectrum of the two -band absorbing polymer P5 reflecting the color blue -green (green curve), superimposed to that of the control magenta polymer P8 (pink curve), as well as the hypothetical optical spectrum of a polymer hybrid comprising blocks of the monomers constitutive of P5 and P8 (M5 and M8 ) with various composition ................................................................ 125 4 10 Synthetic route to ProDOT BTD based donor acceptor bla ck polymer electrochromes ...................................................................................................................... 125 4 11 Spectroelectrochemistry of P9 ............................................................................................ 126 4 12 a) Cyclic voltammetry (CV) of a thin film of P9 drop -c ast on a Pt button electrode (10 scans) b) Differential Pulse Voltammetry (DPV) on a thin film of P9 drop -cast on a Pt button electrode ....................................................................................................... 127 4 13 Relative luminance as a function of applied pot ential and film thickness for spray coated P9 ............................................................................................................................... 128 4 14 Square -wave potential step absorptometry of P9 spray -coated on ITO and monitored a) at 636 nm and b) 428nm in 0.1 M LiBF4/PC solution, 0 .6 V to +0.7 V vs. Fc/Fc+ ... 129 4 15 Long -term stability study via square -wave potential stepping electrochromic switching of drop -cast P9 onto glassy carbon button electrode, in 0.1 M LiBF4/pro pylene carbonate solution switching between 0.35 V and +0.65 V (vs. Fc/Fc+), and with a switch time of 4s .................................................................................. 130 4 16 Photographs of the smart window type device assembled at ambiance with P9 as the polymer electrochrome (A = 1.3) ........................................................................................ 132 4 17 Tandem chronoabsorptometry/chronocoulometry of a P9 -based window device (A = 0.6) switched from 0.8 V to +1.6 V ( ............................................................... 132

PAGE 17

17 4 18 a) Attenuation coefficients as a function of wavelength among the DA copolymer series P1 -P3 along with those for the control polymer P4 b) Attenuation coefficients as a function of wavelength among the DA -copolymer s eries P5 -P7 along with those for the control polymer P8 ...................................................................... 135 4 19 Proposed Model #1, a) Schematic representation of the evolution of the two-band spectral absorption in a series of DA conjugated polymers with varying concentration of electronrich and electron -deficient substituents along the backbone, b) The DA conjugated polymers are seen with distinct chromophores along the backbone .............................................................................................................................. 137 4 20 Proposed Model #1, the electronaccepting units are seen as isolated dopants which induce the formation of low lying charge separated energy states with the neighboring electron donating units ................................................................................... 138 4 21 Proposed Model #2, a) Schematic representation of the evolution of the two-band spectral absorption in a series of DA conjugated polymers with varying concentration of electronrich and electron -deficient substitu ents along the backbone, b) The DA conjugated polymers are seen as following Frontier Molecular Orbital Theory principles ................................................................................................................ 141 5 1 Designing 3,4 -dioxythiophene (DOT) and 2,1,3 -benzothiadiazol e (BTD) based polymers ( P1 -P6 ) whereby the DOTs are spaced by an unsaturated spacer to limit the extent of steric hindrance along the polyheterocyclic backbo ne and offer an alternative pathway to narrowing bandgaps in ECPs ......................................................... 155 5 2 a) Synthetic route to DalkOT BTD based DA -conjugated copolymers P1 P3 containing unsaturated linkers ( P2 and P3 ) or not ( P1 ) b) Synthetic route to ProDOT BTD based DA -conjugated copolymers P4 -P6 containing unsat urated linkers ( P5 and P6 ) or not ( P4a P4b ) ................................................................................ 156 5 3 MALDI -MS of DA -copolymers a) P1 b) P4a c) P3 and d) P6 ...................................... 158 5 4 Normalized solution optical absorption (in Toluene) for the trans -ethylene unsaturated DA -c opolymers P3 and P6 compared to their ethylhexyl -substituted parent copolymers P1 and P4a (respectively) .................................................................... 161 5 5 a) Cyclic voltammograms of P4a P4b and P6 drop -cast onto platinum disk electro des (0.02 cm2) in 0.1M TBAP/PC electrolyte solution, using a scan rate of 50 mV/s, b) Differential pulse voltammetry of P1 P4a and P6 drop -cast onto platinum disk electrodes (0.02 cm2) in 0.1M TBAP/PC electrolyte solution using a step time of 0.1 s, a ste p size of 2 mV, and an amplitude of 100 mV ................................................... 164 5 6 Spectroelectrochemistry of DA -copolymer a) P1 b) P4a c) P2 d) P5 e) P3, f) P6 and g) P4b ............................................................................................................................. 170 5 7 Relative luminance as a function of applied potential and film thickness for spraycoated DA polymers a) P1 b) P4a, c) P3 and P6 d) P4a and P4b ................................. 172

PAGE 18

18 5 8 Color matching corresponding to the plots of relative luminance as a function of applied potential and film thickness described in Figure 5 7 for spray -coated DA polymers a) P1 b) P4a c) P3 and P6 d) P4a and P4b .................................................... 174 5 9 Square -wave potential step absorptometry of spray -coated a) P1 (monitored at 653 nm, 0.6 V +), b) P4a (monitored at 677 nm, 0.6 V 0.6 V versus Fc/Fc+), c) P3 (monitored at 648 nm, 0.6 V /Fc+) and d) P6 (monitored at 685 nm, 0.6 V +), onto ITO in 0.1 M LiBF4/propylene carbonate solution. Switch times chosen among: 10 s step for 40 s (2 cycles), 5 s step for 20 s (2 cycles), 2 s step for 20 s (5 c ycles) and/or 1s ste p for 20 s (10 cycles) ..................................................................................................................... 177 5 10 Long -term switching of spray -cast a) P1 (monitored at 653 nm, 0.4 V versus Fc/Fc+, square -wave potential steps of 4s, complete cycle is 8s), b) P4a (monitored at 677 nm, 0.4 V +, s quare -wave potential steps of 6s, complete cycle is 12s), and c) P6 (monitored at 685 nm, 0.4 V V versus Fc/Fc+, square -wave potential steps of 6s, full cycle is 12s), on ITO in 0.1 M LiBF4/propylene carbonate solution............................................................................... 180 5 11 Overview representation of the color states attained among the 3,4 -dioxythiophene (DOT) and 2,1,3 -benzothiadiazole (BTD) based polymer series ( P1 -P6 ) .................... 184 6 1 Green -colored semiconducting materials addressing the requirements for use in photovoltaic devices could be especially useful in aesthetically pleasing light harvesting technologies spanning a) photovoltaic trees (pho tograph shows one of the photovoltaic trees standing in Styria, Austria), and b) photovoltaic turfs, where each leaf like or grass like light -harvesting pixel would act as an individual solar cell ......................................................................................................................................... 186 6 2 Green -colored solar cells combined into an earth toned and lightweight flexible photovoltaic device, for example, could find potential utility in military gears, chameleonic fabrics, and other camouflag e related applications, and can be envisaged to limit the use of conventional batteries, over the course of military operations for instance ......................................................................................................... 186 6 3 Schematic illustration of the proposed desi gn for the synthesis of dioxythiophene benzothiadiazole (DOT BTD) DA copolymers with tunable absorption spectra and variable charge transport properties .................................................................................... 189 6 4 Synthetic route to the dioxythi ophene benzothiadiazole (DOT BTD) DA copolymers P1 -P3 .................................................................................................................................... 192 6 5 Solution (in toluene) and thin film optical absorption spectra of DA -copolymers a) P1 b) P2 and c) P3 .............................................................................................................. 194 6 6 a) Cyclic (scan rate of 50 mV/s) and differential pulse voltammograms (step time of 0.1 s) of P2 drop -cast onto a platinum button electrode (0.02 cm2) in 0.1M TBAP/PC electrolyte solution b) Differential pulse v oltammog ram (step time of 0.1 s) of P3

PAGE 19

19 drop -cast onto a platinum button electrode (0.02 cm2) in 0.1M TBAP/PC electrolyte solution .................................................................................................................................. 196 6 7 Schematic representation of the energy -band structure of P1 -P3 showing the polymer HOMO and LUMO energies as estimated by DPV (green filled rectangles) with respect to those of an ideal polymer ( i.e. designed to be integrated in BHJ solar cells employing PC60BM and PEDOT PSS) ............................................................. 198 6 8 IV curves for BHJ solar cells made of P1 -P3 (at best polymer:PC60BM composition) in the dark (black curve, only P1 is represented) and under AM 1.5 solar illumination, 100 mW cm2 (blue, green and red curves) .................................................. 200 6 9 Superimposed EQE (blue curves) and polymer:PC60BM blend absoprtion (red curves) for P2 (filled squares) and P3 (empty circles) based BHJ solar cells (at best polymer:PCBM composition) and under AM 1.5 solar illuminat ion, 100 mW cm2 ........................................................................................................................................ 202 6 10 Energy -level diagram correlating the energy band structure of P3 and the work function of MoO3 (here replacing PEDOT PSS as the hole -tra nsporting interface layer) and PC70BM (here replacing PC60BM) .................................................................... 204 6 11 IV curves for BHJ solar cells made of P3 (at best polymer:PCBM composition, i.e 1:8) in the dark (black curve) and unde r AM 1.5 solar illumination, 100 mW cm2, for different device configurations ...................................................................................... 206 6 12 Superimposed EQE (blue curves) and polymer:PCBM blend absoprtion (red curves) for BHJ solar cells made of P3 (at best polymer:PCBM composition) with the architecture ITO/PEDOT/ P3 :PC60BM/LiF/Al ( empty circles ) and ITO/M oO3/ P3 :PC70BM/LiF/Al (filled circles), under AM 1.5 solar illumination, 100 mW cm2 ............................................................................................................................... 206 6 13 AFM tapping-mode images of P1 -P3 in blend with PC60BM (best polymer:PC60BM compositions are represented) ............................................................................................. 208 6 14 a) Current densities for hole -only devices of copolymers P1 -P3 a s a function of the effective electric field b) J -V characteristics of hole -only devices at best polymer:PCBM blend composition ..................................................................................... 211 7 1 a) The thieno[3,2 b ]thiophene based all -donor polymer analogues developed by McCulloch et al. (namely poly(2,5 -bis(3 alkylthiophen 2 yl)thieno[3,2 b ]thiophenes)s demonstrate up to 0.6 cm2 V1s1 of hole mobility (on/o ff ratio ~107) in FETs, b) I -V curves of an optimized polymer:PC70BM BHJ device in the dark (black) and under 1 sun illumination (blue) ....................................................................... 217 7 2 a) Series of PTs and PTVs as designed by Li et al. aimed at exhibiting the substantial charge transport properties of P3HT, while broadening the polymer optical spectrum over the UV visible region via the incorporation of conjugated side -chains and

PAGE 20

20 unsubstituted thiophenes (see polymers 1 12) or via cross -linking with conjugated bridges b) Some representative corresponding thin -film absorption spectra ................... 218 7 3 a) (2 -ethylhexyl) -substituted cyclopentadithiophene (CPDT) and 2,1,3 benzothiadiazole (BTD) based DA copolymer as initially reported by Brabec et al. are proving especially effective in BHJs, with PCEs exceeding 5%, o wing to their high charge -carrier mobilities (as high as 2 x 102 cm2 V1 s1 in bottom -gate FETs) and b) their long -wavelength absorption band extending into the near IR ...................... 221 7 4 Semiconducting silole -based all -donor -conjugated polymers developed by Marks et al. for use in p -channel FETs .......................................................................................... 223 7 5 Silole, Dithienosilole (DTS), Dibenzosilole (DBS), and their respective all -carbon counterparts, Cyclopentadie ne (Cp), Cyclopentadithiophene (CPDT) and Fluorene (F) .......................................................................................................................................... 223 7 6 Proposed silole based donor acceptor -conjugated semiconducting polymers combining electron rich DTS moieties and unsubstituted thiophene spacers with the electron -deficient BTD ........................................................................................................ 224 7 7 Synthetic route to silole -based donor acceptor -conjugated copolymers P1 P4 ........... 226 7 8 Normalized solution optical absorption (in Toluene) for the DTS BTD copolymers P1 P4 .................................................................................................................................... 229 7 9 Schematic representation of the energy -band structure of P1 -P4 showing the polymer H OMO and LUMO energie s as estimated by CV and DPV (green filled rectangles) with respect to those of an ideal polymer .................................................... 232 7 10 2D wide angle X -ray scattering ( 2D -WAXS) of P1 -P4 whereby the po lymer samples consisted of fibers prepared by mechanical extrusion ....................................................... 233 7 11 Field effect transist or characteristics of P4 : a) output curves taken at different gate voltages and b) transfer cur ves ............................................................................................ 235 7 12 Integration plots of the equatorial scattering intensities for P2 and P4 as determined from the 2D WAXS patterns of the polymer fibers prepared by mechanical extrusion 236 7 13 X ray diffraction (XRD ) of P1 P4 whereby the polymer samples consisted of thinfilms drop -cast onto HMDS treated silicon ........................................................................ 237 7 14 a) IV curves for a BHJ solar cell made of P4 (at best polymer:PCBM composition, i.e 1:1) in the dark (blue curve) and under AM 1.5 solar illumination, 100 mW cm2, for t he device structure ITO/MoO3/ P4 :PC70BM/LiF/Al (red curve), b) EQE response (blue curve ) of the polymer:PCBM blend of the same P4 -based device .......................... 239 8 1 Dissertation Work Overview: Spectral Engineering in CathodicallyColoring Conjugated Electrochromic Polymers ................................................................................ 247

PAGE 21

21 8 2 Dissertation Work Overview: Spectral Engineering i n Conjugated Polymers with Substantial Charge Transport Properties for Photovoltaic Applications .......................... 251

PAGE 22

22 LIST OF ABBREVI ATIONS AFM Atomic force microscopy AFP Amplifying fluorescent polymer BTD 2,1,3 Benzothiadiazole CV Cyclic voltammetry DOT 3,4 Dioxythiophene DalkOT 3,4 Dialkoxythiophene DPV Differential pulse voltammetry EC Electrochromic ECD Electrochromic device ECP Electrochromic polymer ETM Electron transport material Fc/Fc+ Ferrocene/Ferrocemium FET Field effect transistor GPC G el permeation chromatography HTM Hole transport material MALDI Matrix assisted laser desorption/ionization Mn Number average molecular weight Mw Weight average molecular weight NMR Nuclear magnetic resonance OFET Organic field effect transistor PCE Power conversion efficiency PDI Polydispersity index ProDOT Propylenedioxythiophene PSC Polymer solar cell

PAGE 23

23 PV Photovoltaic SCLC Space charge limited current SCE Saturated calomel electrode

PAGE 24

24 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy SPECTRAL ENGINEERING IN CONJUGATED ORGANIC POLYMERS By Pierre M. Beaujuge August 2009 Chair: John R. Reynolds Major: Chemistry In this dissertation original design rules for the synthesis of conjugated semiconducting polymers with tailored optical properties are proposed, and oriented towards producing materials finding applications in electrochromic and photovoltaic devices without excluding further implications in organic thin -film transi stors and photodetection systems In brief, the said design rules encompass a strat egy for controlling the concentration of electronrich and electron deficient heterocycles incorporated in the polymer repeat unit, as well as some critical insights into the extent / position / nature of the solubilizing groups to be appended to the main-chain depending on the electronic / physical properties targeted Importantly, the synthesized materials are systematically characterized in terms of their fundamental physical and electrochemical properties, with a special emphasis on elucidating their en ergy band structure ( following frontier molecular orbital theory principles ) when the focus is placed on photovoltaic applications. In addition several approaches are investigated to produce multifunctional materials combining optical features commonly di fficult to achieve with substantial charge transport properties as specifically desired in light -harvesting related processes The relevance of the synthetic approaches intr oduced is systematically challenged by in -depth analytical studies, as well as prot otype construction, via collaborative work with various well -established research groups.

PAGE 25

25 In a first instance two -band absorbing cathodically -coloring electrochromic polymers reflecting (or transmitting) the color green are described, along with their swi tching performance in three -electrode electrochemical cells. A second project introduces a synthetic methodology for manipulating this dual -band optical absorption with the goal of generating new color patterns, and demonstrates the synthesis of the first polymer electrochrome with an absorption spectrum extending homogenously over the entire visible region, hence being essentially black in the neutr al state, while switching to a transmissive state on application of an external bias. In a third project a m ore fundamental approach aims at probing the potential of unsaturated linkers (namely eth ylene and ethynylene) in cathodically -coloring polymer electrochromes, with the perspectives of copolymerizing a multiplicity of chromophores via the said linkers to p roduce unprecedented palettes of colors. The last two projects developed in this dissertation are directed to the synthesis, characterization, and testing of semiconducting polymers addressing the requirements of organic electronics suitable for solar cell applications. The emphasis is placed on two -band absorbing polymers reflecting the color green, or broadly absorbing across the visible spectrum to be essentially black, and combining substantial charge -carrier mobilities to promote the charge generation / transport processes at work in photovoltaic devices.

PAGE 26

26 CHAPTER 1 THE DONOR -ACCEPTOR APPROACH IN CONJUGATED POLYMERS: ORIGIN, PRINCIPLES AND APPLI CATIONS 1.1 The Fundamentals of Bandgap Control in Conjugated Polymers Following the first reports of metallic conductivity in iodine doped polypyrrole in 1963 by Weiss et al. ,13 and in iodine doped polyacetylene in 1977 by Alan J. Heeger, Alan MacDiarmid and Hideki Shirakawa,4 conjugated polymers have become a mainstay of research and development finding application in field effect transistors,5 light emitting diodes,6 chemical sensors,7,8 memories,9 photovoltaic cells,10,11 and electrochromic devices ,12 14 for the most cited. In particular, the perspectives of achieving synthetic metals via the polymerization of well chosen unsaturated organic precursors into conjugated species in which the charge carriers are readily transferred from the valence to the conduction band have widely driv en the research field during the firs t two decades. Considering the potential of conjugated polymers as organic metals and their perspectives in the various electronic and photonic devices mentioned above the concept of bandgap -engineering has rapidly emerged and the several variables infl uencing the bandgap in semiconducting polymers have been examined In fact, a primary parameter in the origin of an energy gap between valence and conduction bands in conjugated system s is considered to be the bond length alternation that results from the presence of both single and double covalent bonding between atoms (e.g. polyacetylene has bond lengths of ca. 1.35 and 1.45 ) More precisely, electronic repulsions analogous to those observed in cyclobutadiene take place upon conjugation, creating a geom etrical distortion, namely a Jahn Teller distortion (or Peierls distortion in condensed matter physics), thus opening up the energy gap. Energy bands arise from hybridization of the orbitals associated with the presence of double covalent bonding which i ntroduce a multiplicity of new energy states in the energy band structure of the conjugated material (see Figure 1 1). In the ideal case, the electronic

PAGE 27

27 delocalization of the -system is ev en along the conjugated backbone hence minimizing the bond length alternation such that the different mesomeric forms of the conjugated species would be equivalent in energy (see Figure 1 2) While t his hypothetical case would produce a zero bandgap system, it has practically never been attained to date and conjugated polymers have been employed for the ir semiconducting properties instead. Figure 1 1 Schematic representation of the evolution of the bandgap with the extent of conjugation in -conjugated polymers Figure 1 2. The two mesomeric forms of polythiophene are not equivalent in energy (non degenerate) due to the bond length alternation that results from the presence of both single and double covalent bonding between atoms. In brack ets, the hypothetical bond -equal situation would produce a zerobandgap system with its electrons evenly delocalized along the main -c hain

PAGE 28

28 In semiconductors ( Eg < 3 eV), the top of the valence band corresponds to the highest occupied molecular orbital (HOMO) and the bottom of the conduction band to the lowest unoccupied molecular orbital (LUMO). Both the HOMO and the LUMO levels can be adjusted through synthetic modifications, although a number of parameters adding to the bond length criteria and influ encing the width of the bandgap must be taken into account (see Figure 1 3). As a result, the overall energy gap of a -conjugated system can be obtained by adding the different contributions described above according to the following equation :15 Eg = EBLA + E + ESub + ERes + EInt Figure 1 3 S tructural factors influencing the extent of the bandgap in semiconducting polymers Irregioregular poly(3 -hexylthiophene) (P3HT) serves as an illustration 1.2 High -Performance -Conjugated Polymers for Device Applications In this section, a selected number of established or promising application s for organic electronics are presented with a stress on solution -processable -conjugated polymers in general and a specific look at the pers pectives of donor acceptor systems when appropriate

PAGE 29

29 1.2.1 In Field Effect Transistors Solution -processable semiconducting conjugated polymers are f inding application in field -effect transistors (FETs) for which mechanic al flexibility and processing costs have become a sine qua non of technological development and highthroughput device production. Relying on the application of an electric field to control the conductivity of a channel of charge carriers in a semiconducto r, FETs are employed in the design of logic circuit s In comparison wit h their inorganic counterparts which processing require elevated temperatures (>350oC), solution processable organic semiconductors can be applied onto plastic substrates ( e.g. PET) usi ng ambient conditions. A large number of deposition techn iques have been described in this area .16 18 Sirringhaus has detailed the device physics of solution-processed organic FETs (OFET s).19 In this area, polymeric systems are of particular interest owing to the possibility of producing ink formulations which rheological properties ca n be adjusted as a function of the manufacturing process.20 The t op-gate and bottom -gate device architectures illustrated in Figure 1 4 have been the most explored configurations so far. In addition, a bottom -gate device can be constructed with b ottom -contacts or top -contacts. While top -gate FETs have two electrodes (source and drain) deposited on a substrate, and are subsequently covered by a semiconductor (the polymer) and a dielectric layer (see Figure 1 4 a), bottom -gate devices have the gate e lectrode deposited directly on the substrate ( or alternatively serving as the substrate ), and their dielectric layer directly above (see Figure 1 4 b) In a bottom -gate configuration constructed with bottom -contacts, the source and drain are placed on top o f the dielectric and covered by the semiconducting layer. In contrast, a bottom gate configuration constructed with top-contacts has its semiconducting layer on top of the dielectric, and the source and drain deposited last (see Figure 1 4 c). The power

PAGE 30

30 cha nnel of the FET is defined as the region comprised between source, drain and gate The channel length corresponds to the distance between source and drain and its width is determined by the length of these two electrodes. Yang et al. and Rinzler et al. hav e recently independently introduced a vertical transistor architecture (VOFET) and demonstrated its utility for conjugated semiconducting polymers (see Figure 1 4 d) .21 23 One clear advantage of this approach is represented by the ease in obtaining especially small channel lengths which, in this case, correspond to the thickness of the organic semiconducting thin layer itself. a) b) c) d) Figure 1 4 Representative architectures employed in the construction of organic field-effect transistors ( OFETs): a) top -gate, b) bottom -gate / bottom -contact c) bottom -gate / top contact, and d) vertical transistor configuration as recently independently proposed by Yang et al. and Rinzler et al.21 23 Over the past few years, s olution processed O FETs with field mobilities as high as those obtained for amorphous siliconbased transistors (1011 cm2 V1s1), and above have been described. In this area, soluble oligoacenes ( e.g. pentacene), oligothiophenes and their fused hybrid derivat ives have shown the highest performance in devices so far, as illustrated by the

PAGE 31

31 dithieno[2,3 d ;2,3 d ]benzo[1,2b ;4,5 b ]dithiophene s (see Figure 1 5 a ) exhibiting up to 1.7 cm2 V1s1 of hole mobility (on/off ratio 107) recently reported by Mllen et al.24 The same group has achieved soluble discotic hexa peri -benzocoronenes which are liquid -crystalline, can organize into columnar superstructures and form 2D lattices (see Figure 1 5 b).25,26 When measured along their columnar axis, some of the ir structures have show n hole mobilities as high as 1.1 cm2 V1s1, which go down to 0.01 cm2 V1s1 (on/off ratio 104) in bottom gate FETs likely due to the diff iculty in controlling the macroscopic order of the columns that must be aligned in plane with the substrate to favor the charge transport. In parallel, when carefully designed, conjugated polymers have recently revealed performances approaching, sometimes exceeding, those of their small molecule counterparts in FETs assembled via conventional processing techniques For example, McCu lloch et al. have designed and synthesized pol y(quaterthiophene) analogue s possessing thieno[3,2 b ]thiophene building blocks namely poly(2,5 bis(3 alkylthiophen2 -yl)thieno[3,2 b ]thiophenes) s (see Figure 1 5 c) assembling into large crystalline domains upon annealing treatment and exhibiting up to 0. 6 cm2 V1s1 (on/off ratio ~ 107) in top gate FETs .17 As a comparison, regioregular poly(3 -hexylthiop hene) (P3HT) has shown up to 0.1 cm2 V1s1 of hole mobility in top -contact FET .27 In more disordered polythienyl systems, Marks et al. and McCullough et al. have demonstrated hole mobilities in the range 101102 cm2 V1s1 in bottom gate / top-contact and bottom -gate / bottom -contact devices respectively (see Figure 1 5 d and 1 5 e ), hence introducing the idea that the degree of macroscopic order may not be a determining factor in device performance and that other variables such as close -stacking interactions and polymer molecular weight can be the key -components for excellent charge transport Similarly, in absence of pronounced macroscopic order, Mllen et al. have monitored

PAGE 32

32 hole mobilities as high as 0.17 cm2 V1s1 in cyclopentadithiophene -based polymer s which only correlated to a close stacking (see Figure 1 5 f). Fig ure 1 5. Chart of the representative semiconducting -conjugated polymers with state -of the art performance in solution-processed FET devices If OFETs with relatively high electron mobilities have been in general more challenging to achieve due to the comp lexity in designing semiconducting polymers with especially high electron affinities, Facchetti et al. have shown 0.45 0.85 cm2 V1 s1of electron mobility in a solution -processable naphtalenedicarboxyimide derivative (Polyera ActivInk N2200, see Figure 1 5 g) offering perspectives for polymeric complementary circuit technologies, where electron and hole transport organic thin -film transistors are combined to operate in concert.20

PAGE 33

33 Importantly, thin film transistors can be assembled with the purpo se of evaluating the charge -carrier mobility of a semiconducting component. This aspect will be further discussed in Chapter 2 1.2.2 In Light Emitting Diodes Light emission is a nother field of application where -conjugat ed polymers are especially desired as both emitters and charge transport materials with the perspective of being applied in flexible -panel displays and lightning.6 F ollowing the discovery of light emission in conjugated polymers by Tang et al. ,28,29 significant research effort has been initiated with the intention of probing the full potential of organic materials in simple vertical device architectures (see Figure 1 6 ). In parallel designing suitable organic electronics for red, blue, green and white light emitting diodes (OLEDs) has been the mainstay of research and development in this area so far. a) b) c) d) Figure 1 6 Representative architectures employed in the construction of organic light -emitting diodes (OLEDs): a) without carrier -transport layer, b) with hole transport material (HTM), c) with electron transport material (ETM), and d) with HTM and ETM

PAGE 34

34 F or a single layer configuration to be efficient, the emitting material should combine high luminescence quantum yield and substantial charge transport properties, which are two variables that have long been thought to be exclusive.30,31 As a result, materials enhancing the electron and hole injection s are generally integrated in the device structure, resulting in reducing the nonradiative loss mechanism s, hence improving the devic e efficiency, brightness and operating voltage. In addition, a matchup between the various constitutive components of an OLED is required as large injection barriers to overcome generally limit the device performance. Figure 1 7 illustrates how the introdu ction of an electronor hole transporting layer (ETM or H TM) can favor the charge transport across the device by introducing intermediate energy levels aligning the emitter energy band s tructure with the work -functions of the injection electrodes. In the area of OLEDs, the polymer band structure is generally defined in terms of electronic affinity (EA) and ionization potential (IP) instead of their frontier molecular orbitals ( LUMO and HOMO, respectively ). For exampl e, poly[2 methoxy5 (2' -ethyl hexyloxy) -1,4 phenylene vinylene] (MEH -PPV see Figure 1 8 a) and poly(9,9 -dioctylfluorene) (PFO see Figure 1 8 b ) have significant barriers for hole injection and particularly large ones for electron injection and necessitat e the introduction of an ETM It is worth noting that an ETM for instance does not only lower the barrier for electron injection ( Ee) but also acts as a hole blocking layer since its IP is commonly larger than that of the emitter. As a result the ETM will balance the overall charge flow across the device by limiting the hole currents as most semiconductors suitable for light -emissi on show relatively low electron mobilities. In addition to the well -established PFO and MEH PPV, -conjugated polymers exhibiting higher EA have been explored in OLEDs such as the alternating 2,1,3 benzothiadiazole / 9,9 dioctylfluorene analogue (EA = 3.2 3.5 eV and IP = 5.9 eV) (see Figure 1 8 c ) demonstrating a

PAGE 35

35 brightness of ca. 10, 000 cd m2 and an efficiency of ca. 14.6 cd A1 when used as a green emitting ETM and up to ca. 153, 000 cd m2 of brightness in blend with PFO in micro -sized pixels .32,33 a) b) Figure 1 7. Schematic representations of a) singlelayer OLED device energy band diagram, b) two layer OLED device energy band diagram including a p -type ( i.e. hole transporting) emitter and an electron transport material (ETM) Figure 1 8. Chart of commonly employed emitting -conjugated polymers with well -behaved light emission in solution -processed OLED devices While much progress remains to be made in the field of polymer LEDs with the synthesis of intrinsically more efficient emitting -materials and the careful design of suitable device

PAGE 36

36 architectures, some trends are now emerging in device optimization with the most conventional electroluminescent polymers as described by Jenekhe et al. for instance.6 1.2.3 In Chemical Sensors With more than just promises, -c onjugated polymers have emerged as transduction materials for chemical sensing where by a chemical signal associated to the detection of an analyte is transformed into an electrical or optical event. For example amplifying fluorescent polymers (AFP s ) r ely on the transport of electronic excited states as opposed to the transport of charges (electrons or holes) in s ilicon -based systems. Excited states in AFP s are also referred to as ex citons owing to the quasipartic le character of the local structural c hange (finite in length) induced in the polymer back bone upon excitation/relaxation. The propagation mechanisms involve through-space dipolar interactions and mixing of excited states. The amplifying propensity of conjugated polymers arises from this effic ient transport mechanism which endows them with high levels of sensitivity to the smallest perturbations when compared to small er molecules. A representative approach developed by Swager et al. for metal ion sensing has been the design, synthesis and use of c rown ether -based polymeric materials .7,8 In such systems, analyte induced aggregation is the driving force for importa nt optical changes comprising shifts in optical absorption and fluorescence quenching. For example, the 15-crown -5 -functionalized poly( p -phenylene ethynylene) illustrated in Figure 1 9 is prone to aggregate through intermolecular complexation of K+ ions.34 This process results in a bathochromic shift of the polymer maximum of absorption from ca. 25 nm consistent with the formation of -stacked aggregates. In parallel, the emission is quenched from 82% at the polymer emission maximum.

PAGE 37

37 While K+ produces 1:2 complexes with 15-crown 5, Li+ and Na+ can only form 1:1 complexes wit h the same complexing moiety and cannot lead to metal ion bridged polymer aggregates. Figure 1 9 K+ induced aggregation of 15 -crown 5 -functionalized poly( p phenylene ethynylene) via intermolecular complexation as K+ produces 1:2 complexes with 15 -crown 5 (Adapted with permission from Ref.34 Copyright 2000 WILEY -VCH Verl ag GmbH & Co. KGaA, Weinheim) The synthetic design and mechanistic hypothesis illustrated in Figure 1 10 have been introduced to demonstrate how a binding event can directly affect the conformation of a conjugated backbone.35 In this case, the ion complexation induces the polymer backbone to twist out of planarity with the extent of this twist depending on the diameter of the metal ion and the size of the complexing macrocycle. As a result, both intrachain (conjugation length related) and interchain conductivity (interchain spacing -dependent) are reduced and strong ionochromic effects are observed. In 2000, Swager et al. have described the broad topic of chemical se nsing from the perspective of conjugated polymers in a comprehensive review article.7 The same group has more recently reviewed the various approaches employing fluorescent conjugated polymers that have been developed and investiga ted for chemical sensing.8

PAGE 38

38 Figure 1 10. K+ complexation induces a polythiophene backbone to twist out of planarity with the extent of this twist depending on the diameter of the metal ion and the size of the complexing macrocycle (Adapted with permission from Ref.35 C opyright 1993 American Chemical Society) 1.2.4 In Photovoltaic Cells In contrast with OLEDs where small molecules and -conjugated oligomers have proven significantly more effective so far, semiconducting polymers are pres ently revealing better results in thin film light -harvesting devices. With state of -the art power conversion efficiencies (PCEs) exceeding 5% under AM 1.5 solar illumination, polymer solar cells (PSCs) as polymer FETs, capitalize on low -cost solution proc essing techniques applicable at ambiance as well as on the mechanical flexibility and lightweight of the substrates from which they can be fabricated. First introduced by Heeger et al. in 1995,36 active layers consisting of interpenetrating networks of semiconducting -conjugated materials and fullerene derivatives (e.g. [6,6] phenyl C61butyric acid methyl ester, PCBM) namely bulk heterojunctions (BHJs), have rapidly become the cornerstone of organic solar cell development In comparison to their bilayer analogues (s ee Figure 1 1 1 a) BHJs offer greater surface areas between the electron rich ( e.g. polymer) and electron -deficient ( e .g. PCBM ) components (see Figure 1 1 1 b) thereby offering the potential for higher densities of photogenerated charges and more effective c harge ex tracti on

PAGE 39

39 a) b) Figure 1 11. Comparative illustration of a) a bilayer single junction polymer:PCBM based device and b) a bulk -heterojunction polymer:PCBM based device as first proposed by Heeger et al. in 1995.36 In comparison to their bilayer analogues, BHJs offer greater surface areas between the electr onrich ( e.g. polymer) and electron-deficient (e.g. PCBM) components, thereby offering the potential for higher densities of photogenerated charges and more effective charge extractions

PAGE 40

40 In conventional BHJ type devices, t he active layer across which the ch arge dissociation occurs under illumination is sandwiched between two electrodes collect ing the dissociated charges, here Al ca. 4.3 eV vs. vacuum) as the cathode and ITO ( ca. 4.8 eV vs. vacuum) for the anode At the polymer anode interface, PEDOT:PSS (LUMO = ca. 5.0 5.1 eV vs. vacuum ca. 25 nm ) is commonly spin-coated from aqueous solution whereas LiF ( ca. 3.8 eV vs. vacuum, 1 nm ) is employed at the cathode to f acilitate the electron -extraction It is worth noting that low work -function metals such as Ca ( ca. 2.9 eV vs. vacuum ) have shown excellent results in organic solar cells, although their atmospheric inst ability causes the subsequent devices to degrade in absence of encapsulation. In contrast, low work -function metals are well suited for electron -injection and are commonly used in OLEDs and FETs. Marks et al. have recently commented on the use of PEDOT:PSS which is described as a material highly corrosive, hence setting the progressive degradation of both the semiconducting polymer active layer and the ITO electrode over time. They have proposed to replace the acidic composite with different organic and inorganic hole -injection layers (e.g. NiO ) .37,38 In spite of projected power conversion efficiencies of 10% in single layer solar cells ,39 and 15% in tandem devices ,40 the highest reported values surprisingly remain in the 5 6% range.10,11,41 Of all the variables influencing BHJ device performance, the polymer energy band structur e is determining as it governs the width and position of the spectral absorption, it impacts t he device open -circuit voltage and it controls the photoinduced electron transfer to the strongly accepting fullerene analogue. In these conditions the donor acceptor (DA) approach introduced by Havinga et al. in macromolecular systems ,42 and consisting in alternating electron rich and poor building units along the conjugated backbone has attracted a good deal of attention and narrow -bandgap -conjugated polymers absorbing at longer -wavelengths than their wi der -

PAGE 41

41 bandgap all donor parents ( e.g. P3HT, MEH -PPV and MDMO PPV) have been synthesized.4348 In particular, Heeger et al. have demonstrated how donor a cceptor polymer chromophores possessing longer -wavelengths optical transi tions (where the photon flux is the most intense see Figure 1 1 2 c ) can be employed as complementary light absorbing materials in tandem solar cells with an all donor analogue harvest ing in the shorter -wavelength region (i.e. where the photon flux is the most energetic, see Figure 1 1 2 a and 1 1 2b ) .49 a) b) c) Figure 1 1 2. a) Solar s pectral irradiance (under Air Mass 1.5 conditions) as a function of photon wavelength (bottom) and energy (top). This quantit y defines the amount of radiation received by a surface at a particular wavelength (W m2 nm1). b) The s olar spectral irradiance is converted into power density (W m2). c) T he photon flux (number of photons received per unit area and per second, m2 s1) is calculated by simply dividing the power density by the energy (in Joule s as opposed to eV !) of the corresponding photon Note that the maximum photon flux becomes more intense past 600 nm.

PAGE 42

42 On the other hand, beside the requirement for a broad spectral absorption across the visible and into the near IR, organic polymers suitable for BHJ applications with PCBM must exhibit balanced charge carrier mobilities over the polymer and PCBM rich phases (less than two orders of magnitude difference) to avoid fall ing into a space -charge limited device performance. In this respect, since the electron mobility of PCBM is commonly reported in the order of 103 cm2 V1s1,50 the polymer should not exhibit hole mobilities lower than 105 cm2 V1s1 (as vertically measur ed i.e. direction normal to the substrate). In spite of this known requirement, it is well accepted that the charge transport in organic solar cells remains the limiting -factor in device performance. In general, the exciton diffusion length in conjugated polymers is low in comparison to the photon absorption length ( ca. 100 nm), and the commonly reported values range from 50 to 200 for the highest estimations (i.e. 5 20 nm). The mobility and lifetime of the exciton are the parameters that define its diffusion length. Figure 1 1 3 illustrates the processes at work in a polymer solar cell, where 1) light absorption, 2) charge genera tion, 3) charge diffusion, 4) charge dissociation, and 5) charge collection constitute the key mechanistic steps. In the favorable situation, charge transfer occurs within an encounter donor acceptor complex (involving either a donor or an acceptor singlet exciton), hence forming a geminate electron -hole pair, which can dissociate and produce free charges. Alternatively, the geminate pair can recombine in a non radiative loss mechanism, and the exciton is lost. Geminate recombinations are especially promine nt in low dielectric -constant media dominated by strong Coulombic binding energies between holes and electrons. The charge dissociation step is commonly believed to be rate determining, and can be enhanced with increased concentration of a higher dielectri c -constant medium such as electron accepting fullerenes. An example of a well -behaved donor acceptor

PAGE 43

43 system is P3HT/PCBM for which up to 90% of the geminate pairs undergo effective dissociation upon post deposition annealing of the blend.51 In contrast, dissociation rates of less than 40% are commonly observed in other polymer based solar cells. Following the dissociation process, an additional loss mechanism is represent ed by the bimolecular recombinations taking place as free oppositely charged species have large interface areas to meet in BHJ devices in particular. Bimolecular recombinations occur in high concentrations in low charge -carrier mobility media where the bui ldup of charges favors the side processes as well as in blends of charge transport balanced low charge -carrier mobility media in general. a) b) Figure 1 1 3 Overview mechanism illustrating the solar power -conversion taking place in organic solar cells a) Photoinduced charge transfer and formation of a geminate electron -hole pair and b) Dissociation of the intermolecular exciton

PAGE 44

44 In addition to the charge transport aspects developed above, optimizing the device performance should include minimizing the l oss of energy in the mismatch between the polymer energy band structure and the work -functions of the various device components. Figure 1 1 4 illustrates the device energy conf iguration when a typical polymer donor is introduced in a blend with a fullerene acceptor as the active layer. While the device is in the open -circuit mode (flat band conditions) allowing the determination of its opencircuit voltage ( VOC) in Figure 1 1 4a (note the degeneracy lift in both the donor and the acceptor upon photoinduced el ectron transfer as radical ions are forming!) the band s of the semiconducting components bend in the short circuit conditions as the Fermi levels of the metal electrodes align as described in Figure 1 1 4b a) b) Figure 1 14. Schematic representations of a) PV device energy band diagram in open-circuit conditions, b) PV device energy band diagram in short -circuit conditions Beside those well -established and simple physical aspects, a closer look at the actual energetics of the photoinduced energy transf er occurring within the donor acceptor blend placed in those conditions has very recently been proposed by Janssen et al. in a contribution looking to understand the energy losses in organic solar cells.52 Figure 1 1 5 summarizes the impor tant conclusions of their work.

PAGE 45

45 Figure 1 15. Schematic illustration of the loss mechanism encountered in polymer solar cells, and leading to the device open-circuit voltage ( VOC) being ultimately controlled by the optical bandgap of the smallest gap component in the donor acceptor blend. (Adapted with permission from work presented by Ren Janssen at University of Florida in March -April 2009) Throughout their analysis, it appears that as little 0.1 eV is sufficient to induce charge transfer from the donor to the acceptor, and access the charge separated state (CT) (i.e. to form an inter molecular charge transfer exciton) Importantly, an additional 0.3 eV was proposed to account for the Coulombic energy difference between the CT state and a lower energy state corresponding to the formation of an intra molecular c harge transfer exciton ( i.e. considering the donor acceptor exciton as its own entity). With an extra 0.2 eV used in the charge transport, it was concluded that at least 0.6 eV is lost between the optical bandgap of the smallest -gap component in the blend ( EgOPT, most commonly that of the polymer donor) and the actual device VOC, meaning that the maximum VOC attainable is ultimately set by the bandgap according to e VOC = EgOPT 0.6 eV. This important remark is expected to impact the design and synthesis of -conjugated semiconducting polymers with optimized energy band structure including low lying HOMOs as desired for photovoltaic cells with higher VOC, and thereby possibly higher

PAGE 46

46 power conversion efficiencies (see parameters of importance in solar cells d escribed in Chapter 2 of this dissertation). In addition to those essential considerations, the redox thresholds beyond which semiconducting polymers are susceptible to degrade (i.e. s elf -dope ) under environmental conditions should be kept in mind (see Fig ure 1 16). Figure 1 16. Schematic illustration of the possible redox degradation mechanism s undergone by semiconducting polymers under environmental conditions 1) Polymers with a LUMO deeper than 3.6 eV undergo ntype doping, 2) Polymers with a HOMO hi gher than 5.25.3 eV undergo ptype doping (Figure created with contribution from Ken Graham ) With the main focus placed in tailoring the band struc ture of the polymer donor fullerene acceptors have received only limited attention so far,53 likely due to the complexity in synthesizing and isolating deriv atives with tailored energy band configurations. To date, in BHJ solar cells, the most conventionally employed electron accepting fullerenes remain PC60BM and PC70BM (see Figure 1 1 7 ) which possess the same LUMO energy levels (see Table 1 1, and see Chapte r 2 for more details on the electrochemical measurements performed) hence the same electron affinity

PAGE 47

47 a) b) Figure 1 17. a) [6,6] -phenyl C61butyric acid methyl ester or PC60BM, b) [6,6]-phenyl C71butyric acid methyl ester or PC70BM Table 1 1. Electroc hemically determined HOMO and LUMO energy levels by cyclic voltammetry (CV) and by differential pulse voltammetry (DPV) for the fullerenes PC60BM and PC70BM, corresponding electrochemical bandgaps, and comparison with their optically estimated values (from thin films) Fullerene Eoxonset (V) HOMO (eV) Eredonset (V) LUMO (eV) Egap (V) Egap (V) CV DPV CV DPV CV DPV CV DPV CV DPV Optical PC60BM 1.22 1.10 6.32 6.20 0.94 0.80 4.16 4.30 2.16 1.90 1.89 PC70BM 1.03 0.91 6.13 6.01 0.97 0.81 4.13 4.29 2.00 1.72 1.74 a oxidation ( Eoxonset ) and reduction ( Eredonset ) potentials are reported vs. Fc/Fc+. HOMO and LUMO energy levels are derived from the electrochemical data ( Eoxonset and Eredonset respectively) considering that the SCE is 4.7 eV vs. vacuum and Fc/F c+ is 0.38 eV vs SCE, i.e. 5.1 eV relative to vacuum. As opposed to PC60BM which absorbs mainly in the UV where the photon flux is minimal (see Figure 1 1 2c) PC70BM absorbs in the blue region of the visible (peak at ca. 500 nm ), and thereby contributes more effectively to the photogenerated current This phenomenon is commonly attributed to symmetry differences inducing distinct spin allowed and spin-forbidden transitions in PC60BM and in PC70BM Then again, beside the symmetry considerations, it is wort h noting from Table 1 1 that even though the fullerene counterparts show near identical LUMOs, their respective HOMOs differ by ca. 0.2 eV, with the lower estimated bandgap value for PC70BM also 0.2 eV lower than that of PC60BM. Although never mentioned as a contributing -factor t he lower overall bandgap of PC70BM can be expected to correlate to a certain extent with the bathochromic shift observed

PAGE 48

48 1.2.5 In Electrochromic Devices (ECDs) 1.2.5.1 Electrochromism Defined as a n optical change operating in a material upon electron transfer (redox), or by application of an external bias, the phenomenon of electrochromism was first observed by Deb et al. in the transition metal oxide WO3 in 1969.54 Featuring high photochemical stability and switching reversibly from a transparent to an intense blue color via electrochemically induced intervalence chargetransfer optical transitions, cathodically col oring WO3 has become the material of choice for smart windows .55 Since then, numerous other metal oxides have been studied in the context of black to colorles s' switching electrochromics including oxides of molybdenum (MoO3), iridium (Ir(OH)3) and nickel (NiO) .13 In addition to being inherently brittle, metal oxides often exhibit long switching times of multiple seconds to minutes along with low coloration efficiencies and contrast ratios .13 Alternatively, by exhibiting higher contrasts, a variety of colors, and the capability to perform on flexible substrates, organic electrochromics are now being applied to numerous appl ications including smart windows and mirrors ,12,56 switchable displays ,57 and e lectrochromic inks .13,57 In parallel, they could possibly meet the growing demand for optical modulators and shutters in message -laser communications and optical data storage .13 1.2.5.2 Electrochromic p olymers (ECPs) Of all the possible applications, light emitting and nonemissive electrochromic devices (ECDs) require precise control of the colors displayed in terms of their hue, saturation, intensity and/or their brightness. Besides the requirements for ae sthetically pleasing color patterns, displays, for instance, are expected to include emissive or non-emissive chromophores which can be brought together as pixels or superimposed to recreate several other desirable colors via the principles of color mixing theory. In this context, -conjugated polymers combining mechanical

PAGE 49

49 flexibility, and ease in bandgap/color tuning via structural control, along with the potential for low -cost scalability and processing, are attractive. In both light emitting and electrochromic (EC) technologies, changing the composition of the polymers at the molecular level has been the most extensively investigated approach to color tuning so far. Hence, a variety of synthetic strategies have been described over the years spanning varying the overall planarity of the backbones as a function of steric hindrance, changing the electronrich and poor character of the building blocks incorporated in the repeat unit, increasing the conjugation lengths via the use of fused heterocycles able to reduce the polymers bond-length alternation, copolymerizing different monomers randomly and in different feed ratios, and so on.14 Alternatives have consisted in blending different electroactive components, or creating laminates and composites with other types of chromophores or insulating materials.14 In fully -conjugated organic polymers, as the oxidation level is progressively in creased, charged carriers balanced with counter ions are created along the backbone which have long wavelength optical absorptions.13 For a sufficiently low energy -gap polymer, as oxidative (p type) doping generates the formation of radical cations (polarons) and further dications (bipolarons) the absorption can be transferred into the near IR with depletion of the ground-state transition in the visible. The ability of the backbone to assume a stable quinoidal geometry, as well as the position of the bipolaronic transition relative to the visible region, governs the transmissivity of the fully oxidized state. Combining high contrast ratios, fast response times, narrow potential windows of application with the perspective for long term optical stability, processable electrochromic polymers (ECPs) are now impacting the development of both transmissive and reflective EC technologies. With the same ubiquitous target applications as those described above for their

PAGE 50

50 inorganic counterparts, ECPs have been subjected to significant synthetic effort during the last decade, and soluble systems with new colors, higher contrasts and improve d ambient stability have been successively developed.12,14 In parallel, some of the most effective strategies described in the construction of organic electronic -based devices have been applied to polymer ECDs, enhancing their lifetime and performance considerably.12 Given their relative ease of synthesis in general, ECPs are now only a step away from being accessible for technological development and transferred to everyday products. The potential of ECPs for practical scalability is an additional major appeal when compared to their inorganic counterparts12,13,58 in applications comprising large area information displays or portable devices made of finely printed pixel arrays. 1.2.5.3 Electrochromic d evices (ECDs) An electrochromic device consists of an electrochem ical cell where one or several electroactive components undergo an optical change (occurring in the visible or not) on application of an external bias.13 While an ionic conducting interface allows the various electroactive constituents of the device to communicate, the performance of an ECD is commonly diffusion -limited and depends on the nature of this interface, as well as that of each electroactive constituent employed. For instance, liquid -crystal displays (LCDs) where the molecules react quasi instantaneously to the electrical bias applied exhibit much shorter switching times than metal -oxide based ECDs which commonly require multiple seconds to fully switch between colored states. On the other hand, unlike their LCD and LED counterparts, non -emissive electrochromic technologies present the advantage of being viewable from practically any angle or lighting conditions including direct sunlight. Considering the potential of -conjugated polymers for cost -effective processing of multicolored patterns which can be

PAGE 51

51 obtained without further addition of dyes (required in LCD technologies for example), the escalating research effort observed in device architecture over the last dec ade, enhancing response times and lifetimes, is impressive, yet not surprising. B eing operational under a wide range of viewing angles and lighting conditions including direct sunlight, low voltage driven polymeric ECDs promise to impact the development of transmissive and reflective color changing systems spanning smart polychromic glass technologies12,56 and e -papers.13,57 1.2.5.4 Absorption/Transmission ECDs An absorption/transmission ECD switches reversibly between a state of high optical density (colored) and a more transmissive state (ideally transparent) on application of an electrical bias. To take advantage of the full contrast in an EC cell, the color -changing materials are sandwiched between two transparent substrates as illustrated in Figure 1 1 8 a. For the same reason, the electrode materials must be transmissive to the wavelength range of interest. Smart windows, optical shutters and window type displays are the main target applications.13 For reasons relating to chemical compatibility and redox stability issues, ECDs with good switching performances can only be obtai ned when using a counter electrode stable in the potential window of application of both the EC material deposited at the working electrode and the supporting electrolyte.59 61 Hence, pairs of complementary EC materials must be found to not only address the color requirements according to the device made but also balance the redox processes occurring in the cell to enhance its lifetime on repeated cycling. In most cases, the constraint for contrast maximization prevails and one EC constituent has to be cathodically -coloring while the second is anodically -coloring, thereby exhibiting simultaneous colored to transmissive switches (under opposite redox processes). Alternatively, the charge storage layer can be non-coloring, yet electroactive

PAGE 52

52 1.2.5.5 Reflective ECDs With most -conjugated ECPs possessing high contrast ratios in the near IR owing to the charge -carrier associated low -energy transitions arising and depleting on doping and de -doping, electrochromism is not limited to the visible region of the electromagnetic spectr um. In fact, various reports have proposed to take advantage of these important optical changes occurring in the near -, mid IR and microwave regions. For example, Reynolds et al. have incorporated high contrast polymeric electrochrome s62,63 in reflective device architecture s consisting of an outward facing active electrode bearing the ECP of interest ,64 and originally reported by Bennett et al.65 and Chandrasekhar et al.66 in the patent literature ( as illustrated in Figure 1 -1 8 b ). a) b) Figure 1 18. a) Schematic illustration of a convention al dual polymer transmission ECD (not to scale) b) Schematic illustration of a reflectivetype PECD using a porous membrane electrode (not to scale) In the presented device, the electrodes were made of gold -coated Mylar, the working electrode was slitted to account for ion diffusion and a highly viscous polymeric electrolyte consisting of LiClO4, ACN, PC and PMMA was employed that further induced self -sealing of the structure (by evaporation of ACN). In analogy with absorption/transmission devices, a comple mentary ECP is deposited onto the counter -electrode to account for the redox balance and

PAGE 53

53 Faradaic reversibility of the structure. R edox switchable refle ctivity extending into the IR is, for example, desired for thermal control applications (though reflecti ve devices are also suitable for displays). 1.3 Thienylene Benzothiadiazole Based Polymers (PTBTDs) As succinctly mentioned in the solar cell device section above, the donor acceptor (DA) approach introduced by Havinga e t al. in macromolecular systems,42 brings electron -rich and poor heterocycles along the same backbone and produces semiconducting -conjugated polymers with high electron affinity and low ionization potentials Their resulting narrow energy gaps allows them to absorb at longer -wavelengths than their wide r bandgap all -donor parents such as P3HT, MEH -PPV and MDMO -PPV .4348 Of all DA copolymers present in the literature, those containing thienyl ene building units and 2,1,3 benzothiadiazole (PTBTDs) are meeting significant impact in the field of -conjugated organic electronics for high -performance device applications. Starting with work from Yamashita et al. on the elect ropolymerization of narrow bandgap copolymers of aromatic donors and quinonoidacceptors,6769 PTB TD -type semiconducting polymers have been used as a platform to demonstrate the relevance of the donor acceptor approach in producing pol ymers with especially narrow bandgaps. Figure 1 -1 9 illustrates the representative analogues initially developed by Yamashita and coworkers I mportantly, it was observed that the strength of the acceptor (i.e. its electron affinity ) increases with increasin g number of fused heterocycles and higher nitrogen content the most electron -defi cient being the 3 ring benzobisthiadiazole fused heterocycle For example, a difference of ca. 1eV was found in the reduction potentials electrochemically estimated for the b enzothiadiazole and the benzobisthiadiazole -containing polymer ana logues whereas less than 0.3 eV separated their

PAGE 54

54 oxidation processes. Overall, w ith the presence of an electron -poor substituent every two electron -rich heterocycle s the polymer LUMO (as re flected by the onset of reduction) is significantly represse d and its b andgap reduced when compared to the all donor parent polymer s As a result, the optical absorption spectrum of the subsequent polymers exceeds the visible and extends into the near -IR. Figure 1 1 9 Representative electropolymerized PTBTD type narrow bandgap polymer analogues initially developed by Yamashita et al. from aromatic -donors and quinonoidacceptors67 69

PAGE 55

55 Based on this design, Krebs et al. have produced soluble PTBTD derivatives (see Figure 1 20), and have reported on their optical and physical properties, revealing bandgaps ra nging from 1.7 to 2.1 eV for the BTD -based polymer analogues and as low as 0.7 eV for the ones containing the strongest acceptor ( i.e. benzobisthiadiazole).47 Figure 1 20. Representative chemically po lymerized PTBTD type narrow bandgap polymer analogues developed by Krebs et al. from aromatic -donors and quinonoid acceptors47 Employed as semiconductor s in O FETs and PSCs by various research groups from both academia and industry ,41,44,49,7074 polymers of dialkyl -substituted cyclopentadithiophene (CPDT) and 2,1,3 -benzothiadiazole (see Figure 1 2 1 ) are proving especially effective for their high charge -carrier mobilities (h ighly dependent on their solubilizing groups) and their long wavelength absorption bandwidths extending into the near -IR. Further, their energy band structures are particularly well suited for conventional device architectures comprising ITO, PEDOT:PSS, Al and PCBM in solar cells, and Au in FETs. For example, in bulk -heterojunctions with PCBM, CPDT BTD based semiconducting polymers have shown power conversion efficiencies as high as ca. 5.5% through controlling the thin-film blend morphology with the prop er additives .71

PAGE 56

56 Figure 1 2 1 Solution -processable CPDT BTD type narrow bandgap polymer analogu es as de veloped by Brabec et al. and further investigated by various groups such as Bazan, Heeger, Mllen et al.41,44,49,7074 Alternatively, in work from Janssen et al. t hienylene diketo pyrrolo -pyrrole copolymers (PTDPP, see Figure 1 2 2 a ) have appeared as excellent candidates for solar cell applications with PCEs as high as 4.0 % under AM 1.5 solar illumination when processed into semi -crystalline thin -films from a polymer dispersion.43 Further, with an analogou s copolymer hybrid (see Figure 1 2 2 b ), Winnewisser et al. have reported field effect mobilities as high as 0.1 cm2 V1s1 and 0.09 cm2 V1s1 for holes and electrons, respectively .75 In t hese case s BTD was replaced by the accep tor 3,6 -diaryl 2,5 dihydro -pyrrolo[3,4c ]pyrrole 1,4 -dione (DPP) an electron -withdrawing moiety that has commonly been employed in the synthesis of high -performance dyes and pigments for paints, plastics and inks .76 Importantly, t he photochemical, mechanical and thermal stability of DPP -based materials in general along with their strongly absorbing and fl uorescent character m ak e them suitable for high performance electronic devices For instance, a number of oligomers have also recently been probed in solution -processed FETs and organic solar cells by Nguyen et al. (see Figure 1 2 2 c ).7678 Overall, DPP -based conjugat ed systems introduce new synthetic design to the donor acceptor polymer toolbox with the perspectives of enhancing the performance of their BTD -containing counterparts in devices necessitating efficient charge transport

PAGE 57

57 Figure 1 2 2 C hemicall y polymer ized PTDPP type narrow bandgap poly mer analogues developed by Janssen et al.43 and Winnewisser et al.75 have recently emerged as an alternative to the now more established PTBTD s The well -defined DPP -based oligomers developed by Nguyen et al. constitute an attractive alternative to their relatively disordered polymer homologues7678 1.4 -Conjugated Polymers in this Disssertation With s pectral e ngineering in -c onjugated organic polymers as the main approach angle the following Chapters of this dissertation encompass t he synthesis and characterization of novel -conjugated semiconducting polymers all thienylene benzothiadiazole derivatives (PTBTDs) exhibiting original optical properties, and intentionally designed to find applications in optoelectronic systems such as electrochromic and photovoltaic devices While Chapter 2 describe s the toolbox of characterization techniques and experimental methods essential to understand the various directions taken throughout this dissertation work, Chapters 3 to 7 ar e dedicated to discussing the specificities and the results obtained for a number of research projects achieved over the years More precisely, Chapter 3 introduces a synthetic approach to producing

PAGE 58

58 neutral state green cathodically -coloring electrochromic polymers ( poss essing a dual band of absorption in the visible spectrum) with especially high contrast ratios and longterm stability on repeated electrochemical switching. Chapter 4 proposes a synthetic methodology to manipulate dual -band optical absorption s such as tho se exhibited by the polymers described in Chapter 3 and demonstrates its impact on the neutral state color of the resulting polymers. This approach involving careful structural modifications of the polymer repeat unit is further extended to the synthesis of the first polymer electrochrome absorbing homogenously across the visible spectrum, hence being essentially black in the neutr al state. Importantly (when considering display applications), the same black electrochrome can be easily oxidized on applicati on of an external bias to reveal a transmissive state with minimal residual hue Chapter 5 investigates a more fundamental approach to producing new color patterns which involves probing the effect of incorporating unsaturated linkers (namely ethylene and ethynylene) in cathodically -colorin g polymer electrochromes. Here, t he interest lies in the perspectives of copolymerizing a multiplicity of chromophores via the said linkers to produce unprecedented palettes of colors. Based on the design rules developed in Chapter 3 and 4, Chapter 6 and 7 (respectively) propose synthetic methodologies directed towards producing photovoltaic materials combining: 1) either a two -band optical absorption with a window of transmission restricted to the green region, or a spect rum homogenously extended across the visible to reflect the color black, and 2) substantial charge -carrier mobilities in order to promote the charge generation / transport processes at work in organic solar cells. In general, the polymers are systematicall y characterized in terms of their fundamental physical and electrochemical properties, with a special emphasis on elucidating their energy band structure (following frontier molecular orbital theory principles) when the focus is placed

PAGE 59

59 on light -harvesting device applications All devices are constructed in collaborative effort with various well -established research groups. Finally, Chapter 8 gives an overview of the work achieved, and suggests further perspectives in the field.

PAGE 60

60 CHAPTER 2 CHARACTERIZATION T ECHNIQUES AND E XPERIMENTAL METHODS 2.1 Introduction In Chapter 1, the specifics of bandgap control in -conjugated polymers including the fundamentals of the donor acceptor approach as introduced in macromolecular systems by Havinga et al. and pursued by Yamashita et al. and others, were described with an emphasis on the possible device applications for t he subsequent semiconducting materials. From the selected number of applications presented, it appears that conducting polymers have gained in visibility over the past few years as the level of understanding of their s tructure -property relationships improv ed Considering the l arge number of synthetic and characterization tools which have been developed by chemists, physicists and material scientists among the various fields of investigation in order to reach or exceed the expected material performance, the useful approaches must be identified and appreciated in terms of the nature and accuracy of the information that they provide In this section, the toolbox of characterization techniques and experimental methods employed during the course of this dissertat ion work is described. More details on the synthetic aspects related to the monomer production can be found in various dissertations from the same group including that of Benjamin D. Reeves (Processable disubstituted poly(propylenedioxythiophenes, Ph.D. Dissertation, University of Florida, Gainesville, FL, 2005) and Christophe R. G. Grenier (Synthetic control of order in soluble dioxythiophene polymers, Ph.D. Dissertation, University of Florida, Gainesville, FL, 200 6 ). 2.2 General Synthetic Methods The solvents were distilled and dried using known methods.79 All reactions were carried under argon atmosphere (using schlenk line techniques) unless otherwise mentioned.1H NMR and 13C NMR spectra were collected on a Mercury 300 MHz using CDCl3 and were referenced

PAGE 61

61 to the solven t residual peak (CDCl3: 1H: = 7.26 ppm, 13C: = 77.23 ppm). Elemental analyses were carried out by Atlantic Microlab, Inc. High resolution mass spectrometry was performed by the spectroscopic services at the Department of Chemistry of the University of Fl orida with a Finnigan MAT 96Q mass spectrometer. Gel permeation chromatography (GPC) was performed using a Waters Associates GPCV2000 liquid chromatography system with its internal differential refractive index detector (DRI) at 40 oC, using two Waters Sty ragel HR 5E columns rate of 1.0 mL / min. Injections were made at 0.05 0.07 % w/v sample concentration using a ated against narrow molecular weight polystyrene standards (Polymer Laboratories; Amherst, MA). Important mechanistic details about some of the most representative synthetic steps mentioned throughout this dissertation are illustrated below: 2.2.1 Selected Palladium -Mediated Cross -Couplings 2.2.1.1 Stille cross -c ouplings While Eaborn et al. reported the first palladium catalyzed reaction of an organotin (organostannane) as early as in 1976, it was only a year later that Kosugi and Migita described the cross coupling of organotins with aryl halides assisted by transition metal. During the early 1980s, Stille et al. invested significant effort in developing mild conditions and promoting the yields for the Pd(0)-mediated reaction between organotins and electrop hiles. The subsequent C C sigma bond formation is now commonly named Stille cross -coupling. The catalytic cycle for the cross coupling of organotins with aryl halides is illustrated in Figure 2 1. The co upling consists of a three-step mechanism involvin g: 1) oxidative addition of the aryl halide (generally the most electron deficient species) onto the Pd(0), 2) transmetallation as the organostannane enters the catalytic cycle (rate -determining) and 3) reductive elimination regenerating the

PAGE 62

62 catalyst Imp ortantly, aryls transmetallate to the Pd(0) much faster than alkyls which accounts for the trialkyltin halides being the outcome of the transmetallation step. It is worth noting that, in contrast with conventional Suzuki conditions (see below) Stille cros s -couplings do not necessitate the use of a Lewis base to activate the tin moi ety for transmetallation Further, the reacti on is functional group tolerant and can be run at ambient conditions without altering the reaction yields significantly. However, the toxicity of the tin reagents employed in the synthesis of organotins (trialkyltinchlorides are common) as well as that of the tin halides produced during the transmetallation step, constitute an important obstacle to large scale manufacturing. The tin by-products are further often difficult to isolate from the final product, which often hinders the use of Stille cross -couplings in the synthesis of active molecules with medical applications Finally, as organotins can undergo oxidative homocouplings, the t in reagent is often loaded in excess. Despite the above mentioned hindrances, the Stille cross -coupling approach remains one of the most effective and extensively used method in the C -C sigma bond formation. Figure 2 2 shows the step -growth mechanism commo nly proposed to illustrate Stille type polycondensations. Over the course of this dissertation work, several parameters were found to impact the outcome of palladium -mediated Stille -type polymerizations, including 1) the choice of the palladium catalyst 2 ) the quantity of added catalyst, 3) the monomer stoichiometry, 4) the solvent mixture employed, 5) the reaction temperature, 6) the r eaction atmosphere, and 7) the polymerization time. In general, 2,1,3 benzothiadiazole -containing oligomers and polymers w ere found sensitive to refluxing conditions in THF (degradation observed by sudden or progressive color change of the reaction mixtures, e.g. from dark blue to dark purple -red), in contrast with mixtures of dry and degassed toluene and DMF (8:1.5 volume ra tio ) which were found

PAGE 63

63 especially suitable at reaction temperatures on the order of 85 90 oC. For the catalyst, a combination of Pd2dba3 (2mol% to the monomer) and the bulky ligand P( o -Tol)3 (8mol% to the monomer) was found especially reactive with respect to the substrates considered (see Chapter s 6 and 7) when compared to the more traditional PdIICl2( PPh3)2 and Pd0(PPh3)4. Using those conditions, in most instances, only 15 20 min were necessary to form the desired polymer in high molecular weights (with t he highest molecular weight precipitating out of solution rapidly), whereas the more conventional conditions would usually require at least 24 hours to yield reasonable molecular weight ( Mn > 10,000 g mol1). This could be due to the high turn over of the bulky ligand based catalyst, further reducing the concentration of coupling by-products and ligand end -capping processes. Expectedly, longer reaction times were found to narrow the polymer polydispersity ( PDI ) to values on the order of 2 3 (in absence of s ubsequent polymer fractionation beyond isolation by precipitation in MeOH ). As in any palladium -mediated step growth polymerization process, careful respect of the monomer to -comonomer stoichiometry was essential, and an oxygen-free environment increased t he lifetime of the catalyst. Figure 2 1. Catalytic cycle for the cross -coupling of organotins with aryl halides (electrophiles) according to the Stille cross -coupling conditions

PAGE 64

64 Figure 2 2. Step growth mechanism commonly proposed to illustrate Stille type polycondensations 2.2.1.2 Suzuki cross -c ouplings The palladium -mediated and stereoselective C -C bond forming reaction between an aryl halide and an alkenylborane was first described in 1979 by Miyaura and Suzuki. The subsequent sigma bond formation i s now commonly named Suzuki cross -coupling. The catalytic cycle for the cross coupling of organoboranes with aryl halides is illustrated in Figure 2 3 The coupling consists in a four -step mechanism involving: 1) oxidative addition of the aryl halide (ge nerally the most electron deficient species, although the swap is possible as shown below) onto the Pd(0), 2) metathesis (anion exchange between Pd(II) and the base) 3) transmetallation as the base activated organoborane (organoborate) enters the catalytic cycle (rate -determining), and 4) reductive elimination regenerating the catalyst. Importantly, aryls transmetallate to the Pd(0) much faster than alkyls which accounts for the trialkyltin halides being the outcome of the transmetallation step. Organoboronic acids and esters do not transmetallate readily, and require quaternization of the boron.

PAGE 65

65 In contrast with the Stille approach (see above), Suzuki cross -couplings which necessitate the use of a Lewis base are less tolerant to the presence of unprotected functional groups. However, organoboranes are often commercially available and are considerably less toxic than organotins, such that this technique is commonly employed for large scale manufacturing. The inorganic salts generated over the course of the re action can be easily removed and the product islotated. Then again, the reaction cannot be run at ambient conditions, and has to be kept under inert atmosphere at all times to avoid altering the reaction yields significantly as self -coupling by -products ar ise from the presence of oxygen. The oxidative addition of aryl halides to the Pd(0) can be an expecially slow process, such that the cross -coupling tends to be highly substrate dependant. Despite the above mentioned hindrances, the Suzuki cross -coupling a pproach probably remains the most effective and extensively used method in the C C sigma bond formation between aryls Suzuki -type polycondensations are commonly described to follow a step growth mechanism equivalent to that depicted in Figure 2 2, albeit following the catalytic cycle shown in Figure 2 3. Despite the general case, it is worth noting that Suzuki polycondensation have recently been shown to undergo chain -growth polycondensations on addition of a carefully chosen initiator.80 It is also worth no ting that base -free Suzuki polymerization conditions have recently been developed b y Brookins et al. whereby fluorine based nucleophilic sources can be used to activate the organoborane.81 In analogy to the comments previously made with regard to Stille type polycondensation reactions, over the course of this dissertation work, several parameters were also found to impact the outcome of palladium -mediated Suz uki type polymerizations, including 1) the choice of the palladium catalyst, 2) the choice of the base 3) the monomer stoichiometry, 4) the solvent mixture employed, and 5) the re action atmosphere The same combination of Pd2dba3 (2mol% to

PAGE 66

66 the monomer) an d the bulky ligand P( o Tol)3 (8mol% to the monomer) was found especially reactive with respect to the substrates considered, when compared to the more traditional Pd0(PPh3)4, although, the best reaction solvent medium remained the biphasic toluene/water mi xture (9:3 volume ratio) gently refluxed at 95 oC. Expectedly, the presence of a phase transfer agent (Aliquat 336) was found to accelerate the polymerization process. Similarly longer reaction times were found to narrow the polymer PDI to values on the order of 2 3 (in absence of subsequent polymer fractionation). As in any palladium -mediated step -growth polymerization process, careful respect of the monomer to -comonomer stoichiometry was essential, and a drastic oxygen-free environment was a sine qua non condition for the reaction to occur Although no synthetic example will be detailed within the next Chapter s of this dissertation work, Figure 2 3 illustrates the initiation step (formation of AB block) for one of the most successful Suzuki polymerization carried out along this graduate work The halide presented in Figure 2 3 was further replaced by an ester -functionalized analogue and successfully polymerized with the same organoborane via a base -free condition inspired from those developed by Brookin s et al. (here CsF was used) .81 Importantly, the residuals of palladium catalyst potentially remaining after precipitation of the polymer s in MeOH, and subsequ ent overnight wash with MeOH in a Soxhlet extraction vessel, were removed using a palladium scavenger (diethylammonium diethyldithiocarbamate) in an approach similar to that previously described by Krebs et al.82 Traces of palladium are frequently pointed as being particularly difficult to remove efficiently. For instance, an investigation of several interesting methods including the use of activated ch arcoal in reflux with the polymer, is available from the dissertation of Peter B. Balanda (Synthetic of Functionnalized Poly p Phenylenes via Palladium

PAGE 67

67 Acetate Catalyzed Suzuki Cross Coupling Polymerization, Ph.D. Dissertation, University of Florida, Gai nesville, FL, 1997). Figure 2 3 Catalytic cycle for the cross -coupling of organoborane s with aryl halides (electrophiles) according to the Suzuki cross -coupling conditions 2.2.2 Oxidative Polycondensation s The f erric chloride -mediated oxidative polym erization of thiophene monomers was proposed by Sugimoto in 1986.83 With the perspectives of using ambient conditions (room temperature, non oxygen/ moisture -free atmosphere) in the synt hesis of -conjugated polymers, the method gained in popularity, and H.C. Star c ks for example is now producing conducting polymers such as poly(3,4 -ethylenedioxythiophene) poly(styrenesulfonate) (PEDOT -PSS), namely Baytron P, based on this scalable tech nique In brief, oxidative polycondensations are believed to occur by successive combinations of cation radicals formed at the monomer stage, and then at the chain -ends of the growing macromolecules in the presence of chemical oxidizing

PAGE 68

68 agent (see Figure 2 4 ). The dicationic species formed shall subsequently release two molecules of hydrochloric acid (HCl) in order for the overall aromaticity of the system to be reestablished. As a result, it is worth noting that large quantities of HCl are produced during the course of a chemically induced oxidative polymerization, which may arguably overoxidize the growing polymer, introduce molecular defects ( e.g. via the presence of cross -links) and alter its overall quality if the molecular structure of the resulting po lymer is prone to undergo such irregularities or if the experimental conditions are poorly designed. It is worth noting that if the polymer chain -ends should expectedly consist of hydrogen atoms, the chloride anions within the reaction mixture have been s hown to constitute a source of end-capping for the growing polymer,84 hence potentially limiting its molecular weight. Over the course of this dissertation work, a large number of parameters were found to impact the outcome of ferric chloride -mediated oxidative polymerizations, including 1) the method of adding ferric chloride (add oxidizing agent to monomer vs. monomer to oxidizing agent, addition solvent), 2) the rate of addition of the oxidizing agent, 3) the quantity of oxidizing agent added, 4 ) the reaction temperature, 5 ) the reaction atmosphere, and 6 ) the reacti on time. In general, longer reaction times were found to narrow the polymer polydispersity (PDI ) to values as low as 1.5, while room temperature reaction conditions were sufficient to initiate and carry the polymerization to completion ( Mn > 10,000 g mol1). In parallel, the HCl formed during the polymerization process was continuously removed by passing a relatively strong air flow thro ugh the reaction solvent during the first couple of hours, which air flow was then reduced to limit excessive solvent ev aporation. Ferric chloride (ca. 5 equivalents) dissolved in nitromethane (NO2CH3), was added dropwise (over a period of 30 min to 1 hour) to the vigorously stirred monomer/oligomer solution T he oxidizing agent would then recrystallize

PAGE 69

69 throughout the ad dition process, hence potentially increasing the surface area of the catalyst for the reaction. Figure 2 4 Proposed mechanism for the oxidative chemical polycondensation of easily oxidized monomers 2.3 Electrochemical Methods The redox properties of -conjugated polymers encompassing onsets of oxidation and reduction, half -wave potentials and electrochemical bandgaps, for instance, can be conveniently examined in simple electrochemical cells consisting of an electroly te solution, a working electrode, a counter -electrode and a reference electrode. Throughout this dissertation, the three electrode cell is generally composed of a platinum button or an ITO (Indium Tin Oxide) -coated glass slide as the working electrode (7 50 0.7 mm, sheet resistance, Rs = 8 2), a platinum wire or flag as the counter electrode, and a silver wire pseudo reference electrode calibrated vs. Fc/Fc+ or a Ag/Ag+ reference electrode. The ITO electrodes were purchased from Delta Technologies, Ltd. Acetonitrile (ACN, dried over CaH2) and propylene carbonate (PC)

PAGE 70

70 were distilled prior to use. Tetra n -butylammonium perchlorate (TBAP) was synthesized by the metathesis of tetra n -butylammonium bromide (98 %, Sigma -Aldrich) and concentrated perchloric acid ( uct was then recrystallized three times from isopropyl alcohol and dried in a vacuum oven for 3 days. Lithium tetrafluoroborate (LiBF4) (anhydrous, 99.998%, Aldrich) was used as received. The electrolyte salts were stored in an argon -filled dry box. All ot her reagents and starting materials were purchased from commercial sources and used without further purification, unless otherwise noted. All electrochemical measurements were collected via a EG&G Princeton Applied Research model 273A potentiostat / galvan ostat operated under the control of Corrware II software from Scribner and Associates. Polymer thinfilms were obtained by drop -casting a toluene solution of the polymer onto the working electrode. HOMO and LUMO energy levels are derived from the electroch emical data considering that the SCE is 4.7 eV vs. vacuum85 and Fc/Fc+ is 0.38 eV vs. SCE,86 i.e. 5.1 eV relative to vacuum (see Figure 2 5 ). Figure 2 5. Converting electrochemical potentials measured as a function of a particular reference electrode to a different reference electrode (left axis). HOMO and LUMO energy levels (relative to vacuum) can also be extrapolated from this diagram (right axis). T he reported values were extracted from the literature8587

PAGE 71

71 Here, it is worth noting that this value of 5.1 eV for the energy level of Fc/Fc+ relative to vacuum is significantly higher than the 4.8 eV commonly used by a large number of research groups in the field. An extensive discussion regarding the ori gin of this likely underestimated value of 4.8 eV can be found in the dissertation of Barry C. Thompson (Variable bandgap poly(3,4 alkylenedioxythiophene) -based polymers for photovoltaic and electrochromic applications, Ph.D. Dissertation, University of Florida, Gainesville, FL, 2005, p.5962). 2.4 Spectroscopic Methods suitable for ECP Characterization 2.4.1 Spectroelectrochemistry vs. Luminance Measurements Commonly reported as a perce nt transmittance change ( %T ) at a given wavelength, the electrochromic contrast is a primary tool in the overall characterization process of an electrochrome. While the control wavelength chosen is often the one at which the electrochrome exhibits its highest optical contrast, th is EC contrast can practically be monitored and reported at any local absorption maximum, when desired. Transmittance values are generally recorded upon application of square -wave potential steps to the electroactive film placed in the beam of a spectrophotometer. Alternatively, spectroelectrochemical experiments monitoring the formation of ionic states, namely polarons (radical -cations/anions) and bipolarons (dications/dianions), upon progressive application of an electrical bias, can be used to evaluate t he transmittance changes undergone over a broad range of wavelengths (see Figure 2 6 ). As the long-wavelength optical transitions associated with the charge -carrier formation arise, EC contrasts can generally be evaluated in the near -IR as well. In -conju gated polymers, the extent of depletion of the neutral state interband(s) transition(s) greatly depends on the degree of stability attained by the polymeric backbone in its quinoidal geometry which relates to the formation of charge -carriers.

PAGE 72

72 Here, the spe ctroelectrochemical data were obtained using a Varian Cary 500 UV -vis NIR spectrophotometer. The rapid absorbance measurements performed at the polymer max in order to control the thickness of the films while spraying were carried out on a Genesys 20 Spectrophotometer (Thermo Electron Corporation). Figure 2 6 Spectroelectrochemical experiments monitoring the formation of ionic states, namely polarons (radical -cations/anions) and bipolarons (dications/dianions), upon progressive application of an electrical bias, can be used to evaluate the transmittance changes undergone over a broad range of wavelengths Another practical approach to evaluate EC cont rasts in the case of colored -to transmissive switching electrochromes is provided by measuring the relative luminance change ( Y ) as a function of redox doping (see Figure 2 7). Relative luminance values describe the overall contrast of a material across the visible region of the electromagnetic spectrum taking into account the sensitivity of the human eye in this range and rel y on a calibrated light source. Y values become particularly useful in the characterization of broadly absorbing materials where no distinct absorption maximum can be unambiguously chosen.

PAGE 73

73 Figure 2 7. Relative luminance as a function of applied potent ial for various film thicknesses of a polymer solution-cast on ITO -coated glass (with the thickest film represented by the green curve). Relative luminance values describe the overall contrast of a material across the visible region taking into account the sensitivity of the human eye 2.4.2 Colorimetry In 2000, Reynolds et al. introduced a methodology aimed at describing the different color states exhibited by an ECP on electrochemical switching based on the CIE (Commission Internationale de lEclairage) standards for color matching.63,88 As for a dye or pigment to be employed in commercial products, the potential of -conjugated polymers for application in multi -chromic display devices relies on the ability to precisely define their color based on the use of at least one of the several color spaces available.89 Fur ther, to avoid falling into a mode of trial and error when fabricating devices, it is essential to have rapid access to the color state of a deposited polymer film which is to an extent controlled by its thickness. Finally, since ECPs are primarily desirab le for their color changes, it is essential to map the color path followed by the polymer as well as its light transmission properties as a function of the doping level attained. These aspects have encouraged the group to develop a protocol especially suit able for polymeric electrochromes and consisting in using a portable colorimeter calibrated to the

PAGE 74

74 sensitivity of the human eye which not only accounts for the changes of color and light transmission observed but allows selecting and varying the illuminati on conditions as well. Further, the measurements can be subjected to practically any cell configuration, whereas obvious constraints and limitations arise from the use of a spectrophotometer as previously proposed.90,91 In particular, the use of a colorimeter allows visual tracking of the sample color both in a reflectance and a transmission mode depending on the considered application. The CIE defined L*a*b* color space introduced in 1976 is employed as a means to define the color/light transmission changes undergone by the ECP as a function of the external bias applied. The L*a*b* values associated with a colored state are calculated from the Yxy tristimulus values determined by the colorimeter where Y is the luminance and xy the two dimensional set of coordinates defining the hue and saturation of the color state at the given Y value. Note that the relative lightness of a color Y is commo nly described as a percentage of light transmitted with respect to that coming from a light source according to Equation 2 1 :89 % Y = ( Y / Y0) x 100 2 1 The proposed in situ colorimetric analysis has pr oven to be useful in the systematic study of electro -optical property relationships of -conjugated polymers in general as well as in understanding the complex color changes in multilayer electrochromic devices. This method allows the construction of color databases from single colored polymers and their resulting hues when in combination in devices, hence completing the toolbox for polymer ECD characterization. In this dissertation work colorimetry was carried out using a Minolta CS 100 Chroma Meter. The samples were illuminated from behind with a D50 (5000K) light source in a light booth designed to exclude external light.

PAGE 75

75 2.4.3 Coloration Efficiency The proportionality factor that relates the optical absorbance change of an electrochrome at a given wavelength ( A ) to the density of injected/ejected electrochemical charge necessary to induce a full switch (Qd) is called coloration efficiency (CE). CE values are inherent to the electrochromic material under characterization, some being intensely colored by nature in at least one redox state while others display fainter tones over their doping/dedoping cycles, regardless of the deposited film thickness. Derived from the Beer -Lambert law, the quantitative measure of the electrochromic contrast observed CE is estimate d through E quation 2 2 : CE A / Qd = log10 ( Tox/ Tneut) / Qd 2 2 where Qd is in C cm2, CE is in cm2 C, Tox and Tneut are the dopedand dedoped-state transmittance values, respectively. Given the above equation, a parameter affecting CE values substanti ally is the level of transmissivity attained in the one extreme state of the electrochrome, such that CEs become especially useful for the characterization of colored to transmissive switching materials. Upon doping, the charge -carrier transitions with sig nificant overlap in the visible region tend to lower the CEs of the corresponding electrochromes. However, it is worth noting that low CE values do not always translate into modest optical changes as in the case of colored EC species switching to a differe nt color state on doping. In spite of their well -established utility in this field, the determination and further comparison of CE values between works suffer from the different methods available to measure them. In addition, most of the perceptible EC con trast is obtained within the first 90 95% of overall optical change such that smaller charge density windows could be used to calculate more representative CE values. One approach for the characterization of the extent of optical change

PAGE 76

76 undergone by an ele ctrochrome at a given wavelength consists in estimating a composite coloration efficiency (CCE) measured at a representative percentage ( e.g. 95%) of the total optical change. Quantitative comparisons between electrochromes can be achieved precisely and re liably by tandem chronocoulometry/chronoabsorptometry during which variations of film transmittance and charge density are simultaneously monitored while the potential is repeatedly stepped from one redox state to the other.12 Fabretto et al. have recently proposed a new approach for measuring the CEs of polymers in ECDs by extracting the Faradaic charge from the total charge involved during a device switch.92 The Faradaic corrected CEs of several control polymers were found to be significantly larger than the CE values commonly described for th e same materials. The same group has also commented on the importance of precisely identifying the conditions within which CE values are reported ( e.g. full contrast vs. portions of the full contrasts).93 2.5 Surfa ce Morphology Characterization Tools Following development of the scanning tunneling microscope (STM) by Binnig and Rohrer in the 1980s, atomic force microscopy (AFM) has rapidly become the imaging approach of choice in the resolution of nanoscale morphologies in polymer networks. As a mechanical probe (cantilever equipped with a nanometer -sized tip) is used to sense the matter by detecting the various forces present at the immediate proximity of the surface ( mechanical contact, Van der Waals, capillary, chemical bonding, electrostatic, among others ), a laser beam is reflected from the top of the probe into a photodiode array detector. In general, a piezoelectr ic stage holds the sample, maintaining a constant force ( and distance) between the sample and the probe, hence preventing the tip from colliding with the matter. The lateral resolution of an AFM can be as high as 1020 when the sharpest tips are used, alt hough lateral resolutions in the order of 100 are common with more conventional tips The depth resolution is always higher and commonly

PAGE 77

77 in the order of 0.1. Imaging can be achieved in contact (static) mode for which the forces induced by the surface ont o the cantilever are repulsive and kept constant or in noncontact (dynamic) mode whereby the cantilever oscillates at a resonance frequency which is modified by the sample -probe interaction (along with the signal amplitude and phase) Alternatively, in t apping mode, the force between an oscillating probe (mounted on a piezoelectric component) in intermittent contact with the sample surface is monitored This technique is particularly suitable for the characterization of soft matter such as polymer network s which can be altered by the AFM tip in the contact mode there by changing the physical significance of the data collected. For this particular reason, the AFM data presented in this dissertation were collected in tapping mode. In recent years, AFM imagin g has been applied to the characterization of donor acceptor blend morphologies in bulk heterojunction solar cells. For instance, Heeger et al. have evidenced improved nanoscale morphology upon annealing devices based on the semi -crystalline semiconducting polymer P3HT in blend with PCBM at elevated temperatures ( ca. 150oC).94 In this important contribution, the post production thermal annealing step lead to a diminution of the series resistance across the device by enhancing the quality of the interface between active layer and collection ele ctrodes. In parallel Yang et al. have monitored the e ffect of growth rate and thermal annealing on the morphology of P3HT/PCBM composite films The corresponding tapping mode AFM images are presented in Figure 2 8 where distinct blend morphologies can be observed, with the most effective in terms of photovoltaic performance being the ones grown at a slow rate and then further annealed to maximum the degree of microstructural organization across the active layer, and further enhance the contact interfaces.95

PAGE 78

78 Figure 2 8. Effect of growth rate and thermal annealing on the morphology of P3HT/PCBM composite films (PCBM concentration = 50 wt%) The AFM height images of the a ) Slow -grown film before thermal annealing. b ) Slow -grown film after thermal annealing at 110oC for 10 min, c ) Fast -grown film before thermal annealing, and d ) Fast grown film after thermal annealing at 110 a) and b) is 0 100 nm, whereas that for films c) and d) it is 0 10 nm (Adapted with permission from Ref.95 Copyright 2005 Nature Publishing Group, a division of Macmillan Publishers Limited) In contrast with AFM, scanning electron microscopy (SEM ) produces two-dimensional images from a high -energy electron beam directed onto samples which must be prepared by metal or carbon evaporation before being placed under high vacuum an approach that can arguably alter the sample morphology. However, it is worth noting that SEM can scan millimeter -sized areas with millimeter -depths of field whereas the maximum image sizes obtained by AFM dont generally exceed 150 m x 150 m with micrometer resolution depths. On the other hand, transmission electron microscopy (TEM) transmits a beam of electrons through an ultra thin layer of matter which beam interacts with the matter itself to generate patterns reflecting the thickness and composition of the material (with varying absorption of electrons in the material). This technique has been widely employed by Heeger et al. over the

PAGE 79

79 past few years in the nanoscale characterization of BHJs .94 In particular, one of the most impressive results obtained by the grou p so far may have been the cross -sectional characterization of the P3HT/PCBM morphology in optimized PV devices, providing information on the charge percolation pathway in BHJs (see Figure 2 9 ).96 a) b) Figure 2 9 a) Schematic of the cross -section of a P3HT/PCBM BHJ device (1:0.7) and photographs representing the focused (left) and defocused (right) cross -sectional TEM ima ges of the corresponding device. b) TEM of a 120 nm thick slice of active layer with the corresponding binary image showing the continuous interpenetrating network favoring the assumption of a direct percolation pathway across the device (Adapted with perm ission from Ref.96 Copyright 2009 American Chemical Society) Probing the density of states in matter via tunneling currents a scanning tunneling microscope (STM) allows surface imaging at the atomic level. In this approach, a bias is applied between a conducting tip held by a piezoelectric tube mounted with electrodes and the metallic or semiconducting sample surface, induci ng electrons to tunnel through the vacuum interspace. The resulting image is then reflected from the variations in current measured as the probe is scanning the sample surface. The resolution of an STM is meant to be in the order of 1 in the lateral direc tion and as high as 0.1 in depth, such that the extent of cleanliness of the substrate,

PAGE 80

80 as well as the quality of the tip are of particular importance. High -vacuum atmospheres give the best results. For instance, STM imaging reveals to be especially usefu l in the visualization of the self assembly of large polycyclic aromatic hydrocarbons (PAH) including 2 -dimensionnal graphenes as frequently demonstrated by M llen et al.25,97100 2.6 Structural Analysis by 2D WAXS T wo -dimensional wide an gle X ray scattering experiments (2D -WAXS) whereby a fiber made from the -conjugated mat erial is placed in t he field of an X ray beam have become a mainstay in the charact erization of the level of organization in semiconducting polymers Throughout several important contributions,73,74 Pisula et al. have demonstrated how aligned polymer samples can be prepared by filament extrusion using the homebuilt mini extruder illustrated in Figure 2 1 0 a ,101 and be further subjected to 2D WAXS analysis. In their device the -conjugated semiconductor is heated to a state (above Tg) in which it can undergo plastic deformations and is then extruded via a piston applying a constant -rate motion (ca. 5 mm s1) along the cylinder -shaped inner part of the nozzle of the extruder Se veral variables rule the extrusion process such as the rate of the piston motion, the inner -diameter and shape of the nozzle, as well as the shape of the die Using their setup, Pisula et al. extrude fibers that are 0.7 mm thick, which diameter correlates with the size of the nozzle orifice. When the semiconducting fiber is placed perpendicular to the X -ray beam, a scattering pattern is produced from which information about the microstructural organization and the degree of macroscopic order present in the extruded organic materia l can be extracted As exemplified by the X ray pattern of Figure 2 1 0 b the level of organization attainable by the semiconducting component (e.g. oligomer, polymer) subjected to the thermal treatment describe above is qualitativel y estimated as a function of -stacking (if any) and chain to -chain distances

PAGE 81

81 (also referred to as lamellar distances or d -spacing ) or intercolumnar distances in the case of a columnar arrangement of discotic species (e.g. hexa peri -benzocoronenes) For a -conjugated polymer, the -stacking distances ( ca. 3.5 to 4.5) are commonly found along the equatorial axis of the pattern and their intensity varies with the degree of crystallinity attained by the polymer for instance (although crystallinity is not a requirement for -stacking ). The chain -to -chain distances are also generally found in the equatorial plane of the pattern (since the chains align in the direction of the extrusion) and greatly depend on the length, bulkiness and concentration of the solubilizing side -chains along th e polymer backbone In -between, an amorphous halo (generally isotropic for polymers) can usually be seen which relates to the polymer side -chains. In contrast, in the case of columnar architectures, intercolumnar interactions are found along the equatorial axis whereas intracolumnar interactions ( -stacking) appear along the meridional axis, which corresponds to the direction of the extrusion. a) b) Figure 2 10. a) Schematic illustration of a semiconducting fiber as extruded by Pisula et al. (Adapted with permission from Ref.101 Copyright 2005 American Chem ical Society) b) Characteristic features of a 2D -WAXS pattern for a -conjugated polymer exhibiting relatively pronounced -stacking interactions One common misunderstanding in our field concerns the variations in molecular ordering /interactions that could potentially arise as a function of the deposition/processing conditions employed. Here, it is essential to realize that only the macroscopic arrangement might be influenced by the deposition/substrate, not the molecular packing itself. 2D -WAXS

PAGE 82

82 experiments provide additional information about the -stacking of the polymers, as opposed to mono -dimensional analysis such as XRD, which can only provide indications relative to the out of -plane configuration in the thin film with respect to the substrate (i.e. re lative to the lamellar spacing) In general, 2D -WAXS experiments performed from extruded filaments are also useful in probing the possibilities of a novel material in terms of macroscopic order, i.e. what degree of order can potentially be observed in devi ces upon controlled processing conditions. It is clear that only a careful processing of the polymer during the device construction can enable the same macroscopic order to be reproduced. 2.7 Charge -Carrier Mobility Measurements in FETs The most conventi onal FET architectures employed with conjugated organic polymers were illustrated in Chapter 1 The next few paragraphs briefly describe the fundamentals of conventional FETs, their mode of operation and how charge -carrier mobilities can be calculated fr om the ir characteristic plots. As the source electrode is at ground, the gate ( VG) and drain voltages ( VSD) can be described as a function of the source With application of a negative gate voltage, an electric field is generated which is perpendicular to the device constitutive layers and produces a buildup of positive charges at the dielectric -semiconductor interface (when semiconductor is p type see Figure 2 1 1 ). Applying a drain voltage simultaneously can induce the charge carriers to migrate from the source to the drain, hence setting the on state (as opposed to the off state when VG = 0 ). Further, a threshold voltage ( Vt) characterizes the drain voltage at which the charge carriers start flowing between source and drain and gives an estimate of t he concentration of traps in the semiconductor. The source -drain current ratio between on and off state s must be as high as possibl e for the information to be easily readable from the transistor (on/off ratios of 103106 are

PAGE 83

83 usually desired). Associate d to the device on/off ratio, the extent of voltage to be applied to operate the device (and its power consumption) depends on the semiconductor charge -carrier mobility (FET) among other variables such as the dimensions of the device Figure 2 1 1 Sche matic illustration of p type (left) and n type (right) field effect transistors (FETs) which, in turn, can be used to determine the hole and electron mobility (respectively) of semiconducting polymers The performance of an FET is commonly illustrated by it s output characteristic curves representing the pl ot of the source -drain current as a function of the drain voltage for various gate voltages applied (see Figure 2 1 2a) At low VSD values, ISD increases in a near linear fashion, and then saturates at highe r VSD. Equation 2 3 where L is the channel length, W the channel width, Ci the capacitance of the dielectric per unit area and FET,lin the charge carrier mobility in the linear regime, describes ISD in the linear region. Hence, plotting ISD as a function of VG at constant VSD allows the determination of FET,lin with the slop of the resulting transfer characteristic curve (see Figure 2 1 2b) being given by Equation 2 4 (when VSD << VG Vt). In the saturation regime (where VSD > VG Vt), the charge-carrier mobility at saturation FET,sat can be calculated from the slope of ISD 1/2 as a function of VG at constant VSD (using Equation 2 5) SD = FET lin [ ( G t ) SD SD 2 2 ] 2 3

PAGE 84

84 SD G = FET lin SD 2 4 SD = FET sat 2 ( G t ) 2 2 5 Figure 2 1 2. a) Output characteristic curves of the source drain current ( ISD) as a function of the drain voltage ( VSD) for various gate ( VG) voltages applied b) Transfer plot (ISD as a function of VG) from which the charge -carrier mobility at saturation can be determined for the semiconducting polymer according to Equation (2) (Adapted with permission from Ref.102 Copyright 2009 American Chemical Society) 2.8 Photovoltaic Devices Figure 2 1 3 illustrates the typical response of a cell under solar illumination (red curve) from which the short circuit current ( JSC), open-circuit voltage ( VOC), and the current and voltage at the maximum power point (Jm and Vm), can be defined. F rom these parameters, the fill factor (FF), and the power conversion efficiency ( PCE ) can be calculated for the de vice: = m SC OC 2 6 = P out P in = SC OC P in 2 7 Importantly, as illustrated in Figure 2 14 the highest device fill factors (0 < FF < 100%) are obtained when the shunt resistance (Rshunt) in reverse bias is maximal ( JSC relatively constant with V ), whereas the series resistance ( Rseries) must be a s negligible as possible at higher voltages

PAGE 85

85 (JSC goes to infinity at VOC) In this ideal situation, the maximum power point is attained by intersecting the straight passing by VOC with that passing by JSC. As previously stated in this dissertation, a low FF is an indication that the charge transport a cross the device is dominated by high concentrations of electronhole recombinations. Figure 2 1 3 Current density response of a PV cell in the dark (blue curve) and under solar illumination (red curve) from which the short -circuit current ( JSC), open-c ircuit voltage ( VOC), and the current and voltage at the maximum power point ( Jm and Vm) can be determined Figure 2 14. The highest device fill factors (0 < FF < 100%) are obtained when the shunt resistance (Rshunt) in reverse bias is maximal ( JSC relat ively constant with V ), whereas the series resistance must be as negligible as possible at higher voltages ( JSC goes to infinity at VOC)

PAGE 86

86 In this dissertation work, t he polymer solar cells were constructed using pre -patterned ITO coated glass substrates (sh eet resistance: 20 2). First, a thin layer ( ca. 30nm) of PEDOT:PSS (Baytron AI4083 from HC Starck) was spincoated, followed by the mixed solution of polymer (4 mg mL1) and PCBM (99% pure, Solenne BV) in chlorobenzene. Alternatively, a MoO3 interlaye r ( ca. 10 nm) was thermally evaporated on the pre -cleaned ITO substrate under a vacuum of 106 torr. After the active layer was cast, it was subjected to a thermal annealing step of 70 C for 45 min. LiF ( ca. 1 nm) and aluminum ( ca. 100 nm) were thermally evaporated at a vacuum of ~106 mbar on top of the polymer:PCBM active layer. Each device pixel has a surface area of 0.04 cm2. J V measurements were carried out using a Keithley 4200 semiconductor characterization system under AM1.5G, 100 mW cm2 illumina tion from a 150 W ozone free xenon arc lamp (from Newport). A Newport 70260 radiant power meter was combined with a Newport 70268 probe used to measure the power densities of the white light illumination. National Institute of Standards and Technology (NIS T) calibrated UV -enhanced silicon and germanium photodetectors were used to calibrate the measurements. The device fabrication was performed under nitrogen atmosphere and the measurments were carried out in air (and without device encapsulation). 2.9 Space Charge Limited Current (SCLC) Modeling A useful approach in the characterization of the charge transport taking place in conventional (vertically stacked) BHJ photovoltaic devices consists in assembling a hole -only device of the donor acceptor blend ( e.g. p olymer and PCBM) along with a control device comprising the semiconducting material alone (see Figure 2 1 ). H ole -only device s rely on the use of a high -lying work -function cathode such as gold (~5.2 eV vs. vacuum) or pal ladium (~5.4 eV vs. vacuum) to prevent the electron injection (into the polymer LUMO) while measuring the

PAGE 87

87 concentration of holes flowing across the device. The anode is generally made of PEDOT:PSS (~5.1 eV vs. vacuum) -c oated ITO (~5.0 eV vs. vacuum). Fi gure 2 1 5 Schematic configuration of a hole only device used in the characterization of the hole transport taking place in the active layer of conventional vertically stacked devices such as BHJ solar cells (e.g. composed of polymer and PCBM) In conventio nal BHJ photovoltaic devices, the device photoresponse is primarily governed by the charge transport in the active layer. Depending on the generation rate of geminate electron -hole pairs and on the mean carrier drift length (or diffusion length of the dis sociated charge carriers), various densities of photocurrent can be collected In particular, t he carrier drift length itself depends on the charge -carrier mobility (of both electrons and holes), as well as on the lifetime of the dissociated carrier s to w hich relates the concentration in recombination of electron and holes representing an important non radiative loss mechanism in BHJ solar cells .50,103 As previously stated (see Chapter 1 ), the charge transport is in fact often limited by the donor component (polymer) in a donor acceptor BHJ device using PCBM such that polymers with low intrinsic hole mobility commonly induce an imbalance with the higher intrinsic electron mobility of PCBM hence producing a buildup of charges and increasing the electron hole recombination s By observing the space -charge limited current (SCLC) in the active layer ( e.g. pristine polymer polymer :PCBM blend) ,103 it is possible to determine the hole mobility in the

PAGE 88

88 phase with the lowest intrinsic charge carrier mobility From a simple J V measurement the hole mobility can be estimated using the Mott Gurn ey equation (2 8) for trap -free SCLC: 3 28 9 d V J 2 8 where is the dielectric constant, is the charge -carrier mobility, d is the sample thickness. From the slop of the device J -V2 plot (see Figure 2 16 as an example ), a charge carrier mobility can be calculated in the vertical direction which is more representa tive than t hat given by a laterally -configured FET when studying the photovoltaic performance of semiconducting polymers as the orientation of the chains relative to the substrate can influence the efficiency of the charge transport with respect to the dif ferent directions considered for the measurements In this example the applied voltage V has been corrected by accounting for the built -in voltage Vbi resulting from the difference in the work -function between electrodes, and the current density is in fac t plotted versus the effective electric field ( Eeff) rather than the applied voltage in order to allow comparison of devices possessing active layers of different thicknesses. Figure 2 16. Current density as a function of the effective electric field in a polymer -based hole only device Inset shows the quadratic dependence of the current density on applied bias; the red line is a fit for using the SCLC model (Adapted with permission from Dr Kaushik Roy Choudhury)

PAGE 89

89 In this dissertation work, the hole -only devices were constructed using the same substrate preparation as that used for the solar cells, using either solutions of the pristine polymers in chlorobenzene (16 mg ml1) or polymer:PCBM mixed solutions (as optimized for solar cell performance). Pd (as opposed to Au, for reasons related the metal work-function as outlined later on in this dissertation) was thermally deposited to serve as the top counter electrode.

PAGE 90

90 CHAPTER 3 SPRAY PROCESSABLE GREEN TO TRANSMIS SIVE SWITCHING 3,4 DIOXYTHIOPHENE 2,1,3 -BENZOTHIADIAZOLE DONOR ACCEPTOR ELECTROCHROMIC POLYM ERS 3.1 Context, Proposed Design and Rational T he last two decades have seen much emphasis be placed on incorporating electrochromic polymers (ECPs) as fast and reversib le color -changing materials in devices such as windows, mirrors and displays.13 While a number of neutral state red and blue conjugated polymers were synthesized and their properties investigated,14 attempts in making saturated green polymers, chemically or electrochemically synthesized, have met limited success104 probably due to the complex nature of the re quired absorption spectrum expected to contain at least two bands in the neutral state of the material with a window of transmission in the 480560 nm region of the visible The concept of valence and co nduction band broadening based on alternating electron rich and poor moieties in conjugated polymers was introduced in 1993 by Havinga et al.42 In spite of the considerable impact of this discovery for device a pplications such as field-effect transistors, light em itting diodes and photovoltaics (see Chapter 1 of this dissertation) only recently has the use of donor acceptor based conducting polymers attracted the attention of research groups with an interest in ECPs For instance, i n 2004, Wudl et al. reported on completing the additive primary color space: red, green and blue (RGB), with the first promising neutral green conjugated ECP, a donor acceptor based material which exhibited a high degree of stabilit y upon electrochemical switching.104 More than just broadening the range of red and blue ECPs available to the color green, in 2004 Wudl et al. introduced the idea that the DA approach could be judiciously utilized in engineering novel EC systems exhibiting hues and saturati ons generally difficult to access. Beside its alternating DA structure, the reported electroactive polymeric

PAGE 91

91 material was designed with two distinct chromophores in order to exhibit the two -band absorption in the visible essential for a system reflecting/t ransmitting green light. While its longer wavelength absorption band was found to deplete extensively upon oxidative doping, the optical transition associated to the smaller chromophore (the one in broken conjugation with the main -chain) could not be bleac hed effectively leaving an oxidized state with remnant color ( see Figure 3 1 ). Even though a soluble version of the neutral green polymer was achieved,105 the persistent brown hue of its oxidized state remained an obstacle that could not be overcome hence hindering the use of this polymer in colored to transmissive switching display devices .105,106 Figure 3 1. The first promising neutral green conjugated ECP by Wudl et al. a donor acceptor based material exhibiting two bands of absorption in the visible spectrum. Even though a soluble version of the neutral green polymer was achieved (left photograph), the persistent brown hue of its oxidized state (right photograph) remained an obstacle that could not be overcome (Adapted with permission from Ref.104 Copyright 2004 WILEY -VCH Verlag GmbH & Co. KGaA, Weinheim, and from Ref.105 Copyright 2005 American Chemical Society) As mentioned above, t o achiev e a neutral state green polymer, it is desired to have a material that absorbs both red and blue light. Fine tuning of the energy gap, along with introduction of an adequate set of absorption bands in the visible spectrum, can be obtained by controlling the structure of the repeat unit that composes the conjugated material. Based on the theoretical work of Salzner et al. in a ddressing the relevance of the donor a cceptor approach for designing organic metals,107,108 the strictly linear and alternating DA systems involving two

PAGE 92

92 distinct energy transitions have been described in the literature, and can be rationalized by considering the presence of low lying unoccupied and localized energy levels within the band gap of the conjugated system in its ground -state. These localized energy levels are easily accessed from the valence band and should have little electron densi ty on the donor substituents thus forming a discrete energy band (whose contribution would stem from the most electron poor heterocycle) lying beneath the conduction band in the orbital correlation diagram of the conjugated polymer.108 The consequences of these assumptions on the optical spectrum of the undoped material described are two -fold: 1) low energy transitions from the valence band to the low lying unoccupied and localized energy levels result in a long wavelength absorption band 2) higher energy transitions from the valence band to the true conduction band are responsible for the presence of a distinct blue -shifted absorption band in the visible spectrum ( see proposed interpretation in Figure 3 2, see Cha pter 4 for a more complete discussion on how twoband absorption s pectra may arise in DA polymers) Figure 3 2. Strictly linear and alternating DA systems involving two distinct energy transitions have been described by Salzner et al.107,108 This figure is proposed here as an illustration of their interpretation based on Molecular Frontier Orbital Theory principles (see Chapter 4 for a more complete discussion on how two -band absorption spectra may arise in DA polymers)

PAGE 93

93 In addition, for a polymeric material transmitting / reflecting green light, obtaining a highly transmissive state upon oxidation is an additi onal major difficulty due to the requirement for simultaneous and efficient bleaching of the two absorption bands that occurs in the blue and red portions of the visible region. An excellent example of the existence of strictly linear and alternating DA polymers involving two distinct energy transitions can be found in work from Krebs et al. on the synthesis and fundamental characterization of low bandgap conjugated polymers based on thiophenes and 2,1,3 benzothiadiazole (BTD) .47 As shown in Figure 3 3 polymer 20 (so-called in the corresponding paper) exhibited a first band of absorption peaking in the UV, and a second one somewhat more intense peaking at about 540 nm in the visible, leaving a window of t ransmission in the 400 500 nm range. Though, i t is w orth noting that no particular emphasis was placed on the two -band spectral features of the corresponding polymer throughout this study. Figure 3 3 An example of the existence of strictly linear and alternating DA polymers involving two distinct energy transitions can be found in work from Krebs et al. on low bandgap conjugated polymers based on thiophenes and 2,1,3 -benzothiadiazole (Adapted with permission from Ref.47 Copyright 2006 American Chemical Society) Following up on work from Wudl et al. a novel approach to solution-processable neutral state green conjugated polymers of tunable hues which possess highly transmissive oxidized states was de veloped in our laboratories. This approach is il lust ra ted in Figure 3 4 and will be discussed below.

PAGE 94

94 Figure 3 4 Engineering solution-processable neutral state green donor acceptor -conjugated polymers exhibiting fast switching properties and highly tr ansmissive oxidized states. The polymers are synthesized on the sole basis of a specific linear alternation of electron -rich and electron -poor building units (Adapted with permission from Ref.109 Copyright 2008 WILEY -VCH Verlag GmbH & Co. KGaA, Weinheim) Figure 3 4 suggests the design of conjugated polymers with a two -band absorption in the visible so as to reflect and transmit the color green in the n eutral state. In this case the acceptor 2,1,3 benzothiadiazole (BTD) is symmetrically fun ction alized with different thiophene -based donor heterocyclic units While the first substitution is expected to have the largest impact on the energy gap of the resulting donor acceptor polymer a second substitution with a more electron rich heterocycle such as 3,4 -propylen edioxythiophene (ProDOT) could be sufficient to induce the bathochromic shift necessary to transfer the two-band optical absorption of the polymer presented in Figure 3 3 from the UV visible to the visible NIR portion of the spectrum. T his would also red-shift the inter -band window of transmission from the blue to the green region, thereby providing the resulting polymer with the desired color. In addition, as demonstrated earlier in our group,110 alkoxy -derivatized ProDOT -based ECPs exhibit sub -second switching times, high EC contrast ratios, redox switching stability, transmissivity of the oxidized state, and

PAGE 95

95 a particular ease in sy nthesis which make it a logical building unit in the design of heterocyclic precursors for colored to transparent switching ECPs. Regiosymmetric alkoxy -substituted ProDOTs also possess a low oxidation potential so that mild oxidizing agents can be used for polymerizing the designed oligomers which limits the risks of cross -linking occuring during the polymerization and induces solution-processability to the subsequent polymers. 3.2 Synthesis and Characterization of 3,4 -Dioxi thiophene 2,1,3 -Benzothiadiazole Polymers 3.2 .1 Synthesis and Physical Characterization Based on the considerations highlighted above, ProDOT building blocks bearing solubilizing side -chains aimed at imparting the subse quent polymers with a dequate solutionprocessability were prepared Figure 3 5 illustrates the synthetic pathway developed to produce the DA copolymers PA and PB Following Ullmanns conditions, 3,4-dibromothiophene 1 was first converted to 3,4 dimethoxyth iophene 2 in a refluxing mixture of sodium methoxide containing methanol and in the presence of a Cu(II) source (here CuO) Next, the Williamson transetherification conditions were employed to form the propylenedioxy bridge of the ProDOT intermediate 3 Oc tyloxy substituents were appended to 3 by nucleophilic substitution of a sodium octyloxy salt in DMF to yield the dioctyloxy -substituted ProDOT derivative 4 in ca. 70% yield. The trimethyltin derivative 5 was obtained upon reacting a monolithiated ProDOT 4 with trimethyltinchloride. The BTD -containing building blocks 6a and 6b were then symmetrically substituted with two electron rich ProDOT stannane species ( 5 ) via Pd -mediated cross -coupling Stille conditions111 in gently refluxing THF to yield oligomers 7a and 7b respectively Column chromatography over silica using mixtures of he xanes and dichloromethane afforded 7a and 7b which were isolated as purple tacky solid s The DA oligomers were homopolymerized at room temperature using the mild oxidizing agent FeCl3 and the resulting polymer s were subsequently

PAGE 96

96 reduced with hydrazine to yield P A and PB Both polymers were precipitated in methanol and puri fied by Soxhlet extraction with methanol before further charact erization Figure 3 5. Synthetic route to ProDOT BTD based donor acceptor -conjugated copolymers PA and PB The structures of the repeat unit of P A and PB were supported by 1H NMR and the quality of the polymers was probed by elemental analysis (see Experimental Section below ). The polymer molecular weights were determined via polystyrene -calibrated GPC with THF as the mobile phase and are summarized in Table 3 1 Number average molecular w eights ( Mn) of 18,900 g mol1 ( PDI =3.2) and 10,300 g mol1 ( PDI =1.7) were estimated for PA and PB respectively It is worth noting that the reported molecular weights were estimated from polymers non -fractionated beyond isolation by precipitation in MeOH hence the relatively large molecular weight polydispersity observed for PA T he minimum average number of repeat units for all the polymers was found to be 8 (P B), correspond ing to a main -chain of nearly 40 aromatic units, which is more than 2.5 times as high as the value for which the electronic properties of DA conjugated polymers are commonly found to saturate (~15 aromatic rings).112 On the other hand,

PAGE 97

97 the average number of repeat units for P A was found to be 16 corresponding to a main -chain of nearly 80 aromatic units, which demonstrated the possibility of producing high molecular weight polymers from the polycondensation of relatively large functional building blocks (as opposed to a conventional copolymerization of single heterocycles). Table 3 1. Number average molecular weight ( Mn, g mol1), weight average molecular weight (Mw, g mol1), polydispersity index (PDI), average number of repeat units, and average number of rings for the donor acceptor copolymers PA and PB Polymer Mn (g mol1) Mw (g mol1) PDI Avg. No. of Repeat Units Avg. No. of Rings PA 18,900 60,500 3.2 16 80 PB 10,300 17,500 1.7 8 40 3.2 .2 Optical Characterization Experimentally, it was found that symmetric addition of ProDOT on the BTD -containing donor acceptor donor precursors (highlighted orange in Figure 3 4 ) induces a bathochromic shift in th e range 55 62 nm for the shorter wavelength absorption band and 75 78 nm for the longer wavelength absorption band (see Figure 3 6 a ). While the first substitution was expected to have the largest impact on the energy gap of the donor acceptor system, the second substitution with ProDOT does account for fine energy tuning and substantial control over the two-band a bsorption spectrum of the resulting oligomers. As conveniently anticipated by results from various research groups,47,112114 polymerization of the resulting heterocyclic pentamers can be expected to give a distinct optical (bathochromic) shift. As illustrated in Figure 3 6 b, p olymerizat ion adds a more significant red -shift of about 80 nm to the shorter wavelength optical transition, while ranging from 120 to 150 nm for the longer wavelength optical transition In solution, both polymers are green and show minor optical changes upon incre ase of the temperature b etween room temperature and 100oC indicating neither significant disruption of effective conjugation ,115,116 nor major aggregation of t he backbones once solvated.117,118

PAGE 98

98 a) b) Figure 3 6. a) Solution optical absorption spectra (in toluene) of Pentamer A and Pentamer B (the oligomer precursors for polymer PA and PB) superimposed to those of their Trimer precursor (Trimer A and Trimer B), b) Solution optical absorption spectra (in toluene) of DA -copolymers PA and PB superi mposed to those of their Pentamer precursor (Pentamer A and Pentamer B). The spectrum of each system is normalized at the longer wavelength absorption maximum (Adapted with permission from Ref.109 Copyright 2008 WILEY -VCH Verlag GmbH & Co. KGaA, Weinheim) Table 3 2 summarizes the local absorption maxima for PA and PB in toluene and as cast. It is worth noting that the window of transmission of PB is shifted towards the higher wavelengths ( local minimum of absorption at 528 nm) when compared to that of PA (local minimum of absorption at 505 nm) owing to the more electronrich character of EDOT in comparison with thiophene, hence the yellow hue in the green color of PB Further, it is worth noting that all spectra in Figure 3 6 were normalized at the longer wavelength absorption maximum of the oligomers or polymers, such that the peak intensities should not be considered in terms of their physical significance in this Figure. As a matter of fact, it is likely that the second absorption band gro ws with increasing extent of conjugation, in the order Trimer>Pentamer>Polymer (see in Chapter 4 how measuring the mass absorption coefficients of each component could give further insight into this assumption) The key incorporation of regiosymmetric ProD OTs be aring two octyloxy solubilizingchains induces sufficient solubility

PAGE 99

99 (up to ca. 8 mg mL1 in the case of PA ) for solid thin -film processing using common organic solvents. While PA shows good to excellent solubility at room temperature in toluene, chl oroform and ortho-dichlorobenzene (ODCB), P B requires hot mixtures of toluene and chloroform (60oC) or hot ODCB (80 100oC) to be efficiently dissolved in solution. Table 3 2 Local absorption maxima (solution in toluene and solid state) and film optical bandgap ( Eg, eV) for the donor acceptor copolymers PA and PB Polymer abs (nm) In Toluene abs (nm) Thin Film E g (eV) 1 2 1 2 PA 449 643 458 672 1.47 PB 464 704 467 710 1. 42 3.2 .3 Polymer Redox and Spectroelectrochemical Characterization Films of PA and PB were spray -cast onto ITO -coated glass slides from organic solvents, and subsequently redox cycled until they reached a stable and reproducible switch prior to spectroelectrochemical analysis. As illustrated in Fig ures 3 7 a and 3 7 b, as the potential applied to the film was increased from 0.3 V ( vs. Fc/Fc+) for PA 0.6 V for PB simultaneous bleaching of both the low and high energy transitions was initiated. These values are consistent with the low onsets of oxid ation of PA and PB respectively estimated at 0.3 V and -0.55 V via differential -pulse voltammetry (see Figures 3 8 b and 3 8 d) As the short and long wavelength absorption bands depleted, new transitions progressively ar o se in the NIR indicating the format ion of the charge carriers. Determined from the onset of their lowest energy transitions, PA exhibits a band gap at 1.47 eV and PB shows a more narrow value of 1.42 eV as expected considering that 3,4 ethylenedioxythiophene (EDOT) brings a higher energy HOMO than thiophene to the conjugated donor acceptor system. When fully ox idized, both polymers show a high level of transparency to the human eye. Upon full oxidation the bipolaronic transition of PB clearly more red -shifted in the NIR, tails less into the visible hence inducing a

PAGE 100

100 higher electrochromic contrast. From this data and taking the longer wavelength absorption maximum as a reference as it interacts with the near infrared tail of the fully doped polymer, a transmittance change up of about 32 % for PA versus 53% for PB wa s calculated, indicative of the higher contrast observed for the latter polymer. From the cyclic and differential -pulse voltammetry plots presented in Figure 3 8, two oxidation processes can be seen (along with their corresponding reductions on the reverse cycle). These two processes can be attributed t o the formation of polaronic and bipolaronic species (respectively) on application of the external bias (see supporting EPR work from Irvin et al.119), and is in good agreement with the spectroelectrochemical data shown in Figure 3 7. While it is worth noting that the current peak associated to the first oxidation process (formation of cat ion radicals) in PB is especially sharp, the reasons inducing the observed phenomenon remain unclear. This phenomenon was found to be perfectly reproducible over the various thin-films investigated. Figure 3 7. Spectroelectrochemistry of a) PA and b) P B. Films were spray -cast onto ITO coated glass from solution (2 mg mL1) in toluene ( PA ) or a hot mixture of toluene and chloroform ( PB ). Electrochemical oxidation of the films was carried out in 0.1 M LiBF4/ACN supporting electrolyte using a silver wire a s a quasi -reference electrode (calibrated against Fc/Fc+) and a platinum wire as the counter electrode. In both cases, the applied potential was increased in 25 mV steps from 0.4 V to +0.95 V for PA and from 0.65 V to +0.65 V for PB (Adapted with permiss ion from Ref.109 Copyright 2008 WILEY -VCH Verlag GmbH & Co. KGaA, Weinheim)

PAGE 101

101 a) b) c) d) Figure 3 8 Cyclic voltammetry (CV) of a thin film of a) PA and c) PB drop -cast on a Pt button electrode (9 scans) Electrochemical oxidation of t he films was carried out in 0.1 M LiBF4/ACN supporting electrolyte using a silver wire as a quasi -reference electrode (calibrated against Fc/Fc+) and a platinum flag as the counter electrode. Differential Pulse Voltammetry (DPV) on a thin film of b) PA and d) PB drop -cast on a Pt button electrode. Electrochemical oxidation was carried out in conditions identical to CV (Adapted with permission from Ref.109 Copyright 2008 WILEY -VCH Verlag GmbH & Co. KGaA, Weinheim) 3.2.4 Polymer Colorimetric Analysis As shown in Figure 3 9 the relative luminance change (estimating the brightness of the transmitted light as a percentage of the brightness of the light source calibrated to the sensitivity of the human eye) was measured as the doping level induced by electrochemical oxidation was progressively increased From their neutral green to fully oxidized transmissive state, the spraycast polymers exhibit a relative luminance change of approximately 30 %, a promising va lue considering that the light adapted human eye is most sensitive to forest -green, a tone that almost

PAGE 102

102 equally stimulates two of the three kinds of cone cells in the eye and that is consequently likely to be perceived regardless of its intensity. It is wor th noting that relative luminance can also be modulated and optimized as a function of deposited film thickness. The materials in their neutral state exhibit a saturated green, a fact supported by the very negative a* values obtained, but they show differe nt hues as characterized by the significantly different b* values.12 For instance, the large positive b* value for PB translates the yellow tone of its green. L* values (ranging from 0 to 100) such as thos e exhibited by PA (L* = 91) and PB (L* = 89) upon oxidation demonstrate their aptitude to nearly reach the white point (W) of color space. In the oxidized state, a* and b* drop down to orders as low as 3 ( PA ) and 5 ( PB ) yielding corresponding tints tha t can hardly be perceived by the human eye. The polymers complete their full switches in a potential window of less than 1.5 V, a parameter of significant interest for low -voltage device applications. The window necessary to achieve more than 90 % of the o ptical change is considerably narrower: as low as 1.05 V for PA and 1.15 V for PB Figure 3 9. Relative Luminance (%) as a function of applied potential for both spray -coated PA and PB Pictures show the hues of green perceived and give an indication of the degree of transmissivity obtained upon complete oxidation. For color matching, L*a*b* values (in the sense of the CIE 1976 L*a*b* color model) of fully neutral and oxidized states are reported for the films (Adapted with permission from Ref.109 Copyright 2008 WILEY -VCH Verlag GmbH & Co. KGaA, Weinheim)

PAGE 103

103 3.2.5 Polymer Switching Study As shown in Figure 3 10a and 3 10b t he electrochromic contrasts % T ) were monitored as a function of time at the longer wavelength absorption maximum of each material applying square -wave potential steps of 10 s, 2 s and 1 s in order to characterize their respective response time Figure 3 10. a) Square -wave p otential step absorptometry of spray -coated PA (monitored at 672 nm, from 0.5 V to +1.05 V vs. Fc/Fc+) and b) PB (monitored at 710 nm, from 0.7 V to +1 V vs. Fc/Fc+) onto ITO in 0.1 M LiBF4/ACN solution. Switch times: 10 s step for 40 s (2 cycles), then 2 s step for 20 s (5 cycles) and 1 s step for 20 s. (10 cycles). c) Square -wave potential stepping EC switching of drop -cast PA onto Pt in 0.1 M LiBF4/propylene carbonate solution switching between 0.4 V and +0.85 V ( vs. Fc/Fc+) with a switch time of 4 s (Adapted with permission from Ref.109 Copyright 2008 WILEY -VCH Verlag GmbH & Co. KGaA, Weinheim)

PAGE 104

104 As predicted by the spectroelectrochemical study, the film of PB exhibited higher contrasts (44 40 %) than PA ( 39 36 %), although in eith er case a variation in EC contrast of only 3 to 4 % was observed as the switch time is increased from 10 to 1 second. Measured at the shorter wavelength absorption maximum of PA and PB %T was estimated at 27% and 45% respectively. A new film of PA was spray -coated onto ITO and its stability upon square -wave potential stepping EC switching was investigated in 0.1 M LiBF4 / propylene carbonate solution using a switch time of 2 s The corr esponding data is summarized in Table 3 3 Table 3 3. Square -wave potential stepping EC switching of spray -cast PA onto ITO in 0.1 M LiBF4 / propylene carbonate solution switching between 0.5 V and +1.05 V ( vs. Fc/Fc+) with a switch time of 2 s (Adapted with permission from Ref.109 Copyright 2008 WILEY -VCH Verlag GmbH & Co. KGaA, Weinheim) Cycle [a] [b] Q (mC/cm 2 ) [c] CE (cm 2 /C) [d] %Decay [e] 1 34.6 2.7 126 0 100 34.1 2.77 121 1.5 250 34 2.71 123 1.7 500 33.8 2.64 124 2.3 750 33.4 2.6 124 3.5 1000 32.7 2.55 123 5.5 [a] Complete cycle is 4 s [b] % Optical change in transmittance at 672 nm [c ] Injected/ejected charge per unit area of spray cast film [d] CE = OD/Q = log(%Tox/%Tred)/Q [e] T ) At the longer wavelength absorption maximum, a loss of EC contrast of only 5.5 % was estimated after 1000 c ycles along with no significant change in coloration efficiency ( CE ) which indicates the proper relationship between charge stored (calculated from integration of the current) and contrast observed. This result is significant considering the nature of the substrate used, and the redox induced solubility of the polymer in the electrolyte, as there is little to prevent the chains from slowly migrating away from the working electrode to the platinum

PAGE 105

105 counter electrode during the experiment. In order to highligh t the influence of the substrate on stability studies, long term switching was also investigated on a Pt button electrode, as reported in Figure 3 10c While the decay in charge stored per unit area of spray -coated film of PA is in the order of 5.5 % onto ITO, a film drop -cast on Pt showed about 2 % loss after 1000 cycles, about 3 % after 2000 cycles and less than 5 % after 10 000 cycles (nearly 24 hours) which reinforces the potential of this ECP for commercial applications. 3.2 .6 Synthetic Details Compoun d 2 -3 : Details relative to the synthesis and characterization of 2 -3 can be found in Reeves, B. D.; Grenier, C. R. G.; Argun, A. A.; Cirpan, A.; McCarley, T. D.; Reynolds, J. R. Macromolecules 2004, 37, 75597569. Compound 4 : A 500 mL flame dried single ne ck round bottom flask was filled with 250 mL of anhydrous DMF, NaH (60% in mineral oil) (5.6 g, 140 mmol) and octan1 ol (12.2 g, 94 mmol). The mixture was heated at 100 C over a period of 4 hours, then compound 3 (8 g, 23.4 mmol) was added by portions to the hot reagent. After stirring overnight at 100 C, the reaction mixture was cooled down to room temperature, added to brine (700 mL) and extracted 3 times with ethyl ether (500 mL). The organic phase was washed with water, dried over magnesium sulfate a nd the solvent was removed affording a dark orange oil. The resulting oil was purified by column chromatography on silica with hexane/dichloromethane (3:2) as eluent. T he solvent was evaporated and compound 4 was obtained as a clear oil (7 g, 68 %). 1H NMR (300 MHz, CDCl3, ): 6.44 (s,2H), 4.01 (s, 4H), 3.48 (s, 4H), 3.40 (t, J = 6.4 Hz, 4H), 1.571.56 (m, 4H) 1.291.27 (m, 20H), 0.88 (t, J = 6.8 Hz, 6H); 13C NMR (75MHz, CDCl3, ): 149.9, 105.2, 73.9, 71.9, 69.8, 47.9, 32.1, 29.7, 29.6, 29.5, 26.3, 22.9, 14.3; HRMS (TOF, m/z): [M+H]+ calculated for C25H44O4S, 441.3033; found, 441.2992.

PAGE 106

106 Compound 5 : Compound 4 (2.12 g, 4.81 mmol ) was dissolved in dry THF (17 mL) and cooled down to 78 C. A solution of nbutyllithium in hexanes (2.51 mL, 6.26 mmol ) was added o ver a 2 hour period and the mixture was stirred for 1 hour. Trimethyltinchloride (6.3 mL, 6.3 mmol ) was subsequently added, the mixture was allowed to warm up to room temperature and stirred for 16 hours. The solvent was evaporated obtaining a yellow -brown oil (2.67 g, 92 %) which was used for the next step without further purification. 1H NMR (300 MHz, CDCl3, ): 6.70 (s, 1H), 3.98 (s, 2H), 3.93 (s, 2H), 3.48 (s, 4H), 3.39 (t, J = 6.4 Hz, 4H), 1.571.56 (m, 4H), 1.291.27 (m, 20H), 0.87 (t, J = 6.8 Hz, 6H) 0.32 (s, 9H). Com pound 6a : Tributyl(thiophen2 -yl)stannane (7.46 g, 20 mmol), 4,7 dibromobenzo[ c ][1,2,5]thiadiazole (2.45 g, 8.33 mmol) and Pd(PPh3)2Cl2 (333 mg, 4 mol %) were dissolved in 80 ml of THF. The mixture was stirred for 36 hours at 60 C then the solvent was evaporated and the product was precipitated twice in MeOH. After filtering through a Bchner funnel 4,7 di(thiophen2yl)benzo[ c ] [1,2,5]thiadiazole was collected as an oran ge -red solid (1.85 g, 74 %) 1H NMR (300 MHz, CDCl3, ): 8.12 (dd, J = 0.9, 3.7 Hz, 2H), 7.83 (s, 2H), 7.45 (dd, J = 0.9, 4.9 Hz, 2H), 7.20 (dd, J = 0.9, 4.9 Hz, 2H); 13C NMR (75 MHz, CDCl3, ): 152.81, 139.57, 128.22, 127.72, 127.00, 126.15, 125.94; HRMS (TOF, m/z): [M+] calcd. for C14H8N2S3, 300.9922; foun d, 300.9936. 4,7 di(thiophen2yl)benzo[ c ] [1,2,5]thiadiazole (2.5 g, 8.32 mmol) was dissolved in 200 ml of CHCl3 and a catalytic amount of acetic acid. Then NBS (4.44 g, 24.97 mmol) was added portion wise and the mixture was stirred for 36 h at room temper ature under light protection. The precipitate was filtrated through a Bchner funnel, washed with copious amounts of water, methanol, acetone and compound 6a was collected as an red solid (2.75 g, 72 %).1H NMR (300 MHz, CDCl3, ): 7.82 (d, (t, J = 3.9 Hz, 2H), 7.79 (s, 2H), 7.16 (d, (t, J = 3.9 Hz, 2H). 13C NMR

PAGE 107

107 could not be determined because of poor solubility; HRMS (TOF, m/z): [M+] calcd. for C14H6N2Br2S3, 456.8133; found, 456.8157. Compound 6b : (2,3 -dihydro thieno[3,4 -b][1,4]dioxin5 yl)trimethylstannane (11.58 g, 37.93 mmol), 4,7 dibromobenzo [ c ][1,2,5]thiadiazole (3.99 g, 13.55 mmol) and Pd(PPh3)2Cl2 (380 mg, 4 mol %) were dissolved in 200 ml of THF. The mixture was stirred for 72 hours at 60 C, the solvent was evaporated a nd the product was precipitated twice in MeOH. The dark red solid 4,7 -bis(2,3 dihydrothieno[3,4b ][1,4]dioxin5 -yl)benzo[ c ][1,2,5]thiadiazole was filtrated through Bchner funnel and collected (3.63 g, 64%). 1H NMR (300 MHz, CDCl3, ): 8.38 (s, 2H), 6.56 (s, 2H), 4.40 4.30 (m, 8H); 13C NMR (75 MHz, CDCl3, ): 152.58, 141.90, 140.49, 126.86, 123.93, 113.69, 102.19, 65.23, 64.58; HRMS (TOF, m/z): [M+] calcd. for C18H12N2O4S3, 417.0032; found, 417.0066. 4,7 bis(2,3 dihydrothieno[3,4 b ][1,4]dioxin 5 -yl)benzo[ c ][1,2,5]thiadiazole (3.49 g, 8.38 mmol) was dissolved in 800 ml of CHCl3 and a catalytic amount of acetic acid is added. Then NBS (3.73 g, 20.95 mmol) was added portion wise and the mixture was stirred for 48 h at room temperature un der light protection. The dark red precipitate was filtrated through a Bchner funnel, washed with water, methanol, acetone and compound 6b was collected (4 g, 83 %).1H NMR (300 MHz, CDCl3, ): 8.35 (s, 2H), 4.39 (m, 8H); 13C NMR could not be determined because of poor solubility; HRMS (TOF, m/z): [M+] calcd. for C18H10Br2N2O4S3, 572.8242; found, 572.8225. Compound 7a ( Pentamer A ): Compound 5 (4.24 g, 6.97 mmol), compound 6a (1.3 g, 2.81 mmol ) and Pd(PPh3)2Cl2 (81 mg, 4 mol %) were dissolved in 70 ml of THF. The mixture was stirred for 48 hours at 60 C, the solvent was evaporated and the product was purified by column chromatography on silica with hexane/dichloromethane (2:1) as eluent. T he s olvent was

PAGE 108

108 evaporated and compound 7a was obtained as a purple oil ( 2.34 g, 70 %). 1H NMR (300 MHz, CDCl3, ): 8.10 (d, J = 3.8 Hz, 2H), 7.84 (s, 2H), 7.32 (d, J = 3.8 Hz, 2H), 6.40 (s, 2H), 4.23 (s, 4H), 4.08 (s, 4H), 3.57 (s, 8H), 3.43 (t, J = 6.4 Hz, 8H), 1,55 (m, 8H), 1.30 (m, 40H), 0.87 (t, J = 6.3 Hz 12H); 13C NMR (75MHz, CDCl3, ): 152.85, 150.27, 145.87, 138.18, 136.60, 128.02, 125.83, 125.33, 124.14, 117.33, 102.72, 74.32, 74.17, 72.03, 69.93, 48.11, 32.10, 29.81, 29.68, 29.54, 26.41, 22.90, 14.32; HRMS (TOF, m/z): [M+H]+ calcd. for C64H92N2O8S5, 1177.5530; found, 1177.5547. Anal. calcd. for C64H92N2O8S5: C 65.27, H 7.87, N 2.38; found: C 65.78, H 8.03, N 2.32. Compound 7b (Pentamer B) : Compound 5 (4.28 g, 7.09 mmol), compound 6b (1.68 g, 2.95 mmol) and Pd(PPh3)2Cl2 (85 mg, 4 mol %) were dissolved in 70 ml of THF. The mixture was stirred for 48 hours at 60 C, the solvent was evaporated and the product was purified by column chromatography on silica with hexane/dichloromethane (2:1) as eluent. T he solvent was evaporated and compound 7b was obtained as a dark blue tacky solid ( 1.08 g, 29 %). 1H NMR (300 M Hz, CDCl3, ): 8.34 (s, 2H), 6.45 (s, 2H), 4.41 (s, 8H) 4.25 (s, 4H), 4.08 (s, 4H), 3.60 (m, 8H), 3.43 (t, J = 6.4 Hz, 8H), 1,55 (m, 8H), 1.30 (m, 40H), 0.87 (t, J = 6.3 Hz, 12H); 13C NMR (75MHz, CDCl3, ): 152.63, 149.84, 145.49, 140.07, 137.61, 126.86, 123.65, 115. 41, 113.08, 112.54, 103.69, 74.71, 74.32, 72.04, 70.14, 65.15, 6492, 48.09, 32.08, 29.81, 29.68, 29.52, 29.40, 22.88, 14.31; HRMS (MALDI TOF, m/z): [M+] calcd. for C68H96N2O12S5, 1292.5561; found, 1292.5557. Anal. calcd. for C68H96N2O12S5: C 63.13, H 7.48, N 2.17; found: C 63.32, H 7.50, N 2.14. Polymer PA : Compound 7a (459 mg, 0.39 mmol) was dissolved in chloroform (70 mL). The monomer solution was placed under air bubbling. A solution of FeCl3 (316 g, 1.95 mmol, 5eq) in nitromethane (2 mL) was added dropw ise to the stirred monomer at room temperature

PAGE 109

109 over a 1 hour period. The dark purple monomer solution turned dark green with addition of oxidizing agent. The mixture was stirred 48 hours at room temperature under constant air bubbling. It was then precipit ated into methanol (500 mL). The precipitate was filtered, redissolved in chloroform (300 mL) and stirred for 6 hours with hydrazine monohydrate (6 mL). After concentration by evaporation, the polymer solution (green) was precipitated into methanol (500 mL ), the precipitate was filtered through a Soxhlet thimble and purified via Soxhlet extraction for 48h with methanol. The polymer was extracted with chloroform for 6 hours, the solvent was evaporated, the polymer was precipitated in methanol and collected a s a black solid (404 mg, 88 %). 1HNMR (300 MHz, C2D2Cl4 (100 C), ): 8.16 (bs, 2H), 7.89 (bs, 2H), 7.43 (bs, 2H), 4.34 (bs, 8H), 3.723.56 (m, 16H), 1.69 (bs, 8H), 1.38 (bs, 40H), 0.96 (bs, 12H). GPC analysis: See Table 3 1 Anal. calcd. for C64H90N2O8S5: C 65.38, H 7.72, N 2.38 Found: C 66.08, H 7.75, N 2.21. Polymer P B: Compound 7b (578 mg, 0.45 mmol) was dissolved in chloroform (70 mL). Following the proced ure described for synthesizing P A P B was obtained as a black solid (480 mg, 83 %). 1HNMR (300 MHz, C2D2Cl4 (100 C), ): 8.40 (bs, 2H), 4.50 (bs, 8H), 4.28 (bs, 8H ), 3.70 (bs, 8H), 3.55 (bs, 8H), 1.55 0.93 (m, 60H). GPC analysis: See Table 3 1 Anal. calcd. for C68H94N2O12S5: C 63.22, H 7.33, N 2.17 Found: C 62.87, H 7.26, N 2.27. 3.3 Conclusions and Outlook In summary, t wo symmetri cal donor acceptor heterocyclic pentamers were designed, synthesized and chemically polymerized affording organic soluble neutral state green conjugated polymers of different hues adequate for multi -colored display applications Optical and electrochemical investigation of the green conjugated polymers revealed highly transmissive oxidized states, excellent optical contrasts both in the visible and in the NIR, small potential windows of operation, fast switching times and long term stability.

PAGE 110

110 It is worth noting that t he synthesized green polyheterocyclics exhibit a repeat unit whose molecular structure can be synthetically adjusted in order to tailor the energy gap and control the green hue of the polymer, a desirable property for applications in displays. C omplete characterization in terms of structural and physical properties, as well as practical integration of ECPs into devices, requires solution -processability, which is not attained with the commonly insoluble electropolymerized materials. Here, the repo rted materials can easily be processed into electrochromic devices using convenient deposition methods such as spin coating and spray casting. This combination of useful properties suggest s a new avenue towards dual ECP, photovoltaic and near IR device app lications. In recent work120 from Toppare et al. publishe d during the preparation of this dissertation work the donor acceptor 4,7 -di(2,3 dihydro-thieno[3,4 b][1,4]dioxin 5 yl)benzo[1,2,5]thiadiazole (BEDOT BTD), the synthesis and electrochemical propertie s having previously been reported,121123 was electropolymerized and its subsequent polymer (PBEDOT BTD) was re investigated as a possible candidate for electrochromic applications. The insoluble polymer film was revealed to be a neutral state green material switc hing rapidly and reversibly to a transmissive light blue oxidized state,120 an experimental result strongly supporting the utility of the donor acceptor theory in the design of neutral green ECPs. Hence, a number of works reporting on neutral state green polymer electrochromes with oxidized states of various colors and transmissivities are now available.104,105,109,120,124 132

PAGE 111

111 CHAPTER 4 SYNTHETIC CONTROL OF THE SPECTRAL ABSORPT ION IN 3,4 DIOXYTHIOPHENE 2,1,3 BENZOTHIADIAZOLE DONOR ACCEPTOR ELECTROCHRO MIC POLYMERS: FROM GREEN TO BLACK 4.1 Conte xt and Motivations for Tailoring a Two -Band Absorption in the Visible Considering the fast -growing demand for innovative high -performance display technologies, the perspective of manufacturing low -cost functional materials that can be easily processed over large areas (see Figure 4 1a) or finely printed into individual pixels, while being mechanically deformable (see Figure 4 1b) has motivated the development of novel electronically active organic components fulfilling the requirements for flexible display s and portable applications. Among all technologies relying on a low -power stimulated optical change, non -emissive organic electrochromic devices (ECDs) offer the advantage of being operational under a wide range of viewing angles and lighting conditions s panning direct sunlight as desired for various applications including signage, information tags and electronic paper (see Figure 4 1 b ). a) b) Figure 4 1. a) Switchable windows ( courtesy of Sage Electrochromics Inc.) and b) Electronic paper ( courtesy of Plastic Logic )

PAGE 112

112 Combining mechanical flexibility, high contrast ratios, and fast response times, along with color tunability via structural control, polymeric electrochromes constitute the most attractive organic electronics for tomorrows reflective / tran smissive ECDs and displays .12,104 Of the organic electrochromes, electroactive conducting polymers show relatively facile synthetic access, environmental stability and solution processability. By controlling the repeat unit structure, a range of colors spanning the full visible spectrum can be attained.12 For instance, polyaniline can attain a transmissive yellow to black switch, though it is unstable to repeated switching,133 hence the necessity for new polymer design. While red, blue14 and most recently green104,105,109,120 electrochromic polymers (including cathodically -coloring and multichromic!) desired for additive primary color space were investigated, attempts in making saturated black EC Ps have not been reported, likely due to the complexity in designing materials absorbing effectively over the whole visible spectrum. Processable black to transmissive ECPs could impact the development of both reflective and transmissive ECDs by providing lower fabrication and processing costs via printing, spraying, and coating methods, along with good scalability when compared to their traditional inorganic counterparts (including WO3).12,13,58 In spite of significant effort in the field of polymer based solar cells towards developing materials absorbing homogeneously over a broad bandwidth of the UV -visible spectrum,49,134141 the abili ty to design strongly absorbing ECPs that can be switched to a highly transmissive state has remained a challenge due to the additional requirement for simultaneous and efficient electrochemical bleaching of all absorption bands over the visible region. In light of the results obtained throughout Chapter 3 of this dissertation, whereby neutral state green -to -transmissive switching polymers with a two -band spectral absorption in the visible were produced, a logical step forward would consist in examining car efully the parameters

PAGE 113

113 governing the existence of both a short and a long-wavelength transition in the optical spectrum of the cathodically -coloring -conjugated polymers synthesized As also stated in Chapter 3 the presence of a high energy transition in strictly linear and alternating DA systems involving two distinct energy transitions could be explained by the presence of low -lying unoccupied and localized energy levels within the band gap of the conjugated system in its ground-state.107,108 In the case where dual absorption bands do result from the linear -conjugation of electron rich and electron -deficient heterocycle s along the polymer backbone, one may wonder how varying the relative contribution of donor and acceptor units incorporated in the repea t unit of the corresponding polymers could impact the polymer two-band absorption. In particular, considering a hypothetical situation where low and high energy tra nsitions could be controlled in terms of their intensity and extent of mutual overlap so as to balance the optical pattern of the subsequent chromophore (see Figure 4 2a) while extending it across the entire visible spectrum, deep -blue and black ink -like electrochromes could possibly by produced (see Figure 4 2 b ). a) b) Figure 4 2 a) Schemati c representation of the optical spectrum of a two -band absorbing chromophore reflecting the color green (green curve) Could l ow and high energy transitions be controlled in terms of their intensity and extent of mutual overlap (dotted black curve) so as t o balance the optical pattern of the subsequent chromophore? b) Schematic representation of the optical spectrum of an ink like blue (blue curve) and black chromophores (black curve) Note that for a chromophore to be exactly black, its optical spectrum mu st extend homogeneously over the entire visible spectrum

PAGE 114

114 4.2 Proposed Design and Polymer Optical Characterization 4.2.1 Design and Synthesis of 3,4-Dialkoxythiophene 2,1,3 -Benzothiadiazole Polymers As mentioned above, in the case where dual absorption bands (such as those observed for the neutral state green polymer electrochromes described in Chapter 3 ) do result from the simple linear -conjugation of electronrich and electron deficient heterocycles along the polymer backbone, varying the relative contribution of donor and acceptor units incorporated in the repeat unit of the corresponding polymers could significantly impact the polyme r two band absorption. In order to probe this assumption, a series of 3,4 -dialkoxythiophene (DalkOT) and 2,1,3 benzothiadiazole (BTD) based polymers was designed (see Figure 4 3 ), whereby the acceptor core is increasingly substituted with a various numbe r of donor substituents hence producing the series of oligomers M1 -M3 which oligomers are subsequently polymerized, hence producing polymers P1 -P3 along which the BTDs are increasingly spaced by the DalkOTs. The control all D alk OT based polymer P4 was sy nthesized for comparison of its optical pattern with that of the DA polymers P1 P3 Figure 4 3 Designing 3,4 dialkoxythiophene (DalkOT) and 2,1,3 -benzothiadiazole (BTD) based polymers ( P1 -P3 ) whereby the BTDs are increasingly spaced by the DalkOTs al ong the backbone As illustrated in Figure 4 4 DalkOT building blocks bearing 2 -ethylhexyl solubilizing side -chains ( 3,4 -bis(2 -ethylhexyloxy)thiophene, DEtHxOT ), and aimed at imparting the subsequent polymers with adequate solutionprocessability, were pre pared.

PAGE 115

115 Figure 4 4. Synthetic route to DalkOT BTD based donor acceptor conjugated copolymers P1 -P3 Following Ullmanns conditions, 3,4 dibromothiophene 1 was first converted to 3,4 dimethoxythiophene 2 in a refluxing mixture of sodium methoxide -conta ining methanol and in the presence of a Cu(I) source (here CuI ). Next, the Williamson transetherification conditions were employed to append the 2 -ethylhexyl solubilizing chains of the DalkOT intermediate 3 in ca. 90% yield The trimethyltin derivative 4 w as obtained upon reacting a monolithiated D EtHx OT 3 with trimethyltinchloride. The halogenated BTD unit 5a was then symmetrically substituted with two electron rich D EtHx OT stannane species ( 4 ) via Pd -mediated cross coupling Stille conditions111 in gently refluxing THF to yield oligomer 6a ( M1 ). Column chromatography over silica using mixtures of hexanes and dichloromethane afforded 6a which was iso lated as a bright orange oil. A fraction of 6a was subjected to bromination conditions at room temperature and via NBS to afford oligomer 5b Pd -mediated cross -coupling and

PAGE 116

116 bromination of the resulting oligomer were repeated to afford oli gomers 6b (M2 ), 5c and 6c (M3 ). In other words, t he soluble oligomeric analogues M1 -M3 were synthesized by symmetrical and repeated substitution of the donor DEtHxOT onto the BTD acceptor The DA oligomers were next homopolymerized at room temperature using the mild oxidizi ng agent FeCl3 and the resulting macromolecules were subsequently reduced with hydrazine to yield polymers P 1 -P3 The p olymers were precipitated in methanol and purified by Soxhlet extraction with methanol before further characterization. The structures of the repeat unit of P 1 -P 3 were supported by 1H NMR and the quality of the polymers was probed by elemental analysis (see Experimental Section below). The polymer molecular weights were determined via polystyrene -calibrated GPC with THF as the mobile phase and are summarized in Table 4 1. Number average molecular weights ( Mn) ranging from 16,300 g mol1 to 47,900 g mol1 were estimated ( for P 1 and P 3 respectively ). Table 4 1. Number average molecular weight ( Mn, g mol1), weight average molecular weight (Mw, g mol1), polydispersity index (PDI), average number of repeat units, and average number of rings for the donor acceptor copolymers P1 -P3 Polymer Mn (g mol1) Mw (g mol1) PDI Avg. No. of Repeat Units Avg. No. of Rings P1 16,300 42,700 2.6 20 60 P2 26,300 69,200 2.6 17 85 P3 47,900 98,300 2.1 22 154 4.2.2 3,4 -Dialkoxythiophene 2,1,3 -Benzothiadiazole Polymer Optical Characterization Experimentally, it was confirmed that symmetric successive addition s of DOT s on to BTD induce a progressive bathochrom ic shift of both the shorter wavelength absorption band and the longer wavelength absorption band of the growing oligomers (see Figure 4 -5 a). It is also clear that the first substitution has the largest impact on the energy gap of the donor acceptor system while the second substitution account s for a fine r energ y tuning over the two -band absorption

PAGE 117

117 spectrum of the resulting oligomers. As illustrated in Figure 4 5 b, polymerization adds a more significant red -shift to the set of optical transition s when comp ared to the precursor oligomers a) b) Figure 4 5. a) Solution optical absorption spectra (in toluene) of M1 -M3 (oligomer precursors for polymer P1 -P3 ), b) Solution optical absorption spectra (in toluene) of DA copolymers P1 -P3 along with that of the control polymer P4 The spectrum of each system is normalized at the longer wavelength absorption maximum

PAGE 118

118 From Figure 4 5 b it can further be seen that diminishing the concentration of the electronpoor heterocycle along the conjugated backbone ( i.e. increasing the length of the oligo thienylene electron -rich block) gives a simultaneous bathochromic shift of the high energy transition and hypsochromic shift of the low energy transition with both peaks tending towards the max of the P4 control homopolymer resulting in the merging of the two transitions. At the same time, the relative intensity of the low and high energy peaks evolve towards balancing each other. Hence P3 exhibits a broad absorption spectrum almost entirely covering the visible region with the exception of a dark red portion between 625 700 nm and the blue -green gap located at 450 550 nm; the colors of P1 through P4 in toluene solution are shown on the photographs of Figure 4 5 b T he dark brown hue of P3 can be attributed to its overall high visible absorbance with minimal red light transmission. Table 4 2 summarizes the local absorption maxima for P1 P3 in toluene and as cast. In toluene solution, all polymers show ed only minor optical changes upon incre ase of the temperature between room temperature and 100oC indicating neither significant disruption of effective conjugation115,116 nor major aggregation of th e backbones once solvated.117,118 The incorporation of regiosymmetric D EtHx OTs bearing two ethylhexyl solubilizing -chains induces sufficient solubility (up to ca. 10 mg mL1 in the case of P3 ) for solid thin -film processing using common organic solvents such as t oluene. Table 4 2. Local absorption maxima (solution in toluene and solid state ), and film optical bandgap ( Eg, eV) for the donor acceptor copolymers P1 -P3 Polymer abs (nm) In Toluene abs (nm) Thin Film E g (eV) 1 2 1 2 P1 391 638 399 653 (708) a 1.6 P2 4 10 591 4 24 612 1.69 P 3 4 18 568 416 567 1. 77 a Shoulder

PAGE 119

119 4.2.3 Design and Synthesis of 3,4Propylenedioxythiophene 2,1,3 -Be nzothiadiazole Polymers With the absorption spectrum of P3 extending over the entire visible region with the exception of a dark -red portion between 625 700 nm and the blue -green gap located at 450 550 nm (to a lesser extent) a pol ymer design that wou ld produce material s exhibiting narrower bandgap s, thereby reducing the extent of red -light transmission, was envisaged. To this end, the DOT moieties of P1 -P3 were replaced with the more strongly donating 2 ethylhexyloxysubstituted ProDOT which possesses a higher HOMO and shoul d substantially narrow the bandgap of the subsequent polymer without the need to exchange BTD for a stronger acceptor Following these considerations, t he series of ProDOT BTD based polymers presented in Figure 4 6 was proposed w hereby the acceptor core is increasingly substituted with a varied number of donor substituents, hence producing the series of oligomers M5 -M 7 which oligomers are subsequently polymerized, hence producing polymers P5 -P 7 where the BTDs are increasingly spa ced by the ProD OTs. The control all Pro DOT based polymer P 8 was synthesized for comparison of its optical pattern with that of the DA polymers P 5 P 7 Figure 4 6 Designing 3,4 propylenedioxythiophene (ProDOT) and 2,1,3 -benzothiadiazole (BTD) based pol ymers ( P5 -P 7 ) whereby the BTDs are increasingly spaced by the ProD OTs along the backbone

PAGE 120

120 As illustrated in Figure 4 7 ProDOT building blocks bearing 2 ethylhexyl oxy solubilizing side -chains ( D OEtHx ProD OT ), and aimed at imparting this analogous polymer ser ies with adequate solution -processability, were prepared. The synthesis of 1 -3 was described in Chapter 3 of this dissertation. Figure 4 7. Synthetic route to ProDOT BTD based donor acceptor -conjugated copolymers P5 -P7 The halogenated BTD unit 4 a wa s symmetrically substituted with two electron -rich D OEtHxProDOT stannane species ( 3 ) via Pd -mediated cross -coupling Stille conditions111 in gently refluxing THF to yield oligomer 5 a (M 5 ). Column chromatography over silica using mixtures of hexanes and dichloromethane afforded 5 a which was isolated as a bright orange oil. A fraction of 5 a was subjected to bromination conditions at room temperature and via NBS to

PAGE 121

121 afford oligomer 4 b Pd -mediated cross coupling and bromination of the resulting oligomer were repeated to afford olihgomers 5 b (M 6 ), 4 c and 5 c (M 7 ). In other words, the soluble oligomeric analogues M5 -M 7 were synthesized by symmetrical and repe ated substitution of the donor D OEtHxProDOT onto the BTD acceptor. The DA oligomers were next homopolymerized at room temperature using the mild oxidizing agent FeCl3 and the resulting macromolecules were subsequently reduced with hydrazine to yield polyme rs P 5 P 7 The polymers were precipitated in methanol and purified by Soxhlet extraction with methanol before further characterization. The structures of the repeat unit of P 5 -P 7 were supported by 1H NMR and the quality of the polymers was probed by element al analysis (see Experimental Section below). The polymer molecular weights were determined via polystyrene -calibrated GPC with THF as the mobile phase and are summarized in Table 4 3 Number average molecular weights ( Mn) ranging from 1 2,5 00 g mol1 to 27 ,8 00 g mol1 were estimated (for P 7 and P 5 respectively). Interestingly, in contrast with what was observed among the D EtHx OT -based polymer series represented by P1 P3 the GPC estimated Mn decreased with increasing number of DOEtHxProDOT inserted in the backbone among the polymer series represented by P5 -P7 (i.e. with diminishing number of acceptors) Although apparently counter intuitive since DOEtHxProDOT bears solubilizing groups (not BTD) this empirical observation points toward a stiffening of the conjugated backbone under the influence of the oxygen-sulfur interactions between adjacent ProDOT units (as seen in PEDOT142), which could induce the early precipitation of the growing polymer on oxidative polymerization. The presence of bulky in-plane solubilizing groups in the D EtHx OT based polymers, effectively twisting the backbone out of planarity, is believed to be the reason for the difference in solubility observed (incl uding the solubility of the growing species on oxidative polymerization) hence for the difference in molecular weight attained among the

PAGE 122

122 series P1 -P3 This interpretation is confirmed by the especially high solvent -solubility exhibited by the control poly mer P4 which remains a tacky solid at room temperature while possessing a GPC -estimated Mn greater than 50,000 g mol1. Table 4 3. Number average molecular weight (Mn, g mol 1), weight average molecular weight (Mw, g mol 1), polydispersity index (PDI), ave rage number of repeat units, and average number of rings for the donor acceptor copolymers P5 -P7 Polymer Mn (g mol1) Mw (g mol1) PDI Avg. No. of Repeat Units Avg. No. of Rings P5 27,800 59,000 2. 1 2 7 81 P6 16,000 40,800 2.6 8 40 P7 12,500 23,800 1 9 4 28 4.2.4 3,4 -Propylenedioxythiophene 2,1,3 -Benzothiadiazole Polymer Optical Characterization Similarly it was found that symmetric successive additions of Pro DOTs onto BTD induce a progressive bathochromic shift of both the shorter wavelength absorp tion band and the longer wavelength absorption band of the growing oligomers (see Figure 4 8 a). Likewise, the first substitution had the largest impact on the energy gap of the donor acceptor system, while the second substitution account ed for a finer ener gy tuning over the two-band absorption spectrum of the resulting oligomers. As illustrated in Figure 4 8 b, similar absorption changes and merging of both optical transitions take place upon increasing the length of the ProDOT electronrich block. Diminishing the concentration of the electron-poor heterocycle along the conjugated backbone gives a simultaneous bathochromic shift of the high energy transition and hypsochromic shift of the low energy transition with both peaks tending towards the max of the P8 control homopolymer resulting in the merging of the two transitions. In particular, the polymers are now all strongly absorbing at 650nm, thus covering the targeted red portion of the visible. However, in the case of P7 the higher energy tr ansition undergoes a more extensive redshift than that of the DalkOT -based polymer P3 thus opening a blue transmission in the 400

PAGE 123

123 450 nm region and does not balance the lowest energy transition as effectively as in P3 ; the colors of P5 through P8 in tol uene solution are shown on the photographs of Figure 4 8 b. Table 4 8 summarizes the local absorption maxima for P5 -P7 in toluene and as cast. a) b) Figure 4 8 a) Solution optical absorption spectra (in toluene) of M5 -M 7 (oligomer precursors for polyme r P 5 -P 7 ), b) Solution optical absorption spectra (in toluene) of DA copolymers P 5 -P 7 along with that of the control polymer P8 The spectrum of each system is normalized at the longer wavelength absorption maximum (Adapted with permission from Ref.143 Copyright 2008 Nature Publishing Group, a division of Macmillan Publishers Limited)

PAGE 124

124 In toluene solution, all polymers showed only minor optical changes upon increase of the temperature between room temperature and 100oC indicating neither significant disruption of effective conjugation115,116 nor major aggregation of the backbones once solvated.117,118 The incorporation of regiosymmetric DOEtH xProDOT s bearing two ethylhexyloxy solubilizingchains induces sufficient solubility (up to ca. 6 mg mL1 in the case of P7 ) for solid thin -film processing using common organic solvents such as toluene. Table 4 4. Local absorption maxima (solution in tolue ne and solid state) and film optical bandgap ( Eg, eV) for the donor acceptor copolymers P5 -P7 Polymer abs (nm) In Toluene abs (nm) Thin Film E g (eV) 1 2 1 2 P 5 396 644 4 08 677 (722) a 1.47 P6 4 35 6 19 4 42 642 1. 62 P 7 4 61 607 470 644 1. 63 a Shoulder 4.3 Extension of the Proposed Approach to the Synthesis of Black to Tr ansmissive Switching P olymer Electrochromes 4.3 .1 Design, Synthesis and Physical Characterization As a method to overcome the persistent lack of cover age in the blue and green regions of the visible, it was reasoned that by copolymerizing M5 and M8 (monomer of the subsequent control polymer P8 ) in a random fashion but in an appropriate feed ratio, the resulting hybrid would exhibit the extensive absorpt ion from 400 to 700 nm required for a black chromophore (see Figure 4 9 a ). This is indeed the case (see Figure 4 9 b) and an optimal ratio was found to be one equivalent of M5 to four equivalents of M8 as illustrated in Figure 4 10 and corresponding to that of the repeat unit composing the well defined analogue P7 in terms of electron rich to poor substituents incorporated. The molecular composition of the resulting polymer P9 was confirmed by NMR and elemental analysis ( see Experimental Section below) A nu mber average molecular weight of 13.6 kDa and a polydispersity of 2 were estimated by GPC. In toluene

PAGE 125

125 solution, P9 exhibits an intense ink -like dark -blue color and shows no significant optical change upon increase of the temperature b etween room temperatur e and 100oC indicating neither major aggregation of the backbone, nor disruption of effective conjugation. a) b) Figure 4 9 a) Schematic representa tion of the optical spectrum of the two band absorbing polymer P5 reflecting the color blue -green (green curve) superimposed to that of the control magenta polymer P8 (pink curve), as well as the hypothetical optical spectrum of a polymer hybrid comprising blocks of the monomers constitutive of P5 and P8 (M5 and M8 ) with various composition T he resulting hybrid (blue curve) would exhibit the extensive absorption from 400 to 700 nm required for a ink like blue or black chromophore b) Thin -film optical absorption of P9 resulting from the copolymerization of M8 and M5 (4:1 molar ratio) Figur e 4 10. Syntheti c route to ProDOT -BTD based donor acceptor black polymer electrochromes Table 4 5. Number average molecular weight ( Mn, g mol1), weight average molecular weight (Mw, g mol1), polydispersity index (PDI), average number of repeat units, and average numbe r of rings for the donor acceptor copolymer hybrid P9 Polymer Mn (g mol1) Mw (g mol1) PDI Avg. No. of Repeat Units Avg. No. of Rings P9 13,600 26,700 2 5 35

PAGE 126

126 4.3.2 Polymer Redox and Spectroelectrochemical Characterizat ion A film of P9 (Abs. Max. = 1.13 a.u.) was spray-cast onto ITO -coated glass at room temperature from toluene (4 mg mL1), and redox cycled until a stable and reproducible electrochromic switch was reached. Electrochemical oxidation of the film was carri ed out in 0.1 M LiBF4 / propylene carbonate supporting electrolyte using a silver wire as a quasi -reference e lectrode (calibrated against Fc/ Fc+) and a platinum wire as the counter electrode. As shown in Figure 4 11, when the potential applied to the film is increased from +0.04 V, simultaneous bleaching of the full set of visible absorption bands is obser ved. Figure 4 11. Spectroelectrochemistry of P9 Film was spray -cast onto ITO -coated glass from solution (4 mg mL1) in toluene. Electrochemical oxida tion of the film was carried out in 0.1 M LiBF4/PC supporting electrolyte using a silver wire as a quasi reference electrode (calibrated against Fc/Fc+) and a platinum wire as the counter electrode. The applied potential was increased in 25 mV steps from + 0.04 V to +0.74 V (Adapted with permission from Ref.143 Copyright 2008 Nature Publishing Group, a division of Macmillan Publishers Limited) This ons et of oxidation is consistent with the low value estimated at 0.08 V via cyclic and differential -pulse voltammetry (see Figures 4 12a and 4 12b) As the visible absorption bands are depleted, polaronic and bipolaronic transitions progressively arise in th e near -IR as expected

PAGE 127

127 for a colored -to -transmissive ECP. Determined from the onset of absorption of the solid thin film, the band gap of P9 was estimated at 1.64 eV. Considering the absorption maximum at 592 nm as a reference, a transmittance change of 51.5 % was measured upon electro chemical oxidation, indicative of the high degree of electrochromic contrast observed. As shown in the photographs of Figure 4 11, the spray deposited thin film of P9 (Abs. Max. 1.13 a.u.) is essentially black in its neutral sta te and attains a remarkably high level of transparency to the human eye when fully doped a) b) Figure 4 12. a) Cyclic voltammetry (CV) of a thin film of P 9 drop -cast on a Pt button electrode (10 scans). Electrochemical oxidation of the films was carri ed out in 0.1 M LiBF4/ PC supporting electrolyte using a silver wire as a quasi reference electrode (calibrated against Fc/Fc+) and a platinum flag as the counter electrode. b) Differential Pulse Voltammetry (DPV) on a thin film of P 9 drop -cast on a Pt butt on electrode. Electrochemical oxidation was carried o ut in conditions identical to CV (Adapted with permission from Ref.143 Copyright 2008 Nature P ublishing Group, a division of Macmillan Publishers Limited) 4.3.3 Polymer Colorimetric Analysis Considering the transparency observed in the doped state by spectroelectrochemical analysis the colorimetrically determined relative luminance change of P9 ( brightness of the transmitted light as a percentage of the brightness of the light source calibrated to th e sensitivity of the human eye) was measured as a function of film thickness (Abs. Max. = 1.04, 1.14 and 2.58

PAGE 128

128 a.u.) a nd doping level induced by electrochemic al oxidation as shown in Figure 4 13. The polymer exhibits an excellent relative luminance change varying from 52 % for the thinnest to 40 % for the thickest deposited film, indicating how these values can be modulat ed and optimized as a function of thickness as desired for a specific application (high color density vs. high transparency). P9 completes its full switch in a potential window of less than 0.8 V and exhibits L* values (ranging from 0 to 100) in the range 72 85 upon oxidation, demonstrative of its aptitude to nearly reach the white point (W) of color space .144 Although the residual blue hue of the neutral state characterized by the negative b* values als o reported in Figure 4 13 is not readily perceived by the human eye at the considered film thicknesses, the choice of the monomers as we ll as the feed ratios to be employed in the copolymerizations, could possibly be optimized in order to meet the most exactly black ECP. Figure 4 13. Relative luminance as a function of applied potential and film thickness for spraycoated P9 Legend ind icates the absorbance of the deposited films (estimated at the absorption maximum) and gives an indication of the thickness obtained upon spraying. For color matching, L*a*b* values (in the sense of the CIE 1976 L*a*b* color model) of fully neutral and oxi dized states are reported for the films (Adapted with permission from Ref.143 Copyright 2008 Nature Publishing Group, a division of Macmillan Publis hers Limited)

PAGE 129

129 4.3.4 Polymer Switching Study The response time of the same broadly absorbing polymer P9 sprayed onto ITO (Abs. Max. estimated at 0.9 a.u.) was characterized throughout a series of square -wave potential stepp ing electrochromic switching experiments in a LiBF4/PC solution electrolyte as described in Fig ure 4 14. By monitoring the transmittance of the spray-cast film as a function of time at both the shorter (428 nm) and longer (636 nm) wavelength absorption max ima for potential steps ranging from 10 s to 1.5 s, contrast ratios ( % T where % T is the percent optical transmittance of the film at a given wavelength) as high as 35% (428 nm) and 40% (636 nm) were observed. While a variation of electrochromic contrast of about 9% was observed as the switch time was progressively decreased to 1.5 s, it is likely that sub -second switching times will be attained upon optimization of the switching conditions (nature of the electrochemical device, identity and size of the ionic species constitutive of the electrolyte, quality / thickness of the spray -cast film, etc.). a) b) Figure 4 14. Square -wave potential step absorptometry of P9 spray -coated on ITO and monitored a) at 636 nm and b) 428nm in 0.1 M LiBF4/PC solution, 0. 6 V to +0.7 V vs. Fc/Fc+. Switch times ( ): 10 s step for 40 s (2 cycles), 5 s step for 30 s (3 cycles), 2 s step for 20 s (5 cycles), then 1.75 s step for 20 s and 1.5 s step for 20 s from the left to the right (Adapted with permission from Ref.143 Copyright 2008 Nature Publishing Group, a division of Macmillan Publishers Limited)

PAGE 130

130 As illustrated in Figure 4 1 5 the longterm switching stability of P9 was probed upon application of square -wave potential steps of 4s (complete cycle is 8s) to a film of P9 drop -cast on a glassy carbon button electrode (as P9 showed better adhesion properties on carbon than on platinum) and in 0.1 M LiBF4/PC electrolyte s olution. The film of P9 was repeatedly switched between 0.35 V and +0.65 V (vs. Fc/Fc+) and subjected to a total of 10,000 cycles upon which the film underwent less than 20% of charge density loss in atmospheric conditions. Those results should improve a s the polymer electrochrome is integrated in an electrochromic device, as opposed to an electrochemical cell where there is little to prevent the chains from slowly migrating away from the working electrode to the platinum counter electrode given the redox induced solubility of the polymer in the electrolyte during the experiment. Figure 4 1 5 Long -term stability study via square -wave potential stepping electrochromic switching of drop -cast P9 onto glassy carbon button electrode, in 0.1 M LiBF4/propylene carbonate solution switching between 0.35 V and +0.65 V (vs. Fc/Fc+), and with a switch time of 4s (Adapted with permission from Ref.143 Copyright 2008 Nature Publishing Group, a division of Macmillan Publishers Limited)

PAGE 131

131 4.4 Electrochromic Devices ( Experimental w ork was achieved and results kindly supplied by Dr Svetlana Vasilyeva) To demonstrate how P9 can be suita bly employed in ECDs, a smart window type device was made based on a dual polymer absorption/transmission architecture. T he prepared ECD consisted of t wo ITO -coated glass slides with the working electrode involving P9 spray -cast from toluene and with the counter electrode consisting of a blend of poly(2,2,6,6 tetramethylpiperidinyloxy 4 -yl methacrylate) (PTMA provided by Ciba Specialty Chemicals) and PMMA (1:4 wt ratio ) also spray -deposited from toluene ( electroactive area : 2.4 cm2) The w orking electrode bearing the cathodically -coloring polymer P9 was electrochemically oxidized prior to device assembly, while the counter electrode bearing PTMA /PMMA was neutralized to ensure the electronic balance for the redox system on repeated cycling, and increase it s long -term stability The gel electrolyte was h ighly transparent and conducting (6.8 wt% of ionic liquid 1 ethyl 3methylimidazolium bis(trifluoromethyl sulfonylimide EMI -BTI, 8.0 wt % of PMMA, 48.0 wt% PC and 37.2 wt% EC ; 5 mg of glass beads /17.5 mL gel ; c onductivity: 6.8 mS cm1), although not yet optimized for device performance The d evice w as encapsulated using a commercial curable epoxy As illustrated in Figure 4 1 6 a smart -window type device made with P9 (A = 1.3) was found to switch from a dark, es sentially black color state to a light blue oxidized state in good agreement with the spectroelectrochemical data previously shown (see F i gure 4 11). A relative luminance change of ca. 30% was estimated for a device (A = 0.6) switching between its colored (L* = 76, a* = 0, b* = 8) and clear state s (L* = 90, a* = 2, b* = 3) As shown in Figure 4 1 7 the same device (A = 0.6) showed a transmittance change of ca. 28% at the polymer absorption maximum and a subsecond switching time ( t0.95 = 0.67 s) at 95% of its full switch.

PAGE 132

132 a) b) Figure 4 1 6 Photographs of the smart window type device assembled at ambiance with P9 as the polymer electrochrome (A = 1.3) a) device in its colored state, and b) device in the bleached state (courtesy of Dr Svetlana Vasilyeva ) Figure 4 1 7 Tandem chronoabsorptometry/chronocoulometry of a P9 -based window device (A = 0.6) switched from 0.8 V to +1.6 V ( max2 = 550 nm (courtesy of Dr Svetlana Vasilyeva ) 4.5 Further Insight into the Optical Changes Observed When Using the Donor Acceptor Approach to Tailor a Two -Band Absorption in the Visible 4.5.1 Determination of the A ttenuation Coefficients of P1 P8 In Figure 4 5 b and 4 8 b of this Chapter 4 diminishing the concentration of the electronpoor heterocycle along the conjugated backbone ( i.e. increasing the length of the oligo thienylene electron -rich block) was shown to induce a simultaneous bathochromic shift of the short wavelength absorption band and hypsochromic shift of the long wavelength counterpart, with both peaks tending towards the max of the all -donor control homopolymer This phenomenon resulted in the merging of the two optical transitions, with the relative intensity of the low and high energy peaks evolv ing towards balancing each other.

PAGE 133

133 In an effort to reach a better understanding of the subtle spectral variations und ergone among the two polymer series on changing the contribution of donors and acceptors along the main -chain, polymer P1 to P8 were investigated in terms of their mass attenuation coefficient s (in g L1 cm1) in to luene which were subsequently converted to thin -film attenuation coefficient values (in cm1) by assuming a density of 1 g cm3 for all polymers. Figure 4 1 8 a and 4 1 8 b propose to illustrate the corresponding variations among P1 -P4 and P5 -P8 respective ly. Here, it is worth noting that in spite of the possible discrepancies found throughout the literature, conjugated polymers cannot be considered as small -molecule chromophores, and do not possess a well defined molecular weight which could be used in the determination of molar attenuation coefficie nts (in mol L1 cm1). In most cases, the reported molar attenuation coefficients are estimated considering the molecular weight of the polymer repeat unit as the mass of the chromophore for the calculations. H owever, conjugated polymer chromophores are obviously not limited to the polymer repeat unit, as the electronic delocalization of the system is known to extend beyond a single repeat unit. In addition, the polymers herein described possess repeat units o f various lengths ( i.e. of various masses), and the same calculations would certainly lead to results with out physical significance. A potentially more elegant alternativ e would consist in reporting attenuation coefficients directly measured from the polym er thin -films. I n this study involving subtle changes in the molecular structure of conjugated polymers the subsequent spectral changes are expected to be relatively subtle as well, and it is likely that the error arising from the film thickness measureme nts be sufficient to produce results without physical significance, here again. Following those considerations, the attenuation coefficients estimated for the polymers across the visible region, and represented in Figure 4 1 8 point towards a depletion of the longer

PAGE 134

134 wavelength absorption band with diminishing the concentration of electronaccepting heterocycle along the polymer main -chain. This depletion seems especially sudden from P1 to P2 as well as from P5 to P6 which indicate s a n especially strong donor acceptor interaction as the electron -deficient units remain in sufficiently close proximity ( i.e. as the all -donor spacers remain sufficiently short) In parallel, the shorter -wavelength absorption bands shifts toward longer wavelengths ( i.e. towards t he optical spectrum of the all -donor control polymer) with a relatively constant intensity according to the same optical data presented in Figure 4 1 8 The relative leveling off of this short -wavelength optical transition is somewhat unexpected although t he same transition can be expected to grow as rapidly as the long -wavelength transition depleted on further reducing the concentration of electron-poor heterocycle along the polymer main-chain. The distinct optical feature observed in the 750 800 nm range in the case of P5 is attributed to an intramolecular interaction as both solvent and temperature were found to alter its intensity, whereas no visible aggregation was detected. It is worth noting that the absorption coefficients of all the polymers chara cterized remain particularly high (on the order of 10,000 to 30,000 cm1) due to the extended electronic delocalization proper to conjugated in general. Finally it is essential to keep in mind that the presented optical data was obtained from mass attenua tion coefficients estimated from a given mass of material subsequently dissolved in a given volume of organic solvent. At the same time, all polymers do not possess the same concentration in solubilizing -group and as a result, the various weighted quantit ies of polymers do not contain the same quantity of chromophore (conjugated segment ideally unsubstituted). For instance, a given mass of P1 should contain a larger extent of chromophores than the same mass of P3 since one heterocycle out of three is unsubstituted in the case of P1 whereas only

PAGE 135

135 one heterocycle out of seven is unsubstituted in the case of P3 This important caveat may further be a contributing factor in the leveling off observed for the short -wavelength optical transition of P3 and P2 relat ive to P1 and of P7 and P6 relative to P5 Although the approximation made throughout this study are believed acceptable, hence leading to a fair approximation of the reality, this last observation highlights the need within the research community for a m ore accurate approach to the determination of polymer attenuation coefficients based on the concentration of actual chromophores ( i.e. without the side -chain dependence). a) b) Figure 4 1 8 a) Attenuation coefficients as a function of wavelength among t he DA -copolymer series P1 -P3 along with those for the control polymer P4 b) Attenuation coefficients as a function of wavelength among the DA -copolymer series P5 P7 along with those for the control polymer P8 The attenuation coefficients (in cm1) are derived from the mass attenuation coefficients (in g L1 cm1) measured in toluene solution by assuming a density of 1 g cm3 for all polymers

PAGE 136

136 4.5.2 Donor -Acceptor Conjugated Polymers Seen with Distinct Chromophores along the Backbone (Model #1 and #1 ) Following the considerations developed throughout the above sub-section, a model consisting in envisaging the described conjugated polymers as involving two distinct chromophores can be proposed. The corresponding model is shown as Figure 4 19. In a simp listic approximation, the optical changes schematically described in Figure 4 19a which correlate to the structural modifications induced in the DA conjugated backbone ( P1 -P3 and P5 P7 respectively), can be seen (Figure 4 19b) as the coexistence of an ele ctron rich chromophore (which set of energy bands is represented in blue) increasingly longer ( i.e. which energy gap is gradually narrowed, hence red -shifting the corresponding optical transition in the spectrum), with a chromophore (which set of energy ba nds is represented in red) resulting from the donor acceptor interaction induced by the presence of an electron deficient heterocycle. In this model, the spectral absorption relative to the chromophores produced by the DA interaction should remain relative ly constant in terms of its position on the spectrum (as each flanked accepting unit should only be able to see a limited number of neighboring donors), and should deplete as the DA chromophores are dispersed along the main-chain ( i.e. as the all -donor spa cers become longer). The diminution of DA chromophores is schematically represented by the narrowing of the width of their corresponding energy band diagram (in the order case #1, to case #2, to case #3, Figure 4 19b), to which a diminution of the number o f optical transitions occurring between the corresponding valence and conduction bands is expected to correlate. In parallel, an increase in the number of optical transitions occurring between the valence and the conduction bands of the growing electron ri ch segments (which energy gap is gradually narrowed, hence red -shifting the corresponding optical transition in the spectrum) can reasonably be expected.

PAGE 137

137 a) b) Figure 4 19. Proposed Model #1, a) Schematic representation of the evolution of the two-band spectral absorption in a series of DA conjugated polymers with varying concentration of electron -rich and electron -deficient substituents along the backbone, b) The DA conjugated polymers are seen with distinct chromophores along the backbone As a close alternative to Model #1, Model #1 proposes to see the electronaccepting heterocycles as isolated dopants which can interact with the neighboring electron -donating heterocycles to induce the formation of low -lying charge separated energy states (CT). T he corresponding model is shown as Figure 4 20. This would imply an effective formation of intramolecular charge transfer excitons stabilizing the covalently bound donor acceptor

PAGE 138

138 complexes along the polymer backbone. In this case, the long -wavelength absor ption band can be seen as the result of low -energy transitions occurring from the ground state of the polymer (S0) to the CT. And as the concentration in dopants decreases along the backbone, the concentration of accessible CTs decreases as well to the p oint where most transitions are localized on the donor segments. The diminution of dopants is schematically represented by the narrowing of the width of their energy correlation diagram (in the order case #1, to case #2, to case #3, Figure 4 20). In para llel, an increase in the number of optical transitions occurring between the valence and the conduction bands of the growing electron -rich segments (which energy gap is gradually narrowed, hence red -shifting the corresponding optical transition in the spec trum) can reasonably be expected. Additionaly, higher -energy optical transitions from energy states localized on the dopants could possibly arise in the UV. Figure 4 20. Proposed Model #1, the electronaccepting units are seen as isolated dopants which induce the formation of low lying charge separated energy states with the neighboring electron donating units

PAGE 139

139 Such a model based on the assumption of an optical transition with charge transfer characteristics is routinely implied. For instance, Jenek he et al. have strongly suggested that the two -band optical absorption of their DA systems is associated to a transition localized either on the donor or the acceptor (high -energy transition, chromophore #1), along with a transition arising from the intramolecular charge transfer inherent to the DA system (low -energy transition, chromophore #2), and identified as a spinallowed transition from the ground state of the system (S0) to a charge separated state (CT).145,146 Importantly, this interpretation assumes th at the internal electron accepting moiety is strong enough to induce the formation of an actual charge separated state which further implies that a true charge transfer process is at work within the polymer main -chain (in analogy with that described in Cha pter 1 between P3HT and PCBM in BHJs for example, although not inter but rather intra molecular). In this regard, Jenekhe et al. have occasionally provided further experimental insight probing the existence of a true charge transfer using solvatochromism s tudies on the model oligomers correlating with the repeat of the parent polymers.146 Their interpretations of the presented data and compelling conclusions remain arguable to date. In addition, a disambiguation of intramolecular charge -transfer should be provided to in dicate whether the transfer of an actual electron is involved and whether a pair of ion radicals is formed or not (in analogy with that described in Chapter 1 between P3HT and PCBM in BHJs for example). 4.5.3 D onor -Acceptor Conjugated Polymers Seen as Following Fro ntier Molecular Orbital Theory Principles (Model #2) An alternative to the model described above relies on basic Frontier Molecular Orbital Theory principles. In this instance, the DA conjugated polymer is seen as an extended array of orbitals delocalizing electrons along the backbone. The said orbitals are hybridized to produce a unique energy band diagram for the DA molecular system. The corresponding model is shown as

PAGE 140

140 Figure 4 21. In this case, the presence of two bands of absorpti on in the optical spectrum of the corresponding polymer is tentatively explained by the introduction of a discrete energy band (represented in red) in the bandgap of the semiconducting material, easily accessed from the valence band (represented in blu e and purple), and which can act as a conduction band for the system. Here, the existence of low lying unoccupied and localized energy levels within the band gap of the conjugated system in its ground -state would stem from the incorporation of electron -deficient heterocycles (bringing especially low LUMOs), with a greater concentration of accessible energy states with increasing number of accepting moieties along the main -chain (in the order case #3, to case #2, to case #1, Figure 421b). As a result, the low -energy transitions to the discrete energy band can be envisaged to produce a long wavelength absorption band, whereas high -energy transitions arising from the valence band (which is governed by the donors, and represented in blue) to the electron ric h segments first unoccupied energy levels (also represented in blue) can be envisaged to produce a blue -shifted absorption band in the visible spectrum. In this model, the optical absorption emerging from the low -energy transitions could undergo a progres sive hypsochromic shift in the visible spectrum, and depletion in terms of its intensity, with diminishing number of electron -deficient heterocycles along the main -chain ( i.e. as the all -donor spacers become longer, the extent of hybridization between acce pting units is reduced, and the concentration of mid-gap energy levels accessible decreases). The diminution of accessible energy levels within the discrete energy band is schematically represented by the narrowing of its size in the energy correlation diagram of the DA molecular system (in the order case #1, to case #2, to case #3, Figure 4 21b), to which a diminution of the number of related optical transitions is expected to correlate. In parallel, an increase in the number of optical transitions occ urring between the valence band and the first unoccupied energy levels governed

PAGE 141

141 by the growing electron -rich segments can reasonably be expected (with the energy gap between the valence band and those unoccupied energy levels being gradually narrowed, henc e red shifting the corresponding optical transition in the spectrum). ) b) Figure 4 21. Proposed Model #2, a) Schematic representation of the evolution of the two-band spectral absorption in a series of DA conjugated polymers with varying concentration of electron -rich and electron -deficient substituents along the backbone, b) The DA conjugated polymers are seen as following Frontier Molecular Orbital Theory principles

PAGE 142

142 This approach has been investigated by Salzner et al. for example, throughout theo retical work aimed at addressing the relevance of the donor acceptor approach for designing organic metals (see Chapter 3).107,108 It has been considered as the most likely approach so far throughout this dissertation work. 4.5.4 In Summary Although distinct, n either of the two models developed above should be excluded at this point i n the description of the em pirical results obtained from the two analogous polymer series investigated In fact, i t is likely that the reality lies somewhere in between those two distinct models, the examination of which could create an exciting challenge for the community of theore ticians looking to better understand the specificities of conjugated systems in general .107,108,147 152 4.6 Synthetic Details 4.6 .1 Synthesis of P1 -P3 Compound 2: Sodi um metal (40 g) was dissolved in anhydrous methanol (500 mL). 3, 4 dibromothiophene 1 (102.6 g, 428 mmol) was added to the alkaline solution at room temperature. CuI (20.4 g, 107 mmol) was then added quickly, and the solution heated to reflux for 72 hours. The reaction mixture was cooled down to room temperature, added to water and extracted multiple times with diethyl ether. The organic phase was washed with brine, water, dried over magnesium sulfate and the solvent was removed affording a dark -yellow resi due. The resulting oil was purified via distillation under reduced pressure and compound 2 was obtained as a clear oil (62.5 g, 68 %). 1H NMR (300 MHz, CDCl3) = 6.19 (s, 2H), 3.86 (s, 6H); 13C NMR (75MHz, CDCl3) = 150.78, 96.48, 57.80. Compound 3 (M4): Compound 2 (62.37 g, 433 mmol), 2 -ethylhexan 1 -ol (225.74 g, 1.73 mol) and p -toluenesulfonic acid (8.24 g, 43.3 mmol) were charged into 500 mL of dry toluene

PAGE 143

143 and placed under reflux. The refluxing mixture was strongly flushed with argon every 24 hours, hen ce evacuating the generated methanol and displacing the equilibrium to the formation of 3 This process was controlled by TLC and repeated until completion of the reaction. The dark yellow organic phase was washed with water (became dark green) and dried o ver anhydrous magnesium sulfate Toluene was removed by rotary evaporation and the resulting dark oil was purified by column chromatography on silica with hexanes as eluent. T he solvent was evaporated and compound 3 was obtained as a clear oil ( 90 %). 1H N MR (300 MHz, CDCl3) = 6.16 (s, 2H), 3.86 (d, J = 5.9 Hz, 4H), 1.77 1.71 (m, 2H), 1.551.30 (m, 18H), 0.94 (m, 12H) ; 13C NMR (75MHz, CDCl3) = 148.25, 97.08, 73.35, 39.47, 30.81, 29.30, 24.17, 23.27, 14.30, 11.38. Anal. calcd. for C20H36O2S1: C 70.53, H 10.65 Found: C 70.52, H 10.76. Compound 4 : Compound 3 (4.31 g, 12.62 mmol ) was dissolved in dry THF (50 mL) and cooled to 78 C. A solution of n-butyllithium in hexanes (7.12 mL, 16.12 mmol ) was added over a 2 hour period and the mixture was stirred for 1 h our at 78 C. Trimethyltinchloride (3.23 g, 16.12 mmol ) was subsequently added, the mixture was allowed to warm up to room temperature and stirred for 16 hours. The solvent was evaporated affroding a yellow -brown oil (6.0 g, 95 %) which was used for the n ext step without further purification. 1H NMR (300 MHz, CDCl3) = 6.40 (s, 1H), 3.88 (d, J = 5.92 Hz, 2H), 3.83 (d, J = 5.2 Hz, 2H), 1.771.30 (m, 18H), 0.90 (t, J = 6.8 Hz, 12H), 0.32 (s, 9H). Compound 6a (M1 ): Compound 4 (4.3 g, 8.53 mmol), 4,7 dibromobe nzo[ c ][1,2,5]thiadiazole 5a (1.03 mg, 3.54 mmol) and Pd(PPh3)2Cl2 (100 mg, 4 mol %) were dissolved in 60 mL of THF. The mixture was stirred for 36 hours at 60 C, the solvent was evaporated and the product was purified by column chromatography on silica wi th hexane/dichloromethane (3:2) as eluent. T he solvent was evaporated and compound 6a was

PAGE 144

144 obtained as a bright orange oil ( 2.53 g, 88 %). 1H NMR (300 MHz, CDCl3) = 8.35 (s, 2H), 6.37 (s, 2H), 4.01 (d, J = 5.6 Hz, 4H), 3.92 (d, J = 5.5 Hz, 4H), 1.770.81 ( m, 60H); 13C NMR (75MHz, CDCl3) = 153.11, 150.72, 145.35, 128.12, 124.80, 120.62, 98.13, 75.42, 72.46, 40.49, 39.75, 30.88, 30.65, 29.36, 29.33, 24.19, 23.31, 23.27, 14.32, 14.27, 11.41, 11.28. HRMS (TOF) [MH+] m/z calcd. for C46H72N2O4S3: 813.4688 Found: 813.4675. Anal. calcd. for C46H72N2O4S3: C 67.93, H 8.92, N 3.44 Found: C 68.26, H 8.92, N 3.44. Compound 5b : Compound 6a (4.6 g, 5.66 mmol) was dissolved in 200 mL of CHCl3. Then NBS (2.22 g, 12.45 mmol) was added by portions and the mixture was stirred for 24 h at room temperature under light protection. The organic layer was washed with water, dried over magnesium sulfate the solvent was evaporated and the product was purified by column chromatography on silica with hexane/dichloromethane (3:1) as elue nt. T he solvent was evaporated and compound 5b was obtained as a dark red oil ( 10.5 g, 96 %). 1H NMR (300 MHz, CDCl3) = 8.35 (s, 2H), 4.06 (d, J = 5.7 Hz, 4H), 3.97 (d, J = 5.6 Hz, 4H), 1,761.22 (m, 36H), 1.200.81 (m, 24H); 13C NMR (75MHz, CDCl3) = 152.52, 148.69, 148.05, 127.61, 124.62, 121.06, 100.35, 76.78, 76.51, 40.53, 40.49, 30.50, 30.47, 29.36, 29.30, 23.82, 23.36, 23.29, 14.39, 14.28, 11.40, 11.26. HRMS (MMI TOF) [MH+] m/z calcd. for C46H70Br2N2O4S3: 969.2937 Found: 969.2938. Anal. calcd. for C46H70Br2N2O4S3: C 56.90, H 7.27, N 2.91 Found: C 57.17, H 7.20, N 2.91. Compound 6b (M2) : Compound 4 (5.52 g, 10.96 mmol), compound 5b (4.42 g, 4.55 mmol) and Pd(PPh3)2Cl2 (130 mg, 4 mol %) were dissolved in 60 mL of THF. The mixture was stirred for 36 hou rs at 60 C, the solvent was evaporated and the product was purified by column chromatography on silica with hexane/dichloromethane (3:1) as eluent. T he solvent was evaporated and compound 6b was obtained as a purple oil ( 4.16 g, 62 %). 1H NMR (300 MHz,

PAGE 145

145 CD Cl3) = 8.36 (s, 2H), 6.13 (s, 2H), 4.103.87 (m, 16H), 1.971.22 (m, 72H), 1.000.81 (m, 48H); 13C NMR (75MHz, CDCl3) = 153.04, 150.21, 148.42, 145.96, 143.09, 127.64, 124.55, 120.22, 119.14, 117.38, 95.30, 75.80, 72.40, 40.51, 39.79, 32.86, 31.82, 30.50, 30.47, 29.36, 29.30, 23.82, 23.36, 23.29, 14.39, 14.28, 11.42, 11.41 11.23. HRMS (MMI TOF) [MH+] m/z calcd. for C86H140N2O8S5: 1489.9240 Found: 1489.9191. Anal. calcd. for C86H140N2O8S5 C 69.31, H 9.47, N 1.88 Found: C 69.68, H 9.70, N 1.92. Compound 5c : Compound 6b (2.22 g, 1.5 mmol) was dissolved in 150 mL of CHCl3. Then NBS (587 mg, 3.3 mmol) was added by portions and the mixture was stirred for 24 h at room temperature under light protection. The organic layer was washed with water, dried over magnes ium sulfate the solvent was evaporated and the product was purified by column chromatography on silica with hexane/dichloromethane (4:1) as eluent. T he solvent was evaporated and compound 5c was obtained as a dark purple oil ( 2.1 g, 85 %). 1H NMR (300 MHz CDCl3) = 8.37 (s, 2H), 4.043.90 (m, 16H), 1.971.22 (m, 72H), 1.000.81 (m, 48H); 13C NMR (75MHz, CDCl3) = 152.91, 148.31, 148.12, 146.14, 145.14, 127.64, 124.42, 119.53, 119.12, 109.87, 96.78, 40.48, 40.45, 34.89, 30.50, 30.47, 29.36, 29.30, 23.82, 23.36, 23.29, 14.37, 14.37, 14.29, 11.42, 11.22. HRMS (MMI TOF) [MH+] m/z calcd. for C86H138N2O8S5: 1644.7418 Found: 1644.7525. Anal. calcd. for C86H138N2O8S5: C 62.67, H 8.44, N 1.70 Found: C 62.58, H 8.40, N 1.81. Compound 6c (M3 ): Compound 4 (400 mg, 0. 79 mmol), compound 5c (512 mg, 0.31 mmol) and Pd(PPh3)2Cl2 (10 mg, 4 mol %) were dissolved in 40 mL of THF. The mixture was stirred for 36 hours at 60 C, the solvent was evaporated and the product was purified by column chromatography on silica with hexane/dichloromethane (4:1) as eluent. T he solvent was evaporated and compound 6c was obtained as a dark blue tacky solid ( 356 mg, 53 %). 1H NMR

PAGE 146

146 (300 MHz, CDCl3) = 8.40 (s, 2H), 6.10 (s, 2H), 4.01 3.86 (m, 24H), 1.91 1.23 (m, 108H), 1.30 (m, 72H); 13C NMR (75 MHz, CDCl3) = 152.99, 150.13, 148.43, 146.56, 146.27, 145.59, 142.99, 127.53, 124.41, 119.59, 119.26, 117.27, 117.00, 115.31, 95.08, 76.52, 65.67, 72.40, 40.44, 40.33, 39.72, 30.84, 29.81, 29.68, 29.54, 26.41, 22.90, 14.35, 14.32, 14.26, 11.43, 11.38, 11.22. HRMS (MMI TOF) [M+] m/z calcd. for C126H208N2O12S7: 2165.3767 Found: 2165.3856. Anal. calcd. for C126H208N2O12S7: C 69.82, H 9.67, N 1.29 Found: C 70.50, H 9.78, N 1.33. Polymer P1 (7a): Compound 6a (445 mg, 0.64 mmol) was dissolved in chloroform (45 mL). A solution of anhydrous FeCl3 (444 mg, 2.74 mmol, 5eq) in nitromethane was added dropwise over a period of 45 minutes to the stirred monomer at room temperature (the bright orange monomer solution turned progressively dark blue with addition of oxidiz ing agent). The mixture was stirred 48 hours at room temperature. It was then precipitated into methanol (300 mL). The precipitate was filtered, redissolved in chloroform (300 mL) and stirred for 6 hours with hydrazine monohydrate (6 mL). After evaporation the concentrate (dark blue) was precipitated into methanol (300 mL), the precipitate was filtered through a Soxhlet thimble and purified via Soxhlet extraction for 48 hours with methanol. The polymer was extracted with chloroform, concentrated by evapora tion, the polymer was precipitated in methanol (300 mL) and collected as a dark blue solid (373 mg, 84 %). 1H NMR (300 MHz, CDCl3) = 8.47 (bs, 2H), 4.2 3.9 (br, 8H), 2.2 1.1 (br, 36H), 1.1 0.7 (br, 24H) GPC analysis: See Table 4 1 Anal. calcd. for C46H70N2O4S3: C 68.10, H 8.70, N 3.35 Found: C 67.63, H 8.58, N 3.51. Polymer P2 (7b): Compound 6b (606 mg, 0.41 mmol) was dissolved in chloroform (60 mL). Following the procedure described for synthesizing polymer P1 ( 7a ) and using methanol/water mixtures for the successive precipitations, polymer P2 was obtained as a dark green tacky solid (430 mg, 71 %). 1H NMR (300 MHz, CDCl3) = 8.42 (bs, 2H), 4.13.9 (br,

PAGE 147

147 16H), 2.11.1 (br, 72H), 1.10.7 (br, 48H) GPC analysis: See Table 4 1 Anal. calcd. for C86H138N2O8S5: C 69.40, H 9.35, N 1.88 Found: C 69.58, H 9.54, N 1.95. Polymer P3 (7c): Compound 6c (205 mg, 0.1 mmol) was dissolved in chloroform (30 mL). Following the procedure described for synthesizing polymer P1 ( 7a ) and using methanol/water mixtures for the suc cessive precipitations, polymer P7 was obtained as a black tacky solid (150 mg, 73 %). 1H NMR (300 MHz, CDCl3) = 8.41 (bs, 2H), 4.13.9 (br, 24H), 2.1 1.1 (br, 108H), 1.1 0.7 (br, 72H) GPC analysis: See Table 4 1 Anal. calcd. for C126H206N2O12S7: C 69.89, H 9.59, N 1.29 Found: C 69.55, H 9.60, N 1.54. 4.6 .2 Synthesis of P5 -P7 Compound 2 (M8): A 1000 mL flame dried single neck round bottom flask was filled with 450 mL of anhydrous DMF, NaH (60% in mineral oil) (23.85 g, 596 mmol) and 2ethylhexan 1 -ol (51.8 g, 398 mmol). The mixture was heated at 90 C over a period of 3 hours, then 3,3-bis(bromomethyl) 3,4 -dihydro 2H thieno[3,4 b][1,4]dioxepine ( 1 ) (34 g, 99.4 mmol) was added by portions to the hot reagent. After stirring 24 hours at 95 C, the reaction mixture was cooled down to room temperature, added to brine and extracted 4 times with diethyl ether. The organic phase was washed with water, dried over magnesium sulfate and the solvent was removed affording a brown oil. The resulting oil was purified by column chromatography on silica with hexane/dichloromethane (3:2) as eluent. T he solvent was evaporated and compound 2 was obtained as a clear oil (62 g, 71 %). 1H NMR (300 MHz, CDCl3 = 6.42 (s,2H), 4.02 (s, 4H), 3.49 (s, 4H), 3.33 (d, J = 5.6 Hz, 4H), 1.571.29 (m, 18H) 0.94 (m, 12H); 13C NMR (75MHz, CDCl3 23.36, 14.35, 11.40. HRMS [M+Na+] m/z calculated fo r C25H44O4S: 463.2853 Found: 463.2939.

PAGE 148

148 Compound 3 : Compound 2 (2.13 g, 4.83 mmol ) was dissolved in dry THF (30 mL) and cooled down to 78 C. A solution of nbutyllithium in hexanes (3.32 mL, 5.94 mmol ) was added over a 2 hour period and the mixture was st irred for 1 hour. Trimethyltinchloride (1.2 g, 5.99 mmol ) was subsequently added, the mixture was allowed to warm up to room temperature and stirred for 16 hours. The solvent was evaporated affording a yellow brown oil (2.76 g, 95 %) which was used for the next step without further purification. 1H NMR (300 MHz, CDCl3) = 6.70 (s, 1H), 3.98 (s, 2H), 3.93 (s, 2H), 3.46 (s, 4H), 3.29 (d, J = 5.6 Hz, 4H), 1.501.27 (m, 18H), 0.94 (m, 12H), 0.32 (s, 9H). Compound 5 a (M5 ): Compound 10 (3.2 g, 5.31 mmol), 4,7 dibromobenzo[ c ][1,2,5]thiadiazole 4 a (623 mg, 2.81 mmol) and Pd(PPh3)2Cl2 (63 mg, 4 mol %) were dissolved in 60 mL of THF. The mixture was stirred for 36 hours at 60 C, the solvent was evaporated and the product was purified by column chromatography on silic a with hexane/dichloromethane (4:1) as eluent. T he solvent was evaporated and compound 12a was obtained as a bright orange oil ( 2.27 g, 80 %). 1H NMR (300 MHz, CDCl3) = 8.30 (s, 2H), 6.66 (s, 2H), 4.21 (s, 4H), 4.11 (s, 4H), 3.55 (s, 8H), 3.33 (d, J = 5.6 Hz, 8H), 1.30 (m, 36H), 0.92 (m, 24H); 13C NMR (75MHz, CDCl3) = 152.88, 149.98, 148.08, 127.79, 124.44, 117.82, 106.71, 74.56, 74.51, 74.05, 70.19, 48.14, 39.87, 30.90, 29.81, 29.35, 29.54, 24.26, 3.34, 14.35, 11.40. HRMS (MMI TOF) [MH+] m/z calcd. for C56H88N2O8S3: 1013.5776 Found: 1013.5745. Anal. calcd. for C56H88N2O8S3: C 66.36, H 8.75, N 2.76 Found: C 66.39, H 8.88, N 2.71. Compound 4 b: Compound 5 a (5.48 g, 5.4 mmol) was dissolved in 250 mL of CHCl3. Then NBS (2.10 g, 11.88 mmol) was added by portions and the mixture was stirred for 24 h at room temperature under light protection. The organic layer was washed with water, dried over magnesium sulfate the solvent was evaporated and the product was purified by column

PAGE 149

149 chromatography on silica with hexane/dichloromethane (3:1) as eluent. T he solvent was evaporated and compound 4 b was obtained as a dark red oil ( 5.2 g, 82 %). 1H NMR (300 MHz, CDCl3) = 8.30 (s, 2H), 4.23 (s, 4H), 4.18 (s, 4H), 3.56 (s, 8H), 3.43 (d, J = 5.5 Hz, 8H), 1.521.32 (m, 36H), 0.92 (m, 24H); 13C NMR (75MHz, CDCl3) = 152.38, 147.81, 147.24, 127.40, 123.70, 117.54, 96.45, 74.56, 74.51, 74.37,74.24, 70.13, 48.16, 39.88, 30.90, 29.93, 29.36,24.26, 23.34, 14.36, 11.42. HRMS (MMI TOF) [M+Na+] m/z calcd. for C56H86Br2N2O8S3: 1191.3805 Fo und: 1191.3786. Anal. calcd. for C56H86Br2N2O8S3: C 57.42, H 7.40, N 2.39 Found: C 57.86, H 7.40, N 2.39. Compound 5 b (M6 ): Compound 3 (3.18 g, 5.26 mmol), compound 4 b (2.94 mg, 2.51 mmol) and Pd(PPh3)2Cl2 (75 mg, 4 mol %) were dissolved in 60 mL of THF. The mixture was stirred for 36 hours at 60 C, the solvent was evaporated and the product was purified by column chromatography on silica with hexane/dichloromethane (2:1) as eluent. T he solvent was evaporated and compound 5 b was obtained as a purple oil ( 3.51 g, 74 %). 1H NMR (300 MHz, CDCl3) = 8.28 (s, 2H), 6.43 (s, 2H), 4.23 (s, 4H), 4.20 (s, 8), 4.05 (s, 4H), 3.58 (s, 8H), 3.55 (s, 8H), 3.33 (m, 16H), 1.521.27 (m, 72H), 0.90 0.83 (m, 48H); 13C NMR (75MHz, CDCl3) = 152.96, 149.65, 147.86, 145.72, 145.50, 127.74, 124.25, 117.15, 115.93, 115.47, 103.81, 74.54, 74.30, 70.32, 48.23, 48.17, 39.89, 39.86, 34.89, 31.80, 30.92, 29.36, 25.50, 24.25, 23.34, 23.31, 22.87, 21.37, 14.34, 14.32, 11.62, 11.40. HRMS (MMI TOF) [M+] m/z calcd. for C106H172N2O16S5: 1889.1305 Found: 1889.1324. Anal. calcd. for C106H172N2O16S5: C 67.33, H 9.17, N 1.48 Found: C 67.48, H 9.28, N 1.49. Compound 4c : Compound 5 b (1 g, 0.53 mmol) was dissolved in 150 mL of CHCl3. Then NBS (207 mg, 1.16 mmol) was added by portions and the mixtur e was stirred for 24 h at room temperature under light protection. The organic layer was washed with water, dried over

PAGE 150

150 magnesium sulfate the solvent was evaporated and the product was purified by column chromatography on silica with hexane/dichloromethane (1:1) as eluent. T he solvent was evaporated and compound 4c was obtained as a dark purple tacky solid ( 975 mg, 90 %). 1H NMR (300 MHz, CDCl3) = 8.28 (s, 2H), 4.22 (s, 4H), 4.20 (s, 8), 4.13 (s, 4H), 3.58 (s, 8H), 3.56 (s, 8H), 3.32 (m, 16H), 1.52 1.27 (m 72H), 0.900.83 (m, 48H); 13C NMR (75MHz, CDCl3) = 152.96, 147.86, 147.17, 145.72, 145.50, 127.64, 124.35, 116.85, 114.76, 114.27, 92.93, 84.61, 83.00, 74.54 70.32, 48.23, 48.17, 39.89, 39.86, 34.89, 31.80, 30.92, 29.36, 25.50, 24.25, 23.36, 23.31, 22. 87, 14.63, 14.32, 11.47, 11.36. HRMS (MMI TOF) [M+] m/z calcd. for C106H170Br2N2O16S5: 2044.9515 Found: 2044.9443. Anal. calcd. for C106H170Br2N2O16S5: C 62.15, H 8.36, N 1.37 Found: C 62.63, H 8.40, N 1.37. Compound 5 c (M7 ): Compound 3 (404 mg, 0.67 mmol), compound 4 c (370 mg, 0.18 mmol) and Pd(PPh3)2Cl2 (8 mg, 4 mol %) were dissolved in 60 mL of THF. The mixture was stirred for 36 hours at 60 C, the solvent was evaporated and the product was purified by column chromatography on silica with hexane/dichl oromethane (2:1) as eluent. T he solvent was evaporated and compound 5 c was obtained as a dark blue -green tacky solid ( 324 mg, 65 %). 1H NMR (300 MHz, CDCl3) = 8.28 (s, 2H), 6.40 (s, 2H), 4.27 (s, 4H), 4.23 (s, 8H), 4.16 (s, 8), 4.07 (s, 4H), 3.673.57 (m 24H), 3.32 (m, 24H), 1.521.29 (m, 108H), 0.90 0.87 (m, 72H); 13C NMR (75MHz, CDCl3) = 152.97, 149.71, 148.13, 147.97, 145.79, 145.72, 145.47, 145.32, 145.23, 127.62, 124.20, 117.52, 116.32, 115.79, 114.31, 113.55, 103.57, 74.56, 74.48, 48.24, 48.17, 48.13, 39.84, 30.94, 30.87, 29.94, 29.38, 24.26, 23.36, 14.38, 11.43, 11.40, 11.36. HRMS (MMI TOF) [MH+] m/z calcd. for C156H256N2O24S7: 2766.6991 Found: 2766.7058. Anal. calcd. for C156H256N2O24S7: C 67.69, H 9.23, N 1.01 Found: C 68.11, H 9.66, N 1.06.

PAGE 151

151 Po lymer P5 (6 a): Compound 5 a (650 mg, 0.64 mmol) was dissolved in chloroform (65 mL). A solution of anhydrous FeCl3 (519 mg, 3.2 mmol, 5eq) in nitromethane was added dropwise over a period of 45 minutes to the stirred monomer at room temperature (the bright orange monomer solution turned progressively dark blue with addition of oxidizing agent). The mixture was stirred 48 hours at room temperature. It was then precipitated into methanol (300 mL). The precipitate was filtered, redissolved in chloroform (300 mL ) and stirred for 6 hours with hydrazine monohydrate (6 mL). After evaporation, the concentrate (dark blue) was precipitated into methanol (300 mL), the precipitate was filtered through a Soxhlet thimble and purified via Soxhlet extraction for 48h with met hanol. The polymer was extracted with chloroform, concentrated by evaporation, the polymer was precipitated in methanol (300 mL) and collected as a dark blue solid (475 mg, 73 %). 1HNMR ( 300 MHz, CDCl3) = 8.35 (bs, 2H), 4.30 (bs, 8H), 3.66 (bs, 8H), 3.36 (bs, 8H), 1.31 (bs, 36H), 0.91 (bs, 24H). GPC analysis: See Table 4 3 Anal. calcd. for C56H86N2O8S3: C 66.50, H 8.57, N 2.77 Found: C 66.21, H 8.48, N 2.73. Polymer P6 (6 b): Compound 5 b (465 mg, 0.25 mmol) was dissolved in chloroform (45 mL). Following the procedure described for synthesizing polymer P5 ( 6 a ), polymer P6 was obtained as a dark green solid (400 mg, 86 %). 1HNMR ( 300 MHz, CDCl3) = 8.33 (bs, 2H), 4.23 (bs, 16H), 3.63 (bs, 16H), 3.35 (bs, 16H), 1.31 (bs, 72H), 0.90 (bs, 48H). GPC analysis: See Table 4 3 Anal. calcd. for C106H170N2O16S5: C 67.40, H 9.07, N 1.48 Found: C 66.35, H 9.15, N 1.57. Polymer P7 (6 c): Compound 5 c (400 mg, 0.14 mmol) was dissolved in chloroform (40 mL). Following the procedure described for synthesizing polymer P5 ( 6 a ) and using methanol/water mixtures for the successive precipitations, polymer P7 was obtained as a dark

PAGE 152

152 solid (480 mg, 83 %). 1HNMR ( 300 MHz, C DCl3) = 8.30 (bs, 2H), 4.20 (bs, 24H), 3.62 (bs, 24H), 3.33 (bs, 24H), 1.29 (bs, 108H), 0.90 (bs, 72H). GPC analysis: See Table 4 3 Anal. calcd. for C156H254N2O24S7: C 67.74, H 9.26, N 1.01 Found: C 67.28, H 9.19, N 1.54. 4.6 3 Synthesis of P9 Polymer P9: Compound 5 a (M5) (150 mg, 0.15 mmol) and compound 2 (M8) (260.9 mg, 0.59 mmol, 4eq) were dissolved in chloroform (45 mL). A solution of anhydrous FeCl3 (600 mg, 3.7 mmol, 5eq) in nitromethane was added dropwise over a period o f 1 hour to the stirred monomer at room temperature (the bright orange monomer solution turned progressively opaque with addition of oxidizing agent). The mixture was stirred 24 hours at room temperature and 6 additional hours at 40 C It was then precipi tated into methanol (300 mL). The precipitate was filtered, redissolved in chloroform (300 mL) and stirred for 6 hours with hydrazine monohydrate (6 mL). After evaporation, the concentrate was precipitated into methanol (300 mL), the precipitate was filter ed through a Soxhlet thimble and purified via Soxhlet extraction for 48 hours with methanol. The polymer was extracted with chloroform, concentrated by evaporation, the polymer was precipitated in methanol (300 mL) and collected as a black solid (378 mg, 9 2 %). 1HNMR ( 300 MHz, CDCl3) = 8.35 (bs, 2H), 4.17 (bs, 24H), 3.62 (bs, 24H), 3.34 (bs, 24H), 1.29 (bs, 108H), 0.90 (bs, 72H). GPC analysis: See Table 4 5 Anal. calcd. for C156H254N2O24S7: C 67.74, H 9.26, N 1.01 Found: C 66.89, H 9.00, N 1.28. 4. 7 Conclusions and Outlook Following up on the donor acceptor theory introduced in 1993 by Havinga et al. ,42 this simple approach towards making easily oxidized conjugated polymers wit h broad and homogeneous absorption bandwidths extending over the entire visible spectrum which can be fully bleached has implications that span numerous applications including electrochromic windows and displays. In brief, this investigation of the use of the DA approach in designing

PAGE 153

153 novel ECPs exhibiting a two -band absorption in the visible, so as to reflect and / or transmit colors commonly difficult to achieve such as saturated greens, has led to the conclusion that low and high energy transitions can be controlled in a substantial and interconnected fashion by varying the relative contribution of electron rich and poor moieties incorporated in the repeat unit. In addition, a merging of the bands was observed that offers the potential for the synthesis of neutral state colored materials possessing either highly saturated or darker colors. The solution processability of all of the polymers prepared is important when considering ultimate utility relative to the often studied electropolymerized systems.104 Taking advantage of these empirical considerations, a copolymer possessing an absorption spectrum extended over the entire visible region (400 700nm) was synthesized and its redox p roperties investigated. It was found that the copolymer achieves reversible EC switching from a strong opaque neutral black color state to a highly transmissive oxidized state over a potential window of less than 1 V, a desirable parameter for low -voltage device applications. T he first neutral state black polymeric electrochrome with potential applications in e -papers and smart windows was thereby demonstrate d Finally the band-merging effect outlined above impacts other potential uses, such as the development of solution-processable and narrow band -gap precursors absorbing throughout the visibl e and into the near IR for organic polymer solar cells

PAGE 154

154 CHAPTER 5 FINE BANDGAP TUNING IN 3,4 DIOXYTHIOPHENE 2,1,3 BENZOTHIADIAZOLE DONOR ACCEPTOR ELECTROCHRO MIC POLYMERS VIA THE USE OF UNSATURATED LINNKAGE S: FROM BLUE TO GREE N 5.1 Motivations for Simplifyi ng the Access to Colors Difficult to Achieve and Proposed Design As described in Chapter 3 of this dissertation, t he perspective of generating new colors commonly difficult to attain such as cyan blue and greens o f va rious tones and saturations constitutes a motivation for designing conjugated polymer s finely tuned in their molecular structure to reassemble the optical features desired in the context of non -emis sive ECD applications. While Chapter 3 set the context and introduced a method to overcome the technological obstacles hindering the synthesis of neutral state green polymer electrochromes with highly transmissive oxidized states efforts to simplify the synthetic access to those colors should be envisaged. At th is point, it is also worth noting that neutral state deep -blue PEDOT which switches to a transmissive sky-blue color state is not a solution -processable ECP, and is commonly electropolymerized in -situ (in contrast with its doped aqueous dispersion153) In addition, its near -IR optical transitions overlap with the visible, decreasing the transmissivity of the p -doped state considerably in devices. Here, an alternative blue -to -colorless and solutionprocessable ECP candidate ( see work initiated by Thompson et al.112) could be optimized to anticipate the path to commercialization. In the following Chapter 5 a serie s of soluble DA conjugated polymers involving 3,4 dioxythiophenes (DOTs) and 2,1,3 -benzothiadiazole (BTD) constructed in combination with unsaturat ed linkages, namely ethynylene and trans -ethylene (see Figure 5 1) will be presented and compare d to their fully polyheterocyclic DA control analogues with careful emphasis on optical, elect rochemical, and electrochromic properties. In parallel, this syst ematic study should

PAGE 155

155 give further insight into the effect of introducing unsaturated linkers in the backbone of cathodically -co loring ECPs In particular, the presence of a spacer relaxing the extent of steric hindrance along the polyheterocyclic backbone w hile retaining its conjugation could offer an alternative pathway to narrowing bandgaps in polymer electrochromes, in the situation where the said spacer does not impede the polymer EC performance. Importantly, m ore than just finely tuning the optical prop erties of fully polyheterocyclic D A polymers incorporation of the unsaturated linker could open a pathway toward the synthesis of ECPs made of multiple chromophores randomly copolymerized to give access to any possible hue, including the color black such as desired for electronic ink -based applications. Figure 5 1. Designing 3,4 dioxythiophene (DOT) and 2,1,3-benzothiadiazole (BTD) based polymers ( P1 -P6 ) whereby the DOTs are spaced by an unsaturated spacer to limit the extent of steric hindrance along the polyheterocyclic backbone and offer an alternative pathway to narrowing bandgaps in ECPs (Adapted with permission from Ref.128 Copyright 2009 American Chemical Society) 5.2 Synthesis and Characterization of 3,4 -Diox ythiophene 2,1,3 -Benzothiadiazole Polymers Containing Unsaturated Linkages 5. 2 .1 Synthesis and Physical Characterization ( Experimental w ork on MALDI -MS was achieved and results kindly supplied by Dr Tracy McCarley)

PAGE 156

156 The fully heterocyclic model polymers P1 P4a and P4b were synthesized from the oxidative polymerization of macromonomers M1 and M2 (respectively), using the mild oxidizing agent FeCl3, before being subsequently reduced with hydrazine (See Figure 5 2 ). The d et ails concerning the experimental conditions employed in the synthesis of M1 M2 P1 P4a and P4b can be found in Chapter 4 of this dissertation. The unsaturated spacer -containing polymers P2 P3 P5 and P6 were accessed from the halogenated analogues of M1 and M2 namely M1 and M2 using traditional Stille cross coupling polymerization conditions111 between the stannane derivatives of ethynylene an d trans ethylene and the macromonomers (See Figure 5 2 ). Refluxing mixtures of toluene and dimethylformamide (DMF) were found suitable in the case of these condensations between the bulky, sterically hindered heterocyclic macromonomers and their ethylenic/ ethynylenic comonomers, conversely small in size, hence relatively prone to transmetallation mechanisms. Figure 5 2 a) Synthetic route to DalkOT BTD based DA -conjugated copolymers P1 -P3 containing unsaturated linkers (P2 and P3 ) or not ( P1 ) b) Synthetic route to ProDOT BTD based DA -conjugated copolymers P4 P6 containing unsaturated linkers ( P5 and P6 ) or not ( P4a P4b ) (Adapted with permission from Ref.128 Copyright 2009 American Chemical Society)

PAGE 157

157 The repeat unit structures of all polymers a re supported by 1H NMR (see Experimental Section) and matrix assisted laser desorption / ionization mass spectrometry (MALDI -MS). Figure 5 3 shows the MALDI mass spectra of P1 P4a P3 and P6 as representative examples. Using terth iophene84,154 or trans 2 -[3 (4 tert -butylphenyl) 2 -methyl 2 propenylidene]malononitrile (DCTB)155 as the matrix, ions with masses up to nearly 9,000 amu, 12,000 amu and 20,000 amu were respectively detected for the control polymers P1 P4a and P4b As illustrated in Figure 5 3 a, the mass spectrum of P1 shows an ion series with a spacing of ~1,622 amu between each ion, corresponding to the calculated mass of two repeat units ( i.e. one oxidative condensation). Each major ion detected consists of an odd number of repeat units (n=3,5,7,9) terminated by two chlorines (Note that the reasons for the detection of an odd number of repeat units only are not clear to date, and a ra tionale is currently being explored by our collaborator Dr Tracy McCarley). End -capping via chlorine atoms is not unusual during FeCl3 assisted oxidative polymerizations and has been demonstrated in previous work from McCarley et al.84 In Figure 5 3 b, P4a is represented by very intense peaks at m/z 3,037 and m/z 3,107 which correspond respectively to trimeric chains possessing H/H and Cl/Cl end groups. The peak at m/z 3,107 is part of an ion series with intervals matching the mass of the polymer repeat unit (~1011 amu). This series ranges from m/z 3,000 up to m/z 12,000 (n=3 12), thereby supporting the presence of a higher molecular weight fraction. In the case of P4b at least two ion series spanning from m/z 3,000 to 20,000 (n=319) with ions separated by the mass of the polymer repeat unit (~1,011 amu) were identifie d. Similarly to P1 and P4a the dominant series is consistent with oligomers endcapped by chlorine atoms. The unsaturated linker -containing DA polymers P2 P3 P5 and P6 were investigated employing DCTB as a matrix. Both the structures of ethynylene -linke d P2 and P5 were

PAGE 158

158 supported by series of ions (n=1 5 and n=1 2) with spacing matching the mass of the repeat unit of the polymers, ~835 amu and ~1,036 amu respectively. The residual masses of these series indicate that the endgroups consist of a hydrogen and an additional monomer unit M1 (in the case of P2 ) or M2 (in the case of P5 ) (s ee Figure 5 2 ) which can be end -capped with a phenylphosphine derivatized moiety as reported in several other studies.156 In addition both of these series have related, lower -intensity series in which a bromine is substitut ed for the hydrogen end group. Figure 5 3. MALDI MS of DA -copolymers a) P1 b) P4a c) P3 and d) P6 DCTB was used as the matrix in each case (Adapted with p ermission from Ref.128 Copyright 2009 American Chemical Society)

PAGE 159

159 A large number of ion series were detected in the range m/z 1,000 9,000 for both P3 and P6 as illustrated in Figure 5 3 c and 5 3 d. Each series exhibited a spacing of ~837 amu for P3 and ~1 ,038 amu in the case of P6 matching the masses of the expected repeat units. Residual masses of some of the ion series suggest endgroups of hydrogen and phenyl or phenyl and styryl, an experimental observation supported by work from Janssen et al. for inst ance.157 Other ion series appear to have endgroups of hydrogen and phenylphosphine analogue moieties or hydrogen and an extra aromatic building block M1 (in the case of P3 ) or M2 (in the case of P6 ) (See Figure 5 2 ). It is worth not ing that ion series detected in MALDI -MS analyses of polymers commonly occur at much lower masses than those estimated employing conventional gel permeation chromatography (GPC). In fact, discrimination of the higher mass ions take place both during ioniza tion and detection and the phenomenon is amplified with increasing polymer polydispersity.158160 Molecular weights were fur ther investigated via polystyrene -calibrated GPC employing tetrahydrofuran (THF) as the mobile phase. As illustrated in Table 5 1, polymers with weight average molecular weight (Mw) ranging from 11,000 to 60,000 g mol1 with relatively narrow PDIs (1.7 2.6 ) and number average molecular weight (Mn) values up to 28,000 g mol1 for the oxidatively polymerized analogues were estimated. Importantly, all polymers had a minimum average number of repeat units of 7 corresponding to a backbone of about 28 unsaturated building units (including heterocycles and unsaturated linkages) which is nearly twice as much as the value at which the electronic properties of conjugated polymers commonly saturate (~15 aromatic rings).112 The hig hest average numbers of unsaturated building units incorporated into a

PAGE 160

160 single polymer chain (~45, 60 and 81) were obtained upon using oxidative polymerization conditions (in synthesizing P4b P1 and P4a respectively). Thermogravimetric analysis in a nitro gen atmosphere of the all -heterocyclic analogues P1 P4a and P4b revealed only negligible weight loss below 320 oC, demonstrating their high thermal stability. On the other hand, the polymers containing unsaturated linkages exhibited only limited thermal s tability in general with clear signs of degradation starting at 200 oC in the case of P5 and P6 for instance. Table 5 1. Weight average molecular weight ( Mw, g mol1), polydispersity index (PDI), average number of repeat units, average number of rings, and Onset of decomposition temperature (as measured by TGA under nitrogen atmosphere) for the donor acceptor copolymers P1 -P6 Polymer Mw (g mol1) PDI Avg. No. of Repeat Units Avg. No. of Rings Td (oC)a P1 42,700 2.6 20 60 321 P2 11,800 1.8 7 21 213 P3 13,000 1.7 9 27 265 P4a 59,000 2.1 27 81 350 P4b 32,100 2.2 15 45 348 P5 16,800 1.9 8 24 201 P6 15,500 1.8 9 27 200 a Onset decomposition temperature measured by TGA under nitrogen. 5.2 .2 Polymer Optical Characterizatio n All polymers exhibit a two -band absorption in the UV -visible spectral range with a peak to -peak distance larger than 190 nm opening a gap of low optical density in the blue, blue -green or green region. While the lower energy transition can be attributed to the intramolecular donor acceptor interaction arising from the bonding of electron rich (DOTs and unsaturated linkages) and electron deficient (BTDs) building units, the higher energy transition can be assigned to the electron -rich portion of the alter nating polymer backbone (see Chapter 4 for a more complete discussion on how two band absorption spectra may arise in DA polymers) .143 As can be see n

PAGE 161

161 from the higher energy absorption maxima of P3 and P6 relative to P1 and P4a in Figure 5 4 the insertion of an ethylenic spacer between the two electron-donating substituents of these donor acceptor systems induces a bathochromic shift which also moves the absorption gap from the blue or blue -green to the blue -green or green regions of the visible, hence predicting some significant differences in neutral state colors To a lesser extent, the high wavelength optical transitions also shift towards longer w avelengths although the onsets of absorption remain nearly identical here in solution and in spite of the spacer insertion. It is expected to see a further bathochromic shift as we examine thin -films where the degree of coplanarity of the polymer main -chai ns influences the optical bandgap. We attribute the shoulders tailing between 750 and 800 nm in the case of P4a and P6 to the aggregation of the highest molecular weight fraction of the polymers in solution (although no visible aggregation could be perceived) as elevating the temperature of the polymer/toluene solutions had the effect of reducing their intensity. The polymers maxima of optical absorption are reported in Table 5 2. Figure 5 4 Normalized solution optical absorption (in Toluene) for the t rans-ethylene unsaturated DA -copolymers P3 and P6 compared to their ethylhexyl -substituted parent copolymers P1 and P4a (respectively) (Adapted with permission from Ref.128 Copyright 2009 American Chemical Society)

PAGE 162

162 Table 5 2. Local absorption maxima (so lution in toluene and solid state) and film optical bandgap ( Eg, eV) for the donor acceptor copolymers P1 -P6 (Adapted with permission from Ref.128 Copyright 2009 American Chemical Society) Polymer abs (nm) In Toluene abs (nm) Thin Film E g (eV) 1 2 1 2 P1 391 638 399 653 (708) a 1.6 P2 362 553 373 577 1.75 P3 401 625 416 648 (706) a 1.6 P4a 402 674 408 677 (722)a 1.54 P4b 405 678 407 685 1.52 P5 346 578 357 578 1.65 P6 419 677 433 685 (745) a 1.5 a Shoulder 5.2 .3 Polymer Redox Properties ( Experimental w ork was achieved and results kindly supplied by Dr Svetlana Vasilyeva) The redox properties of all polymers were investigated via cyclic voltammetry (CV) and differential pulse voltammetry (D PV) with the goal of determining how polymer oxidation/reduction potentials, associated HOMO/LUMO energy levels, and relative electrochemical bandgaps vary as a function of the structural modifications. In Table 5 1 3, the corresponding polymer energy level s (HOMO and LUMO) are given relative to the vacuum level, considering that the SCE is 4.7 eV vs vacuum and Fc/Fc+ is 0.38 eV vs SCE, i.e. ~5.1 eV relative to vacuum. In comparison to CV, DPV offers a higher sensitivity and sharper redox onsets throughout the electrochemical process owing to a reduced contribution of the charging background currents which, in turn, enhances the accuracy and reliability of the bandgaps estimated electrochemically. Thin films of the polymers were drop -cast onto platinum disk electrodes from toluene solutions (6 mg mL1) for characterization in an argon -filled dry box. A platinum flag was used as the counter electrode in combination with a Ag/Ag+ reference

PAGE 163

163 electrode and all results were subsequently calibrated to Fc/Fc+ (for c onsistency with the rest of this Chapter ). Prior to characterization, the polymer films were cycled at a scan rate of 50 mV s1 between 0.2 and 0.9 V until they reached a stable and reproducible electrochemical response.3 The polymer films were found sta ble to p type doping during extensive cycling from 0.2 to 0.9 V retaining 90% of their peak currents after 100 switches with the exception of polymers P2 and P5 We attribute the more rapid loss of electroactivity of P2 and P5 to the sensitivity of the et hynylene spacers to overoxidation side -reactions. In general, the polymers were found to be less stable to n -doping processes. In 0.1M LiBF4/propylene carbonate (PC) and 0.1M TBAP/PC electrolyte solutions, both CV and DPV experiments show similar and repro ducible redox behavior for each polymer upon oxidation and subsequent neutralization. However, we found that upon reduction of the polymer films the current responses could only be observed when using TBAP supporting electrolyte. It is worth noting that su ch specificity to the positively -charged balancing counter ion has already been reported in the case of various conjugated polymers.161,162 In fact, during the reduction process, hard electrophilic cations such as lithium induce only minimal conductivity due to pinning of the doped state and insufficient delocalization of the charge carrier along the polymer main -chain.163,164 For consistency, the electrochemical results summarized in Table 5 3 refer to measurements collected in 0.1M TBAP/PC solution electrolyte. Careful examination of the data present ed in this table highlights the correlations between polymer architecture and electrochemical properties.

PAGE 164

164 For example, the CVs of polymer P4b (possessing linear octyl solubilizing side -chains) and its branched analogue P4a (possessing 2 -ethylhexyl substitu tents instead) illustrated in Figure 5 5 a show a more reversible redox process with potential peak difference ( Ep) of 150 mV in the case of P4a while P4b has a significantly broader current response of ca. 260 mV. The increase of interchain spacing result ing from the branched nature of the alkyl chains of P4a could promote the insertion of charge compensatingion and lead to a faster redox response (this hypothesis will also be backed up throughout the next sections of this Chapter ). However, P4b shows a l ower electrochemical bandgap as compared to P4a probably due to its more extended conjugation in the solid state when compared to that of the branched derivative which is also likely more disordered.110 Figure 5 5. a) Cyclic voltammograms of P4a P4b and P6 drop -cast onto platinum disk electrodes (0.02 cm2) in 0.1M TBAP/PC electrolyte solution, using a scan rate of 50 mV/s, b) Differential pulse vol tammetry of P1 P4a and P6 drop -cast onto platinum disk electrodes (0.02 cm2) in 0.1M TBAP/PC electrolyte solution using a step time of 0.1 s, a step size of 2 mV, and an amplitude of 100 mV (Adapted with permission from Ref.128 Copyright 2009 American Chemical Society) When comparing the dialkoxythiophene (DalkOT) derivatized polymers P1 P2 and P3 with their propylenedioxythiophene (ProDOT) analogues P4a P5 and P6 respectively, a clear trend is observed in terms of oxidation potential. Upon replacing the DalkOT units with the more

PAGE 165

165 electron -rich ProDOTs, lower onsets of oxidation are obtained which result in higher HOMO level estimated values and lower electrochemical bandgaps. Figure 5 5 b shows the DPVs of P1 (a DalkOT based derivative) and P4a (ProDOT -based) using a step time of 0.1 s, a step size of 2 mV, and an amplitude of 100 mV. P4a has an oxidation onset 350 mV less positive and a reduction onset only 120 mV more negative, than P1 Such a drastic increase of the HOMO energy in the case of P4a lea ds to a lower magnitude of the DPV estimated band gap (1.54 eV compared to 1.77 eV for P1 ). P2 and P5 contain ethynylene linkages which reduce the degree of conjugation of the subsequent donor acceptor systems. This effect appears to be the most significan t in the case of the less electron rich DalkOT derivatized polymer P2 Thus after only several cycles, P2 loses its electroactivity as CV and DPV current responses become indiscernible from background currents, though a visible thin polymer film (now insul ating) is still present at the electrode surface. The DPV of P6 is also shown in Figure 5 5 b in order to be compared to that of P4a The observed differences that relate to the electrochemical properties of these polymers, such as a decreased amplitude of the bandgap and a more reversible redox switching in the case of P6 ( Ep (P6 ) = 70 mV and Ep (P4a ) = 150 mV) can be rationalized by considering the enhanced planarity and extended conjugation length of the P6 backbone in which a trans -ethylene linkage was inserted between the donor acceptor building blocks. P1 and P3 exhib it similar variations in their redox properties, though the differences appear less pronounced. In general, the electrochemically determined bandgaps were found to be slightly larger than the optically estimated values, but remain in good agreement.

PAGE 166

166 Table 5 3. Electrochemically determined HOMO and LUMO energy levels (by CV and by DPV), Corresponding electrochemical bandgaps, and Comparison with their optically estimated values for the copolymers P1 P6 (Adapted with permission from Ref.128 Copyright 2009 American Chemical Society) Polymer Eoxonset (V) HOMO (eV) Eredonset (V) LUMO (eV) Egap (V) Egap (V) CV DPV CV DPV CV DPV CV DPV CV DPV Optical P1 0.25 0.25 5.35 5.35 1.57 1.52 3.53 3.59 1.82 1.77 1.6 P2 0.51 0.4 5.61 5.5 1.33 1.38 3.77 3.72 1.84 1.78 1.75 P3 0.12 0.14 5.22 5.24 1.46 1.51 3.54 3.59 1.58 1.65 1.6 P4a 0.06 0.1 5.04 5.0 1.7 1.64 3.4 3.46 1.64 1.54 1.54 P4b 0.0 0.02 5.1 5.12 1.6 1.57 3.48 3.49 1.6 1.59 1.52 P5 0.21 0.24 5.31 5.34 1.48 1.47 3.62 3.63 1.69 1.71 1.65 P6 0.07 0.0 5.17 5.1 1.56 1.5 3.54 3.6 1.63 1.5 1.5 a oxidation ( Eoxonset ) and reduction ( Eredonset ) potentials are reported vs Fc/Fc+. HOMO and LUMO energy levels are derived from the electrochemical data ( Eoxonset and Eredonset respectively) considering that the SCE is 4.7 eV vs. vacuum and Fc/Fc+ is 0.38 eV vs. SCE, i.e. 5.1 eV relative to vacuum. 5.2.4 Polymer Spectroelectrochemical Characterization Figure 5 6 uses spectroelectrochemistry to illustrate the propensity of each of the DA polymers to form positively-charged carriers upon electrochem ical oxidation (p type doping). All DA -polymers were spray -cast from toluene (4 6 mg mL1) onto indium tin oxide (ITO) coated glass slides at room temperature. Electrochemical o xidation of the films was carried out in 0.1 M LiBF4/propylene carbonate supporting electrolyte using a silver wire as a quasi reference electrode (calibrated against Fc/Fc+) and a platinum wire as the counter electrode. Figures 5 6 a and 5 6 b show the spec tral changes occurring in the 300 1600 nm range upon electrochemical oxidation of P1 and P4a which both exhibit a two -band absorption in the visible spectrum with a major contribution from the lower energy transition attributed to the intramolecular donor acceptor electronic interaction. With an open gap in light absorption in the range 455 470 nm, coupled to a moderate absorption between 350 and 450 nm and a broad coverage of the red portion of the visible, these polymers exhibit different hues of blue While P1 shows a deep cyan blue neutral state, P4a is blue green due to the more red -shifted low energy

PAGE 167

167 transition allowing a minor green transmission. The spectral differences observed in their neutral state, along with the estimated variation of 0.06 e V in optical bandgaps (calculated from the onset of their longer wavelength absorption band) relate to the difference in molecular orbital energy levels with ProDOT bringing a higher energy HOMO to the donor acceptor conjugated system. Oxidation of the films induces simultaneous and extensive bleaching of both the short and long wavelength absorption bands. Taking the two local absorption maxima as references, P1 shows a transmittance change ( T ) of 41.7% ( abs1) and 50.3% ( abs2) compared to 49.6% (abs1) and 56.1% ( abs2) for P4a indicative of the superior electrochromic contrast observed in the case where DalkOT s are substituted with ProDOTs. When fully oxidized, bipolaronic states peaking beyond 1600 nm in the infrared with almost no visible absorption govern their optical spectra and as a result, both polymers show a remarkably high level of transparency to the human eye. Figures 5 6 c and 5 6 d show the spectroelectrochemical behavior of p olymers P2 and P5 the ethynylene -linked derivatives of P1 and P4a P2 and P5 exhibit similar solid -state absorption spectra involving a relatively intense absorption in the UV region and a broad band spanning 450 700 nm due to the dioxythienyl BTD DA ch romophore. Their open gap at about 440 and 450 nm (blue transmission), along with the red transmission between 700 and 800 nm (note especially for P2 ) gives a deep purple hue to these polymers. The overall hypsochromic shift observed when compared to P1 an d P4a is a consequence of the insertion of sptype unsaturated spacers along the DA backbone. However, more significant spectral changes relative to P1 and P4a are evident during oxidation. Examining the depletion of the two absorption maxima, P2 and P5 sh ow smaller transmittance changes of 14.3% ( abs1), 23.3% ( abs2) and 3.8% ( abs1), 11.8% ( abs2) respectively, indicating a limited electrochromic performance. In the case of P2

PAGE 168

168 the transition arising in the near IR (~850 nm) supports the formation of rad ical cation states which do not ultimately convert into dicationic species and the long wavelength absorbance beyond 1400 nm is low. Upon further increase of applied potential, P2 loses its reversible redox activity and the film breaks down. Similarly, P5 shows a broadly defined pattern in the near IR involving a polaronic transition and does not result in a well defined or intense generation of bipolarons. In both cases, the lack of stable bipolaronic states forming on electrochemical oxidation prevents ef fective bleaching of the UV -visible absorption bands and hinders access to a transmissive doped state. As such, the ethynylene -containing P2 and P5 do not address the requirements of a useful system for electrochromic device applications. Figures 5 6 e and 5 6 f illustrate the effect of trans -ethylene spacers incorporated between the DAD macromonomers on the electrochromicity of the subsequent polymers P3 and P6 As for P1 and P4a P3 and P6 feature a two -band absorption in the visible with a main contributi on from the donor acceptor transition. However, in these instances the two transitions have broadened and merged together and now possess a more significant overlap, thus predicting two neutral state polymers of darker tone. While P3 features a dark blue-green hue in the neutral state, the gradual application of potential first induces the growth of charge -carriers in the near IR with simultaneous depletion of the visible absorption. Considering the two local absorption maxima, P3 shows substantial transm ittance changes of 24.3% ( abs1) and 50.3% ( abs2) leading to a clear -grey state upon oxidation. Nonetheless, the redox process is hampered by an irreversible degradation of the active layer as illustrated in Figure 5 6 e (see the final blue curve). From th is experimental result, a small potential window of reversible operation is anticipated for this polymer which is undesirable for EC materials (such that further work with this polymer will not be pursued).

PAGE 169

169 As a consequence of the significant bathochromic shift (about 25 nm) and broadening undergone by its short wavelength absorption band, the open gap of polymer P6 exhibits a local minimum of absorption shifted from 36 nm compared to its parent P4a Likewise, its long wavelength absorption band shifted to wards the lower energy region and covers a more extensive bandwidth of the optical spectrum. As a result, P6 exhibits a saturated green color which is particularly useful in the construction of EC displays. The optically determined energygap of P6 (1.50 e V) is slightly narrower than that for P4a (1.54 eV). The spectral differences observed between P6 and P4a in their neutral state along with the estimated variation of 0.04 eV in optical bandgaps can be rationalized considering an increase in planarity of t he polymer backbone triggered by the introduction of the ethylenic conjugated spacers which minimize the steric interactions between the propylenedioxy bridges of the electronrich ProDOT units. Upon electrochemical oxidation, P6 shows excellent transmitta nce changes of 29.7% ( abs1) and 37.9% (abs2) and switches to a light grey transmissive doped state. While the transmittance ratios observed are notably lower than the ones measured for P4a (49.6% ( abs1), 56.1% ( abs2)), the relatively uniform bleaching observed across the vi sible region, along with the formation of stable polaronic and bipolaronic states in the near IR, projects the potential of this green cathodically -coloring polymer for EC applications. Finally, Figure 5 6 g gives insight into the extent to which the nature of the solubilizing side -chains can affect the spectroelectrochemical performance of otherwise identical polymers. In this study, P4b the linear alkyl -substituted DA analogue of P4a shows a transmittance change of 40.3% in the blue region ( abs1) that is lower by about 9% in comparison with P4a The transmittance change for the long -wavelength absorbance remains high (48.9% ( abs2) versus 56.1%) in correlation with the near IR charge -carrier transitions. This variation in contrast can be

PAGE 170

170 ten tatively rationalized considering an increase of interchain spacing in the case of the branched ethylhexyl -substituted polymer P4a as respect to the linearly octyl-substituted polymer P4b that allows improved ion diffusion and dopant insertion. Overall, P4 a appears more transmissive than P4b in its fully doped state. Figure 5 6. Spectroelectrochemistry of DA -copolymer a) P1 b) P4a c) P2 d) P5 e) P3, f) P6 and g) P4b (Spectra a), b), e), f) and g) are normalized at the polymer longer wavelength a bsorption maximum). Films were spray -cast onto ITO -coated glass from toluene solutions (~4 mg mL1). Electrochemical oxidation of the films was carried out in 0.1 M LiBF4/PC supporting electrolyte using a silver wire as a quasi reference electrode (calibra ted against Fc/Fc+) and a platinum wire as the counter electrode. The applied potential was increased a) in 25 mV steps from +0.3 V to +0.5 V and in 10 mV steps from +0.5 V to +0.76 V b) in 25 mV steps from 0.05 V to +0.7 V c) in 50 mV steps from 0.17 V to +0.23 V and in 25 mV steps from +0.23 V to +0.8 V d) in 25 mV steps from +0.23 V to +0.7 V.e) in 25 mV steps from +0.33 V to +0.73 V f) in 25 mV steps from +0.2 V to +0.83 V g) in 25 mV steps from +0.17 V to +0.72 V (Adapted with permission from Ref.128 Copyright 2009 American Chemical Society)

PAGE 171

171 5. 2 .5 Polymer Colorimetric Analysis Films of polymer P1 P4a P3 and P6 were spray -cast onto ITO coated glass, and the color changes occurring on redox switching were cha racterized based on the Commission Internationale de l'Eclairage 1976 L*a*b* color standards. Each film was redox -cycled several times and, as shown in Figure 5 7 the colorimetrically determined relative luminance change, estimating the brightness of th e transmitted light as a percentage of the brightness of the light source calibrated to the sensitivity of the human eye, was measured as a function of the doping level induced by electrochemical oxidation (the corresponding absorbance of the films at abs 2 is reported in the legends). Generated using the L*a*b* color coordinates estimated under constant illumination of the polymer thin -films, the color swatches shown in Figure 5 8 give insight into how the structural modifications induced in the DA backbon es influence their color states on p -

PAGE 172

172 doping. For consistency in comparing the various polymers, the film depositions were here followed by coarse luminance measurements allowing to approximately match the polymer optical transmission over the visible. Figure 5 7. Relative luminance as a function of applied potential and film thickness for spray coated DA polymers a) P1 b) P4a, c) P3 and P6 d) P4a and P4b The legends indicate the absorbance of the deposited films (estimated at the longer wavelength absorption maximum) and give an indication of the thickness obtained on spraying. For color matching, L*a*b* values (in the sense of the Commission Internationale de l'Eclairage 1976 L*a*b* color model) of fully neutral and oxidized states are reported fo r the films (Adapted with permission from Ref.128 Copyright 2009 American Chemical Society) Fi gure 5 7 a and 5 7 b compare films of about the same neutral state luminance values of polymer P1 and P4a respectively. Interestingly, while the monochromatic tr ansmittance ratios

PAGE 173

173 estimated by spectroelectrochemical analysis at the local absorption maxima of P1 and P4a predicted slightly higher contrasts in the case of P4a here the colorimetric measures lie in favor of P1 which, based on the perception of the bro ad visible region, exhibits luminance -change values as high as ca. 50% and L* values (ranging from 0 to 100) up to 94, indicative of its aptitude to nearly reach the white point of color space. Nonetheless, the contrast ratios and transmissivities of P4a upon full oxidation remain outsta nding as illustrated in Figure 5 7 b. In particular, the luminance change of ca. 45% observed for the thicker deposited film of polymer P4a (A( abs2) = 2.4 a.u.) demonstrates exceptional switching characteristics, suggestin g good diffusion of the elect rolyte through the active layer. Overall, both P1 and P4a attain remarkable transparencies with almost no residual hue in their fully oxidized state as indicated by the estimated a* and b* values close to those of the clear sta te of coordinates a*=0 and b*=0 (See Figure 5 8 a and 5 8 b). The large b* values observed in the neutral state of P1 account for its deep blue saturation whereas the stronger a* values obtained for P4a justify the green tone of its blue -green neutral state. These two polymers complete their full switch in a potential window of less than 0.4 V. Figure 5 7 c evaluates how the incorporation of a trans -ethylene spacer affects the colorimetric properties of the fully heterocyclic control polymers P1 and P4a P3 an d P6 were spray -cast at thicknesses representative of their optimal colorimetric performance. Accordingly, a luminance change of 26% was observed along with little residual hue in the oxidized state as attested by the low a* and b* values. Although this lu minance change is lower than those observed for the control polymers, such contrasts remain adequate for electrochromic polymers with application in window devices.

PAGE 174

174 A direct effect of the presence of unsaturated spacers regularly inserted in the polymer b ackbones is the lower b* values of neutral state P3 (b*= 11) and P6 (b*= 1) in comparison with P1 (b*= 41) and P4a (b*= 31) which translates to a progressive loss of blue in their tone with P6 exhibiting a near -perfect green saturation (See Figure 5 8 c). S imilar to P1 and P4a these two polymers complete their full switch in a potential window of less than 0.4 V. Figure 5 8 d suggests a slightly sharper color change when branched alkyl solubilizing groups replace the linear ones. At the deposited film thickn ess (A( abs2) of ca. 0.6 a.u.), the colorimetrically estimated contrasts are near -identical (about 28%), as are the color coordinates of both the neutral and the oxidized states (See Figure 5 8 d). Only a modest improvement in transparency can be observed for P4a as predicted by the spectroelectrochemical study. The ease in solution processing of P4a makes this polymer electrochrome stand out when compared to P4b and adds advantageously to the moderate optical enhancement observed here. Figure 5 8 Color matching corresponding to the plots of relative luminance as a function of applied potential and film thickness described in Figure 5 7 for spray -coated DA polymers a) P1 b) P4a c) P3 and P6 d) P4a and P4b The associated L*a*b* values (in the sense of the Comm ission Internationale de l'Eclairage 1976 L*a*b* color model) of fully neutral and oxidized states are reported for the films (Adapted with permission from Ref.128 Copyright 2009 American Chemical Society)

PAGE 175

175 5.2.6 Polymer Switching Stud ies New films of po lymer P1 P4a P3 and P6 were carefully spray -coated onto ITO and their deposition simultaneously monitored by transmittance measurements at their long wavelength absorption maximum ( abs2). Taking the optical density as representative of the thickness of active layer deposited, polymer thin-films of near identical consistencies were obtained (% T (abs2) = 17 20%), allowing reliable comparison of their electrochromic performance. Each film was redox -cycled until a stable and reproducible switch was reached prior to data collection and comparison. Figure 5 9 shows the transmittance change (or EC contrast, T ) attained as a function of time and at the longer wavelength absorption maximum of each material applying square -wave potential steps in the range 10 s t o 1 s ( ). In general, square -wave potential stepping experiments are preferred over potential cycling when characterizing the response time of a thin film electrochrome as significant optical change occurs in less than 0.1 s. In this study, the lower ener gy transition was chosen to be monitored due to its more intense absorption in the visible region, and it is, in turn, expected that its contrast ratio and rate of depletion will be representative of the polymer s overall EC switching ability. Figures 5 9 a and 5 9 b compare the response of polymer P1 and P4a respectively. As predicted by the spectroelectrochemical study, P4a exhibits a higher EC contrast (50.6%) at the longer wavelength absorption maximum than that of P1 (41.8%) on slow electrochemical switc hing ( = 10 s). As the switch time is decreased from 10 s to 2 s, a difference of 29% in the contrast estimated is noted in the case of P1 whereas P4a undergoes a contrast change of less than 4.5%. It is only by changing the switch time from 10 s to 1 s that P4a showed a significant loss of contrast, 19% lower in value. Given the similar oxidation potentials, such a notable

PAGE 176

176 variation in switching performance between the two polymers can be attributed to the differences in ease of attaining the planar quin oidal geometry characteristic of the oxidized state. With its solubilizing side -chains in the plane of the conjugated main -chain generally inducing a relative twisting of the polymer backbone, it is likely that P1 does not possess the same extent of conjug ation than P4a and is, in turn, not as prone to radical cation (polaron) coupling into stable dications (bipolarons). On the contrary, the ProDOT units of P4a exhibit a spiro like center which maintains the branched solubilizing pendant groups out of the plane of the main -chain and increases the interchain spacing allowing ions to penetrate into and out of the film more freely. Figures 5 9 c and 5 9 d report the response of the two trans -ethylene substituted analogues of P1 and P4a respectively P3 and P6 to square -wave signals of 10 s, 5 s and 2 s. The EC contrasts of ca. 34% obtained at the longer wavelength absorption maximum on slow switching ( = 10 s) are in agreement with the spectroelectrochemical estimated values ranging from 35 to 50%. Similar to the parent polymers ( P1 and P4a ), as the switch time is increased from 10 s to 2 s, a decrease of about 42% in the contrasts is noted in the case of P3 whereas P6 undergoes a 2% contrast change from 10 s to 5 s and 24% from 5 s to 2 s. Overall, in analogy with its control polymer P4a P6 exhibits higher EC contrasts than P3 as well as shorter response times. It is worth noting that P3 was also more sensitive to high oxidation potentials than P6 and lost its redox reversibility rather rapidly. The difference in switching speed between P6 and the fully heterocyclic parent P4a is tentatively attributed to the decrease of interchain spacing arising from an increase of planarity of the polymer backbone upon insertion of ethylene substituents between the electron -donating building units. It is possible that the reduced interchain spacing hinders the diffusion of the counter -balancing ions from the electrolyte to the active layer. An additional

PAGE 177

177 reason could lie in the rotational freedom of the double bond which woul d inhibit the combination of the radical cationic species. Figure 5 9 Square -wave potential step absorptometry of spray -coated a) P1 (monitored at 653 nm, 0.6 V +), b) P4a (monitored at 677 nm, 0.6 V 0.6 V versus Fc/ Fc+), c) P3 (monitored at 648 nm, 0.6 V +) and d) P6 (monitored at 685 nm, 0.6 V +), onto ITO in 0.1 M LiBF4/propylene carbonate solution. Switch times chosen among: 10 s step for 40 s (2 cycles), 5 s step f or 20 s (2 cycles), 2 s step for 20 s (5 cycles) and/or 1s step for 20 s (10 cycles) (Adapted with permission from Ref.128 Copyright 2009 American Chemical Society) Of all the candidates for EC applications synthesized in this study, P1 P4a and P6 stan d out in terms of color contrasts, switching -times and short term redox reversibility as

PAGE 178

178 demonstrated throughout the previous sections. To reinforce the potential of these ECPs for integration into ECDs and commercial applications, their long -term switchin g was investigated on ITO -coated glass applying square -wave potential steps of a duration adapted to the polymers individual response time. Figure 5 10 demonstrates the performance of the selected polymers as the spray -deposited thin-films are subjected to 3,000 redox cycles in 0.1 M LiBF4/propylene carbonate solution and under atmospheric conditions Measured at its longer wavelength absorption maximum ( abs2= 653 nm) and using a switching step of 6s, a transmittance change similar to that previously monit ored ( T (abs2) ~ 40) extends with consistency over the first 2,000 redox cycles (6h 40min) for P1 as described in Figure 5 10a. Between 2,000 and 2,500 cycles, a progressive loss of transparency is observed on doping whereas the film in its neutral state sees only little change in coloration. Such a sudden loss of electrochromicity is a common indication that the contact between the ITO -coated working electrode and the polymer thin-film is altered. In fact, a major difficulty in the synthesis of solution -processable ECPs lies in the tradeoff that must be found in the choice of the solubilizing groups considering that the subsequent electroactive polymer thin-film should be sufficiently insoluble and compact to remain strongly adhered to the working elect rode while being able to swell effectively and allow rapid diffusion of the charge balancing counter ions throughout the polymeric network. In the case of P1 the combination of the branched nature of the substituents with their in plane configuration induces an excess of solubility in the context of long -term EC switching. Nonetheless, in devices where gel electrolytes are used, the extent of delamination or dissolution is generally reduced in comparison with a conventional solution electrochemical cell su ch that it is not always necessary to re -synthesize a polymer analogue with shorter solubilizing side -chains.

PAGE 179

179 In contrast, P4a exhibits a remarkably stable switch over 3,000 cycles (10h based on a 4s switching step, one cycle is 8s) as demonstrated in Figu re 5 10b. Measured at its longer wavelength absorption maximum ( abs2= 677 nm), a variation in electrochromic contrast of only 1% is observed along with no significant lose of transparency in the doped state or color in the neutral state. This outstanding enhancement when compared to P1 can be attributed to the reduced solubility of P4a in solid thin -films. The out of -plane configuration of the solubilizing side -chains of P4a accounts for an increase of planarity of the conjugated backbone which in turn affects both inter and intra -chain interactions. This last conside ration becomes especially obvious as the somewhat reduced solubility of P4a necessitated the use of elevated temperatures to dissolve the polymer rapidly in the casting solvent (toluene) while P1 dissolves instantaneously at room temperature using the same solvent. The same long -term stability study (using a 6s switching step, one cycle is 12s) performed on the ethylene unsaturated polymer P6 revealed a more conventional behavior for an ECP. Thus, as depicted in Figure 5 10c, a slow but constant variation o f transparency in the oxidized state ( T (DS)= 7.9 after 3,000 cycles) and color in the neutral state ( T (NS)= 9.1 after 3,000 cycles) was monitored with increasing number of cycles, whereas the overall transmittance ratio (% T (DS) -% T (NS)) remains relatively constant during the experiment ( T = 1.2 after 3,000 cycles). As stated previously, this result remains significant considering the nature of the substrate used and the redox induced solubility of the polymer in the solution electrochemical cell as there is little to prevent the chains from slowly migrating towards the platinum counter electrode during the experiment. Nonetheless, this phenomenon could be overcome (at least greatly minimized) by casting the higher molecular weight fraction of P6 hence improving both film quality and resista nce of the polymer network to repeated switching.

PAGE 180

180 Figure 5 10. Long -term switching of spray -cast a) P1 (monitored at 653 nm, 0.4 V versus Fc/Fc+, square -wave potential steps of 4s, complete cycle is 8s), b) P4a (monitored at 677 nm, 0. 4 V +, square -wave potential steps of 6s, complete cycle is 12s), and c) P6 (monitored at 685 nm, 0.4 V versus Fc/Fc+, square -wave potential steps of 6s, full cycle is 12s), on ITO in 0.1 M LiBF4/propylene carbonate solut ion (Adapted with permission from Ref.128 Copyright 2009 American Chemical Society) 5.2.7 Synthetic Details Synthesis of Macromonomers M1, M1, M2, M2, P olymer s P1, P4a and P4b: Refer to Chapter 4 of this dissertation. P4b was synthesized using conditions analogous to those described for P4a 1HNMR ( 300 MHz, CDCl3) = 8.34 (bs, 2H), 4.29 (bs, 8H), 3.67 (bs, 8H), 3.47 (bs, 8H), 1.27 (bs, 36H), 0.86 (bs, 24H). P olymer P2 : Macromonomer M1 (245 mg, 0.252 mmol), 1,2 bis(tri methylstannyl)ethyne (88.7 mg, 0.252 mmol) and tetrakis(triphenylphosphine)palladium (4

PAGE 181

181 mol%) were cycled (argon/vacuum, 3x) and subsequently dissolved in a mixture of dry toluene (8 mL) and deoxygenated dimethylformamide (1.5 mL). The solution was heated to 100oC. After 72 hours, the reaction mixture was cooled to room temperature and poured into methanol deionized water (2:1, 200 mL). The precipitate was filtered through a Soxhlet thimble, purified via Soxhlet extraction for 12 h with methanol, and extrac ted for 2 hours with chloroform. The chloroform fraction was concentrated by evaporation and the polymer was precipitated into deionized water, collected as a dark solid and dried under vacuum overnight (81 %). 1H NMR (300 MHz, CDCl3) = 8.47 (bs, 2H), 7.7 2 (m, chain ends), 7.41 (m, chain-ends), 4.1 3.9 (br, 8H), 1.8 1.2 (br, 36H), 1.1 0.6 (br, 24H) Polymer P5: Macromonomer M2 (245 mg, 0.209 mmol), 1,2 bis(trimethylstannyl)ethyne (73.5 mg, 0.209 mmol) and tetrakis(triphenylphosphine)palladium (4 mol%) wer e cycled (argon/vacuum, 3x) and subsequently dissolved in a mixture of dry toluene (8 mL) and degassed dimethylformamide (1.5 mL). The solution was heated to 100oC. After 72 hours, the reaction mixture was cooled to room temperature and poured into methano l -deionized water (2:1, 200 mL). The precipitate was filtered through a Soxhlet thimble, purified via Soxhlet extraction for 12 h with methanol, and extracted for 2 hours with chloroform. The chloroform fraction was concentrated by evaporation and the polymer was precipitated into deionized water, collected as a dark solid and dried under vacuum overnight (72 %). 1HNMR ( 300 MHz, CDCl3) = 8.29 (bs, 2H), 7.72 (m, chain-ends), 7.41 (m, chain-ends), 4.20 (bs, 8H), 3.58 (bs, 8H), 3.32 (bs, 8H), 1.28 (bs, 36H), 0.89 (bs, 24H). P olymer P3 : Macromonomer M1 (315 mg, 0.324 mmol), 1,2 -bis(tributylstannyl)ethene (196.6 mg, 0.324 mmol) and tetrakis(triphenylphosphine)palladium (4 mol%) were cycled (argon/vacuum, 3x) and subsequently dissolved in a mixture of dry toluene (15 mL) and

PAGE 182

182 degassed dimethylformamide (3 mL). The solution was heated to 100oC. After 72 hours, the reaction mixture was cooled to room temperature and poured into methanol -deionized water (2:1, 200 mL). The precipitate was filtered through a Soxhlet th imble, purified via Soxhlet extraction for 12 h with methanol, and extracted for 2 hours with chloroform. The chloroform fraction was concentrated by evaporation and the polymer was precipitated into methanol deionized water (1:1, 100 mL), collected as a d ark solid and dried under vacuum overnight (70 %). 1H NMR (300 MHz, CDCl3) = 8.41 (bs, 2H), 7.72 (m, chain-ends), 7.41 (m, chain-ends), 4.1 3.9 (br, 8H), 1.8 1.2 (br, 36H), 1.1 0.6 (br, 24H) P olymer P6 : Macromonomer M2 (300 mg, 0.256 mmol), 1,2 -bis(trib utylstannyl)ethene (155.2 mg, 0.256 mmol) and tetrakis(triphenylphosphine)palladium (4 mol%) were cycled (argon/vacuum, 3x) and subsequently dissolved in a mixture of dry toluene (15 mL) and degassed dimethylformamide (3 mL). The solution was heated to 100oC. After 72 hours, the reaction mixture was cooled to room temperature and poured into methanol -deionized water (2:1, 200 mL). The precipitate was filtered through a Soxhlet thimble, purified via Soxhlet extraction for 12 h with methanol, and extracted for 2 hours with chloroform. The chloroform fraction was concentrated by evaporation and the polymer was precipitated into methanol deionized water (1:1, 100 mL), collected as a dark solid and dried under vacuum overnight (78 %). 1HNMR ( 300 MHz, CDCl3) = 8. 42 (bs, 2H), 7.72 (m, chain-ends), 7.41 (m, chain-ends), 4.25 (bs, 8H), 3.60 (bs, 8H), 3.35 (bs, 8H), 1.31 (bs, 36H), 0.91 (bs, 24H). 5. 3 Conclusions and Outlook A series of alternating conjugated DA polymeric hybrids cont aining 3,4 -dioxythiophenes, 2,1,3 benzothiadiazole and unsaturated spacers (either ethynylene or trans -ethylene) were synthesized (P2 P3 P5 and P6 ) electrochemically characterized and their electrochromic

PAGE 183

183 performance evaluated with the goal of attaining a better understanding of the structural requirements to achieving reversibly switching cathodically -coloring ECPs of high contrasts and tunable optical performance. As demonstrated by spectroelectrochemical analysis, ethynylene linkers hindered the forma tion of a defined bipolaronic transition in the near IR and were thus found disruptive with respect to the EC potential of their subsequent alternating copolymers (P2 and P5 ). In contrast, the presence of trans -ethylene spacers regularly inserted to enhanc e the coplanarity of the DA polymeric backbones induced narrowing of the energygaps affording new colors (P3 and P6 ) distinct from that exhibited by the control polymers (P1 P4a and P4b ), including saturated green, a complementary color in the realizati on of polymeric EC displays (see Figure 5 11). The increased degree of conformational freedom provided allowed stable quinoidal geometries to be formed upon doping and a transmissive oxidized state to be reversibly attained. Further, the more saturated gre en analogue ( P6 ) showed excellent optical and redox stabilities to repeated switching (over 3000 cycles onto ITO) in comparison to its control parent, hence demonstrating the potential of this ECP for device applications. Figure 5 11 overviews the color pa lette attained throughout this study involving simple structural modifications, and spanning cyan blue, intermediate blue -green s and saturated green. Beside the fine tuning of optical properties afforded by the simple structural modifications described in this study, trans -ethylene linkers -containing ECPs open an obvious pathway toward the synthesis of polymeric electrochromes synthesized via the copolymerization of multiple chromophores. Our group is currently investigating this promising access to new col or palettes, including the color black such as desired for electronic ink-based applications.

PAGE 184

184 Figure 5 11. Overview representation of the color states attained among the 3,4 dioxythiophene (DOT) and 2,1,3 -benzoth iadiazole (BTD) based polymer series ( P 1 -P6 )

PAGE 185

185 CHAPTER 6 STRUCTURE PROPERTY RELATIONSHI PS AND DESIGN RULES IN 3,4 DIOXYTHIOPHENE 2,1,3 -BENZOTHIADIAZOLE DONOR ACCEPTOR GREEN POLYMERS FOR PHOTOVOLTAIC APPLICATIONS 6.1 Motivations for a Structure -Relationship S tudy of 3,4-Di oxythiophene 2,1,3 Benzothiadiazole Polymers With t he perspective s of producing power -generating displays taking advantage of the variety of colors accessible through -conjugated semiconducting polymers the synthetic design of polymeric systems based on the donor acceptor approach should take into account multiple requirements including 1) a wide ranging optical spectrum reflecting the desired shade while maintaining large absorption coefficients, and 2) adequate charge -carrier mobilities in thin -film devices. As discussed throughout the previous Chapter s of this dissertation, g reen is a color difficult to achieve in conjugated polymers and green -colored high-perfor mance semiconducting materials addressing the requirements for use in photovoltaic devices could be especially useful in aesthetic ally -pleasing light -harvesting technologies spanning photovoltaic trees (see Figure 6 1 a ), and photovoltaic turf s (see Fig ure 6 1b) where by each leaf like or grass -like light harvesting pixel would act as an individual solar cell for instance Importantly, each individual solar cell could be either a polymer solar cell, or a dye sensitized solar cell, and could be made fr om plastic substrates to be mechanically flexible as desired for applications associated with physical movements and mechanical constraints. The access to this technology is currently being explored by Sestar Technologies LLC who is supporting our researc h effort in this direction In a different context the utility of green -colored solar cells combined into an earth toned and lightweight photovoltaic device, potentially flexible, which could subsequently be incorporated in military gears, chameleonic fab rics, and other camouflage related applications

PAGE 186

186 (s ee Figure 6 2 a and Figure 6 2b ), can be envisaged to limit the use of conventional batteries, over the course of military operations for instance. a) b) Figure 6 1. G reen -colored semiconducting materia ls addressing the requirements for use in photovoltaic devices could be especially useful in aesthetically pleasing light harvesting technologies spanning a) photovoltaic trees (photograph shows one of the photovoltaic trees standing in Styria, Austria) and b) photovoltaic turfs where each leaf like or grass like light -harvesting pixel would act as an individual solar cell a) b) Figure 6 2. G reen -colored solar cells combined into an earth toned and lightweight flexible photovoltaic device fo r example could find potential utility in military gears, chameleonic fabrics, and other camouflage related applications and can be envisaged to limit the use of conventional batteries, over the course of military operations for instance

PAGE 187

187 While dye -sensi tized solar cells of practically any color can be made based on the use of small molecules,165 167 the most established photovoltaic polymers can only yield red ( e.g. P3HT),168 orange ( e.g. MDMO or MEH -PPV),40 or blue ( e.g. PCPDTBT)44 light -harvesting modules. Though structure -property correlation studies can help establish parallels and trends between molecular structure, color, charge transport and PV performance, the number of works describing a specific polymer design and the rationale that led to it remains limited. As described throughout Chapter s 3 5 of this dissertation, solution -processable neutral state green -conjugated polymers can be produced based on electronrich 3,4 -dioxythiophenes (DOTs) and the electron -deficient 2,1,3 benzothiadiazole (BTD). However, to date, semiconducting polymers employing DOTs and BTD have shown limited performance in photovoltaic devices likely due to a combination of unbalanced charge-carrier mobilities, and un favorable morphology when blended with the strong electron acceptor PCBM (interpenetrating network not forming).169 Further, it can be argued that the pre sence of a high concentration of aliphatic solubilizing side -groups in the proposed conjugated backbones, inducing steric hindrance and forcing the main-chain out of planarity, may have been a contributing factor in their moderate photovoltaic efficiencies Following these consideration, careful design of DOT BTD based semiconducting polymers suitable in the context of solar cell applications is desired and should be pursued with the goal of producing neutral state green photovoltaic polymers, for example. 6.2 Proposed Design and Rational Interpenetrating and bicontinuous networks of semiconducting polymers and fullerene derivatives, namely bulk heterojunctions (BHJs), have rapidly become the mainstay of research in the area of organic solar cell a fter their first description by Heeger et al. in 1995,36 promising

PAGE 188

188 higher densities of photogenerated charges, improved charge extractions and low cost solution processing. While power conversion effici encies as high as 10% in solar cells made of a single active layer39 and 15% in tandem devices40 have been anticipated, the highest reported values surprisingly remain in the 5 6% range.41 As described earlier in this dissertatio n (see Chapter 1 ), o f all the parameters governing BHJ device performance, the polymer energy band structur e prevails as it sets the breadth and location of the spectral absorption (via the bandgap) directly relates to the device open -circuit voltage ( VOC, via the HOMO) and imparts the photoinduced electron transfer to the strongly accepting fullerene analogue. In 2006, Janssen et al. reported on the performance of a branched di alkyl -substituted 3,4 dioxythiophene and 2,1,3 -benzothiadiazole -containing cop olymer in BHJ solar cells with PC60BM as the fullerene electron acceptor .169 The low bandgap DA hybrid (~1.55 eV) was found to exhibit a spectral response extending from 350 to 800 nm (with a main optical transition located in the 500 to 800 nm range of the visible) along with a FF of 42% and an overall PCE of 0.9% under AM 1.5 solar illumination at optimized polymer:PCBM blend composition. Importantly, the DOT BTD derivative described allowed the construction of cells exhibiting a relati vely high VOC of 0.77 V, but the device EQE remained low (maxima of 20% at ~450 nm and 13% at ~630 nm), therefore suggesting a performance limi ted by the charge transport With this important contribution169 in mind, a series of DOT -BTD c opolymers could be envisaged and designed to provide insight into how careful stru ctural modifications can be achieved to enhance the charge transport properties and photovoltaic response of DA conjugated polymers. The polymer series illustrated in Figure 6 3 was developed c onsidering the DOT BTD polymer analogue independently developed by Janssen et al. for solar cells and by our group for electrochromic applications (see Chapter 4 ) as a platf orm for n ew synthetic design.

PAGE 189

189 This proposed design is envisioned to demonstrate how the polymer optical absorption can be fully transferred into the visible (hence reflecting the color green) according to the principles previously described in the manuscri pt (see Chapter s 3 ), and its charge transport properties simultaneously optimized to improve the polymer EQE in devices and, in turn, the device PCE owing to the increase in photogenerated current Figure 6 3 Schematic illustration of the proposed desi gn for the synthesis of dioxythiophene benzothia diazole (DOT BTD) DA copolymers with tunable absorption spectra and variable charge transport properties (Adapted with permission from Ref.170 Copyright 2009 American Chem ical Society) 6.3 Synthesis and Characterization of the 3,4 -Dioxythiophene 2,1,3 -Benzothiadiazole Polymer Hybrids 6.3.1 Design, Synthesis and Physical Characterization T he utility of the donor acceptor approach in the design of solution-processable polymer electrochromes has been examined throughout the previous Chapter s of this dissertation (see Chapter s 3 5 ), demonstrating color states commonly difficult to achieve in -conjugated polymers, such as greens of t unable hues. While neutral state green polymers were a chieved

PAGE 190

190 relying on a two band absorption in the visible attainable on the sole basis of a linear combination of well chosen electronrich and poor building uni ts ,109 the molecular structure of the resulting polymers was not optimized to provide the materials with substantial thin -film charge transport properties as desired for photovoltaic devices. For example, since soluble polymers intentionally designed for electrochromic device applications gene rally incorporate a large number of appended aliphatic side -chains (insulating), they do not commonly show a pronounced degree of microstructural order (including -stacking ) when cas t and their hole mobilities remain modest in general.171173 Given that pendant groups alter the backbone planarity,173 175 the macroscopic organization,176 and governs the intermolecular distances30 in -conju gated polymers, two repeat units with limited extent in aliphatic solubilizing chains namely M2 and M3 in Figure 6 4 were designed. The same scheme illustrates the synthetic access to the copolymers P1 -P3 Here, the electron accepting building unit BTD w as first symmetrically substituted with two electron rich 3,4 dialkoxythiophenes comprising 2 -ethylhexyl pendant solubilzing groups to yield the intermediate M1 (as a bright orange oil). M2 and M3 were subsequently obtained in a Stille -type palladium -media ted cross coupling111 in gently refluxing THF between the halogenated precursor M1 and either (3,4 -dimethoxythiophen 2 -yl)trimethylstannane ( M2 ) o r tributyl(thiophen 2 yl)stannane ( M3 ). C olumn chromatography over silica using mixtures of hexanes and dichloromethane afforded M2 and M3 which were isolated as tacky solids. P2 and P3 were produced by room temperature oxidative polymerization of the easi ly oxidized macromonomers M2 and M3 using the mild oxidizing agent FeCl3 and were subsequently reduced with hydrazine P2 and P3 were precipitated in methanol and purified by Soxhlet extraction with methanol before further characterization and device fabri cation. It is worth noting

PAGE 191

191 that while P1 was first produced by Janssen et al. via a nickel -catalyzed Yamamoto polycondensation approach to be investigated as a photovoltaic semiconducting polymer ,169 nickel -mediated polymerizations requir e elevated temperatures and oxygen -free environments. Here, oxidative polymerization s at ambiance proved to be an effective alternative. In the resulting copolymers P2 and P3 the DOTs symmetrically functionalized onto BTD and bearing the solubilizing side -chains are spaced by a small conjugated block composed of two thienylene heterocycles possessing either short pendant groups ( P2 ) or no substituent ( P3 ). The presence of said conjugated spacers is expecting to increase the backbone planarity appreciably i n comparison with the control polymer P1 ( directly obtained from the self -condensation of M1 as described in Chapter 4 ). In particular it can be anticipated that the steric hindrance related to the in -plane branched alkoxy substituents as in P1 forces tw o adjacent DOT building units out of planarity, therefore minimizing the overall polymer conjugation length. In contrast, reducing t he concentration of aliphatic side groups along the main-chain to the necessary extent as in P2 and P3 should reduce the c hain -to -chain distances, favor the -stacking in teractions and, in turn, improve the charge transport properties of the corresponding materials.177179 Importantly, the electron -rich character of M2 and M3 was anticipated to enable the polymer two-band ab sorption in the visible suitable to set the color green previously observed for the neutral state green -to -transmissive swicthing polymer electrochrome s disc ussed throughout Chapter s 3 and 4 of this dissertation The structures of the repeat unit of P1 -P3 were supported by 1H NMR and the quality of the polymers was probed by elemen tal analysis (see Experimental s ection below) The polymer molecular weights wer e determined via polystyrene -calibrated GPC with THF as the mobile phase and are summarized in Table 6 1 N umber average molecular weight s (Mn) ranging from

PAGE 192

192 16,300 g mol1 (P1 ) to 69,300 g mol1 ( P3 ) and relatively narrow PDIs (1.6 -2.6) were estimated for the polymers It is worth noting that the reported molecular weights were estimated from polymers non-fractionated beyond isolation by precipitation in MeOH h ence demonstrating the effectiveness of the proposed synthetic route involving long oligomers whi ch can be easily oxidized as the precursors for polymerization. This last comment is especially relevant given that no consideration of the monomer -to -comonomer stoi chiometry is to be taken into account following this polymerization pathway Importantly, t he minimum average number of repeat units for all the polymers was found to be 20 ( P1 ), corresponding to a main -chain of nearly 60 aromatic units, which is 4 times as high as the value for which the electronic properties of DA conjugated polymers are commo nly found to saturate (~15 aromatic rings).112 However, it is also worth noting that the polymerization degrees calculated for P2 and P3 appear especially large considering the conventional polymerization techniques u sed in this work, which may point towards an overestimation inherent to the instrument/method employed in the determination of the corresponding molecular weights. Thermogravimetric analysis of P1 -P3 showed only negligible weight loss below ~320oC (under n itrogen atmosphere ), hence demonstrating their excellent thermal stability. Figure 6 4 Synthetic route to the dioxythiophene -benzothiadiazole (DOT BTD) DA copolymers P1 -P3 (Adapted with permission from Ref.170 Copyr ight 2009 American Chemical Society)

PAGE 193

193 Table 6 1. Number average molecular weight ( Mn, g mol1), Weight average molecular weight (Mw, g mol1), Polydispersity index (PDI), Average number of repeat units, Average number of rings, and Onset of decomposition te mperature (as measured by TGA under nitrogen atmosphere) for the donor acceptor copolymers P1 -P3 (Adapted with permission from Ref.170 Copyright 2009 American Chemical Society) Polymer Mn (g mol1) Mw (g mol1) PDI Avg. No. of Repeat Units Avg. No. of Rings Td (oC)a P1 16,300 42,400 2.6 20 60 321 P2 43,000 90,300 2.1 39 195 322 P3 69,300 110, 9 00 1.6 71 355 321 a Onset decomposition temperature measured by TGA under nitrogen. 6.3.2 P olymer Optical Characterization As illustrated in Figure 6 5 w hile P1 -P3 all exhibit a two -band absorption in the UV visible spectrum their short -wavelength transition differs in relative intensity and position in the spectrum with respect to the long -w avelength transition among the polymer series In toluene solution, P1 shows a first local absorption maximum in the UV (391 nm) and at a distance of ~250 nm from the second absorption maximum (638 nm). In contrast, in the case of P2 and P3 the first loca l absorption maximum has been shifted to the visible (430 and 422 nm, respectively) and is separated by less than 190 nm from the second absorption maximum (613 and 591 nm, respectively) hence producing a significant overlap between the corresponding opti cal transitions, along with shifting the window of transmission towards the higher wavelengths (484 and 475 nm, respectively) In the solid state, the inter -band windows of transmission are further transferred in the gree n region of the visible (480 550 nm), which is sufficient to provid e P2 and P3 with a neutral state green color In comparison, the shorter wavelength abso rption band of P1 remains less intense and less shifted than that of P2 and P3 and a ccordingly, P1 shows a more conventional saturated blue color in its neutral state. Table 6 2 summarizes the local absorption maxima for P1 P3 in toluene and as cast.

PAGE 194

194 While the low energy transition in these spectra can be attributed to the intramolecular donor acceptor interaction arising from the coval ent bonding of electron -donating (DOTs) and electron accepting (BTDs) aromatic units, the higher energy transition can be assigned to the electron -donating contribution of the alternating p olymer backbone (i.e. the all -thienylene segments) as described ear lier in this dissertation (see Chapter s 3 and 4). Figure 6 5 Solution (in toluene) and thin film optical absorption spectra of DA -copolymers a) P1 b) P2 and c) P3 The spectra of each system are normalized at the longer wavelength absorption maxi mum (Adapted with permission from Ref.170 Copyright 2009 American Chemical Society)

PAGE 195

195 Table 6 2. Local absorption maxima (solution in toluene and solid state) and film optical bandgap ( Eg, eV) for the donor acceptor copo lymers P1 -P3 (Adapted with permission from Ref.170 Copyright 2009 American Chemical Society) Polymer abs (nm) In Toluene abs (nm) Thin Film E g ( e V) 1 2 1 2 P1 391 638 399 653 (708) a 1.6 P2 430 613 446 651 1.55 P 3 422 591 448 636 1.65 a Shoulder 6.3.3 Polymer Electrochemistry Cyclic voltammetry (CV) and differential pulse voltammetry (DPV) were used to investigate t he redox properties of P1 P3 including determining how polymer oxidation and reduction potentials, associated HOMO and LUMO energy levels and electrochemical bandgaps vary with presence/absence of the bithienyl conjugated spacer. Table 6 3 summarizes these results, along w ith the optical bandgap of the polymers as es timated from the onset of their low energy transition. Polymer t hin films w ere drop -cast from toluene solutions ( 6 mg mL1) onto platinum button electrodes and characterized in an electrochemical cell comprising a platinum counter -electrode and a silver wi re as the reference electrode, with the said electrochemical cell set in an argon -filled dry box All estimated potentials were subsequently calibrated versus Fc/Fc+. The polymer films were systematically cycled prior to characterization, until a stable an d reproducible redox response was reached T he redox characteristics of P1 have been described earlier in this dissertation (see Chapter 5 ). In brief P1 revealed a low oxidation potential of +0.25 V by CV (and supported by DPV) which correspon ded to a rel atively high HOMO energy leve l of 5.35 eV relative to vacuum The onset of reduction was found at 1.57 V by CV and 1.52 V by DPV, which corresponded to a LUMO of 3.53 eV by CV and 3.59 eV by DPV, respectively, and to electrochemical bandgaps of 1.82 and 1.77 eV.

PAGE 196

196 Figure 6 6 a shows t he CV and DPV traces of P2 P2 revealed a lower oxidation potential than that of P1 +0.18 V by CV and +0.14 V by DPV, and thus a higher lying HOMO, 5.28 eV from CV and 5.24 eV from DPV, in agreement with the introduction of the 3,4 dimethoxythiophene based conjugated spacer increasing the electron rich contribution in the DA backbone. It is worth noting that t he signal corresponding to the first reduction process observed was relatively weak in comparison to that of the second reduction seen by CV, the second one inducing rapid degradation of the polymer on repeated cycling. The onset estimated from the first reductive process, 1.63 V by CV and 1.48 V by DPV, was consistent with the existing literature relating to BTD contain ing DA copolymers.44,45,180 From the associated LUMOs, 3.47 eV by CV and 3.62 eV by DPV, energy gaps of 1.81 and 1.62 eV, respectively, were determined in agreement with those found for P1 a) b) Figure 6 6. a) Cyclic (scan rate of 50 mV/s) and differential pulse voltammograms (step time of 0.1 s) of P2 drop -cast onto a platinum button electrode (0.02 cm2) in 0.1M TBAP/PC electrolyte solution b) Differential pulse voltammogram (step time of 0.1 s) of P3 drop -cast onto a platinum button electrode (0.02 cm2) in 0.1M TBAP/PC electrolyte solution (Adapted with permission fro m Ref.170 Copyright 2009 American Chemical Society) In contrast with P2 the incorporation of a bithiophene conjugated spacer in P3 represse d the polymer HOMO by ca. 0.3 eV (5.60 eV by CV and 5.53 eV by DPV). This impor tant result can be attributed to the less pronounced electronrich character of thiophene when compared to

PAGE 197

197 that of 3,4 -dimethoxythiophene In comparison with P1 and P2 polymer P3 revealed a higher oxidation potential (+0.50 V by CV and +0.43 V by DPV) wh ile retaining a near identical onset of reduction ( 1.6 V by CV and 1.47 V by DPV). As illustarted in Figure 6 -6 b, and i n analogy with P2 the signal corresponding to the first reduction process observed was weaker than that of the second reduction which induced rapid degradation of the polymer on repeated cycling. In general, the electrochemical bandgaps were found to be slightly smaller when estimated by DPV (by up to 0.2 eV lower than those determined by CV ) As previously mentioned in this dissertation (see Chapter 5 ), DPV tends to provide more accurate estimates when evaluating the redox potentials of conjugated polymers 1) by reducing the contribution of t he charging background currents, and 2) by inducing sharper redox onsets than CV.128,181 Although the electrochemical ly determined bandgaps we re systematically larger than the optically estimated values this experimental observation is in good agreement with work on DA polymers from various groups .46,112,182 For instance, the bandgap of P2 estimated by CV (1.81 eV) differs from its optically determined analogue (1.55 eV) from 0.26 eV. Similarly, the CV estimated gap of P3 (2.1 eV) differs significantly from its optically determined analogue (1.55 eV), though the gap estimated by DPV (1.9 eV) only differs from 0.35 eV from the optical analogue Table 6 3. Electrochemically determined HOMO and LUMO energy levels (by CV and by DPV), Corresponding electrochemical bandgaps, and Compariso n with their optically estimated values for the copolymers P1 P3 (Adapted with permission from Ref.183 Copyright 2009 American Chemical Society) Polymer Eoxonset (V) HOMO (eV) Eredonset (V) LUMO (eV) Egap (V) Egap (V) CV DPV CV DPV CV DPV CV DPV CV DPV Optical P1 0.25 0.25 5.35 5.35 1.57 1.52 3.53 3.59 1.82 1.77 1.60 P2 0.18 0.14 5.28 5.24 1.63 1.48 3.47 3.62 1.81 1.62 1.55 P3 0.50 0.43 5.60 5.53 1.60 1.47 3.50 3.63 2.10 1.90 1.65 a oxidation ( Eoxonset ) and reduction ( Eredonset ) potentials are reported vs Fc/Fc+. HOM O and LUMO energy levels are derived from the electrochemical data ( Eoxonset and Eredonset respectively) considering that the SCE is 4.7 eV vs. vacuum and Fc/Fc+ is 0.38 eV vs. SCE, i.e. 5.1 eV relative to vacuum.

PAGE 198

198 Given the above discussion, the energy di agram presented as Figure 6 7 was constructed by employing the DPV estimated energy levels of the polymers. In short, Figure 6 7 highlights the appropriate match between the energy band structure of the designed polymers and the key energy levels of the va rious components commonly utilized in the fabrication of BHJ photovoltaic test -devices ( e.g. ITO, PEDOT:PSS, PCBM and Al).44,95 Importantly, the rather low lying HOMO of P3 (when comp ared to that of P1 and P2 ) is expected to have a favorable impact on the open circuit voltage of the solar cells made thereof, which could ultimately enhance the overall power conversion efficiency of the corresponding devices.10,11 Figure 6 7 Sch ematic representation of the e nergyband structure of P1 P3 showing the polymer HOMO and LUMO energies as estimated by DPV (green filled rectangles) with respect to those of an ideal polymer ( i.e. designed to be integrated in BHJ solar cells employing PC60BM and PEDOT -PSS). The optically determined bandgaps of P1 P3 were placed at the baricenter of the DPV estimated bandgaps, and a second approximated set of HOMO and LUMO levels could be defined assuming the energy levels equidistant from the baricenter ( in green dotted rectangle) (Adapted with permission from Ref.170 Copyright 2009 American Chemical Society)

PAGE 199

199 6.3.4 Photvoltaic Device Performance (Experimental work was achieved and results kindly supplied by Dr Jegadesan Subbiah) The PV performance of P1 P3 in donor acceptor BHJ photovoltaic devices employing PC60BM as the acceptor were investigated under simulated AM 1.5 G solar illumination (irradiation intensity: 100 mW cm2). At first, an optimization study of the pol ymer:PCBM blend composition was carried out with P1 P3 whereby the active layer was spin -cast from chlorobenzene in devices using PEDOT -PSS (~30 nm) coated ITO Figure 6 8 overlays the current density-voltage ( J V ) responses of devices employing P1 (blue curve), P2 (green curve) and P3 (red curve) at opti mized polymer:PCBM composition. Table 6 4 summarizes the PV data for device s based on the neutral state blue DA polymer P1 D evice s were made with a polymer:PC60BM ratio of 1:4 (by weight) with the goal of reproducing the results attained by Janssen et al. in an earlier work.169 In our laboratories, the 1:4 polymer:PC60BM device revealed ne ar -identical PV characteristics, with a JSC of 2.92 mA cm2, a VOC of 0.80 V and a moder ate FF of 3 8%. The same device exhibited an overall PCE of 0.8 8 % which is in excellent agreement with the PCE of 0.9% reported by Janssen et al. Importantly, the proximity of our results in device performance also confirmed the quality of o ur polymer P1 although pr epared under the relatively harsh oxida tive polymerization conditions (due to the large quantities of HCl generated and the possibility for cross links to be produced, as opposed to a Ni catalyzed polycondensation following Yamamoto conditions whereby the polymer grows in its neutral form ) and subsequently reduced with hydrazine before being p urified according to different methods (see Experimental below) Nevertheless in spite of the relatively high VOC measured, the performance of the low -bandgap polym er P1 remains moderate at the best polymer:PCBM blend composition (1:4) which could be the result of the formation of an unfavorable blend morphology ( e.g. poor interpenetrating network, pronounced phase

PAGE 200

200 segregation between the polymer and PC60BM ), or the consequence of a charge transport limitation across the thin -film devices, or possibly the combination of both. Figure 6 8. IV curves for BHJ solar cells made of P1 P3 (at best polymer:PC60BM composition) in the dark (black curve, only P1 is represent ed) and under AM 1.5 solar illumination, 100 mW cm2 (blue, green and red curves). The photograph illustrates the green color attained in the case of P3 at the polymer:PCBM ratio of 1:8 (device is backlit and the Al contacts appear black). Devices with pos t -polymer processing thermal treatment at 70oC (30 min). The device structure is ITO/PEDOT/PX:PC60BM/LiF/Al with PX = P1 P2 or P3 (Adapted with permission from Ref.170 Copyright 2009 American Chemical Society) Table 6 4. Solar cell device performance for P1 (under AM 1.5 illumination at an irradiation intensity of 100 mW cm2) as a function of polymer:PC60BM blend composition. Devices with post -polymer processing thermal treatment at 70oC (30 min). The device structure is ITO/PEDOT/P1:PC60BM/LiF/Al (Adapted with permission from Ref.170 Copyright 2009 American Chemical Society) P1:PC60BM (w/w) Jsc (mA cm2) Voc (V) FF (%) PCE (%) 1:4 2.92 0.80 0.38 0.88 1:6 2.15 0.80 0.42 0.73 Tabl e 6 5 summarizes the PV data for devices based on the neutral state green DA polymer P 2 comprising a bi dioxythienyl spacer with short pendant -groups (methoxy). The best P2 -based device was reached with a polymer:PCBM ratio of 1:5 (by we ight ). This P2 -base d device revealed a higher JSC of 3.56 mA cm2, but a markedly lower VOC of 0.62 V which likely contributed to the drop of 6% in FF (32%) when compared to the best device made with P1 ( FF = 38%). T he lower VOC observed in all P2 based devices is in agreeme nt with the high -lying

PAGE 201

201 HOMO of the polymer estimated electrochemically (5.24 eV relative to vacuum by DPV). The PCE reached 0.70% under AM 1.5 at the best device composition which was by ca. 0.2 eV lower than that calculated for the best P1 -based device. Here, in spite of its two band absor ption extended across the visible spectrum and its reduced insulating character ( limited number of solubilizing groups along the backbone) P2 did not imply an overall improvement over P1 This result was supported by th e relatively low EQE of ca. 24% obtained at 480 nm (see Figure 6 9 a). Here again, the performance limitation could result from a high concentration of cha rge recombinations due to the buildup of space charges throughout the active layer a poor interpenetr ating network, or the combination of both. Table 6 5. Solar cell device performance for P2 (under AM 1.5 illumination at an irradiation intensity of 100 mW cm2) as a function of polymer:PC60BM blend composition. Devices with post -polymer processing therma l treatment at 70oC (30 min). The device structure is ITO/PEDOT/P2:PC60BM/LiF/Al (Adapted with permission from Ref.170 Copyright 2009 American Chemical Society) P2:PC60BM (w/w) Jsc (mA cm 2 ) Voc (V) FF (%) PCE (%) 1:4 1.90 0.58 33 0.37 1:5 3.56 0.62 32 0.70 1:6 3.08 0.58 33 0.58 1:8 2.60 0.58 36 0.54 Following those results the copolymer analogue comprising an unsubstituted bithiophene spacer P3 was investigated as an alternative to P2 with the hope that it would retain the targeted neutral green color state Table 6 6 summarizes the PV data for devices based on P3 The corresponding BHJ devices proved to be especially sensitive to the PC60BM loading and at the best polymer:PC60BM composition (1:8), a JSC as high as 5.56 mA cm2, a large VOC of 0.77 V near -equal to that of P1 and an improved FF of 44% were estimated. The VOC observed in all P3 -based devices similar to that observed in the P1 -based devices is in excellent agreement with the low lying HOMO of the polymer estimated electrochemically (5.53 eV relative to

PAGE 202

202 vacuum by DPV) The PCE reached up to 1.90% under AM 1.5. T he EQE data of P3 is compared to that of P2 in Figure 6 9 and consists of a two -band res ponse with the long wavelength local maximum of th e photocurrent corresponding to the absorption maximum of the long -wavelength absorption band of P3 The short -wavelength maximum of the photocurrent is dominated by the absorption of PC60BM added to that of the short -wavelengt h absorption band of P3 The on set of photocurrent at 750 nm was in agreement with the polymer optical absorption data (overlaid in Figure 6 9 ). I mportantly, the EQE response of the P3 -based device containing 80% PC60BM (blend ratio of 1:4) presented a maximum of 28% at 460 nm wherea s that of the device containing 88.9% PC60BM (blend ratio of 1:8) showed a maximum of 54% at t he same wavelength, hence confirming the devi ce sensitivity to the PC60BM loading in the blend that was observed throughout the polymer:PC60BM device optimization stud y. Figure 6 9. Superimposed EQE (blue curves) and polymer:PC60BM blend absoprtion (red curves) for P2 (filled squares) and P3 (empty circles) based BHJ solar cells (at best polymer:PCBM composition) and under AM 1.5 solar illumination, 100 mW cm2 (Adapted with permission from Ref.170 Copyright 2009 American Chemical Society) When comparing the EQE data to the optical absorption spectra of the 1:4, 1:6 and 1: polymer:PC60BM blends (measured from the actual sol ar cells) the minimal differences between absorption spectra of all the blends did not correlate to the large enhancements observed in EQE from one device to the next such that an increase in charge -carrier mobility, similar to that observed in MDMO -PPV: PC60BM systems, could explain this important result.

PAGE 203

203 In order to probe the effect of a significant raise in PC60BM loading on the PV properties of P1 a test -device was constructed employing a polymer: PC60BM ratio of 1:6 (see Table 6 3). In this case, the increase of fullerene in the blend (from 80% to 85.7%) did not increase the PCE (0.73 %) in contrast with the results obtained for P3 In parallel JSC decreased to an average of 2.15 mA cm2, hence confirming the blend composition (1:4) initially propose d by Janssen et al. to be optimum in the PV performance of P1 Table 6 6. Solar cell device performance for P3 (under AM 1.5 illumination at an irradiation intensity of 100 mW cm2) as a function of polymer:PC60BM blend composition. Devices with post -polym er processing thermal treatment at 70oC (30 min). The device structure is ITO/PEDOT/P3:PC60BM/LiF/Al (Adapted with permission from Ref.170 Copyright 2009 American Chemical Society) P3:PC60BM (w/w) Jsc (mA cm 2 ) Voc (V) FF (%) PCE (%) 1:4 1.99 0.78 41 0.64 1:6 3.93 0.77 46 1.39 1:7 4.79 0.76 46 1.68 1:8 5.56 0.77 44 1.90 1:9 3.97 0.73 35 1.00 1:10 2.10 0.71 40 0.60 Beside the effect of the polymer: fullerene bl end composition on the charge -transport properties and nanoscale morphology of the subsequent active layer the efficiency of the hole extraction process depends on the nature (and work -function) of the interface material employed between the conducting metal oxide electrode (anode) and the organic active laye r. For example, as recently demonstrated by various groups, metal oxides including NiOX, MoO3 and V2O5 can be used as the hole transporting interface layer ( replacing the often used PEDOT PSS) to enhance the performance of polymer solar cells significantly .37,95 Here, we have investigated the benefit of incorporating MoO3 as an interface layer between ITO and the polymer: PC60BM active layer in devices made with P3 and the results are presen ted in Table 6 7 As PEDOT PSS (~25 nm) was replaced by MoO3 (~10 nm) in a device made with the best polymer: PC60BM

PAGE 204

204 blend composition (1:8) the FF increased from 46 % to 51 % and the device PCE was in turn, raised from 1.90 % to 2.12 %. Since low FFs are commonly explained by a relatively high concentration in electron -hole recombinations across the active layer where the excitons only have limited lifetimes ,11 it is essential to facilitate the extraction of the charge -carrier by enhancing the quality of the polymer -electrode interface. C onsidering the work -function of MoO3 at about 5.3 eV (relative to vacuum) as often stated in the literature, the favorable alignment in energy levels between ITO MoO3 and P3 (see Figure 6 10) could account for the reduced series resistance and the FF improvement Expectedly, in this device configuration, JSC (5.38 mA cm2) and VOC (0.77 V) remain ed near identical to the values obtained for the PEDOT based device configuration. Figure 6 10. Energy -level diagram correlating the energy band structure of P3 and the work function of MoO3 (here replacing PED OT PSS as the hole -transporting interface layer) and PC70BM (here replacing PC60BM). When P3 was blended with PC70BM and the blend was cast onto MoO3, the JSC increased again to an average of 5.96 mA cm2 (owing to PC70BM) and the VOC remained the same (0. 80 V). Importantly, the corresponding FF increased from 48% to 57% (owing to MoO3), which significantly enhanced the PCE to 2.71% (Adapted with permission from Ref.170 Copyright 2009 American Chemical Society)

PAGE 205

205 A s illustrated in Figure 6 1 1 and summarized in Table 6 7, the use of PC70BM as the electron transport material along with either PEDOT PSS or MoO3 as the hole transporting interface layer was further investigated in the P3 based solar cell. In comparison with PC60BM, PC70BM possesses a visible absorption (due to symmetry differences accounting for distinct sets of spin allowed and spin-forbidden transitions) which is often referred to in order to explain the performance enhancements typically monitored in BHJ devic es employing donor acceptor semiconducting polymers.43,46 W hen a blend of P3 and PC70BM (1:8) was cast onto PEDOT PSS, JSC reached an average of 5.90 mA cm2, and a somewhat higher VOC of 0.80 V was measured possibly due to PC70BM having a slightly higher LUMO in comparison to that of PC60BM (by ca. 0.1 0.2 eV).184 The corresponding FF remained high (48%) and the PCE increased to 2.31%. When a blend of P3 and PC70BM (1:8) was cast onto MoO3, JSC increased even more to reach 5.96 mA cm2 and the VOC remained the same (0.80 V) Importantly, the corresponding FF increased from 48% to 57%, which significantly enhanced the PCE to a value of 2 .71 %. Table 6 7. Optimization of the solar cell device performance for P3 (under AM 1.5 illumination at an irradiation intensity 100 mW cm2) by varying the hole -transporting interface layer (PEDOT or MoO3) and the type of PCBM used (PC60BM or PC70BM). Dev ices with post -fabrication thermal treatment at 70oC (30 min) (Adapted with permission from Ref.170 Copyright 2009 American Chemical Society) P3 Jsc (mA cm 2 ) Voc (V) FF (%) PCE (%) MoO3/PC60BM 5.38 0.77 51 2.12 PEDOT /PC70BM 5.90 0.80 48 2.31 MoO3/PC70BM 5.96 0.80 57 2.71 Figure 6 1 2 compares t he EQE responses of P3: PC70BM and P3: PC60BM based devices Interestingly, a broad -band resp onse reaching ca. 60% at 480 nm was observed when the active layer consisted of a b lend of P3 and PC70BM as the optical absorption spectrum of PC70BM

PAGE 206

206 complemented that of P3 in the 410620 nm range (the device took a brown hue). The onset of photocurrent at 750 nm was also in agreement with the blend optical absorption (overlaid in Figu re 6 1 2 ). Here, the notable EQE enhancement can reaso nably be attributed to the improved spe ctral absorption in the visible owing to the use of PC70BM. Figure 6 1 1 IV curves for BHJ solar cells made of P3 (at best polymer:PCBM composition, i.e 1:8) i n the dark (black curve) and under AM 1.5 solar illumination, 100 mW cm2, for different device configurations. The device structures are ITO/MoO3/ P3 :PC60BM/LiF/Al (blue curve), ITO/PEDOT/P3 :PC70BM/LiF/Al (green curve), and ITO/MoO3/ P3 :PC70BM/LiF/Al (red c urve). Devices with post -polymer processing thermal treatment at 70oC (30 min) (Adapted with permission from Ref.170 Copyright 2009 American Chemical Society) Figure 6 1 2 Superimposed EQE (blue curves) and polymer:P CBM blend absoprtion (red curves) for BHJ solar cells made of P3 (at best polymer:PCBM com position) with the architecture ITO/PEDOT/ P3 :PC60BM/LiF/Al ( empty circles ) and ITO/MoO3/ P3 :PC70BM/LiF/Al (filled circles), under AM 1.5 solar illumination, 100 mW cm2 (Adapted with permission from Ref.170 Copyright 2009 American Chemical Society)

PAGE 207

207 6.3.5 Device Morphology Study (Experimental work was achieved and results kindly supplied by Dr Jegadesan Subbiah) The active layer of th e PV devices at their optimum blend composition was characterized by AFM ( tapping -mode ) with the purpose of understand ing how the blend morphology changes as a function of the copolymer employed in the donor acceptor bulk -heterojunction Each device was su bjected to a mild thermal treatment (70oC, 30 min) prior to Al contact deposition. PC60BM was employed to produce the photovoltaic blends. Figure 6 1 3 a and 6 1 3 b show the 3D and 2D -images, respectively, of the P1 based device surface. The coarse morphology revealed in the presented patterns could result from a rather pronounced phase segregation (demixing) between PCBM and the polymer, as described in existing literature on the topic .72,185,186 Assuming a demixing of the polymer and the fullerene the lack of interpenetrating network would validate the performance limitation seen for devices involving P1 Based on earlier work from various groups ,187,188 th e domains of higher heights (up to 60 nm) can be attributed to the formation of PCBM clusters excluding the presence a large concentration of mixed polymer Figure 6 1 3 c shows the 3D -image of the P2 based device surface. In this pattern, a significantly mo re homogeneous film morphology can be seen where the heights are limited to ca. 6 nm thereby pointing towards a lack of phase separation across the active layer (in contrast with the lack of phase mixing between P1 and PCBM) As previously mentioned in this dissertation (see Chapter 1 ), a certain extent of phase separation between donor and acceptor in the BHJ remains a necessary condition in the formation of the bicontinuous network desired to attain efficient PV devices.72 In contrast, Figure 6 1 3 d and 6 1 3 e showing the 3D and 2D -images, respectively, of the P3 -based device surface suggest a more hete rogeneous morphology where relatively large domains of higher heights (ca. 6 nm outlined in black in the 2D -AFM image ) are dispersed across the active lay er. Considering the consequent PCBM

PAGE 208

208 loading (~88.9%) of the P3 -based device at best comp osition (1:8 by weight, polymer: PCBM), it is reasonable to assign these domains to a PCBM -rich phase, where o vergrown PCBM -clusters are not produced. A dditiona l infor mation on the variations observed in the PV performance of P1 P3 in BHJ could be revealed by further characterization of the charge -transport properties of the active layer composing the devices at best composition. Figure 6 1 3 AFM tapping-mode images of P1 -P3 in blend with PC60BM (best polymer:PC60BM compositions are represented). Film surfaces from devices with post -polymer processing thermal treatment at 70oC (30 min). a) and b) P1:PC60BM, c) P2:PC60BM, d) and e) P3:PC60BM. All images are 2 x 2 m (A dapted with permission from Ref.170 Copyright 2009 American Chemical Society)

PAGE 209

209 6.3.6 Charge Transport in Devices (Experimental work was achieved and results kindly supplied by Dr Kaushik Roy Choudhury) As previously ment ioned in this manuscript (see Chapter 1 ), for semiconducting polymers with relatively low charge -carrier mobilities, whereby the dark current in devices is space charge limited (restricted by the buildup of charges) the hole mobility can be estimated from simple I -V measurements using the Mott -Gurney equation for trap-free SCLC. For this purpose, hole only devices were fabricated for the polymers based on the following architecture: ITO/PEDOT PSS/Polymer(or P olymer :PCBM) /Pd where Pd ( ca. 5.4 eV) has a higher lying work -function than Au ( ca. 5.2 eV), hence blocking the electron injection into the PCBM rich phase more effectively Figure 6 1 4 shows the dark current densities measured from the hole -only devices made with the pristine polymers P1 -P3 (see Figur e 6 1 4 a) and from those made with the polymer:PCBM blends (see Figure 6 1 4 b) optimized for the best device performance. The applied voltage V is corrected by accounting for the built in voltage Vbi resulting from the difference in the work -function between electrodes, and the current density is plotted versus the effective electric field (Eeff) rather than the applied voltage in order to allow comparison of devic es possessing active layers of different thicknesses As seen in Figure 6 1 4 a, the current densi ties measured at any given electric field increase from P1 to P2 to P3 including at low device bias. This result is in good agreement with the favorable changes anticipated in charge transport throughout the polymer series as reducing the concentration of solubilizing side chains along the polymer main -chain via the introduction of short conjugated spacers should simultaneously increase the backbone planarity from P1 to P2 to P3 In parallel, i t cannot be excluded that such conjugated spacers reduce the c hain to -chain spacing, and enhance the -

PAGE 210

210 stacking although this should be confirmed by X ray scattering experiments (not in the scope of this study) The inset of Figure 6 1 4 a shows the quadratic variation of the current densities as a function of the app lied voltage from which the hole mobility of the polymers can be estimated according to SCLC modeling Table 6 8 summarizes the room temperature zero -field hole mobility values along the polymer series, which increase by almost two orders of magnitude g oi ng from P1 to P3 As illustrated in Figure 6 1 4 b, the dark current densities of hole -only devices for the optimized polymer:PCBM blend composition are significantly higher in comparison to the devices with the pristine polymers. T he resulting hole mobility in the corresponding polymers is raised by an order of magnitude or more in comparison to the value s obtained from the PCBM free devices (see Table 6 8) While this result may seem somewhat surprising considering that the hole percolation pathway of the polymer network could be unfavorably altered by the presence of high concentrations in PCBM, such enhancement s in hole mobility for polymers blended with PCBM has been described before .189 In particular, it has been suggested that a favorable morphology change occurring upon mixing the polymer and the fullerene can result in a situation where the polymer intermolecular interactions are promoted, hence facilitating the charge transport in the polymer -rich phase.189 Here, it is worth noting that the zero -field hole mobility of P3 reaches 105 cm2 V1s1 when blended with PC60BM (1:8) which is only one to two orders of magnitude lower from that of PC60BM alone or in a blend with low polymer content (104103 cm2 V1s1).190 This improved balance between electron and hole mobilities in the P3 -based device s suggest an additional explanation in the PV device performance enhancement seen with P3 As P3 is mixed with PC70BM (1:8), the zero -field hole mobility in the polymer phase increases by about an order of magnitude (when compared to that in the

PAGE 211

211 polymer:PC60BM blend), and reaches 9 105 cm2V1s1, such that the 28% improvement of overall PCE in the devices employing PC70BM and MoO3 likely result s from the combination of an adequate device energy band diagram, a favorable morphology, an improved visible absorption (owing to the presence of PC70BM) along with a higher hole mobility. a) b) Figure 6 1 4 a) Current densities for hole -only devices of copolymers P1 -P3 as a function of the effective electric field. Inset shows the quadratic dependence of the curr ent density on applied bias; the red line is a fit for using the single -carrier SCLC model. b) J -V characteristics of hole -only devices at best polymer:PCBM blend composition. The legends show the thickness of the respective active layers (Adapted with per mission from Ref.170 Copyright 2009 American Chemical Society) Table 6 8 Zero -field hole mobility in the pristine copolymers P1 P3 and in the polymer phase of the optimized blends, derived from fitting the J -V data to trap -free single -carrier SCLC model (Adapted with permission from Ref.170 Copyright 2009 American Chemical Society) Device Composition P1 P1:PC60BM P2 P2:PC60BM P3 P3:PC60BM Zero Field Hole Mobility (cm 2 V -1 s -1 ) 5.4 x 10-8 2 x 10-6 6.9 x 10-7 5 x 10-6 3.9 x 10-6 1 x 10-5 6.3.7 Synthetic Details Synthesis of Macrom onomers M1, M1, and Polymer P1: Refer to Chapter 5 of this dissertation. Compound M2 : Macromonomer M1 (1.3 g, 1.08 mmol), 3,4 -dimethoxythiophen 2 -yl) trimethylstannane (1.55 g, 5.04 mmol) and Pd(PPh3)2Cl2 (4 mol %) were cycled (argon/vacuum,

PAGE 212

212 3x) and subsequently dissolved in 30 mL of THF. The mixture was stirred for 24 hours at 60 C, the solvent was evaporated and the produ ct was purified by column chromatography on silica with hexane/dichloromethane (1:1) as the eluent. T he solvent was evaporated and M2 was obtained as a purple -blue tacky solid ( 0.97 g, 43 %). 1H NMR (300 MHz, CDCl3) = 8.30 (s, 2H), 6.18 (s, 2H), 4.06 (d, J = 6.6 Hz, 4H), 3.98 (s, 6H), 3.93 (m, 4H), 3.87 (s, 6H), 2.00 1.10 (m, 36H), 1.000.79 (m, 24H); 13C NMR (75MHz, CDCl3) = 153.07, 150.60, 148.39, 146.36, 143.46, 128.02, 124.65, 119.48, 117.70, 95.43, 76.61, 60.40, 57.46, 40.54, 40.48, 34.89, 31.82, 30. 51, 29.30, 25.51, 23.86, 23.77, 23.40, 23.30, 22.88, 20.92, 14.37, 14.35, 14.27, 11.36, 11.23. HRMS (TOF) [MH+] m/z calcd. for C58H84N2O8S5: 1097.4904 Found: 1097.4948. Anal. calcd. for C58H84N2O8S5: C 63.47, H 7.71, N 2.55 Found: C 64.03, H 7.89, N 2.56. Compound M3: Macromonomer M1 (1.4 g, 1.44 mmol), tributyl(thiophen2 -yl)stannane (1.61 g, 4.32 mmol) and Pd(PPh3)2Cl2 (4 mol %) were cycled (argon/vacuum, 3x) and subsequently dissolved in 30 mL of THF. The mixture was stirred for 24 hours at 60 C, the s olvent was evaporated and the product was purified by column chromatography on silica with hexane/dichloromethane (3:1) as the eluent. T he solvent was evaporated and M3 was obtained as a purple tacky solid ( 0.72 g, 51 %). 1H NMR (300 MHz, CDCl3) = 8.41 (s 2H), 7.39 (dd, J = 1.2, 2.4 Hz, 2H), 7.31 (dd, J = 1.2, 3.9 Hz, 2H), 7.07 (dd, J = 3.6, 1.5 Hz, 2H), 4.053.97 (m, 8H), 1.911.22 (m, 36H), 1.00 0.81 (m, 24H); 13C NMR (75MHz, CDCl3) = 152.85, 149.23, 145.99, 134.81, 127.74, 127.17, 125.29, 124.23, 123. 98, 122.21, 118.45, 76.66, 40.61, 40.55, 30.52, 29.33, 23.86, 23.37, 23.31, 14.35, 14.28, 11.36, 11.27. HRMS (TOF) [MH+] m/z calcd. for C54H76N2O4S5: 977.4481 Found: 977.4456. Anal. calcd. for C54H76N2O4S5: C 66.35, H 7.84, N 2.87 Found: C 67.10, H 7.93, N 2.94.

PAGE 213

213 Polymer P2 : Macromonomer M2 (260 mg, 0.24 mmol) was dissolved in chloroform (25 mL). A solution of anhydrous FeCl3 (195 mg, 1.2 mmol, 5eq) in nitromethane was added dropwise over a period of 45 minutes to the stirred monomer at room temperature (the dark purple monomer solution turned progressively dark blue green with addition of oxidizing agent). The mixture was stirred 24 hours at room temperature. It was then precipitated into methanol (200 mL). The precipitate was filtered, redissolved in chloro form (200 mL) and stirred for 3 hours with hydrazine monohydrate (6 mL). After evaporation, the concentrate (dark blue -green) was precipitated into methanol (200 mL), the precipitate was filtered through a Soxhlet thimble and purified via Soxhlet extractio n for 24 hours with methanol. The polymer was extracted with chloroform, concentrated by evaporation, the polymer was precipitated in methanol (200 mL) and collected as a dark solid (145 mg, 48 %). 1H NMR (300 MHz, CDCl3) = 8.33 (bs, 2H), 4.23.8 (br, 20H ), 2.0 1.1 (br, 36H), 1.1 0.7 (br, 24H) GPC analysis: Mn = 43 000 g mol1, Mw = 88 600 g mol1, PDI = 2.1 Anal. calcd. for C54H72N2O4S5 C 63.58, H 7.54, N 2.56 Found: C 64.68, H 7.69, N 2.53. Polymer P3 : Macromonomer M3 (300 mg, 0.307 mmol) was dissolve d in chloroform (50 mL). A solution of anhydrous FeCl3 (248 mg, 1.53 mmol, 5eq) in nitromethane was added dropwise over a period of 45 minutes to the stirred monomer at room temperature (the dark purple monomer solution turned progressively dark blue green with addition of oxidizing agent). The mixture was stirred 24 hours at room temperature. It was then precipitated into methanol (200 mL). The precipitate was filtered, redissolved in chloroform (200 mL) and stirred for 3 hours with hydrazine monohydrate ( 6 mL). After evaporation, the concentrate (dark blue -green) was precipitated into methanol (200 mL), the precipitate was filtered through a Soxhlet thimble and purified via Soxhlet extraction for 24 hours with methanol. The polymer was extracted with

PAGE 214

214 chlor oform, concentrated by evaporation, the polymer was precipitated in methanol (200 mL) and collected as a dark solid (145 mg, 48 %). 1H NMR (300 MHz, CDCl3) = 8.44 (bs, 2H), 7.57.0 (br, 4H), 4.3 3.8 (br, 8H), 2.0 1.1 (br, 36H), 1.1 0.7 (br, 24H) GPC anal ysis: Mn = 69 300 g mol1, Mw = 111 600 g mol1, PDI = 1.6 Anal. calcd. for C54H74N2O4S5 C 66.49, H 7.65, N 2.87 Found: C 66.69, H 7.67, N 2.77. 6.4 Conclusions and Outlook With solution -processable organic electronics su itable for combined photovoltaic/display devices, power -generating ornaments (such as photovoltaic plants and trees) or light harvesting window applications, now emerging, a necessary step forward in this area consists in developing new synthetic strategie s taking bandgap/color -engineering along with charge transport into account. Here, we have carefully investigated the structure -performance relationships in PV devices of a series of soluble donor acceptor -conjugated polymers (P1 P3 ) comprising electron -rich 3,4-dioxythiophenes (DOTs) and the electron -deficient 2,1,3 benzothiadiazole (BTD), with an emphasis on correlating molecular structure, energy band characteristics and charge transport. In particular, the synthesis and chemical polymerization of two polyheterocyclic and regio symmetric DA oligomers ( M2 and M3 ) afforded two -band absorbing photovoltaic polymers reflecting the color green ( P2 and P3 ). As the polymer structures were successfully confirmed, we realized that the approach consisting in polycondensing oligomers with low oxidation potentials using mild oxidative reagents could be employed to produce novel polymer repeat units otherwise synthetically demanding as well as high molecular weight poly mers without the need for tedious monomer purifications and respect of the monomer to comonomer stoichiometry.

PAGE 215

215 As t he PV properties of P1 -P3 were investigated in BHJ devices with PC60BM it was found that the neutral state green copolymer P3 comprising an unsubstituted bithiophene spacer exhibited a two -fold enhancement in PCE (1.90%) over the all dioxythiophene -based copolymers P1 and P2 (0.88% and 0.70%, respectively). AFM imaging of the corresponding polymer:PCBM blends (best composition) and careful ana lysis of their charge transport in hole only devices suggest essential differences in morphology and electronic properties throughout the polymer series, and support our reasoning for the synthetic design employed. By further substituting PEDOT:PSS with Mo O3 and PC60BM by PC70BM in the P3 based devices to demonstrate up to 2.71% of PCE in the absence of any tedious solvent or thermal annealing treatment. These results confirm the importance of adjusting the device configuration as a function of the intrins ic properties ( e.g. energy band structure, mobility of the charge -carriers) of the semiconducting polymers investigated. In the case of P3 the optical absorption of PC70BM came to complement the two -band absorption of the polymer and to balance the charge transport of the blend, thereby extending the photon collection over the entire visible spectrum while raising the EQE to a maximum of ca. 60% at ~ 480 nm.

PAGE 216

216 CHAPTER 7 SYNTHETIC CONTROL OF THE SPECTRAL ABSORPT ION AND CHARGE TRANSPORT PROPERTIES IN DITHIENO SILOLE 2,1,3 BENZOTHIADIAZOL E DONOR ACCEPTOR POLYMERS FOR PHOTOVOLTAIC APP LICATIONS 7.1 Context and Motivations for the Design of Broadly Absorbing Polymers with High Charge -Carrier Mobilities In spite of considerable effort invested in the design and synthesis of high-performance semiconducting materials addressing the requirements for BHJ photovoltaic devices (spanning solution -processability, wide ranging optical spectra and large absorption coefficients, suitable e nergy band structure, efficient charge transport, and environmental stability) the access to conjugated polymers combining substantial charge -carrier mobilities, with a broad absorption spectrum extending across the visible and into the near IR ha s rem ained elusive to date. An excellent illustration of the complexity to address the two crucial parameters mentioned above within the same material is represented by the thieno[3,2 b ]thiophene -based all -donor polymer analogues developed by McCulloch et al. ( namely poly(2,5 -bis(3 alkylthiophen 2 yl)thieno[3,2 b ]thiophenes)s shown in Figure 7 1a ), which demonstrate up to 0.6 cm2 V1s1 of hole mobility (on/off ratio ~107) in top -gate FETs when large crystalline domains are formed upon thermal treatment ,17 but surprisingly, exhibit a limited PCE of 2.3 % in BHJ solar cells when blended with PC70BM.191 I n this instance, a moderate hole mobility of 3.8 x 10 cm2 V s was measured in the direction normal to the substrate by space charge limited current modeling for the best polymer:PCBM device composition (1:4), hence hindering the incorporation of an active layer thicker than 115 nm as desired to absorb light effectively and achieve especially efficient solar cell devices. Such differences between in -plane charge -carrier mobility and that measured in a vertical device geometry is commonly attributed to a high extent of edge -on molecular alignment which re duces the dimensionality of the charge transport through the active layer in highly crystalline semiconducting polymers. Further the relatively

PAGE 217

217 high HOMO of the polymer ( 5.1 eV relative to vacuum) inherent to most all -thienylene conjugated polymers minim izes the device VOC, which further limits the PCE. Finally, a nother important contributing factor consists in the lack of spectral coverage as the onset o f absorption of the polymer lies in the 650 680 nm range while its absorption maximum is on the ord er of 540 nm. As a result and as shown in Figure 7 1b, the measured EQE remains low at higher wavelengths (beyond 600 nm) where the photon flux is maximum (see solar photon flux under AM 1.5 represented in Chapter 1) a) b) Figure 7 1 a) The thieno[3, 2 b ]thiophene -based all donor polymer analogues developed by McCulloch et al. (namely poly(2,5 -bis(3 alkylthiophen 2 yl)thieno[3,2 b ]thiophenes)s demonstrate up to 0.6 cm2 V1s1 of hole mobility (on/off ratio ~107) in FETs17 b) IV curves of an optimized polymer:PC70BM BHJ device in the dark (black) and under 1 sun illumination (blue). The PCE remains limited to 2.34% (JSC=9.37 mA cm2, VOC=0.525 V and FF=0.48) (Adapted with permission from Ref. 191 Copyright 2008 American Institute of Physics ) In this context some of the most successful results have been obtained and described by Li et al. throughout their work on poly(thienyl ene vinylene) polymers (PTVs) possessing conjugated side -chains, and cross linked polythiophenes (PTs) with conjugated bridges ( see Figure 7 2) .192

PAGE 218

218 a) b) Figure 7 2. a) Series of PTs and PTVs as designed by Li et al. aimed at exhibiting the substantial charge transport properties of P3HT, while broadening the polymer optical spectrum over the UV visible region via the i ncorporation of conjugated side -chains and unsubstituted thiophenes (see polymers 1 12) or via cross -linking with conjugated bridges b) Some representative corresponding thin -film absorption spectra (Adapted with permission from Ref.192 Copyright 2008 WILEY -VCH Verlag GmbH & Co. KGaA, Weinheim)

PAGE 219

219 Considering the 3 -hexyl substituted derivative of polythiophene (P3HT) as a benchmark in terms of performance for polymer solar cell applications49,94,95 ( exceeding 5 % of PCE reproducibly under AM 1.5, 100 mW cm2 solar illumination ),41 Li and coworkers designed a series of PTs aimed at exhibiting the substantial charge transport properties of P3HT while broadening the polyme r optical spectrum over the UV -visible region via the incorporation of conjugated side -chains and unsubsti tuted thiophenes (see polymers 1 12, Figure 7 2a) In this case, the conjugated pendants (phenylene vinylene or thienylene vinylene side -groups) introduced a secondary chromophore absorbing in th e short -wavelength region of the spectrum (e.g. 300 400 nm) and in broken conjugation with the polymer main -chain absorbing at longer wavelength s (e.g. 400 700 nm). Unsubstituted thiophenes commonly planarize the polymer main -chain, extending the conj ugation, and induc ing a bathochromic shift in the visible (as discussed in the previous Chapter s of this manuscript) desired to overcome the mis match with the solar spectrum As illustrated in Figure 7 2b, replacing the bi(phenylenevinylene) side -chains by the bi(thienylenevinylene) analogues improved the visible absorption by inducing coalescence of both the short and long-wavelength absorption bands, hence producing a broad absorption plateau extending from 350 to 650 nm (see polymer 8, Figure 7 2a) Fur thermore, the HOMO of the corresponding PTs was found to be ca. 0.2 eV deeper than that of P3HT (since all heterocycles are not substituted in their 3 -position by an aliphatic side -chains, relatively electron donating in those systems) which is expected t o increase the VOC in BHJ with PCBM in solar cell devices (see Chapter s 1, 2 and 7) In the case of polymer 8, PCEs as high as 3.18% were measured under AM 1.5 solar illumination which represented a 38% increase compared to that of P3HT following the same experimental conditions, and in devices made by the same group.193 A different polymer series was investigated which consisted of PTV derivatives with

PAGE 220

220 bi(thienylenevinylene) conjugated side -chains (see polymers 14 16, Figure 7 2a). In particular, polymer 16 exhibited a remarkably broad and homogeneous absorption spectrum extending from 350 to 780 nm which was unprecedented for an all -organic -conjugated semiconducting polymer. However, surprisingly, this polymer series showed very limited performance in solar cells (with PCEs < 0.5%). With the hope of enhancing the charge transport in the most promising PT analogues designed, small fractions of comonomers containing conjugated bridges (typically < 10%) were subsequently incorporated in to the polymer backbones, which produced moderately cross -linked materials (see polymers 1725, Figure 7 2a) with hole mobilities as high as 7 x 103 cm2 V1 s1 (for polymer 19, Figure 7 2a) by SCLC modeling. It is expected that conjugated bridges reduce the interchain distances between molecules and favor the charge hopping mechanism throughout the active layer in devices. However, it was also found that excess cross linking reduced the solubility of the resulting polymers, while causing a decrease in hole mobility likely due to the higher degree of backbone distortion.194 As p reviously described in this manuscript, an attractive alternative to all-donor semiconducting polymers for photovoltaic applications consists in the incorporation of electron deficient building blocks alo ng the backbone, hence shifting the polymer absorp tion towards higher wavelengths ( i.e. 500 800 nm ) where the solar photon flux is the most intense (peak photon intensity at ca. 700 nm) While this goal as widely motivated the development of donor acceptor -conjugated polymers with various electron -do nating and electron accepting units, the vast majority of them have shown a relativel y narrow spectral absorption, along with limited charge -carrier mobilities.11,192 As a result, donor acceptor semiconducting polymers have almost always failed to compete with the most efficient all thienylene counterparts, including P3HT. An exception to this is represented by the s trictly alternating co polymer composed of di (2 -

PAGE 221

221 ethylhexyl ) -substituted cyclopentadithiophene (CPDT) and 2,1,3 benzothiadiazole (BTD) initially developed by Brabec et al. (see Figure 7 3a ) which are proving especially effective in photovoltaic devices wit h PCEs exceeding 5%, owing to their high charge -carrier mobilities (as high as 2 x 102 cm2 V1 s1 in bottom -gate FETs) and their long-wavelength absorption band extending into the near -IR (see plain black curve in Figure 7 3b) .44,70 a) b) Figure 7 3 a) (2 ethylhexyl) -substituted cyclopentad ithiophene (CPDT) and 2,1,3benzothiadiazole (BTD) based DA copolymer as initially reported by Brabec et al. are proving especially effective in BHJs, with PCEs exceeding 5%, owing to their high charge -carrier mobilities (as high as 2 x 102 cm2 V1 s1 in bottom -gate FETs) and b) their long -wavelength absorption band extending into the near IR(Adapted with permission from Ref.44,70 Copyright 2006 WILEY -VCH Verlag GmbH & Co. KGaA, Weinheim) Nonetheless, in contrast with the recent results reported by Li and coworkers on PTs with precisely engine ered repeat unit patterns, no clear design rules have been proposed to date for the synthesis of donor acceptor semiconducting polymers absorbing across the entire visible spectrum and into the near IR, while meeting the charge transport properties favorab le when in BHJs with PCBM. Given the results described in Chapter 4 of this manuscript, whereby solutionprocessable DA systems absorbing over the whole visible spectrum were designed with perspectives in electrochromic display applications (not limited by charge transport), the next step forward would consist in probing the possibilities of combining these unconventional optical properties with substantial hole mobilities within the same semiconducting polymers.

PAGE 222

222 7.2 Propose d Design and Rational In order to produce -conjugated semiconducting polymers possessing the wide spectral coverage attained via the methodology described in Chapter 4 of this manuscript along with large charge -carrier mobilities in thin -film devices a n original synthetic design in spired by recent work from Marks et al. on silole -based copolymers with high-performance in FETs was developed. Throughout their work on high -mobility systems for FET devices, Marks et al. have extended the use of silicon-deri vatized organic molecules to a series of conjugated polymers (see Figure 7 4 ) based on dialkyl -substituted dithienosilol e (DTS) and dibenzosilole (DBS) (see Figure 7 2 ), two fused aromatic systems bridged by a silicon atom (tetravalent) bearing the solubilizing groups.178,179 Interestingly, in silole type systems the exocyclic Si C -orbital mixes with the -orbital of the adjacent diene, hence lowering the LUMO of the resulting heterocycle when compared to that of cyclopentadiene .195,196 In th e all donor polymers described by Marks et al. the Si involvement in the -conjugation was revealed by the presence of bathochromic shifts in the optical absorption of the silole derivatives relative to their control all carbon polymer analogues.178 The perspective of enhancing the -electron overlap of a conjugated backbone via the presence of electron -withdrawing silole -based arenes, thereby increasing the electronic delocalization, has motivated the same group to investigate the charge transport properties of the polymers describe d in Figure 7 5 In top -contact FETs hole mobilities as high as 0.08 cm2 V1s1 (on/off ratio 5 x 105) were found for the DTS bithiophene analogue (m=2) in ambient conditions and for unaligned films. As illustrated in Table 7 1, the hole mobility increas ed with increasing number of unsubstituted thiophene in the backbone.

PAGE 223

223 Figure 7 4 Semiconducting silole based all -donor -conjugated polymers developed by Marks et al. for use in p -channel FETs Figure 7 5 Silole, Dithienosilole (DTS), Dibenzosilole (DBS) and their respective all -carbon counterparts Cyclopentadiene (Cp), Cyclopentadithiophene (CPDT) and Fluorene (F) Ta ble 7 1. Weight average molecular weight ( Mw, kDa), polydispersity (PDI), number average molecular weight ( Mn, kDa, calculated from Mw and PDI), film optical bandgap (Eg, eV), -stacking ( nm) and chain to -chain distances ( d nm), FET charge carrier mobi lities at saturation (sat, cm2 V1 s1), and current on/off ratios ( Ion: Ioff) for the all donor copolymers developed by Marks et al. (Adapted with permission from Ref.179 Copyright 2009 American Chemical Society) P olymer Mw (PDI) Mn Eg d sat Ion: Ioff TS6 26 (2.9) 9 1.8 TS6T1 30 (2.9) 10.3 1.8 (1. 5 5 4 ) 5 x 102 1 x 105 TS6T2 41 (3.0) 13.6 1.9 (1. 538) 8 x 10 2 5 x 10 4 BS8 32 (3.4) 9.4 2.9 4.0 4.4 (1. 776) BS8T1 112 (3.1) 36.1 2.5 4.0 4.4 (2.091) 6 x 10 5 3 x 10 4 BS8T 2 127 (3.7) 34.3 2.3 4.0 4.4 (1.870) 6 x 103 4 x 106 F8T1 17 (2.6) 6.5 2.5 (1.717) 9 x 10 5 2 x 10 5 F8T2 80 (3.3) 24.2 2.4 (1.554) 6 x 10 3 2 x 10 5 As highlighted above and in Chapter 1 conjugated polymers based on fused and coplanar heterocycl ic building blocks such as CPDT and DTS show several advantages over their more conformationally flexible analogues composed of s ingle aromatic rings (non-fused) such as enhanced packing ability pronounced macroscopic ordering and superior charge -carrier mobilities in thin -film devices Beside the increase in backbone rigidity, the coplanarity of fused

PAGE 224

224 aromatic systems promotes close -stacking interac tions between chains, which have been shown ( by 2D WAXS ) to arrange in lamellar superstructures.73,74 By analogy with the synthetic approach used in Chapter 4 to produce polymeric systems wit h broad spectral absorption over the visible, a polymer repeat unit can be envisaged in which DTS building blocks are employed in combination with unsubstituted thiophenes to space the electrondeficient unit BTD along the polymer main -chain. The structure s proposed in Figure 7 6 illustrate how a polymer series incorporating an increasing number of electronrich and coplanar substituents along the backbone, hence reducing the relative concentration of electron-deficient units, can be designed in order to pr ovide insight into how careful structural modifications can be employed to tune the spectrum and charge -carrier mobilities in -conjugated semiconducting polymers simultaneously. Figure 7 6 Proposed silole -based donor acceptor -conjugated semiconducting polymers combining electron rich DTS moieties and unsubstituted thiophene spacers with the electron -deficient BTD BTD is i ntroduced in the all donor backbones developed by Marks et al. to produce broadly absorbing materials with absorption maxima in the red portion of the visible and efficient charge transport properties as desired for photovoltaic applications

PAGE 225

225 7.3 Synthesis and Characterization of Dithienosilole 2,1,3 -Benzothiadiazole Polymers 7.3.1 Synthesis and Physical Characterization The utility of the donor acceptor approach in the design of solution-processable electrochromic and photovoltaic polymer s has been demonstrated throughout the previous Chapter s of this dissertation (see Chapter s 3 6 ). Based on the considerations and experimental results highlighted throughout Chapter 6 building blocks offe ring perspectives for limiting the steric hindrance (mainly related to the nature, configuration and concentration of the solubilizing side -chains) hence inducing more planarity in the subsequent polymer backbones, were prepared with the purpose of providing the materials with enhanced charge t ransport properties. Figure 7 7 illustrates the synthetic access to the DA copolymers P1 P 4 The unsubstituted 2,2 -bithiophene 1 was first subj ected to bromination conditions carried out at elevated temperature in o rder to functionalize the positions 3,3 and 5,5 quantitatively. Selective debromination of the positions 5,5 was subsequently achieved in acidic conditions employing zinc powder. The resulting intermediate 3 was then lithiated and ring -closed with dich lorodioctylsilane to yield species 4 in 40 60% yield. The trimethyltin derivative 5 was obtained upon reacting a monolithiated DTS block with trimethyltinchloride. T he electron accepting building unit BTD was then symmetrically substituted with two elec tron rich DTS stannane species (5 ) via Pd -mediated cross -coupling Stille conditions111 in gently refluxing THF to yield oligomer 8 Column chromato graphy over silica using mixtures of hexanes and dichloromethane (12:1) afforded 8 which was isolated as a purple tacky solid. Alternatively, 4 was brominated, subsequently lithiated and stannylated with ca. 2 equivalents of trimethyltinchloride in order t o be further copolymerized with BTD, affording a control polymer, namely P1 8 was either homopolymerized using the mild oxidizing agent FeCl3 (room temperature oxidative polymerization) and the resulting polymer subsequently reduced with

PAGE 226

226 hydrazine to yiel d P2 or was brominated ( at the external positions only) and copolymer ized following modified Stille conditions with 2,5 bis(trimethylstannyl)thiophene to yield copolymer P3 or with 5,5' bis(trimethylstannyl) 2,2' -bithiophene to yield copolymer P4 All pol ymers were precipitated in methanol and purified by Soxhlet extraction with methanol before further characterization and device fabrication. It is worth noting that the repeat unit structure of P2 necessitated using an oxidative polymerization route as opp osed to the Pd -mediated Stille type polymerization route employed to synthesize P1 -P3 Our attempts to produce P2 using Ni mediated Yamamoto conditions le d to rapid precipitation of an insoluble material, likely crosslinked via competitive insertion of Ni at the silole center. Figure 7 7. Synthetic route to silole -based donor acceptor conjugated copolymers P1 -P4 (Adapted with permission from Ref.102 Copyright 2009 American Chemical Society) In analogy with the approach described in Chapter 6 of this dissertation to enhance the charge transport properties in DA polymers, in P 3 and P 4 the DTD s symmetrically functionalized onto BTD and bearing the solubilizing side -chains are spaced by a small

PAGE 227

227 conjugated block composed of one or two thienylene heterocycles possessi ng no substituent Similarly it can be anticipated that reducing the concentration of aliphatic side -groups along the main -chain to the necessary extent, as in P 3 and P 4 should diminish the chainto -chain distances, favor the -stacking interactions and, in turn, improve the charge transport properties of the corresponding materials.177179 The structures of the repeat unit of P1 -P 4 were supported by 1H NMR (see Experimental Section ) and the quality of the polymers was probed by elemental analysis (see Table 7 2 ). The polymer molecular weights were determined via polystyrene -calibrated GPC with THF as the mobile phase and are summa rized in Table 7 3 Number average molecular weights ( Mn) ranging from 1 0 8 00 g mol1 (P1 ) to 1 9, 8 00 g mol1 ( P 4 ) were estimated for the polymers. It is worth noting that the reported molecular weights were estimated from non -fractionated polymers hence t he relatively large molecular weight polydispersity observed in general which could ultimately me reduced to refine the physical properties and performance in electronic devices of these materials. Importantly, the minimum average number of repeat units f or all the polymers was found to be 15 (P1 ), corresponding to a main -chain of nearly 57 aromatic units, which is about 4 times as high as the value for which the electronic properties of DA conjugated polymers are commonly found to saturate (~15 aromatic r ings).112 On the other hand, the average number of repeat units for P4 was found to be 1 7 corresponding to a main -chain of nearly 119 aromatic units which demonstrated the possibility of producing high molecular wei ght polymers from the polycondensation of relatively large functional building blocks (as opposed to a conventional copolymerization of single heterocycles) Thermogravimetric analysis of P1 P 4 showed onl y negligible weight loss below ca. 420oC (under nitr ogen atmosphere), hence demonst rating their remarkable thermal stability. Only P2 which possesses two DTSs connected to each other as a

PAGE 228

228 spacer for the BTDs shows a relatively small shoulder of decomposition at ca. 312 oC, hence pointing towards a potential ly reduced resistance to harsh thermal treatments. Table 7 2. Elemental Analysis (C, H, N) as calculated (left) and experimentally found (right) by Atlantic Microlab, Inc. for the donor acceptor copolymers P1 -P4 Polymer EA (Calcd/Found) C H N P1 65.40/64.89 6.95/6.99 5.08/4.67 P2 67.02/66.61 7.71/7.52 2.89/2.91 P3 66.36/66.81 7.30/7.45 2.67/2.51 P4 65.79/66.13 6.95/7.19 2.44/2.47 Table 7 3. Number average molecular weight ( Mn, g mol1), Weight average molecular weight (Mw, g mol1), Polydispersity i ndex (PDI), Average number of repeat units, Average number of rings, and Onset of decomposition temperature (as measured by TGA under nitrogen atmosphere) for the donor acceptor copolymers P1 -P4 Polymer Mn (g mol1) Mw (g mol1) PDI Avg. No. of Repeat Uni ts Avg. No. of Rings Td (oC)b P1 10,800 34, 2 00 3.2 19 57 422 P2 17,700 59, 2 00 3.3 18 90 (312) 429 a P3 16,000 60,100 3.8 15 90 424 P4 19,800 64, 1 00 3.2 17 119 426 a main decomposition onset 7.3.2 Polymer Optical Charac terization While P2 P4 all exhibit a two -band absorption in the visible spectrum, their short wavelength transition strongly overlap with the long-wavelength transition, and the first one differs in relative intensity with the second one among the polymer series (see Figure 7 8) In toluene solution, P1 shows a first local a bsorption maximum in the UV (392 nm) and at a distance of ca. 185 nm from the second absorption maximum ( 577 nm). In contrast, in the case of P2 -P 4 the first local absorption maximum h as been shifted to the visible (458, 451 and 494 nm, respectively) and greatly overlap with the second absorption maximum (627, 602 and 620 nm, respectively ) Table 7 4 summarizes the local absorption maxima for P1 P4 in toluene and as cast. It is worth noting that P4 possesses the more extended and homogeneous absorption

PAGE 229

229 spectrum in the polymer series (when normalized to the longer wavelength absorption maximum) and is essentially dark ink -blue to black in solution Similar considerations arise in the so lid state where the spectra undergo a bathochromic shift of 1020 nm in average According to the discussions lead throughout the previous Chapter s of this dissertation (see Chapter s 3 4 ), the low energy transition in these spectra can be attributed to the intramolecular donor acceptor interaction arising from the covalent bonding of electron -donating ( DTSs ) and electron accepting (BTDs) aromatic units, the higher energy transition can be assigned to the electron -donating contribution of the alternating pol ymer backbone ( i.e. the all -thienylene segments) The optically determined energy gaps were all relatively low and in the same range (1.511.56 eV). Figure 7 8 Normalized solution optical absorption (in Toluene) for the DTS BTD copolymers P1 P4 (Adapt ed with permission from Ref.102 Copyright 2009 American Chemical Society) Table 7 4. Local absorption maxima (solution in toluene and solid state) and film optical bandgap ( Eg, eV) for the donor acceptor copolymers P1 -P4 Polymer abs (nm) In Toluene abs (nm) Thin Film E g (V) 1 2 1 2 P1 392 577 404 583 1.53 P2 458 627 459 643 1.53 P3 451 602 4 64 611 1.56 P4 ~480 620 4 83 6 45 1.51

PAGE 230

230 7.3.3 Polymer Electrochemical Characterization and Band Structure Cyclic voltammetry (CV) and differential pulse voltammetry (DPV) were used to investigate the redox properties of P1 P4 including determining how polymer oxidation and reduction potentials, associated HOMO and LUMO energy levels, and electrochemical bandgaps vary as the acceptor (BTD) is increasingly spaced with electron rich thienyl units along the polymer main -chain Table 7 5 summarizes these results, along with the optical bandgap of the polymers as estimated from the onset of their low energy tr ansition. Polymer t hin films were drop -cast from toluene solutions ( 6 mg mL1) onto platinum button electrodes and characterized in an electrochemical cell comprising a platinum counter -electrode and a silver wire as the reference electrode, with the said electrochemical cell set in an argon -filled dry box. All estimated potentials were subsequently calibrated versus Fc/Fc+. The polymer films were systematically cycled prior to characterization, until a stable and reproducible redox response was reached. Ta ble 7 5. Electrochemically determined HOMO and LUMO energy levels (by CV and by DPV), Corresponding electrochemical bandgaps, and Comparison with their optically estimated values for the copolymers P1 P4 Polymer Eoxonset (V) HOMO (eV) Eredonset (V) LUMO (e V) Egap (V) Egap (V) CV DPV CV DPV CV DPV CV DPV CV DPV Optical P1 0.44 0.39 5.54 5.49 1.58 1.38 3.52 3.72 2.02 1.77 1.53 P2 0.39 0.32 5.49 5.42 1.68 1.48 3.42 3.62 2.07 1.80 1.53 P3 0.34 0.33 5.44 5.43 1.63 1.45 3.47 3.65 1.97 1.78 1.56 P4 0.42 0.45 5.52 5.55 1.62 1.46 3.48 3.64 2.04 1.91 1.51 aoxidation ( Eoxonset ) and reduction ( Eredonset ) potentials are reported vs. Fc/Fc+. benergy level derived from the electrochemical data ( Eoxonset and Eredonset ) considering that the SCE is 4.7 eV vs. vacuum and Fc/Fc+ is 0.38 eV vs. SCE, i.e. 5.1 eV relative to vacuum. cenergy level derived from the DPV estimated HOMO value of the polymer and its optical bandgap. P1 revealed a moderate oxidation potential of +0.44 V by CV and +0.39 V by DPV which corre sponded to a relatively low -lying HOMO energy leve l of 5.54 eV relative to vacuum by CV and 5.49 eV by DPV The onset of reduction was found at 1.58 V by CV and 1.38 V by DPV,

PAGE 231

231 which corresponded to a LUMO of 3.52 eV by CV and 3.72 eV by DPV, respectively and to electrochemical bandgaps of 2.0 2 and 1.77 eV. As mentioned in the previous Chapter s of this dissertation (see Chapter s 5 and 6), DPV tends to provide more accurate estimates when evaluating the redox potentials of -conjugated polymers, by inducing sharper redox onsets than CV in particular.128,181 In the case of P1 the differences between the CV and DPV estimated values highlight well the care that should be taken when limiting an electrochemical study of a conjugated polymer energy band structure to the use of cyclic voltammetry. P2 and P3 revealed slightly lower oxidation potential s than that of P1 +0.32 V and +0. 33 V respectively by DPV, an d thus a highe r -lying HOMO, 5.42 eV and 5.43 eV ( respectively ) from DPV, in agreement with the introduction of thienylene building units increasing the electron -rich contribution in the DA backbone. The onset s estimated from the first reductive process (the second one i nducing rapid degradation of the polymer on repeated cycling) 1. 48 V and 1.4 5 V respectively by DPV, were consistent with the existing literature relating to BTD containing DA copolymers.44,45,180 From the associated LUMOs, 3.62 eV and 3.65 eV (respectively) by DPV, energy gaps of 1.80 and 1.78 eV (respectively) from DPV were determined in agreement with that found for P1 (1.77 eV) In contrast with P2 the incorporation of a bithiophene conjugated spacer in P 4 repressed the polymer HOMO by ca. 0.1 eV (5.55 eV by DPV) which drop was not as significant as that observed al ong the polymer series discussed in Chapter 6 Similarly here, t his r esult can be attributed to the less pronounced electron rich character of thiophene when compared t o that of DTS In comparison with P 2 and P 3 polymer P 4 revealed a higher oxidation pote ntial (+0. 42 V by CV and +0. 45 V by DPV), while retaining a near -identical onset of reduction ( 1.6 2 V by CV and 1.4 6 V by DPV).

PAGE 232

232 In general, the electrochemical bandgaps were found slightly smaller when estimated by DPV (by up to ca. 0.3 eV lower than tho se determined by CV) DPV tending to provide more accurate estimates when evaluating the redox potentials of -conjugated polymers Although the electrochemically determined bandgaps were systematically larger than the optically estimated values, this expe rimental observation is in good agreement with work on DA polymers from various groups.46,112,182 Given the above discussion, the energy dia gram presented in Figure 7 5 was constructed by employing the CV and DPV estimated energy levels of the polymers. In short, Figure 7 9 highlights the appropriate match between the energy band structure of the designed polymers and the key energy levels of the various components utilized in the fabrication of BHJ photovoltaic test -devices in Chapter 6 for instance ( e.g. ITO, MoO3, PC70BM and Al).44,95 Figure 7 9 Schematic representa tion of the energyband structure of P1 -P4 showing the polymer HOMO and LUMO energies as estimated by CV and DPV (green filled rectangles) with respect to those of an ideal polymer ( i.e. designed to be integrated in BHJ solar cells employing PC70BM and P EDOT PSS). The optically determined bandgaps of P1 P4 were placed at the baricenter of the DPV estimated bandgaps, and a second approximated set of HOMO and LUMO levels could be defined assuming the energy levels equidistant from the baricenter (in green dotted rectangle)

PAGE 233

233 7.3.4 Polymer Structural Analysis The microstructural organization of P1 P4 was investigated by 2 D w ide a ngle X -ray s cattering (2D WAXS, see Figure 7 10), whereby the polymer samples consisted of fibers pr epared by mechanical extrusion (as described in Chapter 2) .73,74,101 The X -ray data including -stacking ( nm) and chain -to chain distances ( d nm), are summarized in Table 7 6 Figure 7 10. 2D wide angle X ray scattering (2D -WAXS) of P1 P4 whereby the polymer samples consisted of fibers prep ared by mechanical extrusion (Adapted with permission from Ref.102 Copyright 2009 American Chemical Society) From the X ray data presented as Figure 7 10, P1 and P2 were found to organize in lamellae structures with an identical chain to -chain distance of 1.95 nm calculated from the sma ll angle equatorial reflections. In the meantime, P1 did not show any evidence for the presence of -stacking interactions as supported by the isotropic small angle reflection and the lack of corresponding scattering intensity in its X ray pattern. In contrast the pattern of P2 showed wide angle reflections in the equatorial plane corresponding to -stacking distances of 0.36 nm which pointed towards increased microstructural order. Similarly to P1 the reflections

PAGE 234

234 in the X -ray pattern of copolymer P3 did not indicate the presence of a pronounced degree of order, though a lamellar spacing of 1.94 nm could be calculated from the small angle equatorial reflections In contrast, a higher degree of crystallinity and alignment of the lamellae structures a long the fiber direction were demonstrated by t he distinct scattering intensities (wide angle reflections) in the pattern of P4 In particular, a chain to -chain spacing of 1.84 nm and -stacking distances of 0.36 nm were calculated in the case of P4 which represent some of the closest intermolecular interactions reported to date for a solution processable -conjugated polymer.17,197 Table 7 6 -stacking ( nm) and chain to -chain distances ( d nm) for the donor acceptor copolymers P1 -P4 P olymer nm d (nm) P1 a 1.95 P2 0.36 1.95 P3 -a 1.94 P4 0.36 1.84 a not observed 7.3.5 Field Effect Transistors The FET performance data for the donor acceptor copolymers P1 -P4 in bottom gate / bottom contact devices a re summarized in Table 7 7 The corresponding FETs were obtained by drop -casting the copolymers on HMDS treated SiO2 covering heavily doped Si (as the gat e electrode) from a 2 mg mL1 chlorobenzene solution. In agreement with their differences in microstructural organization polymer P1 did not reveal any transistor behavior on HMDS treated SiO2, whereas P2 showed a substantial field -effect with a moderate hole mobility of 3 x 104 cm2 V1s1 at saturation and an on/off current ratio of 6 x 103. B y substituting HMDS with phenyltriethoxysilane (PTES), a transistor behavior was observed for P1 which revealed a very limited hole mobility of 2 x 106 cm2 V1s1 at saturation and a low on/off current ratio of 1 x 102.

PAGE 235

235 Table 7 7. Number average molecular weight ( Mn, kDa), polydispersity (PDI), film optical bandgap (Eg, eV), -stacking ( nm) and chain to chain distances ( d nm), FET charge carrier mobilities at saturation (sat, cm2 V1 s1), and current on/off ratios (Ion: Ioff) for the donor acceptor copolymers P1 -P4 (Adapted with permission from Ref.102 Copyright 2009 American Chemical Society) P olymer nm d (nm) sat Ion: Ioff P1 a 1.95 2 x 10 6 1 x 10 2 P2 0.36 1.95 3 x 10 4 6 x 10 3 P3 a 1.94 3 x 10 3 2 x 10 4 P4 0.36 1.84 2 x 10 2 1 x 10 3 a not observed Figure 7 11. Field -effect transistor characteristics of P4 : a) output curves take n at different gate voltages and b) transfer curves (Adapted with permission from Ref.102 Copyright 2009 American Chemical Society) In spite of its lack of microstructural organization, the copolymer analogue P3 showed enhanced FET performance when compared to that of P1 and P2 w ith a hole mobility of 3 x 103 cm2 V1s1 at saturation and an on/off current ratio of 2 x 104. This result supported the idea that unsubstituted heterocyclic units can be inserted in the backbones of -conjugated polymers to enhance their charge -carrier mobilities effectively .178,179 In correlation with the increase o f order and molecular packing observed for the copolymer analogue P4 an enhanced hole mobility of

PAGE 236

236 0.02 cm2 V1s1 at saturation and an on/off current ratio of 1 x 103 were achieved as illustrated in Figure 7 11. T he especially close intermolecular distances and notable FET performance enhancement in the case of P4 can be attributed to the presence of unsubstituted bithiophene spacers creating effective hopping sites for the charge -carriers, reducing the concentration of solubilizing side -chains, while inducing an increase in backbone pl anarity. In order to explain the two orders magnitude difference in mobility between P2 and P4 in spite of the close -stacking distances found in both cases, the integration plots of the equatorial scattering intensities for P2 and P4 were extracted from the 2D -WAXS patterns of the polymer fibers prepared by mechanical extrusion. While t he relatively broad reflections and the lack of higher order scattering intensities in the corresponding 2D pattern s pointed towards a lower degree of crystallinity in the case of P2 the differences in scattering intensity illustrated in their integration plots in Figure 7 12 confirmed the pronounced variation in macroscopic order from P2 to P4 T he higher degree of solid -state alignment in P4 is further indicated by higher order reflections (see (200) in Figure 7 12) for the chain to -chain correlation, while P2 shows only a unique and broad reflection. Figure 7 12. Integration plots of the equatorial scattering intensities for P2 and P4 as determined from the 2D WAXS pa tterns of the polymer fibers prepared by mechanical extrusion

PAGE 237

237 Importantly, as illustrated in Figure 7 13, the same trend was observed by X ray diffract ion in thin film (the polymers were drop -cast onto HMDS treated silicon) with P4 showing the highest degree of crystallinity, while the X ray diffractogram of P2 revealed only a weak reflection indicative of a low degree of macroscopic order. As expected, i dentical d -spac ing distances were determined for the thin films an d from the bulk, suggesting no variat ion in molecular packing upon thin film processing. Figure 7 13. X ray diffraction (XRD) of P1 P4 whereby the polymer samples consisted of thinfilms drop -cast onto HMDS treated silicon (Adapted with permission from Ref.102 Copyright 2009 American Chemical Society) 7.3.6 Photovoltaic Devices Taking into consideration the results obtained in Chapter 6 polymer P4 was processed into a BHJ device composed of PC70BM as the electron accepting material, along with MoO3 (ca. 10 nm) as the hole transpo rting interface layer and the resulting solar cell was tested under AM 1.5 solar illumination ( irradiation intensity: 100 mW cm2) T he active layer (ca. 120 nm) was spin cast from chlorobenzene and su bjected to a thermal treatment step (16 0oC, 10 min) be fore Al

PAGE 238

238 contact deposition. Figure 7 14a shows the current density-voltage ( J V ) responses of the device employing P 4 at optimized polymer:PCBM composition (1:1 by weight as rapidly determined from a series of test -devices made with PC60BM ) At this early stage, the 1:1 polymer:PC70BM device revealed promising PV characteristics, with a n encouragingly high JSC of 7 86 mA cm2 supporting our synthetic strategy to enhancing the charge transport in DA conjugated polymers, but a moderate VOC of 0. 59 V and a m oderate FF of 44 %. Although higher VOC values should be targeted in organic BHJs as it directly relates to the device PCE this estimated value was in agreement with the HOMO of the polymer estimated electrochemically (5.5 5 eV relative to vacuum by DPV). Th e same device exhibited an overall PCE of 2 04 % which was reasonable considering the novelty of the material produced It is worth noting that at the 1:1 polymer:PCBM ratio employed to construct this device, the resulting cell was essentially black in col or as desired to absorb light with the most efficacy. Figure 7 1 4 b shows t he EQE data of P 4 In this case, the response consists of a broad-band extending across the entire visible spectrum with a long -wavelength local maximum of photocurrent of ca. 28% c orresponding to the absorption maximum of the long -wavelength absorption band of P 4 The short er -wavelengt h maximum of the photocurrent reaches ca. 43% and is dominated by the absorption of PC70BM added to that of the short -wavelength absorption band of P3 The onset of photocurrent at ca. 800 nm was in agreement with the polymer optical absorption dat a It is expected to see the PV performance of this polymer improve as a function of the processing conditions employed In addition, narrowing the polydisper sity of the corresponding polymer (here PDI was 3.2 in absence of fractionation) to select only the higher molecular weight fraction, along with optimizing the purification process according to the most recent methods developed to this effect, should widel y contribute to reaching the expected performance enhancement. Further

PAGE 239

239 synthesis, characterization and device fabrication (including the remaining polymers in the series, P1 P3 ) is underway. a) b) Figure 7 1 4 a) I V curves for a BHJ solar cell made of P 4 (at best polymer:PCBM composition, i.e 1:1) in the dark (blue curve) and under AM 1.5 solar illumination, 100 mW cm2, for t he device structure ITO/MoO3/ P 4 :PC70BM/LiF/Al ( red curve) b) EQE response (blue curve) of the polymer:PCBM blend of the same P4 -b ased device 7.3.7 Synthetic Details Compound 2 -4 : Details relative to the synthesis and characterization of 2 -4 can be found in Usta, H.; Lu, G.; Facchetti, A.; Marks, T. J. J. Am. Chem. Soc. 2006, 128, 90349035. Compound 5 : Compound 4 (1.644 g, 3.93 mmol) was dissolved in dry THF (120 mL) and cooled down to 78 C. A solution of nbutyllithium in hexanes (2.04 mL, 5.1 mmol ) was added over a 30 min period and the mixture was stirred for 3 hours at 78 C. A solution of tri methyltinchloride in hexanes (5.5 mL, 5.5 mmol ) was subsequently added, the mixture was allowed to warm up to room temperature and stirred for 16 hours. The solvent was evaporated affording a yellow -brown residue which was used for the next step without further purification (~95 %). 1H NMR (300 MHz, CDCl3) = 7.17 (d, J = 4.8 Hz, 1H), 7.09 (s, 1H), 7.03 (d, J = 4.8 Hz, 1H), 1.500.60 (m, 34H), 0.37 (s, 9H).

PAGE 240

240 Compound 6 : Compound 4 (1.542 g, 3.7 mmol) was dissolved in 100 mL of CHCl3. Then NBS (1.442 g, 8.1 m mol) was added by portions and the mixture was stirred for 24 h at room temperature under light protection. The organic layer was washed with water, dried over magnesium sulfate the solvent was evaporated and the product was purified by column chromatogra phy on silica with hexane as the eluent. T he solvent was evaporated and compound 6 was obtained as a light yellow viscous oil ( 1.974 g, 93 %). 1H NMR (300 MHz, CDCl3) = 6.99 (s, 2H), 1.500.80 (m, 34H), 0.37 (s, 18H).; 13C NMR (75MHz, CDCl3) = 149.15, 141.25, 132.38, 111.64, 33.30, 32.06, 29.40, 29.32, 24.23, 22.88, 14.33, 11.84. Compound 7 : Compound 6 (0.3 g, 0.52 mmol) was dissolved in dry THF (10 mL) and cooled down to 78 C. A solution of nbutyllithium in hexanes (0.499 mL, 1.25 mmol ) was added over a 10 min period and the mixture was stirred for 2 hours at 78 C. A solution of trimethyltinchloride in hexanes (1.35 mL, 1.35 mmol ) was subsequently added, the mixture was allowed to warm up to room temperature and stirred for 2 hours. The solvent was e vaporated affording a yellow -brown residue ( 7 ) which was used for the next step without further purification (~95 %). 1H NMR (300 MHz, CDCl3) = 7.08 (s, 2H), 1.50 0.60 (m, 34H), 0.37 (s, 18H). Compound 8 : Compound 5 (3.93 mmol), 4,7 -dibromobenzo[ c ][1,2,5] thiadiazole (0.525 g, 1.78 mmol) and Pd(PPh3)2Cl2 (4 mol %) were dissolved in 30 mL of dry THF. The mixture was stirred overnight at 60 C, the solvent was evaporated and the product was purified by column chromatography on silica with hexane/dichlorometha ne (12:1) as the eluent. T he solvent was evaporated and compound 8 was obtained as a dark purple viscous oil ( 1 g, 58 %). 1H NMR (300 MHz, CDCl3) = 8.11 (s, 2H), 7.85 (s, 2H), 7.28 (d, J = 4.8 Hz, 2H), 7.11 (d, J = 4.8 Hz, 2H), 1.500.80 (m, 68H); 13C NMR (75MHz, CDCl3) = 152.76, 150.92, 149.46, 143.20, 142.65,

PAGE 241

241 140.40, 130.29, 130.02, 126.19, 125.91, 125.23, 33.43, 32.09, 29.45, 29.41, 24.45, 22.88, 14.32, 12.16. HRMS (MMI TOF) [MH+] m/z calcd. for C54H76N2S5Si2: 969.4223 Found: 969.4197. Anal. calcd. for C54H74N2S5Si2: C 66.89, H 7.90, N 2.89 Found: C 66.94, H 7.95, N 2.84. Compound 9 : Compound 8 (0.963 g, 0.99 mmol) was dissolved in 100 mL of CHCl3. Then NBS (0.354 g, 1.99 mmol) was added by portions and the mixture was stirred for 24 h at room temperatu re under light protection. The organic layer was washed with water, dried over magnesium sulfate the solvent was evaporated and the product was purified by column chromatography on silica with hexane/dichloromethane (12:1) as the eluent. T he solvent was e vaporated and compound 9 was obtained as a dark purple tacky solid ( 0.98 g, 87.5 %). 1H NMR (300 MHz, CDCl3) = 8.09 (s, 2H), 7.82 (s, 2H), 7.05 (s, 2H), 1.50 0.80 (m, 68H); 13C NMR (75MHz, CDCl3) = 152.65, 150.26, 149.86, 142.89, 142.22, 140.78, 132.63, 130.14, 125.84, 125.25, 112.25, 33.40, 32.09, 29.45, 29.38, 24.37, 22.89, 14.33, 12.04. HRMS (MMI TOF) [MH+] m/z calcd. for C54H74Br2N2S5Si2: 1127.2419 Found: 1127.2360. Anal. calcd. for C54H74Br2N2S5Si2: C 57.52, H 6.62, N 2.48 Found: C 57.86, H 6.73, N 2.44. Copolymer P1: Compound 7 (3.93 mmol), 4,7 dibromobenzo[ c ][1,2,5]thiadiazole (0.525 g, 1.78 mmol), Pd2(dba)3 (2 mol %) and P( o tol)3 (8 mol %) were dissolved in 10 mL of dry toluene and 1 mL of degassed anhydrous DMF. The mixture was stirred for 3 days at 90 C. It was then precipitated into methanol (200 mL). The precipitate was filtered through a Soxhlet thimble and purified via Soxhlet extraction for 24 hours with methanol. The polymer was extracted with chloroform, concentrated by evaporation, and treated with a strongly complexing ligand ( diethylammonium diethyldithiocarbamate) in order to remove any residual trace of catalyst before being precipitated in methanol (200 mL) and collected as a dark purple solid (230 mg, 80 %). 1H NMR (300 MHz, CDCl3) = 8.20 (bs, 2H), 7.91 (bs, 2H), 1.500.80 (m, 34H)

PAGE 242

242 GPC analysis: Mn = 10 800 g mol1, Mw = 34 200 g mol1, PDI = 3.8 Anal. calcd. for C30H38N2S3Si1: C 65.40, H 6.95, N 5.08 Found: C 64.89, H 6.99, N 4.67. Copolymer P2: Compound 8 (415 mg, 0.428 mmol) w as dissolved in chloroform (50 mL). A solution of anhydrous FeCl3 (347 mg, 2.14 mmol, 5eq) in nitromethane was added dropwise over a period of 30 minutes to the stirred monomer at room temperature (the dark purple monomer solution turned progressively dark blue green with addition of oxidizing agent). The mixture was stirred overnight at room temperature. It was then precipitated into methanol (200 mL). The precipitate was filtered, redissolved in chloroform (200 mL) and stirred for 3 hours with hydrazine m onohydrate (6 mL). After evaporation, the concentrate (dark blue -green) was precipitated into methanol (200 mL), the precipitate was filtered through a Soxhlet thimble and purified via Soxhlet extraction for 24 hours with methanol. The polymer was extracte d with chloroform, concentrated by evaporation, the polymer was precipitated in methanol (200 mL) and collected as a dark solid (170 mg, 41 %). 1H NMR (300 MHz, CDCl3) = 8.14 (bs, 2H), 7.87 (bs, 2H), 7.00 (br, 2H), 1.500.80 (br, 68H) GPC analysis: Mn = 17 700 g mol1, Mw = 59 200 g mol1, PDI = 3.35 Anal. calcd. for C54H74N2S5Si2: C 67.02, H 7.71, N 2.89 Found: C 66.61, H 7.52, N 2.91. Copolymer P3: 2,5 dibromothiophene (0.101 g, 0.421 mmol) was dissolved in dry THF (10 mL) and cooled down to 78 C. A solution of nbutyllithium in hexanes (0.405 mL, 1.011 mmol ) was added over a 10 min period and the mixture was stirred for 1 hour at 78 C. A solution of trimethyltinchloride in hexanes (1.09 mL, 1.09 mmol ) was subsequently added, the mixture was allowed to warm up to room temperature and stirred for 1 hour. The solvent was evaporated affording a light orange residue which was used for the next step without further purification (~98 %). Compound 9 (0.475 g, 0.421 mmol), Pd2(dba)3 (2 mol %) and P( o tol)3 (8

PAGE 243

243 mol %) were added to the residue and dissolved in 12 mL of dry toluene and 1 mL of degassed anhydrous DMF. The mixture was stirred for overnight at 90 C. It was then precipitated into methanol (200 mL). The precipitate was filtered through a Soxhlet thi mble and purified via Soxhlet extraction for 24 hours with methanol. The polymer was extracted with chloroform, concentrated by evaporation, and treated with a strongly complexing ligand ( diethylammonium diethyldithiocarbamate) in order to remove any residual trace of catalyst before being precipitated in methanol (200 mL) and collected as a dark solid (180 mg, 41 %). 1H NMR (300 MHz, CDCl3) = 8.12 (bs, 2H), 7.84 (bs, 2H), 7.00 (br, 4H), 1.500.80 (m, 68H) GPC analysis: Mn = 15 900 g mol1, Mw = 60 100 g mol1, PDI = 3.8 Anal. calcd. for C58H76N2S6Si2: C 66.36, H 7.30, N 2.67 Found: C 66.81, H 7.45, N 2.51. Copolymer P4: 5,5' dibromo2,2' -bithiophene (0.141 g, 0.434 mmol) was dissolved in dry THF (10 mL) and cooled down to 78 C. A solution of n-butyllit hium in hexanes (0.416 mL, 1.04 mmol ) was added over a 10 min period and the mixture was stirred for 2 hours at 78 C. A solution of trimethyltinchloride in hexanes (1.13 mL, 1.13 mmol ) was subsequently added, the mixture was allowed to warm up to room te mperature and stirred for 2 hours. The solvent was evaporated affording a yellow orange residue which was used for the next step without further purification (~98 %). Compound 9 (0.489 g, 0.434 mmol), Pd2(dba)3 (2 mol %) and P( o tol)3 (8 mol %) were added to the residue and dissolved in 12 mL of dry toluene and 1 mL of degassed anhydrous DMF. The mixture was stirred for overnight at 90 C. It was then precipitated into methanol (200 mL). The precipitate was filtered through a Soxhlet thimble and purified vi a Soxhlet extraction for 24 hours with methanol. The polymer was extracted with chloroform, concentrated by evaporation, and treated with a strongly complexing ligand ( diethylammonium diethyldithiocarbamate) in order to remove any residual trace of catalys t before being

PAGE 244

244 precipitated in methanol (200 mL) and collected as a dark solid (250 mg, 51 %). 1H NMR (300 MHz, CDCl3) = 8.13 (bs, 2H), 7.85 (bs, 2H), 7.00 (br, 4H), 1.500.80 (m, 68H) GPC analysis: Mn = 19 800 g mol1, Mw = 64 100 g mol1, PDI = 3.2 An al. calcd. for C62H78N2S7Si2: C 65.79, H 6.95, N 2.44 Found: C 66.13, H 7.19, N 2.47. 7.4 Conclusions and Outlook In conclusion, a series of four DTS BTD copolymers differing by the concentration of electron -donating and -withdrawing substituents along the backbone have been synthesized and characterized by 2D -WAXS, and in bottom -gate/ bottom -contact FETs. While all copolymers were able to self assemble into lamellar superstructures, only P2 and P4 showed a propensity to -stack. The highest hole mobility of 0.02 cm2 V1s1 was observed for P4 in excellent agreement with the close -stacking and lamellar distances found by structural analysis (0.36 nm and 1.84 nm, respectively). It is expected to see the FET performance im prove with controlled polymer molecular weight and macroscopic order .74 In parallel, P4 absorbs homogeneously across t he entire visible spectrum as solar cell applications require. From the initial PV data obtained the need for a new series of DTS BTD semiconducting polymer derivatives possessing lower lying HOMO s, and thus providing light -harvesting devices with higher VOC values, can be anticipated. In combination with high current densities in absence of applied bias such as that observed with P4 (ca. 8 mA cm2), the subsequent materials could exceed the state -of the art performance of P3HT, while exhibiting the ambie nt longterm stability required for commercial application.

PAGE 245

245 CHAPTER 8 PERSPECTIVES AND OUT LOOK 8.1 -Conjugated Polymers in Context In an emerging era of flexible, rollable or foldable high performance electronic displays the perspective of manufacturing low -cost functional materials that can be easily processed over large areas along with being mechanically tolerant has become a sine qua non to commercial viability. In parallel, the idea of designing multifunctional semi conducting materials that can be integrated in various operating systems by inducing simple changes in device architectures has triggered new challenges for materials scientists who are now trying to answer the following: How can organic electronics reach or exceed the performance of their inorganic based counterparts for a specific technology? Should they complement existing technologies in target applications or should they seek to replace them in the most conventional areas of use? Can they remain cost e ffective regardless of the number of manufacturing steps? In portable applications where lightness, thickness and low power consumption are further desired, organic electronics that can be precisely printed, stamped, sprayed, drop -cast or spin coated into pre -defined patterns, offer a competitive alternative to their conventional inorganic homologues. However, solution-processability, solid -state ordering, efficient electroluminescence, stable color -switching and high charge -carrier mobility for instance, a re a variety of interconnected factors difficult to match up for the synthetic chemist, when not a priori exclusive. In the fast -growing area of plastic electronics, -conjugated organic polymers combining mechanical flexibility, and ease in bandgap/color tuning via structural control, along with the potential for low -cost scalability and processing, are particularly attractive. As mentioned throughout the introduction chapters of this dissertation (see Chapter s 1 and 2), in considering

PAGE 246

246 the most established applications, polymeric semiconductors have found widespread application in thin film transistors,5,19 photovoltaic cells,10,11,41 radio -frequency identification tags, sensor s,7,8 memories,9,198 and light -emitting diodes.6 8.2 The Donor Acceptor Approach in -Conjugated Polymers with Perspectives in Electrochromic Applications Considering the growing interest in solution-processable ECPs, an essential step forward is to develop a better understanding of their fundamental structur e property relationships. To date, few reports address the general characteristics of conjugated polymers suitable for use in EC applications. Published work often refers to the EC potential of a newly synthesized system without providing any structural re asoning that lead to the desired properties. Throughout this dissertation work, carefully designed sets of conjugated polymers were synthesized and characterized to draw parallels and trends between molecular structures and EC performance. As illustrated in Figure 8 1a, solution -processable neutral state green conjugated polymers possessing a two -band optical absorption in the visible spectrum were accessed based on a specific linear combination of electron rich dioxythiophene (DOT) and electron -deficient benzothiadiazole (BTD) heterocycles according to the DA approach. Optical and electrochemical investigation of the green conjugated polymers revealed highly transmissive oxidized states, excellent optical contrasts both in the visible and in the NIR, small potential windows of operation, fast switching times and long -term stability (see Chapter 3 ). While further investigating the influence of the content in electron-donating and electron accepting building units along the backbone of two analogous series of DOT -BTD copolymers, it was found that low and high energy transitions in two -band absorbing linear DA polymers can be controlled in a substantial and interconnected fashion by varying the relative contribution of electron rich and poor moieties incorporat ed in the repeat unit (see Chapter 4 ) In addition, a merging of the bands

PAGE 247

247 was observed that offers the potential for the synthesis of neutral state colored materials possessing either highly saturated or darker colors. a) b) c) Figure 8 1. Disserta tion Work Overview: Spectral Engineering in Cathodically-Coloring Conjugated Electrochromic Polymers As illustrated in Figure 8 1b, the approach developed for the synthesis of conj ugated polymers possessing a two-band absorption in the visible was e xpanded to the synthesis of a

PAGE 248

248 copolymer possessing an absorption spectr um extended over the entire visible region (400 700nm) T he copolymer was found to achieve reversible EC switching from a strong opaque neutral black color state to a highly transmissive oxidized state over a potential window of less than 1 V hence repr esenting the first neutral state black polymeric electrochrome with potential applications in e -papers and smart windows F in ally, as illustrated in Figure 8 1c a series of DOT BTD copolymers containing unsaturated spacers (either ethynylene or trans eth y lene) was electrochemically characterized and their electrochromic performance evaluated with the goal of simplifying the access to colors difficult to access such as green and cyan blue (see Chapter 5 ). Here, it was found that DA building blocks can be conveniently polymerized with trans -ethylene spacers to produce EC polymers with red -shifted optical spectra when compared to their all heterocyclic control parent polymers. Beyond the fine bandgap tuning afforded by this simple structural modification tran s -ethylene linkers -containing ECPs introduce a potential pathway toward the synthesis of polymeric electrochromes synthesized via the copolymerization of multiple chromophores to create new color palettes, including the color black such as desired for electronic ink -based applications. It is evident from the number of electroactive materials described in the literature14 that adding one more ECP candidate to the available library may not represent a substantial contribution aimed at moving the field ahead. If electrochemical polymerizations can be employed t o demonstrate a concept, the same approach suffers from important limitations spanning large -scale applicability, tedious purifications or the introduction of molecular defects, and importantly, the subsequent polymers do not always reach their optical sat uration threshold due to the lack of solubility of the growing backbone. In contrast, further research effort should now be directed towards designing solution -processable ECPs with improved optical contrasts,

PAGE 249

2 49 switching ability and optimized performance in devices, hence facilitating technology transfer, and promoting the perspectives for mass produced displays. In particular, the next generation of electrochromic polymers should intend to address a number of essential aspects including (1) the compromise between solution-processability and film to -substrate adhesion upon repeated switching, (2) the presence of at least one accessible highly transmissive state allowing the combination and separate redox control of several ECPs to generate multiple colors and shades otherwise difficult to achieve, (3) the polymer response time at various desirable film thicknesses, as well as (4) the long term environmental stability. These goals are not exclusive and will be attained by understanding the structure property re lationships governed by the careful structural modifications operated at the molecular level, and by proposing new methodologies for the study of systems with more complex energy band patterns / optical features. The correlation between optical density, sw itching performance, and film thickness -quality has been widely ignored so far in spite of the obvious interplay. Similarly, coloration efficiencies and color coordinates for the polymer electrochromes are not systematically reported, which complicates fur ther comparison between works. Correct use and implementation of the existing characterization toolbox for polymer electrochromism will be a key variable in the overall equation leading to widespread applications of this fascinating technology. 8.3 The Donor Acceptor Approach in -Conjugated Polymers with Perspectives in Photovoltaic Applications Relying on the concept of energy band broadening induced in a macromolecular conjugated system involving alternating electron rich and poor substituents,42 the donor acceptor approach has found its foremost application in the design and synthesis of low -bandgap

PAGE 250

250 photovoltaics absorbing effectively over the visible spectrum with the goal of improving the solar energy conver sion in bulk -heterojunction solar cells.10,11,41 As narrow -bandgap -conjugated polymers absorb at longer -wavelengths than their wider -bandgap all donor counterparts ( e.g. P3HT, MEH PPV and MDMO -PPV) they have the ability to harvest light over the range of wavelengths where the photon-flux is maximum (i.e. 500800 nm) However, absorption of sunlight is far from being the only parameter of importance and the charge transport taking place across the solar cell d evices remains the primary factor governing their performance. Then again, despite the continuous research effort committed to their design and synthesis, high -performance semiconducting polymers fulfilling the many requirements for use in photovoltaic devices (including solution -processability, wide ranging optical spectra and large absorption coefficients, suitable energy band structure, efficient charge transport, and environmental stability), the access to materials combining substantial charge transpo rt properties, with a broad absorption spectrum extending homogeneously across the visible and into the near IR, has remained a challenge to date. As illustrated in Figure 8 2a, solution-processable and neutral state green DOT BTD based conjugated polyme rs exhibiting a two band optical absorption in the visible spectrum were designed with an emphasis on correlating molecular structure, energy band characteristics and charge transport. In particular, their structure performance relationships in solar cell devices were carefully investigated and the subsequent device architectures optimized to reach power conversion efficiencies as high as ca. 3% (see Chapter 6). In analogy with Chapter 3, t he synthetic approach consisting in polycondensing oligomers with el ectron rich chain -ends using mild oxidative reagents produce d novel polymer repeat units otherwise synthetically demanding as well as high molecular weight polymers without the need for tedious monomer purifications

PAGE 251

251 and respect of the monomer to -comonomer stoichiometry Thereby, Chapter 6 introduced a new synthetic strategy taking bandgap/color -engineering along with charge transport into account for the production of s olution -processable organic electronics suitable for combined photovoltaic/display device s, power -generating ornaments (such as photovoltaic plants and trees) or light -har vesting window applications. a) b) Figure 8 2. Dissertation Work Overview: Spectral Engineering in -Conjugated Polymers with Substantial Charge Transport Properties for P hotovoltaic Applications As illustrated in Figure 8 2b, the approach developed in Chapter 4 to access polymer electrochromes possessing a tunable tow -band absorption in the visible, whereby the relative intensity and overlap of the optical transitions are controlled by the concentration of donors and acceptor along the backbones, was expanded to the synthesis of broadly absorbing DTS -BTD copolymers (400 700nm) with the best chromophore producing thin -films essentially black in

PAGE 252

252 color (see Chapter 7) The s ynthesized materials where further investigated in terms of their microstructural organization which was correlated to their charge transport in FETs to reveal hole mobilities increasing with the proportion of unsubstituted heterocycles inserted along the backbone and reaching values as high as ca. 2 x 102 cm2 V1s1 in excellent agreement with the close -stacking and lamellar distances found by structural analysis (0.36 nm and 1.84 nm, respectively). Importantly, a s the polymers were found to form oriented lamellar superstructures in the solid state the charge transport can be anticipated to differ depe nding on the direction considered for the measurements ( e.g. in -plane or normal to the device substrate) As a result, further investigation of the charge transport of the materials performance in vertically stacked photovoltaic devices should be directed to address these considerations as well In general, both Chapter 6 and Chapter 7 point towards the need for semiconducting polymers d erivatives possessing lower lying HOMOs with the perspective of providing light harvesting devices with higher open-circuit voltages ( VOC). This should be possible by decreasing the concentration of electronrich building blocks along the conjugated backbone. Since the most conventional less electron rich heterocycles ( e.g. phenylene) also tend to produce materials with low er charge -carrier mobilities the said less electron rich units should be carefully designed to not lower the polymer intrinsic charge transport properties. In parallel, a further increase in device VOC could be induced by bringing a higher lying LUMO full erene. The combination of a polymer low -lying HOMO, a fullerene highlying LUMO, and an intimate phase mixing between the two components in bulk heterojunction PV devices appears to be the next logical joint effort to initiate between the synthetic chemist and the material scientist. Here, it is worth noting that the polymer LUMO energy level should be repressed along with its HOMO energy level in order to maintain the narrow bandgap of the polymer, and thereby its

PAGE 253

253 improved light absorption properties when compared to the high lying HOMO all -donor polymer analogues such as P3HT After about 15 years of research and development, the ca. 5.5 % of overall power conversion efficiency attained in blends of P3HT with PCBM remains state -of the art This sole obser vation should justify device optimization a nd extra synthetic effort such as that proposed above.

PAGE 254

254 LIST OF REFERENCES (1) McNeill, R.; Siudak, R.; Wardlaw, J. H.; Weiss, D. E. Australian Journal of Chemistry 1963, 16, 1056. (2) Bolto, B A.; Weiss, D. E. Australian Journal of Chemistry 1963, 16, 1076. (3) Bolto, B. A.; McNeill, R.; Weiss, D. E. Australian Journal of Chemistry 1963, 16, 1090 (4) Chiang, C. K.; Druy, M. A.; Gau, S. C.; Heeger, A. J.; Louis, E. J.; MacDiarmid, A. G.; Park, Y. W.; Shirakawa, H. J. Am. Chem. Soc. 1978, 100, 1013. (5) Allard, S.; Forster, M.; Souharce, B.; Thiem, H.; Scherf, U. Angew. Chem. Inter. Ed. 2008, 47, 4070. (6) Kulkarni, A. P.; Tonzola, C. J.; Babel, A.; Jenekhe, S. A. Chem. Mater. 2004, 16, 4556. (7 ) McQuade, D. T.; Pullen, A. E.; Swager, T. M. Chem. Rev. 2000, 100, 2537. (8) Thomas, S. W.; Joly, G. D.; Swager, T. M. Chem. Rev. 2007, 107, 1339. (9) Moller, S.; Perlov, C.; Jackson, W.; Taussig, C.; Forrest, S. R. Nature 2003, 426, 166. (10) Gunes, S.; Neugebauer, H.; Sariciftci, N. S. Chem. Rev. 2007, 107, 1324. (11) Thompson, B. C.; Frchet, J. M. J. Angew. Chem. Int. Ed. 2008, 47, 58. (12) Argun, A. A.; Aubert, P. H.; Thompson, B. C.; Schwendeman, I.; Gaupp, C. L.; Hwang, J.; Pinto, N. J.; Tanner, D. B.; MacDiarmid, A. G.; Reynolds, J. R. Chem. Mater. 2004, 16, 4401. (13) Monk, P. M. S.; Mortimer, R. J.; Rosseinsky, D. R. Electrochromism and electrochromic devices ; Cambridge University Press: Cambridge, 2007. (14) Skotheim, T. A.; Reynolds, J. R. Handbook of Conducting Polymers, Third Edition; CRC Press, 2007. (15) Roncali, J. Macro. Rapid Comm. 2007, 28, 1761. (16) Ling, M. M.; Bao, Z. Chem. Mater. 2004, 16, 4824. (17) McCulloch, I.; Heeney, M.; Bailey, C.; Genevicius, K.; MacDonald, I.; Shkunov, M.; Sparrowe, D.; Tierney, S.; Wagner, R.; Zhang, W.; Chabinyc, M. L.; Kline, R. J.; McGehee, M. D.; Toney, M. F. Nat. Mater. 2006, 5 328. (18) Menard, E.; Meitl, M. A.; Sun, Y.; Park, J. -U.; Shir, D. J. L.; Nam, Y. -S.; Jeon, S.; Rogers, J. A. Chem. Rev. 2007, 107, 1117. (19) Sirringhaus, H. Adv. Mater. 2005, 17, 2411. (20) Yan, H.; Chen, Z.; Zheng, Y.; Newman, C.; Quinn, J. R.; Dotz, F.; Kastler, M.; Facchetti, A. Nature 2009, 457, 679. (21) Sheng Han, L.; Zheng, X.; Liping, M.; Chih Wei, C.; Yang, Y. Appl. P hys. Lett. 2007, 91, 083507. (22) Sheng Han, L.; Zheng, X.; Guanwen, Y.; Liping, M.; Yang, Y. Appl. Phys. Lett. 2008, 93, 213301. (23) Liu, B.; McCarthy, M. A.; Yoon, Y.; Kim, D. Y.; Wu, Z.; So, F.; Holloway, P. H.; Reynolds, J. R.; Guo, J.; Rinzler, A. G. Adv. Mater. 2008, 20, 3605. (24) Gao, P.; Beckmann, D.; Tsao, H. N.; Feng, X.; Enkelmann, V.; Baumgarten, M.; Pisula, W.; Mllen, K. Adv. Mater. 2009, 21, 213. (25) Wu, J.; Pisula, W.; Mullen, K. Chem. Rev. 2007, 107, 718. (26) Feng, X.; Marcon, V.; Pisul a, W.; Hansen, M. R.; Kirkpatrick, J.; Grozema, F.; Andrienko, D.; Kremer, K.; Mullen, K. Nat. Mater. 2009, 8 421. (27) Sirringhaus, H. Nature 2009, 457, 667. (28) Tang, C. W.; VanSlyke, S. A. Appl. Phys. Lett. 1987, 51, 913. (29) Tang, C. W.; VanSlyke, S A.; Chen, C. H. J. Appl. Phys. 1989, 65, 3610.

PAGE 255

255 (30) Yap, B. K.; Xia, R.; Campoy Quiles, M.; Stavrinou, P. N.; Bradley, D. D. C. Nat. Mater. 2008, 7 376. (31) Berggren, M.; Nilsson, D.; Robinson, N. D. Nat. Mater. 2007, 6 3. (32) Alasdair, J. C.; Donal, D. C. B.; Homer, A. Appl. Phys. Lett. 2001, 79, 2133. (33) Wilkinson, C. I.; Lidzey, D. G.; Palilis, L. C.; Fletcher, R. B.; Martin, S. J.; Wang, X. H.; Bradley, D. D. C. Appl. Phys. Lett. 2001, 79, 171. (34) Kim, J.; McQuade, D. T.; McHugh, Sean K.; Swag er, Timothy M. Angew. Chem. Int. Ed. 2000, 39, 3868. (35) Marsella, M. J.; Swager, T. M. J. Am. Chem. Soc. 1993, 115, 12214. (36) Yu, G.; Gao, J.; Hummelen, J. C.; Wudl, F.; Heeger, A. J. Science 1995, 270 1789. (37) Irwin, M. D.; Buchholz, D. B.; Hains, A. W.; Chang, R. P. H.; Marks, T. J. Proc. Natl. Acad. Sci. 2008, 105, 2783. (38) Alexander, W. H.; Tobin, J. M. Appl. Phys. Lett. 2008, 92, 023504. (39) Scharber, M. C.; Mhlbacher, D.; Koppe, M.; Denk, P.; Waldauf, C.; Heeger, A. J.; Brabec, C. J. Adv. M ater. 2006, 18, 789. (40) Dennler, G.; Scharber, M. C.; Ameri, T.; Denk, P.; Forberich, K.; Waldauf, C.; Brabec, C. J. Adv. Mater. 2008, 20, 579. (41) Dennler, G.; Scharber, M. C.; Brabec, C. J. Adv. Mater. 2009, 21, 1323. (42) Havinga, E. E.; Hoeve, W.; W ynberg, H. Synth. Met. 1993, 55, 299. (43) Wienk, M. M.; Turbiez, M.; Janssen, R. A. J. Adv. Mater. 2008, 20, 2556. (44) Mhlbacher, D.; Scharber, M.; Morana, M.; Zhu, Z.; Waller, D.; Gaudiana, R.; Brabec, C. Adv. Mater. 2006, 18, 2884. (45) Blouin, N.; Mi chaud, A.; Leclerc, M. Adv. Mater. 2007, 19, 2295. (46) Hou, J.; Park, M. H.; Zhang, S.; Yao, Y.; Chen, L. -M.; Li, J. H.; Yang, Y. Macromolecules 2008, 41, 6012. (47) Bundgaard, E.; Krebs, F. C. Macromolecules 2006, 39, 2823. (48) Ergang, W.; Li, W.; Linfe ng, L.; Chan, L.; Wenliu, Z.; Junbiao, P.; Yong, C. Appl. Phys. Lett. 2008, 92, 033307. (49) Kim, J. Y.; Lee, K.; Coates, N. E.; Moses, D.; Nguyen, T. Q.; Dante, M.; Heeger, A. J. Science 2007, 317 222. (50) Mihailetchi, V. D.; Koster, L. J. A.; Hummelen, J. C.; Blom, P. W. M. Phys. Rev. Lett. 2004, 93, 216601. (51) Mihailetchi, V. D.; Xie, H. X.; de Boer, B.; Koster, L. J. A.; Blom, P. W M. Adv. Funct. Mater. 2006, 16, 699. (52) Veldman, D.; Meskers, S. C. J.; Janssen, R. A. J. Adv. Funct. Mater. 2009, 9999, NA. (53) Yang, C.; Kim, J. Y.; Cho, S.; Lee, J. K.; Heeger, A. J.; Wudl, F. J. Am. Chem. Soc. 2008, 130 6444. (54) Deb, S. K. Appl. Opt. Suppl. 1969, 192. (55) Gustafsson Carlberg, J. C.; Ingans, O.; Andersson, M. R.; Booth, C.; Azens, A.; Granqvist, C. G. Electrochim. Acta 1995, 40, 2233. (56) Argun, A. A.; Cirpan, A.; Reynolds, J. R. Adv. Mater. 2003, 15, 1338. (57) Mller, M.; Asaftei, S.; Corr, D.; Ryan, M.; Walder, L. Adv. Mater. 2004, 16, 1558. (58) Somani, P. R.; Radhakrishnan, S. Mater. Chem. Phys. 2003, 77, 117. (59) Zotti, G.; Schiavon, G. Synth. Met. 1989, 30, 151. (60) Jelle, B. P.; Hagen, G.; Sunde, S.; degrd, R. Synth. Met. 1993, 54, 315. (61) Sapp, S. A.; Sotzing, G. A.; Reddinger, J. L.; Reynolds, J. R. Adv. Mater. 1996, 8 808.

PAGE 256

256 (62) Welsh, D. M.; Kumar, A.; Meijer, E. W.; Reynolds, J. R. Adv. Mater. 1999, 11, 1379. (63) Thompson, B. C.; Schottland, P.; Zong, K.; Reynolds, J. R. Chem. Mater. 2000, 12, 1563. (64) Schwendeman, I.; Hwang, J.; Welsh, D. M.; Tanner, D. B.; Reynolds, J. R. A dv. Mater. 2001, 13, 634. (65) R. B. Bennett, W. E. K., M. J. Hannan, L. G. Boxall US Patent 5 446 577 1995. (66) Chandrasekhar, P. US Patent 5 995 273 1999. (67) Tanaka, S.; Yamashita, Y. Synth. Met. 1995, 69, 599. (68) Kitamura, C.; Tanaka, S.; Yamashita Y. J. Chem. Soc., Chem. Commun. 1994, 13, 1585 (69) Kitamura, C.; Tanaka, S.; Yamashita, Y. Chem. Mater. 1996, 8 570. (70) Mauro Morana; Wegscheider, M.; Bonanni, A.; Kopidakis, N.; Shaheen, S.; Scharber, M.; Zhu, Z.; Waller, D.; Gaudiana, R.; Brabec, C. Adv. Funct. Mater. 2008, 18, 1757. (71) Peet, J.; Kim, J. Y.; Coates, N. E.; Ma, W. L.; Moses, D.; Heeger, A. J.; Bazan, G. C. Nat. Mater. 2007, 6 497. (72) Lee, J. K.; Ma, W. L.; Brabec, C. J.; Yuen, J.; Moon, J. S.; Kim, J. Y.; Lee, K.; Bazan, G. C.; Heeger, A. J. J. Am. Chem. Soc. 2008, 130, 3619. (73) Zhang, M.; Tsao, H. N.; Pisula, W.; Yang, C.; Mishra, A. K.; Mullen, K. J. Am. Chem. Soc. 2007, 129, 3472. (74) Tsao, H. N.; Cho, D.; Andreasen, J. W.; Rouhanipour, A.; Breiby, D. W.; Pisula, W.; Mlle n, K. Adv. Mater. 2008, early view N/A. (75) Brgi, L.; Turbiez, M.; Pfeiffer, R.; Bienewald, F.; Kirner, H. J.; Winnewisser, C. Adv. Mater. 2008, 20, 2217. (76) Tamayo, A. B.; Walker, B.; Nguyen, T. Q. J. Phys. Chem. C 2008, 112, 11545. (77) Tantiwiwat, M.; Tamayo, A.; Luu, N.; Dang, X. -D.; Nguyen, T. Q. J. Phys. Chem. C 2008, 112 17402. (78) Peet, J.; Tamayo, A. B.; Dang, X. D.; Seo, J. H.; Nguyen, T. Q. Appl. Phys. Lett. 2008, 93, 163306. (79) Armarego, W. L. F.; Chai, C. L. L. Purification of Laboratory Chemicals, 5th Edition ; ELSEVIER, 2003. (80) Yokoyama, A.; Suzuki, H.; Kubota, Y.; Ohuchi, K.; Higashimura, H.; Yokozawa, T. J. Am. Chem. Soc. 2007, 129, 7236. (81) Brookins, R. N.; Schanze, K. S.; Reynolds, J. R. Macromolecules 2007, 40, 3524. (82) Nie lsen, K. T.; Spanggaard, H.; Krebs, F. C. Macromolecules 2005, 38, 1180. (83) Sugimoto, R.; Taketa, S.; Gu, H. B.; Yoshino, K. Chem. Express 1986, 1 635 638. (84) McCarley, T. D.; Noble, C. O.; DuBois, C. J.; McCarley, R. L. Macromolecules 2001, 34, 7999. (85) Hansen, W. N.; Hansen, G. J. Phys. Rev. A 1987, 36, 1396. (86) Pavlishchuk, V. V.; Addison, A. W. Inorganica Chimica Acta 2000, 298 97. (87) Bard, A. J.; Faulkner, L. R. Electrochemical Methods:Fundamentals and Applications, 2nd Ed., Wiley: New York 2001. (88) Thompson, B. C.; Schottland, P.; Sonmez, G.; Reynolds, J. R. Synth. Met. 2001, 119, 333. (89) Overheim, R. D.; Wagner, D. L. Wiley: New York 1982. (90) Meeker, D. L.; Mudigonda, D. S. K.; Osborn, J. M.; Loveday, D. C.; Ferraris, J. P. Macromole cules 1998, 31, 2943. (91) Hyodo, K. Electrochim. Acta 1994, 39, 265.

PAGE 257

257 (92) Fabretto, M.; Vaithianathan, T.; Hall, C.; Mazurkiewicz, J.; Innis, P. C.; Wallace, G. G.; Murphy, P. Electrochim. Acta 2008, 53, 2250. (93) Fabretto, M.; Vaithianathan, T.; Hall, C .; Murphy, P.; Innis, P. C.; Mazurkiewicz, J.; Wallace, G. G. Electrochem. Commun. 2007, 9 2032. (94) Ma, W.; Yang, C.; Gong, X.; Lee, K.; Heeger, A. J. Adv. Func. Mater. 2005, 15, 1617. (95) Li, G.; Shrotriya, V.; Huang, J.; Yao, Y.; Moriarty, T.; Emery, K.; Yang, Y. Nat. Mater. 2005, 4 864. (96) Moon, J. S.; Lee, J. K.; Cho, S.; Byun, J.; Heeger, A. J. Nano Lett. 2009, 9 230. (97) Lei, S.; Tahara, K.; Feng, X.; Furukawa, S.; De Schryver, F. C.; Mullen, K.; Tobe, Y.; De Feyter, S. J. Am. Chem. Soc. 2008, 130, 7119. (98) Yang, X.; Dou, X.; Rouhanipour, A.; Zhi, L.; Rader, H. J.; Mullen, K. J. Am. Chem. Soc. 2008, 130 4216. (99) Samori, P.; Yin, X.; Tchebotareva, N.; Wang, Z.; Pakula, T.; Jackel, F.; Watson, M. D.; Venturini, A.; Mullen, K.; Rabe, J. P. J Am. Chem. Soc. 2004, 126, 3567. (100) Watson, M. D.; Fechtenkotter, A.; Mullen, K. Chem. Rev. 2001, 101, 1267. (101) Pisula, W.; Tomovic, Z.; Simpson, C.; Kastler, M.; Pakula, T.; Mullen, K. Chem. Mater. 2005, 17, 4296. (102) Beaujuge, P. M.; Pisula, W.; Tsao, H. N.; Ellinger, S.; Mullen, K.; Reynolds, J. R. J. Am. Chem. Soc. 2009, 131, 7514. (103) Koster, L. J. A.; Mihailetchi, V. D.; Blom, P. W. M. Appl. Phys. Lett. 2006, 88, 052104. (104) Sonmez, G.; Shen, C. K. F.; Rubin, Y.; Wudl, F. Angew. Chem. Int er. Ed. 2004, 43, 1498. (105) Sonmez, G.; Sonmez, H. B.; Shen, C. K. F.; Jost, R. W.; Rubin, Y.; Wudl, F. Macromolecules 2005, 38, 669. (106) Sonmez, G. Chem. Commun. 2005, 5251. (107) Salzner, U. J. Phys. Chem. B 2002, 106, 9214. (108) Salzner, U.; Kose, M. E. J. Phys. Chem. B 2002, 106, 9221. (109) Beaujuge, P. M.; Ellinger, S.; Reynolds, J. R. Adv. Mater. 2008, 20, 2772. (110) Reeves, B. D.; Grenier, C. R. G.; Argun, A. A.; Cirpan, A.; McCarley, T. D.; Reynolds, J. R. Macromolecules 2004, 37, 7559. (111) Stille, J. K. Angew. Chem. Int. Ed. 1986, 25, 508. (112) Thompson, B. C.; Kim, Y. G.; McCarley, T. D.; Reynolds, J. R. J. Am. Chem. Soc. 2006, 128, 12714. (113) Mullekom, H. A. M. v.; Vekemans, J. A. J. M.; Meijer, E. W. Chem. Eur. J. 1998, 4 1235. (114) Jayakannan, M.; Hal, P. A. V.; Janssen, R. A. J. J. Polym. Sci. Part A:Polym. Chem. 2002, 40, 2360. (115) Rughooputh, S. D. D. V.; Hotta, S.; Heeger, A. J.; Wudl, F. J. Polym. Sci., Part B: Polym. Phys. 1987, 25, 1071. (116) Faied, K.; Frechette, M.; Rang er, M.; Mazerolle, L.; Levesque, I.; Leclerc, M.; Chen, T. A.; Rieke, R. D. Chem. Mater. 1995, 7 1390. (117) Yue, S.; Berry, G. C.; McCullough, R. D. Macromolecules 1996, 29, 933. (118) Apperloo, J. J.; Janssen, R. A. J.; Malenfant, P. R. L.; Frechet, J. M. J. Macromolecules 2000, 33, 7038. (119) Irvin, J. A.; Piroux, F.; Morvant, M. C.; Robertshaw, V. L.; Angerhofer, A.; Reynolds, J. R. Synth. Met. 1999, 102, 965.

PAGE 258

258 (120) Durmus, A.; Gunbas, G. E.; Camurlu, P.; Toppare, L. Chem. Commun. 2007 3246. (121) Ra imundo, J. -M.; Blanchard, P.; Brisset, H.; Akoudad, S.; Roncali, J. Chem. Commun. 2000, 11, 939. (122) Blanchard, P.; Raimundo, J. M.; Roncali, J. Synth. Met. 2001, 119, 527. (123) Aldakov, D.; Palacios, M. A.; Anzenbacher, P. Chem. Mater. 2005, 17, 5238. (124) Sonmez, G.; Sonmez, H. B.; Shen, C. K. F.; Wudl, F. Adv. Mater. 2004, 16, 1905. (125) Sonmez, G.; Wudl, F. J. Mater. Chem. 2005, 15, 20. (126) Gunbas, G. E.; Durmus, A.; Toppare, L. Adv. Mater. 2008, 20, 691. (127) Gunbas, G. E.; Durmus, A.; Toppare, L. Adv. Funct. Mater. 2008, 18, 2026. (128) Beaujuge, P. M.; Vasilyeva, S.; Ellinger, S.; McCarley, T. D.; Reynolds, J. R. Macromolecules 2009, 42, 3694 3706. (129) Celebi, S.; Balan, A.; Epik, B.; Baran, D.; Toppare, L. Organic Electronics 2009, 10, 631. (130) Durmus, A.; Gunbas, G. E.; Toppare, L. Chem. Mater. 2007, 19, 6247. (131) AlgI, F.; Cihaner, A. Organic Electronics 2009, 10, 704. (132) Atilla Cihaner, F. A.; inodot Advanced Functional Materials 2008, 18, 3583. (133) Kobayashi, T.; al. J. Electroa nal. Chem. 1984, 161, 419. (134) Kim, Y.; Cook, S.; Choulis, S. A.; Nelson, J.; Durrant, J. R.; Bradley, D. D. C. Chem. Mater. 2004, 16, 4812. (135) Friend, R. H.; Granstrom, M.; Petritsch, K.; Arias, A. C.; Lux, A.; Andersson, M. R. Nature 1998, 395 257. (136) Holmes, A. B.; Halls, J. J. M.; Walsh, C. A.; Greenham, N. C.; Marseglia, E. A.; Friend, R. H.; Moratti, S. C. Nature 1995, 376, 498. (137) Alam, M. M.; Jenekhe, S. A. Chem. Mater. 2004, 16, 4647. (138) Veenstra, S. C.; Wienk, M. M.; Verhees, W. J. H.; Kroon, J. M.; Koetse, M. M.; Sweelssen, J.; Bastiaansen, J. J. A. M.; Schoo, H. F. M.; Yang, X.; Alexeev, A.; Loos, J.; Schubert, U. S. Chem. Mater. 2004, 16, 2503. (139) Hou, J.; Tan, Z.; He, Y.; Yang, C.; Li, Y. Macromolecules 2006, 39, 4657. (140) K ietzke, T.; Egbe, D. A. M.; Horhold, H. H.; Neher, D. Macromolecules 2006, 39, 4018. (141) Marder, S. R.; Zhan, X.; Tan, Z.; Domercq, B.; An, Z.; Zhang, X.; Barlow, S.; Li, Y.; Zhu, D.; Kippelen, B. J. Am. Chem. Soc. 2007, 129, 7246. (142) Groenendaal, L.; Jonas, F.; Freitag, D.; Pielartzik, H.; Reynolds, J. R. Adv. Mater. 2000, 12, 481. (143) Beaujuge, P. M.; Ellinger, S.; Reynolds, J. R. Nat. Mater. 2008, 7 795. (144) CIE No. 15.2004, Colorimetry, 3rd Edition; Commission Internationale de l'Eclairage: Vi enna, Austria, 2004. (145) Kulkarni, A. P.; Zhu, Y.; Babel, A.; Wu, P. T.; Jenekhe, S. A. Chem. Mater. 2008, 20, 4212. (146) Jenekhe, S. A.; Lu, L.; Alam, M. M. Macromolecules 2001, 34, 7315. (147) Bredas, J. L.; Beljonne, D.; Coropceanu, V.; Cornil, J. Ch em. Rev. 2004, 104, 4971. (148) Coropceanu, V.; Cornil, J.; da Silva Filho, D. A.; Olivier, Y.; Silbey, R.; Bredas, J. -L. Chem. Rev. 2007, 107, 926. (149) Grabowski, Z. R.; Rotkiewicz, K.; Rettig, W. Chem. Rev. 2003, 103, 3899. (150) Mikkelsen, K. V.; Ratn er, M. A. Chem. Rev. 1987, 87, 113. (151) Bredas, J. L.; Street, G. B. Acc. Chem. Res. 1985, 18, 309.

PAGE 259

259 (152) Bredas, J. L.; Cornil, J.; Beljonne, D.; dos Santos, D. A.; Shuai, Z. Acc. Chem. Res. 1999, 32, 267. (153) Heuer, H. W.; Wehrmann, R.; Kirchmeyer, S Adv. Funct. Mater. 2002, 12, 89. (154) McCarley, T. D.; McCarley, R. L.; Limbach, P. A. Anal. Chem. 1998, 70, 4376. (155) Wyatt, M. F.; Stein, B. K.; Brenton, A. G. Anal. Chem. 2006, 78, 199. (156) Zhang, X.; Qu, Y.; Bu, L.; Tian, H.; Zhang, J.; Wang, L. ; Geng, Y.; Wang, F. Chemistry A European Journal 2007, 13, 6238. (157) Jayakannan, M.; van Dongen, J. L. J.; Janssen, R. A. J. Macromolecules 2001, 34, 5386. (158) Montaudo, G.; Montaudo, M. S.; Puglisi, C.; Samperi, F. Rapid Communications in Mass Spec trometry 1995, 9 453. (159) Schriemer, D. C.; Li, L. Anal. Chem. 1997, 69, 4169. (160) Axelsson, J.; Scrivener, E.; Haddleton, D. M.; Derrick, P. J. Macromolecules 1996, 29, 8875. (161) Zotti, G.; Schiavon, G.; Zecchin, S. Synth. Met. 1995, 72, 275. (162) DuBois, C. J.; Abboud, K. A.; Reynolds, J. R. J. Phys. Chem. B 2004, 108, 8550. (163) Yamamoto, T.; Lee, B. L. Macromolecules 2002, 35, 2993. (164) Lee, B. L.; Yamamoto, T. Macromolecules 1999, 32, 1375. (165) Gledhill, S. E.; Scott, B.; Gregg, B. A. J. M ater. Res. 2005, 20, 3167. (166) Kong, F. T.; Dai, S. Y.; Wang, K. J. Advances in OptoElectronics 2007, 2007, 1. (167) Gratzel, M. J. Photochem. and Photobio. A. 2004, 164, 3. (168) Huang, J.; Li, G.; Yang, Y. Adv. Mater. 2008, 20, 415. (169) Wienk, M. M.; Struijk, M. P.; Janssen, R. A. J. Chem. Phys. Lett. 2006, 422, 488. (170) Beaujuge, P. M.; Subbiah, J.; Choudhury, K. R.; Ellinger, S.; McCarley, T. D.; So, F.; Reynolds, J. R. Chem. Mater. 2009. (171) Campos, L. M.; Mozer, A. J.; Gnes, S.; Winder, C.; N eugebauer, H.; Sariciftci, N. S.; Thompson, B. C.; Reeves, B. D.; Grenier, C. R. G.; Reynolds, J. R. Solar Energy Materials and Solar Cells 2006, 90, 3531. (172) Galand, E. M.; Kim, Y. G.; Mwaura, J. K.; Jones, A. G.; McCarley, T. D.; Shrotriya, V.; Yang, Y.; Reynolds, J. R. Macromolecules 2006, 39, 9132. (173) McCullough, R. D.; Lowe, R. D. J. Chem. Soc., Chem. Commun. 1992, 1 70 (174) McCullough, R. D.; Lowe, R. D.; Jayaraman, M.; Anderson, D. L. J. Org. Chem. 1993, 58, 904. (175) SanMiguel, L.; Matzger A. J. Macromolecules 2007, 40, 9233. (176) Prosa, T. J.; Winokur, M. J.; Moulton, J.; Smith, P.; Heeger, A. J. Macromolecules 1992, 25, 4364. (177) Osaka, I.; McCullough, R. D. Acc. Chem. Res. 2008, 41, 1202. (178) Usta, H.; Lu, G.; Facchetti, A.; Marks, T. J. J. Am. Chem. Soc. 2006, 128, 9034. (179) Lu, G.; Usta, H.; Risko, C.; Wang, L.; Facchetti, A.; Ratner, M. A.; Marks, T. J. J. Am. Chem. Soc. 2008, 130, 7670. (180) Ach, R. B.; Blouin, N.; Bouchard, A.; Leclerc, M. Chem. Mater. 2009, 21, 751. (181) DuBois, C. J.; Abboud, K. A.; Reynolds, J. R. J. Phys. Chem. B 2004, 108, 8550. (182) Hou, J.; Tan, Z.; Yan, Y.; He, Y.; Yang, C.; Li, Y. J. Am. Chem. Soc. 2006, 128, 4911. (183) Subbiah, J.; Beaujuge, P. M.; Choudhury, K. R.; Ellinger, S.; Reynolds, J. R. ; So, F. ACS Applied Materials & Interfaces 2009, 1 1154 1158. (184) Brabec, C. J.; Cravino, A.; Meissner, D.; Sariciftci, N. S.; Fromherz, T.; Rispens, M. T.; Sanchez, L.; Hummelen, J. C. Adv. Funct. Mater. 2001, 11, 374.

PAGE 260

260 (185) Zhang, F.; Jespersen, K. G .; Bjrstrm, C.; Svensson, M.; Andersson, M. R.; Sundstrm, V.; Magnusson, K.; Moons, E.; Yartsev, A.; Ingans, O. Adv. Funct. Mater. 2006, 16, 667. (186) Wienk, M. M.; Kroon, J. M.; Verhees, W. J. H.; Knol, J.; Hummelen, J. C.; Hal, P. A. v.; Janssen, R. A. J. Angew. Chem. Int. Ed. 2003, 42, 3371. (187) Martens, T.; D'Haen, J.; Munters, T.; Beelen, Z.; Goris, L.; Manca, J.; D'Olieslaeger, M.; Vanderzande, D.; De Schepper, L.; Andriessen, R. Synth. Met. 2003, 138, 243. (188) Hoppe, H.; Niggemann, M.; Winde r, C.; Kraut, J.; Hiesgen, R.; Hinsch, A.; Meissner, D.; Sariciftci, N. S. Adv. Funct. Mater. 2004, 14, 1005. (189) Mihailetchi, V. D.; Koster, L. J. A.; Blom, P. W. M.; Melzer, C.; de Boer, B.; van Duren, J. K. J.; Janssen, R. A. J. Adv. Funct. Mater. 2005 15, 795. (190) Mihailetchi, V. D.; Duren, J. K. J. v.; Blom, P. W. M.; Hummelen, J. C.; Janssen, R. A. J.; Kroon, J. M.; Rispens, M. T.; Verhees, W. J. H.; Wienk, M. M. Adv. Funct. Mater. 2003, 13, 43. (191) Jack, E. P.; Alex, C. M.; Brian, E. H.; Shawn R. S.; Michael, D. M.; Martin, H.; Iain, M. Appl. Phys. Lett. 2008, 92, 113309. (192) Li, Y.; Zou, Y. Adv. Mater. 2008, 20, 2952. (193) Hou, J.; Tan, Z. a.; Yan, Y.; He, Y.; Yang, C.; Li, Y. J. Am. Chem. Soc. 2006, 128, 4911. (194) Zhou, E.; Tan, Z. a.; Yang, C.; Li, Y. Macro. Rapid Comm. 2006, 27, 793. (195) Zhan, X.; Risko, C.; Amy, F.; Chan, C.; Zhao, W.; Barlow, S.; Kahn, A.; Bredas, J. -L.; Marder, S. R. J. Am. Chem. Soc. 2005, 127, 9021. (196) Risko, C.; Kushto, G. P.; Kafati, Z. H.; Bredas, J. L. J. Chem. Phys. 2004, 121, 9031. (197) Osaka, I.; Sauv, G.; Zhang, R.; Kowalewski, T.; McCullough, R. D. Adv. Mater. 2007, 19, 4160. (198) Ling, Q. D.; Song, Y.; Lim, S. -L.; Teo, E. Y. H.; Tan, Y. P.; Zhu, C.; Chan, D. S. H.; Kwong, D. -L.; Kang, E. T.; Neoh, K. G. Angew. Chem. Int. Ed. 2006, 45, 2947.

PAGE 261

261 BIOGRAPHICAL SKETCH A native of France, Pierre M. Beaujuge began his graduate studies at the Graduate School of Chemistry and Physics of Bordeaux (ENSCPB) in France in 2002. As part of his graduate program, he joined in 2004 the research laboratories of Arkema, Inc. in King of Prussia PA, where he worked under the supervision of Dr. Gary S. Silverman and Dr. Roman Y. Korotkov. In 2005, Pierre moved to the University of Florida Gainesville, where he joined the research group of Prof. John R. Reynolds to become a PhD candidate In 2006, while at UF, he received his Diplme dIngnieur from the Graduate School of Chemistry and Physics of Bordeaux Hi s doctoral research has been d irected towards developing novel -conjugated semiconducting polymers with tunable optical and charge transport properties for optoelectronic applications He has been involved in device fabrication aspects via collaboration with Prof. Franky So at the Mat erials Science a nd Engineering department of UF