|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 SOLUTION-PROCESSABLE ORGANIC SEMICONDUCTING MATERIALS FOR PHOTOVOLTAIC APPLICATIONS By JIANGUO MEI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010
2 2010 Jianguo Mei
3 To my wife, our soon-to-come son, and my parents
4 ACKNOWLEDGMENTS I would like to thank my advisor Dr. John R Re ynolds. After four year s of practicing wet chemistry in the lab, I am still optimistic and en thusiastic about what I am doing, for which he deserves a lot of credit. He has been a good advi sor, wise listener and generous supporter. He made my experience at UF enjoyable and help ed shape me in many aspects. As a teaching coordinator for his Organic Chemistry class, he taught me how to make and grade an exam, organize an office hour and communi cate with students. As a research er in the lab, he trained me how to properly write a technical report and a grant, draft a cove r letter and a manuscript, review a scientific paper and give a public presentation. As an individual person, I gratefully benefited from the dedication and caring nature he demons trated. He was always there to advice and support when it came to my career development. I would also like to thank th e members of my committee, Prof. Kirk S. Schanze, Prof. Daniel R. Talham, Prof. Ronald K. Castella no, Prof. Franky So, and Prof. Jiangeng Xue, for contributing their valuable time a nd expertise to my education at th e University of Florida. I also wish to thank Prof. Ken Wagener for the many thoughtful pieces of advice and stimulating discussions he has given over the ye ars, and for his efforts to create an excellent environment in the Butler Polymer Laboratory. I am also grateful to my many collaborators w ho have contributed to the projects I have carried out at UF. Dr. Kirk Schanze, Dr. Young-Gi Kim, Dr. Katsu Ogawa, Dr. Nathan Heston and Dr. Jarrett Vella contributed significantly to my very first project on platinum-acetylide low bandgap polymers. Dr. Schanze has been a passi onate supporter and on my reference list since then. Dr. Nathan Heston, Dr. Svetlana Vas ilyeva, Dr. Michael F Durstock (AFRL) and Christopher A Bailey (AFRL) helped to devel op the project on the vinylene-linked donoracceptor conjugated polymers. Dr. Bernard Kippele n, Dr. Shree Prakash Tiwari, Dr. Hyeunseok
5 Cheun and Jaewon Shim at Georgia Tech contribu ted to the projects on small molecule-based field-effect transistors and or ganic solar cells. Dr. Franky S o, Dr. Jegadesan Subbiah and Dr. Kaushik Roy Choudhury in Department of Material Science and Engineering at UF have worked on polymer-based devices together with me. And I also wish to thank Dr. Franky So for his insightful advice, hospitality and willingness to serve as my refe rence. Dr. Ronald Castellano made great efforts to the project on self-assembled amphiphilic cruciform shaped oligothiophenes. I also thank him for the valu able advice and servi ng as a reference. I am deeply indebted to my two Reynolds group collaborators: GRIT fellow Romain Stalder and GRIT fellow Kenneth Graham. We ha ve worked together on a number of projects, and our collaborations still keep moving ahead. I would also like to thank Brian Atiken in the Wagener group. We have worked closely together on regioregular electro active polyolefins, the first collaboration between the two groups. Many people have been influential in my develo pment as a graduate student. My masters degree advisor Prof. Shaoming Yu at the Hefe i University of Technology encouraged and brought me into the scientific field. With his uns elfish support and encouragement I was able to step out and make my move to the States. Prof. Steven P Nolan at the University of New Orleans was actually the first one who trained me as a chemist. Knowi ng that I would transfer to UF one year later, he still accepted me into hi s group, and provided me with various training opportunities. In his group, I, formerly trained as a chemical engineer, was trained by a number of experienced chemists, including Dr. Oscar Na varro, Dr. Roy A. Kelly, III, Dr. Natalie M. Scott, Dr. Nicolas Marion, Dr. Rohit Singh. I sti ll remembered when Dr. Kelly showed me how to use a cannula to transfer liquids under iner t atmosphere and Dr. Singh demonstrated the proper way to make a TLC spot.
6 After joining the Reynolds group, many people cont ributed to my work and made my stay at UF a pleasant journey of l earning. I am grateful for the help and advice over the years from Dr. Stefan Ellinger, Dr. Prasad Taranekar, Dr. Robert Brookins, Dr. Timothy Steckler, Dr. Ryan Walzack, Dr. Nathan Heston, Dr. A ubrey Dyer, Dr. Svetlana Vasily eva, Dr. Dan Patel, Dr. Mike Craig, Dr. Chad Amb, Dr. Pierre Beaujuge and Pe ngjie Shi. I also thank the younger generation of the Butler Lab members, Ken Graham, Romain Stalder, David Liu, Frank Arroyave, Yuying Wei, Bora Inci, Brian Atiken and Paula Delgado for their service and friendship. Special thanks go to Coralie Richard for occasionally bri nging me her delicious home-made cookies. I would also like to show my appreciation to Cheryl Googi ns, Gena Borrero and Sara Klossner for their great service, which made ev erything easy and conveni ent. Additionally, I will take this opportunity to thank George and Jospehine Butler. I have enormously benefited from the environment and facility brought by their generous gifts. I am also grateful for the Butler Polymer Award funded by the Butler Foundation. Finally, I would like to thank my parents Shixin Mei and Jinhua Zhu, and my wife, Yanrong Wu for their unconditional support and love. I am extremely indebted to my wife, who is carrying our first baby hundreds of miles away and asks little from me while I am working to prepare this dissertation. We ar e looking forward to the future and our soon-to-come son.
7 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4LIST OF TABLES ................................................................................................................ .........10LIST OF FIGURES ............................................................................................................... ........11LIST OF ABBREVIATIONS ........................................................................................................1 9ABSTRACT ...................................................................................................................... .............21 CHAPTER 1 INTRODUCTION ................................................................................................................ ..241.1 Organic Semiconducting Materials ..................................................................................241.1.1 Inorganic and Organic Semiconducting Materials ...............................................251.1.1.1 Conductivity ..........................................................................................2220.127.116.11 Polaron, bipolaron and soliton ..............................................................218.104.22.168 The width of energy bands ....................................................................322.214.171.124 Charge carrier, charge transpor t, mobility and charge trap ...................321.1.2 P-, N-type and Ambipolar Semiconducting Materials .........................................351.1.3 Molecular and Polymeric Semiconducting Materials ..........................................361.2 Bandgap Engineering ....................................................................................................... .381.2.1 Bandgap Control ..................................................................................................391.2.2 The Donor-Acceptor Approach ............................................................................401.3 Photovoltaic Devices ...................................................................................................... ..421.3.1 Silicon-Based Solar Cells .....................................................................................431.3.2 Organic Solar Cells ..............................................................................................471.3.3 Bilayer Organic Solar Cells .................................................................................531.3.4 Bulk Heterojunction Organic Solar Cells ............................................................5126.96.36.199 Polymer/PCBM bulk hete rojunction solar cells ....................................5188.8.131.52 Polymer/polymer bulk heterojunction solar cells .................................6184.108.40.206 Molecular bulk heterojunction solar cells .............................................6220.127.116.11 Organic-inorganic hybrid solar cells .....................................................661.3.5 Tandem Bulk-heterojucntion Solar Cells .............................................................681.4 Objectives of this Dissertation ..........................................................................................6 92 EXPERIMENTAL METHODS AND CHARACTERIAZATIONS .....................................712.1Materials Characterization ............................................................................................712.1.1 Structural Characterization ..................................................................................712.1.2 Molecular Weight Characterization .....................................................................712.1.3 Thermal Characterization .....................................................................................722.1.4 Electrochemical Characterization ........................................................................72
8 2.1.5 Optical Characterization.......................................................................................722.1.6 Morphology Characterization by Atomic Force Microscopy ..............................732.1.7 Single Crystal X-ray Diffraction ..........................................................................742.1.8 2-D wide-angle X-ray scattering. .........................................................................752.2 Bulk Heterojunction Solar Cells .......................................................................................752.3 Charge Mobility Measurements .......................................................................................773 LOW BAND GAP PLATIMUN-ACETYLI DE POLYMERS FOR PHOTOVOLTAIC APPLICATIONS .................................................................................................................. ..803.1 Introduction .............................................................................................................. .........803.2 Synthesis of Platinum-acetylide Model Complexes and Polymers ..................................853.3 Structural Characterizations .............................................................................................. 893.4 Photophysical and Electrochemical Studies .....................................................................913.4.1 Photophysical Studies ..........................................................................................913.4.2 Electrochemical Studies .......................................................................................983.5 Solar Cells ............................................................................................................... ........1013.5.1 Optical Properties and Hole Mobility of Polymer Films ...................................1013.5.2 Solar Cell Studies ...............................................................................................1033.7 Mechanism and Energetics of Charge Separation in the Pt-polymer/PC61BM Blends ..1063.6 Conclusion ................................................................................................................ ......1103.7 Experimental Details ...................................................................................................... 1124 VINYLENE-LINKED DONOR-ACCEPTOR POLYMERS FOR PHOTOVOLTAIC APPLICATIONS .................................................................................................................. 1224.1 Introduction .............................................................................................................. .......1224.2 Synthesis of P3HTV and PTVBT ...................................................................................1254.3 Structural Characterizati ons and Optical Studies ...........................................................1274.4 Electrochemical and Spectro electrochemical Studies ....................................................1304.5 Organic Solar Cells ....................................................................................................... ..1334.6 Conclusion ................................................................................................................ ......1354.7 Experimental Details ...................................................................................................... 1365 DIKETOPYRROLOPYRROLE-BASED SEMICONDUCTING MATERIALS FOR PHOTOVOLTAIC APPLICATIONS ..................................................................................1445.1 Introduction .............................................................................................................. .......1445.2 Diketopyrrolopyrrole-based Amphiphilic Oligothiophene .............................................1475.2.1 Synthesis of DPP-1 ............................................................................................1475.2.2 Structural Characterization, Optic al and Electrochemical Studies ....................1495.2.3 Thermal Analysis, Polarized Light Microscopy and X-ray Analysis ................1525.2.4 Morphology Studies ...........................................................................................1565.2.5 Organic Field-Effect Tr ansistors and Molecular Heterojunction Solar Cells ....158 5.3 Diketopyrrolopyrrole-based Hydrogen-bonded Amphiphilic Oligothiophenes .............1615.3.1 Synthesis of DPP-OH-n (n = 0, 1, 2 and 3) and DPP-TEG ...............................1625.3.2 Structural Characterization, Optic al and Electrochemical Studies ....................164
9 5.3.3 Thermal Analysis ...............................................................................................1675.4 Thermally Cleavable DPP-Based Low Bandgap Polymer .............................................1695.5 Conclusion ................................................................................................................ ......1775.6 Experimental Details ...................................................................................................... 1786 ISOINDIGO-BASED SEMICONDUCTING MATERIALS FOR PHOTOVOLTAIC APPLICATIONS .................................................................................................................. 1906.1 Introduction .............................................................................................................. .......1906.2 Isoindigo-based Oligothiophenes ...................................................................................1916.2.1 Synthesis of Isoindi go-Based Oligothiophenes .................................................1916.2.2 Structural Characterization, Optic al and Electrochemical Studies ....................1946.2.3 Molecular Bulk-Heterojunction Solar Cells.......................................................1966.3 Molecular Engineering in Isoindigo-based Oligomers ...................................................1996.3.1 Synthesis of Isoindigo-Based Oligomers ...........................................................1996.3.2 Structural Characterization, Optic al and Electrochemical Studies ....................2006.3.3 Thermal Analysis ...............................................................................................2076.3.4 Field-Effect Transistors and Molecu lar Bulk-Heterojunction Solar Cells .........2086.4 Isoindigo-Based Polymers with Di-block Solubilizing Groups .....................................2126.4.1 Synthesis of Isoindigo-Based Polyme rs with Biphasic Solubilizing Groups ....2136.4.2 Structural Characterization, Optic al and Electrochemical Studies ....................2146.4.3 Current-Voltage Measurements .........................................................................2206.5 N-type Isoindigo-Based Conjugated Polymers ..............................................................2226.5 Conclusion ................................................................................................................ ......2276.6 Experimental Details ...................................................................................................... 2287 PERSPECTIVES AND OUTLOOK ....................................................................................241LIST OF REFERENCES ............................................................................................................ .247BIOGRAPHICAL SKETCH .......................................................................................................265
10 LIST OF TABLES Table page 1-1 Photovoltaic data of repres entative high performance OPVs ............................................613-1 Photophysical properties of model compounds and polymers ..........................................923-2 Photophysical properties of M3, M4, M5, P3, and P4 in ODCB ......................................983-3 Electrochemical properties of model compounds and polymers .....................................1013-4 I-V characteristics of P1/PC61BM photovoltaic devices ..................................................1053-5 Summary of I V characteristics of polymer/PC61BM OPVs ..........................................1065-1 Summary of DPP-1 based OFET devices ........................................................................1595-2 Electrochemical energy levels and gaps of DPP-OH-n (n = 1-4) from DPV ..................1675-3 Photovoltaic performance of PDPP/PC71BM solar cells with different blend ratio. .......1755-4 Transport properties of PDPP-Boc and PDPP polymer and their blends with PC71BM in a ratio of 1:2 ............................................................................................................. ....1766-1 Solid state optical and electrochemical properties and calculated energy levels. ............1966-2 Performance of I-1/I-2:PC60BM sola r cells before and after annealing .........................1976-3 Electrochemically determined HOMO a nd LUMO energy levels of IsoI-N, IsoI-O and IsoI-S (by CV and by DPV). .....................................................................................2066-4 Bottom-Gate Top-Contact OFET Characteri stics of IsoI-N, Is oI-O, and IsoI-S. ............2106-5 Performance of IsoI-N/IsoI-O/IsoI-S:PC61BM solar cells. ..............................................2126-6 Zero-field hole mobility in PIsoIAM -1, PIsoIAM-2 and PIsoIAM-3, derived from fitting J-V data to the trap-free si ngle-carrier SCLC model. ......................................................2216-7 Electrochemically determined energy le vels and gaps of PIsoI-C16 and PIsoI-BTC16 by DPV. ................................................................................................................... .226
11 LIST OF FIGURES Figure page 1-1 One silicon atom has been replaced by a) a pentavalent antimony atom, and b) a trivalent aluminum atom. ...................................................................................................271-2 a) Band gaps of metals (Eg ~ 0 eV), semiconductors (Eg < 3 eV) and insulator (Eg > 3 eV); b) illustration of the energies involv ed in a molecular ionization process; c) schematic illustration of the one-electron ener gy levels for an organic molecule in its ground state electronic conf iguration adopting: the equi librium geometry of the ground state (left) and the e quilibrium geometry of the first ionized state; d) schematic illustration of the band structure of a polymeric chain in the case of a vertical ionization proce ss (left) and the formation of a polaron (right). EIP-v is the vertical ionization energy, Erel, the relaxation energy gained in the ionized state, Edis the distortion energy to be pa id in the ground state in orde r that the molecule adopts the equilibrium geometry of the ionized state, EIP-d, the ionization energy of the distorted molecule ......................................................................................................... ..291-3 a) Spectroelectrochemical experiments monitoring the formation of ionic states, namely polarons (radicalcations) and bipolarons (dic ations), upon progressive increase of an electrical bias in an ox idation process; b) progressive doping of poly(3,4-propylenedioxythiophene) (For simplicity, OH tails are omitted). .....................311-4 Charge mobilities in different semiconduc tors and the necessary mobilities for their relevant applications ......................................................................................................... .341-5 a) The schematic diagram of the energy dist ribution of localized el ectronic states in the energy gap between the HOMO and LUMO bands (Adapted from Ref. 30 with permission); b) Distribution of HOMO and LUMO levels in disordered organic semiconducting materials...................................................................................................351-6 a) representative p-type organic semiconducting materials; b) representative n-type organic semiconducting materi als; c) representative ambipolar semiconducting materials. .................................................................................................................... ........371-7 a) Simplified diagram of the buildup of energy band of conjugate polymer chain; b) Illustration of the nondegenerate ground stat es of polythiophene as a representation for all aromatic conjugated polymer s; c) the origin of bandgap. .......................................391-8 Schematic illustration of donor-accep tor interaction. The HOMO of the donor segment interacts with the HOMO of the acceptor segment to yield two new HOMOs for the D-A polymer, so does LUMO leve ls. After the electrons redistribute themselves from their original non-interacti ng orbitals to the new hybridized orbitals of the polymer, a higher lying HOMO and a lower lying LUMO are formed. This leads to a narrowing of the optical band gap. ....................................................................411-9 a) Representative electron-deficient un its; b) Representative electron-rich units. ............42
12 1-10 a) Representation of a s ilicon-based p-n heterojunction solar cell under short circuit conditions, illustrating the built -in electric field that exis ts in the depletion region near the p-n junction and the direction of current flow upon irradiation with light; b) The density of charge carriers in n-type doping, p-type doping and depletion region. .....441-11 Current-voltage characteristic of a silicon solar cell for dark and light conditions with illustration of the fill factor ( FF ), showing the open circu it voltage (Voc), the short circuit current (Isc) and the voltage and current at the maximum power point. ...............451-12 a) Schematic representation of the irradi ation geometry; b) Comparison of the solar spectrum under AM0 and AM1.5 conditions. ...................................................................471-13 a) Energy diagram showing three possibl e arrangements of the lowest singlet (S1), triplet (T1), and charge transfer (CT) excited states relative to the singlet ground state (S0) for DA blends: Type I represents DA blends in which photoinduced electron transfer (PET) is absent because th e CT state is situated at an energy higher than the lowest S1 state. Types IIa and IIb show s ituations in which PET does occur: with (Type IIa) and without (Type IIb) charge recombination to the lowest T1 state (CRT). Note that Eg and ET represent the lowest energies of Eg(D) or Eg(A), and ET(D) or ET(A), respectively; b) Jablonski diagram with energies of Eg, ET, ECT, EHOMO(D) ELUMO(A) and eVoc relative to the ground state (rounded to a tenth of an eV). The double headed arrow between Voc and Eg indicates the minimum energy difference for which efficient PET is expected and that between Voc and ET the minimum energy difference that prevents CRT. ................................................................501-14 A schematic sketch of electronic proce sses in an organic photovoltaic device. ................521-15 a) The original Tan cell; b) a solution processed polymer-polymer bilayer device. .........531-16 Bulk heterojunction solar cells with different compos itions: a) polymer/ PCBM; b) polymer/polymer; and c) small molecule/PCBM. .............................................................541-17 Representative p-type co njugated polymers with power conversion efficiency over 4% in blend of fullerene derivatives ..................................................................................571-18 Donor and acceptor combinations us ed in all-polymer solar cells ....................................631-19 Solution-processable p-in three layered molecular bul k heterojunction solar cells based on TBP and SIMEF..................................................................................................651-20 a) Energy-level diagram showing the HOMO and LUMO energies of each of the component materials; b) the device struct ure (right) and TEM cross-sectional image (left) of the polymer tandem solar cell. Scale bars, 100 nm (lower image) and 20 nm (upper image). ................................................................................................................ ....682-1 AFM images of self-assembled DPP-based nanowires on mica (10 x 10 m scan size): a) amplitude image, and b) height image. ................................................................74
13 2-2 Four possible FET architectures (in crosssection): a) top contact b) bottom contacts, c) top contacts/top gate, d) bottom contacts/top gate .........................................................793-1 Structures of Pt-acetylide model compounds and polymers. .............................................833-2 Synthesis of M-1, M-2, P-1 and P-2: a) tributyl(thiophen-2-yl)stannane, Pd(PPh3)2Cl2, THF, 76 %; b) NBS, DMF, 92%; c) trimethylsilylacetylene, Pd(PPh3)2Cl2, CuI and iPr2NH-THF, 90%; d) K2CO3, MeOH, 97%; e) transPt(PBu3)2Cl2, CuI and piperdine-CH2Cl2; f) trans -Pt(PBu3)2Cl2, TBAF, CuI and piperidine-CH2Cl2. .............................................................................................................863-3 Synthesis of compound 3-12: a) fuming H2SO4-HNO3, 39 %; b) tributyl(thiophen-2yl)stannane,Pd(PPh3)2Cl2, THF, 76 %; c) Iron-acetic aci d, 65 %; d) tetradecane-7,8dione, p-TSA, CHCl3, 79 %. e) NIS, CHCl3, 87 %; f) trimethylsilylacetylene, Pd(PPh3)2Cl2, CuI and iPr2NH-THF, 84 %; .....................................................................873-4 Synthesis of M-3, M-4, M-5 a nd P-3: a) 1 eq trans-Pt(PBu3)2Cl2, TBAT, CuI and Et3N-CH2Cl2;b) 5eq trans-Pt(PBu3)2Cl2, TBAT, CuI and Et3N-Ch2Cl2; c) CuI and Et3N-CH2Cl2. .....................................................................................................................873-5 Synthesis of P-4: a) CuI and Et3N-CH2Cl2. .......................................................................883-6 31P-NMR spectra of model complexes of 3-13, 3-14, and 3-15. .......................................893-7 MALDI-TOF mass spectra of P1 and P2. ..........................................................................913-8 Absorption and emission spectra of a) M-1 and P1, and b) M-2 and P2. ..........................933-9 Transient absorption difference spectra of a) M-1and b) P1. Excited at 550 nm with 5 ns pulses. Spectra obtained with an init ial 60 ns delay and with succeeding 1 s delay increments. Arrows indicate the dire ction of change of spectra with increasing delay time. ................................................................................................................... .......963-10 Fluorescence emission quenching of M-1by PC61BM. The legend shows the concentration of PC61BM in solution, and the plot in the inset shows the SternVolmer plot of Io/I vs. [PC61BM]. ......................................................................................973-11 UV-vis absorption spectra of P1, P3, and P4. ....................................................................983-12 Cyclic voltammograms of a) M-1, b) P1, c) M2, and d) P2 in CH2Cl2 with 0.1 M TBAPF6 as supporting electrolytes, s canned at 100 mV/s. Potentials are referenced to Fc/Fc+ as an internal standard. .........................................................................................1003-13 I-V characteristic curves of a) P1/PC61BM and b) P2/PC61BM photovoltaic cells under AM 1.5 simulated solar irradiation (100mW-cm-2) ...............................................1043-14 External quantum efficiencies and absorption spec tra of a) P1/PC61BM blend, and b) P2/PC61BM. .....................................................................................................................106
14 3-15 Energy level diagram for P1. ...........................................................................................10 84-1 Chemical structures of P3HTV, PTVBT and bisTVBT ..................................................1244-2 Synthesis of regioregul ar poly(3-hexylthienylenevinyl enes) (P3HTV). a) NBS, DMF, 88%; b) acrylic acid, Pd(OAc)2, P(o-tol)3, Et3N-CH3CN-THF, 95%; b) NBS, LiOAc, CH3CN-H2O, 48%. .............................................................................................1254-3 Synthesis of PTVBT and bisTVBT. a) 1) C8H17Br, Mg, Et2O; 2) Ni(dppf)Cl2, Et2O, reflux, 88% b) 4,4,4',4',5 ,5,5',5'-octamethyl-2,2'-b i(1,3,2-dioxaborolane) [Ir(OMe)COD]2-dtbpy Heptane, 50 oC, 81%; c)Pd(OAc)2, P(o-tol)3,Et3N, CH3CNTHF, reflux, 65%; d) NBS, LiOAc, CH3CN-H2O, rt, 60%; e) Pd2(dba)3,P(otyl)3,Et4NOH, toluene-water, 60 oC, 87%; f) 2-(3,4-dioctylthiophen-2-yl)-4,4,5,5tetramethyl-1,3,2-dioxaborolane, Pd2(dba)3, [(tBu3)PH]BF4, CsF, THF-water, rt, 86%; .......................................................................................................................... .......1274-4 1H-NMR spectra of bisTVBT in deuterated chloroform (top) at room temperature and PTVBT in deuterated tetrachloroethane at 100 oC ....................................................1284-5 IR spectra of bisTVBT and PTVBT ................................................................................1284-6 Absorption spectra of bisTVBT and PTVBT in THF ......................................................1304-7 a) Differential pulse vol tammetry of PTVBT on a platinum working electrode in 0.1M TBAP/PC solution with a step time of 0.02 s, a step size of 2 mV, and amplitude of 100 mV; b) Cyclic voltamme try of PTVBT on a platinum working electrode (0.02 cm2) in 0.1M TBAP/PC solution at 50mV s-1 .........................................1314-8 Spectroelectrochemistry of PTVBT sp ray cast on ITO/glass from 3 mg/mL solution of the polymer in toluene in 0.1M TB AP/PC between 0.2 and 0.85V in 50mV steps (vs Fc/Fc+). The thick line corres ponds to the neutral state of the polymer at 0.2 V. ....1324-9 a) A.M. 1.5 J-V characteristics of devices with varied PTVBT/PC61BM weight percentages; b) A.M. 1.5 Efficiencies measured from cells of differing PTVBTPC61BM weight percentages and also differing active layer thicknesse s; c) IPCE of a representative device. .......................................................................................................1 344-10 Tapping mode atomic force microscopy im ages of spin coated blend films from solutions of varying PTVBT:PCBM weight ratios. The z-scal e factor is 20. .................1355-1 Chemical structure and model of DPP-1. ........................................................................1475-2 Synthesis of DPP-1. a) potassium tert -butoxide, tert -amyl alcohol, 95 oC, ~30%; b) 1-bromo-2-(2-(2-methoxyethoxy)ethoxy)ethane potassium carbonate, DMF, 46%; c) NBS, chloroform, 61%; d) tributyl(5'dodecyl-2,2'-bithiophen-5-yl)stannane, Pd2(dba)3, P(o-tyl)3, THF, 49%........................................................................................1485-3 1H-, and 13C-NMR of DPP-1 in CDCl3 ............................................................................150
15 5-4 a) Absorption and emission of DPP-1 in toluene; b) absorp tion of DPP-1 (2.2 x 10-5 M) in NMP, DCM, THF and DME; c) the images of DPP-1 ) in NMP, DCM, THF and DME ....................................................................................................................... ...1515-5 a) Cyclic voltammetry, and b) differentia l pulse voltammetry of DPP-1 measured in a 0.1M solution of TBAPF6/DCM (scan rate 25mV/s) vs Ag/Ag+ (EFc/Fc + = EAg/Ag + + 0.16 V) ....................................................................................................................... ......1525-6 Differential scanning calorimetr y thermograms. a) DPP-1, b) PC61BM, and c) DPP-1 and PC61BM mixture (1:1, wt %). ...................................................................................1535-7 a) Polarized optical microscope (POM ) images of DPP-1 under 170 C at two different angels; b) Snapshots of the dire cted growth of a DPP-1 plastic phase as viewed by POM at 170 C after shearing (p ictures taken at 5 s intervals and the arrow indicates the shearing direction); c) POM images of DPP-1 under 80 C, and d) RT ......................................................................................................................... .......1545-8 Diffraction patterns obtained fr om DPP-1 at 170, 80, and 50 C. ...................................1555-9 a) tapping mode AFM image (20 20 m, 250 nm in height) of DPP-1 as deposited from THF-hexane onto mica. b) UV-vis sp ectra of DPP-1 in methylcyclohexane (2.2 10-5 M, at 30 C) recorded at 45 s intervals; .................................................................1575-10 Tapping mode AFM images (5 5 m) with z-height line scan profiles of DPP-1 drop-cast from chlorobenzene (2.2 10-4 M) on mica. Solvent evaporation times are a) 3.5 min, b) 5 min and c) 8.5 min. .................................................................................1575-11 a) Output and b) transf er characteristics of a repr esentative DPP-1 field-effect transistor device ............................................................................................................. ..1595-12 a) J-V characteristics and b) exte rnal quantum efficiencies of a representative DPP-1 : PC61BM photovoltaic device before and after annealing at 90 oC ................................1605-13 Molecular structures of DPP-OH-n and DPP-TEG. ........................................................1615-14 Synthesis of donor building blocks for DPP-OH-n. a) C12H25Br, K2CO3, DMF, 60 oC, 94%; b) 4,4,4',4',5,5,5',5'-octame thyl-2,2'-bi(1,3,2-dioxaborolane), Ir[OMe(COD)]2, heptane, 80 oC, 91%; c) 2-bromothiophene, Pd2(dba)3, P(o-tyl)3, aqueous Et4NOH, toluene, 98%; d) 2-bromobithiophene, Pd2(dba)3, P(o-tyl)3, aqueous Et4NOH, toluene, 71%; e) 2-bromoterthiophene, Pd2(dba)3, P(o-tyl)3, aqueous Et4NOH, toluene, 72%; f) 4,4,4',4',5,5, 5',5'-octamethyl-2,2'-bi(1,3,2dioxaborolane), Ir[OMe(COD)]2, heptane, 80 oC, 50%; g) 4,4,4',4',5,5,5',5'octamethyl-2,2'-bi(1,3,2-dioxa borolane), Ir[OMe(COD)]2, heptane, 80 oC, 50%; h) 4,4,4',4',5,5,5',5'-octamethyl-2,2'-bi( 1,3,2-dioxaborolane), Ir[OMe(COD)]2, heptane, 80 oC.............................................................................................................................. ...1625-15 Synthesis of 3,4,5-tris(dodecy loxy)phenylboronic acid. a) K2CO3, C12H25Br, DMF, 90oC, 24h, 85%; b) NaNO2, HNO3, H2O, CH2Cl2, rt, 3h, 90%; c) hydrazine, graphite,
E n B 5-16 S D 5-17 1 H 5-18 A 5-19 D 5-20 D 5-21 T 5-22 a ) P 5-23 a ) P c o 5-24 a ) ( r h h y s p h y 5-25 C 5-26 T r a 6-1 C 6-2 S > 6-3 S ( c 6-4 S 5 6-5 U E tOH, 110 o C B uLi, THF, ynthesis of D D MF, 90oC, 2 H -NMR spe c A bsorption s p D SC thermo g D SC thermo g T hermal tran s ) di-tertb ut y roDOT-Sn 2 ) Weight lo s DPP-Boc i n o mpound 5) Weight lo s r ate, 10 C m eight, 40.7 % y droxy-pol y p in-coated ( y y drogenb o n C yclic volta m T he J-V curv a tio of a) 1: 1 C hemical str u ynthetic sc h > 95%; b) R B ynthetic sc h c at. ), acetic a ynthesis of I 7% for I-1, a U V-vis spect r C 24h, 88% ; -78 oC foll o D PP core fo 2 4h, 42%; c ) c tra of DPP p ectra of D P g rams of DP g rams of DP s ition of PD P y l dicarbona t P d 2(dba)3 s s of PDPPB n solution, t h 15. ............. s s of N, N' b m in-1; onset % ); b) Ultra v y styrene fil m y ellow) and n ding netwo m mograms a e of a DPP p 1 b) 1:2, an d u ctures of Ih eme of isoi n B r, K 2CO3, D h eme of mo n a cid, reflux, I -1, and I-2. a nd I-2, res p r a of I-1 an d ; d) NaNO2, o wed by B( O r DPP-OHn ) NBS, chlo r C12, DPPT P P-OH-n (n = P-C12, DP P P-OH-n (n = P P-Boc ...... t e, DMAP, T P(o-tyl)3, t o B oc as a fun h in-film bef o ................... b is-( tert b ut o 163.6 C; m v iolet-visibl e m containing afte r 2 min rk of DPP. nd different i p olymer sol a d c) 1:3. ...... 1, and I-2. .. n digo buildi n D MF, 100 o C n o-functiona l > 95%; b) R a) P d 2(dba) 3 p ectively. .... d I-2 in solu t 16 HCl, H2O, 0 O Me)3, -78 o n a) Et3N, C r oform, the n T EG and D P = 1-4) in C H P -TEG and D = 1, 2 3 and 4 .................. T HF, rt, 60 % o luene, 85 o C ction of te m o re and after .................. o xycarbony l m idpoint, 17 e absorption 40% N, N' at 180 C (r e .................. i al pulse vo l a r cell under .................. .................. n g blocks. a ) C ................ l ized isoind i R Br, K 2CO3 3 P(o-tyl)3 E .................. t ion and in t h 0 oC and the n o C and HCl, C H2Cl2, rt, 9 5 n HCl, H2O, P P-OH-2 ..... H 2Cl2 and th e D PP-OH-2. ). ................ .................. % ; b) NBS, C C 85%. ...... m perature; b) cleavage; c ) .................. l )-DPP as a f 7.7 C; end spectra of a bis-( tert b u t e d); c) Sche m .................. l tammogra m AM 1.5 co n .................. .................. ) con HCl ( c .................. i go building DMF, 100 E t4NOH, tol u .................. h in-film ...... n KI, rt, 12 h H2O, 80%. 5 %; b) 5-1 K rt, 60%. ..... .................. e ir optical i m .................. .................. .................. C HCl3, rt, 7 7 .................. Absorption ) Crystal st r u .................. f unction of t point 191.8 1.5-m-thi c t oxycarbon y m atic illustr .................. m s of PDPPB n ditions wit h .................. .................. c at. ), acetic a .................. block. a) c o oC. ............. u ene, 85 oC, .................. .................. h 49%; e) .................. K 2CO3, .................. .................. m ages ......... .................. .................. .................. 7 %; c) .................. spectra of u cture for .................. t emperature C; step c k p y l)-DPP, ation of the .................. B oc. ........... h a blend .................. .................. a cid, reflux, .................. o nc. HCl .................. 83% and .................. .................. ..163 ..164 ..164 ..166 ..168 ..169 ..170 ..170 ..171 ..172 ..173 ..174 ..191 ..192 ..193 ..193 ..194
17 6-6 Cyclic voltammetry of I-1 (left) and I-2 (right) measured in a 0.1 M solution of TBAPF6/DCM (scan rate 25 mV/s) vs Fc/Fc+. ................................................................1956-7 J-V characteristics of I-1/I-2:PC61BM solar cells under 100 mW/cm2 white light illumination annealed at 100 oC for 20 min. ....................................................................1976-8 AFM height images of I-1:PCBM (50:50) sp in-coated from chlor obenzene a) as cast, and b) annealed at 100 oC for 20 min; I-2:PCBM (60:40) spin-coated from chlorobenzene c) as cast, and d) annealed at 100 oC for 20 min. All images are 1 x 1 um with 5 nm height scales. RMS surface roughness values of the AFM images are 0.15 nm, 0. 98 nm 0.14 nm and 0.95 nm from left to right..........................................1986-9 Chemical structures of IsoI-N, IsoI-O and IsoI-S. ...........................................................1996-10 Synthesis of IsoI-N, IsoI-O and IsoI-S. a) Pd2(dba)3, P(o-tyl)3 Et4NOH, toluene, 85 oC, 88%, 92%, and 84% for IsoI-N, Is oI-O and IsoI-S, respectively. .............................2006-11 1H-NMR spectra of IsoI-N, IsoI-O and Iso I-S including a) the full spectra and the aromatic regions of b) IsoI-N c) IsoI-O, and IsoI-S. .....................................................2016-12 UV-vis spectra of IsoI-N, IsoI-O and IsoI-S a) in so lution, and b) in thin-film. .............2036-13 Temperature-dependent UV-vi s spectra of a) IsoI-N, b) IsoI-O and c) IsoI-S. ...............2046-14 Cyclic voltammograms (scan rate = 50 mV/s) and differential pulse voltammetry (step size of 2 mV and step time of 0.1 s econds) of IsoI-N, IsoI-O and IsoI-S in 0.1 M TBAPF6-CH2Cl2 electrolyte solution. a) CV of Is oI-N; b) DPV of IsoI-N; c) CV of IsoI-O; d) DPV of IsoI-O; e) CV of IsoI-S; f) DPV of IsoI-S. ....................................2056-15 a) Repetitive scan electropolymeri zation of 5 mM IsoI-N on ITO in 0.1 M TBAPF6/DCM ; b) UV-Vis spect ra of IsoI-N in CH2Cl2 and electropolymerized Poly(IsoI-N) on ITO ........................................................................................................2066-16 Thermogravimetric analysis of a) IsoI-N, b) IsoI-O and c) IsoI-S. .................................2076-17 DSC thermograms of a) IsoI-N, b) IsoI-O and c) IsoI-S (kJ mol-1, unit in parendissertation after transition temperature) ................................................................2086-18 a) Output and b) transfer characteristics of a repres entative IsoI-S field-effect transistor device processed from chloroform solution. ....................................................2106-19 J-V characteristics of IsoI-O/IsoI-S:PC61BM solar cells .................................................2126-20 Polymer structures of PIsoIAM-1, PIsoIAM-2 and PIsoIAM-3. ...........................................2136-21 Synthesis of IsoIAM building blocks and PIsoIAM. a) K2CO3, DMF, 100 oC; b) Pd2(dba)3, P(o-tyl)3, toluene, 85 oC. .................................................................................2146-22 GPC traces of PIsoIAM-1, PIsoIAM-2 and PIsoIAM-3 ........................................................215
18 6-23 UV-vis spectra of PIsoIAM-1, PIsoIAM-2 and PIsoIAM-3 in thin film. ..............................2166-24 Temperature-dependent UVVis spectra of a) PIsoIAM-1; b) PIsoIAM-2; c) PIsoIAM-3; and d) PIsoIAM-1, PIsoIAM-2 and PIsoIAM-3 in dichlorobenzene at 95 oC. ......................2176-25 Cyclic voltammetry of a) PIsoIAM-1; b) PIsoIAM-2; c) PIsoIAM-3 measured in a 0.1M solution of TBAPF6/ACN vs Fc/Fc+. ...............................................................................2186-26 Differential pulse voltammetry of a) PIsoIAM-1; b) PIsoIAM-2; c) PIsoIAM-3 measured in a 0.1M solution of TBAPF6/ACN vs Fc/Fc+. ...............................................................2186-27 2D wide-angle X-ray scattering of PIsoIAM-1, PIsoIAM-2 and PIsoIAM-3. .......................2206-28 a) Experimental dark current dens ities for hole-only devices of PIsoIAM-1, PIsoIAM-2 and PIsoIAM-3 as a function of the effective elec tric field; b) Experimental dark current densities for electron-only devices of PIsoIAM-1, PIsoIAM-2 and PIsoIAM-3 as a function of the effec tive electric field. ..........................................................................2216-29 Polymer structures of PIsoI-C16 and PIsoI-BT-C16. ......................................................2236-30 Synthesis of PIsoI-C16 and PIsoI-BT-C16. a) Pd(dppf)Cl2, KOAc, 1,4-dioxane, 80 oC, 75%; b) Pd2(dba)3, P(o-tyl)3 Et4NOH, toluene, 85 oC, 95% and 93% for PIsoIC16 and PIsoI-BT-C16, respectively. ..............................................................................2246-31 1H-NMR spectrum of compound 6-11. ............................................................................2246-32 Differential pulse voltammetry of PI soI-C16 and PIsoI-BT-C 16. Electrochemical reduction and oxidation of the films on a pl atinum button was carried out in 0.1 M TBAP6/ACN supporting electrolyte using Ag/Ag+ reference electrode (calibrated against Fc/Fc+) and a platinum flag as the counter electrode. .........................................225
19 LIST OF ABBREVIATIONS ACS American Chemical Society AFM Atomic Force Microscopy BTD 2,1,3-benzothidiazole CRT Charge recombination to triplet CST Charge separated state CT Charge transfer CV Cyclic voltammetry DMF Dimethylformamide DPP Diketopyrrolopyrrole D-A-D Donor-Acceptor-Donor dppf 1,1-diphenylphosphino-ferrocene DPV Differential pulse voltammetry Fc/Fc+ Ferrocene/Ferrocenium FF Fill factor GPC Gel permeation chromatography Jsc Short current density MALDI Matrix assisted laser desorption/ionization NBS N-bromosuccimide OPVs Organic photovoltaics OLEDs Organic light-emitting diodes OFETs Organic field-effect transistors Pd2(dba)3 Trisbenzylidene acetone dipalladium (0) PCE Power conversion efficiency PCBM [6,6]-phenyl-C61-butyric aci d methyl ester fullerene
20 PDI Polydispersity index PET Photoinduced electron transfer P3HT Poly(3-hexylthiophene) PTV Poly(thienylene vinylene) SCLC Space-charge limited current TBAF Tetrabutyammonium fluoride TBAT Tetrabutyammonium triphenyldifluorosilicate THF Tetrahydrofuran Voc Open circuit voltage
21 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy SOLUTION-PROCESSABLE ORGANIC SEMICONDUCTING MATERIALS FOR PHOTOVOLTAIC APPLICATIONS By Jianguo Mei August 2010 Chair: John R. Reynolds Major: Chemistry This dissertation documents our efforts in the development of solution-processable organic semiconducting materials for photovoltaic applications. Th e operation of organic solar cells generally involves the following proce sses: light absorption, exciton generation and diffusion, charge separation and transport, as well as charge collection. Thus, the organic molecules and polymers reported are designed to meet one or more criteria imposed by these processes for the purpose of understanding fundame ntal aspects and potentially achieving high performance photovoltaic materials. Incorporation of heavy transi tion metals into polymer backbones theoretically imparts the resulting metallated conjugated polymers with l onger triplet exciton lifetimes, and less geminate recombination compared to pure hydrocarbon conjugated polymers when used as active materials in bulk-heterojunction so lar cells. Toward this end, a family of platinum-acetylide polymers with bandgaps ranging from 1.4 to 2.0 eV has been developed. Extensive photophysical on the polymers and their blends with PCBM (phe nyl-C61-butyric acid methyl ester) revealed that although a triplet excited state is produced following light absorption, it is too low in energy to undergo photoinduced electron tran sfer with PCBM. It is suggested that the
22 photovoltaic response of the solid ma terials arises due to charge se paration from the singlet state of the polymers. The second part of this dissert ation deals with a general appr oach to prepare a class of vinylene-linked donor-acceptor low bandgap conjugated polymers. By using a simple sequence of Heck coupling, Hunsdiecker d ecarboxylation bromination and Su zuki coupling, we were able to prepare a vinylene-linked thiophene-benzoth iadiazole polymer (PTVBT). By removing the torsion and planarization of polymer backbones, PTBVT shows a red shift (~20 nm) in its absorption in comparison to its parent analogue. Using diketopyrrolopyrrole (DPP) as an elec tron acceptor, amphiphilic discrete oligomers that can self-assemble into highly ordered nanos tructures for organic fi eld-effect transistors (OFETs) and molecular bulk-heterojunction solar cel ls (OPVs) were studied. Charge mobility as high as 4 x 10-3 cm2V-1s-1 was obtained from OFET measurement and PCEs of 0.7% were reported with a high fill factor of 0.58 in molecu lar BHJ solar cells with PCBM as an electron acceptor. A DPP-based thermocleavable polymer was also prepared and OPVs based on this polymer demonstrated an enhanced PCE of 1.44% upon cleavage. In chapter 6, isoindigo was introduced as an electron acceptor in -conjugated materials. Isoindigo-based oligothiophenes were prepared a nd used as donor materials in molecular bulkheterojunction OPVs and power conversion efficiencies up to 1.85% were achieved. Three isoindigo-based polymers were synthesized to va lidate the hypodissertation that charge mobility in conjugated polymers can be enhanced via enforcement of interactions by means of introducing biphasic solubilizing gr oups. The results from SCLC m odeling of J-V characteristics of single-carrier diodes are consis tent with the hypodissertation pr esented where a nearly 10-fold increase in hole mobilities was observed for polymers with biphasic solubilizing group. In
23 addition, a facile approach to isoindigo-based ntype conjugated polymers was also reported, in which the LUMO level as deep as -4.1 eV and deep HOMOs of ~6.0 eV were found. These polymers can be considered as an alternativ e to commonly used acceptors such as PCBM derivatives currently employed in polymer-based solar cell devices
24 CHAPTER 1 INTRODUCTION 1.1 Organic Semiconducting Materials Organic semiconducting materials have been focus of inquiry in the development of potential low-cost, large-area, flexible and li ght-weight optoel ectronic devices, such as lightemitting diodes,1 solar cells,2 field-effect transistors,3 organic lasers,4 and electrochromics.5 Compared with their widely known inorganic co unterparts; mainly silicon, germanium and metal oxide semiconductors, organic semiconducting materials present some intrinsic merits. For instance, materials properties can be fine tuned via stru ctural modifications that can be easily achieved through intelligent molecular design. Solutio n-processing can be realized for low cost and large-area devices via numerous techniques (spin-coating, spray coating, ink-jet printing, roll-to-roll printing and Langmui r layer-by-layer assembly, etc).6-8 The low processing temperature, combined with the mechanical flex ibility of organic mate rials, provides great opportunities to access flexible in tegrated circuits, electronic pa per (or fabric), and foldable organic electronics.9-10 To chemists, physicists, materials scientists a nd electrical engineers who currently work in this field, the question then simply becomes how to turn these great vision s and perspectives into reality. Arguably the most important merit of or ganic semiconducting materials lies in the ability to impart functionality by rational molecular de sign with chemists synthetic capabilities. The aspect of materials and electrical engineering is also an indispensi ble component of the successful development of 21st century organic electronics. The performance of these sophisticated electronics depends not only on the intrinsic properti es of the organic semiconducting materials, but also the device configuration and th e technique to assemble the devices. In addition, understanding the operation princi ples is a prerequisite in order to achieve
25 sustainable development in organic devices. In a word, synergetic efforts from different disciplines are required for the realization of reliable and high-perfo rmance opto-electronics using organic materials. The first step woul d be to learn from the success of inorganic semiconductors and understand the unique features of organic materials. Therefore, this chapter will address this critical issue by making compar isons of inorganic semiconductors and organic semiconducting materials, followed by categorizi ng them by either structure or function to present the intrinsic features of organic material s. The subsequent section will then address some basics in organic semiconducting materials such as the origin of the bandgap. After having the information on the materials side, the discussi on will then move to device fabrication and physics. A variety of devices will be cove red and an emphasis will be put on organic photovoltaics (OPVs). Finally, the cent ral topic of this dissertation will be discussed; approaches to design solution-processible organic semiconduc ting materials for photovoltaic applications, and the principles for cons tructing these materials. 1.1.1 Inorganic and Organic Semiconducting Materials Recently, there has been an increasing debate as to whether the next generation semiconductors, organic semiconducting materials, will replace or be a vi able alternative to silicon in the electr onic industry. To set aside whether it is a valid clai m or not, however, it is always important for organic semiconductor to learn from the success of inorganic semiconductor in electronic industr y. This section is intended to present a condensed collection of concepts in modern solid stat e theory, primarily developed for inorganic materials, to explore the applicability of these concep ts in organic semiconductors, a nd to make a comparison between the two types of materials.
26 18.104.22.168 Conductivity Conductivity ( ) is defined as the charge transporte d across a unit cross-sectional area per second per unit electric field, and can be written as: = ze n (1-1) where z is ion charge per elemental charge, e is elemental charge, n is the concentration and is mobility. The conductivity is thus proporti onal to the amount of charge ( ze n) and to the velocity of charge transport () in a field, where ze represents a net charge of electronic charges for each carrier. If it is assumed to travel under unit el ectric field, the velocity is called the mobility Conductivity also shows anisotropic nature in organic materials, because the mobility is usually different with adapting di fferent spatial orientations. The major breakthrough in conductance of or ganic semiconducting materials came from the discovery in 1977 that polyacetylene could be made highly conducting, ~ 103 ( cm)-1, from its intrinsic conductivity, ~ 10-5 -1 cm-1, by exposing it to oxidizing or reducing agents.11 The room temperature conductivity of doped polyacetylene is ther efore comparable with their inorganic counterparts (p olymers), for instance polysulfur nitride (SN)x, whose conductivity is in the order of 103 ( cm)-1 (also to be compared with 5 x 106 ( cm)-1 for copper). It must be stressed that the metallic character of (SN)x is an intrinsic property of the material, related with the presence of one unpaired el ectron for each S-N unit. As a result, the highest occupied molecular orbitals are only ha lf-filled. Since there is no fo rbidden gap (see section 22.214.171.124 for discussion of this term) between the highest occupied and lowest unoccupied levels, the electrons can move freely under an applied elec tric field to give rise to an electrical conductiv ity. In most organic semiconducting materials, they have closed-shell, meani ng all the electrons are paired. Doping is required to bring up the co nductivity as observed in polyactylene.11
27 The doping mechanism in -conjugated organic films differs from that of inorganic semiconductors. In order to get conductivities ap proaching the metallic range, high doping levels are typically required, as high as 20% in the case of pol yacetylene when it is doped by iodine.1113 This is in sharp contrast with inorganic semiconductors wher e doping levels are several orders of magnitude lower. In inorganic semiconductors, the conductivity can be intrinsic or extrinsic. Any conductivity from intrinsic semiconductors is due to electrons acquiring sufficient energy to cross the forbidden gap. The energy can be suppl ied from thermal, photon or electrical field excitation. Excitation of electrons into the conduc tion band in an intrinsic semiconductor leaves behind holes in the valence band. Th at is to say excitation of an intrinsic semiconductor creates two charge carriers. Extrinsic conductivity is due to lattice defect s, resulting from lattice imperfections or impurities. For instance, one silicon atom has been replaced by a pentavalent antimony atom or trivalent aluminum atom, as il lustrated in Figure 1-1. The form er leads to one free electron, called n-doped. It presents n-ty pe conductivity. The latter creates an electron vacancy, namely a hole. This type of material is called p-type and gives rise to p-type conduction. Figure 1-1. The origin of extrinsic conductivity. One silicon atom has been replaced by a) a pentavalent antimony atom, and b) a trivalent aluminum atom.
28 In organic semiconductors, optical absorption ca n lead to formation of spatially localized electrically neutra l electron-hole pairs (called excitons). The exciton binding energy that holds electron-hole pairs is typically hi gh around ~0.5 eV in organic materials.14 Exciton motions are not affected by an electrical field, since they are neural species. In order to generate an electrical current, excitons are required to be split to genera te free charge carriers. This feature herein leads to a different operation mechanism in organic solar cells. More analysis will be given in section 1.4. Initially, high conductivity observed in doped organic semiconductor was ascribed to the formation of unfilled electronic bands, resulting from electrons being removed from the top of the valence band for p-doping or being added to the bottom of conduction band for n-doping.15 This is in analogy to the m echanism of charge generation mechanism in doped inorganic semiconductors. The validation of this theory was questioned when unpaired electrons were not observed in doped polyacetylene, polypyrrole and poly( paraphenylene), etc. Instead, polaronic models and disordered charge-transport mechan isms are consistent with the experimental findings.16-18 126.96.36.199 Polaron, bipolaron and soliton In a conjugated molecule (polymer), the inte raction of a unit cell wi th all its neighboring units lead to the formation of electronic bands.19 The highest occupied molecular orbitals (HOMO) form the valence band and the lowest u noccupied molecular orbitals (LUMO) create conduction band. The width of the forbidden ba nd, namely energy gap (bandgap), between the valence band and conduction band determines the intr insic electrical properti es of a material as illustrated in Figure 1-2a. Upon doping, high conductiv ity can be observed in organic materials. As discussed in section 188.8.131.52. however, conductivity in highl y doped organic materials does
29 not seem to be associated with unpaired electro ns but rather with spin less charge carriers, characteristic of form ation of bipolarons. Figure 1-2. a) Band gaps of metals (Eg ~ 0 eV), semiconductors (Eg < 3 eV) and insulator (Eg > 3 eV); b) illustration of the energies involv ed in a molecular ionization process; c) schematic illustration of the one-electron ener gy levels for an organic molecule in its ground state electronic conf iguration adopting: the equi librium geometry of the ground state (left) and the e quilibrium geometry of the first ionized state; d) schematic illustration of the band structure of a polymeric chain in the case of a vertical ionization proce ss (left) and the formation of a polaron (right). EIP-v is the vertical ionization energy, Erel, the relaxation energy gained in the ionized state, Edis the distortion energy to be pa id in the ground state in orde r that the molecule adopts the equilibrium geometry of the ionized state, EIP-d, the ionization energy of the distorted molecule (Adapted fr om Ref. 19 with permission). It is helpful to elucidate dopi ng process in order to understa nd this phenomenon. In organic molecules, it is usually the case that the equili brium geometry in the i onized state is different than in the ground state, e.g., benzen oid-like geometry in ground state vs quinoid-like geometry in the ionized state. Figure 1-2b shows an ioni zation process of a molecule (polymer). There are
30 two possible pathways, a verti cal ionization process and a di storted ground-state ionization process. In the former process, an energy EIP-d has to be initially paid for ionized states by keeping the geometry of a ground state. Through relaxing down to the bottom of the potential energy surface of the first (lowest) ionized state, a relaxation energy Erel can be gained back and the ionized state reach its equilibrium geomet ry. The second pathway involves the molecule adapting the equilibrium geometry of the ionized states in its ground state via a lattice distortion. In this case, this distortion l eads to an upward shift of the HOM O level and a down shift of the LUMO level, as illustrated in Figure 1-2b. Only EIP-d is required to proceed to the ionized states. In organic materials, the latter case is energetically favored and leads to the localization of the charge on the chain through a local distortion of th e lattice. This proce ss produces the localized electronic states in the ga p due to a local upward shit of the HOMO and down shit of the HOMO (Figure 1-2c). With the removal of an electron from a polymer chain upon oxidation, we lower the ionization energy by an amount of (Figure 1-2d). If is larger than the distortion energy Edis, this charge localization process will be favored compared to the band process. We say a polaron is created. In chem ical terminology, the polaron is a radical ion with spin number of half. These electronic states can be called polaronic states with a lattice di stortion. When a second electron is removed from the polymer ch ain, a bipolaron is created. A bipolaron is defined as a columbically bound pair of charges with the same sign. In case of p-type doping, the bipolaron in the gap are empty, th e bipolarons are thus spinless With the increase of doping level, polaron becomes bipolaron on the polymer chain, and eventually bipolaron bands are formed. This evolution is consistent with ESR measurements, where the ESR signals grow, saturate, decrease and eventually vanish. Th e optical spectra upon doping provide additional information on this evolution. Figure 1-3 shows spectroelectrochemical experiments upon
31 progressive increase of an electr ical bias in an oxidation proc ess, monitoring the formation of polarons and bipolarons on a donor-acceptor conjugated polymer.20 At very high p-doping level, the absorption completely moves towards near-IR region with the depletion of visible absorption. This can be explained by the broadening of the bipolaron states in the forbidden gap upon increasing the voltage (doping level) and even tually the merging of the lower and upper bipolaron bands with the valence and conducti on band, respectively. For the polymers with small bandgap in their neutral state (< 2.0 eV), th is process can result in a conventional metalliclike conduction mechanism (band theory). Figure 1-3. a) Spectroelectrochemical experiment s monitoring the formation of ionic states, namely polarons (radicalcations) and bipolarons (dic ations), upon progressive increase of an electrical bias in an ox idation process; b) progressive doping of poly(3,4-propylenedioxythiophene) (For simplicity, OH tails are omitted, adapted from Ref 20. with permission). A soliton is a charge associated with a boundary and has the pr operties of a solitary wave which can propagate without deformation and dissipation.19 Therefore, it requires a degenerate ground sate within the polymer. trans -Polyacetylene is such a polymer.21 Other than this unique polymer, most polymers we enc ounter (e.g., polythiophene, polypy rrole) have a non-degenerate ground state. The charge-carrying states are polaroni c, and differ significantly in energy to the HOMO and LUMO edges of the neutral polymer. More details about soliton can be found elsewhere.22-23
32 184.108.40.206 The width of energy bands In inorganic solids, atoms are covalently bonded in an ordered high dens ity crystal lattice. The strong interactions of massive atoms lead to considerably wide energy bandwidths for both valence and conduction band, much wider than kBT ( kBT is the product of the Boltzmann constant, k, and the temperature. At RT kBT is about 0.025 eV). Organic semiconductors, e.g., molecular crystals and conjugated polymers, are c onsidered as disordered materials. They have weak crystal lattices, held together by van der Waals forces.14 The interaction among the adjacent cell units is therefore weak, resu lting in a narrow intermolecular bandwidth ( smaller compared to kBT ). In other words, electronic structure of an organic solid preserves that of a molecule or a single polymer chain. This ha s a pronounced influence on charge transport in organic solids. For instance, th e population of thermally excited charge carriers from valence band to conduction band is exceedingly low. The validity of band theory, e.g., band transport, is thus problematic in most conjugated organic molecules and polymers, except for the case of some highly crystalline high-mobility organic film s, such as single crystals of rubrene or pentacene.24 Instead, charge-carrier mobilities m ove around by a hopping mechanism between localized states in organic semiconductors.25 Therefore, they are inhe rently low, with typical values of <102 cm2/Vs.26 220.127.116.11 Charge carrier, charge transport, mobility and charge trap Both electron and hole are effective charge carriers. A hole is the region of space from which a negative charge has been removed. A prim e parameter in determin ing the successful use of organic compounds as the active layers in op toelectronic devices is the mobility of these charge carriers within the materials. This parame ter controls, for example, the switching speed of field effect transistors, the intensity of light-e mitting diodes, and the separation of charge in photovoltaic cells. Since the organic systems ha ve a closed-shell configuration, the charge
33 carriers are thus either injected into the organic semiconducto rs from the electrodes or generated with photo-induced charge separa tion within the materials at the interface between electrondonor and electron acceptor components. Al so to be noted, pristine films of -conjugated molecules and polymers, other than their doped forms, are used in most devices. As mentioned in the previous se ction, band transport is often no t valid in organic materials. As a matter of fact, conjugated molecules and po lymers are known to transport via a thermally activated hopping-type mechanism, which depends on the interplay between the intrinsic features of the individual molecule and polymer chain a nd their relative orientations and solid state packing.18 There are two major mechanisms proposed for for charge transport in disordered organic materials. In the systems where the electron-phonon interaction and the polaron effect are significant, e.g., small molecules, Marcus theory provides a good estimation for charge transfer between polaronic st ates of different molecules.27 For the systems with weak electronphonon coupling, phonon-assisted hopping mechanism can be accountable for the charge transfer among localized active energy sites (Many models have been developed to describe the hopping mechanisms).18,27 The mobility can be expressed as in the presence of an external electric field: = /E (1-2) where is the velocity of the charges and E is the amplitude of the applied field. Charge mobilities can be measured experimentally by a variety of techniques.28 Many factors affect charge carrier mobility, such as molecular pack ing, the presence of impurities, charge-carrier density, electric field, temperat ure and pressure. Compared to band transport in inorganic materials, the thermally assisted hopping mobility in disordered organic materials is many orders of magnitude lower (t ypically less than 10-2 cm2V-1S-1 as compared to 102 cm2V-1S-1 of Si single
34 crystal).29 Figure 1-4 shows the rough charge mob ilities in different semiconductors and the necessary mobilities for thei r relevant applications. Figure 1-4. Charge mobilities in different semic onductors and the necessary mobilities for their relevant applications Charge carriers and mobilities ar e accompanied by charge traps, the localized energy states where a charge carrier will be reta ined temporally or permanently.18 In other words, once a carrier is trapped, one of two events will eventu ally take place: either the carrier acquires sufficient energy, usually by means of a lattice intera ction, to climb out of the trap and rejoin the conduction band after a specific re tention period, or it recombin es by attracting a carrier of opposite sign. More specifically in the former ca se, if the energies of localized states are separated from the mobility edge by more than a few kBT, the states act as deep traps: once trapped in deep traps, the ch arge carriers have to be rel eased by thermal excitations or photoexcitations. On the other side, trap states with energies within a few kBT of the mobility edge are characterized as shallow trap; afte r being trapped for a characteristic time ( trap), the charges can be thermally activated and released to the band wit hout applying external forces. Electron traps are localized states below the conduction band edge and hole traps are the ones above valence bands. In organic materials the width of the bands is usually very small, less than kBT, therefore HOMO and LUMO are typically used to replace the valence and conduction
35 bands, respectively. The density of states (DOS) in organic thin films can be represented by a Gaussian-like distribution of localized molecular orbitals of individual molecules as shown in Figure 1-5.30-31 The charge transport is dominated by the hopping processes between localized energy states. There is no distinguishing line be tween a charge trap en ergy state and a charge transport state, which are both te mperature-dependent. The origin of trap states can result from impurities, structural defects, geminate pairs, as well as self trapping polarons and biporalons. The more detailed discussions on charge traps an d their measurements are out of the scope of this dissertation and can be found elsewhere.32-33 Figure 1-5. a) The schematic diagram of the energy distribution of localized electronic states in the energy gap between the HOMO and LUMO bands (Adapted from Ref. 30 with permission); b) Distribution of HOMO and LUMO levels in disordered organic semiconducting materials (Adapted from Ref 31 with permission) 1.1.2 P-, N-type and Ambipola r Semiconducting Materials P-type semiconducting materials can stabilize positive charges and transport holes; while n-type materials are the ones that can stab ilize negative charges a nd transport electrons. Ambipolar semiconductors are the combinations that they can transport both charge carriers. Currently, the majorities of conjugated small mo lecules and polymers are used as p-type materials in organic electronic devices. In photovoltaic devices, thes e p-type conjugated
36 materials are also called electron donors. A fe w representative solution processable p-type materials are listed in Figure 1-6a.34-36 Research in the field of so lution processable n-type mate rials (in particular n-type conjugated polymers) lags be hind, compared to the existence of hundreds of p-type semiconductors. In the case of OFET devices, one of the most challenging tasks is to overcome the stability obstacle for n-channel materials. On ly a handful of n-type materials have been reported with fairly good electron mobility under am bient conditions. Besides, little is currently known about electron transport and field-effect mobility of electrons in polymeric semiconductors. With the successful development of high-mobility p-type materials, n-type conductors have recently drawn increasing attenti on due to the potentials to couple with p-type materials and make all-polymer complementary ci rcuits. Figure 1-6b shows some of advanced solution processable n-type semiconducting materials.37-40 Ambipolar semiconducting materials also attract a growing interest as active materials.41-48 Ambipolar semiconductors support both hole and el ectron accumulation. The transistors thereof can operate in pure p-type mode, in pure n-ty pe mode or in a mode where hole and electron accumulations coexist. Ambipolar materials impart ed by this dual nature are suitable for the fabrication of complementary in tegrated circuits based on a si ngle organic semiconductor. This can provide substantial processing advantages over CMOS-type logics (CMOS stands for complementary metal-oxide semiconductor) using on two organic semiconductors, one of p-type and the second of n-type. Figure 1-6c exhibits a couple of examples that have been implemented in ambipolar logics.41,49 1.1.3 Molecular and Polymeric Semiconducting Materials Molecular semiconducting materials in this dissertation refer to those materials that have a well-defined molecular structure. Polymeric mate rials simply refer to conjugated polymers.
37 Figure 1-6. a) representative p-t ype organic semiconducting material s; b) representative n-type organic semiconducting materi als; c) representative ambipolar semiconducting materials.
38 Both molecular and polymeric materi als have been extensively studied.2,50 They have their own intrinsic advantages and di sadvantages regarding their use as active materials in organic electronic devices. As far as molecular semicon ductors are concerned, th eir monodisperse nature with well-defined chemical structures impart s no end-group contaminants and little batch-tobatch variations beyond impurities to these materi als. Much existing knowledge and techniques in the field of small molecules can be applie d to purify them and obtain electronic grade materials. In addition, the physic al and electronic prope rties of molecular materials can be much more easily controlled, even at molecular level. However, the use of solution processing to obtain high-quality thin-films can be problematic. Polymeric materials, on the other hand, have an excellent thin-film formation capability and mechanical flexiblity. They can also be func tionalized and modified w ith great freedom. The biggest disadvantage for polymeric materials is lack of efficient and effective purification techniques that can produce high purity materi als. Trace amounts of metal catalyst residue trapped inside the polymer matrix can be detrim ental to the device perf ormance. It is worth mentioning that recycling gel permeation chro matography (GPC) has recently been used to obtain high purity polymers.51 1.2 Bandgap Engineering An advantage for working with organic materi als is that there are countless numbers of them. Compared with inorganic semiconductors, they are easy to modify or design entirely from the scratch. This implies that their properties, e .g., absorption, emission and charge transport, can be customized for specific applications. The focu s in the previous section (1.1) has been to consider the collective pr operties of organic semiconductors (e.g., charge transport) as solid state materials. In this section, attention will be given to understand and control the fundamental properties of conjugated polymers and molecule s, such as bandgap and energy levels.
39 1.2.1 Bandgap Control The band structure of conjugated polymers results from the interactions of -orbitals of the repeat unit through the entire chain, exemplified in Figure 1-7a. The energy difference between valence band and conduction band is defined as the bandgap. The presence of bandgap is mainly caused by bond length alternation, a product of Peierl s instability, as shown in Figure 1-7b. As a matter of fact, most conjugated polymers have non-degenerate ground stat es between aromatic and quinoid forms (An exception is polyacetyle ne). Although bandgap is mainly dependent on bond length alternation, it is also affected by other factors, including planarity, substitution, aromaticity and interchain interac tion, as sketched in Figure 1-7c.52 Figure 1-7. a) Simplified diagram of the buildup of energy band of conjugate polymer chain; b) Illustration of the nondegenerate ground stat es of polythiophene as a representation for all aromatic conjugated polym ers; c) the origin of bandgap.
40 Specifically, EBLA is related to the difference in b ond length between the single and double bonds. The donor-acceptor appr oach is probably the most effec tive tool to reduce the energy gap and will be discussed in the next section. E is correlated to the torsion angle between the two adjacent repeat units, usually caused by steric in teractions between the rings. To use a linkage such as a double bond or a triple bond can d ecrease the torsion and enhance the planarity, resulting in a smaller bandgap. Esub is caused by electron-donati ng or electron-withdrawing effects of substituents. For inst ance, an electron-donating substitu ent, ( for example, an alkoxyl group), will raise the HOMO level while an electron-withdrawing group, (such as a cyano group), will lower the LUMO leve l. In practice, peripheral subs tituents (R in Figure 1-7c) are often introduced to increase the solubility of th e (otherwise intractable) polymer. These passive substituents also affect Eint through their influence on morphologi cal properties of the polymer. Eres is connected to the aromatic resonance en ergy (Dewar Resonance Energy, DRE). The DRE for benzene, thiophene, pyrole and furan are 22.6, 6.5, 5.3 and 4.3 kcal/mole, respectively.53 The higher this value, the larger the bandgap of the corresponding polymer has. a high DRE value disfavors delocalization. Eint is determined by intermolecular in teractions and can be affected by many factors, including the aforementione d substituent effects. Strong interchain interactions can reduce torsion angle and thus reduce the bandga p. These five correlated parameters all have influence on energy level and bandgap of conj ugated polymers (molecules). To achieve the desired energy levels and bandgaps typically re quire tuning one or more these parameters.52 1.2.2 The Donor-Acceptor Approach The donor-acceptor approach, also known as the push-and-pull method, is based on the incorporation of electron-rich un it (donor) and electron-deficient uni t (acceptor) in an alternating fashion along the polymer backbone. Through the introduction of push-pull driving forces to favor electron delocalization and the forma tion of quinoid mesomeri c structures (D-A<->D+=A-)
41 over the conjugated main chain, the BLA can be si gnificantly reduced. This effectively leads to a compressed bandgap. Photoinduced intramolecular ch arge transfer (ICT) correlated with the highlying HOMO of the donor unit and the lowlying LUMO of the acc eptor unit can also account for the reduced optical band gap.2 Figure 1-8 shows the prin ciple of the donor-acceptor concept.54 Here it can be seen that a hybridizat ion of HOMO and LUMO energy levels of the donor and the acceptor results in the formation of the compressed band gap. One experimental characteristic of an ICT absorption band is its sensitivity to solvent polarity. In general, the energy required for excitation decreases as the solvent polarity increases. The donor-acceptor approach has been by far the most successful synthetic tool to control energy levels and bandgaps while avoiding the necessity of control ling interchain effects or generating insoluble, rigidly planar polymer backbones. A soluble and stable conjugated polymer with bandgap as low as 0.5 eV has been prepared w ithin this group, using bisbenzothiadiazole as an acceptor and dithienopyrrole as a donor.44 Figure 1-8. Schematic illustration of donor-ac ceptor interaction. Th e HOMO of the donor segment interacts with the HOMO of the acceptor segment to yield two new HOMOs for the D-A polymer, so does LUMO leve ls. After the electrons redistribute themselves from their original non-interacti ng orbitals to the new hybridized orbitals of the polymer, a higher lying HOMO and a lower lying LUMO are formed. This leads to a narrowing of the op tical band gap (Adapted from Ref. 54 with permission).
42 We noticed there is one exception that is poly( isothianaphthene). Its bandgap is estimated to be ~ 1eV, and it is nearly 1 eV lo wer than that of poly(thiophene) (~2 eV).55 This is simply realized by stabling quinoid structures without adapting donor-acceptor approach. Figure 1-9a shows a set of repr esentative electron-deficient ar omatic cores that have been used as acceptors in the literature.56-58 A series of commonly-used electron-rich moieties are provided in Figure 1-9b.59-60 In addition, more and more new electron-rich and electrondeficient aromatic heterocyclic compounds are currently developed for the purpose of designing materials for optoelectronic applications.61 The numerous combinations of donors and acceptors have built a rich library of conjugated polymers (molecules) a nd therefore specified properties may be tailored by judicious choices of donor and acceptor moieties. Figure 1-9. a) Representative electron-deficient units; b) Representative electron-rich units. 1.3 Photovoltaic Devices With increasing energy demand for human activ ities and the gradual depletion of nonnon-renewable fossil-based fuels, the search for renewable energy becomes increasingly important, if not critical. Renewable energy, fr om abundant natural sources, e.g., sunlight, geothermal heat, tide and wind, can be used with little or no adverse effects to the environment and no concerns of depletion.
43 Solar energy is the energy from solar radia tion can be regarded as clean and renewable energy. There are two principle types of solar energy: active and passive.62 In this dissertation, we focus on active solar energy; relying on the generation of elect ricity from solar irradiation. The device operated in such transformation is th erefore called photovoltaic device (solar cell). Based on the functional (light-absorbing) material s in these devices, phot ovoltaic devices can be further categorized into inorga nic and organic (OPV) technologies In the following sections, a brief introduction will be given to silicon-based solar cells, including their operating mechanism and some important parameters that define th e performance of such a device. An extensive description of OPVs will be subsequently provided, including their operating mechanism and device configuration. The problems and challenges of these devices will be addressed in the discussions. It is worth menti oning that the research-cell effici encies are different than the module efficiencies. A module is composed of a series of solar cells that are connected and encapsulated. In reality, module efficiency is much lower than the single cell efficiency. 1.3.1 Silicon-Based Solar Cells The first silicon-based solar cell was invented by Russel Ohl at Be ll Laboratory in 1941. There have been many approaches to improve the efficiency of this technology over the years, such as textured front surfaces for enhanced li ght absorption, extremely thin cells with backsurface reflectors for internal li ght trapping through total intern al reflection, and passivated cell surfaces to reduce losses due to recombination effects.63 The highest measured efficiency for a large-area (i.e., 5 inch2) crystalline silicon solar cell sta nds at 21.5%. Due to the defects associated with the grain boundaries the best polycrystalline silicon (p-Si) solar cell efficiencies stand at 19.8%. Amorphous silicon thin-film so lar cells have been reported around 13% with multi-junction architecture.
44 The standard silicon solar cell is also called a p-n junction solar cell, as shown in Figure 110a. A p-n junction is formed where the doping status is abruptly switched fr om p-type to n-type at the interface. At the p-n junction interface, free el ectrons from the n-type side will move to fill the holes from the p-type side. The net outcome is creating a region that is deficient of majority charge carriers and is known as a depletion regi on (Figure 1-10b). Additiona lly there is a built-in electric field resulting from generation of a net positive charge on the n-type side of the depletion region and a net negative charge on the ptype side of the depletion region. Figure 1-10. a) Representation of a silicon-based p-n he terojunction solar cell under short circuit conditions, illustrating the built-i n electric field that exists in the depletion region near the p-n junction and the direc tion of current flow upon irra diation with light; b) The density of charge carriers in n-type doping, p-type doping and depletion region.
45 Upon irradiation, an electron is excited fr om valence band to conduction band when the absorbed energy from incident pho tons is larger than the band gap. An electron-hole pair is formed at the interface w ith relatively weak columbic intera ctions (< 0.05 eV), which can be fully dissociated into a free electron and hole w ith the assistance of th ermal activation. In the field of a built-in potential ac ross the depletion region, electron s are swept toward the n-type region, and holes are swept toward the p-type re gion. If the heterojunction is connected to an external circuit, a current will be measured flowing in the reverse bias sense, from p to n, and the resulting current will be the combination of the photocurrent and the equi librium current (which in solar cell terminology is of ten called the dark current). Figure 1-11. Current-voltage characteristic of a silicon solar cell for da rk and light conditions with illustration of the fill factor ( FF ), showing the open circ uit voltage (Voc), the short circuit current (Isc), and the voltage and current at the maximum power point. The important characteristics of defining a sola r cell are listed in Fi gure 1-11. The value of the current is called the short-circuit current when the bias voltage is zero. It is the current that results from the direct connection of the p-side and n-side with a wire (short-circuited), while the device is under illumination. The op en circuit voltage is the value of the voltage where the curve crosses the bias voltage axis. This is the vo ltage where the device is illuminated, but not
46 connected. The point of maximum power is indi cated where the product of I and V on the curve is maximum. The fill factor is defined as th e maximum power divided by the product of the open circuit voltage and the s hort circuit current. The efficiency of the device is determined by the ratio of the maximum output power of the device to the input power from the incident photons. External quantum efficiency (EQE), define d as the ratio of charges extracted from a device to the number of incident photons (E q. 3), is the key benchmark of solar cell performance. The significance of this characterist ic is that it is a single-wavelength measurement that indicates the number of photogenerated char ge carriers (electrons) produced by a photon of a given energy. When the EQE is examined as a function of wavelength, a photocurrent action spectrum is obtained that determines the relative contribution of various spectral regions (and thus the materials absorbing in th ose regions) to the overall photocur rent generated in the device. ] ) / ( ) ( ) / ( 124 0 [ 100 ] # # [ (%)2 2cm W P nm cm A I Electrons Photons EQEin sc (1-3) Internal quantum efficiency, i.e. the ratio of charges extract ed from a device and the number of photons absorbed by the active layer, provides a useful way to isolate electronic loss mechanism from light coupling and parasi tic absorption losses in a solar cell. In order to evaluate the solar cell performan ce under the same standard, it is necessary to define the radiation sources. Fi gure 1-12a illustrates three cond itions where the solar radiation interacts with different air mass (AM0, AM1 a nd AM1.5G). The letters AM stand for air mass and indicate the amount of atmosphere through wh ich the radiation passes. AM0 indicates the radiation passes through zero air mass and is therefore used for outer space. AM1.5 represents light that travels through approximately one and a half earth atmospheres, where the States and Europe are approximately located when the sun light hits the surface. Figure 1-12b shows the relationship of AM0 and AM1.5G radiation. Here it can be seen that the atmosphere reduces the
47 intensity of solar radi ation in general, especially in the UV region. Additionally, atmospheric components such as water, carbon dioxide, oxygen, ozone, and methane, introduce several characteristic absorption bands into the spectru m. If the total amount of energy is summed over the frequency range the intensity of light out side the earths atmosphere is about 1.35 KW/m2. A similar sum for AM1.5 radiation results in about 890 W/m2. For a standard repor ting, an intensity of 1 KW/m2 is usually selected for a solar simulator. Figure 1-12. a) Schematic representation of the i rradiation geometry; b) Comparison of the solar spectrum under AM0 and AM1.5 conditions. Other than silicon semiconductors, group III-V semiconductors, such as CdTe, GaAs and CuInSe2, have also drawn considerable attention for developing the third generation of inorganic photovoltaics, aimed at reducing thermal and op tical losses in the devices. Significant improvements have been made and more information can be found elsewhere.64-66 1.3.2 Organic Solar Cells Organic semiconducting materials have ga ined momentum for implementation in photovoltaic devices over the past few years in the context of increase demand for renewable energy. Since the inception of a molecular bila yer thin-film organic so lar cell by Tang (with PCE over 1%),67 several concepts have evolved using di fferent device confi gurations, including small molecule-small molecule bilayer,68 conjugated polymer-conj ugated polymer bilayer,69
48 small molecule-conjugated polymer bilayer,70 small molecule-small molecule bulk heterojunction,71-83 small molecule-polymer bulk heterojunction,84 all-polymer bulk heterojunction,40,85-107 as well as combinations of or ganic-inorganic hybrid systems. 108-121 No matter what device configurations are, all OPVs share something in common that is the excitonic character of their optical properties. Optical excitation in organic materials results in the formation of a spatially confined electron-hole pair (Frenkel type exciton), as mentioned free charge carriers are generated in the case of conventional inorganic semiconductors. The exciton binding energy is usually large, on the scale of 0.5 eV and above (note: less than 0.025 eV for silicon-based semiconductors).14 This value is much larger than the thermal activation energy and therefore the formation of heterojunction between electron-donor material and electronacceptor material is necessary to dissociate th e exciton into free elec tron and hole carriers. Below, a full description of electronic and optical processes that take place during the operation of an organic solar cell will be pr esented in the order of photoinduced charge separation (optical absorption a nd exciton formation, exciton migr ation, exciton dissociation at the donor-acceptor interface), char ge carrier transport and char ge collection electrodes. Conjugated materials typically have large absorp tion extinction coefficients in the range of 105 L mol-1cm-1, resulting from the large wave func tion overlap between the electronic ground state and the lowest excited state. The high absorp tivities enable efficient light harvesting even in thin films with thickness of 100-200 nm for or ganic materials. As mentioned earlier, conjugated material presents st rong electron-vibration coupling. On ce promoted to an electronic excited state, the excited system relaxes down to the bottom of the potentia l energy surface of the lowest excited state and the whol e excited state reaches equilibr ium geometry. A new exciton is
49 generated in this event. This process is associ ated with certain energy loss, especially in low bandgap conjugated polymer systems. The neutral exciton will then diffuse randomly within the materials. Typically, the exciton diffusion length, defined by diffusion coe fficient and exciton life time (L = (D )1/2), is in the order of 10-20 nm (highly crystalline pentacene:C60 device has shown an exciton diffusion length as high as 71 nm).122 In order to generate free charge ca rriers, the exciton has to reach the donor and acceptor heterojunction be fore it decays back to the gr ound state. It thus requires the length (thickness) of individual phase (either donor or acceptor phase) should be comparable to exciton diffusion length. This is the ma jor drawback for bilayer organic devices. Upon reaching the donor-acceptor interface, excit on is now able to dissociate into an electron and a hole. It is rather complicated event. At the present, no clear picture has emerged to describe the exciton dissociation at the molecu lar level. A two-step mechanism has been proposed to interpret this event. Initially, an excitonic state at the interface evolves into a chargetransfer state (D+/A-), which then either recombines to the ground state or dissociates into free charge carriers via a char ge-separated state. A closer look at the role of the charge transfer state at donor-acceptor interface and its relationship with ope n circuit voltage (Voc) has been given by Janssen (Figure 1-13).123 In their analysis, the energy difference between elec trochemical bandgap and optical bandgap is introduced to estimate so called effective optical HOMO (Opt HOMOE) and LUMO (Opt LUMOE) energies of the individual donor and acceptor materials in thin films. They are derived from Eox and Ered and using the value of -5.23 eV for Fc/Fc+ versus vacuum level123: ) ( 2 / 1 23 5g sol cv ox Opt HOMOE E eE eV E (1-4) ) ( 2 / 1 23 5g sol cv red Opt LUMOE E eE eV E (1-5)
50 Figure 1-13. a) Energy diagram showing three po ssible arrangements of the lowest singlet (S1), triplet (T1), and charge transfer (CT) excited states relative to the singlet ground state (S0) for DA blends: Type I represents D A blends in which photoinduced electron transfer (PET) is absent because the CT stat e is situated at an energy higher than the lowest S1 state. Types IIa and IIb show situat ions in which PET does occur: with (Type IIa) and without (T ype IIb) charge recombin ation to the lowest T1 state (CRT). Note that Eg and ET represent the lowest energies of Eg(D) or Eg(A), and ET(D) or ET(A), respectively; b) Jablonski diagram with energies of Eg, ET, ECT, EHOMO(D) ELUMO(A) and eVoc relative to the ground state (rounded to a tenth of an eV). The double headed arrow between Voc and Eg indicates the minimum energy difference for which efficient PET is expected and that between Voc and ET the minimum energy difference that prevents CRT. (Adapt ed from Ref. 123 with permission) The energy level of the intermolecular charge tran sfer (CT) state at the donor-acceptor interface can be accordingly described as: ) ( ) ( A E D E EOpt LUMO Opt HOMO CT (1-6) Where is a term that describes a Coulomb force re sponsible for higher energy in CT state than in charge separated (CS) state. By the sa me argument, the driving force for photoinduced electron transfer (PET), GCT, and the driving force for charge recombination to triplet state (CRT), GCRT, can also correlate to this coulomb term, as follows: g Opt LUMO Opt HOMO CTE A E D E G ) ( ) ( (1-7) ) ( ) ( A E D E E GOpt LUMO Opt HOMO T CRT (1-8) Three statements from the semi-experimental analysis are 1) a minimal energy loss of 0.6 eV (Eg eVoc) in organic heterojunction solar cells, 2) a minimal driving force of 0.1 eV for PET
51 process and 3) the considerations of charge recombination into trip let states when there is a large gap between singlet stat e and triplet state. Once the charges (electron and hole) are separate d from each other, they will move to their respective electrodes via hopping in a typically di sordered system. As a result, the chargemobilities, as described earlier, strongly rely on the local morphology and can vary over several orders of magnitude from ordered crystalline organic materials (10-31 cm2 V-1 s-1 ) to highly disordered amorphous conjugated polymers (~ 10-6 -10-3 cm2 V-1 s-1). Upon reaching the electrode/organ ic layer interfaces, charges ar e efficiently collected in most cases. It is noted that the efficiency of the charge collection process can not simply be determined from the work function of the isolat ed electrode and the io nized potential of donor materials or electron affinity of acceptor materi als. With the close proximity between organic layers and electrodes, implica tions, e.g. interfacial charge-density redistribution, geometry modifications and even chemical reactions, will have strong influence the alignment of the organic frontier orbitals and th e electrode Fermi level. Figure 1-14 sketches the electronic processe s above mentioned in OPVs. The solid line represents a favorable process; while the dotted line indicates a recombination event. Step 1 involves light absorptio n to generate an exciton. A geminate recombination event is accompanied as indicated by step 2. If the exciton can safely drift to the DA interface, a process called exciton dissociation will occur and free ch arge carriers are generated in step 3. The generated hole and electron will face two choi ces. The charge carriers can ether hop along their own domain to the electrodes as shown in step 4, or dilapidate via ch arge recombination as indicated in step 7. While the free charge ca rriers move to the electrode, there is still a possibility that a charge recombination event can occur as demonstrated in step 8. If the charge
52 carriers survive, a hole will even reach the anode and an electr on will hit the cathode. These are called effective charge carriers. As many people point out, the optimization of OPVs is a fine balancing act among the various competing event. For instance, the presen ce of an interfacial dipole at the D-A interface will likel y improve the open circuit voltage (Voc), but also reduce the driving force for exciton dissociation. For photo voltaic material point of view, alternating donor and acceptor low bandgap conjugated polymers have been used extensively to harvest longwavelength sunlight. The low-lying LUMO in th ese conjugated materials will decrease the likelihood of charge separation and the high-ly ing HOMO will reduce the open circuit voltage. There are many more trade-off events in the whol e process. Clearly, it is important to bear in minds that careful balances and collective understanding of a ll electronic processes is a prerequisite step to design mate rials for high performance OPVs. Figure 1-14. A schematic sketch of electronic pr ocesses in an organic photovoltaic device.
53 1.3.3 Bilayer Organic Solar Cells A bilayer heterojunction solar cell wi th a layer of vacuum-deposited copper phthalocyanine (CuPc) as a donor and a layer of vacuum-depos ited perylene tetracarboxylic derivative (PV) as an acceptor reported by Tang in 1986 embarked a new era in the development of organic-based photovoltaic devices (Figure 1-15a), even though quite a few studies had previously been performed using single layer configuration. In this seminal work, an open circuit voltage (Voc) of 0.45 V, a short circuit current (Isc) of 2.3 mA/cm2, a high fill factor of 0.65 and a PCE of 0.95% were obtained on a typi cal device illuminated under AM2 (it means the efficiency should be higher under standard AM1.5).67 Many efforts have been made to improve the performance of vacuum-depos ited bilayer solar cells. Currentl y, the state-of-the-art such organic solar cells have reached PCEs ove r 4%, claimed by Xue and Forrest, et al.68 Figure 1-15. a) The original Tan cell; b) a so lution processed polymer-polymer bilayer device. Conjugated polymer-based bilayer devices have also drawn certain interests. These devices have been made by either laminating donor (p-t ype) layer and acceptor (n-type) layer or placing a p-type polymer layer and then adding n-type materials such as fullerenes, semiconductor nanocrystals, or conjugated polymers via incomp atible processing condi tions. For instance, Jenekhe et al. reported the solu tion processed polymer-polymer b ilayer solar cells using the p-
54 type poly(p-phenylene vinylene) (PPV) and th e n-type conjugated ladder polymer poly (benzimidazobenzophenanthroline ladder) (BBL) (Figure 1-15b).69 In their devices, PPV thin films were spin coated from the sulfonium pr ecursor solution in methanol, followed by heating in vacuum at 250 C for 1 h. On top of the resu lting PPV thin film a layer of BBL was spin coated from its solution in GaCl3/nitromethane and decomplexe d by immersing in deionized water for 8 h to remove the gallium chloride. In Jenekhes report, th ey found that the device performance is highly dependent on layer thickne ss. A 40% drop from 49% to 10% in IPCE at maximum point was observed when BBT layer thic kness increased to 75 nm from 50 nm. This highlights the tradeoff between optical length a nd exciton diffusion length that is the created exciton in the donor (acceptor) phase may not re ach D-A interface due to the presence of thick donor (acceptor) phase (a large optical layer). Th is leads to the loss of absorbed photons and external quantum efficiency. The fundamental c onflict has pushed resear chers to explore other potential device configurations. Up to this point, less efforts have been put forth to develop high performance bilayer devices, even though this configuration is sti ll frequently used to diagnose and identify whether charge transfer can o ccur between new donor and acceptor material. 1.3.4 Bulk Heterojunction Organic Solar Cells Three types of bulk heterojunction organic (B HJ) solar cells will be addressed in this section, as shown in Figure 16, namely polym er/PCBM, polymer/polymer and molecular BHJ solar cells. Figure 1-16. Bulk heterojunction solar cells with different compositions: a) polymer/ PCBM; b) polymer/polymer; and c) small molecule/PCBM.
55 18.104.22.168 Polymer/PCBM bulk heterojunction solar cells Sariciftci, et al. reported efficient photoi nduced electron transfer (PET) in a conjugated polymer-fullerene (Buckminsterfullerene, C60) composite in 1992.124 In this report a time scale of 45 fs PET was observed from a conjugated polyme r to fullerene, several orders of magnitude faster than any photoexcitation radiative decay or back electron transfer in the process. Consequently, the quantum efficiency of charge separation in such a composite can approach unity. The tendency for fullerene to crystallize in organic solv ents and on surfaces, however, leads to unfavorable phase separa tion in the composite. This implies charge carriers do not have the necessary channels to reach the electrodes. He nce, efficient solar cells are still not achievable in this context. The conceptually new photovoltaic device, na mely bulk heterojunction (BHJ) solar cell, was demonstrated by Wudl, Heeger and cowork ers in 1995, simply using the blend of poly(2methoxy-5-(2`-ethyl-hexyloxy)-1,4 -phenylene vinylene) (MEH-PPV ) and a fullerene derivative ([6,6]-phenyl-C61-butyric acid methyl ester, PC61BM) (Figure 1-16a).125 The replacement of fullerene with more soluble and less symmetri cal PCBM decreases the formation of large fullerene clusters, and instead increases the possib ility to form a D-A interpenetrating network in the composite. Thus this network with a large co njugated polymer-fullerene interfacial area and the appropriate phase domain size, at lease in one-dimension comparable to the exciton diffusion length, enables the required co mpromise between optical length and exciton diffusion length, as mentioned in the discussion of bilayer heterojunc tion solar cells. In other words, the absorbing sites in the blend are most likely within exciton drifting length to the D-A interface. Meantime, efficient hole and electron transport can also be realized with the likeli hood of the formation of bicontinuous donor and acceptor network in such a bl end. It is also noted that the fast PET can effectively improve the photostability of the c onjugated polymers, due to the fast quenching of
56 the highly reactive ex cited states and thus the reduc tion of any possi ble photooxidation associated with oxygen and water, etc. In a ddition, the use of a single active layer greatly simplify the solution processing, a great advantag e over solution-processed bilayer devices. Since then, many efforts have been put forth to design new soluble -conjugated polymers as donor material for BHJ solar cells, where fullere ne derivatives are primarily used as electron acceptor, e.g. PC61BM and PC71BM. Figure 1-17 shows a list of conjugated polymers from which PCE over 3% have been ach ieved in blend with either PC61BM or PC71BM. Regioregular Poly(3-hexylthiophene) (RRP3HT) is perhaps the most investigated conjugated polymer in the field of BHJ solar ce lls. RR-P3HT has HOMO and LUMO levels at 5.2 and -3.2 eV, respectively, with an optical bandgap of ~ 2.0 eV. Using RR-P3HT as donor and PC61BM as acceptor with the device st ructure of Glass/ITO/P3HT:PC61BM/TiOx/Alumina, BHJ solar cells have been able to exhi bit EQE of 75% and PCE up to 5%.126 Unfortunately, the success of RR-P3HT has not been repeated in other homo-conjugated polymers. The high efficiency of RR-P3HT/PC61BM devices may result from a unique microcrystalline lamellar stacking in the blends. Sariciftci et al. ha s studied a series of regioregular poly(3alkylthiophenes), with butyl, hexyl, octyl, decyl and dedecyl as so lubilizing groups. They observed that chain lengths longer than eight carbons fac ilitate diffusion rates of PC61BM in the blend during the thermal annealing.127 This leads to unfavored phase separation and thereby lowers the device performance. From this study, it is reasonable to conclude that the passive solubilizing chains also affect the optical and electronic propertie s of a conjugated material and hence its performance in a device. In the course of studying RR-P3HT/PC61BM BHJ solar cells, people have learned many useful processing techniques to improve device pe rformance. It has been demonstrated that
57 device performance based on RR-P3HT/PCBM ca n be dramatically enhanced by careful selection of processing solvent,128 solvent vapor annealing,129 thermal annealing,130 and the addition of high-boiling point additives.131 (Note: the additive effect was first reported in the blend of PCPDTBT.132) These strategies are now widely applied in other conjugated polymerbased BHJ solar cells. Figure 1-17 Representative p-type conjugated polymers with pow er conversion efficiency over 4% in blend of fullerene derivatives It is arguable that the further improvement of PCEs in RR-P3HT/PCBM cells is hindered by its relatively large band gap (~ 1.9-2.0 eV). That is to say a RR-P3HT/PCBM blend only
58 absorbs lights with wavelength shorter than 650 nm, at best about 22.4% of total amount of photons under AM 1.5G. It is therefore important to design polymers that can harvest more photons from the available sunlight. It also needs to consider that narrowing bandgap will consequently cause a decrease in open circuit voltage and thus a decr ease in PCE. Through a careful estimation, an optical bandgap of 1.4 eV will be ideal for polymer /PCBM BHJ solar cells provided the appropriate en ergy offset is present.133-134 We are proud to point out the Reynolds group was among the first one to provide an analysis on this issue.134 In practice, an optimal bandgap of 1.3 to 1.8 eV has been reported for conjugated polymers as the active absorbing materials blended with PCBM in high performan ce BHJ solar cells. A few examples are listed below to illuminate the common feat ures that these polymers share. PCPDTBT is donor and acceptor al ternating conjugated polymer. It possesses an optical bandgap (Eg opt) of 1.4 eV (absorption onset 890 nm), and an electrochemical band gap (Eg echem) of 1.7 eV with a HOMO level of -5.3 eV and a LUMO level of -3.6 eV.135 This polymer is the first low bandgap polymer with highly efficient photovoltaic response in near-IR region and has a PCE of 3.2% blended with PC71BM (a PEC of 2.7% blended with PC61BM). The Voc is typically at 0.65 V; the highest observed values approach 0.7 eV With the addition of small amount alkanedithiols in the solvent, PCPDTBT/PC71BM devices have shown PCEs up to 5.5%, with Voc of ~0.62, Jsc of ~16.2 mA cm 2, and FF of ~0.55.132 The role of dithiols is to alter the bulk heterojunction morphology.136-137 PSiF-DBT is a 2,7silafluorene (SiF) and 4,7di(2-thienyl)-2,1,3-benzothiadiazole (DBT) alternating polymer with an optical bandgap of 1.8 eV. The HOMO level from electrochemical measurements was -5.4 eV. In blend with PC60BM, the optimized PSiF-DBT BHJ solar cells exhibited a PCE up to 5.4 %, a large Voc of 0.9 V, a Jsc of 9.5 mA cm 2, and a FF of 0.51.138
59 Noticeably, no thermal or solvent annealing was performed on these devices. In addition, this combination is also the first low ba ndgap polymer with a PCE over 5%. PSPTPB is a combination of PCPDTBT and PSi F-DBT from the material design point of view. From the study of Psi-DBT, it has shown Si atom has a pronounced effect on the device performance. Yang et al. theref ore replaced the bridge carbon atom in cylcopentadithiophene (CPDT) with silicon atom and obtained poly[(4,4 -bis(2-ethylhexyl )dithieno[3,2-b:2 ,3 -d]silole)2,6-diyl-alt-(2,1,3-benzothiadi azole)-4,7-diyl] (PSPTPB). 59 The HOMO and LUMO levels are 5.1 and -3.3 eV measured by cyclic voltammetry (Not e: these values are calculated using -4.8 eV as ferrocene/ferrocenium standard ve rsus vacuum level. It can be converted to -5.4 and -3.6 eV for a comparison reason, since -5.1 eV as ferrocene/f errocenium standard versus vacuum level is used in this dissertati on). The optical band gap (Eg opt) of 1.45 eV is very similar to that of PCPDTBT. The photovoltaic devi ce with a structure of ITO/ PEDOT-PSS/PSBTBT: PCBM /Ca /Al showed a maximum PEC up to 5.1%, with a Voc of 0.68V, a Jsc of 12.7 mA cm 2, and a FF of 0.55. PPtTTBT is a metallated polymer with an Eg opt of 1.8 eV. HOMO and LUMO levels are 5.1 and -3.3 eV, respectively.139 The best solar cell performance is based on the PPtTTBT and PC71BM blend and yields an open circuit voltage ( Voc) of 0.79, a short circuit current ( Jsc) of 10.1 mA cm 2, a FF of 0.51, and a PCE of 4.13% under simulated AM 1.5 G illumination. In an early report, Wong et al. using a similar polymer as a donor and PC61BM as an acceptor demonstrated PCEs up to ~5%, with EQEs as high as 87% at 570 nm.140 Serious doubts, however, have been raised, suggesting that the reported efficiencies are significantly overestimated (In chapter 3, more details will be given).141 Nevertheless, the concept of de signing materials that demonstrate triplet excitons is st ill interesting, as demonstrated in PPtTTBT.
60 PBDTTBT is different from most of linear donor-acceptor type conjugated polymers in that it possesses a cross-conjugation segment in its donor. This polymer has three absorption bands, a distinction from a twoband donor-acceptor system, and th erefore absorbs broadly from 300 to 700 nm with an Eg opt of 1.75 eV. PBDTTBT presents HO MO and LUMO levels of -5.6 and -3.7 eV (after correction) respectively. The HOMO level is about 0.3 eV deeper than PSPTPB and PCPDTBT.142 The deeper HOMO level leads to a higher Voc of 0.9 V, an expected result, considering the difference between the HOMO level of a donor and the LUMO level of PCBM. PCEs up to 5.66% have been obtained from the PBDTTBT-based device, with a Voc of 0.92 V, a Jsc of 10.7 mA cm-2, and a fill factor (FF) of 0.58, which is one of the highest PCEs for single-active-layer OPVs. PTPT is perhaps the only existing random donor-acceptor copolymer that exhibits efficient photovoltaic response, with EQE as high as 63% at 540 nm and over 50% for a broad range. The PTPT/PC71BM cell has highest AM1.5G PCE of 4.4%, with a Voc of 0.81 V, a Jsc of 10.2 mA cm2, and a FF of 0.53.143 The devices also showed encourag ing stability after encapsulation under ambient conditions, where only 15% loss was obse rved after two months storage in air. It is interesting to notice that all th e high performance D-A conjugated polymers mentioned above contain benzothiadiazole as an acceptor. Thieno[3,4-b]thiophene (TT) and diketopyrrolopyrrole (DPP) are the two other known acceptors being used in D-A conjugated polymers that demonstrate their applicab ility in high performance solar cells. PBDTTT and its derivatives are benzo[1,2b :4,5b ]dithiophene (BDT) and Thieno[3,4b]thiophene (TT) alternating polymers, with HOMO levels from -5.20 to -5.5 eV, and LUMO levels from -3.6 to -3.8 eV.144-146 The PBDTTT-CF3/PC71BM-blend BHJ solar cells prepared from chlorobenzene (CB), show a Voc of 0.76 V, a Jsc of 10.2 mA cm-2, a FF of 0.51, which
61 corresponds to a PCE of 3.92%. Upon using a mixture of chlorobenzene (CB) and 1,8diiodoctane (DIO) (97:3 by volume) as a co-solvent, the PCE of this blend is almost doubled to 7.4%, with a increase in Jsc to 14.5 mA cm-2 and a boost in FF to 0.69. This is the highest PCE in a polymeric solar cell reported up to date.146 PDPP3T is an ambipolar conjugated polymer. Field-effect transi stors based on PDPP3T exhibit nearly balanced electron and hole mobilities of 0.01 and 0.04cm2 V-1 s-1, respectively. The best cells obtained for PDPP3T/PC61BM in a 1:2 weight ratio have a Voc of 0.68 V, a Jsc of 8.3 mA cm-2, and a FF of 0.67, yielding a PCE of 3.8%.41 The PCEs of the devices increase to ~4.7% due to a jump in photocurrent to 11.8 mA cm-2, switching PC61BM to PC71BM. In addition, it needs to point that th e EQE of the optimized device is still relatively low, about 35% on average. It can be ascribed to inefficient electron transfer, due to approaching the minimum offset of eVoc = Eg opt -0.6 eV with Voc of 0.65~0.68 V and a small optical bandgap of 1.3 eV.123 An interesting observation for PDPP3T is that it s molecular weights aff ect the performance of photovoltaic devices, but have little or no influence on charge mobilities. Table 1-1. Photovoltaic data of repr esentative high performance OPVs
62 The photovoltaic data of some representative high performance OPVs are summarized in table 1-1. As one can see, the efficiencies for OPVs are steadily approaching the predicted values of 10~12%. Designing new polymers will still be the focus in realizing such a grand goal. 22.214.171.124 Polymer/polymer bulk heterojunction solar cells Friend and Holmes reported the first all-polym er bulk-heterojunction solar cells in 1995, blending MEH-PPV as a donor with CN-PPV as an acceptor.147 Compared to polymer/fullerene photovoltaic devices, all-polymer BH J solar cells have several fore seeable advantages. In order to maximize PCEs of solar cells, it is necessary to absorb acro ss the whole visible and near-IR solar spectrum. In a polymer/polymer blend, both donor and acceptor components contribute to harvest light. It is possible to make a compleme ntary polymer pair that absorb across the field. Although PC71BM provides enhanced absorption in the blue region (peak at ca 500 nm) with respect toPC61BM (peak at ca 349 nm), it contributes little in the red and near-IR region in the case of polymer/fullerene systems. In addition, po lymer Synthesis offers much higher versatility and flexibility than fullerene chemistry, as demons trated in the development of polymer/fullerene solar cells. Thus, it enables the control the ener gy levels to a great extent. The energy offset between the LUMO level of a donor and the LUMO level of an acceptor can be pre-customized to obtain the highest possible Voc that is dependent on the differe nce between the LUMO level of an acceptor and the HOMO level of a donor without sacrificing the driving force for charge transfer and separation. Figure 1-18 shows some donor and acceptor polym er combinations used in all-polymer BHJ solar cells. The state-of-the-art all-polym er cells exhibit PCEs of 1.5-1.9%, using the combination ranging from POPT/MEH-CN/PPV,88,90 P3HT/F8BT,100 M3EH-PPV/CN-etherPPV to TVPT/PDIDTT.40,106 These efficiencies are sign ificantly less than those of polymer/PCBM BHJ solar cells. The efficiency discre pancy has tentatively been attributed to the
63 lower electron mobility of most conjugated polymer s compared to fullerene derivatives which is consequently led by the poor phase segrega tion resulted from binary polymer demixing and the limited availability of ntype conjugated polymers. Figure 1-18 Donor and acceptor combinations used in all-polymer solar cells Polymer blends have an intrinsic proclivity to demixing (phase separate) into their individual pure phase due to low entropy of mixing. There is no exception for rigid rod-like conjugated polymers. Binary random coil-like polymers and their demixing have been
64 extensively studied and ar e relatively well understood.148-149 Phase separation in conjugated polymers is instead much less understood, ev en though some demixing mechanisms, e.g. nucleation and growth of one phase in a surr ounding phase, or by the process of spinodal decompositions, have been proposed based on what has been learned from random coil polymer systems. It also needs to be noted that these models are suitable for bulk phase separation, but may not be applicable to thin-film phase separation.95 Nevertheless, people have applied these principles to control the phase se paration process in orde r to obtain the characteristic domain size of 10-20 nm for photovoltaics. Certain progress ha s been achieved through techniques such as thermal annealing, and the use of mixed solvents.101,103 Many efforts have been devoted to design p-t ype conjugated polymers, resulting in a large pool of p-type conjugated polymers. In contra st, the number of solution-processable n-type conjugated polymers is still limited, as men tioned in section 1.1.2. The available n-channel polymeric materials are still dominated by peryle ne diimide and naphthalene diimide-containing polymers.89,150-151 To design and prepare other n-type pol ymers with variable electron affinity, high electron mobility, and good ambient stability is thereby becoming the biggest challenge for all-polymer BHJ solar cells. With the increased synthetic efforts and the better understanding of phase-separation for rigid-rod type polymers, all-polymer BHJ solar ce lls have the potential to push PCEs to higher levels in organic photovoltaics. 126.96.36.199 Molecular bulk heterojunction solar cells Molecular bulk heterojunction (MBHJ) solar cells refer to photovoltaic devices based on monomeric semiconductors as the donor and the accep tor materials, and fullerene derivatives are still the dominant electron acceptors at this stage. Solution-proce ssable MBHJ solar cells did not draw attention until 2006,74,152 about ten years after the debut of the polymer/PCBM and all-
65 polymer solar cells. Nevertheless, MBHJ solar ce lls have made significa nt progress in a few years,71,73,75-79,81-82 and the PCEs of a singl e layer device have climbed up to 4.4% in a recent report.80 Matsuo and Nakamura et al. have recently reported a solution-pro cessable three-layered p-i-n organic photovoltaic device using tetraben zophorphrin (TBP) as a molecular donor and silymethylfullerene (SIMEF) as an electron acceptor (Figure 1-19).72 This device shows a PCE of 5.2%, a value that brings MBHJ solar cells into the family of the best performance BHJ solar cells that is usually composed of polymer /fullerene combinations. This study consolidates the need for further investigation of sma ll molecules and new device configuration for development of efficient organic solar cells. Figure 1-19. Solution-processable p-i-n three layered molecular bulk heterojunction solar cells based on TBP and SIMEF (Adapted fr om Ref. 72 with permission) The knowledge accumulated in the developmen t of polymer/fullerene solar cells has greatly facilitated the design of small molecule s for MBHJ photovoltaic devices. On the other side, one also needs to realize that crystallin e small molecules and semi-crystalline conjugated
66 polymers are two distinctive types of materials and have different phase separation behaviors. To explore the solution-processi ng conditions and post-annealing t echniques suitable for small molecule/fullerene blends becomes important and needs to be addressed in future study. 188.8.131.52 Organic-inorganic hybrid solar cells Grtezel et al. in 1991 reporte d an efficient photo voltaic device based on organic dye molecules absorbed on TiO2 nano-crystalline thin films, widely described as dye sensitized solar cells (DSSCs).116 This type of photovoltaic device di ffers from the later developed bulkheterojunction solar cells, as they replace organi c electron acceptors, e.g. fullerene derivatives, with inorganic TiO2 nano-crystalline material s, and rely on liquid redox electrolytes to transfer charges from the dye molecules after charge se paration. Other than its own significance as a new type of photovoltaic device, 109,118-119 the study of DSSCs also stimulated the interest in organic-inorganic hybrid solar cells, because it ha s been proven that charge separation can be efficient at the hybrid interface in DSSCs. The hybrid solar cells are of interest for several reasons. First, the energy gap of nanocrystals is a function of their particle size. For instance, the energy gap can be varied from 2.6 to 3.1 eV in CdS and from 2.0 to 2.6 eV in CdSe when changing the size from 6 to 2 nm.108 Therefore, their absorption profile can be tuned in the visible region. This quantum size effect can also be applied to tandem solar cells. Second, the electron affinity of inorganic nanocrystals is usually significantly lower than organic conjugat ed polymers. In the case of widely used CdSe, its electron affinity ranges from 3.5 to 4.5 eV,113 suitable to function as acceptors in inorganic/polymer hybrid solar cells. Third, nano crystals have a large surface area that can interact with polymers to lead to a nanoscale mixing, which is important for efficient charge separation. Last, but the most important, inorganic nanocrystal s have high intrinsic carrier mobilities.
67 There are also some disadvantages for deve lopment of hybrid solar cells. Nanocrystals tend to coalescence into large particles by a process called Ostwald ripening,110 due to high surface tension. The surface of the nanocrystals is thereby typica lly capped by an organic ligand, which passivates the surface electronically. Greenha m et al observed significant quenching of the luminance in the blend of polymer/CdSe when the nanocrystals surface is not coated with trioctylphosphineoxide (TOPO).112 It is much less efficient quenching when the nanocrystals are coated with TOPO, which forms a barrier of 1.1 nm thickness between the nanocrystal core and the polymer. In addition, the ligand is also a barrier for transport of charges from one nanocrystal to the other. If such ligands are to be rem oved, it causes a different problem that electronic defects will be created at the in terface when polymer has an intim ate contact with the nanocrystal in the blend. Since the first report on CdSe/MEH-PPV system, where PCE of 0.1% has been achieved,112 hybrid solar cells have mana ged their way to obtain higher efficiencies. Alivisator et al. reported CdSe-nanorods/P3HT hybrid sola r cells, for which a PCE of 1.7% has been achieved. Greenham used the high boiling solvent trichlorobenzene to process CdSe/MODOPPV blends and obtained a vertic all phase separation. In this de vice, the best PCE of 2.8%, on average 2.1%, has been obtained. By usi ng a low bandgap polymer poly[2,6-(4,4-bis-(2ethylhexyl)-4 H -cyclopenta[2,1b ;3,4b ]dithiophene)alt -4,7-(2,1,3-benzothiadiazole)] (PCPDTBT) blended with CdSe nanocrystals, Dayal demonstrated a PCE of 3.1% for hybrid solar cells. In order to continue to improve the PCEs of hybrid solar cells, a few issues require special attention. The ill-balanced charge carrier mobilities in the hybrid blends typically lead to the decreased EQE and fill factor. Further, the electronic defects on the nanocrystals surface
68 generate electron traps and hinder th e efficient charge transport. It is thereby necessary to come up with a solution either chemically or physically to remove these traps. In addition, designing suitable conjugated polymers specifically for hybrid solar cells is also important to pursue higher efficiencies, considering high electron a ffinity in inorgani c nanocrystals. 1.3.5 Tandem Bulk-heterojucntion Solar Cells In a single layer solar cell, a sizable porti on of energy is inevitabl y lost. Photons with energy smaller than the bandgap of active mate rials will not be abso rbed; while high energy photons will lose the excess energy via thermalizaion. One way to minimize the energy loss is to adapt tandem configuration, as demonstrated in small molecule hete rojucntion solar cells,153 i.e. stack two cells in series with complementary wavelength absorption. Anot her feature of tandem cells is that their large open circuit voltage, in principle, is the sum of the individual subcells. Figure 1-20. a) Energy-level di agram showing the HOMO and LUMO energies of each of the component materials; b) the device struct ure (right) and TEM cross-sectional image (left) of the polymer tandem solar cell. Scale bars, 100 nm (lower image) and 20 nm (upper image). (Adapted from Ref. 153 with permission) Heeger et al. first reported tandem bulk-heter ojucntion solar cells with device structure as shown in Figure 1-20.154 The PCPDTBT/PC61BM single cell yields Jsc = 9.2mA/cm2, Voc = 0.66 V, FF = 0.50, and = 3.0%, and the P3HT/PC71BM single cell yields Jsc = 10.8 mA/cm2, Voc = 0.63 V, FF = 0.69, and = 4.7%. In a typical tandem cell, it gives Jsc = 7.8 mA/cm2, Voc = 1.24
69 V, FF = 0.67, and = 6.5%. By using the similar st rategy, Yang et al. have successfully demonstrated PSBTBT/P3HT/PC71BM tandem cells.155 And the optimal PCE of 5.84% has been obtained, compared to 3.77% for P3HT and 3.94% for PSBTBT individual cells. Janssen et al. showed that the combination of a wide bandgap material (PFTBT) and a small band gap polymer (PBBTDPP) for a tandem cell exhib ited a PCE of 4.9%, with a large Voc of 1.58 V. Brabec et al. estimated that a 15% tandem cell can be indeed achieved through care selection of the polymer pair and device structure.156 Clearly, using a tandem configuration can be a very effective approach to improve the device performance. It also need s to be clear that introducing a large number of different layers via solution processing can be challenging and cos tly in such a tandem de vice. It is therefore important to reach a balance between processing and efficiency. 1.4 Objectives of this Dissertation It is obvious that designing high perfor mance organic semiconducting materials for photovoltaic applications is a ch allenging goal that requires de licate balance of all involved factors. It therefore becomes necessary to unde rstand the underlying design principles and their consequent interplay, apply these interacting rules to material selection, and reconsolidate our comprehension on design criteria, so on and so forth. It is also important to point out that we use chemistry as a tool to make materials a nd develop new chemistry in this process. In this dissertation, I intend to outline the fi eld of organic semiconduc ting materials with a comparison with inorganic material s, and appraise the current stat us of various-types of organic solar cells. Chapter 2 introduces all the character ization tools and experimental setups. Chapter 3 focuses on a series of low bandgap platinum acety lide polymers with the intention to understand whether the introduction of triplet excitons will have an influence on solar cell performance. In chapter 4, a new approach has been demonstrat ed to prepare vinylene-linked donor-acceptor
70 conjugated polymers, and these low bandgap co njugated polymers have been tested in photovoltaic devices. Chapter 5 targets sm all molecules and polymers based on diketopyrrolopyrrole conjugated core and will describe how amphiphilic molecular design has been applied to achieving the first conjugated plastic crystal. In chapter 6, the preparation of a large number of donor-acceptor type small molecules and polymers is described using the newly designed electron-deficient isoindigo as acceptor co re. This dissertation stops at chapter 7 with our perspectives for organic solar cells.
71 CHAPTER 2 EXPERIMENTAL METHODS AND CHARACTERIAZATIONS 2.1 Materials Characterization All reagents and starting materials were pur chased from commercial sources and used without further purification, unless otherwis e noted. The poly-(3,4-ethylenedioxythiophene) /poly(styrenesulfonate) (PEDOT-PSS ) used was Baytron P VP Al 4083. PC61BM was purchased from SES Research, Houston, TX. All procedures involving airand moistu re-sensitive reagents were performed using standard Schlenk techni ques. 1,2-Dichlorobenzene (ODCB) was distilled from calcium hydride under a nitrogen atmosphe re. Other anhydrous solvents were obtained from an anhydrous solvent system. 2.1.1 Structural Characterization All 1H NMR (300 MHz) and 13C NMR (75 MHz) spectra were recorded on a Varian Mercury 300 spectrometer. Chemical shifts for 1H and 13C NMR were referenced to residual signals from CDCl3 (1H NMR =7.25 ppm and 13C NMR =77.23 ppm). Mass spectrograms were recorded on either a Finni gan MAT95Q Hybrid Sector (EI, HRMS) or a Bruker Reflex II (MALDI-TOF) mass spectrometer operated in linear mode with delayed extraction. Elemental analyses were carried out by Atlantic Microlab, Inc. 2.1.2 Molecular Weight Characterization Gel permeation chromatography (GPC) was performed at 40 oC using a Waters Associates GPCV2000 liquid chromatography system with an in ternal differential refr active index detector and two Waters Styragel HR-5E columns (10 m PD, 7.8 mm ID, 300 mm length) using HPLC grade THF as the mobile phase at a flow rate of 1.0 mL/min.
72 2.1.3. Thermal Characterization Thermogravimetric analysis (TGA) was perf ormed on TA Instruments TGA Q1000 Series using dynamic scans under nitrogen. Differentia l scanning calorimetry (DSC) analysis was performed using a TA Instruments Q1000 series equipped with a controlled cooling accessory (liquid nitrogen cooling system). 2.1.4 Electrochemical Characterization Anhydrous TBAPF6 salt, freshly distilled propylene carbonate (PC) and dichloromethane (DCM) were transferred to an argon-filled dry box (OmniLab model, Vacuum Atmospheres). Cyclic voltammetry (CV) and differential pulse voltammetry (DPV) studies were performed using an EG&G Princeton Applied Research model 273A potentiostat/galvanostat in an argonfilled dry box. Solution CVs were carried out in a 0.1M TBAPF6/DCM solution at a scan rate of 25mV/s. Solid state DPVs were recorded in a 0.1M TBAPF6/PC (ACN) solution from drop cast films on a platinum disk electrode, wh ere platinum disk electrode (0.02 cm2), platinum flag and Ag/Ag+ electrode were the working, counter and reference electrode s respectively. All potentials are reported with respect to the potential of Fc/Fc+ redox couple (-5.1 eV vs. vacuum).157 2.1.5 Optical Characterization UV-vis spectra were recorded on a Va rian Cary 500 Scan UV-vis-near-IR spectrophotometer. Solid state absorption spectra were recorded for films s with a PerkinElmer Lambda 25 UV-vis spectrometer. The oligomer f ilms were spin coated on 1 x 1 PEDOT:PSS coated glass slides. Fluorescence spectra were recorded with an ISA SPEX Triax 180 spectrograph coupled to a Spectrum-1 liquid nitr ogen cooled silicon charge coupled device detector. Steady-state UV-visible absorp tion spectra were obtained on a Perkin-Elmer Lambda 25 dual beam absorption spectrometer using 1 cm qu artz cells. Steady-stat e fluorescence emission
73 spectra were recorded on a SPEX TRIAX 180 sp ectrograph coupled with a Spectrum One CCD detector. Steady-state near-IR fluorescence spectra were recorded on SPEX-2 fluorescence spectrophotometer with an Indium-Gallium-Arse nide (InGaAs) detector. Fluorescence decays were obtained by time-correlated single photon coun ting on an instrument that was constructed in-house. A violet diode lase r (405 nm, IBH instruments, Ed inburgh, Scotland, pulse width 800 ps) was used as the excitation source. Transient absorption difference spectra were collected using a 2 mm path length cell on an apparatus described elsewhere.158 Solutions were prepared in ODCB and purged with argon for 30 min before each measurement. The 3rd harmonic (355 nm) of a Continuum Surelite II-10 ND:YAG laser was augmented with a Continuum Surelite OPO Plus optical parametric oscillator to provide 550 nm laser pulses (10 mJ-pulse-1) as the excitation source. 2.1.6 Morphology Characterization by Atomic Force Microscopy Atomic force microscopy (AFM) was extensiv ely used to characterize film morphology in this dissertation. Atomic Force Microscopy (AFM ) relies on the repulsiv e or attractive forces between a finely pointed tip and the surface of a sample. There are 3 main modes of AFM which rely on different interactions between the atom s on the tip and the atoms on the sample. Tapping mode is the most commonly used mode of AF M, especially for our samples, due to its non destructive nature vs. contact mode and higher resolution vs. non contact mode. In tapping mode the tip oscillates and a se t point amplitude is chosen. The tip is then scanned across the surface and the tip sample height is constantly adjusted by the piezoelectric scanner to maintain the constant set point amplitude. In tapping mode the oscillation of the tip is driven by a small piezoelectric in the chip carrier. The difference be tween the oscillation of this piezoelectric and the oscillation of the tip is known as the phase shift. The phase shift can be used to construct an
74 image which provides information about the natu re of the material or the sloping features present. If the surface is biphasi c the phase shift will likely vary between materials and therefore regions of each material can be identified. The am plitude of the tips oscillation can also be used to create an image. The amplitude will vary th e most when the height of the surface changes more abruptly. This can be used to iden tify underlying features which may have been overshadowed in the height image. Figure 2-1 shows two AFM images obtained from our selfassembly studies on diketopyrro lopyrrole-based supramolecular nanostructures. These images are contributed by Kenneth Graham. Figure 2-1. AFM images of self-assembl ed DPP-based nanowires on mica (10 x 10 m scan size): a) amplitude image, and b) height image. 2.1.7. Single Crystal X-ray Diffraction The crystals were submitted to the Center for X-ray Cryptography at UF. Data were collected at 173 K on a Siemens SMART PLATFO RM equipped with A CCD area detector and a graphite monochromator utilizing MoK radiation ( = 0.71073 ). Cell parameters were refined using up to 8192 reflections. A full sphere of data (1850 frames) was collected using the -scan method (0.3 frame width). The firs t 50 frames were re-measured at the end of data collection to monitor instrument and crystal st ability (maximum correction on I was < 1 %).
75 Absorption corrections by integration were applie d based on measured indexed crystal faces. The structure was solved by the Direct Methods in SHELXTL6, and refined using full-matrix least squares. The non-H atoms were treated anis otropically, whereas the hydrogen atoms were calculated in ideal posi tions and were riding on their respective carbon atoms. A total of 158 parameters were refined in the final cycle of refinement using 2169 reflections with I > 2 (I) to yield R1 and wR2 of 4.85% and 12.26%, respectively. 2.1.8 2-D wide-angle X-ray scattering. 2-D wide-angle X-ray scattering (WAXS) expe riments were performed at Max-PlanckInstitute for Polymer Research. The fiber WAXS experiments were performed using a rotating anode (Rigaku 18 kW) Xray beam with pinhole collimation and a 2D Siemens detector. A double graphite monochromator for the Cu K radiation ( = 0.154 nm) was used. The fibers were prepared by filament extrusion using a hom e-built mini-extruder at elevated temperatures. The extrusion was performed for 0.7 mm th in fibers by a constant-rate motion. A Siemens D500 Kristalloflex with a graphite-monochromatized CuK X-ray beam was used for the investigation of the structure in the thin film The diffraction patterns were recorded in the 2 range from 1 to 34 and are presented as a functi on of the scattering vector s; with s = (2 sin )/ where 2 is the scattering angle. 2.2 Bulk Heterojunction Solar Cells Solar cells were fabricated on indium tin oxi de (ITO) covered gla ss substrates (Delta Technologies, Rs = 8-12 /square). The ITO/glass substrates were etched by exposure to aqua regia vapor and subsequently cleaned in an ul trasonic bath for 15 min with aqueous sodium dodecyl sulfate (SDS, Fisher Scientific), de-ion ized water (Milli-Q), a cetone, and isopropanol. The substrates were then trea ted with oxygen plasma for 15 min in a Plasma Cleaner (HARRICK
76 PDC-32G). An aqueous PEDOT-PSS solution (Bayer Baytron P VP Al 4083) was spin coated at 4000 rpm onto the cleaned glass substrates and the resulting polymer film was dried under vacuum for 10 min at 150 C. A solution of the photoactive material (P1 or P2 and PCBM, 1:4 weight ratio of polymer:PCBM) was prepared in toluene or o -dichlorobenzene with solids added at 1 2 % by weight. The solution of the phot oactive materials was th en spin-coated onto the PEDOT-PSS coated substrate in an inert atmosphere box (M-Braun) under argon, and the resulting films were dried under high vacuum ov ernight at room temper ature. The thickness of the active layer was measured using a Dektak 30 30 (Veeco Instruments Inc.) profilometer. (Each thickness reported in this study is the average of at least 3 diffe rent measurements on different regions of the film.) Alumin um (Al, 100 nm) was deposited by thermal evaporation on the photoactive layer. The active area of the devices was 0.25 cm2. Note that the entire process of active layer spin-coating and electrode evaporat ion were carried out inside of the inert atmosphere glove box. The current-voltage (I-V) characteristics were measured with a Keithley SMU 2400 source measurement unit under the illumination of AM 1.5 with an incident power density of 100 mW/cm2 using a 150 W Xe arc lamp power supply (O riel instruments). The external quantum efficiency of the photovoltaic de vices was evaluated by measuring the incident photon to current efficiency (IPCE, %). For IPCE measuremen ts, device pixels were irradiated with monochromatic light through an ISA H20 monochrom ator with a 75 W Xe arc lamp as a light source. The intensity of the source at each wave length was determined using an energy meter (S350, UDT Instruments) equipped with a ca librated silicon detector (Model 221, UDT Instruments). The current response under short circuit conditions was then recorded for each pixel at 10 nm intervals using a Keithley 2400 SMU (positive lead to ITO and negative lead to
77 aluminum). The current-voltage (I-V) curve and IPCE plots were measured for at least three pixels, and the data represent the average of the three measurements. The measurements were performed in air without encapsulation. 2.3 Charge Mobility Measurements Charge mobility is perhaps one of utmost important parameters in determining the applicability of organic semi conducting materials in photovolta ic applications. Typically, organic semiconducting materials (o ther than highly crystalline molecules, e.g. pentacene and rubrene) have low mobility (<<1 cm2V-1s-1). The method for measuring mobility of organic materials is therefore different from that of i norganic materials. In other words, the models developed for inorganic crystalline materials ha ve to be modified before being applied to disordered organic materials. Quite a few su itable methods have been evolved over the past several decades to deal with organic materials, including timeof-flight method, analysis of steady-state trap-free space-charge lim ited current diodes (SCLC), modeling of J V characteristics from field-effect transistors (FET) and pulse radi olysis time-resolved microwave conductivity techniques. These methods have their own featur es and also present their challenges. In this dissertation, SCLC and FET me thods are used to evaluate the mobilities of some newly prepared polymers and small molecules. The applicability of the SCLC technique is built on the assumptions that ohmic contacts are present across semiconducting materials and meta l electrodes, and only one type of charge carrier is present under the sele cted device construction. Therefor e, one should be cautious to apply this model to analyze mobility of a materi al when these two assumptions are not granted. The sandwich device structure is ra ther simple, composed of two el ectrodes that have the same or very similar work function and one layer of organic semiconducting ma terials that can be either nor p-type. When the material is n-chan nel, low-work function metal is chosen to create
78 an injection barrier for holes. For a hole-only device, typical for conjugated polymers, high-work function metal, e.g. gold and platinum, will be used as electrodes to create an injection barrier for electrons. Charge mobility can be extracted from J V characteristics of a de vice in dark. Typically, there are two regions in the J V characteristics. At the low applied voltage, J V characteristics show ohmic response. At the high applied voltage, the J V characteristics exhibit space-charge limited current behavior because charge is in jected only from one electrode. Under ohmic contact conditions, the current J is transport-limited instead of injection-limited and can be expressed as 3 2 08 9 d V J (2-1) In this equation, and d are permittivity and thickness of a material, respectively. Since SCLC diode construction is very similar with the photov oltaic configuration, this model is frequently used and very useful for determining charge mobility of a material that is for photovoltaic applications. It also needs to keep in mind th at the mobility by this method is dependent on the third power of the film thickne ss. Inaccuracy of film thickne ss measurement will lead to a devastating deviation. The FET technique is currently a common meth od to measure charge mobility in organic materials. There are four possible OFET archit ectures, including top contacts/bottom gate, top contacts/top gate, bottom contacts/bottom gate a nd bottom contacts/top gate All four of these OFETs architectures are shown in Figure 2-2. In this dissertation, top contacts/bottom gate is chosen for its structure simplicity to evaluate the charge mobility.
79 Figure 2-2. Four possible FET archit ectures (in cross-section): a) top contact, b) bottom contacts, c) top contacts/top gate, d) bottom contacts/top gate The charge mobility can be extracted from J V characteristics of a FET device. In the linear and saturated regimes, it can be expressed as SD T G SDV V V C L W ) ( I and 2) ( 2 IT G SDV V C L W (2-2) Here ISD and VSD denote the current and voltage bias between source and drain. VG denotes the gate voltage and VT is the threshold voltage at which the current starts to rise. C is the capacitance of the gate dielectric, and W and L are the width and length of the conducting channel. The mobility resulting from the TFT measur ements is often dependent on many other factors, other than the intrinsi c mobility of a semiconducting materi al. The dielectric constant of the gate insulator greatly affects the mobility. The mobility is even sometimes gate-voltage dependent. In a word, it requires extra caution to make a comparison among mobilities values. Another notion is that in-plane mobility al ong a very thin channel is obtained in FET measurements, as opposed to vertical mobility. This makes FET mobility less applicable to photovoltaic devices, where the charges ar e perpendicular across the entire film.
80 CHAPTER 3 LOW BAND GAP PLATIMUN-ACETYLIDE POLYMERS FOR PHOTOVOLTAIC APPLICATIONS 3.1 Introduction Organic -conjugated polymers and oligomers, due their potential in the development of plastic solar cells that are lightwe ight, mechanically flexible and low cost, have drawn significant attention as active materials in organic photovoltaic devices (OPVs).159-161 Since the discovery of rapid and efficient photoinduced electron transfer from poly( 2-methoxy-5-(2'-ethyl-hexyloxy)para -phenylenevinylene (MEH-PPV) to C60 in early 1990s,162-163 the use of fullerene derivatives as electron acceptors in OPVs has become ubiqu itous. The so-called bulk-heterojunction OPVs have been most studied, using a single layer blend of a conjugated polymer and a fullerene derivative such as [6,6]-phenylC61butyric acid methyl ester (PC61BM) or [6,6]-phenylC71 butyric acid methyl ester (PC71BM). In this type of device, a ppropriate phase segregation enables the formation of a bulk-heterojunction at the interface of the donor and acceptor components. Current state-of-the-art single layer conjugated polymer-based OPVs achieve power conversion efficiencies in the range of ~ 7 %, while the highest re ported efficiency of 6.5% is for a tandem solar cell.144-145,164-171 Solar cells with optim ized performance have been developed using P3HT/PC61BM blends,172 and these cells feature nearly ideal photon-to-c urrent quantum efficiency (IPCE) in the mid-visible region. However, to produce highly efficient OPVs light absorption by the active layer must be extended in to the near-infrared region while at the same time preserving the high IPCE and open circuit voltage. In principl e, the overall power conversion efficiency of the devi ce is mainly determined by severa l individual efficiencies: light harvesting efficiency across the visible and near -infrared regions, exciton diffusion to the donoracceptor interface, photoinduced charge separati on, and the mobility of the charge carriers produced by photoinduced charge se paration and charge collection.159
81 In the Reynolds and Schanze groups there is an interest in exploring whether triplet excited states can be harnessed to increase the e fficiency of charge generation in OPV active materials.173-174 In theory, there are several possible reas ons why triplet states are expected to give rise to efficient charge ge neration. First, the long lifetime (up to the s time regime) of the triplet state (triplet exciton) may enhance the probability of ex citon diffusion to a donor-acceptor interface.175-176 Second, quantum mechanical spin-restrict ions prevent charge recombination in the geminate ion-radical pair produced as a result of photoinduced charge transfer from a triplet state precursor.177 Conventionally, triplet stat es are not produced efficien tly as a result of direct photo-excitation of mo st organic hydrocarbon -conjugated materials. Therefore, in order to explore the effects of triplet states in OPVs, it is necessary to utilize materials that incorporate heavy atoms which give rise to efficient intersystem crossing by enhancing spin-orbit coupling.178 Platinum acetylide oligomers and polymers have been widely studied because they represent a class of -conjugated materials featuring high qua ntum efficiency for intersystem crossing (to produce the triplet ex cited states) following direct photo-excitation. This provides considerable insight into the electronic structure and the delo calization and dynamics of the triplet exciton.179-184 Platinum acetylides have also b een used as the active materials in organometallic photovoltaic devices. An early ex ample by Khler and co-workers demonstrated that the photocurrent respons e of a blend of a platinum acetylide polymer and C60 was enhanced relative to the pure polymer.185 Evidence that the polymer triple t state was involved in charge generation came from the observation that C60 only partially quenched the polymers singlet emission (fluorescence) but completely quenched the triplet emission (phosphorescence). More recently, our group reported a photophysical a nd OPV device study that focused on active
82 materials consisting of blends of the pl atinum acetylide polymer p-PtTh and PC61BM.173-174 Luminescence quenching and transient absorptio n spectroscopy indicate that photoinduced electron transfer (PET) from the p-PtTh triplet to PC61BM is efficient. Solar cells constructed using a 4:1 PC61BM/p-PtTh blend as the ac tive material exhibit a peak IPCE of ca. 10% and an overall power conversion efficiency of 0.27%. While the IPCE of the p-PtTh/PC61BM cells is respectable, the overall power conversion efficiency is limited because p-PtTh only absorbs blue light ( max 411 nm), and consequently its absorpti on overlaps poorly with the solar emission spectrum. In a natural continuation of our investigations utilizing the triplet excited state in OPVs, we sought to develop platinum acetylide polymers that absorb light strongly throughout the visible and near-IR regions. It was pr oposed that such materials w ould lead to more efficient photocurrent generation due to in creased light harvesting effici ency while preserving the high quantum yield of intersystem afforded by the pres ence of the platinum heavy metal centers in the -conjugated chain. In order to achieve this obj ective, we designed a series of polymers and model oligomers that feature -conjugated segments of the type donor-acceptor-donor alternating with (or end-capped by) trans -Pt(PBu3)2-units (Figure 3-1). These materials were developed on the basis of prior work demons trating that low bandgap polymers and oligomers can be prepared when the conjugated backbone cons ists of alternating re peat units that have electronic donor and acceptor properties.56,186-193 Three pairs of polymers and model oligomers, together with a copolymer, are the focus of the wo rk presented herein. Th e first pair features a -conjugated segment consisting of the 2,1,3-benzot hiadiazole (BTD) accep tor moiety flanked on either side by 2,5-thienyl donor units ( M-1 and P1 ), whereas in the second pair the BTD unit is flanked by (3,4-ethylene dioxy)-2,5-thienyl donors ( M-2 and P2 ). Both oligomer/polymer
83 pairs absorb strongly throughout the vi sible region; however, because the ethylenedioxythiophene moiety is a stronger donor than thiophene, the latter oligomer/polymer pair has a correspondingly lower bandgap, and therefore harvests light more efficiently in the near-infrared region. The third pair use a much stronger accepto r [1,2,5]thiadiazolo[3,4g]quinoxaline to replace 2,1,3-benzothi adiazole leading to a shif t in the absorption maximum into the near-IR region. Combin ing the repeat units found in P1 and P3 polymer P4 absorbs across the entire visible and Near-IR regions. Figure 3-1. Structures of Pt-acetylide model compounds and polymers. In this chapter a complete report of th e syntheses as well as photophysical and electrochemical characterizati on of a set of low bandgap platinum-acetylide model complexes and polymers is provided. In addition, the ability of P1 to undergo photoinduced electron transfer (PET) to PC61BM from both the single t and triplet excited state manifolds is also
84 investigated. Finally, OPVs based on blends of P1 / P2 / P3 / P4 with PC61BM were characterized under monochromatic and simulated solar (AM 1.5) illumination. The photophysical studies reveal that the materials undergo relatively efficient intersystem crossing. It was observed that the singlet state is quenched efficiently by PC61BM, but the triplet state is not quenched, indicating that charge generati on in the photovoltaic materials must ensue from the singlet manifold. A thermodynamic analysis of PET ba sed on the electrochemical and spectroscopic data indicate that charge transf er from the singlet state of the low bandgap platinum acetylides is thermodynamically feasible but not favored from th e triplet state. Neve rtheless, the materials perform well when used in bulk heteroj unction OPV devices. Operating under AM1.5 conditions, optimized devices exhi bit an open circuit voltage ( Voc) of ~0.5 V, a short circuit current density (Isc) of ~7.2 mA-cm-2, and a fill factor of ~35%, which yields overall power conversion efficiencies of 1.1 to 1.4% for P1 During the course of our studies, a report by Wo ng and coworkers appeared describing the fabrication and characterization of OPVs fabricated with blends of PC61BM and P1 .194-195 The OPVs tested by Wongs group operate with remarkab ly high efficiency: a peak IPCE of 87% at 575 nm is reported and the overall efficiency of a cell operating under AM1.5 illumination is reported to be 4.93%.194 It should be noted that the technical accuracy of these reported values have been examined by others as careful analysis of the results is required for these PV cells.196197 Our complete study provides an interesting and useful complement and contrast to the materials-oriented study by the Wong group, as we have fully characterized the photophysics, electrochemistry and photoredox properties of two structurally -related low bandgap platinum acetylide polymers, as well as characterizing th eir performance in bulk heterojunction OPV devices. An important point is that the device s characterized in our hands operate with lower
85 efficiency compared to the previous report.194 Possible origins for this discrepancy in device performance are considered in our discussion. In addition, it is worth mentioning that our photovoltaic results are consistent with the observations from Jen139 and Jenekhe.198 All three of these reports are more in agreement with th e theoretical simulation carried out by Janssen.141 3.2 Synthesis of Platinum-acetylide Model Complexes and Polymers The general synthetic ro utes used to prepare P1 M1 P2, and M2 are illustrated in Figure 3-2. Compound 3-2 was obtained in 59% by Stille coupli ng of dibromo-2,1,3beozothidiazole ( 31) and 2-tributylstannylthiophene. Upon reacting compound 3-2 with N-bromosuccinimide in DMF, the precipitate was collected by suction filtration and washed with methanol to yield compound 3-3 in 90% yield. Trimethylacetylene moietie s were installed on the both sides of 3-3 through the palladium-catalyzed Sonogashira reaction in high yields to give compound 3-4 which is subsequently deprotected by th e removal of TMS groups to give compound 3-5 in almost quantitative yield. Compound 3-5 was reacted with trans -Pt(PBu3)2Cl2 in the presence of a catalytic amounts of CuI and piperidine in toluene to afford the polymer P1 as a material that is soluble in THF, chloroform and toluene. The phenylethynyl end-capped model oligomer M-1 was obtained in a good yield by reacting 3-5 with mono-capped phenylacetylene chloroditributylphosphine platinum complex in the pres ence of CuI as a catalyst. A similar strategy was used to prepare P2 and M-2 with a slight modification in the procedure. Specifically, due to the fact that removal of the trimethylsilyl protecting groups by reaction of 3-6 with tetrabutyammonium fluoride (TBAF) led to decomposition of the product, a methodology was adopted whereby 3-6 was deprotected in situ Thus, reaction of 3-6 with TBAF in the presence of either trans -Pt(PBu3)2Cl2 or mono-chloroplatinum complex afforded the expected products P2 or M2 respectively.
86 The sequence to prepare P3 and M-3 is outlined in Figur e 3-3. Nitration of 3-1 with a mixture of fuming sulfuric acid a nd fuming nitric acid led to compound 3-7 Due to the existence of structural isomers, the purification of 3-7 which is obtained in low yield, is extremely tedious and difficult. Figure 3-2. Synthesis of M-1, M-2, P-1 and P-2: a) tri butyl(thiophen-2-yl)stannane, Pd(PPh3)2Cl2, THF, 76 %; b) NBS, DMF, 92%; c) trimethylsilylacetylene, Pd(PPh3)2Cl2, CuI and iPr2NH-THF, 90%; d) K2CO3, MeOH, 97%; e) transPt(PBu3)2Cl2, CuI and piperdine-CH2Cl2; f) trans -Pt(PBu3)2Cl2, TBAF, CuI and piperidine-CH2Cl2. After successful insta llation of the flanking thiophene rings, compound 3-8 was reduced with iron powder in acetic ac id. Ring closing was acheived via the acid-catalyzed condensation of compound 3-9 and tetradecane-7,8-dione yield compound 3-10 in high yield. Compound 3-12 was then obtained using a strategy si milar to that employed for compound 3-4
87 Figure 3-3. Synthesis of compound 3-12: a) fuming H2SO4-HNO3, 39 %; b) tributyl(thiophen-2yl)stannane,Pd(PPh3)2Cl2, THF, 76 %; c) Iron-acetic aci d, 65 %; d) tetradecane-7,8dione, p-TSA, CHCl3, 79 %. e) NIS, CHCl3, 87 %; f) trimethylsilylacetylene, Pd(PPh3)2Cl2, CuI and iPr2NH-THF, 84 %; Figure 3-4. Synthesis of M-3, M-4, M5 and P-3: a) 1 eq trans-Pt(PBu3)2Cl2, TBAT, CuI and Et3N-CH2Cl2;b) 5eq trans-Pt(PBu3)2Cl2, TBAT, CuI and Et3N-Ch2Cl2; c) CuI and Et3N-CH2Cl2.
88 Figure 3-5. Synthesis of P-4: a) CuI and Et3N-CH2Cl2. With compound 3-12 in hand, we first attempted to remove the TMS protecting group. Little success was achieved afte r a number of trials using standard deprotection methods including TBAF, TBAF-acetic acid, K2CO3-methanol, CsF, KF, and HF-pyridine. As mentioned for P2 TBAF used in the case of the preparation of P2 leads to the decomposition of 3-12 Therefore, in-situ deprotection and poly merization were carried out to obtain P3 A color change from deep blue to yellow upon addi tion of TBAF indicates the decomposition of compound 3-12 It is likely caused by the basic nature of TBAF in solution Fortuitously, it was found that replacing TBAF with tetra butyammonium triphenyldifluorosilicate ( TBAT ) can circumvent this problem as shown in Figure 3-4. TBAT is a commercially available white crystalline solid. For the preparation of M-3 a different approach was take n than that used to make M-1 and M-2 Compound 3-13 was instead prepared, which served two purposes. First, it was used to prepare M-3 through a reaction with phenylacetylene. Additionally, compound 3-13 can react with compound 3-5 to yield P4 as shown in Figure 3-5. In the course of preparing 3-13 compounds 3-14 and 3-15 were obtained as side products, which were then converted into M-4 and M-5 respectively. This set of m onomer, dimer and trimer provides a very good base to analyze the difference between oligomer s and their corresponding polymer.
89 3.3 Structural Characterizations The intermediate and model complexes M-1 M-2 M-3 M-4, and M-5 were purified by silica gel column chromatography and the 1H, 13C, and 31P NMR spectra of the isolated products are consistent with their molecular struct ures. As an example, Figure 3-6 shows the 31P NMR spectra of 3-13, 3-14, and 3-15 In compound 3-13 there exists only one phosphine peak with two satellites, in accordance with the existence four equivalent phosphene ligands bound to each platinum atom. From the big coupling constant of 2230 Hz, we can also conclude that the two phosphene ligands are in a trans -conformation. In contrast, two p eaks appear in the case of compound 3-14 and 3-15 These peaks can be explained by the existence of two different phosphene ligands, which are situated at internal and terminal positions in the molecules. The polymer samples were purifie d by precipitation from CHCl3 solution into methanol and then subjected to Soxhlet extraction with methanol, hexane, and CHCl3 for further purification. The 1H and 31P NMR spectra of the polymers are in agreem ent with the structure of the materials. Further characterization information can be found at the end of this chapter. Figure 3-6. 31P-NMR spectra of mode l complexes of 3-13, 3-14, and 3-15. Matrix assisted laser desorption-ionizati on (MALDI) mass spectroscopy was carried out on samples P1 and P2 in an effort to obtain additional info rmation concerning the molecular weight
90 distribution of the polymers. This data is s hown in Figure 3-7. The MALDI mass spectrometry was performed using DCTB ((2-[(2E)-3 -(4-tert-butylphenyl)-2-methylprop-2enylidene]malonitrile) as the matrix.199 The P1 MALDI mass spectrum exhibits peaks spanning from m/z = 1,000-15,000 with several ion series observed in the spectrum. Ions in each series are separated by 946 amu, which is consistent with the repeat unit structure of the polymer (see Scheme 1). Above m/z = 4100 there are two main series. The most intense is [(AB)n + 2]+. It is suspected that that the residua l mass of 2 amu could represent 2 hydrogen end groups. However, at this mass the accuracy of the instrument is insufficient to distinguish between linear compounds with 2 hydrogen end groups and cyclic compounds. [(AB)n + 302]+ is the other main series at higher masses. In contrast, the P2 MALDI mass spectrum has ion series spanning only from m/z 1,000-9,000. This may result from high mass discrimination on the part of the instrument or polymer with low molecular weight Ions in each series are separated by 1062 amu, consistent with the repeat un it structure of this polymer. Below m/z 5000, there are related series of [(AB)n + 672]+, [(AB)n + 764]+, and [(AB)n + 855]+. These series correspond to B(AB)n polymers with 2 chlorine end groups, 1 chlorine /1 iodine end group, a nd 2 iodine end groups, respectively (n = 1-5). Above m/z 5000, the predominant series is (AB)n + with n=5, 6, 7, 8. Interestingly, end groups of 2 chlorines, 1 chlori ne/1 iodine, and 2 iodines, similar to the end groups in the lower mass series, do not exist. I should be noted that the iodine content could originate from the CuI catalyst). Gel permeation chromatography (GPC) was performed on polymer samples using chloroform as the eluent and molecular weights are referenced to polystyrene standards. The GPC chromatogram of P1 exhibits a relatively narrow mol ecular weight distribution with a number-average molecular weight (Mn) of 22 kDa and polydispersity index (PDI) of 1.97. By
91 contrast, P2 exhibits a comparable Mn of 33 kDa, but the PDI of the polymer is unreasonably high. The broad PDI observed for P2 is likely due to the fact that this polymer has a lower solubility and consequently tends to aggregate. This premise is supported by the observation that the PDI of P2 obtained by GPC varies with the con centration of the solution used for the analysis. P3 has a number-average molecular weight of 27 KDa and a PDI of 3.5. The broad PDI can be explained by the nature of in-situ de protection and polymerization, which would be expected to give a large molecula r weight distribution. In contrast, P4 has a larger Mn of 35 KDa, but a smaller PDI of 2.6. Figure 3-7. MALDI-TOF mass spectra of P1 and P2. 3.4 Photophysical and Electrochemical Studies 3.4.1 Photophysical Studies The photophysical properties of the model co mpounds and polymers were characterized in THF solution. The absorption and photoluminescence spectra of the first tw o pairs are shown in Figure 3-8, and pertinent phot ophysical data are listed in Table 1. In general, all of the materials feature two primary absorption band s with one appearing in the near-UV or blue region of the visible spectrum and the second at lower en ergy with a wavelength maximum in the 550 650 nm region. Several features emerge upon close insp ection of the absorption data. First, for each
92 model complex/polymer pair, the absorption ba nd maxima are red-shifted for the polymers relative to the monomers. This f eature suggests that there is some excited state delocalization in the polymers relative to the monomers. This effect may arise via molecular orbital delocalization via d -p orbital overlap through the Pt centers.200 Alternatively, the red-shift may result from exciton intera ctions between adjacent -conjugated chromophores in the polymer systems. A second feature is that the band maxima for M-2 and P2 are red-shifted significantly compared the band maxima for M-1 and P1 This feature is consiste nt with the electrochemical bandgap calculations discussed later (Table 3-3), and results from the strong effect of the EDOT donor on raising the HOMO level in BTD-EDOT type compounds. The net result is that the presence of the strong EDOT donor lowers the ba nd gap resulting in a substantial red-shift of both of the optical transitions. P3 has a much stronger acceptor th at pushes the absorption even more into near-IR region. Table 3-1. Photophysical properties of model compounds and polymers Materiala max/nm /cm-1M-1 em/nm f b f/ns TA/ s c Eg/eV d M1 372 40,900 678 0.29 9.3 1.8 2.09 549 28,400 M2 389 69,200 710 0.14 6.4 1.4 1.95 596 43,800 P1 378 45,600 683 0.04 1.1 1.1 2.07 563 43,600 408 55,300 P2 621 51,100 717 0.05 0.9 1.2 1.93 P3 405 46,900 1010 0.16 1.30 800 36,900 (a In THF. b Excited at 510 nm, calculated using Rhodamine B ( f = 0.69)ref201 in ethanol as an actinometer. c In ODCB. d Optical.) All of the materials feature a broad, structure-less (red) phot oluminescence that is Stokesshifted relatively little from the lowest energy absorption band. Quantu m yields and emission lifetimes are listed in Table 3-1. The photolumin escence decay lifetimes are in the range of a few nanoseconds, consistent with assignment of the photoluminescence to radiative decay from
93 the singlet excited state (fluorescence). Despite th e general similarity of the fluorescence from the materials, there are several significant differences that emer ge from close inspection of the data. First, the trends noted for the absorption maxima are mirrored in the fluorescence spectra. Specifically, the fluorescence maxima are red-shifted in the polymers compared to the corresponding monomers, a nd the fluorescence for the M-2 and P2 materials is red-shifted compared to the M-1 and P1 materials. The energy of the re laxed (fluorescent) singlet excited state (Es) is estimated for each of the material s by using the onset wavelength for the fluorescence. The values are listed in Table 3-1 and they range from 2.1 eV for the M-1 and P1 materials to 1.95 eV for the M-2 and P2 materials. P3 is much more red-shifted than P1 or P2 with an Es of 1.3 eV. Figure 3-8. Absorption and emission spectra of a) M-1 and P1, and b) M-2 and P2. The fluorescence quantum yields and lifetimes were determined for the materials in solution. The quantum yields are 3 5 times la rger for the model complexes relative to the corresponding polymers. A similar trend is seen in the fluorescence lif etimes, which indicates that non-radiative decay is more efficient in the polymers. The more efficient non-radiative decay of the singlet state in the polymers lik ely arises from quenching sites (traps) in the polymers caused by interchain aggr egation and also possibly by def ects in the chain structure.
94 A key objective of this work is to explore the nature and photoactivity of the triplet states in low-bandgap platinum acetylide materials. The observation of moderately efficient fluorescence and fluorescence (singlet) lifetim es in the nanosecond range suggests that intersystem crossing in the BTDTh and BTD-EDOT materials is not as efficient as has been observed in platinum acetylide materials that ha ve been previously investigated where singlet lifetimes < 100 ps and triplet yields a pproaching unity have been reported.183,202 Nevertheless, as outlined below, there is clear evidence that photoexcit ation of all of the materials leads to at least moderately efficient population of long-lived triplet states. First, in efforts directed towards observing phosphorescence from the triplet states of the materials, a series of careful photolumin escence experiments was performed using a spectrometer equipped with a liqui d nitrogen cooled InGaAs detector. Emission scans over the wavelength range of 0.85 1.5 m were carried out with the samples in solution and solvent glass at room temperature and at 80 K, respectiv ely. Unfortunately, in all cases no detectable phosphorescence emission could be observed. This re sult is consistent with the previous report of Wong and co-workers who also reporte d the absence of phosphorescence from P1 in the nearinfrared region.194 Despite the inability to observe phosphores cence emission, clear evidence for the production of triplet states upon dire ct excitation of all of the mate rials was obtained by ns s time-resolved absorption spectroscopy. Specifi cally, as shown in Figure 3-9, following a 550 nm pulsed laser excitation degassed dichlorobenzene solutions of M-1 and P1 exhibit relatively strong transient absorption signals throughout the 350 800 nm region. The spectra are characterized by broad absorption over th e entire visible region, peaking at 700 nm, with negative (bleaching) bands that correspond to the ground state absorption of the BTD-Th
95 chromophore. For both materials the transient speci es giving rise to the absorption decays with 1 2 s and is consistent with assignment of the spectrum to the absorption of the triplet excited state. Close comparison of the spectra in Figure 3-9 shows that the transient absorption ( OD) is approximately 3 times larger for the m odel complex compared to the polymer. Given that the spectra were obtained on solutions havi ng matched absorption (concentration) and the same laser power, this difference (although qualitative) suggests that the triplet yield is larger for the model complex compared to the polymer. Intere stingly, this difference is consistent with the fluorescence yield and lifetimes, which indicate that the singlet stat e is quenched in the polymers. Apparently singlet que nching in the polymers leads to a reduction in the triplet yield. Time resolved absorption on the BTD-EDOT mate rials produced very similar results; i.e., a triplet state is observed with a lifetime in the microsecond range and the transient absorption is stronger for M-2 compared to P2 suggesting a higher triplet yi eld in the former. Moving to a lower bandgap P3 the lifetime measured from transient absorption experiments are in the range of tenths of microseconds. Photoinduced electron transfer (PET) and charge separation is a key step in the mechanism by which photovoltaic cells convert optical energy to electrical power.159,162 As noted above, in a previous study we confirmed that the triplet st ate of a platinum acetylide polymer was active in photoinduced electron transfer by demonstrat ing that PC61BM quenc hes the phosphorescence and, more importantly, by observing the products of PET via the time-resolved absorption spectroscopy.173 Similar studies were conducted in the pr esent investigation in order to probe the involvement of the triplet state in PET. Thus transient absorption experiments were conducted with the platinum acetylide model complexes and the polymers in the presence of PC61BM (conc. = 1 mM) in dichlorobenzene solution. Th ese experiments showed that in every case
96 PC61BM does not quench the triplet state, and ion radicals resulting from charge separation were not observed. These results conclusively dem onstrate that for the BTD-Th and BTD-EDOT materials the triplet is not involved in charge separation. With even lower energy gap and lowlying LUMO levels, the triplet excited states on P3 can not be quenched by PC61BM. As a matter of fact, the triplet excited states are lower than the lowest lying LUMO of PC61BM. Figure 3-9. Transient absorption diffe rence spectra of a) M-1and b) P1. Excited at 550 nm with 5 ns pulses. Spectra obtained with an initia l 60 ns delay and with succeeding 1 s delay increments. Arrows indicate the direction of change of spectra with increasing delay time. With the above results in hand, we turned to experiments designed to probe whether PC61BM interacts with the singlet state. For each of the materials tested, addition of PC61BM to dichlorobenzene solutions of the models (or polymers) was observed to quench the fluorescence efficiently. For example, Figure 310 shows the fluorescence spectrum of M-1 in dichlorobenzene with PC61BM added over the concentratio n range 0 2 mM. The inset illustrates the Stern-Volmer quenching plot, wh ich is linear and affords a Stern-Volmer quenching constant KSV = 2.0 x 103 M-1. Given the relatively s hort fluorescence lifetime, the large KSV value suggests that static quenching ma y be the dominant quenching pathway. In parallel quenching experiments it was observed that PC61BM does not quench the lifetime of M-
97 1 a result which confirms that static quenc hing occurs. Taken toge ther, the fluorescence quenching results suggest that addition of PC61BM to M-1 results in the formation of a ground state association complex.203-204 Similar results were obtained when quenching experiments were carried out using M-2 indicating that in this case as well, PET to PC61BM only occurs from the singlet state, and that a ground state association complex is produced between the complex and PC61BM in solution. Figure 3-10. Fluorescence em ission quenching of M-1by PC61BM. The legend shows the concentration of PC61BM in solution, and the plot in the inset shows the SternVolmer plot of Io/I vs. [PC61BM]. The absorption and emission data for M3 M4 M5 P3 and P4 are summarized in Table 32. The model complexes and the polymers have very similar photophysical properties with all compounds having two band absorp tions with maxima around 400 and 800 nm. Similarly, all exhibit a broad emission band in the near-IR region around 1010 nm. From the comparison of M3 M4 M5 and P3 it is clear that electr on delocalization does not extend beyond each repeat unit. In contrast, P4 presents a three-band ab sorption pattern, as show n in Figure 3-11. Its absorption curve appears to be the sum of P1 and P3 suggesting the above-mentioned argument that the conjugation is limited within the each repeat unit.
98 Table 3-2. Photophysical properties of M3, M4, M5, P3, and P4 in ODCB Figure 3-11. UV-vis absorption spectra of P1, P3, and P4. 3.4.2 Electrochemical Studies Cyclic voltammetry (CV) and differential pul se voltammetry (DPV) were performed in order to characterize the accessible redox states of the compounds and to obtain the oxidation and reduction potentials for the re dox processes. When combin ed with the photophysical data ( vide infra ) the electrochemical potentials can be us ed to estimate the driving force for photoinduced electron transfer from the polymers to PC61BM. A summary of the electrochemical data is provide d in Table 3-3 and representa tive cyclic voltammograms of M-1 and P1 are shown in Figure 3-12. The cyclic voltammograms of M-2 and P2 are shown in Figure
99 3-12. Figure 3-12 reveals that all of the materi als exhibit a single, re versible reduction at negative potentials and two reversible oxidation wa ves at positive potentials. As outlined below, all of the waves can be attri buted to oxidation/reduction of el ectrophores concentrated on the conjugated segments that contain the three heterocyclic rings. In particular, the reduction waves occur at ca. E1/2 = -1.75 V (-1.85 V for the EDOT series), which corresponds very closely to the reduction potentials previously re ported for the free oligomers ( i.e. Th-BTD-Th and EDOTBTD-EDOT).205-206 The reduction potentials for the free oligomers Th-BTD-Th and EDOTBTD-EDOT are ca. E1/2 = -1.55 V and -1.73 V, respectively, in V vs. Fc/Fc+. Note that the reduction of the BTD-Th materials occurs at a potential approximately 100 mV less negative compared to that for the BTD-EDOT materials. This difference reflects the influence of the EDOT donor moiety, which slightly raises the energy of the LUMO level in the BTD-EDOT system relative to its position in the BTD-Th system. As shown in Figure 3-12, the platinum acety lide oligomers and polymers also exhibit two oxidation waves that occur at potentials moderately anodic relative to Fc/Fc+. The two waves are reversible for the oligomers, whereas they are mu ch less well-defined for the polymers. Like the reductions, the oxidations arise from electrophores concentrated on the -conjugated organic segments. This assignment is supported by the fa ct that the first oxidation potentials for the oligomers and polymers correspond closely with those of the corresponding free oligomers.205-206 Importantly, the potential for the first oxidation waves for the BTD-Th materials are shifted positive by approximately 300 mV relative to thos e for the BTD-EDOT materials. This trend reflects the stronger donor nature of the EDOT moie ties, which leads to a s ubstantial increase in the energy of the HOMO level in the BTD-EDOT materials and allows these compounds to be more easily oxidized.
100 Figure 3-12. Cyclic voltammograms of a) M-1, b) P1, c) M2, and d) P2 in CH2Cl2 with 0.1 M TBAPF6 as supporting electrolytes, s canned at 100 mV/s. Potentials are referenced to Fc/Fc+ as an internal standard. The oxidation and reduction potentials of th e oligomers and polymers were used to calculate the electroche mical HOMO-LUMO gap (Eg) of the materials and the results are listed in the last column of Table 3-3. As expecte d, the BTD-EDOT materials exhibit a smaller Eg (~1.8 eV) compared to the BTD-Th materials (~2 eV). This di fference reflects the strong donor property of the EDOT moieties, which raises the HOMO levels more than the LUMO levels leading to a decrease in the band gap (relative to the BTD-Th materi als). Note that there is only a small difference in Eg between the model compounds and th e corresponding polymers. This
result in d (e.g., Th These E g ( vide inf r potential potential complex e acceptin g lowers it s energy g a Table 33 3.5.1 O p Si n organo m d icates that t BTD-Th or g values agr e r a ). In addit i s of PC61B M s were foun d e s have a m u g ability of [ s energy ga p a p, even th o 3 Electroch e p tical Prope r n ce the mai n m etallic poly m t he HOMO a EDOT-BT D e e very clos e i on to electr o M were mea s d to be -1.1 0 u ch lower l y 1,2,5]thiadi a p to 1.45 eV o ugh their H O e mical prop e r ties and H o n focus of th i m ers in bulk a nd LUMO a D -EDOT) r a e ly with opti c o chemical c h s ured under t 0 V and -1.4 8 y ing LUMO, a zolo[3,4-g] Interesting l O MO and L U e rties of mo d 3.5 o le Mobilit y i s project w a heterojunct i 101 a re concent r a ther than be c al Eg value s h aracterizat i t he same co n 8 V respecti v around -1. 3 quinoxaline l y, all three m U MO level s d el compou n Solar Cells y of Pol y m e a s on exami n i on solar ce l r ated on a si n ing delocali z s obtained b i on of these c n ditions. Th e v ely. P3 an d 3 eV, due to The low l y m odel com p s are differe n n ds and pol y e r Films n ing the per f l ls, experim e n gle chromo z ed over se v y spectrosc o c ompounds, e first and s e d its corresp o the stronger y ing LUMO p lexes have a n t. y mers f ormance of e nts were al s phore segm e v eral repeat u o pic method s reduction e cond reduc t o nding mod e electronaccordingly a lmost the s a the s o carried o u e nt u nits. s t ion e l a me u t to
102 characterize the optical and charge carrier mob ility properties of the polymers as spin-coated films. For these experiments, neat polymer films were spin-cast on borosilicate glass and polymer/PC61BM blend films were cast onto PEDOTPSS coated ITO substrates. First, experiments were carried out to dete rmine the absorption coefficient of P1 films as an example. In these experiments, a series of four films of different thickness we re fabricated and the absorption and reflectance spectra of the films we re measured. Subsequently, the thickness of the same set of films was determined by atomic force microscopy (AFM). A thin film optical analysis was carried out as desc ribed in the Supporting Informa tion, and this analysis provided estimates for the wavelength dependent optical constants for the polymer, as well as the absorption coefficient spectrum. The absorption coefficient spectrum is corrected for reflectance and is consistent across the set of four films. The absorption spectrum of the P1 film is very similar to that of the material in solution. In particular, two primary bands are observed with max = 383 and 588 nm. Note that these bands ar e red-shifted somewhat from their position in THF solution (378 and 563 nm). Analysis of absorption by the thin films provides absorption coefficient values of = 1.4 x 105 and 1.1 x 105 cm-1 at 383 and 588 nm, respectively. The absorption coefficient at 588 nm corresponds to a value of = 0.51 for the imaginary component of the refractive index. A similar set of thickne ss dependent absorption experiments was carried out on a set of spin-coated films co nsisting of 4:1 (w:w) blends of PC61BM and P1 and the results indicate that the absorption coefficient of the blend film is ~20% lower than that of the pure polymer, with a value of = 2.2 x 104 cm-1 at 576 nm. Using the value for the blend film we comput e a value of = 0.10 0.02 at 576 nm. This is slightly lower than the value for a 4:1 blend of PC61BM and P1 reported by Wong and co-workers which was determined by spectroscopic ellipsometry ( 0.14).194
103 In order to characterize the car rier mobility of films cons isting of the organometallic polymers, hole mobilities were determined by space charge limited current (SCLC) measurements. In these measurements a devi ce configuration consisting of glass/ITO/PEDOTPSS/polymer/Au was used, where the polymer is a sp in-coated film of one of the organometallic polymers. In these measurements the zero-field hole-mobility of P1 and P2 were determined to be 1.4 0.3 x 10-7 and 1.1 0.2 x 10-8 cm2-V-1-s-1, respectively. Note that the hole mobility in both polymers is comparatively low compared w ith poly(3-hexylthiophene) which has a mobility of ca. 10-3 cm2-V-1-s-1,207 Also of note is the fact that the mobility of P1 is an order of magnitude higher than that of P2 .139 However, it is important to point out that the SCLC model is very sensitive to the thickness of the measured film. Given these extremely low mobilities, it is difficult to determine whether the results actually re flect the properties of th e materials or if they are artificially low due to experimental errors. 3.5.2 Solar Cell Studies The utility of P1 and P2 as light absorbing and electron donating material s with electron acceptor PC61BM for photovoltaic cells was investig ated. Devices were constructed on PEDOT/PSS coated indium tin oxide (ITO) glass substrates by spin coating blends from toluene or o-dichlorobenzene solutions. Each ITO substrat e was masked and etched to allow an array of 4 active photovoltaic pixels each having an area of 0.25 cm2. An aluminum electrode was then deposited onto the active layer by thermal eva poration under vacuum. Though typically a thin inter-layer of LiF would also be evaporated, but this re sulted in decreased performance. As noted in the results below, cell optimization required ca reful control of spin co ating speed and sample handling. Each entry reported in Tabl e 3-4 is the average of at leas t 3 different pixels. Note that the entire process of active layer spin-coating an d electrode evaporation we re carried out in the
104 inert atmosphere of a glove box, whereas the phot ovoltaic measurements were obtained with the devices in air and withou t electrode encapsulation. As demonstrated by the I-V characteristics of optimized pixels upon AM1.5 irradiation in Figure 3-13, it is evident that P1 has an enhanced PV pe rformance relative to P2 Examination of the data shows that this difference is mainly due to the increased short circuit current (3-4 mA/cm2 for P2 relative to ~7 mA/cm2 for P1 ). This is born out in the spectral efficiency results of Figure 3-14 where P1 shows a peak efficiency of 36% at 570 nm while P2 shows a broadened response between 500 650 nm with an efficiency of ~ 15%. It is interest ing that both of these polymers demonstrate photovoltaic activity to wavelengths l onger than 700 nm, a desired consequence of using the donor-acceptor-donor triad for long wavelength absorption. The distinct long wavelength p eaking of the absorption spectrum and IPCE response for P1 relative to P2 is likely due to film quality and scattering from the P2 /PC61BM film in which aggregation was indicated in the GPC studies. Figure 3-13. I-V characteris tic curves of a) P1/PC61BM and b) P2/PC61BM photovoltaic cells under AM 1.5 simulated solar irradiation (100mW-cm-2) A close analysis of the results of Table 3-4 for the P1 /PC61BM devices indicates how important processing conditions and film quality are with respect to the performance of the resulting photovoltaic devices. Films prepar ed from both ODCB and toluene could present similarly high AM 1.5 efficiencies of 1.3 to 1.4%. In general, we and others165,208-209 find there is
a sweet too thin, in the Isc mobiliti e Unfortu n found th e suggest t 2). As is the be s invarian t P1 /PC61 these cel l Table 34 Exa m efficienc P1 /PC61 B spot in the light absor p is observed e s exhibited b n ately, in ou r e absolute v a t hat the prof i such, the re l s t way to co m t the Isc and BM system l s is not nea r 4 I-V chara c m ination of ies ranging B M cells a n thickness o f p tion is redu c with an inc r b y these pol r studies we a lues to be s i lometer thi c l ative thickn m pare the r e fill factors a has a stron g r ly as high a c teristics of the PV cell r from 0.2 t n d strongly p f the photoa c c ed. More i m r ease in thic k ymers (vide utilized a t h omewhat in a c kness valu e esses showi n e sults. Whil e a re distinctl y g PV activit y a s reported p P1/PC61BM r esults for t h t o 0.8%. T p oints to th e 105 c tive materi a m portantly i k ness. This supra ) mak h in film pro fi a ccurate. ( E e s in Table 3 n g a doubli n e the cells o p y affected. O y ; however, i reviously.19 4 photovolta i h e P2 /PC61 B T hese resu l e importanc e a ls in these d n the cells u is attribute d ing charge c fi lometer for E x-situ AF M -4 may be u n n g in the set p en circuit v O verall, this i mportantly 4 i c devices B M system s h l ts are not e of film q u d evices. W i u nder study h d to the relat i c ollection di f thickness m M thickness m n derestimat e of cells stu d v oltages are r work demo n in our hand s h ows distin c nearly as c u ality due t o i th films tha t h ere, a drop o i vely low h o f ficult. m easurement s m easuremen t e d by a fact o d ied in Tabl e r elatively n strates that s the respon s c tly lower A M c onsistent a o the aggre g t are o ff o le s and t s o r of e 3-4 the s e of M 1.5 s the g ation
issues i n smaller e separati o Figure 3 Table 35 3.7 M e As the tripl e triplet e x b e favor a from the n this poly m e nergy offs e o n efficienci e 14. Extern a b) P2/PC 6 5 Summary e chanism a n noted in th e e t excited st a x cited state t o a ble. Thus, i platinum p o m er. P3 and e ts betwee n e s are expec t a l quantum e f 6 1BM. of I V char a n d Ener g et i e introducti o a te in organ o o be involve i n this secti o o lymers to P C P4 have i n n LUMO l e t ed. f ficiencies a a cteristics o f i cs of Char g o n, a key obj e o metallic ph o d the energ e o n we consi d C 61BM for b 106 n ferior perf o e vels of the a nd absorpti o f polymer/P C g e Separati o e ctive of thi o tovoltaic m e tics for pho t d er the ener g b oth singlet a o rmance to P se polymer s o n spectra o f C 61BM Sola r o n in the Pt s work is to m aterials. H o t oinduced e l g etics of ph o a nd triplet e x P 1 and P2 s and PC61 B f a) P1/PC61 B r Cells pol y mer/P C probe the i n o wever, in o r l ectron trans o toinduced e l x cited state s Considerin B M, low c h B M blend, a C 61BM Ble n n volvement o r der for the fer (PET) m l ectron tran s s of the poly m g the h arge a nd n ds o f m ust s fer m ers.
107 The thermodynamic driving force for PET from th e singlet or triplet st ate of a polymer to PC61BM is given by the expression, G = ECS Ees (3-1) where ECS is the energy of the charge separated state and Ees is the energy of the singlet or triplet excited state (all states are referenced to the ground state of the polymer and PC61BM).174,210-211 The first term in this expression (ECS) is given approximately by th e difference in the oxidation potential of the donor ( P1 or P2 ) and the reduction potential of PC61BM, i.e., ECS Eox(polymer) Ered(PC61BM).211 By using the Eox values for the polymers determined by electrochemistry (Table 3-2) combined with Ered(PC61BM) = -1.10 V, the estimated ECS values are 1.39 eV ( P1 /PC61BM) and 1.11 eV ( P2 /PC61BM). For the singlet excited states of the polymers, the second term in Eq. 1 can be determined from the onset of the fluorescence emission band, and these values are 2.07 eV ( P1 ) and 1.93 eV ( P2 ). Since the lack of phosphorescence emission precluded direct measurement of the polymers triplet energies, the triplet energy levels are estimated by assuming a singlet-triplet splitting of 0.8 eV,180,212-213 affording values of ~1.3 eV ( P1 ) and ~1.1 eV ( P2 ). Using the estimated values for ECS and Ees in Eq. 1, we can now see that for both polymers the PET is exothermic from th e singlet excited states (-0.57 and -0.63 eV for P1 and P2 respectively), whereas PET is weakly endot hermic from the triplet states (+0.1 and +0.05 eV for P1 and P2 respectively). The energetics for the P1 /PC61BM system is summarized in the Jablonski diagram provided in Figure 3-15. This di agram makes it clear that the ener gy of the singlet state of the polymer is above that of the charge separate st ate, and therefore PET is quite exothermic and consequently is anticipated to be rapid. By c ontrast, since the energy of the triplet state is slightly below that of the charge separated state, PET from the triplet manifold is not expected to
108 occur to any appreciable extent. (The Jablonski diagram for the P2 /PC61BM blend is similar, and therefore is not shown.) The energetic considerations outlined here are in direct accord with the experimental observations. In particular, as outlined above, it was observed that in solution PC61BM does not quench the transient absorption arising from the trip let state of either polymer. However, PC61BM does efficiently quench the polymers fluorescence, which indicates that PET from the singlet state is rapid. We conclude from this that in the active materials of the photovoltaic devices studied in this work, only the singlet state is active in giving rise to charge carriers. Although the triplet excited state of the polymer does not contribute to charge separation, it could influence th e efficiency of charge separa tion indirectly by providing a pathway for direct charge recombination to the lower lying triplet state.214-215 Figure 3-15. Energy level diagram for P1. The performance of the photovoltaic devices based on P1 or P2 blended with PC61BM are summarized in Table 3-5. Significant improvement in efficiency is observed for both devices in comparison to the efficiency of the p-PtTh/PC61BM based device we reported previously.173 The improvement in efficiency arises mainly due to a large increase (5~7 fold) in short circuit current density, which results from the considerably en hanced light absorption efficiency by the low-
109 bandgap P1 and P2 materials. This result clearly demons trates that the strategy to improve the photovoltaic performance via red-shifting the ab sorption profile of the active polymer was successful. The second objective of this work was to harnes s the triplet exciton to generate charge in photovoltaic devices that are based on relativel y low bandgap polymers. However, on the basis of the thermodynamic analysis presented above a nd the solution triplet quenching data, it is evident that even though the triplet state is produced with moderate efficiency in both P1 and P2 due to its low energy the trip let exciton is unable to under go charge separation (PET with PC61BM). This result points to a significant prob lem with the concept of harvesting the triplet exciton in photovoltaic devices. Specifically, for materials in which there is limited singlettriplet mixing, the absorption spectrum is dominated by the spin-allowed transitions from the singlet ground state to th e singlet excited state.216 As a result, when considering charge separation from the triplet state, there is a significan t loss of energy that occurs following photon absorption but before charge separation occurs. (T he energy loss is analogous to the Stokes shift, but it includes losses due to both internal conversion within S1 and intersystem crossing, S1 T1. The amount of energy that is lost is approxima tely equivalent to the singlet-triplet splitting energy, which has shown to be a minimum of 0.7 eV in -conjugated polymers.180,212-213 Thus, if the triplet state is to be active in charge gene ration, the material must have a sufficiently large bandgap to overcome the energy loss associated with the singlet-triplet sp litting. Alternatively an acceptor with a less negative reduction potential (l ower LUMO) than PCBM must be used as the acceptor. For a -conjugated material with a HOMO level comparable to P1 the bandgap must be above ca. 2.1 eV to insure that the triplet exciton is sufficiently energetic to undergo charge separation. An alternative approach to solving th is problem would be to use an organometallic
110 material in which there is a much gr eater degree of si nglet-triplet mixing.216 In this case the direct singlet ground state to trip let excited state transition would be sufficiently allowed so that it contributes to the low-energy absorption of the material. 3.6 Conclusion Two types of platinum acetylide polym ers featuring low bandgap donor-acceptor conjugated chromophores were synthesized and characterized by electrochemical and photophysical methods. The polymers were also us ed to construct organometallic photovoltaic devices when blended with PC61BM as an acceptor and electr on transporting material. The photovoltaic devices based on the low bandgap polymers display considerably improved performance compared to devices based on blends of a wide bandgap (blue absorbing) platinum acetylide polymer. The results suggest that charge separation in the photovoltaic materials occurs with high internal quantum efficiency, but incomplete light harvesting and low carrier mobility limit the overall photovoltaic performance. The photophysical studies of the polymers reveal that although a trip let excited state is produced following light absorption, it is too low in energy to undergo photoinduced electron transfer with PC61BM. Studies carried out in solution dem onstrate that quenching of the singlet state of the polymers by PC61BM is efficient, and this leads to the conclusion that the photovoltaic response of the solid ma terials arises due to charge se paration from the singlet state of the polymer. The results point to the fact that, in order to ha rness triplet excitons for charge separation in low bandgap materials, it would be necessary to manipulate the energy levels of either the polymer or the acceptor. For exampl e, decreasing the reduction potential of the acceptor by 0.1 0.2 eV would make charge sepa ration with the triplet state of either P1 or P2
111 energetically favorable.217 Alternatively, lowering the oxida tion potential of the platinum acetylide polymer without changing the bandgap could accomplish the same goal. As noted in the introduction, wh ile the project described in the present manuscript was underway, a manuscript was published reporti ng that photovoltaic devices containing P1 /PC61BM blends as the active laye r exhibit a peak IPCE of 87% at 580 nm and an overall AM1.5 power conversion effici ency in excess of 4.9%.194 While these results are significant and interesting, in the course of our work with th e same active materials we have been unable to attain comparably high photovoltaic device efficiencies. Specifically, as can be seen in Tables 33, optimization of processing conditions and active layer thickness affords P1 /PC61BM based devices that exhibit peak IP CE of 36 % and overall AM1.5 powe r conversion efficiency of 1.39%. Careful consideration of the available optical data suggests that the result s reported herein for the P1 /PC61BM devices are in accord with expectation given the intrinsic limitations of active layer film thickness and light harvesting efficiency. In particular, consider ation of the absorption coefficient for the P1 /PC61BM blend films ( = 1.2 x 104 cm-1 for the 1:4 P1/PC61BM blend) suggests that the active layer film thickness would need to be > 200 nm in order for a device to operate with an IPCE of > 85% at 580 nm.218 Although the measur ements of the film thicknesses for the active layers in the photovolta ic devices reported herein are believed to underestimate the film thickness, the external qu antum yields and power conversion efficiencies that we have measured are consistent with de vices that operate with a high internal quantum efficiency, but are limited in overall performance due to incomplete light absorption by the active material. More specifically, the results ar e consistent with the expected performance of devices with active layers in the range of 75 125 nm that have high internal quantum efficiency
112 but absorb only 30 40% of the incident light in the 550 600 nm region. While it may be coincidental, it is very interesting to note that the performance reported herein for the P1 /PC61BM devices is in very good agreement with the optical modeling results reported by Janssen and co-workers.196-197 Furthermore, in a very recent manuscript, Jen and co-workers report that an optimized devi ce containing a 4:1 blend of P1 /PC61BM as the active layer exhibited an AM1.5 power conversion efficiency of 1.32% with a short circuit current of 4.2 mA cm-2.219 These values are in good agreement w ith the findings reporte d herein, supporting the suggestion that the performance of the P1 /PC61BM devices is limited by incomplete light absorption by the active layer and by the low hole mobility which in turn limits the efficiency of cells with a thicker active layer. 3.7 Experimental Details 4,7-dibromobenzo[ c ][1,2,5]thiadiazole (3-1): benzo[ c ][1,2,5]thiadiazole ( 27 g, 198 mmol) was dissolved in 300 mL of HBr (47%). Then, a solution of Br2 ( 30 ml, 592 mmol) dissolved in 300 mL of HBr (47%) was added dropwise. After complete addition of the Br2 the solution was refluxed at 100-110 oC overnight. After the mixture was cooled down to room temperature, a saturated aqueous solution of Na2S2O3 was added to completely consume the excess Br2. The precipitates were collected by suction filtration and washed with water, acetone, and ether The product was purified by recrystalli zation from THF to yield white crystals (52.4g, 90 %) 1H NMR (300 MHz, CDCl3, ): 7.71 (s, 2H), 13C NMR (75 MHz, CDCl3, ): 152.81, 13217, 113.64. 4,7-di(thiophen-2-yl)benzo[c][1,2,5]thiadiazole (3-2): 4, 7-dibromo-2, 1, 3benzothiadiazole ( 2) (2.94 g, 10.0 mmol), tributyl(2-thien yl)stannane (8.20 g, 22 mmol) and PdCl2(PPh3)2 (140 mg, 2 mol%) were added to Schlenk flask (100 ml) charged with DMF (50
113 ml). The mixture was, after three fr eeze-pump-thaw cycles, heated to 85 oC under argon overnight. After the mixture was cooled to room temperature, aqueous potassium fluoride was added. The solid was filtered and the filtrate was extracted with CH2Cl2 (3 x 50 mL). The combined organic phases were wash ed with brine, dried over MgSO4, and solvent removed under reduced pressure. The re sidue was purified by column chromatography on silica gel eluting with CH2Cl2hexane, 1 : 1. The product was obtaine d as an orange-red solid after removal of solvent (1.75 g, 59 %). 4,7-bis(5-bromothiophen-2-yl)ben zo[c][1,2,5]thiadiazole (3-3) : To a 100 mL flask charged with compound 3 (1.50 g, 5.0 mmol) in CHCl3 (25 mL), N-iodosuccinimide (NIS, 1.92g, 10.8 mmol) was added under a stream of argon. A catalytic amount acetic acid (0.2 ml) was injected through the septum The resulting mixture was stirred at room temperature overnight. The dark red precipitate formed wa s collected by suction filtration, washed with water, methanol, and chloroform. The product was isolated as dark, red crystals after recrystallization from ch loroform (2.05 g, 89.6 %). 4,7-bis(5-((trimethylsilyl)ethynyl)thiophe n-2-yl)benzo[c][1,2,5 ]thiadiazole (3-4) : To a 100 mL Schlenk tube, piperdine-THF (v:v, 1:4, 50 ml) co-solvent was added and degassed for 30 min. by bubbling argon. Compound 4 (2.29g, 5 mmol), Pd(PPh3)2Cl2 (70 mg, 0.05 mmol) and CuI (38 mg, 0.2 mmol) were added subsequen tly under argon. Trimethylsilylacetylene (1.23 g, 12.5 mmol) was injected through the septum and the reaction refluxed overnight. Upon cooling to room temperature the solvent was removed under reduced pressure and the residue was purified by flash chromatography on silica gel eluing with CH2Cl2 and hexane (1:9, silica gel), 2.26 g obtained (91.8 %). 1H-NMR (300 MHz, CDCl3): 7.96 (d, J = 4 Hz, 2H), 7.85 (s, 2H), 7.30 (d, J = 4 Hz, 2H), 0.275 (s, 18H); 13C-NMR (75 MHz, CDCl3): 152.4, 140.4, 133.4, 127.1, 125.6,
114 124.7, 101.1, 97.6, 77.2, 0.15; HRMS (ESI) Calculated for C24H25N2S3Si2 (M + H+): m/z 493.0712. Found: m/z 493.0724. 4,7-bis(5-ethynylthiophen-2-yl)b enzo[c][1,2,5]thiadiazole (3-5) : Compound 5 (1.24g, 2.5 mmol) was dissolved into methanol and TH F (volume?). KOH (560 mg., 10 mmol) aqueous solution was then added. The resulting mixtur e was stirred under room temperature and monitored by TLC. After compound 5 was consumed, the solvent was removed under vacuum. Deionized water (10 mL) was adde d followed by extraction with CH2Cl2 (3 x 10 mL). The combined organic phases were washed with brine and dried over MgSO4. Removal of the solvent gave the desired product as a red solid (845 mg, 97 %). 1H-NMR (300 MHz, CDCl3): 7.98 (d, J = 4 Hz, 2H), 7.87 (s, 2H), 7.36 (d, J = 4 Hz, 2H), 3.49 (s, 2H); 13C-NMR (75 MHz, CDCl3): 152.3, 140.6, 133.9, 127.1, 125.7, 125.6, 123.5, 83.1, 77.2; HRMS (ESI) Calculated for C18H9N2S3 (M + H+): m/z 348.9922. Found: m/z 349.9944. 4,7-bis(2,3-dihydrothieno[3,4-b][1,4]di oxin-5-yl)benzo[c][1,2,5]thiadiazole: To a solution of compound 4 (1.47 g, 5 mmol) and compound 5 (4.27 g, 14 mmol) in DMF (50 mL), PdCl2 (35 mg, 2 mol %), [tBu3PH]BF4 (116 mg, 4 mol %), CuI (10 mol %, 95 mg) and CsF (10 mmol, 1.50 g) were added. The resulting mixture was refluxed overnight under argon. After the mixture was cooled to room temperature, deio nized water (400 mL) was added. The product was then extracted by CH2Cl2 (3 x 200 mL). The combined organic phases was subsequently passed through a a pad of celite, washed with brine and dried over MgSO4. The solvent was concentrated under reduced pressure to afford a dark red solid. The solid was collected by suction filtration and washed with hexane to give the title compound in high purity (1.54 g, 74 %). 1H-NMR (300 MHz, CDCl3): 8.39 (s, 2H), 6.65 (s, 2H), 4.31-4.40 (m, 8H); HRMS (ESI) Calculated for C18H12N2O4S3Na (M + Na+): m/z 438.9953. Found: m/z 438.9937.
115 4,7-bis(7-bromo-2,3-dihydrothi eno[3,4-b][1,4]dioxin-5-yl)b enzo[c][1,2,5]thiadiazole: To a 100 mL flask charged with compound 6 (1.67 g, 4.0 mmol) in CHCl3 (350 mL), Niodosuccinimide (NIS, 1.80 g, 10 mmol) was added. A catalytic amount acetic acid (0.2 mL) was added. The resulting mixture was stirred under ro om temperature for 2 days. The precipitate formed was collected by suction filtration and wa shed with water, methanol, and chloroform. The desired product was obtained as a dark purple solid (2.10 g, 91 %). 1H-NMR (300 MHz, C2D2Cl4, 80 oC): 4.38-4.46 (m, 8H); HRMS (CI) Calculated for C18H10Br2N2O4S3 (M+): m/z 571.8164. Found: m/z 571.8218. 4,7-bis(7-((trimethylsilyl)ethy nyl)-2,3-dihydrothieno[3,4-b ][1,4]dioxin-5-yl)benzo[c] [1,2,5]thiadiazole (3-6) To a 100 mL Schlenk tube, piperdine-THF (v : v = 1:5, 60 mL) was added and degassed for 30 min. Compound 7 (1148 mg, 2 mmol), Pd(PPh3)2Cl2 (112 mg, 0.16 mmol, 4 mol %) and CuI (76 mg, 0.4 mmol, 10 mol %) were a dded subsequently under argon. Trimethylsilylacetylene (588 mg, 6 mmol) wa s injected through the septum. The resulting mixture was refluxed overnight. Upon cooling to room temperature the solvent was removed under reduced pressure and the re sidue was subjected to flash ch romatography on silica eluting with CH2Cl2 and hexane (1:9). After removal of solv ent, the desired product was obtained as purple solid (766 mg, 63%). 1H-NMR (300 MHz, C6D6): 8.50 (s, 2H), 3.40-3.36 (m, 8H), 0.23 (s, 18H); 13C-NMR (75 MHz, C6D6): 152.4, 145.2, 139.9, 126.8, 124.1, 115.6, 103.9, 101.9, 96.7, 64.4, 64.0, 0.03. HRMS (ESI) Calculated for C28H29N2O4S3Si2Na (M + Na+): m/z 632.0720. Found: m/z 632.0742. P1: Compound 1 (0.20 mmol, 69.6 mg), trans -Pt(PBu3)2Cl2 (0.20 mmol, 134.1 mg) and CuI (1.0 mg) were added to a Schlenk flask. Th e reaction flask was evacuated and backfilled with argon three times followed by addition of pipe ridine-toluene (v:v = 1:1, 8 mL), which was
116 injected through the septum. The mixture was st irred at room temperature for 24 h and then passed through a bed of neutra l alumina to remove catalysts. The solvent was removed under reduced pressure. The dark-purple solid (film) was redissolved into a minimum amount of CHCl3 and the solution poured into cold methanol. The precipitate was collected and the precipitation repeated three times. A fibrous purple produc t (144 mg, 73%) was obtained after the final precipitation. Further purificati on was effected through Soxhlet extraction by methanol, hexane, and CHCl3. 1H-NMR (300 MHz, C6D6): 8.11 (d, J = 4 Hz, 2H), 7.40 (s, 2H), 7.27 (d, J = 4 Hz, 2H), 2.20-2.24 (m, 12H), 1.74-1.77 (m, 12H), 1.4 7-1.59 (m, 12H), 1.03 (t, J = 7.5 Hz, 18H); 31PNMR (121 MHz, C6D6): 5.11(JPt-P = 2336 Hz). GPC: Mn = 22, 000 g/mol, PDI = 1.9. FT-IR (KBr): 2955, 2871, 2884, 2084, 1481, 1440, 801 and 667 cm-1. M1: To a 100 mL Schlenk flask, compound 3 (0.1 mmol, 34.8 mg) and trans ethynylphenylchlorobis(trin -butylphosphine)platinum(II) (0.2 mmol, 147.3 mg) were added. The reaction flask was evacuated and backfilled with argon three times followed by addition of degassed piperidine-toluene (v :v = 1:1, 10 mL), which was in jected through a septum. The resulting mixture was stirred under mild reflux overnight. Upon cooling to room temperature, silica gel was added to the pur ple solution, and the solvent wa s evaporated. The product was purified by column chromatography using ethyl acet ate/hexane (5/95) as the eluent. The desired product M-1 was obtained in 87 % yield after drying (143.9 mg). 1H-NMR (300 MHz, C6D6): 8.09 (d, J = 4 Hz, 2H), 7.63 (d, J = 8 Hz, 4H), 7.37 (s, 2H), 7.24 (d, J = 4 Hz, 2H), 7.17 (m, 4H), 7.01 (t, J = 7.5 Hz, 2H), 2.12-2.19 (m, 24H), 1.6 4-1.77 (m, 24H), 1.39-1.51 (m, 24H), 0.94 (t, J = 7.5 Hz, 18H); 13C-NMR (75 MHz, C6D6): 152.9, 136.7, 132.2, 131.1, 129.9, 128.5, 128.4, 125.7, 125.3, 124.8, 119.0 (t), 110.1, 108.2 (t), 102.5, 26.9 (m) 24.8 (m), 24.5, 14.0 (due to superimposition, one carbon is missing); 31P-NMR (121 MHz, C6D6): 4.93 ( JPt-P = 2356 Hz).
117 HRMS (MALDI-TOF, terthiophene as matrix) calculated for C82H122N2P4S3Pt2: m/z 1744.7010. Found: m/z 1744.7170. P2: Compound 2 (0.2 mmol, 122 mg) and trans -Pt(PBu3)2Cl2 (0.2 mmol, 134 mg) were added to a Schlenk flask with magnetic stir bar. After degassing, piperidi ne-toluene (v:v = 1:2, 15 mL) was injected through the septum at 0 C followed by a solution of teterabutylammonium fluoride (TBAF) (0.5 mmol, 1 M in THF). The color of the resulting solution turned from red to blue and the mixture was stirred for 24 h at r oom temperature. After removing the solvent under reduced pressure, the blue solid (film) was dissolved into CHCl3 and passed through a bed of alumina. The filtrate was concentrated and precipitated from methanol. A blue solid was obtained (177 mg, 83%). Further purification was effected through Soxhlet extraction by methanol, hexane, and CHCl3. 1H-NMR (300 MHz, C6D6): 8.68 (s, 2H), 3.57-3.52 (m, 8H), 2.342.30 (m, 12H), 1.81-1.78 (m, 12H), 1.64-1.54 (m 12H), 1.00 (t, J = 7.2 Hz, 18H); 31P-NMR (121 MHz, C6D6): 4.20 ( J = 2365 Hz). Mn = 33, 000 g/mol, PDI = 16. M-2: To a 100 mL Schlenk flask, compound 8 (30.5 mg, 0.05 mmol) and trans ethynylphenylchlorobis(trin -butylphosphine)platinum(II) (0.1 mmol, 73.6 mg) and CuI (19 mg) were added. After degassing, pipe ridine-toluene (v:v = 1:1, 10 ml) was injected through the septum. Upon addition of TBAF (0.2 mL, 1 M in THF), the solution turned from red to purpleblue and the resulting mixture was stirred under ambient temperature overnight. The solvent was evaporated under reduced pressure and the product purified by co lumn chromatography on silica using hexane:ethyl acetate (3:1) as the eluent. The desired product M-2 was obtained in 74 % yield (69.5 mg). 1H-NMR (300 MHz, C6D6): 8.67 (s, 2H), 7.63(d, J = 7.6 Hz, 4H), 7.19 (m, 4H), 7.17 (t, J = 7.0, 2H), 3.51-3.48 (m, 8H), 2.23-2.15 (m, 24H), 1.70-1.64 (m, 24H), 1.54-1.42 (m, 24H), 0.95 (t, J = 1 Hz, 36H); 13C-NMR (75 MHz, C6D6): 153.0, 141.4, 140.5, 131.2, 130.1,
118 126.1, 125.2, 124.0, 121.2 (t, J = 14.7 Hz), 111.5, 110.4, 109.1, 109.1, 108.9(t, J = 14.7 Hz), 99.8 (t, J = 1.8 Hz), 64.6, 64.0, 26.9 (m) 24.7 (m), 24.5 (m), 14.1; 31P-NMR (121 MHz, C6D6): 4.48 ( JPt-P = 2360 Hz). HRMS (MALDI-TOF, terthiophene as matrix) calculated for C86H128N2O4P4S3Pt2 (M+): m/z 1863.7292; Found: m/z 1863.7432. 4,7-dibromo-5,6-dinitro-2,1,3-be nzothiadiazole (3-7): Compound 1-1 (20 g, 68 mmol) was suspended in conc. sulfuric acid (100 mL) at 0C. Then fuming nitric acid (100 mL) is added over 3h. The mixture is stir red for an additional 2 h at room temperature before poured into ice water. The precipitate is collected by su ction filtration and washed with water. The solid was dissolved in THF (30 mL) and adsorbed on SiO2. Column chromatography with a toluene/hexane mixture (1:1) affords a yello wish solid (10.43 g, 39%). HRMS (ESI TOF) m/z calcd. for C6Br2N4O4S: 383.7987 (M+) found 383.7942 (M+). 5.6-Dinitro-4,7dithien-2yl-2,1,3benzothiadiazole (3-8): To a solution of 4,7-dibromo5,6-dinitro-2,1,3-benzothia-diazo le (4.05 g, 10.54 mmol) and tibutyl(thien-2-yl)stannane ( 9.05 g, 24,26 mmol) in THF (40 ml), Pd(PPh3)2Cl2 ( 300 mg, 4 mol %) was added. The mixture was refluxed for 4 h. After cooling to room temperat ure, a orange red solid appeared, which was collected by suction filtration, washed with hexane and CH3CN. Column chromatography on silica eluting with CH2Cl2 and hexane (1:1) gave a orange red solid ( 3.14 g, 76 %). Mp: 256 C 259 C; 1H NMR (300 MHz, CDCl3, ): 7.75 (d, 2H), 7.53 (d, 2H), 7.24 (dd, 2H); HRMS (DIPCI) m/z calcd. for C14H6N4O4S3: 389.9551 (M+) found 389.9426 (M+); Anal. calcd. for C14H6N4O4S3: C 43.07, H 1.55, N 14.35 found C 43.19, H 1.48, N 14.21. 5.6-Diamino-4,7dithien-2yl-2,1,3 -benzothiadiazole (3-9): A mixture of 5.6-dinitro4,7dithien-2yl-2,1,3-benzothiadiazole (3.3 g, 8.4 5 mmol) and iron dust (5.31 g, 97.1 mmol) in acetic acid is stirred at 45 C for 5 h. Th e reaction was poured into 5 % NaOH (300 mL) and
119 extracted with CH2Cl2 (500 mL). The organic layer was dried over MgSO4 and the residue was purified by column chromatography on silica eluting with CH2Cl2 to obtain a bright yellow solid (2.21 g, 79 %). Mp: 240 C 245 C; 1H NMR (300 MHz, CDCl3, ): 7.57 (d, 2H), 7.38 (d, 2H), 7.25 (dd, 2H); 13C NMR (75 MHz, CDCl3, ):143.76, 132.32, 128.25, 121.56, 120.34, 120.21, 100.14; HRMS (DIPCI) m/z calcd. for C14H10N4S3: 301.0101 (M+H)+ found 301.0129 (M+H)+; Anal. calcd. for C14H10N4S3: C 50.88, H 3.05, N 16.95 found C 50.54, H 3.04, N 16.85. 6,7-dihexyl-4,9-di(thiophen-2yl)-[1,2,5]thiadiazolo[3,4g ]quinoxaline (3-10): 5.6diamino-4,7-dithien-2yl-2,1,3-benz othiadiazole (2.08 g, 6.3 mmol) and tetradecane-7,8-dione (1.59 g, 6.9 mmol) in CHCl3 (100 mL) was stirred at room temperature for 6 h. Upon removal of solvent under reduced pressure, the remaining residue was purified by column chromatography on silica eluting with toluene:hexane (1:1) to ob tain a blue violet so lid (2.65 g, 80 %). Mp: 114 C -115 C; 1H NMR (300 MHz, CDCl3, ): 8.97 (d, 2H), 7.64 (d, 2H), 7.30 (dd, 2H); 13C NMR (75 MHz, CDCl3, ):157.31, 152.67, 151.43, 135.87, 135.1, 132.65, 130.57, 126.54, 35.42, 31.82, 29.31, 27.86, 22.61, 14.12; HRMS (DIPCI) m/z calcd. for C28H32N4S3: 521.1862 (M+H)+ found 521.1852 (M+H) +; Anal. calcd. for C28H32N4S3: C 64.58, H 6.19, N 10.76, found C 64.51, H 6.21, N 10.57. 4,9-bis(5-iodothiophen-2-yl)-6,7-dihexyl-[1,2,5 ]thiadiazolo[3,4-g]quinoxaline (3-11). To a 50 mL round flask charged with 6,7-dihe xyl-4,9-di(thiophen-2yl)-[1,2,5]thiadiazolo[3,4g]quinoxaline (0.63 mmol, 330 mg) in DMF (20 mL), N-iodosuccinimide (1.28 mmol, 312 mg) was added. The resulting mixture was stirred at room temperature for 6 h. The resulting violet precipitate was collected by suc tion filtration and washed with copious amounts of water and methanol. The solid was then dried under vacuum to give desired product (428 mg, 87 %). Mp: 146 C -149 C ; 1H NMR (300 MHz, CDCl3): 8.64 (d, J = 4 Hz, 2H), 7.36 (d, J = 4 Hz, 2H),
120 2.94 (t, J = 7.5 Hz, 4H), 1.95-2.03 (m, 4H), 1.40-1.55 (m, 12H), 0.97 (t, J = 7.5 Hz, 6H). HRMS (ESI-TOF) Calculated for C28H30I2N4S3 (M + H+): m/z 772.9795. Found: m/z 772.9825. Anal. Calcd for C12H10N4O: C, 43.53; H, 3.91; N. 7.27; Found: C, 42.87; H, 3.84. N. 7.25. 6,7-dihexyl-4,9-bis(5-((trimethylsilyl)et hynyl)thiophen-2-yl)-[1,2,5]thiadiazolo[3,4g]quinoxaline (3-12). To a 25 ml Schlenk tube, iPr2NH-THF (v : v, 1:5, 15 ml) was added and degassed for 30 min. Compound 2 (298 mg, 0.42 mmol), Pd(PPh3)2Cl2 (4 mol %) and CuI (10 mol %) were added subsequently under argon. Trimethylsilylacetylene (96 g, 1 mmol) was injected through the septum. The resulting mixture was stirred overnight at room temperature. The mixture was then concentrated, passed throu gh a short alumina column eluting with ???, and washed with methylene chloride. The product solution was concentrat ed to minimum volume and the product precipitated into methanol. Th e desired product, a blue solid, was collected by suction filtration and washed with methanol (230 mg, 84 %). 1H-NMR (300 MHz, C6D6, 7.16 ppm): 8.69 (d, J = 4 Hz, 2H), 7.48 (d, J = 4 Hz, 2H), 2.83 (m, 4H), 1.93 (m, 4H), 1.40-1.48 (m, 12H), 0.95-1.00 (t, 6H), 0.32 (s, 18H); 13C-NMR (75 MHz, C6D6, 128.39 ppm): 157.4, 150.36, 137.24, 134.32, 132.94, 132.35, 128.03, 119.56, 101.64, 99.12, 35.83, 32.11, 29.72, 27.93, 22.94, 14.41, 0.25.1. HRMS (ESI) Calculated for C38H48N4S3Si2 (M + H+): m/z 713.2652. Found: m/z 713.2664. Anal. Calcd for C38H48N4S3Si2: C, 64.00; H, 6.78; N. 7.86; Found: C, 63.80; H, 6.66. N. 7.64. P3: Compound 3 (0.1 mmol, 71.32 mg), trans-Pt(PBu3)2Cl2 (0.1 mmol, 67.06 mg) and CuI (0.2 mmol, 19 mg) were added to a Schlenk flask with magnetic stir ba r. After evacuation and backfilling with argon three times, Et3N-CH2Cl2 (degassed with argon) was injected through the septum at 0 oC (ice-water bath). TBAT (0.4 mmol, 216 mg) was then added. The mixture was stir for 24 h at room temperature. After rem oving solvent under reduced pressure, the remaining
121 green solid (film) was dissolved in CHCl3 and passed through a short alumina column. The collected solution was concentrated and then precipitated into metha nol. A green solid was obtained (77 mg, 66%). Further purification was affected by Soxhlet extraction with methanol, hexane, and CHCl3. 1H-NMR (300 MHz, C6D6, 7.16 ppm): 8.93 (d, 2H),7.05 (d, 2H), 3.13-3.15 (m, 4H), 2.55-1.95 (m, 14H), 1.25-1.70 (m, 44H), 1.00-0.75 (m, 24H) 31P-NMR (121 MHz, C6D6): 4.77 ( J = 2894 Hz). Mn = 27 Kda, PDI = 3.5.
122 CHAPTER 4 VINYLENE-LINKED DONOR-ACCEPTOR POLYMERS FOR PHOTOVOLTAIC APPLICATIONS 4.1 Introduction As pointed out in Chapter 3, solution-processable -conjugated polymers have gained considerable interests as active materials in orga nic optoelectronics, such as organic solar cells (OPVs), light-emitting diodes (OLED), thin-film transistors (OTFTs) and electrochromic devices. 220-223 Of particular interest has been th e study of various PPVs [poly(p-phenylene vinylenes)] and regioregular P3ATs [poly(3alkyl-thiophenes)] derivatives, which exhibit remarkable optical and electronic properties. PTVs [poly(thienylene vinylene)s], another class of conjugated polymers with a simila r structure repeat unit, have also drawn a great deal of attention. A variety of methods have been developed for ma king thienylene vinylene-based conjugated polymers, including Gilch reacti on, acyclic diene metadissertation (ADMET), Witting/Witting-Horner reaction, Stille cross-c oupling, Heck reaction, Kumada coupling as well as TiCl4/Zn-promoted coupling. 224-226 Unfortunately, it shoul d be emphasized that no versatile synthetic route has been available fo r making highly regioregular PTVs (with regioregularity higher than 95 %) so far. The highe st regio-regularity wa s around 90 % claimed by McCullough through Stille coupl ing between 2,5-dibromo-3dodecylthiophene with (E)-1,2(bistributylstannyl)ethylene),227 which was reported earlier to ge nerate a region-random PTV. It is now accepted that struct ural homogeneity of polymer ch ains plays a crucial role in device performance, particularly in OTFTs and OPVs. This has been well-examined in the case of poly(3-hexyl-thiophene) that re gioregular P3HTs have much hi gher electrical conductivities, more improved charge carrier mobilities and optical responses than their regio-irregular counterparts. The bulk-heterojunc tion solar cell based on regior egular P3HTs has reached PCEs of 5 %, a milestone in the history of OPVs.126 Considering structure resemblance with P3HTs,
123 P3HTVs with broader absorption are expected to show enhanced solar cell efficiency. In contrast, only less than 1 % efficiency has been obtained for PTVs-based solar cells.228 So it is very intriguing to design a synt hetic route which can lead to PT Vs with high structure regularity. It will be not only beneficial for PTVs-based solar cells, but also ot her PTVs-based organic optoelectronics. Typical donor-acceptor conjuga ted polymers are made of electron-deficient nitrogencontaining aromatic heterocycles (2,1,3benzothiadiazole, [ 1,2,5]thiadiazolo[ 3,4g ]quinoxaline, and benzo[1,2c ;3,4c ]bis[1,2,5]-thiadiazole, etc.) as e ffective acceptors and electron-rich derivatives (thiophene, alkxoybenzene, carba zole and fluorene etc.) as donors.229-231 Vinylenelinked donor-acceptor (VDA) conjugated polymer s are low bandgap polymers that have a vinylene linkage between the donor and acceptor gr oups. An advantage of incorporating these vinylene linkages is that they allow to planarization of the polymer backbone by eliminating torsional interactions between donor and accep tor rings, thus extending conjugation length, which could lead to a decreased bandgap. In addition, the introduction of these vinylene groups into the polymer backbones provides rotational fl exibility which partially increases polymer solubility, allowing chromophore con centration to be increased (def ined as the molecular weight ratio of conjugated backbone to solubilizing side chains).232 Currently, this class of polymers is still largel y unexplored, in contrast to the extensive studies of fully heterocyclic donor-acceptor alterna ting conjugated polymers. This is perhaps, to a great extent, due to the synthetic challenges The well-established Gilch and Witting-Horner routes to PPVs and PFVs as mentioned above, are not suitable for VDA conjugated polymers,224226 since the above mentioned acceptors are usua lly prone to decompose under strong basic conditions. Stille coupling reacti on of dibromoaryl compounds with commercially available (E)-
124 1,2-bis(tributylstannyl)ethene is also us ed to prepare poly(arylenevinylenes)233. However, to install bis(stannylvinyl) functi onal groups on either a donor or an acceptor is not an easy task. Only a few bis(stannylvinyl)a rylene compounds have been reported through hydrostannation.234 Very recently, a boron-protected haloalkenylbor onic acid building block has been reported, which provides a possibility to install a vinylboronic acid functional group.235 This potential Suzuki polymerization route still has a drawback that vinylboronic acid derivatives are usually unstable.236 An alternative means to overcome the probl em has been reported by preparing PPVs via making bromovinylarylene derivatives,237-239 However, this method has not been extendable to prepare VDA conjugated polymers. A notable approach in the litera ture to VDA conjugated polymers is to use Heck polymerization.240-242 Unfortunately, the Heck reaction is notoriously known for producing structural -vinyl cross-coupling defects, 243-244 which even in small amount play a detrimental role in organic optoelectronic devices 245. In this chapter, we report our attempts to prepare defect-free PT Vs and VDA conjugated polymers via a new strategy using a two-step consequent Heck and Hunsdiecker reaction. Poly(3-hexyl thienylenevinylene) ( P3HTV ) and the vinylene-linked benzothiadiazole-thiophene ( PTVBT ) polymer have been chosen to demonstrate the chemistry. For the purpose of structural elucidation, its model compound bisTVBT has also been prepared. Electrochemical studies were performed and photovoltaic cells were fabricated using PTVBT. Chemical structures of them are shown in Figure 4-1. Figure 4-1. Chemical structures of P3HTV, PTVBT and bisTVBT
125 4.2 Synthesis of P3HTV and PTVBT The attempted Synthesis of P3HTV is outlined in Figure 4-2. Starting from 3hexylthiophene ( 4-1 ), 2-bromo-3-hexylthiophene ( 4-2 ) was obtained in good yields using NBS as a bromination source in DMF. Subsequently, palladium-catalyzed Heck reaction was carried out reacting 4-2 with acrylic acid, yi elding (E)-3-(3-hex ylthiophen-2-yl) acrylic acid ( 4-3 ) in 95%. The trans -configuration is confirmed by 1H-NMR, showing a large coupling constant of 16 Hz. Conversion of carboxylic acid group into vinyl bromide was successfully achieved using revised Hunsdiecker reaction, giving the (E)-2-(2-bromovinyl)-3 -hexylthiophene ( 4-4 ) in 48%. The attempts to access the proposed monomer ((E)-2-(5-(2-brom ovinyl)-4-hexylthiophen-2-yl)4,4,5,5-tetramethyl-1,3,2-dioxaborolan e), however, failed after ma ny trials. The successful preparation of 4-4 has opened the door for the study of vinylene-linked donor-acceptor conjugated polymers. Figure 4-2. Synthesis of regior egular poly(3-hexylthienylenevi nylenes) (P3HTV). a) NBS, DMF, 88%; b) acrylic acid, Pd(OAc)2, P(o-tol)3, Et3N-CH3CN-THF, 95%; b) NBS, LiOAc, CH3CN-H2O, 48%. The Synthesis of PTVBT is shown in Figure 4-3. Kumada reaction of 3, 4dibromothiophene with 1-bromooctane gave compound 4-6 in 88 % yield, which after Ircatalyzed borylation afforded compound 4-7 in 81 % yield.246 Compound 4-7 is one the monomer that will be used in polymerizati on via Suzuki coupling. Heck coupling of 4, 7-
126 dibromobenzo[c][1,2,5]thiadiazole directly with acrylic acids afforded compound 4-8 in 65 % yield. Compound 4-8 was poorly soluble in most common solvents. Some exceptions are THF, DMF and DMSO. The poor solubility makes it di fficult to purify through flash chromatography. Fortunately, its triethylamine salt was soluble in water. Hence the purification of compound 4-8 was readily achieved by forming the triethylamin e salt, following by acidification with aqueous HCl. The catalytic Hunsdi ecker reaction of compound 4-8 with NBS in the presence of lithium acetate as catalyst afforded compound 4-9 in 64 % yield as light yellow crystals, which appear to be green fluorescent. Low yield or no reacti on was noticed when triethylamine or potassium acetate were used instead of lithium acetate. Theref ore, it is possible that the small cationic ion size (Li+) has pronounced effect on the transition st ate of decarboxylation. In addition, the choice of solvents also played a critical role in this reaction. It turned ou t that a mixture of acetonitrile and water gave the best results when their volume ratio was around 20 to 25 %, while it was usually around 3 % for other catalytic Hunsdiecker reactions in literature.247 Compound 49 was obtained in 71 % overall yield through two steps. PTVBT was prepared by a Suzuki polycondensation between compound 4-7 and 4-9 in 87 % yield after Soxhlet ex traction and precipitation.248 The polymer obtained from the chloroform fraction was soluble in THF, toluene and chlorina ted solvents (> 10 mg/mL in chloroform). GPC analysis revealed that PTVBT had number-average molecu lar weights ranging 20,000 to 31,000 Da and polydispersity indices fr om 1.7 to 2.4 on the base of batch to batch. As a model compound, bisTVBT was prepared by a similar synthetic route by Suzuki c oupling of compound 4-9 with 2-(3,4-dioctylthiophen-2-yl )-4,4,5,5-tetramethyl-1,3,2-dioxabo rolane. In this reaction, a different catalytic sy stem containing Pd2(dba)3, [(tBu3)PH]BF4, and CsF is used, showing the versatility of Suzuki coupling.
127 Figure 4-3. Synthesis of PTVBT and bisTVBT. a) 1) C8H17Br, Mg, Et2O; 2) Ni(dppf)Cl2, Et2O, reflux, 88% b) 4,4,4',4',5 ,5,5',5'-octamethyl-2,2'-b i(1,3,2-dioxaborolane) [Ir(OMe)COD]2-dtbpy Heptane, 50 oC, 81%; c)Pd(OAc)2, P(o-tol)3,Et3N, CH3CNTHF, reflux, 65%; d) NBS, LiOAc, CH3CN-H2O, rt, 60%; e) Pd2(dba)3,P(otyl)3,Et4NOH, toluene-water, 60 oC, 87%; f) 2-(3,4-dioctylthiophen-2-yl)-4,4,5,5tetramethyl-1,3,2-dioxaborolane, Pd2(dba)3, [(tBu3)PH]BF4, CsF, THF-water, rt, 86%; 4.3 Structural Characteriza tions and Optical Studies The structure of PTVBT was confirmed by 1H-NMR, 13C-NMR, IR and elemental analysis. The 1H-NMR of PTVBT is here compared with that of bisTVBT to illustrate the polymer structure, as shown in Figure 4-4. Model complex bisTVBT shows a very clear splitting pattern. Two doublet s are assigned to two trans -vinyl protons, which have a coupling constant of 16.2 Hz. Two singlets (7.50 and 6.84 pp m) are evident from the protons on the benzothiadiazole and thiophe ne rings, respectively. The 1H-NMR of PTVBT appears broad at room temperature and even at 60 oC, a feature common to high mo lecular weight polymers. In addition, the existence of strong polymer aggrega tion in solution is also responsible for NMR broadening. When heated to 100 oC in deuterated tetrachloroethane, the 1H-NMR of PTVBT clearly exhibits three broad, but distinguish able signals among 7.2 to 8.5 ppm, corresponding respectively to Hb, Ha and Hc in model complex. It is worth mentioning that no 1H-NMR signals are found in the region of 5.0 6.0 ppm, where the 1H-NMR signals of 1,1-diarylenevinylene
128 defects ( -vinyl coupling defects) usually appear in Heck polymerization 22. The presence of the trans -vinylene functionality is also confirmed by the appearance of FT-IR bands at 955 and 947 cm-1 for bisTVBT respectively, as shown in Figure 4-5. These bands are due to the out-of-plane C-H bending of trans -vinylene linkages 22, 249. Figure 4-4. 1H-NMR spectra of bisTVBT in deuterated ch loroform (top) at room temperature and PTVBT in deuterated tetrachloroethane at 100 oC Figure 4-5. IR spectra of bisTVBT and PTVBT Thermogravimetric analysis (TGA) was ca rried out to evaluate the thermal stability of PTVBT A mass loss of 5 % is defined as the threshold for thermal decomposition. PTVBT
129 demonstrated good thermal stability with an onset of decomposition at 380 oC. Differential scanning calorimetry (DSC) was performed to characterize the thermal transitions of PTVBT. PTVBT exhibited a glass transi tion-like feature at 100 oC and a broad endothermic transition at 235 oC on the forward sweep of the first DSC hea ting cycle. No thermal transitions were observed after the first cycle in the range of -100 to 300 oC. The absorption spectrum of PTVBT exhibits a significant change in comparison with that of bisTVBT BisTVBT has a two-band spectrum with absorp tion maxima at 359 and 499 nm in THF, while PTVBT shows absorption maxima at 416 and 618 nm as shown in Figure 4-6. There is a ~119 nm red shift in the max of the low energy peaks between PTVBT and bisTVBT In contrast, due to the steric hindrance imparted by the octyl solubilizing gr oups, only a 60 nm red shift is observed for a directly -linked benzothiadiazole-thiophene alternating polymer and its model compound.250 More strikingly, PTVBT has a more red-shifted absorption maximum than its analogy polymer with a th ienylene linkage (Poly(BTDat -Th), which has a max at 602 nm in chloroform.251 Clearly, the red shift in PTVBT indicates the polymer is highly conjugated through the planarizatio n of the benzothiadiazole and th iophene units along the polymer backbone. The absorption maximum of PTVBT on ITO exhibits another 15 nm red shift compared to its absorption maxi mum in THF, due to enhanced interactions in the solid state. In a fluorescence measurement, bisTVBT shows a well-resolved emission spectrum with a maximum at 610 nm. The photoluminescence efficiency of bisTVBT was relatively high at 51 % compared to that of Rhodamine B in THF. Surprisingly, no fluorescence emission was detected for PTVBT with excitation at 618 nm. Varying the concentration of the polymer solutions yielded no further success. This lack of fluorescence emission can be explained by two facts: the formation of polymer aggregates in solution as illustrated by temperature-dependent
130 1H-NMR study and the low bandgap of PTVBT The former weakens fluorescence emission through self-quenching and the latt er leads to favorable non-radia tive pathways for relaxation of excited states, governed by energy gap law.252 The decay rate of exc ited electronic states in a conjugated polymer is calculated for a mode l of a large number of displaced harmonic oscillators. The rate depends exponentially on th e energy difference (gap ) between the initial and final electronic states.253 Considering dramatically lowing of energy gap from bisTVBT to PTVBT and a very small bandgap of 1.4 eV PTVBT exhibits, it is no surprise that bisTVBT is fluorescent and PTVBT presents no florescence. Figure 4-6. Absorption spectra of bisTVBT and PTVBT in THF 4.4 Electrochemical and Spect roelectrochemical Studies Cyclic voltammetry (CV) and differential pul se voltammetry (DPV) were employed to estimate the HOMO and LUMO levels, and along with the band gap of PTVBT as shown in Figure 4-7. Electrochemical measurements were performed in an argon-filled dry box in 0.1M TBAP/PC electrolyte solution for the polymer th in films, which were solution drop-cast on platinum button electrodes. All the potentials have been calibrated versus ferrocene/ferrocenium. Compared to CV, DPV offers be tter sensitivity and reveals sh arper oxidation and reduction potential onsets, which results in an enhanced accuracy when estimating electrochemical band
131 gaps. As illustrated in Figure 4-7a, there is a broad p-type doping process which occurs at potentials in the range of -0.2 to +0.8 V vs Fc/Fc+ with an oxidation onset at +0.27 V, which is more positive than the onset for the polymer oxi dation determined via CV (Figure 4-7b). Based on the electrochemical results, PTVBT showed relatively good air stability. The HOMO energy value was estimated to be between -5.23 and 5.37 eV, based on the assumption that the Fc/Fc+ redox couple is -5.1 eV relative to vacuum254. After 100 potential switchi ng cycles at 50 mV s-1 between -0.2 and +0.9 V the polymer retained 85 % of its peak current. The reductive process was found to be less stable than the oxidative do ping; the current of the reduction peak at -1.72 V decreased significantly after the third cathodic scan. The onset for reduction is around -1.43 V and about ~100 mV more anodic than the co rresponding CV reduction onset, indicating a LUMO energy level between -3.55 and -3.61 eV. Hence, the electrochemically determined band gaps of PTVBT are 1.76 eV from DPV and1.68 eV from CV, respectively. Figure 4-7. a) Differential pulse voltammetry of PTVBT on a platinum working electrode in 0.1M TBAP/PC solution with a step time of 0.02 s, a step size of 2 mV, and amplitude of 100 mV; b) Cyclic voltamme try of PTVBT on a platinum working electrode (0.02 cm2) in 0.1M TBAP/PC solution at 50mV s-1 Spectroelectrochemical measurements can be used not only to determine the band structure of the polymer, but also to directly evaluate its electrochromic properties, which is important from an application standpoint. Figure 4-7 show s the oxidative spectr oelectrochemistry for
132 PTVBT air-brush spray cast on ITO coated glass electrodes in deoxygenated (via argon purge) 0.1M TBAP/PC supporting electrolyte solution. Th e applied potential was increased in 50 mV intervals from +0.2 V to +0.85 V vs Fc/Fc+. For the neutral form of th e polymer (+0.2 V, thick line in Figure 4-8) two absorption maxima at 416 and 633 nm were observed, which correspond to the transitions of the polymer and induce a d eep blue color in the film. Upon oxidation these transitions vanished with the simultaneous formation of a broad p eak outside the visible region at longer wavelengths, as sociated with low energy ch arge carriers (polarons and bipolarons). When the film is completely oxidi zed (+0.85 V), it is converted into a pale blue color due to the remnant absorption be tween 600-800 nm. The optical band gap ( Eg) has been determined by the onset of the absorption for the neutral form of the polymer and relatively low around 1.52 eV. After performing oxi dative spectroelectrochemical studies no significant degradation of the polymer was obser ved (7% loss of the anodic and cathodic peak currents of the cyclic voltammograms compared to those of the freshly prepared and switched polymer film). The optically estimated band gap was found to be smaller than electrochemically determined values, but remain with in a good agreement within 0.02 eV. Figure 4-8. Spectroelectrochemist ry of PTVBT spray cast on IT O/glass from 3 mg/mL solution of the polymer in toluene in 0.1M TB AP/PC between 0.2 and 0.85V in 50mV steps (vs Fc/Fc+). The thick line corres ponds to the neutral state of the polymer at 0.2 V.
133 4.5 Organic Solar Cells Polymer photovoltaic cells were fabricated with a layered struct ure of glass/ITO/ PEDOT:PSS/PTVBT-PCBM blend/Al. Use of LiF di d not have a noticeable effect on the device performance in our case. Characteristic I-V plots of several representa tive photovoltaic cells with different PTVBT to PCBM weight percentages are shown in Figure 49a. As is illustrated with the 10% polymer blend cell, fill f actors ranged from 33 to 55% percen t. Open circuit voltages in these cells ranged between 0.5 to 0.6 V. Thes e values are typical for low bandgap polymers.255256 Simple metal-insulator-metal models suggest that open circuit voltages should depend only on the difference between the electr ode work functions, however it is well known that the polymer HOMO level can influence the maximum open circuit voltage in polymer cells.133 Given the low-lying HOMO level of PTVBT, it is possibl e that further work will result in improved open circuit voltages. To date, short circuit current densities are relatively low, at ~1 mA/cm2. This is possibly due to ble nd film morphologies. Variation of the blend film weight percentage of PTVBT to PCBM resulted in optimal performance with surprisingly low polymer percenta ges as is illustrated in Figure 4-9b. Power conversion efficiencies (PCE) mos tly fell in the range of 0.2-0.3 %. At low polymer percentages the light absorbing capabilities of the cells could be a limiting factor. Increasing the blend film polymer percentage led to improvement in current densities and overall efficiencies as can be seen in the figures. Upon reaching 10-15 % polymer no further improvements in current densities were noted. At higher polymer concentrat ions we noted a gradual decrease in fill factor resulting in an overall drop in PCE. At just 35% polymer the cell performance dropped by a factor of ~3. This indicate s that light absorption by the polymer was not the performancelimiting factor. The variation in current thr ough the bias voltage sweep remained small, especially at low polymer percentages, indicatin g good transport properties. IPCE measurements
134 showed a broad absorption following the polymer spectrum, but low overall external quantum efficiencies, as shown in Figure 4-9c. Figure 4-9. a) A.M. 1.5 J-V characteristics of devices with varied PTVBT/PC61BM weight percentages; b) A.M. 1.5 Efficiencies measured from cells of differing PTVBTPC61BM weight percentages and also differing active layer thicknesse s; c) IPCE of a representative device. Blend film morphologies are possibly th e dominant factor limiting photovoltaic performance. Tapping mode AFM measurements revealed a strong correlation between blend film surface morphologies and polymer percentage A number of differing surface features resulted with small changes in polymer percenta ges, as shown in Figure 4-10. With a blend of just 4 % polymer and 96 % PCBM, the films were very low in RMS roughness. When the polymer percentage was increased to 8%, the resu lt was the formation of small, round pits. At
135 12% polymer many of these pits became quite d eep ~25nm. Further in creases to the polymer percentage led to shallower pits which begin to coalesce. At 20%, th e morphology appeared to be dominated by multiple small valleys. This effect was much more pronounced at 25% where the valleys deepen and lengthen. These vastly differing morphologies may indicate that the polymer was concentrating at the surface of the anode, an undesirable phase separation that may account for the low IPCE and PCE results. Figure 4-10. Tapping mode atomic force microscopy imag es of spin coated blend films from solutions of varying PTVBT:PCBM weight ratios. The z-scale factor is 20. 4.6 Conclusion We have demonstrated a facile approach to synthesize vinylene-linked donor-acceptor (VDA) conjugated polymers with a method that can find broad applicability for preparing polymers for redox active and organic el ectronic applications. Structural -coupling defect-free low bandgap PTVBT has been successfully obtained. UVVis absorption spectroscopy results show that introducing vinylenelinkage can erase steric hind rance imparted by solubilizing groups and planarize the polymer backbone. Spectroelectrochemistry reveals that PTVBT is an electroactive polymer that can be potentially used in electrochr omic devices. The solar cells made of PTVBT: PCBM blends did not give satisfying performance, probably due to low
136 content of PTVBT used in the blend and the unfavorab le phase separation observed. Future work will be directed towards using the vinyle ne chemistry we developed to prepare other Vinylene-linked donor-acceptor (VDA) conjugated polymers with less intense interchain interactions for OPVs and desi gn VDA materials for Near-IR light -emitting diodes. 4.7 Experimental Details 2-bromo-3-hexylthiophene (4-2) .257 To a 100 mL flask charged with 3-hexylthiophene (2.96 g, 17.6 mmol) in DMF (40 mL), N-bromosu ccinimide (3.30 g, 18.5 mmol) was added at 0 oC. Catalytic amount acetic acid (0.2 mL) was added subsequently. The resulting mixture was warmed up to room temperature after 4h and stirre d in the absence of li ght overnight. After the reaction was completed, the mixture was diluted with water and then extracted with diethyl ether. The organic phase was washed brine, dried ove r magnesium sulfate, and concentrated under vacuum to provide crude product. The crude prod uct was purified by ball-to-ball distillation via a Kugelrohr apparatus to yiel d colorless oil (3.80 g, 88 %). 1H NMR (CDCl3, ppm) 7.18 (d, J = 5.1 Hz, 1H), 6.79 (d, J = 5.1 Hz, 1H), 2.58 (t, J = 7.5 Hz, 2H), 1.57-1.53 (m, 2H), 1.29-1.34 (m, 6H), 0.89 (t, J = 6.9, 3 H); 13C NMR (CDCl3, ppm) 142.2, 128.4, 125.3, 190.0, 31.8, 29.9, 29.6, 29.1, 22.8, 14.3. (E)-3-(3-hexylthiophen-2-yl)acrylic acid (4-3) To a 125 mL Schlenk flask charged with CH3CN (35 mL), THF (15 mL) and Et3N (5.6 mL, 40 mmol), 2-brom o-3-hexylthiophene 1 (3.4 g, 13.7 mmol) was added. The resulting solutio n was degassed through bubbling argon for 25 min and then heated up to 60 oC. Under argon flow, Pd(OAc)2 (2 mol %, 61 mg) and trio tolylphosphine (4 mol %, 167 mg) were added. Ac rylic acid (1.44 g, 20 mmol) was then injected through a septum. The resulting mixture was stirred under vigorous reflux for 36 h (shiny palladium mirror was usually formed during the reaction). When it was still hot, the solution was
137 transferred to a 250 mL clean flask and the solv ent was removed under vacuum to yield a brown viscous oil. The yellow precipitates were form ed upon the addition of aqueous HCl (1 M, 100 mL) and were collected and washed with water. The obtained solids then dissolved into aqueous K2CO3 (1 M). The insoluble solids were removed by filtration. Acidification of the filtrate with aqueous HCl (1 M, 100 mL) gave yellow solids, wh ich were collected and washed with water. The desired products were obtained by drying at 85 oC under vacuum for 2h in the yield of 95 %, 3.10 g. 1H NMR (CDCl3, ppm) 7.98 (dd, J1 = 16 Hz, J2 = 0.6 Hz 1H), 7.31 (d (broad), J = 5.1 Hz, 1H), 6.92 (d, J = 5.1 Hz, 1H), 6.21 (d, J1 = 16 Hz, 1H), 2.74 (t, J = 7.5 Hz, 2H), 1.60-1.54 (m, 2H), 1.25-1.40 (m, 6H), 0.90 (t, J = 6.9, 3 H); 13C NMR (CDCl3, ppm) 172.5, 147.8, 137.9, 133.4, 130.4, 128.1, 115.0, 31.8, 31.3, 29.2, 28.8, 22.7, 14.7. HRMS (ESI-TOF) calculated for C13H18O2S (M+H)+: m/z 239.1100. Found: m/z 239.1089. (E)-2-(2-bromovinyl)-3-hexylthiophene (4-4) To a flask charged with (E)-3-(3hexylthiophen-2-yl)acrylic ac id 2 (2.38 g, 10 mmol) in CH3CN-H2O (96 mL 4 mL), lithium acetate dihydrate (0.1 eq, 1 mmol, 102 mg) was added and stirred for 15 min. NBromosuccinimide (1.82 g, 10.2 mmol) was then added portionwise. Bubbles were observed through bubbler, which disappeared after 30 min. The solution was stirred for another 4 h. The solution was diluted with water and brine and th en extracted with diet hyl ether. The organic phase was washed brine, dried over magnesium sulfate, and concentrated under vacuum to provide crude product. The crude product was purified by column chro matography on silica gel using hexane as eluent to give a pale yellow oil in the yield of 48 %, 1.32 g.1H NMR (CDCl3, ppm) 7.22 (dd, J1 = 13.8 Hz, J2 = 0.6 Hz 1H), 7.10 (d (broad), J = 5.4 Hz, 1H), 6.83 (d, J = 5.4 Hz, 1H), 6.21 (d, J1 = 13.8 Hz, 1H), 2.59 (t, J = 7.5 Hz 2H), 1.56-1.51 (m, 2H), 1.25-1.35 (m, 6H), 0.90 (t, J = 6.9, 3 H); 13C NMR (CDCl3, ppm) 141.3, 133.6, 129.7, 129.2, 123.8, 104.6,
138 31.9, 31.0, 29.2, 28.6, 22.8, 14.3. HRMS (ESI-TOF) calculated for C12H17BrS (M)+: m/z 273.0307. Found: m/z 273.0313. (2E,2'E)-3,3'-(benzo[c][ 1,2,5]thiadiazole-4,7-diyl)diacrylic acid (4-8) To a 250 ml Schlenk flask charged with 4,7-dibromobenzo[c ][1,2,5]thiadiazole (5.85 g, 20 mmol), Pd(OAc)2 (2 mol %, 0.8 mmol, 178 mg) and tri(o-toly l phosphine) (4 mol %, 1.6 mmol, 488 mg) under argon, the degassed mixture of CH3CN (120 mL), THF (60 mL) and Et3N (28 mL, 200 mol) was injected through a septum. The resu lting mixture was heated to 60 oC. Acrylic acid (4.32 g, 60 mmol) was then injected slowly. The solids starte d to disappear and the so lution turned orange. The solution was refluxed at 84 oC for 20 h. The reaction was monitored by TLC. A uniform palladium mirror was formed on the wall of the flask during the period of reaction. While the dark solution was still hot, it wa s transferred to a clean roundbottom flask (500 mL).The solvent was concentrated on rotavap. To the concentrated dark viscous mixture was added hexane and diethyl ether (1:1, 200 mL). The resulting precipit ates were collected and washed with hexane and diethyl ether mixture. The dark red solids we re then dissolved into water (1000 mL, with 7 mL Et3N) and stirred for 1 h. The orange solution was filtered and the filtrate was acidified with 1M HCl (100 mL) to give orange yellow precipitates. The preci pitates were collected, washed with water and dried under air overnight. The obtained crude produc ts were dissolved into hot THF (~ 300 mL) and filtered. Hexane was then slow ly added to the filtrate. The precipitates were filtered and dried under vacuum over 60 oC to afford yellow-orange powder (3.55 g, yield: 65 %).1H-NMR (DMF) : 12.6 (bs, 2H), 8.16 (s, 2H), 8.13 (d, J = 15.9 Hz, 2H), 7.52 (d, J = 15.9 Hz, 2H), 13C-NMR (DMF) : 168.6, 154.3, 139.9, 132.3, 129.7, 125.8; HRMS (DIP-CI-MS) Calculated for C12H8O4N2S (M+): 276.0187 Found: m/z 276.0205
139 4,7-bis((E)-2-bromovinyl)benzo[c][1,2,5]thiadiazole (4-9) To a suspension of (2E,2'E)3,3'-(benzo[c][1,2,5]thiadiazole-4,7diyl)diacrylic acid (5 mmol, 1.38 g) in acetonitrile (50 mL) was added a solution of lithium acetate dihydra te (4 mmol, 408 mg) in water (15 mL). N Bromosuccinimide (10.5 mmol, 1.87 g) was th en added. Carbon dioxide gas was observed through bubbler immediately after the addition NB S. The resulting solution was stirred at room temperature for another 2 h after the evoluti on of gas ceased. Water (100 mL) was added. The yellow orange precipitates were collected, wash ed with water and dried under air. The crude product was then purified by silica gel co lumn chromatography, eluting with CH2Cl2/Hexane (4:96) to yield light yellow solids (1.10 g, 64 %). 1H-NMR (CDCl3) : 8.01 (d, J = 13.8 Hz, 2H), 7.43 (d, J = 13.8 Hz, 2H), 7.37 (s, 2 H); 13C-NMR (CDCl3) : 153.1, 133.6, 128.5, 114.5; HRMS (DIP-CI-MS) Calculated for C10H7N2SBr2 (M+): 343.8624 Found: m/z 343.8618; Anal. Calcd for C10H7N2SBr2: C, 34.71; H, 1.75; N, 8.10. Found: C, 34.88; H, 1.59; N, 7.95. 2,2'-(3,4-dioctylthiophene-2,5-diyl )bis(4,4,5,5-tetramethyl-1,3,2-di oxaborolane) (4-7). To a 100 mL two-neck flask charged with 3,4-octylth iophene (10 mmol, 2.73 g ), 4,4-di-tert-butyl2,2'-bipyridine (3.0 mol %, 81 mg) and bis(pina colato)diboron (8 mmol, 2.03 g) were added subsequently under argon. Three cycles of argonvacuum were then applied. Degassed heptane (50 ml) was injected. [Ir(OMe)(COD)]2 (1.5 mol %, 98 mg) was added under argon and the mixture was stirred at 50 oC for 16 h. After cooling down to room temperature, the mixture was filtered and washed with hot hexane. The filtrate wa s concentrated to yield viscous oil. The crude product was purified by silica ge l column chromatography, eluting with Ethyl acetate/Hexane (5:95) to give colorless oil (4.56 mg, 81 %). 1H-NMR (CDCl3) : 2.80 (t, J = 7.5 Hz, 2H), 1.56 1.20 (m, 48H), 0.90 (t, J = 6.3 Hz, 6H); 13C-NMR (CDCl3) : 154.3, 83.6, 32.7, 32.2, 30.0, 29.7, 29.6, 28.7, 25.0, 22.9, 14.3; HRMS (APCI-TOF) Calculated for C32H58B2O4S (M + H+):
140 561.4326. Found: m/z 561.4383. Anal. Calcd for C32H58B2O4S: C, 68.57; H, 10.43; Found: C, 68.69; H, 10.67. Model Complex: To a flask charged with 4,7-bis((E)-2bromovinyl)benzo[c][1,2,5]thiadi azole (1 mmol, 346 mg), Pd2(dba)3 (2.5 mmol, 23 mg), [(tBu3)PH]BF4 (7.5 mol %, 22 mg), CsF (6 mmol 912 mg) under argon atmosphere, 4,4,5,5tetramethyl-2-(octylthiophen-2-yl )-1,3,2-dioxaborolane (2.4 mmol) in THF (40 ml) was injected. The resulting mixture was stirre d at room temperature for 10 h. Water (100 ml) was added and the organic phase was extracted wi th diethyl ether (2 x 20 mL). The combined organic phase was dried over MgSO4. The solvent was removed and the cr ude product was purif ied by silica gel column chromatography, eluting with Dichlorome thane/Hexane (15: 85) to yield red solids. 4,7-bis((E)-2-(3-octylthiophen-2-yl)vinyl)b enzo[c][1,2,5]thiadiazole. (490 mg, 85%) 1HNMR (CDCl3) : 8.35 (dd, J1 = 15.9 Hz, J2 = 0.6 Hz, 2H), 7.49 (s, 2H), 7.29 (d, J = 15.9 Hz, 2H), 7.17 (d, J = 5.4 Hz, 2H), 6.91 (d, J= 5.4 Hz, 2H), 2.81 (t, J = 7.5 Hz, 4H), 1.691.64 (m, 4H), 1.49-1.25 (m, 20H), 0.90 (t, J = 6.3 Hz, 6H); 13C-NMR (CDCl3) : 153.9, 142.7, 137.2, 130.1, 129.2, 127.4, 125.4, 124.0, 123.8, 32.1., 31.2., 29.6, 29.5, 28.8, 22.9, 14.3; HRMS (APCI-TOF) Calculated for C34H44N2S3 (M + H+): 577.2739. Found: m/z 577.2731. C34H44N2S3 (M + H+): 577.2739. Found: m/z 577.2740. Anal. Calcd for C34H44N2S3: C, 70.78; H, 7.69; N, 4.86. Found: C, 70.79; H, 7.87; N, 4.34. 4,7-bis((E)-2-(4-octylthiophen-2-yl)vinyl)b enzo[c][1,2,5]thiadiazole (465 mg, 81%) 1HNMR (CDCl3) : 8.19 (d, J = 16.2 Hz, 2H), 7.54 (s, 2H), 7.34 (d, J = 16.2 Hz, 2H), 7.06 (s, 2H), 6.85 (s, 2H), 2.61 (t, J = 7.5 Hz, 4H), 1.651.61 (m, 4H), 1.40-1.20 (m, 20H), 0.90 (t, J = 6.3 Hz, 6H); 13C-NMR (CDCl3) : 153.9, 144.3, 143.1, 129.0, 128.8, 127.4, 127.1, 124.0, 120.3, 32.1., 30.7., 30.6, 29.7, 29.54, 29.48, 22.9, 14.3; HRMS (APCI-TOF) Calculated for C34H44N2S3
141 (M + H+): 577.2739. Found: m/z 577.2740. Anal. Calcd for C34H44N2S3: C, 70.78; H, 7.69; N, 4.86. Found: C, 70.77; H, 7.58; N, 4.85.. 4,7-bis((E)-2-(3,4-dioctylthiophen-2-yl)vinyl )benzo[c][1,2,5]thiadi azole (689 mg, 86%) 1H-NMR (CDCl3) : 8.38 (d, J = 16.2 Hz, 2H), 7.50 (s, 2H), 7.29 (d, J = 16.2 Hz, 2H), 6.84 (s, 2H), 2.75 (t, J = 6.6 Hz, 4H), 2.55 (t, J = 7.5 Hz, 4H), 1.701.20 (m, 48H), 0.90-0.98 (m, 12H); 13C-NMR (CDCl3) : 154.0, 143.6, 142.0, 137.8, 129.3, 127.4, 126.2, 123.4, 119.4, 32.2., 32.1, 31.3, 30.0., 29.9, 29.8, 29.7, 29.6, 29.56, 29.5, 29.3, 27.3, 22.9, 14.3; HRMS (APCI-TOF) Calculated for C50H76N2S3 (M + H+): 577.2739. Found: m/z 577.2740. Anal. Calcd for C50H76N2S3: C, 74.94; H, 9.56; N, 3.50. Found: C, 75.08; H, 9.56; N, 3.50. PBTVT : In a 25 mL flame dried Schlenk flask, 4,7-bis((E)-2bromovinyl)benzo[c][1,2,5]thiadiazole (346.0 mg 1 mmol) 2,2'-(3,4-dioctylthiophene-2,5diyl)bis(4,4,5,5-tetramethyl-1,3,2-dioxa borolane) (560.5 mg, 1 mmol), Pd2(dba)3 (13.5 mg, 1.45 mmol %), and tri(o-tolyl)phosphine (18.0 m g, 6 mmol %) were disso lved in 12.0 mL of degassed toluene and degassed 20 % aqueous tetraethylammonium hydroxide (3.4 mL, 4.25 mmol). The reaction mixture was vigorously stirred at 60 oC for 24 h and then heated up to 95 oC for 6 h. The solution turned from orange to red, dark purple and finally blue after an hour. After cooling back to 60 oC, 4-iodotoluene (21.8 mg, 0.10 mmol) was added to the mixture. After 2 h, 4,4,5,5-tetramethyl-2-p-tolyl-1,3,2-dioxaborolan e (21.8 mg, 0.10 mmol) was added and the reaction was stirred for anothe r 6 h to complete the end-cap ping reaction. The polymer was purified by precipitation in meth anol/water (10:1), filtered through 0.45 m nylon filter and purified on a Soxhlet apparatus with methanol, hexanes, and chloroform. To the chloroform fraction was added palladium removal regent (E)-N,N-diethyl-2-phenyldiazenecarbothioamide (22 mg, 0.1 mmol). The resulting solution was sti rred for 2 h. The solution was concentrated
142 under reduced pressure, precipitated in methanol (300 mL), filtered thro ugh 0.45 m nylon filter, washed with methanol and dried under vacuum at 60 oC overnight to afford PBTVT (432 mg, 87.4 %) 1H-NMR (C2D2Cl4, 100 oC) : 8.34 (broad, 2H), 7.62(bs, 2H), 7.39 (broad, 2 H), 2.42.9 (broad, 4H), 1.8-1.20 (m, 24), 0.97 (b, 6H); An al. Calcd : C, 73.10; H, 8.21; N, 5.68; Found: C, 72.64; H, 8.07; N, 5.67; GPC Mn = 31,100, Mw = 73,400, and PDI = 2.36. PF8VBT In a 25 mL flame dried Schlenk flask, 4,7-bis((E)-2bromovinyl)benzo[c][1,2,5]thiadiazole (346.0 mg 1 mmol) 2,2'-(9,9-dioctyl-9H-fluorene-2,7diyl)bis(4,4,5,5-tetramethyl-1,3,2-dioxa borolane) (642.57 mg, 1 mmol), Pd2(dba)3 (13.5 mg, 1.45 mmol %), and tri(o-to lyl)phosphine (18.0 mg, 6 mmol %) we re dissolved in 12.0 mL of degassed toluene and degassed 20 % aqueous tetraethylammonium hydroxide (3.7 mL, 4.6 mmol). The reaction mixture was vigorously stirre d 60 oC for 12 h and then heated up to 95 oC for 24 h. The polymer was purified by precipitati on in methanol/water (9:1), filtered through 0.45 m nylon filter and purified on Soxhlet apparatu s with methanol, hexanes, and chloroform. The chloroform fraction was concentrated under re duced pressure, precipitated in methanol (300 mL), filtered through 0.45 m nylon filter and drie d under vacuum at 60 oC overnight to give metallic-texture solids (100.1 mg, 17.4 % ). 1H-NMR (CDCl3) : 8.19 (d, J = 16.2 Hz, 2H), 7.907.6 (m, 10H), 2.16 (bs, 4H), 1.5-1.1 (m, 20H), 1.00.79 (m, 10H); 13C-NMR (CDCl3) : 154.0, 151.9, 141.0, 136.7, 134.0, 129.5, 126.7, 126.0, 123.9, 121.4, 120.0, 55.1, 40.1, 31.5, 29.9, 29.0. 28.9, 23.8, 22.3, 13.7; Anal. Calcd : C, 81.46; H, 8.12; N, 4.86; Found: C, 81.23; H, 8.21; N, 4.30; Mn = 12,900, Mw = 27, 400, PDI = 2.1 and DP = 22. PF10VBT In a 25 mL flame dried Schlenk flask, 4,7-bis((E)-2bromovinyl)benzo[c][1,2,5]thiadiazole (346.0 mg 1 mmol) 2,2'-(9,9-dioctyl-9H-fluorene-2,7diyl)bis(4,4,5,5-tetramethyl-1,3,2-dioxa borolane) (698.7 mg, 1 mmol), Pd2(dba)3 (13.5 mg, 1.45
143 mmol %), and tri(o-tolyl)phosphine (18.0 m g, 6 mmol %) were disso lved in 12.0 mL of degassed toluene and degassed 20 % aqueous tetraethylammonium hydroxide (3.7 mL, 4.6 mmol). The reaction mixture was vigorously stirred 60 oC for 48 h and then heated up to 95 oC for 4 h. The polymer was purified by precipitatio n in methanol/water (9:1), filtered through 0.45 m nylon filter and purified on Soxhlet apparatus with methanol, hexanes, and chloroform. The chloroform fraction was concentrated under redu ced pressure, precipita ted in methanol (300 mL), filtered through 0.45 m nylon filter and dried under vacuum at 60 oC overnight to give metallic-texture solids (354.0 mg, 56 % ). 1H-NMR (CDCl3) : 8.19 (d, J = 16.2 Hz, 2H), 7.907.65 (m, 10H) 2.16 (bs, 4H), 1.5-1.1 (m, 28H), 1.00.79 (m, 10H); Anal. Calcd : C, 81.85; H, 8.66; N, 4.43; Found: C, 81.44; H, 8.63; N, 4.06; Mn = 19500, Mw = 50300, PDI = 2.6 and DP = 31.
144 CHAPTER 5 DIKETOPYRROLOPYRROLE-BASED SEMICONDUCTING MATERIALS FOR PHOTOVOLTAIC APPLICATIONS 5.1 Introduction In Chapters 3 and 4, we developed two di fferent types of polymeric materials for photovoltaic applications. In this chapter, the focus is switched to molecular materials and thermally cleavable polymers to explore ot her possibilities for obt aining high performance materials for field-eff ect transistors and photovol taic applications. As discussed in Chapter 1, monomeric materi als have some intrinsic advantages over polymeric materials, such as ha ving well-defined molecular struct ures, ease of functionalization, and versatile methods of purificati on. However, PCEs of small molecule solar cells are, to date, not competitive with those of polymer based devices.76,78,83,142,146,258-260 This is due, in part, to homogeneous mixing of donor and acceptor component s in as-spun thin films and/or large phase separation upon crystallization; it also results from the reduced crystallinity in thin films prepared via solution processing (e .g., spin-coating or spray coatin g) versus films prepared via thermal-vacuum deposition.261 It is now widely accepted that charge mobility in organic semiconducting materials, as in inorganic semi conductors, is strongly linked to the degree of long range order (i.e., higher crystallinity co rrelates with larger charge mobility).262 Low charge mobility leads to charge accumulation, and in efficient charge collection, and ultimately decreased fill factors and poor pe rformance in a photovoltaic device.263 Another fundamental characteristic, exciton diffusion length, is also cl osely related to the ex tent of crystallinity. Forrest, et al. has correlated the four-fold increase in the exciton diffusion length of 3,4,9,10perylenetetracarboxylic dianhydride (P TCDA) with the enhanced degree of crystalline order that accompanies vacuum deposition.261 In general, small crystalline molecules exhibit higher intrinsic charge mobilities and exciton diffu sion lengths compared to their polymeric
145 counterparts.264 Therefore, in principle, superior phot ovoltaic performance can be expected for small molecule solar cells provided that nanos cale phase segregation among donor and acceptor -stacks occurs while retaining hi gh-order crystallinity in solutionprocessed thin films. Toward this goal, materials that can undergo self-asse mbly and self-organization have emerged as attractive targets.265-267 Access to various supramolecular nanostructu res via self-assembly of small organic molecules has been the focus of extensive research.268-272 To date, much success has been achieved in this regard. Nanostructures, e.g. one -dimensional rods, tubes, fibrils and helical columns, two-dimensional rings and layered sheets as well as three-dimensional vesicles, have been demonstrated to a large exte nt. In addition, a great deal of a ttention in this field has also been gradually shifting towards development of self-assembled materials for functional applications.273-278 In both cases, amphiphilic molecular design has been shown to be an effective strategy to achieve these goals.275,279 Along these lines, we discuss in this Chapter the Synthesis and characterization of the first cruciform-shap ed amphiphilic diketopyrrolopyrrole-based (DPP) molecule with a rich self-assembly behavior in the bulk, in solution, and on a surface. We also demonstrate that the amphiphili city endows the molecule wi th attractive properties (e.g., solution-processable, self-assemble d long-range order, etc.) for or ganic field-effect transistors and solar cells. One problem associated with the commercializ ation of polymer based solar cells is their instability during prolonged light exposure. In a typical MDMO-PPV/PCBM device, photodegradation of the polymer was observed u nder illumination both in the presence and absence of oxygen. The homolytic scission of the O-CH2 bond creates radical species that attack double bonds in the polymer chains, and lead to the degradation and loss of conjugation.280 The
146 photoinstability of P3HT has also been examined by Gardette, et. al., and it is believed that the instability is caused by the hexyl side chains.281 Thermally cleavable polymers draw attention because of their poten tially better operational stability in OPVs, as a result of their lower dens ity of solubilizing groups in the blend film after cleavage. In addition, this type of polymer is pot entially useful for solution-processed multi-layer devices, realized by repeating the cycle of soluti on processing-thermal cleavage. Lastly, a bonus from thermal cleavage is the improvement in chromophore density in the thin-film, which is beneficial for device performance. At this stage, thermocleavable materials remain underdeveloped and less competitive when comp ared to P3HT and other state-of-the-art materials in OPVs. In this chapter, we disc uss the Synthesis, characterization and use of diketopyrrolopyrrole-based (D PP) thermocleavable conjugated polymers on photovoltaic applications. DPP-containing materials are vibrantly colo red and highly fluorescen t with exceptional photochemical, mechanical, and thermal stability, and are therefore widely used in industrial applications as high-performance pigm ents in plastics, paints, and inks.282 Due to strong interactions and extensive hydrogen-bonding forces, these DPP-based molecules can not be directly used as organic solutio n-processed materials in OPVs, ra ther functionaliz ation of these parent molecules is desired, and is ge nerally accomplished through N-alkylation.283-288 Nguyen, et. al demonstrated a series of DPP-based small molecules used as active materials in BHJ OPVs and achieved PCEs of ~2.2-4.4%.76,78,80,83 Janssen, et al. reported high performance OPVs above 5% based on DPP-based conjugated polymers.41 These results greatly pr omote the study of DPPcontaining molecules and polymers as photovoltaic materials. Here, two types of unconventional DPP-based materials will be studied as follows.
147 5.2 Diketopyrrolopyrrole-based Amphiphilic Oligothiophene The molecule designed in this section has a constrained geometrical configuration and is composed of a diketopyrrolopyrrole oligothiophene ( DPP-1 ) conjugated rigid core, one pair of terminal lipophilic paraffinic chains and one pair of lateral hydrophilic triglyme chains, as shown in Figure 5-1. The extended conjugated co mponent is designed to ensure strong interactions and secure a columnar -stacking arrangement.287-288 At the same time, the molecule is expected to broadly absorb in the visible region of the solar spectrum and have ap propriate energy levels for charge transfer to the acceptor materi als commonly used in molecular BHJ solar cells.76,78,80,83 The long paraffinic chains will confer so me degree of solubility, but should induce crystallization below a certain temperature. Fi nally, orthogonal placemen t of the two flexible triglyme chains should enforce hydrophobic effect s with the two long aliphatic chains, and provide the desired processing solu bility for device fabrication. Figure 5-1. Chemical struct ure and model of DPP-1. 5.2.1 Synthesis of DPP-1 The Synthesis of DPP-1 is outli ned in Figure 5-2. Compound 5-1 was first reported by Ciba in 1986.289 Briefly, thiophene-2-car bonitrile and potassium tert -butoxide were mixed in tert -amyl alcohol and then heated to 95 oC under a nitrogen atmosphere. As soon as this temperature was reached, a solution of din -methyl succinate in tert -amyl alcohol was added
148 using a dropping funnel. When the addition was completed, the reaction mixture was kept at 95 oC for 14 h, then cooled to 50 oC, carefully neutralized with glaci al acetic acid, and boiled briefly at the reflux temperature. The resulting pigment suspension was diluted with water and filtered at room temperature. The filter cake was suspende d in a large amount of water, and the dark pigment was isolated again by filtration, and then fi nally washed with methanol and water. After being dried at 100 oC, crude compound 5-1 was obtained. Due to th e existence of strong hydrogen bonding and extensive pi -pi interactions, compound 5-1 is highly insoluble in common organic solvents and cannot be further purified by conventional me thods. In this study, we use it as collected from filtration after drying. Figure 5-2. Synthesis of DPP-1. a) potassium tert -butoxide, tert -amyl alcohol, 95 oC, ~30%; b) 1-bromo-2-(2-(2-methoxyethoxy)ethoxy)ethane potassium carbonate, DMF, 46%; c) NBS, chloroform, 61%; d) tributyl(5'dodecyl-2,2'-bithiophen-5-yl)stannane, Pd2(dba)3, P(o-tyl)3, THF, 49%. N-alkylation of compound 5-1 with 1-bromo-2-(2-(2-methoxyethoxy)ethoxy)ethane was carried out in the presen ce of potassium carbonate using DMF as solvent at 100 oC. Because of the impurities inherited from the pr ior reaction, purific ation of compound 5-2 was troublesome. Many repeated silica gel columns were performed in order to isolate compound 5-2 a dark red viscous liquid which solidifies af ter a few days. Bromination of 5-2 with N-bromosuccimide in
149 chloroform gave compound 5-3 in a fair yield. Stille couplin g between tributyl(5'-dodecyl-2,2'bithiophen-5-yl)stannane and 5-3 yielded the desired product of DPP-1 5.2.2 Structural Characterization, Op tical and Electrochemical Studies The structure of DPP-1 has been confirmed by 1H-, 13C-NMR, HRMS and elemental analysis. DPP-1 has clear-cut NMR spectra for both pr oton and carbon, as shown in Figure 5-3, even though it contains 92 hydrogen atoms and 68 carbon atoms. In the aromatic region of 1HNMR, the protons on thiophene ri ngs split into four separated peaks with integral ratios of 1:2:2:1 from 6.5 to 9.0 ppm. The downfield sign al at 8.78 ppm can be assigned to the thienyl protons next to the DPP core; while the upfield doublet at 6.66 ppm is as signed to the protons on the thiophene rings attached the alkyl chains. The signals from protons on the triglyme chains are located around 3.3-4.4 ppm. A si nglet at 3.32 ppm is assigned to the terminal methyl group of the triglyme chain. The 13C NMR of DPP-1 also exhibits a similar trend, as evidenced in Figure 5-3. In the upfield region from 108 to 165 ppm, there are 13 carbon signals from the conjugated segment; it should be men tioned that there are totally 15 sp2 carbons in this molecule. It is likely that the middle thi ophene only gives two set of signals provided that this thiophene can be considered nearly centrosymmetric and has very similar chemical surroundings for the 2,5 and 3,4 positions. The downfield signal at 161.4 ppm and the upfield signal at 108.5 ppm are readily assigned to the carbonyl carbon and the adjacent carbon. The signal at 146.4 ppm can be also tentatively assigned to the third carbon on the DPP core. It is difficult to make any further assignments of the peaks resulting from the rema ining carbons within th e thiophene rings. The signals from triglyme chains are easily disti nguished from alkyl chains for their different environments.
150 Figure 5-3. 1H-, and 13C-NMR of DPP-1 in CDCl3 DPP-1 has good solubility (> 10 mg/mL) in mo st common organic solvents, such as methylene chloride, chloroform, THF, toluene and di chlorobenzene. It is also fairly soluble in some polar solvents, including NMP and DMF. Figure 5-4a shows the absorption and emission spectra of DPP-1 in toluene. It has a tw o-band absorption pattern, char acterized by two peaks at ~ 390 and ~640 nm. The high energy band can be at tributed to terthiophene components, and the long wavelength band is associated with charge transfer between terthiophene and the DPP core. This observation is very similar with previously reported DPP based molecules.78 In the solid state, the entire spectrum becomes broader, a nd a shoulder appears at longer wavelengths (~730 nm), indicating the strong molecular interactions. DPP-1 in toluene emits with a max at~687 nm and a broad shoulder at ~732 nm; this shoulder is likely resulting from the forma tion of fluorescent exci mers in solution. The fluorescence is completely quenched in thin-fil ms, suggesting the formation of extensive pi stacks.290 Figure 5-4b shows the absorption spectra of DPP-1 in four different solvents, THF, dimethoxyethane (DME), DCM and N-methylpyrrolidone (NMP), in the or der of their relative
151 polarities 0.207, 0.231, 0.309, and 0.355, and photographs of the solutions are shown in Figure 5-4c. As seen from Figure 5-4b, the high energy bands, associated with the terthiophene units, have relatively even absorption intensities; wh ile the low energy bands, attributed to charge transfer, are more influenced by th e solvent polarity. In the case of this series of solvents, NMP is the most polar while DME is the least polar. The observation reflects this trend well in the low energy bands, in which the absorption intensit ies decrease with incr easing of polarities. Figure 5-4. a) Absorption and emission of DPP-1 in toluene; b) absorp tion of DPP-1 (2.2 x 10-5 M) in NMP, DCM, THF and DME; c) the images of DPP-1 ) in NMP, DCM, THF and DME Electrochemical studies were then performed to obtain in formation on energy levels and gaps for DPP-1 Cyclic voltammetry (CV) of DPP-1 in CH2Cl2 using TBAPF6 as an electrolyte shows four quasireversible consecutive redox pr ocesses with half-wav e potentials around 0.65,
152 0.47, -1.30 and -1.79 V vs Ag/Ag+, respectively. The differentia l pulse voltammograms (DPV) for DPP-1 also reveal two well-defined and highly revers ible anodic and cathodic processes with a high degree of symmetry, and the half-wave poten tials are very close to the potentials observed in CV, as shown in Figure 5-5. This gives LUMO and HOMO levels of DPP-1 3.9 and 5.5 eV, respectively. The potential difference be tween oxidation and reduc tion of 1.77 V is in excellent agreement with th e optical gap (1.77 eV) of the molecularly dissolved DPP-1 in CH2Cl2, which has an onset of absorption near 700 nm. A significant shift of absorption onset to 825 nm was observed for the thin film, giving a cal culated optical energy gap of 1.5 eV, slightly larger than its corresponding polymer. Since solid state electrochemical studies are hindered by the fair solubility of DPP-1 in ACN and PC (two commonly used solvents for E-chem), direct comparison can not be provided between the energy gaps from the solid state optical measurement and from solid state electroc hemical experiment. Nevertheless, all the electrochemical and optical results indicate that DPP-1 is indeed a low energy gap molecule as proposed. Figure 5-5. a) Cyclic voltammetry, and b) differential pulse voltamm etry of DPP-1 measured in a 0.1M solution of TBAPF6/DCM (scan rate 25mV/s) vs Ag/Ag+ (EFc/Fc + = EAg/Ag + + 0.16 V) 5.2.3 Thermal Analysis, Polarized Light Microscopy and X-ray Analysis The identification of DPP-1 mesophases was carried out by differential scanning calorimetry (DSC), polarized light microsc opy and X-ray diffraction. DSC thermograms of
153 DPP-1 reveal reversible phase transition behavior as shown in Figure 5-6a. Upon heating (5 C /min), DPP-1 underwent a first endothermic transition at 68 C (12.4 kJ/mol), followed by a second thermal transition at 98 C (25.2 kJ/mol ), and finally reached a clearing point at 184 oC (33.0 kJ/mol). On the cooling cycle, three corr esponding exothermic transitions were observed at 181, 92 and 65 C with enthalpies of 32.8, 22.2, a nd 17.7 kJ/mol, respectively. After the second scan, the subsequent scans reproducibly overlapped the prior scan. In order to gain insight into its phase transition beha vior in the blend of DPP-1 /PC61BM, PC61BM and the blend were subjected to thermal studies as well, as shown in Figure 5-6b and 5-6c. It is surprising to find that DPP-1 still exhibits relatively well-defined therma l behaviors. The disappearances of the melting and recrystallization peaks for PC61BM are anticipated, since DPP-1 is still in its melting state above 248 oC in the blend. Lowering of the recrystallization temperature for DPP-1 is also expected, considering PC61BM as an impurity in the blend. Figure 5-6. Differential scanning calori metry thermograms. a) DPP-1, b) PC61BM, and c) DPP-1 and PC61BM mixture (1:1, wt %). With the rich phase transition behaviors exhibited by DPP-1 in the DSC study, we were encouraged to further investigate phase transi tions using polarized opt ical microscopy (POM). Upon cooling the sample from the isotropic melt and shearing at 170 C, a highly ordered, slightly birefringent palm-tree-leaf-like te xture appeared under PO M (Figure 5-7a). Its
154 appearance (i.e., domain size and color) and ps eudo-uniaxial solid phase growth could be controlled by external forces (e.g. shearing). Figure 5-7. a) Polarized optical microscope (POM) images of DPP-1 under 170 C at two different angels; b) Snapshots of the dire cted growth of a DPP-1 plastic phase as viewed by POM at 170 C after shearing (pictu res taken at 5 s intervals and the arrow indicates the shearing direction); c) POM images of DPP-1 under 80 C, and d) RT Figure 5-7b shows snapshots of directed soli d phase growth that accompanied external shearing, where the images were taken every five seconds. Noticeably, while the fluidity of
155 DPP-1 was largely reduced after shearing, the mate rial still appeared to be soft and waxy, suggesting that this mesophase is not liquid crys talline, but likely a columnar plastic phase.270,291 Another phase change, possibly from a less ordere d to a more ordered columnar plastic crystal phase, was subsequently observed around 90 C, in good agreement with the DSC results. This lower temperature phase transition may correspond to the crystallization of the two terminal alkyl chains. With the formation of another cr ystalline phase below 65 C, possibly associated with solidification of the flexible triglyme chains, a mosaic-like appearance was observed by polarized light microscopy, as shown in Figure 5-7d. Figure 5-8. Diffraction patterns obtaine d from DPP-1 at 170, 80, and 50 C. Variable temperature X-ray diffraction was us ed to further confirm the formation of mesophases. Figure 5-8 shows the diffraction patterns of DPP-1 obtained at 170, 80 and 50 C (temperatures corresponding to th e phases reported by DSC) u pon cooling from the isotropic melt. Sharp Bragg reflections accompanied bot h high temperature mesophases (170 and 80 C), consistent with long-range ordering a nd the assignment of plastic phases.276 That the substance is shearable at temperatures close to the cleari ng temperature and bears a rather soft and waxy
156 texture excludes the possibility that a true cr ystalline phase exists at 170 C. By the same argument, DPP-1 is likely not a plastic crystal belo w 65 C given that the texture is not deformable. The nature of the mesophase around 80 C remains unclear, since it is rather difficult to examine the texture. Unfortunately, due to the considerable orie ntational disorder, it is generally not very insightful to index the reflection peaks for plastic crystal phases obtained by one-dimensional X-ray diffraction.292-293 5.2.4 Morphology Studies In order to visualize the formation of -stacks (aggregates) in solution and on surfaces, we then examined aggregates and thin-films by atomic force microscopy (AFM). A solution investigation was performed as follows. A 2.2 10-4 M solution of DPP-1 in tetrahydrofuran (THF), a good solvent for this molecule, was ad ded dropwise to hexane, a poor solvent, while vigorously stirring. The solution tu rned from blue to purple after some time (a few minutes), suggesting the formation of DPP-1 aggregates. Deposition of the suspension on mica and analysis by AFM revealed uniform nanoscale fi ber-like assemblies as shown in Figure 5-9a. Similar aggregates were also formed in met hylcyclohexane (MCH). These observations were also consistent with results obtained from UV-vis measurements. Brie fly, UV-vis spectra of DPP-1 in MCH were recorded every 45 seconds at 30 C, after initial e quilibration at 60 C. Figure 5-9b clearly shows the grow th of a broad absorption with max ~ 750 nm accompanied by a solution color change from blue to purple indicating the existen ce of strong excitonic interactions and the formation of -stacks. The isobestic point at 680 nm further demonstrates a stoichiometric conversion from free molecules to crystals (agg regates). These solution studies demonstrate that DPP-1 has a strong tendency to self-a ssemble into highly ordered nanostructures, most likely through synergistic sol vophobic effects and interactions.
Figure 5 To and subs t chlorobe n and stud i showed a to 8.5 m i Figure 5 9. a) tappin g from TH F 10-5 M, investigate t rate, a proc e n zene soluti i e d by AFM a significant i nutes. 10. Tappin g drop-cast a) 3.5 mi n g mode AF M F -hexane on t at 30 C) r e the effect o f e ss relevant ons (2.2 1 Tapping m increase in o g mode AF M from chlor o n b) 5 min a M image (2 0 t o mica. b) U e corded at 4 5 f self-organi z to the fabri c 0-4 M) were m ode AFM i m o rdering as t M images (5 o benzene (2. a nd c) 8.5 m i 157 0 20 m, 2 5 U V-vis spec t 5 s intervals ; z ation of D P c ation of or g drop-cast o n m ages and l i t he solvent e 5 m) wi t 2 10-4 M) i n. 5 0 nm in he i t ra of DPP1 ; P P-1 at the i n g anic electro n n to the surf a i ne traces, a s e vaporation t t h z-height l i on mica. So l i ght) of DP P 1 in methylc y n terface bet w n ic devices, a ce of f r eshl y s exhibited i t ime was in c i ne scan pro f l vent evapo r P -1 as depos i y clohexane w een soluti o DPP-1 y cleaved m i i n Figure 51 c reased fro m f iles of DPP r ation times i ted (2.2 o n i ca 1 0, m 3.5 -1 are
158 At 3.5 minutes, irregular aggregates formed base d on the height fluctua tion of the line trace across the surface. At 5 minutes, sharper feat ures with even steps in height of 3.64.0 nm developed, a much more ordered arrangement. At an evaporation time of 8.5 minutes, the molecules had assembled into larger, belt-like domains with uniform step height of 3.64.0 nm. We propose that the step-hei ght of 3.64.0 nm corresponds to DPP-1 molecules that are packed with their long axes oriented perp endicular to the substrate. 5.2.5 Organic Field-Effect Transistors and Molecular Heterojunction Solar Cells This section describes collaborative work on OFETs and OPVs with Professor Bernard Kippelens group at Georgia Inst itute of Technology. Specificall y, Dr. Shree Prakash Tiwari fabricated field-effect transist or devices, offered the data and helped in its interpretation. Dr. Hyeunseok Cheun and Jaewon Shim carried out solar cell studies and provided the data. To determine whether the proclivity of DPP-1 to from -stacks via self-assembly has a pronounced impact on device performance, top c ontact field-effect de vices with gold top source/drain electrodes and DPP-1 as the organic semiconductor were spun-cast from chlorobenzene on O2 plasma cleaned SiO2. Figure 5-11 shows the output and transfer characteristics of a particular p-channel OFET ( W/L = 1000 m/25 m) with DPP-1 after annealing at 130 oC for 30 min. The operating voltage fo r these devices was 60 V. These OFETs exhibited average hole mobility values of 4 10-3 cm2/Vs, and current on/off ratios of 1 104, with the average threshold voltage of 6.4 V (m ore details in Table 51). These results are comparable to some of the best molecular solution-processable OFET devices.139,294 More importantly, it suggests introducing flexible trigly me chains does not necessarily depreciate the device performance, while imparting the desire d solubility for purification and processing.78
159 Figure 5-11. a) Output and b) transfer characteristics of a re presentative DPP-1 field-effect transistor device Table 5-1. Summary of DPP-1 based OFET devices W/L (cm2/Vs) VTH (V) Ion/off 1000m/25m (3 Dev.) 4.1 ( 0.4) x10-3 6.3 (0.1) 1 x104 1000m/50m (4 Dev.) 3.2 ( 0.8) x 10-3 2.5 (1.3) 8 x103 1000m/100m (4 Dev.) 3.0 ( 0.3) x 10-3 1.5 ( 0.4) 3 x103 Solution-processed molecular bulk-heterojunction solar cells were subsequently fabricated using DPP-1 as an electron donor material and PC61BM as an electron acceptor material in a standard device confi guration ITO/PEDOT:PSS/ DPP-1 :PC61BM/Al. The currentvoltage characteristics of a representative device before and after annealing ar e shown in Figure 5-12. Devices with an active layer thickness of ~ 100 nm yielded a short circuit current density (Jsc) of 1.56 0.06 mA/cm2, an open circuit voltage (Voc) of 0.50 0.01 V, a fill factor ( FF ) of 0.54 0.01, and a power conversion efficiency of 0.43 0.01% under AM 1.5 G, 100 mW/cm2 illumination (the values have been obtained and averaged from five devices). Upon annealing at 90 C, the devices exhibited an enhanced Jsc of 2.41 0.07 mA/cm2, a Voc of 0.55 0.02 V, a FF of 0.55 0.03, and a power conversion effici ency of 0.68 0.02%. The external quantum efficiency increased upon a nnealing from 10% to above 15% in the 415580 nm and 620770
160 nm range. The PCEs of these minimally optimized devices are less than the state-of-the-art solution-processable molecular BHJ solar cells,80 however, their fill factors before and after annealing are much higher than other solution-pr ocessed molecular solar cells reported in the literature,57,71,76-77,295 and are comparable to the best performing vacuum-deposited molecular heterojunction solar cells.153 This observation indicates that ther e is a great deal of potential to improve the overall device effi ciency, provided that a high FF is a prerequisite to achieving high performance from photovoltaic devices. Figure 5-12. a) J-V characteristics and b) exte rnal quantum efficiencies of a representative DPP-1 : PC61BM photovoltaic device before and after annealing at 90 oC In summary, we have demonstr ated the design, Synthesis, and characterization of the first cruciform-shaped amphiphilic -conjugated low energy gap oligot hiophene. We are hopeful that the same design principles used to achiev e the desired self-assembly tendency with DPP-1 can be extended to other polycyclic aromatic cores. We have measured hole mobilities through fieldeffect transistor devices and tested DPP-1 blended with PC61BM in molecular heterojunction solar cells. The high mobility indicates that self-assembly of the molecule indeed has a pronounced influence on device performance. Our preliminary results bode well for the use of this type of molecules (i.e. that is possess so lution processability and long-range order via selfassembly) as active materials in field-eff ect transistors and p hotovoltaic devices.
161 5.3 Diketopyrrolopyrrole-based Hydrogen-bonded Amphiphilic Oligothiophenes The self-assembly properties exhibited by DPP-1 encouraged us to investigate other DPPbased oligothiophenes by applying a similar amphi philic molecular design. In this section, we focus on a new set of amphiphilic oligothiophenes ( DPP-OH-n n=1, 2, 3 and 4), as shown in Figure 5-13. This set of molecules has struct ural features that are unique relative to DPP-1 : first, the hydrophobic effect is much enhanced due to the existence of long paraffinic chains. Second, the hydrophilic chains are term inated with hydroxyl groups; st rong hydrogen bonding forces are preinstalled in this manner. For comparisons purpose, DPP-TEG and DPP-C12 have also been prepared. By making a comparison between DPP-TEG and DPP-OH-n we anticipated getting insight into whether and how hydrogen bonding interactions play a role in self-assembly and their phase transition; the pair of DPP-C12 and DPP-OH-n is expected to provide information on hydrophobic effects, which are obviously absent in DPP-C12 and present in DPP-OH-n Third, a set of DPP-OH molecules is proposed and prepared; in these materials, the strength of pi-pi interactions is varied vi a the number of thiophene rings. Figure 5-13. Molecular structur es of DPP-OH-n and DPP-TEG.
162 5.3.1 Synthesis of DPP-OH-n (n = 0, 1, 2 and 3) and DPP-TEG The Synthesis of the donor building blocks is illustrated in Figure 5-14. Compound 5-5 is a donor for DPP-OH-1 and also acts as an intermediate for other extended pi-conjugated donor building blocks. It is worth mentioning that 5-5* an analog of compound 5-5, has been previously prepared in a multi-step (n > 5) Synthesis, as shown in Figure 5-15.296 Using iridiumcatalyzed borylation,297 we are able to shorten the sequence to two steps from the commercially available pyrogallol (benzene-1,2,3 -triol). And the overall yield has increased to 86% from 26%. Upon reacting 5-5 with 2-bromothiophene, 5-bromo-2,2'-b ithiophene, and 2-Bromo-5,2':5',2''terthiophene via the palladium-catalyzed Suzuki reaction, followed by the iridium-catalyzed selective borylation on the 5 pos ition of the terminal thiophe ne ring, donor building blocks 5-7 5-9 and 5-11 were successfully prepared in good yields Due to the simplicity of the chemistry demonstrated here, these donor build ing blocks can be easily scaled up to tens of grams. Figure 5-14. Synthesis of donor bui lding blocks for DPP-OH-n. a) C12H25Br, K2CO3, DMF, 60 oC, 94%; b) 4,4,4',4',5,5,5',5'-octame thyl-2,2'-bi(1,3,2-dioxaborolane), Ir[OMe(COD)]2, heptane, 80 oC, 91%; c) 2-bromothiophene, Pd2(dba)3, P(o-tyl)3, aqueous Et4NOH, toluene, 98%; d) 2-bromobithiophene, Pd2(dba)3, P(o-tyl)3, aqueous Et4NOH, toluene, 71%; e) 2-bromoterthiophene, Pd2(dba)3, P(o-tyl)3, aqueous Et4NOH, toluene, 72%; f) 4,4,4',4',5,5, 5',5'-octamethyl-2,2'-bi(1,3,2dioxaborolane), Ir[OMe(COD)]2, heptane, 80 oC, 50%; g) 4,4,4',4',5,5,5',5'octamethyl-2,2'-bi(1,3,2-dioxa borolane), Ir[OMe(COD)]2, heptane, 80 oC, 50%; h) 4,4,4',4',5,5,5',5'-octamethyl-2,2'-bi( 1,3,2-dioxaborolane), Ir[OMe(COD)]2, heptane, 80 oC
163 Figure 5-15. Synthesis of 3,4,5-tris(d odecyloxy)phenylboronic acid. a) K2CO3, C12H25Br, DMF, 90oC, 24h, 85%; b) NaNO2, HNO3, H2O, CH2Cl2, rt, 3h, 90%; c) hydrazine, graphite, EtOH, 110 oC, 24h, 88%; d) NaNO2, HCl, H2O, 0 oC and then KI, rt, 12h, 49%; e) nBuLi, THF, -78 oC followed by B(OMe)3, -78 oC and HCl, H2O, 80%. The Synthesis of the DPP core is shown in Figure 5-16. The Synthe sis started with the protection of 2-(2-(2-bromo ethoxy)ethoxy)ethanol with the tert -butyldimethylsilyl (TBDMS)group, yielding 12-bromo-2,2,3,3-te tramethyl-4,7,10-trioxa-3-siladodecane ( 5-12 ). Nalkylation of 5-1 with 5-12 was carried out in DMF using po tassium carbonate as a base. It should be mentioned that direct N-alkylation of 5-1 with 2-(2-(2-bromoethoxy)ethoxy)ethanol was also successful, however, subs equent purification turned out to be extremely difficult as a result of the limited solubility of the resulting product. By attaching TBDMS groups this problem is resolved. The relatively low yield of 42% is in part caused by the poor purity of the starting material 5-1 as was mentioned earlier. We also at tempted an alternative route to obtain 5-13 by replacing compound 5-12 with its tosylate analog, howe ver the yield was even lower. Bromination of 5-13 was problematic in our hands and the reaction yields were extremely low; it was later discovered that the TBDMS groups we re partially removed during this process.298 After optimizing the reaction conditions, we were able to obtain compound 5-14 in ~60% yield. The Synthesis of DPP cores for DPP-TEG and DPP-C12 has been previously described in section 5.2. With the donor building blocks and the DPP cores in hand, DPP-OH-n DPP-TEG and DPP-C12 were synthesized by Suzuki coupling, using Pd2(dba)3 and P(o-tyl)3 as the catalyst and ligand respectively, aqueous Et4NOH as the base, and toluene as the solvent. The yields ranged from 54% to 84%. The synthetic details can be found in the experimental section.
164 Figure 5-16. Synthesis of DPP core for DPP-OH-n. a) Et3N, CH2Cl2, rt, 95%; b) 5-1 K2CO3, DMF, 90oC, 24h, 42%; c) NBS, chloroform, then HCl, H2O, rt, 60%. 5.3.2 Structural Characterization, Op tical and Electrochemical Studies The structures of the molecules under investigation have been supported by 1Hand 13CNMR, as well as HRMS. As an example, Figure 5-17 shows 1H-NMR spectra of DPP-C12 DPP-TEG and DPP-OH-2 Their peak assignments are briefly discussed as follows. Figure 5-17. 1H-NMR spectra of DPP-C12, DPP-TEG and DPP-OH-2 It is clear from Figure 5-17 that the chemical shift patterns in the aromatic region from 6.7 to 8.9 ppm are very similar for all three molecu les, resulting from the same pi-conjugated
165 segment that all three have in common. Four pr otons on the two thiophene rings (one side) are split into four doublets. The most downfield doubl et at ~ 8.8-8.9 ppm corresponds to the proton on the thiophene ring next to the DPP core; while the doublet at 7.1 ppm is assigned to the proton on the thiophene ring next to the benzene ring. Due to their similar chemical surroundings, the two inner protons (3 and 3 positions) on the thio phene rings are partially overlapped and appear as a doublet of doublets (dd) around 7.26 ppm, with the signal fr om deuterated chloroform (residue) also overlapping. The singlet that stan ds at ~6.76 ppm is undoubtedly assigned to the two symmetric protons on the benzene ring. The similarity in the aromatic region among the three NMR spectra suggests that these molecules ar e highly solvated and little or no influence is observed from pi-pi interactions. In the aliphatic region from 3.31 to 4.31 ppm, DPP-TEG and DPP-OH-2 also look very similar except that DPP-TEG has a sharp singlet located at 3. 31 ppm from the terminal methyl group on the triglyme chains, and DPP-OH-2 has a broad singlet located at 2.57 ppm from the terminal hydroxyl group. DPP-C12 has a different appearance in this region; only one broad multiplet sitting around ~3.90-4.15 ppm is observe d, corresponding to the methylene groups connected to the nitrogen atoms. All three molecu les exhibit the same pattern in the region below 2 ppm. These peaks are from the protons on the al iphatic chains attached to the benzene rings. Figure 5-18 shows the absorption spectra of the DPP-OH-n series along with photographs of their methylene chloride so lutions. With increasi ng the numbers of th iophene rings, the absorption onsets move from 642, to 693, to 711 and finally to 721 nm for DPP-OH-1 2 3 and 4 respectively. From DPP-OH-1 to DPP-OH-2 the difference in abso rption onset is about 50 nm. The difference decreases to 18 nm from tw o to three thiophene ri ngs. And it continually decreases to 10 nm when moving from DPP-OH-3 to DPP-OH-4 The trend for high energy
166 bands ( to transition), however, is not clear. DPP-OH-1 and DPP-OH-3 have very similar high energy bands with absorption maxima at ~ 401 nm. The intensity ratio of the high energy band to the charge transfer band fo r these two molecules is different. DPP-OH-3 has a very strong absorption for the to transition; where DPP-OH-1 has a relatively weak to transition. In the case of DPP-OH-2 and DPP-OH-4 they have very simila r intensity ratios of the two bands, but rather different absorpti on maxima at 371 and 421 nm, respectively, which may be caused by the odd/even effect,299-300 which refers to the elec tronic geometry difference between odd and even number of rings. Figure 5-18. Absorption spectra of DPP-OH-n (n = 1-4) in CH2Cl2 and their optical images Table 5-2 shows the electrochemi cal energy levels and gaps of DPP-OH-n (n = 1-4) obtained from DPV by drop casting of the sample s on platinum electrodes. With increasing donor strength, the HOMO levels from DPP-OH-1 to DPP-OH-4 increase from -5.55 to -5.26 eV. Interestingly, the LUMO levels deepen as well from -3.77 to -3.69 eV, which suggests that the LUMO level is affected by the extension of conjugation length as well. The change of HOMO and LUMO levels consequently leads to a narrowing of energy gaps, which drops from 1.78 to 1.57 eV from DPP-OH-1 to DPP-OH-4. It is noticed that the inte rval of energy gaps is consistent with that of the abso rption onset aforementioned; an observation which is also evident
in the co l indicate t sides of t Table 52 5.3.3 Th e D S hydroge n Figure 5 transitio n This tra n oC. In co temperat u triglyme 141 oC f o higher cl individu a network b ehavior l or of these m t he conjuga t t he molecul e 2 Electroch e e rmal Anal y S C studies h a n bonding o n 19a shows t n s. At ~33 o C n sition has a l ntrast, DPP u res betwee n chains in D P o r DPP-OH earing temp a l molecule s of DPP-O H of some hi g m olecules, f t ion length i s e e mical ener g y sis a ve been ca r n the self-as s t he DSC the r C the b road l so been obs TEG melts n DPP-12 a n P P-TEG O 2, in which erature is as s in the form H -2 is indica t g h polymers f rom purple, s almost sat u g y levels an d r ried out to i n s embly and p r mogram o f peak corres p erved in D P around 91 o n d DPP-TE n the contra r triglyme c h sociated wi t of supramo t ed by the a p 301-302 167 to blue and u rated after a d gaps of D P n vestigate t h p hase transi t f DPP-12 It p onds to m e P P-TEG an d o C; there is a G This ca n r y, the melt i h ains are ter m t h the strong lecula r nan o p pearance o f eventually t a ttaching th r P P-OH-n (n = h e influence t ion behavi o has two we l e lting of the t d DPP-OH2 a bout a 30 o C n be ascribe d i ng (clearin g m inated wit h hydrogen b o structures. f a cold recr y t o green. Th e r ee thiophen e = 1-4) from of hydroph o o rs of these m l l-defined th t erminal do d 2 DPP-12 m C decrease i n d to the flexi g ) temperatu r h hydroxyl g onding forc e The hydrog e y stallization e se results e rings to b o DPV o bic effects a m olecules. ermal d ecyl chains m elts aroun d n melting ble nature o r e increases roups. This e s that hold t e n bonded at 77 oC, a o th a nd d 122 f to t he
168 Figure 5-19. DSC thermograms of DPP-C12, DPP-TEG and DPP-OH-2. It is clear that we can alter the phase transi tion behaviors to a larg e degree by altering the passive groups on the functional el ectroactive molecules. Phase tr ansition behaviors are directly associated with the microstructu res (or morphology) in a material. In another words, we can alter microstructures of a material via changing the pa ssive groups. Therefore, this approach seems applicable to designing new electroactive materi als for organic electronics, where material morphology plays a very important role. Figure 5-20 illustrates the thermal transition behaviors of DPP-OH-n (n= 1-4). With the extension of the conjugated segments from one to four thiophene rings, the melting temperature increases from ~100 oC to ~237 oC. A ~40, 45 and 50 oC increase in melting temperature is observed for each thiophene extension. This trend can be explained by the decreasing mass ratio of flexible chains to conjugated rigid core, however, it may also result from the stronger pi-pi interactions that occur on incr easing the pi surface areas (from DPP-OH-1 to DPP-OH-4) This could be easily verified by the pi-pi distance in an X-ray study; unfortuna tely, we do not have this data set on hand at this time. We also studied these molecules by POM. Surprisingly, no obvious birefringence was observed for any of the molecules, while it was clearly shown in DPP-1 (Note: the melting temperature is too high in the case of DPP-OH-4 and is out of the operation range of our instrument. It is thus not included in this disc ussion). These observations once again raise the
169 question of whether these amphiph ilic cruciform shaped molecules are plastic crystals, as we proposed for DPP-1 (Section 5.2.2). To answer this questi on, a series of advanced 2-D X-ray studies should be carried out, however, this has been delaye d due to technical difficulties we have encountered; we will continue to work towa rd the realization of th ese phase assignments in the future. Figure 5-20. DSC thermograms of DPP-OH-n (n=1, 2 3 and 4). 5.4 Thermally Cleavable DPP-Based Low Bandgap Polymer As introduced at the beginning of this chapte r, thermally cleavable conjugated polymers are of interest for a number of reasons. For instance, the loss of solubilizing groups will likely lead to an increase in glass tran sition temperature, which is closel y related to the stable and rigid morphologies desired in blends, hence providing more reliable phot ovoltaic performance. In the course of studying DPP-based materials, we re alized the convenience of functionalization of diketopyrrolopyrrole building block 5-1 with thermocleavable tert -butoxycarbonyl groups (tBoc)
170 at the N,N positions (tBoc groups can be thermally removed around 180-200 oC303-304). In this section, we use the following polymer PDPP-Boc (Figure 5-21) as an ex ample to investigate its performance as an active materials before a nd after cleavage in BHJ solar cells. Figure 5-21. Thermal transition of PDPP-Boc The Synthesis of PDPP-Boc is outlined in Figure 5-22. Fi rst, conversion of insoluble pigment 5-1 into soluble compound 5-15 was successfully achieved at room temperature by stirring 5-1 in THF with ditert -butyl dicarbonate, using 4-dime thylaminopyridine (DMAP) as a catalyst. Compound 5-15 was subsequently brominated with NBS at the 2 and 5 positions on the thiophene rings to give compound 5-16 Due to its proclivity to cleave the tBoc group, acetic acid should be avoided as a catalyst in the bromination step, as is ofte n practiced in previous studies. With the DPP core 5-16 in hand, PDPP-Boc polymer was obtained by Stille coupling the DPPcore with 2,5-tributylstannyl-2-ethylhe xyloxy-substituted-3,4-propylenedioxythiophene (ProDOT-Sn2) in toluene. The number-average molecular weight obtained was around 26.8 kDa, with a PDI of 2.2 by GPC measurement. Figure 5-22. a) di -tert -butyl dicarbonate, DMAP, THF, rt, 60%; b) NBS, CHCl3, rt, 77%; c) ProDOT-Sn2, Pd2(dba)3, P(o-tyl)3, toluene, 85 oC, 85%.
171 A thermal cleavage study of PDPP-Boc was carried out by TGA, as shown in Figure 523a. The onset of weight loss starts from ~155 oC and ends at 210 oC with a rate of 50 oC min-1. The percent mass loss for this temperature windo w is about 23.1%, in good agreement with the theoretical value of 21.4%, thus conf irming the successful cleavage of the tBoc groups. Figure 5-23b shows the UV-vi s absorption spectra of PDPP-Boc in chloroform solution and as thin films (before and after cleavage). In solution, PDPP-Boc has a single broad absorption starting from 450 and extending to 103 9 nm, with an absorption maximum at ~783 nm and an absorption shoulder around ~ 887nm. Upon spin-casting from chlorobenzene solution on a glass substrate, the absorption maximum has a red-shift to 809 nm, together with a bathochromic shift to 1068 nm for the onset. Surp risingly, the absorption maximum of the film has a hypsochromic shift back to 800 nm upon annealed at 180 oC for 20 min. Similarly, the absorption onset moves back to 1036 nm. Figure 5-23. a) Weight loss of PDPP-Boc as a f unction of temperature; b) Absorption spectra of PDPP-Boc in solution, thin-film before and after cleavage; c) Crystal structure for compound 5-15. These observations were not expected at first glance for a number of reasons. PDPP-Boc has a fairly planar backbone be fore the cleavage, as indicated by the crystal structure of 5-15 (Figure 5-23c). The tBoc sticks out and causes little distor tion. Therefore, the contribution from planarization is expected to be minimal upon cl eavage and will not likely lead to a significant
b athochr may cre a illustrate cleavage planizati o cleavagi n Figure 5 Th of hydro g substitut e solubiliz i omic shift. O a te a large b a d in Figure 5 of the tBoc o n of the m o n g reaction, 24. a) Wei g (rate, 10 height, 4 0 hydroxyp spin-coat e hydrogen e hypsochr o g en bondin g ed -3,4-prop y i ng groups o O n the other a thochromic 5 -24, where group, acco m o lecule, and leads to en h g ht loss of N C min-1; o n 0 .7%); b) Ul t p olystyrene e d (yellow) a b onding ne o mic shift in g and a wea k y lenedioxyt h o n the dioxy t han d the e x shift. Iqba l the bathoch r m panied wi t concurrent f h anced J agg r N' -bis-( te r n set, 163.6 t raviolet-vis film contai n a nd after 2 m twork of D P the PDPPB k ening in ac c h iophene m o t hiophhene l 172 x pected for m l et al studie d r omic shift w t h a color c h f ormation o f r egates and c r t b utoxycar b C; midpoin t ible absorpt i n ing 40% N, m in at 180 C P P. (Adapte d B oc polyme r c eptor stren g o nomers inc o l ikely block m ation of hy d d DPPb ase d w as found t o h ange from y f hydrogen b c auses the c o b onyl)-DPP t 177.7 C; e i on spectra o N' -bis-( ter t C (red); c) S d with perm i r can then b e g th. PDPPB o rporated al o the formati o d rogen bon d d latent pig m o be as large y ellow to r e d b onding, res u o lor change as a functi o e nd point 1 9 o f a 1.5-m t b utoxycar b S chematic ill u i ssion from R e possibly as B oc has 2-et h o ng the bac k o n of the de s d ing interact i m ent, as as 100 nm a d .305 The u lting from t o n of temper a 9 1.8 C; ste p thick p b onyl)-DPP, u stration of R ef 248.) cribed to a l h ylhexyloxy k bone. The b s ired hydrog e i ons a fter t he a ture p the ack b ulky e n
173 bonding interaction, or at least lower it to some extent. In addition, the tBoc group is an electronwithdrawing substituent, which can possibly lowe r the LUMO level of the DPP core. After the cleavage of the tBoc, its contribution as an electr on-withdrawing group disappears. The combination of the two factors may eventually account for the blue shift upon cleavage. CV and DPV analyses of drop cast thin films of PDPP-Boc were also performed, as shown in Figure 5-25. Applying an anodic potential to PDPP-Boc shows a single oxidation at an E1/2 of 0.59 V with an onset of oxidation at 0.26 V. Reduction of PDPP-Boc (Figure 5-25a) shows a first reduction at an E1/2 of -1.23 V, with an onset for the fi rst reduction at -1.08 V. This reduction is quasi-reversible with adequate charge comp ensation on the reverse s can. Asecond reduction is observed that is chemically irreversible. The el ectrochemical band gap is taken as the onset of oxidation minus the onset of reduction. Thus, PDPP-Boc has an electrochemical band gap of 1.34 eV estimated by CV. One of the advantages of DPV is the increased sensitivity allowing for a more accurate estimation of the onsets of oxidation and reduction. DPV was performed on PDPP-Boc (Figure 5-25b) and shows an onset for reduction at -1.02 V, and an onset for oxidation at 0.07 V, leading to an estima ted band gap of 1.09 eV. In comparison, PDPP-Boc has an optical band gap of ~1.16 eV, in the middle of the electrochemical band gaps estimated from CV and DPV. Figure 5-25. Cyclic voltammograms and diffe rential pulse voltammograms of PDPP-Boc.
174 The HOMO levels estimated from CV and DP V are -5.36 and -5.17 eV, respectively. This relatively high HOMO level suggests that PDPP-Boc is prone to air oxidation. The LUMO levels estimated from CV and DPV are very cl ose at ~ -3.9-4.0 eV. T hus the energy offset between PDPP-Boc and PCBM is merely 0.2 eV, which may cause inefficient electron transfer in the polymer/PCBM BHJ solar cells. The solar cell study as fo llows was performed by Dr Jegadesan Subbiah. The photovoltaic properties of PDPP-Boc were studied in bulk heterojunction solar ce lls using (6,6)-phenyl-C71butyric acid methyl ester (PC71BM) as the electron acceptor. All so lar cell results were collected under atmospheric conditions, and under 100 mWcm-2 simulated AM 1.5G illumination. The PDPP-Boc /PC71BM blends were spin-cast from ch lorobenzene on PEDOT-PSS coated ITO glass substrates. The photoactive layers were subj ected to thermal anneali ng at either 75 C or 180 C before the top electrode was deposited. Figure 5-26 shows the cu rrent density-voltage characteristics of the PV devices with various polymer:PC71BM compositions and the device performance is presented in Table 5-3. Figure 5-26. The J-V curve of a DPP polymer solar cell under AM 1.5 conditions with a blend ratio of a) 1:1, b) 1:2, and c) 1:3. A polymer:PC71BM ratio of 1:2 was found to be the optimum blend ratio, yielding a power conversion efficiency (PCE) of 1.44%, as s hown in Table 5-3. Here, the device annealed at 75 C exhibit a power convers ion efficiency (PCE) of 1.08%, a short-circuit current density
175 (Jsc) of 4.86 mA/cm2, an open-circuit voltage (Voc) of 0.47 V and a fill factor ( FF ) of 0.48. Annealing the device at 180 C leads to th e thermo-cleavage which improves the cell performance resulting in an increase in PCE, Voc, Jsc and FF to 1.44%, 5.88 mA/cm2, 0.49 V and 0.50, respectively. The enhancement in the device performance is perhaps due to the chemical change in the polymer film due to thermal tr eatment and associated changes in thickness and morphology of the film. Similar enhancement in the performance due to thermocleavage was observed for the solar cell fabricated with the bl end ratios of 1:1 and 1:3. The possible chemical transition from thermal treatment is shown in Figure 5-1. As evident for the clearance, the photoactive layer became insoluble in organic solv ent after annealing at 180 C for 10 min. Here, the performance of the solar cell with a blend ra tio of 1:1 exhibited lower performance with a PCE of 0.78% before annealing and 1.08 % after annealing; a PC E of 0.87% before annealing and 1.21% after annealing was obtai ned in the case of a polymer/PC71BM ratio of 1:3. Table 5-3. Photovoltaic performance of PDPP/PC71BM solar cells with different blend ratio. The overall performance of PDPP-Boc remains moderate with a PCE of 1.44% at the optimum blend ratio, in spite of the enhancement in the PCE after thermocleavage. This can be in part explained by an inefficient char ge transfer between the polymer and PC71BM, due to the small energy offset between their LUMO levels (~0.2 eV). In addition, these results are also likely to correlate with the polymers charge m obility. In this regard, charge mobilities of PDPP-
176 Boc and PDPP in pure films and in blends are measured by modeling of the J-V response of space-charge limited current diode devices. These results are exhibited in Table 5-4. The pure PDPP-Boc polymer film presents a SCLC hole mobility of 1 x 10-5 cm2V-1s-1; while its blend film yields a hole mobility of 8.5 x 10-6 cm2V-1s-1. Compared to the hole mobilities of 4 x 10-2 cm2V-1s-1 exhibited by a known DPP-based polymer in a field effect transistor device,41 our charge mobilities are considerably lower (Note: the direct comparison of mobilities from SCLC and OFET is usually not valid, considering the different devi ce structures for diodes and transistors, but it gives us a rough idea of how bi g the difference is). The low hole mobility in the blend may be partially responsible for the m oderate photovoltaic perfor mance. Upon cleavage, the hole mobilities from PDPP are still relatively low around 6.2 x 10-5 cm2V-1s-1 for the pure film and the blend, but they are si gnificantly higher than those of PDPP-Boc This observation may explain the boost in PC E after cleavage of the tBoc group. Table 5-4. Transport properties of PDPP-Boc and PDPP polymer and their blends with PC71BM in a ratio of 1:2 PDPP gives a PCE of 1.44% in our preliminary study. Interestingly, this value still makes PDPP one of the best performing thermocleav able polymers for organic solar cells.303-304,306-309 This suggests that more effort shoud put fort h to improve the photovoltaic performance of thermocleavable conjugated polymers. As we exp ect higher performance from this type of material is indeed feasible. In the case of DPP based conjug ated polymers, more enhanced performance may be realized if we can improve the charge mobility, as well as enlarge the energy offset to provide for more efficient charge transfer. In a ddition, it may also be interesting
177 to develop a processing method fo r device fabrication that can hi ghlight the solubility switching upon cleavage. In that sense, multi-layer devices with complementary absorbing polymers via solution processing may be an option to improve solar cell performance. 5.5 Conclusion In this chapter, we have explored two cl asses of relatively undeveloped materials based on diketopyrrolopyrrole chemistry fo r field-effect transistor and solar cell applications. Using DPP-1 as an example, we applied an amphiphilic mo lecular design to obtain a discrete oligomer with a well-defined molecular st ructure. With the amphiphilic nature and the special cruciformshape, DPP-1 exhibits a strong tendency to self-assemble into highl y ordered nanostructures in solution, in the bulk, and on surfaces. Solution proce ssed field-effect transistor devices yielded hole mobilities of up to 4 x 10-4cm2V-1s-1. PCEs of DPP-1 /PC61BM solar cell devices were only around 0.7%, but with high fill fact ors of ~0.58, great potential fo r further optimization exists. Encouraged by these findings, DPP-OH-n (n = 1, 2, 3 and 4), together with their control molecules DPP-C12 and DPP-TEG were prepared. DSC studies of these molecules suggests that phase transition behaviors can be largely altered by varyi ng the passive solubilizing groups, which opens vast possibilities to control morphol ogy in thin-film devices via the material design approach. Thermocleavable polymers are of interest, motivated by their photochemical stability, improvement of the chromophore de nsity in the device film, and si gnificant advantages in terms of processing (solubility/insolubi lity switching). In this chapte r, we introduced a DPP-based thermocleavable conjugated polymer ( PDPP-Boc ) and studied its utilizat ion as an electron donor in bulk-heterojunction solar cells. A typical PDPP-Boc /PC71BM bulk-heterojunction solar cell showed a PCE of 1.08%, Voc of 0.47, Jsc of 4.68 mA/cm2 and FF of 0.48. Upon cleavage of the solubilizing (protecting) groups, the PDPP /PC71BM photovoltaic device exhibited a PCE of
178 1.44%, Voc of 0.49, Jsc of 5.88 mA/cm2 and FF of 0.50. A significant enhancement has been observed. This encouraging observation will draw mo re attention to the use of thermocleavable materials for solar cell applications, and furthe r investigation is require d to fully unveil their potential for photovoltaic applications. 5.6 Experimental Details N,N'-bis-(10-(3,6,9-trioxadecyl)-3 ,6-di(5-thienyl)-1,4-diketopyrrolo[3,4-c]pyrole (5-2). To a mixture of 3,6-Bis-(2-thienyl)-1,4-dioxopy rrolo[3,4-c]pyrrole (4.50 g, 15 mmol), tetrabutylammonium bromide (483 mg, 1.5 mm ol) and potassium carbonate (10.35 g, 75 mmol) in DMF (120 mL) and 2-(2-(2-methoxyethoxy)et hoxy)ethyl 4-methylben zenesulfonate (18.1 g, 57 mmol) in DMF (30 mL) was adde d. The mixture was heated at 120 oC and stirred for 40 h. The DMF was removed under vacuum. To the con centrated solution, distilled water (150 mL) was added. The organic phases were extracted with CHCl3 (5 x 50 mL), washed with brine and dried over MgSO4. The viscous purple liquids obt ained were purified by silica gel chromatography, eluting with CH2Cl2 / Acetone (9:1 to 8:2) to give red solids (4.1 g, 46%) 1H-NMR (CDCl3) : 8.69 (dd, J1 = 1.2 Hz, J2 = 3.9 Hz, 2H), 7.56 (dd, J1 = 1.2 Hz, J2 = 5.1 Hz, 2H), 7.17 (dd, J1 = 3.9 Hz, J2 = 5.1 Hz, 2H), 4.19 (t, J = 6.3 Hz, 4H), 3.71 (t, J = 6 Hz, 2H), 3.593.39 (m, 16H), 3.26 (s, 6H). 13C-NMR (CDCl3) : 161.6, 140.5, 135.0, 131.1, 129.9, 128.6, 108.0, 72.1, 70.9, 70.7, 69.1, 59.2, 42.0. HRMS (ESI-TOF) Calculated for C28H36N2O8S2 (M+H)+: 593.1986, found: m/z 593.1982. Anal. Calcd for C28H36N2O8S2: C, 56.74; H, 6.12; N, 4.73. Found: C, 55.88; H, 6.12; N, 4.67. N,N'-bis-(10-(3,6,9-trioxadecyl)-3,6-di(5-bromothi enyl)-1,4-diketo-pyrrolo[3,4-c]pyrrole (53). To a solution of N,N'-bis-(10 -(3,6,9-trioxadecyl)-3,6-di(5thienyl)-1,4-diketo-pyrrolo[3,4c]pyrrole (593 mg 1 mmol) in CHCl3 (50 mL), N-bromosuccimi de (NBS) (374 mg, 2.1mmol) was added in one portion. The solution was stirre d for 48 h at room temperature in the dark
179 (covered by aluminum foil). After the reaction was complete (as indicated by TLC), distilled water was added. The organic phases were washed with brine and dried over MgSO4. The dark purple solids obtained were purified by silica gel chroma tography, eluting with CH2Cl2/ Acetone (9:1) to give violet-pur ple solids (4.56 g, 61%) 1H-NMR (CDCl3) : 8.46 (d, J = 4.5 Hz, 2H), 7.16 (d, J = 4.5 Hz, 2H), 4.12 (t, J = 6.0 Hz, 4H ), 3.73 (t, J = 6.0 Hz, 4H), 3.61-3.43 (m, 16H), 3.10 (s, 6H).13C-NMR (CDCl3) : 161.3, 139.6, 135.0, 131.5, 131.3, 119.4, 108.1, 72.1, 70.9, 70.7, 69.1, 59.1, 42.4. HRMS (ESI-TOF) Calculated for C28H34Br2N2O8S2 (M+H)+: 751.0183, found: m/z 751.0161. Anal. Calcd for C28H34Br2N2O8S2: C, 44.81; H, 4.57; N, 3.73. Found: C, 44.51; H, 5.53; N, 3.71. N,N'-bis-(10-(3,6,9-trioxad ecyl)-3,6-di(5-dodecylter thienyl)-1,4-diketo-pyrrolo[3,4c]pyrrole ( DPP-1 ).To a Schlenk flask (125 mL) charged with a stirring bar were N,N'-bis-(10(3,6,9-trioxadecyl)-3,6-di(5-bromothienyl)-1,4-dik eto-pyrrolo[3,4-c]pyrrole (750 mg, 1 mmol), tributyl(5'-dodecyl-2,2'-bithiophen-5-y l)stannane (1.56 mg, 2.50 mmol), Pd2(dba)3 (20.8 mg) and P(o-tyl)3 (24 mg) was added. Subsequently, it was purged with argon gas, followed by applying vacuum for 10 min; this process was repeated th ree times and the flask was refilled with argon. Degassed toluene (20 mL) was injected through a septum. The resulting mixture was heated up to 80 oC under argon and stirred for 20 h. The reacti on mixture was poured into hexane and the precipitates were collected by v acuum filtration. The dark oliv e-green solids obtained were purified by silica gel chromat ography, eluting with CHCl3 / Acetone (from 96:4 to 92:8) to give purple solids (708 mg, 49%). 1H-NMR (CDCl3) : 8.78 (d, J = 4.2 Hz, 2H), 7.20 (d, J = 4.2 Hz, 2H), 7.18 (d, J = 3.9 Hz, 2H), 7.00 (d, J = 3.9 Hz, 2H), 6.99 (d, J = 3.6 Hz, 2H), 6.67 (d, J = 3.6 Hz, 2H), 4.27 (t, J = 6.0 Hz, 4H), 3.82 (t, J = 6.0 Hz, 4H), 3.67-3.44 (m, 16H), 3.31 (s, 6H), 2.80 (t, J = 7.5 Hz, 4H), 1.70 (quintet, 4H), 1.40-1.20 (m, 36H), 0.90 (t, J = 7.2 Hz, 6H). 13C-NMR
180 (CDCl3) : 161.4, 146.6, 143.0, 139.2, 136.5, 134.2, 128.1, 126.2, 125.2, 124.6, 124.2, 124.0, 108.4, 72.1, 71.0, 70.8, 69.2, 59.2, 42.2, 32.1, 31.8, 30.4, 29.88, 29.85, 29.8, 29.59, 29.57, 29.3, 22.9, 14.3. HRMS (ESI-TOF) Calculated for C68H92N2O8S6 (M+Na)+: 1279.5070, found: m/z 1279.5065. Anal. Calcd for C68H92N2O8S6: C, 64.93; H, 7.37; N, 2.23. Found: C, 64.81; H,7.35; N, 2.23. 1,2,3-tris(dodecyloxy)benzene (5-4) .296 A mixture of pyraga llol (6.35 g, 50 mmol, 1equiv.), n -bromododecane (39.53 g, 38 mL, 15 9 mmol, 3.15 equiv.), anhydrous K2CO3 (41.75 g, 302 mmol, 6 equiv.), and DMF (100 ml) was heated at 65 oC for 18 h under nitrogen atmosphere. Then the reaction mixture was then poured into ice-water (1000 ml) and extracted with dichloromethane (3 100 ml). The combined organic solution was washed with water (3 x 100 ml), brine and dried over anhydrous MgSO4. The solution was concentrated and precipitated into methanol (400 mL), yielding pale white so lids. The crude product was further purified by silica gel chromatography, using CH2Cl2/Hexane (1:1) as eluents. The combined fractions were concentrated and poured into methanol (200 mL), yielding white precipitates (29.8 g, 94%). 1HNMR (CDCl3) : 6.91 (t, J = 8.6 Hz, 1H), 6.55 (d, J = 8.6 Hz, 2H), 4.01-3.32 (m, 6H), 1.85-1.28 (m, 60H), 0.88 (t, J = 6.0 Hz, 9H); 13C-NMR (CDCl3) : 153.6, 138.6, 123.2, 106.9, 73.4, 69.1, 32.1, 30.6, 30.0, 29.9, 29.86, 29.64, 29.61, 29.58, 26.4, 26.3, 22.9, 14.2. 4,4,5,5-tetramethyl-2-(3,4,5-tris(dodecyl oxy)phenyl)-1,3,2-dioxaborolane (5-5) To a 100 mL two-neck flask containing a magnetic st ir bar was added 1,2,3-tris(dodecyloxy)benzene (3.16 g, 5 mmol), bis(pinacolato)diboron (1.27 g, 5 mmol) and 4,4-di-tert-butyl-2,2'-bipyridine (40.3 mg, 0.15 mmol). The flask was evacuated a nd back-filled with ar gon three times, after which degassed heptane (50 mL) was transf erred to the mixture. [Ir(OMe)(COD)]2 (0.075 mmol) was added under argon. The reaction was imme diately immersed in an oil bath at 80 oC
181 and stirred for 24 hours at that temperature. Th e reaction was cooled to RT, the mixture was filtered through a pad of silica gel and washed with methylene chloride. The solution was concentrated and poured into methanol ( 100 mL), yielding white solids (3.48 g, 92%). 1H-NMR (CDCl3) : 6.99 (s, 2H), 4.05-3.94 (m, 6H), 1.92-1.68 (m, 6H), 1.52-1.25 (m, 66H), 0.87 (t, J = 6.0 Hz, 9H); 13C-NMR (CDCl3) : 153.1, 141.4, 113.0, 83.9, 73.6, 69.3, 32.2, 30.6, 30.0, 29.94, 29.90, 29.84, 29.72, 29.66, 29.62, 29.60, 26.4, 25.1, 22.9, 14.3; HRMS (APCI-TOF) Calculated for C48H89BO5 (M + H+): 757.6908 Found: m/z 757.6908. 2-(3,4,5-tris(dodecyloxy)phenyl)thiophene (5-6) In a flame-dried Schlenk flask (50 mL), 4,4,5,5-tetramethyl-2-(3,4,5-tris(dodecyloxy)ph enyl)-1,3,2-dioxaborol ane (5-5) (6.81 g, 9 mmol), 2-bromothiophene (4.40 g, 27 mmol), Tris(dibenzylideneacetone)dipalladium(0) (117 mg) and P(o-tyl)3 (138 mg) were added. The flask was evacuated and back-filled with argon three times, after which degassed toluene ( 15 mL) and tetraethylammonium hydroxide (6.2 mmol, 20 wt aqueous solution) was transferred to the mixture through a septum. The resulting solution was heated up to 90 oC under argon and stirred for 20 h. The solvent was concentrated and poured into cold methanol (200 mL). The preci pitates were filtered a nd purified by silica gel chromatography, eluting with CH2Cl2-hexane (1:2) to give a yellowish oil (the oil solidified over a few days. 6.30 g, 98%).1H-NMR (CDCl3) : 7.24-7.20 (m, 2H), 7.06-7.03 (m, 1H), 6.84 (s, 2H) 4.07-4.03 (m, 6H), 1.92-1.80 (m, 6H), 1.621.30 (m, 54H), 0.97 (t, J = 6.0 Hz, 9H); 13CNMR (CDCl3) : 153.5, 144.9, 138.4, 129.8, 127.8, 124.1, 122.7, 105.2, 73.5, 69.3, 32.1, 30.5, 29.9, 29.86, 29.84, 29.83, 29.63, 29.61, 29.57, 29.55, 26.34, 26.30, 22.8, 14.2; HRMS (APCITOF) Calculated for C46H80O3S (M + H+): 713.5901 Found: m/z 713.5908. 4,4,5,5-tetramethyl-2-(5-(3,4,5-tris(dod ecyloxy)phenyl)thiophen-2-yl)-1,3,2dioxaborolane (5-7) To a 100 mL two-neck flask contai ning a magnetic stir bar was added 2-
182 (3,4,5-tris(dodecyloxy)phenyl)thioph ene (5-6) (3.56 g, 5 mmol), bi s(pinacolato)diboron (1.27 g, 5 mmol) and 4,4-di-tert-butyl -2,2'-bipyridine (40 mg, 0.15 mmol) The flask was evacuated and back-filled with argon three time s, after which degassed heptane (50 mL) was transferred to the mixture. [Ir(OMe)(COD)]2 (0.075 mmol, 49 mg) was added under argon. The reaction was immediately immersed in an oil bath at 80 oC and stirred for 24 hours at that temperature. The reaction was cooled to RT, the mixture was filter ed through a pad of silica gel and washed with methylene chloride. The solution was concentrated and poured into methanol (100 mL), yielding white solids (2.1 g, 50%). 1H-NMR (CDCl3) : 7.57 (d, J = 3.6Hz, 1H), 7.29 (d, J = 3.6Hz, 1H), 6.83 (s, 2H), 4.01-3.94 (m, 6H), 1.85-1.72 (m, 6H), 1.52-1.25 (m, 66H), 0.90 (t, J = 6.6 Hz, 9H); 13C-NMR (CDCl3) : 153.6, 138.6, 129.6, 124.3, 105.2, 84.3, 73.7, 69.4, 32.1, 30.6, 30.0, 29.92, 29.87, 29.62, 29.59, 26.35, 26.30, 25.0, 22.9, 14.3; HRMS (APCI-TOF) Calculated for C52H91BO5S (M+): 839.6712 Found: m/z 838.6757. 5-(3,4,5-tris(dodecyloxy)phenyl)-2,2'-bithiophene (5-8) In a flame-dried Schlenk flask (50 mL), 4,4,5,5-tetramethyl-2-(3,4,5-tris(dod ecyloxy)phenyl)-1,3,2-dioxaborolane (5-5) (2.27 g, 3 mmol), Tris(dibenzylideneacetone)d ipalladium(0) (39 mg), P(o-tyl)3 (46 mg) and 5-bromo2,2-bithiophene (1.10 g, 4.5 mmol), were added. The flask was evacuated and back-filled with argon three times, after which degassed tolu ene (15 mL) and tetraethylammonium hydroxide (4.5mL, 20 wt aqueous solution) was transferred to the mixture through a septum. The resulting solution was heated up to 90 oC under argon and stirred for 20 h. The solvent was concentrated and poured into cold methanol (200 mL). The preci pitates were filtered a nd purified by silica gel chromatography, eluting with hexane/toluene (3:1) to give a yellow solid (2.05 g, 86%).1H-NMR (CDCl3) : 7.20-7.18 (m, 2H), 7.12 (s, 2H), 7.03-7.00 (m, 1H), 6.81 (s, 2H) 4.07-4.00(m, 6H), 1.92-1.78 (m, 6H), 1.62-1.30 (m, 54H), 0.95 (t, J = 6.3 Hz, 9H); 13C-NMR (CDCl3) : 153.6,
183 143.6, 138.4, 137.6, 136.3, 129.4, 127.9, 124.5, 124.2, 123.4, 123.35, 104.7, 73.6, 69.3, 32.1, 30.6, 29.96, 29.94, 29.92, 29.90, 29.86, 29.82, 29.63, 29.60, 29.58, 26.35, 26.31, 22.9, 14.3; HRMS (APCI-TOF) Calculated for C50H82O3S2 (M+): 795.5778 Found: m/z 795.5750. 4,4,5,5-tetramethyl-2-(5'-(3,4,5-tris(dodecylo xy)phenyl)-2,2'-bithiophen-5-yl)-1,3,2dioxaborolane (5-9) To a 100 mL two-neck flask containi ng a magnetic stir bar was added 5(3,4,5-tris(dodecyloxy)phenyl)-2,2'-bithiophene (5-8 ) (1.20 g, 1.5 mmol), bi s(pinacolato)diboron (230 mg, 0.9 mmol) and 4,4-di-tert-butyl-2,2'-bi pyridine (12 mg). The flask was evacuated and back-filled with argon three time s, after which degassed heptane (50 mL) was transferred to the mixture. [Ir(OMe)(COD)]2 (16 mg) was added under argon. The reaction was immediately immersed in an oil bath at 80 oC and stirred for 24 hours at that temperature. The reaction was cooled to RT, the mixture was filtered through a pad of silica gel and washed with methylene chloride. The solution was concentrated and pour ed into methanol (100 mL), yielding yellow solids (700 mg, 50%). 1H-NMR (CDCl3) : 7.55 (d, J = 3.6Hz, 1H), 7.25 (d, J = 3.6Hz, 1H), 7.19 (d, J = 3.9Hz, 1H), 7.13 (d, J = 3.9Hz, 1H ), 6.78 (s, 2H), 4.05-3.96 (m, 6H), 1.86-1.73 (m, 6H), 1.52-1.25 (m, 66H), 0.90 (t, J = 6.6 Hz, 9H); 13C-NMR (CDCl3) : 153.6, 144.4, 138.7, 138.2, 136.2, 129.4, 125.3, 124.8, 123.7, 104.9, 84.3, 73.8, 69.5, 32.1, 30.6, 30.0, 29.95, 29.92, 29.86, 29.83, 29.63, 29.61, 29.58, 26.34, 26.32, 24.9, 22.9, 14.3; HRMS (APCI-TOF) Calculated for C56H93BO5S2 (M+): 920.6562 Found: m/z 920.6616. 2-(3,4,5-tris(dodecyloxy)phenyl )-5,2':5',2''-terthiophene ( 5-10) In a flame-dried Schlenk flask (50 mL), 4,4,5,5-tetramethyl-2-(3,4,5-tris(d odecyloxy)phenyl)-1,3,2-dioxaborolane (5-5) (2.27 g, 3 mmol), 2-Bromo-5,2':5',2 ''-terthiophene (981 g, 3 mmol), Tris(dibenzylideneacetone)dipalla dium(0) (39 mg), and P(o-tyl)3 (46 mg) were added. The flask was evacuated and back-filled with argon three tim es, after which degassed toluene (8 mL) and
184 tetraethylammonium hydroxide (4.5mL, 20 wt aque ous solution) was transferred to the mixture through a septum. The resulting solution was heated up to 80 oC under argon and stirred for 20 h. The solvent was concentrated and poured into cold methanol ( 200 mL). The precipitates were filtered and purified by silica gel chromatography, eluting with hexane/toluene (3:1) to give a yellow solid (1.90 g, 72%).1H-NMR (CDCl3) : 7.20-7.17 (m, 3H), 7.16-7.08 (m, 3H), 7.02-6.99 (m, 1H), 6.79 (s, 2H) 4.07-3.98 (m, 6H), 1.921.78 (m, 6H), 1.62-1.30 (m, 54H), 0.95 (t, J = 6.3 Hz, 9H); 13C-NMR (CDCl3) : 153.6, 143.7, 138.5, 137.3, 136.4, 136.2, 136.0, 129.4, 128.0, 124.5, 124.4, 124.1, 123.7, 123.5, 104.7, 73.7, 69.4, 32.1, 30.6, 29.97, 29.93, 29.91, 29.90, 29.88, 29.65, 29.61, 29.59, 26.36, 26.33, 22.9, 14.3; HRMS (APCI-TOF) Calculated for C54H84O3S3 (M+): 1002.6439 Found: m/z 1002.6464. 12-bromo-2,2,3,3-tetramethyl-4,7,10 -trioxa-3-siladodecane (5-12) Triehylamine (13.7 mL, 98 mmol, 1.2 eq), (N,N-dimethylpyridin -4-amine) DMAP (8.2 mmol, 0.1 eq) and tert butylchlorodimethylsilane (TBDMSCl) (13.56 g, 90 mmol, 1.1 eq) we re dissolved in CH2Cl2 (100 mL). 2-(2-(2-hydroxyet hoxy)ethoxy)ethyl 4-methylbenzen esulfonate (24.90 g, 82 mmol) was added dropwise over 10 min, then the mixture was stirred at rt for 6 h. The reaction was poured into H2O and extracted with ethyl acetate. Th e viscous liquid was purified by silica gel column, yielding a colorl ess liquid (31.2 g, 91%). 1H-NMR (CDCl3) :7.79 (d, J = 6.4 Hz, 2H), 7.33 (d, J = 6.4Hz, 2H), 4.15 (t, J = 4.8 Hz, 2H), 3.74 (t, J = 6.4 Hz, 2H), 3.74 (t, J = 4.8 Hz, 2H), 3. 56 (s, 4H), 3.51 (t, J = 6.4 Hz, 4H), 2.42 (s, 3H), 0.87 (s, 12H), 0.03 (s, 6H); 13C-NMR (CDCl3) : 144.9, 133.2, 130.0, 128.2, 72.9, 71.0, 70.8, 69.4, 68.9, 62.9, 26.1, 25.8, 21.8, 18.5, 5.09. 2,5-bis(2,2,3,3-tetramethyl-4,7,10-trioxa-3-s iladodecan-12-yl)-3,6-di(thiophen-2yl)pyrrolo[3,4-c]pyrrole1,4(2H,5H)-dione (5-13) To a mixture of 3,6-Bis-(2-thienyl)-1,4-
185 dioxopyrrolo[3,4-c]pyrrole (3.0 g, 10 mmol), a nd potassium carbonate (8.3 g, 60 mmol) in DMF (120 mL), 12-bromo-2,2,3,3-tetramethyl-4,7,10-tr ioxa-3-siladodecane (10.87 g, 26 mmol) in DMF (60 mL) was added. The mixture was heated up to 120 oC and stirred for 24 h. The suspension was poured into water (1L). Th e organic phases were extracted by CHCl3 (5 x 50 mL), washed by brine and dried over MgSO4. Viscous dark red liquids obtained were purified by silica gel chromatogr aphy, eluting with CH2Cl2 / Acetone (97:3) to give purple-red solids (5.3 g, 61%). 1H-NMR (CDCl3) : 8.72 (dd, J1 = 1.2 Hz, J2 = 3.9 Hz, 2H), 7.56 (dd, J1 = 1.2 Hz, J2 = 5.1 Hz, 2H), 7.20 (dd, J1 = 3.9 Hz, J2 = 5.1 Hz, 2H), 4.22 (t, J = 6.3 Hz, 4H), 3.75 (t, J = 6 Hz, 4H), 3.66-3.65 (m, 4H), 3.57-3.53 (m, 8H), 3.453.42 (m, 4H), 0.82 (s, 18H). 0.01 (s, 12H); 13CNMR (CDCl3) : 161.6, 140.4, 134.8, 130.9, 129.7, 128.5, 107.9, 72.8, 70.8, 70.7, 69.0, 62.8, 41.9, 26.0, 18.4, -5.1. HRMS (ESI-TOF) Calculated for C38H60N2O8S2Si2 (M+H)+: 793.3402, found: m/z 793.3423. 3,6-bis(5-bromothiophen-2yl)-2,5-bis(2-(2-(2-ydroxyeth oxy)ethoxy)ethyl)pyrrolo[3,4c]pyrrole-1,4(2H,5H)-dione (5-14) To a solution of 2,5-bis( 2,2,3,3-tetramethyl-4,7,10-trioxa-3siladodecan-12-yl)-3,6-di(thiophen-2-yl)pyrrolo[ 3,4-c]pyrrole-1,4(2H,5H)-dione (5-13) (3.0 g, 3.78 mmol) in CHCl3 (40 mL), N-bromosuccimide (NBS) (1.62 g, 9.08mmol) was added in three portions over 1h. The solution was stirred 48 h unde r room temperature in the dark (covered by alumina foil). After the reaction was comple ted (indicated by TLC), HCl (4M, 30 mL) was added to cleave TBDMS protecting groups. The so lution was vigorously s tirred for another hour. The organic phases were then extracted with chloroform, washed by brine and dried over MgSO4. Dark purple solids obtained were purifie d by silica gel chromatography, eluting with CH2Cl2/ Acetone (9:1) to give viol et-purple solids (1.70 g, 63%) Before Cleavage 1H-NMR (CDCl3) : 8.47 (d, J = 5.2 Hz, 2H), 7.17 (d, J = 5.2 Hz, 2H), 4.14 (t, J = 6.0 Hz, 4H), 3.77-3.68
186 (m, 8H), 3.60-3.57 (m, 8H), 3.49 (t, J = 5.4, 4H), 0.85 (s, 18H). 0.02 (s, 12H); 13C-NMR (CDCl3) : 161.4, 139.6, 135.0, 131.5, 131.3, 119.5, 108.1, 72.9, 71.0, 70.8, 69.1, 62.9, 42.4, 26.1, 18.5, 5.1. H HRMS (ESI-TOF) Calculated for C38H58Br2N2O8S2Si2 (M+H)+: 951.1596, found: m/z 951.1530. After cleavage 1H-NMR (CDCl3) : 8.45 (d, J = 5.2 Hz, 2H), 7.16 (d, J = 5.2 Hz, 2H), 4.15 (t, J = 6.0 Hz, 4H), 3.75-3.71 (m, 8H), 3.60-3.57 (m, 8H), 3.45(t, J = 5.4, 4H); 13CNMR (CDCl3) : 161.4, 139.6, 135.2, 131.7, 131.2, 119.6, 108.1, 72.7, 70.9, 70.6, 69.1, 61.8, 42.3. The general method to prepare DPP-OH-n, DPP-TEG and DPP-C12 A Schlenk flask charged with 5-14 (0.2 mmol), Pd2(dba)3 (2.5 mol %), [(C4H9)3PH]+BF4 (5 mol %, ), CuI (10 mol %), and were purged with argon atmosphere and vacuum for three cycles, and then filled with argon. Ar-B (2.1 eq, 5-7, 59 and 5-11, respectivel y) in degassed tetrahydrofuran (5 mL) was added through a septum, followed by an aque ous solution of CsF (6 eq, 1.2 mmol, 0.6 mL). The reaction mixture was heated up to 75 oC and stirred for 48 h. After cooling back to room temperature, the mixture was poured into me thanol (100 mL) and stirred for 30 min. The precipitates were collected by v acuum filtration. The crude products were purified by silica gel chromatography, eluting with CH2Cl2/acetone (95:5 to 90:10). The collected fractions were then concentrated and precipitated into me thanol to yield the desired products. DPP-1 1H-NMR (CDCl3) : 8.80 (d, J = 4.2 Hz, 2H), 7.35 (d, J = 4.2 Hz, 2H), 6.83 (s, 4H), 4.35 (t, J = 6 Hz, 4H), 4.05-3.96 (m, 12H) 3.83 (t, J = 6 Hz, 4H), 3.67-3.55 (m, 12H), 3.50 (t, J = 4.2, 4H), 2.63 (bs, 2H), 1.87-1.70 (m, 12H) 1.57-1.25 (m, 108H), 0.87 (t, J = 6.0Hz, 18H); 13C-NMR (CDCl3) : 161.7, 153.8, 150.8, 139.9, 139.7, 136.6, 128.4, 128.35, 124.5, 108.2, 105.5, 73.9, 72.8, 70.9, 70.6, 69.6, 69.3, 61.9, 32.1, 30.6, 29.98, 29.93, 29.88, 29.67, 29.61,
187 29.59, 26.35, 22.9, 14.3.HRMS (MALDI-TOF) Calculated for C110H184N2O14S2 (M+H)+: 1822.3223, found: m/z DPP-2 1H-NMR (CDCl3) : 8.80 (d, J = 4.5 Hz, 2H), 7.24 (dd, J1 = 4.5 Hz, J2 = 3.9 Hz, 2H), 7.12 (d, J = 3.9 Hz, 2H), 6.75 (s, 4H), 4.31 (t J = 6 Hz, 4H), 4.20-3.94 (m, 12H), 3.83 (t, J = 6 Hz, 4H), 3.69-3.59 (m, 12H), 3.52 (t, J = 4.2, 4H), 2.57 (bs, 2H), 1.87-1.72 (m, 12H), 1.57-1.25 (m, 108H), 0.87 (t, J = 6.0Hz, 18H); 13C-NMR (CDCl3) : 161.5, 153.7, 150.8, 146.0, 143.3, 139.3, 138.7, 136.6, 134.8, 128.9, 127.9, 126.5, 124.6, 123.9, 108.4, 104.8, 73.8, 72.8, 71.0, 70.7, 69.5, 69.3, 61.95, 32.1, 30.5, 29.95, 29.91, 29.89, 29.86, 29.81, 29.65, 29.59, 29.56, 26.3, 22.9, 14.3. H HRMS (MALDI-TOF) Calculated for C118H188N2O14S4 (M+H)+: 1987.0203, found: m/z 1987.3120. DPP-3 1H-NMR (CDCl3) : 8.80 (d, J = 4.2 Hz, 2H), 7.04 (d, J = 4.2 Hz, 2H), 6.95-6.88 (m, 8H), 6.65 (s, 4H), 4.24 (bs, 4H), 3.95-3.91 (m, 12H), 3.82 (m, 4H), 3.75-3.64 (m, 12H), 3.52 (t, J = 4.5, 4H), 3.09 (bs, 2H), 1.80-1.70 (m, 12H) 1.47-1.25 (m, 108H), 0.87 (t, J = 6.0Hz, 18H); 13C-NMR (CDCl3) :161.2, 153.6, 144.3, 142.8, 138.6, 138.5, 136.9, 135.2, 134.4, 129.0, 127.8, 126.2, 125.1, 124.5, 124.2, 123.5, 108.2, 104.4, 73.8, 72.7, 70.9, 70.6, 69.4, 61.9, 32.2, 30.6, 30.0, 30.0, 29.94, 29.76, 29.71, 29.6, 26.4, 22.9, 14.3. H HRMS (MALDI-TOF) Calculated for C126H192N2O14S6 (M+H)+: 2151.2803 found: m/z 2151.2789. DPP-TEG 1H-NMR (CDCl3) : 8.80 (d, J = 4.2 Hz, 2H), 7.26 (dd, J1 = 4.5 Hz, J2 = 3.9 Hz, 2H), 7.13 (d, J = 3.9 Hz, 2H), 6.76 (s, 4H), 4.30 (t, J = 6 Hz, 4H), 4.04-3.97 (m, 12H), 3.83 (t, J = 6 Hz, 4H), 3.69-3.55 (m, 12H), 3.45 (m 4H), 3.31 (s, 6H)1.87-1.72 (m, 12H), 1.57-1.25 (m, 108H), 0.88 (t, J = 6.0Hz, 18H); 13C-NMR (CDCl3) : 161.5, 153.7, 145.9, 143.2, 139.3, 138.8, 136.4, 134.9, 128.9, 128.1, 126.4, 124.5, 123.9, 108.4, 104.7, 73.8, 72.1, 71.0, 70.8, 69.5,
188 69.3, 59.2, 32.1, 30.6, 30.0, 29.92, 29.91, 29.87, 29.66, 29.60, 29.58, 36.3, 22.9, 14.3. HRMS (MALDI-TOF) Calculated for C120H192N2O14S4 (M+H)+:2015.3300, found: m/z 2015.3366. DPP-C12 1H-NMR (CDCl3) : 8.92 (d, J = 4.2 Hz, 2H), 7.29 (d, J = 4.2 Hz, 2H), 7.257.23 (m, 4H), 7.15 (d, J = 3.6 Hz, 2H), 6.77 (s, 4H), 4.15-3.95 (m, 20H), 1.90-1.70 (m, 16H),1.57-1.21(m, 140H), 0.88 (m, 24H); 13C-NMR (CDCl3) : 161.4, 153.7, 145.9, 143.1, 139.0, 138.9, 136.8, 135.0, 129.0, 128.1, 126.2, 124.8, 123.9, 108.4, 104.9, 73.8, 69.5, 32.1, 31.8, 30.6, 30.2, 29.99, 29.94, 29.89, 29.77, 29.67, 29.62, 29.60, 29.47, 27.1, 26.4, 22.9, 14.3. HRMS (MALDI-TOF) Calculated for C130H212N2O8S4 (M+H)+: 2059.5232, found: m/z 2059.5215. N,N'-bis-(t-butoxycarbonyl)-3,6-di(2-thienyl)1,4-diketo-pyrrolo[3,4-c]pyrrole (5-15). This compound was prepared by foll owing the literature procedure.83 Briefly described as follows: To a solution of 3,6-Bis-(2-t hienyl)-1,4-dioxopyrrolo[3,4-c]pyrrole ( 1 ) (3.0 g, 10 mmol) and DMAP (1.22 g, 10 mmol) in anhydrous THF (100 mL) was added a solution of di-tert-butyl dicarbonate (5 g, 22.9 mmol) in THF (20 mL) dropw ise over 1 h. The mixture was stirred for 16 h and poured into water. The organic phase was extracted with chloroform, washed with brine and dried over MgSO4. The solvent was concentrated under reduced pressure. Petroleum ether (100 mL) was added. The resulting solid was filte red, washed with petroleum ether and purified by silica gel column chroma tography, eluting with CHCl3 to give shiny red solids (3.0 g, 60 %). 1H-NMR (CDCl3) : 8.22 (dd, J1 = 3.9 Hz, J2 = 1.2 Hz, 2H), 7.63 (dd, J1 = 4.8 Hz, J2 = 1.2 Hz, 2H), 7.19 (dd, J1 = 4.8 Hz, J2 = 3.9 Hz, 2H), 1.58 (s, 18H); 13C-NMR (CDCl3) : 159.2, 149.0, 138.1, 134.1, 132.0, 129.8, 128.2, 110.4, 86.1, 27.8; HRMS (APCI-TOF) Calculated for C24H24N2O6S2 (M + Na+): 523.0968 Found: m/z 523.0985. N,N'-bis-(t-butoxycarbonyl)-3,6-di(5-bromothieny l)-1,4-diketo-pyrrolo[3,4-c]pyrrole (5-15). This compound was prepared by fo llowing the literature procedure.83 Briefly described as
189 follows: Compound 2 (1 g, 2 mmol) was dissolved in chloroform (30 mL) and cooled down to 0 oC. NBS (748 mg, 4.2 mmol) was added portionw ise over 2h. The reaction was monitored by TLC. After the reaction was complete, the mixtur e was poured into a methanol/water mixture (200 mL 50 mL). The resulting solid was filte red, washed with methanol and purified by silica gel column chromatography, eluting with CHCl3 to yield purple solids (1.0 g, 77 %). 1H-NMR (CDCl3) : 8.08 (d, J = 4.2 Hz, 2H), 7.15 (d, J = 4.2 Hz, 2H), 1.61 (s, 18H); 13C-NMR (CDCl3) : 158.2, 149.1, 137.0, 134.7, 131.4, 131.3, 121.0, 110.7, 86.5, 28.0; HRMS (APCI-TOF) Calculated for C24H22Br2N2O6S2 (M + Na+): 680.9159 Found: m/z 680.9187. PDPP-Boc To a flame-dried Schlenk flask (50 mL), compound 3 (658.38 mg, 1 mmol), Tris(dibenzylideneacetone)dipalladium(0) (36 mg) and P(o-tyl)3 (24 mg) were added. The flask was evacuated and back-filled with argon th ree times, after which degassed (3,3-bis(((2ethylhexyl)oxy)methyl)-3,4-dihydro-2H -thieno[3,4-b][1,4]dioxepine-6,8diyl)bis(trimethylstanna ne) (766.29 mg, 1 mmol) in toluene (30 mL) was transferred to the flask through a septum. The resulting solution was heated up to 85 oC under argon and stirred for 36 h, after which the crude polymer was precipitated in methanol. The precipitate was collected in a Soxhlet thimble and was extracted with metha nol, hexanes, and chloroform. The chloroform fraction was reduced in volume, precipitated in methanol, and filtered over a 4.5 m PTFE filter. The polymer was collected and dr ied under vacuum in a vial at 60 C overnight, resulting in 799 mg (85%) of dark blue solids. 1H-NMR: 8.52 (b, 4H), 4.40-4.0 ( b, 4H), 3.82-3.18 (b, 8H), 2.310.75 (b,48H). GPC (THF): Mn = 26,800 g/mol, PDI = 2.2.
190 CHAPTER 6 ISOINDIGO-BASED SEMICONDUCTING MATERIALS FOR PHOTOVOLTAIC APPLICATIONS 6.1 Introduction The field of organic solar cells is largely driven and fostered by continuously emerging new materials since the discovery of bulk-hete rojucntion (BHJ) for photovoltaic devices. As this dissertation was being prepar ed, several new small molecules and polymers have been have been added to the growing list of active mate rials in BHJ solar cells with PCEs above ~4%,61,310312 which are not listed in Figure 1-18. In ot her words, these high performance photovoltaic materials were released one afte r the other over the cour se of three months. Thus, it is fair to state that design and Synthesis of new materials plays a central ro le in pushing the frontiers of organic solar cells. In the previous several chapters, focus was placed on using known donor and acceptor moieties to build oligomers and polymers with de sired optical and electroc hemical properties. In particular, benzothiadiazole as an electron-accep ting building block is frequently used in the previous studies (Chapters 3 and 4) and in literature.309,312-313 In this chapter, we focus on a new electron-deficient building block, namely isoindigo. Isoindigo is a naturally occurr ing pigment. Indigo is its more famous struct ural isomer. Since its importance in dyes, pigments, as well as intermediates in drug development, a number of approaches have been developed to prepare isoindigo derivatives.314 Among them, the most widely us ed method is the acid-catalyzed condensation of isatin and oxindole derivatives.315-316 Another commonly used approach is to utilize a single electron tr ansfer reaction by reacting isatin derivatives with hexaethylphosphinetriamine.316 Surprisingly, the use of isoindigo as an electr on acceptor has not been explored within the context of functional electroa ctive materials. Based on usi ng isoindigo as a new electron-
191 deficient acceptor, we first initiate the study of isoindigo-based electr oactive pi-conjugated materials. We have a prepared a wide range of isoindigo-based small molecules, donor-acceptor type conjugated polymers, as well as a new cl ass of n-type conjugated polymers containing only electron-deficient repeat units. Through our efforts the scope of isoindigo based electroactive materials has been expanded to a great extent. 6.2 Isoindigo-based Oligothiophenes Portions of the material in this se ction have been previously reported.317 Encouraged by recent reports on diketopyrrolopyrrole oligothiophenes as donor materials in molecular BHJ solar cells, where PCEs of 2.2 4.4 % have been demonstrated,41,60,76,78,83,285,295 we here investigate isoindigo-based oligot hiophenes as electron donors in the same type of photovoltaic device. As shown in Figure 6-1, donor-acceptor-donor (DAD, I-1 ) and acceptor-donor-acceptor (ADA, I-2 ) isoindigo-based oligothiophenes are the two targeted molecules in this section. Their performance as donor materials in molecular BH J solar cells will be discussed as well. Figure 6-1. Chemical structur es of I-1, and I-2. 6.2.1 Synthesis of Isoindigo-Based Oligothiophenes The Synthesis of isoindigo building blocks is outlined in Figure 62. As mentioned earlier, there are two approaches that can lead to isoindigo derivatives. For its simplicity and mild reaction conditions, the acid-catalyzed condensa tion approach was chosen to prepare the isoindigo cores. Briefly, the work starts w ith the acid-catalyzed adol condensation and dehydration of commercially av ailable 6-bromoisatin and 6-bromooxindole in acetic acid under
192 nitrogen atmosphere, yielding 6,6-dibromoisoindigo 6-1 in a quantitative yield. It should be noted that a catalytic amount of concentrated hydrogen chloride is required for the reaction to proceed smoothly and to completion. Due to the existence of strong interactions as well as hydrogen bonding interactions, compound 6-1 is only slightly sol uble in common organic solvents. Subsequently shown is the N -alkylation of 6-1 using various alkyl bromides and efficient formation of soluble 6,6-dibromoisoindigo derivatives 6 2 3 4 and 5 in good to excellent yields (> 65%). As alkylation precludes any hydrogen bonding, the alkylated isoindigo derivatives become readily soluble in common organic solvents and can thus be purified by column chromatography. In the case of 6-3 and 6-4 where linear alkyl chains are used as solubilizing groups, the compounds are highly crystalline and can thus be purified by recrystallization. Figure 6-2. Synthetic scheme of isoi ndigo building blocks. a) con HCl ( cat. ), acetic acid, reflux, > 95%; b) RBr, K2CO3, DMF, 100 oC. The same route is utilized to prepare alkylated 6-bromoisoindigo 6-7 a precursor for the ADA oligothiophene, as illustrated in Figure 6-3. Interestingly, 6,6-dibromoisoindigo 6 2 and isoindigo 6-8 were also observed, in addi tion to the desired product 6 7, upon reaction of 6bromoisatin and oxindole followed by alkylation. Fortunately, the mixture can be easily separated by silica gel chromatography. The molar ratio of 6-2 6-7 and 6-8 is approximately 1:2:1 after purification. Th e unexpected formation of 6-8 and 6-2 in this sequence is likely due to
193 the acid(the first step) or base(the sec ond step) promoted retro-adol/adol reaction that scrambled the product distribution under the experi mental conditions employed. In order to test this hypodisserta tion, an equimolar mixture of 6-8 and 6-2 was stirred in the aforementioned reaction conditions. The experiment s revealed that crossover was indeed observed in the second reaction, and compound 6-7 was produced in DMF with K2CO3 at 100 oC; while no crossover was observed in the acidic condition. It wa s further found that use of anhydrous DMF and freshly dried base can minimize the scrambling. Figure 6-3. Synthetic scheme of mono-functionalized is oindigo building block. a) conc. HCl ( cat. ), acetic acid, reflux, > 95%; b) RBr, K2CO3, DMF, 100 oC. With 6-2 and 6-8 in hand, I-1 and I-2 were successfully obtained by incorporating electron rich bithiophene units via Su zuki coupling, as shown in Figure 6-4. Figure 6-4. Synthesis of I-1, and I-2. a) Pd2(dba)3, P(o-tyl)3 Et4NOH, toluene, 85 oC, 83% and 57% for I-1, and I-2, respectively.
194 6.2.2 Structural Characterization, Op tical and Electrochemical Studies UV-vis absorption spectra of th e oligomers, both in THF solution and as films on indiumtin oxide coated glass were measured and are shown in Figure 6-5. Compound I-1 and I-2 broadly absorb at wavelengths up to 688 and 655 nm in solution with mola r absorptivities of 19 500, and 47 300 L M-1 cm-1 at 579 and 560 nm, respectively, a nd are, thus, strongly absorbing through most of the visible spectrum as desired for solar cell applicati ons. Markedly different from other donor-acceptor systems, there exists only a very shallow absorption valley between two absorption bands for both molecules, which ma kes them appear virtually black in solution. The spectra broaden and extend to 744 and 703 nm in the solid state with a peak emerging at 660 and 636 nm for I-1 and I-2 indicating strong interactions in their crystalline forms. Upon annealing at 100 oC for 20 min, a slight decrease in abso rption intensity was observed for both oligothiophenes. From the absorption onsets in the solid state, the opt ical energy gaps are estimated to be 1.67 eV for the DAD molecule and 1.76 eV for the ADA molecule. Both I-1 and I-2 are fluorescent, emitting with maxima at 780 and 749 nm in chloroform, respectively. Figure 6-5. UV-vis spectra of I-1 and I-2 in solution and in thin-film The redox behavior of the oligomers, shown in Figure 6-6, was investigated by cyclic voltammetry (CV) in dichloromethane (DCM) using tetrabutylammonium hexafluorophosphate
195 (TBAPF6, 0.1M) as the supporting elect rolyte. Both oligothiophenes present two consecutive quasi-reversible reduction peaks with E1/2 red of -1.31 and -1.71 V for I-1 and of -1.35 and -1.74 V for I-2 Two oxidation waves appear in the CV of I-1 with E1/2 ox of 0.47 and 0.58 V; while only one oxidation wave was observed for I-2 with E1/2 ox of 0.52 V. This is in accordance with the more electron rich (easier to oxidize) DAD molecule than the ADA molecule. The LUMO and HOMO energies of I-1 and I-2 estimated by CV, are -3.9 and -5.5 eV, and -3.8 and -5.5 eV, respectively. Figure 6-6. Cyclic voltammetry of I-1 (left) and I-2 (right) meas ured in a 0.1 M solution of TBAPF6/DCM (scan rate 25 mV/s) vs Fc/Fc+. In order to gain more insight into the ener gy levels in solid state, differential pulse voltammetry (DPV) was perfor med on drop-cast films of I-1 and I-2 on Pt button electrodes. From the onset of oxidation by DPV, the HOM O level of 5.6 eV is estimated for both I-1 and I2 Unfortunately, the first reducti on wave was not observed for bot h molecules. Therefore, the LUMO levels were calculated using the values obtained from the solid state optical band gap, giving 3.9 and 3.8 eV for I-1 and I-2. The solid state results are very close to the ones in solution. The appropriate energy levels and gap, togeth er with the high absorption coefficients, suggest the new isoindigo-based oligothiophene s should be effective electron donors in BHJ solar cells. The energy levels of I-1 and I-2 are very similar with those of DPP-1 for which HOMO and LUMO levels are -5.5 and -3.9 eV, re spectively. This observation suggests isoindigo
196 and diketopyrrolopyrrole as electron-accepting build ing blocks are very close regarding their ability to stabilize added electrons given that the LUMO levels ar e usually affected and governed by acceptors in donor-accepto r type materials. Table 6-1. Solid state optical and electrochemi cal properties and calc ulated energy levels. 6.2.3 Molecular Bulk-Heterojunction Solar Cells Molecular BHJ solar cells were fabricated by spin-coating I-1 / I-2 :PC61BM blends from chlorobenzene onto a clean ITO/PEDOT:PSS bo ttom electrode on a glass substrate. Optimization experiments explori ng different blend ratios, annea ling temperatures, and solution concentrations resulted in the highest power conversion efficiencies (PCEs) for I-1 / I-2 :PC61BM blend ratios of 50:50 ( I-1 ) and 60:40 ( I-2 ), an annealing temperature of 100 oC, and a total solution concentration of 18 mg/mL. The BHJ cells made from I-1 performed significantly better than devices made from I-2 The J-V characteristics for annealed devices are shown in Figure 6-7. After annealing at 100oC, solar cells made from I-1 showed a PCE of up to 1.76%, with a Voc of 0.74 V, Jsc of 6.3 mA/cm2, and fill factor of 0.38. Annealed photovoltaic devices (100 oC) of I-2 had PCEs of up to 0.55%, with a Voc of 0.66 V, Jsc of 2.4 mA/cm2, and fill factor of 0.36. Preannealed devices were less efficient with PCEs of 0.72 and 0.11% for I-1 and I-2 respectively, as shown in Table 6-8. It is noted that the fill factors here are about ~0.38, typical for solution-processed small molecule solar cells,71 which is in sharp contrast with the high fill factors obtained for DPP-1 This observation indirectly proves that amphiphilic molecular design is indeed accountable for the
197 high fill factors in DPP-1 Future studies will work towards applying amphiphilic molecular design to isoindigo based oligot hiophenes in order to achieve higher performance photovoltaic devices. Figure 6-7. J-V characteristics of I-1/I-2:PC61BM solar cells under 100 mW/cm2 white light illumination annealed at 100 oC for 20 min. Table 6-2. Performance of I-1/I-2:PC61BM solar cells before and after annealing Considering the similar optical and electronic properties of I-1 and I-2 it is interesting to look into their film morphologies in the device, which may expl ain their different photovoltaic performances. As with many examples found in the organic photovolta ic literature, device performance is highly dependent on thin-film mo rphologies in both small molecule and polymer based solar cells.81,318 For this purpose, AFM was used to examine the thin film morphologies of the optimized devices based on I-1 and I-2 The AFM height images of the blend films before
198 and after annealing are presente d in Figure 6-8. Before anneali ng, both films look very smooth. The desired phase separation is not observed, whic h in part explains the low device performance for both I-1 and I-2 Upon annealing, however, the surface of both films become rougher and phase separation appears with more ordered crystalline domains found in the case of I-1 /PC61BM blend film. Apparently, the morphology changes lead to the enhancement in performance of both I-1 and I-2 based devices. However, it is difficult to draw a conclusion on which morphology is preferred between the crystal line domains (Figure 6-9b) or the amphorous domains (Figure 69d), since several fold incr eases in PCEs have been found in both cases. Figure 6-8. AFM height images of I-1:PCBM (50:50) spin -coated from chlorobenzene a) as cast, and b) annealed at 100 oC for 20 min; I-2:PCBM (60:40) spin-coated from chlorobenzene c) as cast, and d) annealed at 100 oC for 20 min. All images are 1 x 1 um with 5 nm height scales. RMS surface roughness values of the AFM images are 0.15 nm, 0. 98 nm 0.14 nm and 0.95 nm from left to right. We note that polydimethylsiloxane (PDMS) ha s been found in the syringe we used for material processing and discovered to have a profound effect on device morphology. Since this so called macromolecular effect is beyond of the sc ope of this dissertati on, it is not discussed here although more details will be revealed in later studies. In summary, we have introduced isoindigo as an acceptor into the field of electroactive materials, presented the Synthesis and characte rization of isoindigo-base d oligothiophenes, and fabricated molecular BHJ solar cells using thes e oligothiophenes as electron donors for the first time. These devices exhibit some of the best power conversion efficiencies for solution
199 processed small molecules. We are currently wo rking to fine tune th e optical and electronic properties of isoindigo-based donor-acceptor ma terials through molecular design, to better understand the effect of chemical structur e and film morphology on photovoltaic response, which will be discussed in more details in the next section. 6.3 Molecular Engineering in Isoindigo-based Oligomers In the previous section, we reported two isoindigo-base d oligothiophenes and their utilization as electron donor materials in molecula r BHJ solar cells. In order to gain more insight into the structure-property relati onships of isoindigo-based mate rials in solid-state electronic devices (OFETs and OPVs), we take a molecu lar engineering approach to access a set of oligomers that highly resemble each other struct urally. While the proposed structures share the same isoindigo acceptor core, the difference only lies in the donor portion where 1methylbenzopyrrole, benzofuran and benzothio phene are used as electron donors to give IsoI-N IsoI-O and IsoI-S respectively, as shown in Figure 6-9. This set of oligomers is expected to provide an excellent platform to look into how rational design and careful structural modification can be utilized to tune material properties. Figure 6-9. Chemical structures of IsoI-N, IsoI-O and IsoI-S. 6.3.1 Synthesis of Isoindigo-Based Oligomers The Synthesis of IsoI-N IsoI-O and IsoI-S was carried out via Suzuki coupling between 6-2 and commercially available boronic ester or boronic acid functiona lized donor components, namely 1-methyl-2-(4,4,5,5-tetramethyl-1,3,2-di oxaborolan-2-yl)-1H-indole, benzofuran-2-
200 ylboronic acid and benzo[b]thi ophen-2-ylboronic acid, yielding IsoI-N IsoI-O and IsoI-S in excellent yields (84 92%), as shown in Figure 610. It is worth mentioning that the purification process for these compounds is rather simple and can be achieved by a precipitation method (see the synthetic details). Figure 6-10. Synthesis of Iso I-N, IsoI-O and IsoI-S. a) Pd2(dba)3, P(o-tyl)3 Et4NOH, toluene, 85 oC, 88%, 92%, and 84% for IsoI-N, Is oI-O and IsoI-S respectively. 6.3.2 Structural Characterization, Op tical and Electrochemical Studies The structures of IsoI-N IsoI-O and IsoI-S have been confirmed by 1Hand 13C-NMR, HRMS, and elemental analysis. A brief analysis is given as follows to reflect the influence of structural differences on chemical shifts in 1H-NMR. Figure 6-11a shows the full 1H-NMR spectra of IsoI-N IsoI-O and IsoI-S The chemical shifts for the protons on the ethylhexyl chains are very similar in all three oligomers wh ile the signal from the N-methyl group is located at ~3.84 ppm in the 1H-NMR spectrum of IsoI-N The peaks in the aromatic region present different splitting patterns and chemical shifts for IsoI-N IsoI-O and IsoI-S and these regions are shown in Figure 6-11b, c and d, respectivel y. For example, a cl ose look at Figure 6-10b leads to assignment of the doublet (it looks like a singl et because of the small coupling constant) at 6.71 ppm to Ha with a very small coupling constant ~0.6 Hz, a consequence of long range W (zigzag) coupling with Hc; while Hb is shifted downfield to 9.29 ppm because of the ortho bonded electron withdrawing group and possibl e hydrogen bonding interactions with the carbonyl group. The signal from Hd also is a doublet with a sma ll coupling constant of 1.5 Hz, resulting from the interaction with Hc as well. The further assignments of He, Hf, Hg and Hh are
201 not very reliable based on 1-D 1H-NMR in the case of IsoI-N even though we can tentatively assign peaks at 7.69 and 7.41 ppm to He and Hf. Figure 6-11. 1H-NMR spectra of IsoI-N, IsoI-O and Iso I-S including a) the full spectra and the aromatic regions of b) IsoI-N, c) IsoI-O, and IsoI-S. A similar analysis can be extended to IsoI-O and IsoI-S It is found that Ha has shifted downfield to 7.05 ppm in IsoI-O and to 6.89 ppm in IsoI-S relative to the 6.71 ppm in IsoI-N This observation is consistent with the decrea sing electron-donating ability on moving from 1methylbenzopyrrole to benzofuran and benzothiophene. It is also interesting to look at Hd which has shiftted to 7.07 ppm in IsoI-O and 7.53 ppm in IsoI-S from 6.95 ppm in IsoI-N This trend follows exactly the same chemical shift change in the parent compounds 1-methyl-benzopyrrole, benzofuran and benzothiophene, where Hd is in a region of greater ar omatic character. It is particularly evident when a comparison is made between IsoI-O and IsoI-S in which He, Hf, Hg and Hh have a vastly different chemical shift split patterns. In the case of IsoI-S the peaks are
202 much more downshifted and have larger coupling constants due to its higher aromatic character of benzothiophene. The chemical shifts from He and Hf are much closer to those from Hg and Hh in IsoI-O Through the simple NMR analysis, we do fi nd that electronic prope rties displayed in the format of chemical shits here are affect ed by the delicate stru ctural modifications. The solution and solid-state UV-vis spectra of IsoI -N IsoI-O and IsoI-S are exhibited in Figure 6-12. In solution, the spectra look alike with minor difference. IsoI-N absorbs past 669 nm with three absorpti on maxima at ~ 555, 457, and 299 nm in chloroform. A molar absorptivity of 20,400 M-1 cm-1 at 555 nm is obtained. Similarly, IsoI-O absorbs with an onset at ~675 nm and three absorption maxima at ~ 563, 461, and 318 nm. Its molar absorptivity is about 27,300 M-1 cm-1 at 563 nm; while IsoI-S absorbs past 662 nm with th ree absorption maxima at ~ 552, 452, and 319 nm. The molar absorptivity of 26,300 M-1 cm-1 at 552 nm is found for IsoI-S As mentioned earlier, benzofuran is a stronger el ectron donor than benzothiophene. Our observation in solution absorbance for IsoI-O and IsoI-S is consistent with their electron donating ability that is the origin for a longe r wavelength absorption in IsoI-O Interestingly, IsoI-N does not strictly agree with this argument and its absorpti on onset and maxima fall in the middle of IsoI-O and IsoI-S even though benzopyrrole is expected to be the strongest electron donor among the three. This is likely due to a slight distortion between the donor and acceptor caused by the methyl group on the nitrogen atom that weakens th e donor-acceptor interaction an d hence gives IsoI-N a slight hypsochromic shift in absorption. An astonishing contrast for IsoI-N IsoI-O and IsoI-S is observed in their solid-state absorption spectra, as shown in Figure 612b. Most significantly, the absorption of IsoI-S broadens and extends to 719 nm and a nearly 57 nm red shift is observed relative to solution absorbance. In addition, two new absorption maxi ma appear at 666 and 617 nm. In contrast,
203 there is ONLYa 6 nm red shift observed for IsoI-N IsoI-O somehow behaves normally like the previously studied molecules I-1 and a 23 nm red shift is found. It is also noticed that IsoI-O and IsoI-S have a relatively strong vibronic peaks acr oss the spectra, indicating a strong coupling between electronic and vibra tional states. An elegant analysis can be found elsewhere.319 Considering the similarity in their solution absorbance, the difference in solid-state absorption can be only explained by their different crystalli nity and solid state orde ring. More specifically, it is logical to speculate th at the crystallinity in IsoI-S is much higher than in IsoI-N and IsoI-O In other words, the aggregation and interchain interactions are much more severe in IsoI-S thin films, which are supported by the much higher melting point for IsoI-S as will be discussed later. To verify this speculation, however, single crystal X-ray diffractio n experiments should be performed. Figure 6-12. UV-vis spectra of IsoI-N, IsoI-O and Is oI-S a) in chlorobenzen e, and b) in thin-film. The absorption spectra of IsoI-N IsoI-O and IsoI-S present similaroptical responses upon thermal annealing, as shown in Figure 6-13. In the case of IsoI-N a slight decrease in absorption intensity for the high-energy band is observed after annealing. This has also been observed in the DPP-based molecule in Nguyens study.80 Changes in optical response in IsoI-O are relatively minor upon annealing. It first has a slight drop in absorption intens ity from room temperature to 100 oC followed by a significant increas e in optical intensity from 100 oC to 150 oC. At this
204 point, this phenomenon is not fully understood. Likely, there is ener gy barrier to overcome before the molecules can be oriented in an orientation for favorable crystal packing. An interesting observation is found in IsoI-S where the optical change is almost absent upon annealing up to 150 oC. We speculate that the lack of thermally-induced optical change is resulted from high crystallinity for IsoI-S upon spin-casting. Figure 6-13. UV-vis spectra of as cast and thermally annealed a) IsoI-N, b) IsoI-O and c) IsoI-S. Cyclic voltammetry (CV) and differential pulse voltammetry (DPV) were used to investigate the redox properties of IsoI-N IsoI-O and IsoI-S with the goal of determining how oxidation and reduction potentials, together with HOMO and LUMO energy levels, vary from IsoI-N to IsoI-O and to IsoI-S The results are summarized in Table 6-3. The experimental setup consists of a platinum button electrode, a pla tinum counter-electrode, a Ag/Ag+ reference electrode and a electrochemical cell with a thr ee-hole cap. The experiments were performed in 0.1 M TBAPF6/DCM under inert atmosphere in an argonfilled dry box. All estimated potentials were subsequently calibrated versus Fc/Fc+. CV and DPV analysis of IsoI-N IsoI-O and IsoI-S were performed, as illustrated in Figure 6-14. IsoI-N shows an onset of reduction at -1.04 V with E1/2 of -1.13 and -1.56 V from cyclic voltammetry; while IsoI-O exhibits a reduction onset at -1.0 V with E1/2 of -1.08 and -1.48 V, and IsoI-S presents an onset of reducti on at -1.0 V as well, with E1/2 of -1.10 and -1.49 V. These values from CV are consistent with the tre nds in the results obtained from DPV. All three
205 oligomers have stable and reversible scans for both reduction processes as observed in CV and DPV. Similar reduction onsets, together with E1/2, suggest the LUMO levels are localized on the isoindigo core and are not str ongly dependent on the donors used in these oligomers. Figure 6-14. Cyclic voltammograms (scan rate = 50 mV/s) and differential pulse voltammetry (step size of 2 mV and step time of 0.1 sec onds) of IsoI-N, IsoI-O and IsoI-S in 0.1 M TBAPF6-CH2Cl2 electrolyte solution. a) CV of IsoI-N; b) DPV of IsoI-N; c) CV of IsoI-O; d) DPV of IsoI-O; e) CV of IsoI-S; f) DPV of IsoI-S. To determine the exact oxidation onset from CV is difficult for IsoI-N since the CV of IsoI-N appeared highly irre versible upon applying an anodic pot ential and the current intensity increased per scan. It was later discovered that this oligomer actually underwent electropolymerization, as s hown in Figure 6-15a. The electropolymerized poly( IsoI-N ) has a much broader and red-shifted absorption on ITO, as displayed in Figure 6-15b. Nevertheless, an oxidation onset was estimated to be 0.59 V with an E1/2 of 0.71 from the first CV scan. The oxidation onset of 0.32 V from the DPV wa s tentatively assigned. The CV of IsoI-O yielded an oxidation onset of 0.87 V with E1/2 of 1.03 and 1.04 V. In comparison, the onset of oxidation for
IsoI-S m reversibl mechani s yielding oxidatio n b enzothi o Figure 6 Table 63 In a close fro m m oves 0.91 V e CV scans, s m is not un d an oxidatio n n process is l o phene to b e 15. a) Repe t TBAPF6/ D Poly(IsoI 3 Electroch e IsoI-S (b y a ddition, it i m both CV a V slighter hi g a shoulder a d erstood at t n onset of 0. 8 l argely dete r e nzofuran t o t itive scan e D CM ; b) U N) on ITO e mically de t y CV and b y i s also inter e a nd DPV an g her than Is o a ppeared in D t his point. I n 8 2 V. Thro u r mined by t h o benzopyrr o lectropoly m U V-Vis spec t t ermined H O y DPV). e sting to not e alyses. The 206 o I O so do e D PV upon o n contrast, I s u gh the anal y h e donors. O o le, the olig o m erization of t ra of IsoI-N O MO and L U e that the en e energy gaps e s the E1/2. A o xidation an d s oI-S has a p y sis of CV a n O n changing t o mers beca m 5 mM IsoIN in CH2Cl2 a U MO energ y e rgy levels o from CV ( D A lthough Iso I d on neutral i p rototypical D n d DPV, it i s t he donor fr o m e easier to o N on ITO i n a nd electrop o y levels of I s o f IsoI-O a n D PV) are 1. 8 I -O had i zation. Th e D PV scan, s clear that t o m o xidize. n 0.1 M o lymerized s oI-N, IsoIO n d IsoI-S ar e 8 7 (1.86) an d e t he O and e very d
207 1.91 (1.87) eV for IsoI-O and IsoI-S respectively. There values are in good agreement with their optical energy gaps, which are 1.83 and 1.87 eV obtained from the onset of absorption in CH2Cl2 solution. However, there is a large diffe rence between the electrochemical gaps (1.63 and 1.34 eV from CV and DPV) and the optical gap (1.84 eV). The large discrepancy between these gaps is not completely unexpected, as the values measured from CV and DPV may reflect more character of poly( IsoI-N ) than that of the oligomer IsoI-N To support this assertion, it is noted that the optical bandgap of poly( IsoI-N ) is about 1.39 eV obtained from electropolymerized films on ITO. 6.3.3 Thermal Analysis Thermogravimetric analysis was carried out for all three oligomers in order to provide necessary information for differential scanning calorimetry (DSC) studies. As illustrated in Figure 6-16, all these molecules are thermally stable up to more than 300 oC under nitrogen atmosphere. Interestingly, IsoI-O and IsoI-S have almost identical d ecomposition temperatures. Figure 6-16. Thermogravimetric analysis of a) IsoI-N, b) IsoI-O and c) IsoI-S. DSC analysis of IsoI-N IsoI-O and IsoI-S was performed in order to provide phase transition information. Figure 6-17 show s the DSC thermograms. Other than IsoI-O in which a normal crystalline-isotropic ther mal transition is present with a melting temperature at ~225 oC, IsoI-N and IsoI-S both exhibit some unexpected thermal behaviors to a cert ain degree. Upon
208 heating at 10 oC min-1, IsoI-N underwent a very broad endothermic transition around ~192 oC (starting at 185 oC) with an enthalpy change of 38.7 kJ mol-1. Surprisingly, the recrystallization process failed to appear upon cooling. More strikingly, the endothermic transition also disappeared in the second heating sc an. As confirmed by the TGA analysis, IsoI-N is thermally stable up to 300 oC. It is unlikely that IsoI-N experienced the decomposition event during the course of DSC experiment. We suspect that a th ermally-induced Diel-Alder reaction occurs. The behavior of IsoI-S came unexpectedly as a consequence of its polycrystalline-like thermal behavior as shown in Figure 6-17c In retrospect, this actually explains why black powder is obtained for IsoI-S This also supports the solid stat e absorption where two new absorption maxima and an extended red-shift are observed. All these observations can be explained by the existence of two or multi-form crystalline mesophases, which have different crystal orientations and thus lead to their preferred absorbance in solid state. Figure 6-17. DSC thermograms of a) Iso I-N, b) IsoI-O and c) IsoI-S (kJ mol-1, unit in parendissertation after transition temperature) 6.3.4 Field-Effect Transistors and Molecu lar Bulk-Heterojunction Solar Cells The results of this OFET part are contribut ed by Dr. Shree Prakash Tiwari at Georgia Tech Bottom-gate top-contact OFETs were fabricat ed on heavily doped n-type silicon substrate (which also serves the gate electrode ) with 200 nm thick thermally grown SiO2 as the gate dielectric, in top a contact c onfiguration. Ti/Au (10 nm/100 nm) metallization on the backside of
209 the substrate was done to enhance the gate electri cal contact. Firstly, the substrates were cleaned by O2 plasma for three minutes, to insure the proper film formation by changing the surface property of SiO2 towards hydrophilic. The capacitance of the SiO2 layer was ~16.2 nF/cm2. A thin layer of organic semiconductor was formed on the substrates by spin coating with a solution 10 mg/mL in chlorobenzene, or 15 mg/mL in ch loroform. Then, Au (approximately 75 nm-thick film) was deposited through a shadow mask to act as top source/drai n electrode. The device fabrication was done under inert (N2) atmosphere. The samples were transferred in a vacuum tight vessel without being expos ed to atmosphere into a N2-filled glovebox (O2, H2O ~ 0.1 ppm) for electrical testing. An annealing step is done at 130C for 30 minutes before device characterization. Table 6-4 summarizes the FET response for all IsoI-N IsoI-O and IsoI-S devices processed from either chloroform or chlorobenzene and annealed at 130 oC. Representative transfer/output plots for IsoI-S OFET devices processed from ch loroform solution are shown in Figure 6-18. The negative gate and s ource-drain voltages demonstrate that IsoI-S is a p-channel material. This is consistent with previous reports on DPP-1 and the aforementioned HOMO energy values, which are accessible for hole injection from Au contacts (5.1 eV). From the output plot of IsoI-S OFET device, hole mobility up to 4.1 x 10-4 cm2 V-1 s-1 and Ion/Ioff of 103 is obtained under inert atmosphere. In sharp contrast, the mo bility values of IsoI-N and IsoI-O are one order magnitude lower around 3.8 x 10-5 cm2 V-1 s-1. The big difference between IsoIO and IsoI-S is not anticipated, as they are both crys talline materials, and present very close electrochemically and optically determined ener gy levels. We speculate that the difference may arise with their different crystal packing and interactions. A close look into the thin-film morphologies would be beneficial for understanding the difference, for instance, through AFM,
210 TEM and GISAXS ( Grazing -Incidence Small Angle X-ray Scattering) stud ies. In addition, we are also working to obtain single cr ystal structures of these molecu les. A present effort is right now to replace the branched ethylhexyl alky chains with linear short linear alky chains in order to obtain single crystals. Figure 6-18. a) Output and b) tr ansfer characteristics of a repr esentative IsoI-S field-effect transistor device processed from chloroform solution. Table 6-4. Bottom-Gate Top-Contact OFET Char acteristics of IsoI-N, IsoI-O, and IsoI-S. The preliminary solar cell results are ki ndly provided by Kenneth Graham at UF. Molecular BHJ solar cells were fabricated by spin-coating IsoI-N/IsoI-O/IsoI-S :PC61BM blends from chloroform onto a clean ITO/PEDOT:PSS botto m electrode on a glass substrate. A blend ratio of 60:40 was used for all three sets. The devices are annealed at a temperature of 100 oC for 20 min. Performance of IsoI-N/IsoI-O / IsoI-S :PC61BM solar cells is summarized in Table 6-5.
211 The BHJ cells made from IsoI-O and IsoI-S performed significantly better than devices made from IsoI-N The J-V characteristics for annealed de vices are shown in Figure 6-19. The pair of IsoI-N /PC61BM barely showed any photovoltaic re sponse. Solar cells made from IsoI-O showed a PCE of up to 0.46%, with a Voc of 0.71 V, Jsc of 2.0 mA/cm2, and a fill factor of 0.33. Annealed photovoltaic devices of IsoI-S had PCEs of up to 0.36%, with a Voc of 0.54 V, Jsc of 2.0 mA/cm2, and a fill factor of 0.34. We speculate th at the low performance of these oligomers is largely caused by inefficient char ge transfer between the donors and PC61BM, resulting from the extremely low LUMO levels (~-4.1 eV) inhe rent to the oligomers. A possible way to increase solar cell efficiency is to introduce acce ptors with even lower LUMO levels, such as PC84BM (its electron affinity is about 0.35 eV higher than PC61BM),320 which will provide sufficient energy offset (~0.4 eV) between the oligomers and PC84BM and thus leads to the sufficient driving force for charge separation. Ty pically, it will cause sm all open circuit voltages (Voc) with the deepening acceptor LUMO level, since Voc is approximately determined by the offset between the LUMO level of the acceptor and the HOMO level of the donor. In this case, we would not have the concern, consider ing the low HOMO levels (~5.8-6.0 eV) for IsoI-N IsoI-O and IsoI-S A decent Voc is still feasible. Ba sed on Janssen theory,123 Voc of 0.8-1.0 V is attainable. In addition, we once again noted that the fill factors here are about ~0.35, which is in sharp contrast with the high f ill factors (~0.58) obtained for DPP-1 As we pointed out in the earlier section, future studies will work to wards applying amphiphilic molecular design to isoindigo based oligothiophenes in order for ach ieving higher performanc e photovoltaic devices. In this section, molecular engineering of isoindigo-based oligomers provides us an opportunity to closely examine the structure-property relatio nship in molecular organic semiconducting materials. From the optical, electrochemical and th ermal analysis of IsoI-N
IsoI-O a n modific a and IsoI still fruit f is a very Table 65 Figure 6 In t isoindig o relations h organic s solubiliz i groups t o more im m nd IsoI-S w a tions. OFE T S can not b e f ul. It conv e delicate pr o 5 Performa n 19. J-V cha r 6.4 Is o t he previou s o b ased disc r h ip in isoin d s olvents, co n i ng groups.2 o conjugate d m iscible se g w e indeed o b T and OPV s e considere d e ys a very i m o cess and s m n ce of IsoIN r acteristics o o indi g o-Ba s s several sec t r ete oligom e d igob ased p n jugated pol y ,145,321 Here, d polymers. g ments alon g b serve big di f tu d ies supp o d as high pe r m portant me s m all alternati o N /IsoI-O/Iso I o f IsoI-O/Is o s ed Pol y me r t ions we dis c e rs. In this s olymers. In y mers typic a we introdu c Block solu b g a hydrocar b 212 f ferences, w o rt the same r formance o r s sage that d e o ns can lead I -S:PC61B M o I-S:PC61B M r s with Dib c ussed the s t ection, we i n order for so l a lly contain c e the conce p b ilizing grou p b on chain. W w hich are in d argument. E r ganic electr e signing ma t to big chan g M solar cells. M solar cells b lock Solub i t ructure-pro p n vestigate t h l ution proce linear or br a p t of block ( p s, as the n a W e expect t h d uced by the E ven though onic materi a t erials for or g g es. i lizin g Gro u p erty relati o h e structurep ssing to be p a nched alkyl ( or biphasic ) a me implies, h e driving f o small struct u IsoI-N Iso I a ls, this stud y g anic electr o u ps o nships in p roperty p ossible fro m chains as ) solubilizin g contain tw o o rce for pha s u ral I -O y is o nics m g o or s e
213 separation among solubilizing group s will enhance the interchain interactions. This hypodissertation originates from the surfancant e ffect and rich phase separation behavior in block copolymers (e.g. PEOb -PSb -PMMA),322-324 where high ordering (both orientational and positional ordering) can be achieved via the process of self-assembly. To test this hypodissertation, a set of three isoindigo-based donor-acceptor type conjugated polymers have been conceived and prepared. Thei r chemical structures are shown in Figure 6-20. All three polymers are composed of alternating isoindi go acceptor and thio phene donor units in the -conjugated backbone. The difference among these polymers lies in their solubilizing groups. Specifically, PIsoIAm-1 has a biphasic solubilizing gr oup where the hydrophobic end is attached to the nitrogen atom on the isoindigo unit; while PIsoIAm-2 has the hydrophiphilic end attached to the nitrogen atom. PIsoIAm-3 only contains a branched alkyl chain (hexyldecyl chain), serving as a control polymer for comparison. Figure 6-20. Polymer structures of PIsoIAM-1, PIsoIAM-2 and PIsoIAM-3. 6.4.1 Synthesis of Isoindigo-Based Polymers with Biphasic Solubilizing Groups The synthetic scheme is outlined in Figure 6-21. N-alkylation of isoindigo 6-1 with 7(bromomethyl)hexadecane, 2-(2-(2-methoxyet hoxy)ethoxy)ethyl 6-bromohexanoate, and 2-(2-
214 (2-(tosyloxy)ethoxy)ethoxy) ethyl nonanoate gave 6-5 6-9 and 6-10 in 70%, 35% and 49% yields, respectively. The yields for 6-9 and 6-10 are significantly lower than their derivatives with alkyl chains, likely due to side reactions caused by the pres ence of water as we described earlier. With 6-5 6-9 and 6-10 in hand, PIsoIAM-1 PIsoIAM-2 and PIsoIAM-3 can be easily obtained by Stille coupling polymerization be tween their corresponding isoindigo monomer and 2,5-bis(trimethylstannyl)thiophene. Figure 6-21. Synthesis of IsoIAM building blocks and PIsoIAM. a) K2CO3, DMF, 100 oC; b) Pd2(dba)3, P(o-tyl)3, toluene, 85 oC. 6.4.2 Structural Characterization, Op tical and Electrochemical Studies The structures of these polymers were characterized by solution 1H and 13C-NMR, and elemental analysis. Solution 1H NMR gives very broad peaks in the aromatic region for these
215 polymers. Even in the aliphatic region the NMR spectrum is quite br oad, and are thus not insightful for the structural elucidation. Elemental an alysis of C, H, and N is consistent with the polymer structures. The experiment al data are consistent with theoretical calculations. The errors for C, H and N are within 0.4% for all thr ee polymers. The GPC results are exhibited in Figure 6-22. PIsoIAM-1 and PIsoIAM-2 in THF likely contain a certain degree of aggregation as sugge sted by their GPC traces. High temperature GPC may give more insight.Therefore it is difficult to estimate th eir molecular weights with high accuracy. It is believed that PIsoIAM-1 and PIsoIAM-2 have number average molecular weights higher than 25 kDa. PIsoIAM-3 shows a well-defined GPC trace, indicating the absence of severe aggregation. The number-average molecular weight estimated for PIsoIAM-3 is about 20.8 kDa and a PDI of 1.6. Figure 6-22. GPC traces of PIsoIAM-1, PIsoIAM-2 and PIsoIAM-3 As stated in our hypodisserta tion, polymers with block so lubilizing groups would have stronger interactions both in solu tion and in solid state. A convenient solid-state UV-Vis measurement can be carried out to verify this argument, since strong and extensive interactions will typically lead to a red shift in absorbance. Figure 6-23 shows UV-vis spectra of PIsoIAM-1 PIsoIAM-2 and PIsoIAM-3 in thin film cast from chlorobenzene solutions. The
216 absorption onsets for PIsoIAM-1 PIsoIAM-2 and PIsoIAM-3 are 785, 790 and 759 nm, respectively. A 30 nm red shif t is indeed observed between PIsoIAM-2 and PIsoIAM-3 This observation supports our aforementioned argument. In terestingly, a 5 nm redshift is also noticed between PIsoIAM-1 and PIsoIAM-2 This small change can be explained by two possibilities. First, the flexible oligoether chain is trapped be tween two incompatible segments, thus leading to higher degree of ordering in PIsoIAM-2 On the opposite, the extremely flexible oligoether group has higher degree of freedom as a terminal chain in PIsoIAM-1 The likelihood for the crystallization of terminal oli goether is low under ambient conditions. That is to say the first possibility is resulted from a higher degree of ordering in PIsoIAM-2 than in PIsoIAM-1 The second possibility is due to a slightly lower LUMO level for PIsoIAM-2 because of the inductive effect from electronegative at om oxygen via the ethylene bridge connected to the nitrogen atom on isoindigo core. Figure 6-23. UV-vis spectra of PIsoIAM-1, PIsoIAM-2 and PIsoIAM-3 in thin film. Temperature-dependent UV-Vis measurements were carried out in order to provide additional information for their differences in absorbance, as shown in Figure 6-24. Upon heating from 5 to 95 oC in dichlorobenzene, a blue shift is observed for all three polymers and the absorption maxima at the longe r wavelength of 709 and 711 nm for PIsoIAM-1 and PIsoIAM-2
217 almost vanish. However, a noticeable abso rption shoulder remains in both cases of PIsoIAM-1 and PIsoIAM-2 indicating the polymers are stil l not full solvated even at 95 oC. We also notice that PIsoIAM-1 and PIsoIAM-2 are almost superimposed each other at 95oC, which suggests the local microstructures are responsible for the 5 nm di fference in solid state absorption. We are also surprised to find out that PIsoIAM-2 is still much braoder than PIsoIAM-3 This was not anticipated. We had thought hoped that the gap would become smaller when the polymers are solvated, since they have the same conjugated backbone. Figure 6-24. Temperature-dependent UV-Vis spectra of a) PIsoIAM-1; b) PIsoIAM-2; c) PIsoIAM3; and d) PIsoIAM-1, PIsoIAM-2 and PIsoIAM-3 in dichlorobenzene at 95 oC. Cyclic voltammetry analysis was carrie d out to obtain redox properties of PIsoIAM-1 PIsoIAM-2 and PIsoIAM-3 by drop cast on platinum electrodes, as shown in Figure 6-25. From CV, the reduction onsets for PIsoIAM-1 PIsoIAM-2 and PIsoIAM-3 are -1.15, -1.14 and -1.26 V, respectively; while the oxidation onset s are 0.32, 0.20 and 0.59 V, respectively. The
218 electrochemical bandgaps are 1.47, 1.34 and 1.85 eV accordingly. Figure 6-26 shows the DPV of PIsoIAM-1 PIsoIAM-2 and PIsoIAM-3 From DPV, the reduction onset s are -1.14, -1.08 and -1.07 V while the onsets of oxidation are 0.30, 0.20 a nd 0.65 V, respectively. The electrochemical bandgaps obtained from DPV are 1.44, 1.28 and 1.72 eV, respectively. Figure 6-25. Cyclic voltammetry of a) PIsoIAM-1; b) PIsoIAM-2; c) PIsoIAM-3 measured in a 0.1M solution of TBAPF6/ACN vs Fc/Fc+. Figure 6-26. Differential pulse voltammetry of a) PIsoIAM-1; b) PIsoIAM-2; c) PIsoIAM-3 measured in a 0.1M solution of TBAPF6/ACN vs Fc/Fc+. The data obtained from CV and DPV are consistent in the case of PIsoIAM-1 and PIsoIAM2 There is a relatively a large discrepancy between the reduction onsets obtained from CV and DPV in the case of PIsoIAM-3 This is likely due to the sensiti vity issue in CV, where the first reduction peak, clearly seen by DPV, is not obs erved. This explanation is supported by the presence of two oxidation peaks in the neutralization process. The redox properties obtained from DPV are believed to be more reliable wh en making a comparison between three polymers.
219 It is surprising to see that the onsets of oxidation are affect ed to such a great extent by introducing different passive solubilizing groups. Because all three polymers share the same donor they are expected to exhibit very simila r oxidation onsets, given that the oxidation onsets are usually governed by the donor used. It is also interesting that the thre e polymers have similar reduction onsets, which are much less affected by variation in the solubilizing group. These results, taken together, suggest the HOMO is more delocalized and thus more easily affected by the local environment; the LUMO is more localized and thus is not easily perturbed by the local morphology. 2D-WAXS study is contributed by Dr. Wo jtek Pisula at MPI and Yuying Wei at UF In order to explore the local microstructures of PIsoIAM-1 PIsoIAM-2 and PIsoIAM-3 2D-WAXS was performed, as shown in Figure 6-27. The sa mples were prepared by fiber extrusion, a method that induces macroscopic orientation, and the fibers mounted vertically towards the 2D detector. All the polymers assemble into lamellar structures. In the case of PIsoIAM-1 PIsoIAM2 two characteristic spacings are indentified. Th e equatorial small-ange l reflections (b) give chain-to-chain dist ances of 2.35 nm for PIsoIAM-1 and 2.47 nm for PIsoIAM-2 A -stacking distance of 0.36 nm is obtained for both PIsoIAM-1 and PIsoIAM-2 by measuring the wide-angle reflections (a). The observed -stacking distances of 0.36 nm are some of the closest intermolecular interactions reported to date for a solution-processed conjugated polymer. 325 In contrast, PIsoIAM-3 demonstrates a chain-to-chain distan ce of 1.9 nm, due to the shorter chain length of the solubilizing group, and a -stacking distance of 0.40 nm. Additionally, one meridional reflection pair (c) is observed with a distance of ~1.36 nm, which is attributed to interactions between units along the backbone. To our surprise, th e overall crystallinity seems higher for PIsoIAM-3 supported by the sharper reflectio ns of b and c, even though PIsoIAM-
220 1 and PIsoIAM-2 have much shorter -stacking distances and stronge r chain-chain interactions. Thus, charge mobility will be greatly affected by the orientations of the polymer lamella by adopting a path parallel or normal to the substrate.321 Figure 6-27. 2D wide-angle X-ray scattering of PIsoIAM-1, PIsoIAM-2 and PIsoIAM-3. 6.4.3 Current-Voltage Measurements From the 2D-WAXS study, we demonstrate that the -stacking distances can indeed be enforced by introducing biphasi c block solubilizing groups. A -stacking distance as low as 0.36 nm can be obtained. The question is now wh ether the small pi-stacking distance can be transferred to high charge mobility in photovolta ic devices. Here we use the space-chargelimited-current (SCLC) model to determine hole mobility from simple current density-voltage ( J-V curve ) measurements using the Mott-Gurney equa tion, which was described in Chapter 2. The architecture of the hole-only devices studied is ITO/PEDOT:PSS/Polymer/Au. The structures of the electron-only devices were Glass/Al/PEDOT-PSS/Polymer/Al. The films for both types of devices were annealed at 150 oC for 30 min. Figure 6-28a shows the experimental dark current densities in the hole-only devices of PIsoIAM-1 PIsoIAM-2 and PIsoIAM-3 ; while Figure 6-28b exhibits the experimental dark cu rrent densities in the electron-only devices of these polymers. The applied voltage V is corrected for the built-in voltage Vbi that results from the difference in the work function of the electrodes.
221 As illustrated in Figure 6-28, cu rrent densities at given elect ric field inside the device increase from PIsoIAM-3 to PIsoIAM-2 to PIsoIAM-1 in both hole-only and electron-only devices. Table 6-6 summarizes the zero-field hole mobility in PIsoIAM -1 PIsoIAM-2 and PIsoIAM-3 derived from fitting J-V data to the trap-free single-carrier SCLC model. PIsoIAM -1 shows hole mobilities of 9.5 x 10-4 cm2V-1s-1 and electron mobilities of 9.1 x 10-5 cm2V-1s-1. Hole mobilities of 8 x 10-4 cm2V-1s-1 and electron mobilities of 6 x 10-5 cm2V-1s-1 are obtained for PIsoIAM -2 which are slightly lower than those of PIsoIAM -1 PIsoIAM -3 exhibits hole mobilities of 7.2 x 10-4 cm2V-1s-1 and electron mobilities of 2 x 10-5 cm2V-1s-1. Hole mobilities observed from PIsoIAM -3 are about one order of magnitude lo wer than its counterpart polymers with biphasic solubilizing groups Its electron mobilities are al so significantly lower than PIsoIAM -1 Figure 6-28. a) Experimental dark current densities for hole-onl y devices of PIsoIAM-1, PIsoIAM2 and PIsoIAM-3 as a function of the effective elec tric field; b) Experimental dark current densities for electron-only devices of PIsoIAM-1, PIsoIAM-2 and PIsoIAM-3 as a function of the effective electric field. Table 6-6. Zero-field hole mobility in PIsoIAM -1, PIsoIAM-2 and PIsoIAM-3, derived from fitting J-V data to the trap-free single-carrier SCLC model. These observations are consistent with our hypodissertation the charge mobility can be enhanced by introducing biphasic solubilizing groups. The enhanced charge mobilities of
222 PIsoIAM-1 and PIsoIAM-2 are resulted from the short -stacking distances of 0.36 nm, in comparison with the larger -stacking distance of 0.40 nm that PIsoIAM -3 exhibits. Additionally, chain-to-chain di stance also plays a role, ev en though less significant, in influencing charge mobility. This is evident through the comparison of PIsoIAM-1 and PIsoIAM2 They have the same -stacking distances, but different chain-to-chain distances. PIsoIAM-1 has a smaller chain-to-chain distance and thus has a larger charge mobility than PIsoIAM-2 Given the results obtained, we envision extending the biphasi c solubilizing group concept to other systems. 6.5 N-type Isoindigo-Based Conjugated Polymers The research in the field of organic solar cell s has been largely direct ed to the development of suitable p-type materials.83,259-260,284 Many high performance p-type polymers are becoming available after more than a decade of effort.2 In contrast, processable and stable n-type polymers remain limited.39,326-327 Few studies have been reported concerning all-polym er solar cells, 8990,106 mainly due to the limited availability of high electron affinity (EA) n-type polymers. Consequently, polymer-based photovoltaic devices have accordingly focused heavily on acceptor molecules such as methanofullerene phenyl-C61-butyric-acid-methyl-est er (PCBM), which has an electron affinity of 4.2-4.3 eV, and its derivatives.125,154,320,328-329 Access to solution processable high electron affinity -conjugated polymers remains a challenge, and one of the most successful strategies has been the cons truction of conjugated polymers based upon Nheterocyclic electron-deficien t aromatics (e.g. perylene diim ide or naphthalene bisimide).89,150151,330 However, development of imide-based ni trogen-containing polyheterocycles with high electron affinities is greatly hi ndered by their poor re activity. Extremely harsh conditions (for example refluxing under strongly ox idizing conditions) are used to prepare brominated and
223 nitrated monomer precursors.151 In most of cases, they are inco rporated into a polymer chain via Stille coupling with electron rich mono mers (e.g. thiophene and bithiophene).89,150-151,330-331 Soluble n-type conjugated polymers, othe r than imide-based nitrogen-containing polyheterocycles, are also reported and a few examples include indenofluorene-based and bisindenofluorene-based polymers.43,332 Unfortunately, Synthesis of these polymers typically involves multiple synthetic steps (n > 7-10) fr om readily available starting materials. The applicability of these monomer building blocks to obtain target polymers is thus adversely affected the lengthy synt heses involved. Because isoindigo-based donor-acceptor type mate rials have a fairly low electron affinity of ~-3.9 eV317 we were encouraged to explore n-type isoindigo-based conjugated polymers. We report herein the syntheses and electron-accepti ng properties of two isoindigo-based n-type conjugated polymers composed of solely electron -deficient repeat units. The proposed polymers are shown in Figure 6-29. Figure 6-29. Polymer structures of PIsoI-C16 and PIsoI-BT-C16. The Synthesis of PIsoI-C16 and PIsoI-BT-C16 is shown in Figure 6-30. The key compound 6-11 was successfully prepared by palladiu m-catalyzed Miyaura Borylation reaction of previously synthesized 6-5 .333 Purification of 6-11 is very convenient. No chromatography is
224 required in this process as the compound can be precipitated from a good solvent (e.g. CHCl3) into a poor solvent (e.g. methanol). Compound 6-11 is isolated as a shiny, dark-red powder. This process can be repeated two or three time to obtain high purity 6-11 In our hands, a 5.0 g (product) scale experiment was successfully performed. 1H-NMR spectrum is shown (Figure 630) to demonstrate the purity of the mate rial obtained by this simple purification. Figure 6-30. Synthesis of PIsoI-C16 and PIsoI-BT-C16. a) Pd(dppf)Cl2, KOAc, 1,4-dioxane, 80 oC, 75%; b) Pd2(dba)3, P(o-tyl)3 Et4NOH, toluene, 85 oC, 95% and 93% for PIsoI-C16 and PIsoI-BT-C16, respectively. Figure 6-31. 1H-NMR spectrum of compound 6-11.
225 With the key monomer 6-11 in hand, we chose to begin e xploring the chemistry of these electron deficient monomers by inclusion into two polymers PIsoI-C16 and PIsoI-BT-C16 Suzuki polymerization of 6-11 with 6-5 and 4,7-dibromobenzo[c ][1,2,5]thiadiazole ( 3-1 ) under our standard conditions (Pd2(dba)3, P(o-tyl)3 Et4NOH, toluene, 85 oC) gave PIsoI-C16 and PIsoI-BT-C16 in excellent yields with a number-avera ge molecular weights of 28.7 kDa (PDI, 2.4) for PIsoI-C16 and 16.3 kDa (PDI, 3.5) for PIsoI-BT-C16 The most intriguing properties, as expected, are their electron affinity levels and ionization potentials, which determine the applicability of these polymers as a possible alternative to PCBM derivatives. Table 6-6 and Figure 6-32 su mmarize the energy levels and gaps of PIsoI-C16 and PIsoI-BT-C16 measured by differential pu lse voltammetry. Polymer thin films were drop-cast from toluene solutions onto platinum button elect rodes and characterized in an electrochemical cell comprising a platinum count er-electrode and an Ag/Ag+ reference electrode. All the measurements were carried out in an argon-filled dry box. A ll estimated pote ntials were subsequently calibrated versus Fc/Fc+. The polymer films were repeatedly cycled prior to characterization, until a stab le and reproducible redox res ponse was finally obtained. Figure 6-32. Differential pulse voltammetry of PIsoI-C16 and PIsoI-BTC16. Electrochemical reduction and oxidation of the films on a pl atinum button was carried out in 0.1 M TBAP6/ACN supporting electrolyte using Ag/Ag+ reference electrode (calibrated against Fc/Fc+) and a platinum flag as the counter electrode.
226 As shown in Figure 6-32 and Table 6-6, PIsoI-C16 shows a low oxidation onset of 1.1 V and a reduction onset of -1.1 V (versus Fc/Fc+). These results suggest PIsoI-C16 is difficult to oxidize, but easy to redu ce. This implies that PIsoI-C16 has more n-character than p-charter. The HOMO and LUMO levels calculated from th e redox onsets are -6.2 eV and -4.0 eV (versus vacuum), associated with a bandgap of 2.2 eV. These values are very close to the HOMO and LUMO levels of PC61BM and PC71BM, suggesting PIsoI-C16 could find potential use as an excellent alternative to these commonly used a cceptors in polymer-based solar cells. Similarly, PIsoI-BT-C16 has a HOMO level of -6.1 eV and LUMO le vel of -4.1 eV. There is about 0.1 eV energy offset for both the HOMO and LUMO levels between PIsoI-C16 and PIsoI-BT-C16 By inclusion of electron-deficient units other than benzothiadiazole it should be possible to fine-tune the energy levels of resulting poly mers. In that sense, a polymer can be customized with the desired energy levels for n-type channel applications, which is ab sent or difficult to achieve in PCBM derivatives. Table 6-7. Electrochemically determined energy le vels and gaps of PIsoI-C16 and PIsoI-BT-C16 by DPV. In summary, we demonstrat ed a facile approach to access isoindigo-based polymers composed of only electron-defici ent units via environmental frie ndly chemistry (Suzuki coupling versus Stille coupling). This se t of two polymers presents LUMO energy levels as deep as -4.1
227 eV, while still retaining a deep HOMO level. It is conceivable that the scope of this chemistry can be greatly expanded by using other readily available electron-deficient moieties and other isoindigo-based polymers can be readily prepared. We envision these polymers will be of interest in the utilization of n-type FETs and all-polymer solar cells. 6.5 Conclusion This chapter devotes four secti ons to isoindigo-based material s. In Section 1, isoindigo was introduced as electron-accepti ng in electroactive material. DAD and ADA molecules based on this acceptor were prepared and used as active materials in bulk-heteroj unction solar cells where power conversion efficiency as high as 1.85% was achieved. With the establishment of isoindigo-chemistry, molecular engi neering was applied in the sec ond section to design a set of four structurally similar isoindi go-based discrete oligomers. The structure-property relationships of these molecules were discussed in depth, conne cting the significance of material design with achieving high performance photovoltaic materials. In the Section 3, three isoindigo-based polymers were synthesized in order to test a hyp odissertation that charge mobility in conjugated polymers can be enhanced via enforcement of interactions by means of introducing biphasic solubilizing groups. Various methods of characterization were employed to verify enhancement of interactions in conjugated polymers with biphasic solubilizing gr oups, including solidstate UV-vis absorption, temperature-depende nt solution UV-vis absorption, 2D-WAXS and solid-state NMR studies. The results from SCLC modeling of J-V charac teristics of singlecarrier diodes are consistent w ith the hypodissertation presented. Both electron and hole mobility have been significantly enhanced in polymers with biphasic groups, in comparison to the polymer with regular branched alkyl chains. The last section, Section 4, is devoted to a critical challenge in the field of conjugated polymer, access to stable and solution processable n-type
228 polymers. We conceived a facile appro ach to isoindigo-base d electron deficient -conjugated polymers in which the LUMO level is as deep as -4.1 eV while retaining a deep HOMO level as well. These polymers can be considered as an al ternative to commonly us ed acceptors such as PCBM derivatives currently employed in polym er-based solar cell devices. These electron deficient polymers may also find utility in ntype FET devices or co mplimentary integrated circuits. Because of the simp le chemistry involved, we envisi on that the Synthesis of these polymers can be easily moved to large scales for practical applications. 6.6 Experimental Details 6,6-dibromoisoindigo (6-1) To a suspension of 6-bromooxindole (500 mg, 2.36 mmol) and 6-bromoisatin (533 mg, 2.36 mmol) in AcOH (15 mL), conc HC1 solution (0.1 ml) was added and heated under reflux for 24. The mixture was allowed to cool and filtered. The solid material was washed with water, EtOH and AcOEt. After drying under vacuum, it yielded brown 6,6-dibromoisoindigo (951 mg, 95%). 1H-NMR (((CD3)2NCOD), DMF at 80 oC) : 10.7 (bs, 2H), 9.14 (d, J = 8.7 Hz, 2H), 7.22-7.15 (m, 4H). 13C-NMR (CDCl3) : 170.3, 147.2, 134.0, 132.3, 127.0, 125.3, 122.6, 113.9. 6,6-dibromo-N,N-(2-ethylhexyl)-isoindigo (6-2) To a suspension of 6,6dibromoisoindigo (420 mg g, 1 mmol) and potassium carbonate (829 mg, 5 mmol) in dimethylformaldehyde (DMF) (20 mL), 1-bromo2-ethylhexane (425 g, 2.2 mmol) was injected through a septum under nitrogen. The mixture was stirred for 15 h at 100 oC and then poured into water (200 mL). The organi c phase was extracted by CH2Cl2, washed with brine and dried over MgSO4. After removal of the solvent under reduced pressure, the deep-red solids were purified by silica chromatograph y, eluting with (CH2Cl2: Hexane = 1:1) to give 6,6-dibromo-N,N-(2ethylhexyl)-isoindigo (548 mg, 85 %) 1H-NMR (CDCl3) : 9.00 (d, J = 8.7 Hz, 2H), 7.13 (dd, J1
229 = 8.7 Hz, J2 = 1.5 Hz, 2H), 6.81 (d, J = 1.5 Hz, 2H), 3.60-3.48 (m, 4H), 1.90-1.72 (m, 2H), 1.431.20 (m, 16H), 0.95-0.82 (m, 12H). 13C-NMR (CDCl3) : 168.1, 146.2, 132.6, 131.2, 126.8, 125.2, 120.5, 111.6, 44.5, 37.6, 30.7, 28.7, 24.1, 23.2, 14.2, 10.8. HRMS (ESI-TOF) Calculated for C32H40Br2N2O2 (M+H)+: 645.1512, found: m/z 645.1510. Anal. Calcd for C32H40Br2N2O2: C, 59.64; H, 6.26; N, 4.35; Found: C, 59.79; H, 6.30; N, 4.26. 6,6-(N,N-heptyl)-dibromoisoindigo (6-3) .To a suspension of 6,6-dibromoisoindigo (1.39 g 3.31 mmol) and potassium carbonate (2.74 g, 19.85 mmol) in DMF (15 mL), 1-bromoheptane (1.78 g, 9.93 mmol) was injected thr ough a septum under nitrogen. The mixture was stirred for 15 h at 100 oC and then poured into water (200 mL ). The organic phase was extracted by CH2Cl2, washed with brine and dried over MgSO4. After removal of the solvent under reduced pressure, the pink-red solids were pur ified by silica chromat ography, eluting with (CH2Cl2: Hexane = 1:1) to give 6,6-dibro mo-N,N-(2-heptyl)-isoindigo (1.74 g, 85 %) 1H-NMR (CDCl3) : 9.07 (d, J = 9 Hz, 2H), 7.17 (dd, J1 = 9 Hz, J2 = 1.8 Hz, 2H), 6.91 (d, J = 1.8 Hz, 2H), 3.71 (t, J = 7.2 Hz, 4H), 1.75-1.60 (m, 4H), 1.40-1.20 (m, 16H), 0.87 (t, J = 6.6 Hz, 6H); 13CNMR (CDCl3) : 167.9, 146.0, 132.9, 131.4, 127.0, 125.3, 120.6, 111.5, 40.5, 31.9, 29.2, 27.6, 27.2, 22.8, 14.3. HRMS (ESI-TOF) Calculated for C30H36Br2N2O2 (M+H)+: 617.1198, found: m/z 617.1201; Anal. Calcd for C30H36Br2N2O2: C, 58.45; H, 5.89; N, 4.54; Found: 58.56; H, 5.78; N, 4.61. 6,6-(N,N-propyl)-dibromoisoindigo (6-4) 1H-NMR (CDCl3) : 9.08 (d, J = 8.7 Hz, 2H), 7.15 (dd, J1 = 8.7 Hz, J2 = 1.5 Hz, 2H), 6.89 (d, J = 1.5 Hz, 2H), 3.70 (t, J = 5.2 Hz, 4H), 1.76 (sex, J = 5.2Hz, 4H), 1.01 (t, J = 5.2Ha, 6H); 13C-NMR (CDCl3) : 168.0, 146.2, 132.8, 131.6, 127.0, 125.3, 120.8, 111.5, 42.1, 21.0, 11.6.
230 6,6-(N,N-2-hexyldecyl)-dibromoisoindigo (6-5) To a suspension of 6,6dibromoisoindigo (4.20 g, 10.0 mmol) and potass ium carbonate (8.29 g, 60 mmol) in DMF (45 mL), 2-hexyl-bromodecane (9.16 g, 30 mmol) was injected through a septum under nitrogen. The mixture was stirred for 15 h at 100 oC and then poured into water (200 mL). The organic phase was extracted by CH2Cl2, washed with brine and dried over MgSO4. After removal of the solvent under reduced pressure, the dark red liquids were purified by silica chromatography, eluting with (CH2Cl2: Hexane = 1:2) to give 6,6-dibromo-N,N-(2-hexyldecyl)-isoindigo (6.1 g, 70 %). 1H-NMR (CDCl3) : 9.04 (d, J = 9 Hz, 2H), 7.14 (dd, J1 = 9Hz, J2 = 1.8 Hz, 2H), 6.83 (d, J = 1.8 Hz, 2H), 3.56 (d, J = 7.5 Hz, 4H), 1.83 (bs, 2H), 1.40-1.24 (m, 48H), 0.88-0.84 (m, 12H); 13C-NMR (CDCl3) : 168.2, 146.3, 132.6, 131.3, 126.9, 125.2, 120.6, 112.7, 44.9, 36.3, 32.1, 32.0, 31.7, 30.2, 29.87, 29.77, 29.5, 26.6, 22.89, 22.87, 14.34, 14.31. 6-bromoisoindigo (6-6) : To a suspension of oxindole (1.47g, 11.06 mmol) and 6bromoisatin (2.50 g, 11.06 mmol) in AcOH (40 mL), conc. HC1 solution (0.2 ml) was added and heated under reflux for 24. The mixture was allowed to cool and filtered. The solid material was washed with water, EtOH and AcOEt. Af ter drying under vacuum, it yielded deep red 6bromoisoindigo (3.53 g, 94%). 1H-NMR ((CD3)2NCOD), DMF at 80 oC) : 10.7 (bs, 1H), 10.6(bs, 1H), 9.20-9.13 (m, 2H), 7.42-6.95 (m, 5H). 13C-NMR (CDCl3) : 170.5, 170.3, 147.0, 146.0, 135.4, 134.0, 133.3, 132.2, 130.9, 126.6, 125.1, 123.5, 122.7, 122.4, 113.7, 110.8. 6,-bromo-N,N-(2-ethylhexyl)-isoindigo ( 6-7 ): To a suspension of 6-bromoisoindigo (1.0 g, 2.93 mmol) and fresh-dried potassium carbona te (2.43 g, 17.59 mmol) in anhydrous DMF (20 mL), 1-bromo-2-ethylhexane ( 1.70 g, 8.79 mmol) was injected th rough a septum under nitrogen. The mixture was stirred for 15 h at 100 oC and then poured into water (500 mL). The organic phase was extracted by CH2Cl2, washed with brine and dried over MgSO4. After removal of the
231 solvent under reduced pressure the deep-red solids were purified by silica chromatography, eluting with (CH2Cl2: Hexane = 2:3) to give compound 7 (Rf = 0.45), 8 (Rf = 0.35), and 4 (Rf = 0.53), of, 740 mg (44.7 %), 362 mg (25.4 %) and 368 mg (19.5 %), respectively. Compound 6-7: 1H-NMR (CDCl3) : 9.12 (dd, J1 = 7.8 Hz, J2 = 1.2 Hz, 1H), 9.04 (d, J = 8.4 Hz, 1H), 7.34 (dt, J1 = 7.8 Hz, J2 = 1.2 Hz, 1H), 7.13 (dd, J1 = 8.4 Hz, J2 = 1.8 Hz, 1H), 7.03 (dt, J1 = 7.8 Hz, J2 = 0.6 Hz, 1H), 6.83 (d, J = 1.8 Hz, 1H), 6.73 (dd, J1 = 7.8 Hz, J2 = 0.6 Hz, 1H), 3.65-3.45 (m, 4H), 1.95-1.85 (m, 2H), 1.43-1.20 (m, 16H), 0.98-0.82 (m, 12H). 13C-NMR (CDCl3) : 168.3, 168.2, 146.2, 145.3, 134.0, 132.7, 132.2, 131.1, 130.1, 126.4, 125.1, 122.4, 121.7, 120.6, 111.5, 108.3, 44.45, 44.34, 37.70, 37.58, 30.86, 30.74, 28.89, 28.75, 24.22, 24.15, 23.22, 14.22, 10.86, 10.80. HRMS (ESI-TOF) Calculated for C32H41BrN2O2 (M+H)+: 567.2409, found: m/z 567.2416. Anal. Calcd for C32H41BrN2O2: C, 67.95; H, 7.31; N, 4.95; Found: C, 67.86, H, 7.27, N, 4.91. Compound 6-8: 1H-NMR (CDCl3) : 9.17 (dd, J1 = 7.8 Hz, J2 = 1.2 Hz, 2H), 7.35 (dt, J1 = 7.8 Hz, J2 = 1.2 Hz, 2H), 7.05 (dt, J1 = 7.8 Hz, J2 = 0.6 Hz, 2H), 6.76 (dd, J1 = 7.8 Hz, J2 = 0.6 Hz, 2H), 3.75-3.58 (m, 4H), 1.92-1.78 (m, 2H), 1.45-1.20 (m, 16H), 0.98-0.80 (m, 12H). 13CNMR (CDCl3) : 168.4, 145.3, 133.7, 132.4, 130.0, 122.3, 121.9, 108.3, 44.4, 37.7, 30.9, 28.9, 24.2, 23.2, 14.2, 10.9. HRMS (ESI-TOF) Calculated for C32H42N2O2 (M+H)+: 487.3319, found: m/z 487.3351. I-1 : In a flame-dried Schlenk flask (50 mL ), 6,6-dibromoisoindigo (446 mg, 0.69 mmol), 2-(5'-hexyl-2,2'-bithiophen-5-y l)-4,4,5,5-tetramethyl-1,3,2-dioxa borolane (625 mg, 1.66 mmol, 2.4 eq), Tris(dibenzylideneacetone)dipalladium(0) (15 mg) and P(o-tyl)3 (10 mg) were added. The flask was evacuated and back-filled with ar gon three times, after wh ich degassed toluene (15 mL) and tetraethylammonium hydroxide (4.2 mmol, 1M) was transferred to the mixture through
232 a septum. The resulting solution was heated up to 85 oC under argon and stirred for 20 h. The solvent was removed under reduced pressure. Th e dark red solids were purified by silica gel chromatography, eluting with CH2Cl2-hexane (1:1) to give br own solids (562 mg, 83 %). 1HNMR (CDCl3) : 9.01 (d, J = 8.7, 2H), 7.25-7.18 (m, 4H), 7.05 (d, J = 3.6 Hz, 2H ) 7.02 (d, J = 3.6 Hz, 2H ), 6.78 (s, 2H), 6.70 (d, J = 3.3 Hz), 3.68-3.41 (m, 4H), 2.80 (t, J = 4.2 Hz, 4H), 1.821.65 (m, 6H), 1.44-1.20 (m, 28H), 0.98-0.81 (m, 18H) 13C-NMR (CDCl3) : 168.7, 146.3, 145.6, 142.1, 138.9, 137.4, 134.7, 131.6, 130.3, 125.1, 125.07, 124.1, 123.9, 121.1, 118.8, 104.5, 44.2, 38.0, 31.79, 31.75, 31.1, 30.4, 29.2, 29.0, 24.5, 23.3, 22.8, 14.4, 14.3, 11.0. HRMS (ESI-TOF) Calculated for C60H74N2O2S4 (M+H)+: 983.4706 Found: m/z 983.4741. Anal. Calcd for C60H74N2O2S4: C, 73.27; H, 7.58; N, 2.85. Found: C, 73.39; H,7.57; N, 2.80. I-2 : In a flame-dried Schlenk flask (50 mL), 6,6-(N,N-heptyl)-dibromoisoindigo (616 mg, 1 mmol), 2-(5'-hexyl-2,2' -bithiophen-5-yl)-4,4,5,5tetramethyl-1,3,2-dioxaborolane (904 mg, 2.4 mmol, 2.4 eq), Tris(dibenzylideneacet one)dipalladium(0) (15 mg) and P(o-tyl)3 (10 mg) were added. The flask was evacuated and back-fille d with argon three times, after which degassed toluene (20 mL) and tetraethylammonium hydrox ide (3 mmol, 1M) was transferred to the mixture through a septum. The resul ting solution was heated up to 85 oC under argon and stirred for 20 h. The solvent was removed under reduced pressure. The dark red solids were purified by silica gel chromatography, eluting with CH2Cl2-hexane (1:1) to give metallic crystalline solids (795 mg, 83 %). 1H-NMR (CDCl3) : 9.11 (d, J = 8.4, 2H), 7.24 (d, J = 3.6 Hz, 2H), 7.19(dd, J1 =8.4 Hz, J2 =1.8 Hz, 2H ), 7.03 (d, J =3.6 Hz, 2H ), 7.00 (d, J =3.6 Hz, 2H ), 6.79 (d, J = 1.8 Hz, 2H ) 6.68 (d, J =3.6 Hz, 2H ), 3.74 (t, J = 4.2 Hz, 4H), 2.78 (t, J = 7.8 Hz, 4H), 1.75-1.63 (m, 8H), 1.44-1.24 (m, 28H), 0.95-0.82 (m, 12H) 13C-NMR (CDCl3) : 168.4, 146.2, 145.3, 142.1, 138.9, 137.5, 134.7, 131.7, 130.6, 125.2, 125.1, 124.1, 123.9, 121.1, 118.9, 104.3, 40.2, 32.0,
233 31.8, 30.5, 29.2, 29.0, 27.8, 27.3, 22.9, 22.8, 14.3. HRMS (ESI-TOF) Calculated for C58H70N2O2S4 (M+H)+: 955.4393 Found: m/z 955.4382. Anal. Calcd for C58H70N2O2S4: C, 72.91; H, 7.38; N, 2.93. Found: C, 72.93; H,7.51; N, 3.06. I-3 : In a flame-dried Schlenk flask (50 mL), 6-bromoisoindigo (1 g, 1.77 mmol), 5,5'5,5'bis(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl )-2,2'-bithiophene (308 mg, 0.736 mmol), Tris(dibenzylideneacetone)dipalladium(16 mg) and P(o-tyl)3 (12 mg) were added. The flask was evacuated and back-filled with argon three times, after whic h degassed toluene (15 mL) and tetraethylammonium hydroxide (4.5 mmol, 1M) was transferred to the mixt ure through a septum. The resulting solution was heated up to 85 oC under argon and stirred for 20 h. The solvent was removed under reduced pressure. The dark red so lids were purified by silica gel chromatography, eluting with CH2Cl2-hexane (1:1) to give br own solids (476 mg, 57 %). 1H-NMR (CDCl3) : 9.22 (t, J = 9.0 Hz, 4H), 7.34-7.29 (m, 6H), 7.28 (d, J = 8.4 Hz, 2H), 7.20 (d, J = 3.9 Hz, 2H), 7.02 (t, J = 7.5 Hz, 2H), 6.94 (d, J = 1.2 Hz, 2H), 6.75(d, J = 8.4 Hz, 2H), 3.80-3.55 (m, 8H), 1.98-1.83 (m, 4H), 1.55-1.25 (m 32H), 1.05-0.90 (m, 24 H). 13C-NMR (CDCl3) : 168.9, 168.7, 146.2, 145.5, 143.7, 138.0, 137.6, 133.0, 132.8, 132.3, 130.8, 130.1, 125.3, 122.31, 122.28, 121.6, 119.3, 108.3, 104.9, 44.66, 44.60, 38.22, 38.05, 31.33, 31.23, 29.25, 29.10, 24.75, 24.57, 23.33, 23.30, 14.27, 14.17, 11.08, 10.98. HRMS (ESI-TOF) Calculated for C72H86N4O4S2 (M+H)+: 1135.6163, found: m/z 1135.6167; Anal. Calcd for C72H86N4O4S2: C, 76.15; H, 7.63; N, 6-13 Temperature-dependent UV-vis spect IsoI-N : In a flame-dried Schlenk flask (50 mL ), 6,6-(N, N-ethylhexyl-)dibromoisoindigo (644 mg, 1 mmol), 1-methyl-2-(4,4,5,5-tetramethyl -1,3,2-dioxaborolan-2-yl)-1H-indole (669 mg, 2.5 mmol, 2.5 eq), Pd2(dba)3-CHCl3 (18 mg) and P(o-tyl)3 (10 mg) were added. The flask was evacuated and back-filled with argon three times, after whic h degassed toluene (30 mL) and
234 tetraethylammonium hydroxide (3 mmol, 1M) was transferred to the mixture through a septum. The resulting solution was heated up to 85 oC under argon and stirred for 16 h. The solvent was removed under reduced pressure. The dark red so lids were purified by silica gel chromatography, eluting with CH2Cl2-hexane (1:1) to give metallic crystalline solids (656 mg, 88 %). 1H-NMR (CDCl3) : 9.29 (d, J = 8.1, 2H), 7.69 (d, J = 8.1 Hz, 2H), 7.40-7.15(m, 8H ), 6..95 (d, J =1.5 Hz, 2H), 6.71 (d, J =0.6 Hz, 2H ), 3.85 (s, 6H), 3.80-3.62 (m, 4H), 1.95-1.85 (m, 2H), 1.50-1.30 (m, 16H), 0.99-0.85 (m, 12 H). 13C-NMR (CDCl3) : 168.8, 145.7, 141.4, 139.3, 136.7, 132.9, 130.0, 128.1, 122.9, 122.7, 121.3, 121.0, 120.5, 110.0, 108.8, 103.1, 44.6, 38.0, 31.9, 31.1, 29.1, 24.4, 23.3, 14.3, 11.0. HRMS (ESI-TOF) Calculated for C50H56N4O2 (M+H)+: 745.4447 Found: m/z 745.4476. Anal. Calcd for C50H56N4O2: C, 80.61; H, 7.58; N, 7.52. Found: C, 80.56; H,7.63; N, 7.52. IsoI-O : In a flame-dried Schlenk flask (50 mL ), 6,6-(N, N-ethylhexyl-)dibromoisoindigo (616 mg, 1 mmol), benzo[b]thiophen-2-ylbo ronic acid (445 mg, 2.5 mmol, 2.5 eq), Pd2(dba)3CHCl3 (18 mg) and P(o-tyl)3 (10 mg) were added. The flask was evacuated and back-filled with argon three times, after which degassed toluen e (30 mL) and tetraethylammonium hydroxide (3 mmol, 1M) was transferred to the mixture throug h a septum. The resulting solution was heated up to 85 oC under argon and stirred for 16 h. The solven t was removed under reduced pressure. The dark red solids were purified by sili ca gel chromatography, eluting with CH2Cl2-hexane (1:1) to give metallic crys talline solids (688 mg, 92 %). 1H-NMR (CDCl3) : 9.11 (d, J = 8.4, 2H), 7.82-7.72 (m, 4H), 7.53 (s, 2H), 7.38-7.30 (m, 6H), 6.89(t, J =1.5 Hz, 2H ), 3.72-3.45 (m, 4H), 1.83-1.77(m, 2H), 1.42-1.24 (m, 16H), 0.97-0.83 (m, 12H) 13C-NMR (CDCl3) : 168.6, 145.7, 143.9, 140.7, 139.9, 137.7, 132.3, 130.5, 125.2, 125.0, 124.2, 122.5, 121.9, 120.8, 120.0, 105.6, 44.2, 38.0, 31.1, 29.2, 24.5, 23.3, 14.4, 11.1. HRMS (ESI-TOF) Calculated for
235 C48H50N2O2S2 (M+H)+: 751.3386 Found: m/z 751.3382. Anal. Calcd for C48H50N2O2S2: C, 76.76; H, 6.71; N, 3.73. Found: C, 76.53; H,6.65; N, 3.74. IsoI-S : In a flame-dried Schlenk flask (50 mL ), 6,6-(N, N-ethylhexyl-)dibromoisoindigo (616 mg, 1 mmol), benzofuran-2-ylboroni c acid (405 mg, 2.5 mmol, 2.5 eq), Pd2(dba)3CHCl3(18 mg) and P(o-tyl)3 (10 mg) were added. The flask wa s evacuated and back-filled with argon three times, after which degassed toluen e (30 mL) and tetraethylammonium hydroxide (3 mmol, 1M) was transferred to the mixture throug h a septum. The resulting solution was heated up to 85 oC under argon and stirred for 16 h. The solven t was removed under reduced pressure. The dark red solids were purified by sili ca gel chromatography, eluting with CH2Cl2-hexane (1:1) to give metallic crys talline solids (795 mg, 84 %). 1H-NMR (CDCl3) : 9.19 (d, J = 8.4, 2H), 7.58-7.45 (m, 6H), 7.32-7.20 (m, 4H), 7.07-7.04(m, 4H ), 3.76-3.52 (m, 4H), 1.88-1.78(m, 2H), 1.42-1.23 (m, 16H), 0.99-0.87 (m, 12H). 13C-NMR (CDCl3) : 168.6, 155.5, 155.3, 145.7, 133.6, 132.4, 130.3, 129.3, 125.3, 123.5, 122.1, 121.4, 118.7, 111.5, 104.1, 103.6, 44.3, 37.9, 31.0, 29.1, 24.5, 23.3, 14.4, 11.0. HRMS (ESI-TOF) Calculated for C48H50N2O4 (M+H)+: 719.3843 Found: m/z 719.3851. Anal. Calcd for C48H50N2O4: C, 80.19; H, 7.01; N,3.90. Found: C, 79.97; H,7.05; N, 3.91. 2-(2-(2-(tosyloxy)ethoxy)ethoxy)ethyl nonanoate: To a round-bottom flask (250 mL), 2(2-(2-hydroxyethoxy)ethoxy)ethyl 4-methylbe nzenesulfonate (14.13 g, 46.41 mmol) and triethylamine (5.64 g, 7.76 mL, 55.69 mmol) in anhydrous dichloromethane (92 mL) were cooled in ice-water batch. nonanoyl chloride (8.61g, 48.73 mmol) in dichloromethane (8 mL) was added dropwise over 15 min. The mixture was stirred for 1 h and then warmed up to room temperature. After the alcohol was completely consumed, the mixture was poured into water. The organic phase was washed with brine and dried over magnesium sulfate. The solvent was
236 removed under reduced pressure. The visc ous orange oil was purified by silica gel chromatography, eluting with ethyl acetate and he xane (1:1) to yield pa le yellow oil (17.1 g, 83 %). 1H-NMR (CDCl3) : 7.76 (dd, J1 = 7.8 Hz, J2 = 1.5 Hz, 2H), 7.30 (dd, J1 = 7.8 Hz, J2 = 1.5 Hz, 2H), 4.19-4.06 (m, 4H), 3.70-3.55 (m, 4H), 3.53 (bs, 4H), 2.39 (s, 3H), 2.29 (t, J = 7.5 Hz, 2H), 1.61-1.49 (m, 2H), 1.28-1.16 (m, 10H), 0.82 (t, J = 6 Hz, 2H) 13C-NMR (CDCl3) : 173.8, 145.0, 133.0, 129.9, 128.0, 70.8, 70.5, 69.3, 69.29, 68.8, 63.3, 34.2, 31.9, 29.3, 29.2, 25.0, 22.7, 21.7, 14.2. HRMS (ESI-TOF) Calculated for C22H36O7S (M+H)+: 444.2182 Found: m/z 843.3167; Anal. Calcd for C22H36O7S: C, 59.43; H, 8.16; Found: C, 59.78; H, 8.25. 2-(2-(2-methoxyethoxy)ethoxy) ethyl 6-bromohexanoate To a round-bottom flask (250 mL), 2-(2-(2-methoxyethoxy)ethoxy)ethanol (8.0 8 g, 49.18 mmol) and triethylamine (5.69 g, 7.83 mL, 56.21 mmol) in anhydrous dichloromethane (92 mL) were cooled in ice-water batch. 6bromohexanoyl chloride (10g, 46.84 mmol) in dichloromethane (8 mL) was added dropwise over 15 min. The mixture was stirred for 1 h and then warmed up to room temperature. After 6bromohexanoyl chloride was completely consum ed, the mixture was poured into water. The organic phase was washed with brine and dried over magnesi um sulfate. The solvent was removed under reduced pressure. The visc ous orange oil was purified by silica gel chromatography, eluting with ethyl acetate and hexane (1:1) to yiel d colorless oil (13.5 g, 85 %). 1H-NMR (CDCl3) : 4.25-4.15 (m, 2H), 3.70-3.60 (m, 7H), 3.54-3.50 (m, 2H), 3.40-3.34 (m, 4H), 2.32 (t, J = 7.2 Hz, 2H), 1.85 (quin, J = 4.2 Hz, 2H), 1.64 (quin, J = 7.5 Hz, 2H), m 1.491.40 (m, 2H); 13C-NMR (CDCl3) : 173.6, 72.1, 70.78, 70.75, 70.73, 69.3, 63.6, 59.2, 34.1, 33.6, 32.5, 27.8, 24.2. HRMS (ESI-TOF) Calculated forC13H25BrO5 (M+H)+: 341.0958 Found: m/z 341.0964; Anal. Calcd for C13H25BrO5: C, 45.76; H, 7.38; Found: C, 46.03; H, 7.46.
237 Compound 6-9 To a 6,6-dibromoisoindigo mixtur e (841 mg, 2 mmol) and potassium carbonate (1.66 g, 12 mmol) in DMF (10 mL ), 2-(2-(2-methoxyethoxy)ethoxy)ethyl 6bromohexanoate (1.5 g, 4.4 mmol) was injected through a septum under nitrogen. The mixture was stirred for 15 h at 100 oC and then poured into water ( 100 mL). The organic phase was extracted by CH2Cl2, washed with brine and dried over MgSO4. After removal of the solvent under reduced pressure, the deep-red solids were purified by silica chromatography, eluting with pure ethyl acetate to give th e desired compound (665 mg, 35 %). 1H-NMR (CDCl3) : 8.91 (d, J = 8.4Hz, 2H), 7.03 (dd, J1 = 8.4 Hz, J2 = 1.8 Hz, 2H), 6.76 (d, J = 1.8 Hz, 2H), 4.14 (t, J = 4.8 Hz, 4H), 3.62-3.51(m, 20H), 3.50-3.42 (m, 4H), 3.28 (s, 6H), 2.25 (t, J = 7.2 Hz, 4H), 1.63-1.53 (m, 8H), 1.37-1.25 (m, 4H); 13C-NMR (CDCl3) : 173.2, 167.4, 145.5, 132.3, 131.1, 126.7, 125.0, 120.2, 111.0, 71.8, 70.52, 70.47, 69.1, 63.4, 58.9, 39.9, 33.8, 27.0, 26.4, 24.4. HRMS (MALDI-TOF) Calculated for C42H56Br2N2O12 (M+H)+: 941.2258 Found: m/z 941.2220. Compound 6-10 To a 6,6-dibromoisoindigo mixture (2.10 g, 5mmol) and potassium carbonate (4.15 g, 30 mmol) in DMF (20 mL), 2-(2-(2-(tosyloxy)ethoxy)ethoxy)ethyl nonanoate (6.67 g, 15 mmol) was injected through a septum under nitrogen. The mixture was stirred for 15 h at 100 oC and then poured into water (200 mL). The organic phase was extracted by CH2Cl2, washed with brine and dried over MgSO4. After removal of the solv ent under reduced pressure, the deep-red solids were purified by s ilica chromatography, eluting with CH2Cl2: acetone (97:3) to give the desired compound (2.39 g, 49 %). 1H-NMR (CDCl3) : 8.93 (d, J = 9.3 Hz, 2H), 7.037.00 (m, 4H), 4.10 (t, J = 4.8 Hz, 4H), 3.83 (t, J = 5.4 Hz, 4H), 3.66 (t, J = 4.8 Hz, 4H), 3.57 (t, J = 4.8 Hz, 4H), 3.52 (s, 8H), 2.24 (t, J = 7.5 Hz, 4H), 1.60-1.48 (m, 4H), 1.28-1.15(m, 20 ), 0.80 (t, J = 7.2 Hz, 6H);13C-NMR (CDCl3) : 173.67, 167.7, 146.1, 132.2, 130.9, 12637, 125.0, 120.2,
238 112.3, 70.9, 70.6, 69.4, 68.96, 63.3, 40.5, 34.2, 31.8, 29.24, 29.13, 24.9, 22.7, 14.1. HRMS (MALDI-TOF) Calculated for C46H64Br2N2O10 (M+Na)+: 987.2806 Found: m/z 987.2811. The general method to prepare PIsoIAM-1 PIsoIAM-2 and PIsoIAM-3 A Schlenk flask charged with compound 6-5/6-9/6-10 (1 mmol), Pd2dba3 (18 mg), P( o -Tolyl)3 (15 mg), and 2,5bis(trimethyltin)thiophene (1 mmol) was purged three times with successive vacuum and argon filling cycles. Degassed anhydrous toluene (15 mL) was injected through a septum. The polymerization reaction was heated to 95C, and the mixture was stirred for 36 h under argon atmosphere. The mixture was cooled to room te mperature and poured slowly in methanol (300 mL). The precipitates were collected through a 0.45 m PTFE filter. The crude polymer was purified with Soxhlet extraction with methanol, hexane to rem ove low molecular species and catalyst residues. Finally the pol ymer was extracted with chloroform. The polymer solution was concentrated and slowly poured in methanol ( 300 mL). The precipitate s were collected via vacuum filtration through a 0.45 m PTFE filter and dried, yielding a dark purple polymer. PIsoIAM-1 1H-NMR (CDCl3) : 9.19-8.61 (m, 2H), 7.45-5.98 (m, 6H), 4.56-3.20 (m, 34H), 2.82-1.21 (m, 16H). Anal. Calcd for PIsoIAM-1 : C, 63.06; H,6.70; N,3.25. Found: C, 64.00; H, 6.80; N, 3.25. GPC: Mw =35 kDa, PDI = 1.2. PIsoIAM-2 1H-NMR (CDCl3) : 9.10-8.67 (m, 2H), 7.42-6.25 (m, 6H), 4.45-3.28 (m, 24H), 2.45-1.08 (m, 28H), 0.98-0.70 (m, 6H). Anal. Calcd for PIsoIAM-2 : C, 67.04; H,7.33; N,3.08. Found: C, 67.42; H, 7.43; N, 3.20. GPC: Mw =78 kDa, PDI = 1.8. PIsoIAM-3 GPC: Mw =34 kDa, PDI = 1.6. Compound 6-11 6,6-(N,N-2-hexyldecy l)-dibromoisoindigo (6 -5) (4.35 g, 5.0 mmol), pinacol ester of diboron (3.05 g, 12 mmol), [PdCl2(dppf)] (220 mg), and potassium acetate (2.95 g, 30 mmol) were mixed at room temperat ure under an argon atmosphere. Anhydrous 1,4-
239 dioixane (2 mL) was injected with a syringe through a septum. The solution was heated at 80 oC for 30 h and then cooled to room temperatur e. The reaction mixture was filtered by passing through a short pad of silica gel, and washed by a mixture of methylene chloride and hexane (1:1). The collected filtration was concentrated and precipitated into co ld methanol (100 mL). The precipitates was filtered and dried to give dark red products (3.6 g, 75%).1H-NMR (CDCl3) : 9.15 (d, J = 7.2 Hz, 2H), 7.48 (d, J1 = 8.1Hz, J2 = 0.6 Hz, 2H), 7.15 (d, J = 0.6 Hz, 2H), 3.69 (d, J = 7.5 Hz, 2H), 1.95 (bs, 2H), 1.59-1.19 (m, 72H), 0.85 (t, J = 6.6 Hz, 6H); 13C-NMR (CDCl3) : 168.3, 144.7, 134.5, 129.0, 128.9, 124.4, 113.7, 84.2, 44.6, 36.3, 32.1, 32.0, 31.8, 31.2, 29.8, 29.78, 29.5, 26.6, 25.1, 22.9, 22.8, 14.32, 14.30. HRMS (MALDI-TOF) Calculated for C60H96B2N2O6 (M+Na)+: 963.7548 Found: m/z 963.7583. Anal. Calcd for C60H96B2N2O6: C, 74.83; H,10.05; N,2.91. Found: C, 74.91; H,10.(7)5; N, 2.80. PIsoI-C16 In a flame-dried Schlenk flask (50 mL), 6-5 (434.45 mg, 0.5 mmol) and 6-11 (481.52 mg, 0.5 mmol), Pd2(dba)3-CHCl3(18 mg) and P(o-tyl)3 (12 mg) were added. The flask was evacuated and back-filled with argon three tim es, after which degassed toluene (15 mL) and tetraethylammonium hydroxide (3 mmol, 1M) was transferred to the mixture through a septum. The resulting solution was heated up to 85 oC under argon and stirred for 36 h. The mixture was cooled to room temperature and poured slowly in methanol (300 mL). The precipitates were collected through a 0.45 m PTFE filter. The crude polymer was purified with Soxhlet extraction with methanol, hexane to remove low molecu lar species and catalyst residues. Finally the polymer was extracted with chloroform. The po lymer solution was concentrated and slowly poured in methanol (300 mL). The precipitates were collected via vacuum filtration through a 0.45 m PTFE filter and dried, yielding dark brown solids (673 mg, 95%). 1H-NMR (CDCl3) :
240 8.98-8.70 (M, 2H), 7.40-6.80 (m, 4H), 3.60-3.25 (m, 4H), 2.20-0.60 (m, 62H). GPC: Mw = 70.2 kDa, PDI = 2.4. PIsoI-C16-BT 1H-NMR (CDCl3) : 9.42-8.95 (m, 2H), 7.78-5.98 (m, 8H), 3.60-3.20 (m, 4H), 2.20-0.45 (m, 62H). GPC: Mw = 56.9 kDa, PDI = 3.5.
241 CHAPTER 7 PERSPECTIVES AND OUTLOOK The development of solution-processabl e organic semiconducting materials for photovoltaic applications has been the focus throughout this disserta tion. Interest in organic solar cells mainly lies in several facts. First, they can be printed via solution-processing. This makes high-speed manufacturing process po ssible in the fabrication of th e third generation organic solar cells, a distinction between orga nic and silicon-based solar cells Fast processing makes organic solar cells more suitable and attractive as renewable energy sources to meet our energy consumption pace. Second, they can be asse mbled under the ambient conditions, and thus consume very little energy in the process. On th e contrary, the fabrication of silicon-based solar cells involves massive energy consumption. In co mparison, the manufacturing of organic solar cells has a lower environmental impact, an im portant argument when it comes to developing renewable energy sources. Third, solution processed organic materi als are compatible with many flexible and light-weight substrates. This makes th em easily integrated into a variety of products including camping tents, clothi ng articles, and commerc ial boards among others. They are seen rich with the opportunity to create their own niche markets. For inst ance, soldiers in battle fields are constantly on the move and in need of pow er for their equipment. They would greatly benefit from a solar cell co ated tent or clothing th at could produce electric ity that would be used to recharge phones, GPS devices and other electronics. Bearing these features, organi c solar cells present a huge pot ential market. This has drawn enormous research efforts from both ends of th e spectrumacademia and industry. With certified power conversion efficiencies of 7.7 % (certifi ed by the Fraunhofer Institute for Solar Energy Systems ISE in Freiburg) over an active area of 1.1 cm achieved by the Dresden-based Heliatek GmbH and 7.9% (certified by the U.S. Depart ment of Energys National Renewable Energy
242 Laboratory) over 0.1 cm reported by Solarmer Ener gy, the future of organic solar cells seems unprecedentedly promising for practical utilization. On the other side, the competition to be first becomes increasingly fierce in the development of organic solar cells. This inev itably brings some undesi red influences in our academic practices, even though a significant incr ease in photovoltaic efficiencies has been realized during this course. When Alan Heeger Alan MacDiarmind, Hideki Shirakawa and their colleagues reported doped polyacety lene having extraordinarily hi gh conductive properties, as a matter of fact, they had not foreseen organic sola r cells (or any other elec tronic devices) and just wanted to understand the basic phys ics of how electrons were set fr ee in the materials. But they, together with other people, opened the door of organic electronics w ith a simple mind of understanding things behind. Therefore, we need to reassure ourselves that our primary role as a scientific researcher is to explore the unknowns and unders tand the fundamental aspects. The field of organic solar cel ls is indeed full of exciting and important problems which need investigation. Here are a few topics that may worth of our attention: 1) Systematically tuning optical density (abs orption coefficient) of -conjugated materials. It is well-understood that there is a trade-off between exciton diffusi on length and optical abso rption length in organic solar cells. An enhancement in optical density w ill release the tension of this trade-off to certain degree and also increase the po ssibility of free charge carrie rs reaching electrodes. An interesting question from a theoretical point of vi ew arises as well. What are the upper limits for the absorption coefficient in -conjugated materials? What kind of chemical structures will bring high optical densities? 2) Tent atively establishing the relationship between energy levels and charge carrier mobilities in -conjugated materials remains a challenge. Charge mobilities are related to the chemical structur es (inter/intra-chain interactions ) and material morphologies. In
243 fact, are charge mobilities affected by a materials HOMO/LUMO energy since different energy levels correspond to different charge stabiliza tion capabilities? As a chemist, how could we design a system to probe this relationship? 3) Co nstructing 2D or 3D conjugated materials. In 1D rod-like organic materials, anisotropic charge transport is dominant and therefore device performance is strongly dependent on orientat ion (morphology) of the material. Introducing (pseudo) 2D or 3D-conjugated materials will conceivably alleviate the anisotropy nature and enhance charge transport in these materials. Up to date, there are few efforts that have been directed towards in this regard. 4) Preci sely positioning donor and acceptor components at nanoscales. In the natural systems, charge transf er and energy transfer are highly efficient, presumably because of the accurate locations of the components with respect to each other. How could we learn from Nature and build such a prec isely positioned system? That is a challenge we need to face in order to make highly efficient solar cells. There are certainly more exciting and impor tant problems to be asked, answered, and tackled. If we can target these fundamental issues other than making materi als by trial-and-error, we will not only succeed in making high perfor mance photovoltaic materials, but place us on a sound scientific base for unlimited possibilities. To investigate the structur e-property relationships a nd explore high performance photovoltaic materials have been our focus thr oughout this dissertati on. Chapter 3 addressed whether triplet excitons can be generated and ha rvested in the blends of low bandgap platinumacetylide polymers and PCBM. Our results revealed that triplet excited excitons cannot be harvested when triplet excited states are below charge separa tion states. Although platinumacetylide polymers in our hands did not provided the expected photovoltaic properties, this study could become interesting and important if this system (platinum-acetylide polymers/PCBM) can
244 be used to probe where charge transfer states an d/or charge separation stat es are located relative to the LUMO level of PCBM. It is well-known that charge separati on states are closely related to the origin of open circuit voltage in the bulk-he terojucntion solar cells. Currently, there are no readily available approaches to directly measure th ese states. If we are able to make a wide range of variable bandgap platinum acetylide polymers with similar HOMO levels and progressively changed LUMO levels, we should be able to indi rectly estimate charge separation states relative to their singlet and triplet excited states. In other words, we can engineer the LUMO levels of the polymers through a donor-acceptor fashion where selected electron-def icient acceptors are used, and then determine whether triplet excito ns can be generated and harvested. Owing to the existence of a rich library of electron-deficient acceptors, a precise contro l of LUMO levels of the resulting platinum-acetylide polymers can be expected. Therefore, estimation of these excited states can be made with a great accuracy. Chapter 4 focused on the development of a general approach to prepare vinylenecontaining donor-acceptor type low bandgap conjugated polymers. The performance of these vinylene-linked donor-acceptor conjugated polymers in BHJ solar cells were largely limited by their low charge mobilities. However, the scope of this approach can be extended. For instance, we can apply this method to make light-emitting polymers, since our approach can provide structurally defect-free vinylenecontaining polymers. It is well unde rstood that structural defects often act as charge traps and recombination cent ers in light-emitting devices. Using this approach to make near-IR light-emitting polymers is undoubted ly worthy of our immediate investigation. In addition, this approach can be certainly ap plied to make other al ternating polymers (e.g. polyene containing conjugated polymers), even t hough our study up to date has only dealt with donor-acceptor type conjugated polymers.
245 DPP-based conjugated materials are genera lly high performance photovoltaic materials. Chapter 5 reported two types of unconventional DPP-based -conjugated materials, namely amphiphilic DPP-based oligothiophenes and ther mocleavable DPP-based polymers. Amphiphilic molecular design fulfilled our expectation to obtain self-assembled highly ordered nanostrucutres. Both field-effect transistors and solar cell devices based on amphiphilic DPPbased oligothiophenes exhibited highly remarkable properties. With cont inuously fine tuning the chemical structures of these molecules (e.g. adjusting the lengths of hydrophobic and hydrophiphilic chains, and changing the rigid cores), enhanced properties can be conceivably obtained. We should further inve stigate these structurally we ll-defined DPP-based discrete oligomers. From the aspect of device fabr ication, we should deve lop suitable processing protocols for these unconventiona l materials. For instance, mi xed solvents with different polarities can be used to control the materials mo rphologies. The photovoltaic performance of thermocleavable polymer PDPP-Bocs was remarkable with power conversion efficien cies up to 1.5%, in spite of its poorly matched HOMO energy levels relative to PCBM. Consider ing the potential benefits of th eir use in solution-processed bilayer and tandem devices, as well as their remark able thermal and morphological stabilities, thermocleavable DPP-based polymer s with properly tuned energy le vels and gaps are certainly of great interest as active materials in achievi ng high performance solar cells. For example, the alternating dithienosilole (DTS) and DPP-Bo c polymer can be a good starting point. In chapter 6, isoindigo was first intro duced as an electronacceptor moiety in -conjugated materials. With the emphasis on the understanding of structure-property relationships, we have synthesized and studied a wide ra nge of small molecules and polym ers, such as IsoI-O/N/S and PIsoIAM-1/2/3. In the study of IsoI-O/N/S, we showed that little changes in chemical structures
246 can lead to huge differences in materi al properties, and the study of PIsoIAM-1/2/3 exhibited the power of the molecular design in c ontrolling material properties. Power conversion efficiencies up to ~2% demo nstrated the remarkab le applicability of isoindigo-based small molecules in bulk heteroju nction solar cells. Of im portance and urgency in the future study is, however, to develop purely n-type isoindigo-based conjugated polymers. With an established facile synthetic route for borylated isoindigo buildin g blocks, a variety of homoor alternating polymers composed of el ectron-deficient repeat units can be readily prepared via Suzuki polymerization under mild conditions opening up many possibilities. These polymers can play an important role in a number of electronic devices, such as all-polymer solar cells and ambipolar and n-type fi eld-effect transistors. In comb ination with p-dopable conjugated polymers, these n-dopable isoindigo-based polymers can be possibly used in the realization of complementary multi-colored electrochromic devices. Clearly, designing materials for organic photovoltaics (electroni cs) involves delicate balances and a comprehensive understanding of all of the processes. With a firm understanding of -conjugated materials and device operation mechan isms at fundamental levels, we anticipate the era of organic electronics is co ming in the foreseeable future.
247 LIST OF REFERENCES (1) Grimsdale, A. C.; Leok Chan, K.; Martin, R. E.; Jokisz, P. G.; Holmes, A. B. Chem. Rev. 2009 109 897. (2) Cheng, Y.-J.; Yang, S.-H.; Hsu, C.-S. Chem. Rev. 2009 109 5868. (3) Murphy, A. R.; Frchet, J. M. J. Chem. Rev. 2007 107 1066. (4) Samuel, I. D. W.; Turnbull, G. A. Chem. Rev. 2007 107 1272. (5) Beaujuge, P. M.; Reynolds, J. R. Chem. Rev. 2010 110 268. (6) Gelinck, G. H.; Huitema, H. E. A.; van V eenendaal, E.; Cantatore, E.; Schrijnemakers, L.; van der Putten, J. B. P. H.; Geuns, T. C. T.; Beenhakkers, M.; Giesbers, J. B.; Huisman, B.-H.; Meijer, E. J.; Benito, E. M.; Touwslager, F. J.; Marsman, A. W.; van Rens, B. J. E.; de Leeuw, D. M. Nat. Mater. 2004 3 106. (7) Kelley, T. W.; Baude, P. F.; Gerlach, C.; Ende r, D. E.; Muyres, D.; Haase, M. A.; Vogel, D. E.; Theiss, S. D. Chem. Mater. 2004 16 4413. (8) Hines, D. R.; Ballarotto, V. W.; W illiams, E. D.; Shao, Y.; Solin, S. A. J. Appl. Phys. 2007 101 024503. (9) Eder, F.; Klauk, H.; Halik, M.; Zs chieschang, U.; Schmid, G.; Dehm, C. Appl. Phys. Lett. 2004 84 2673. (10) Blanchet, G. B.; Loo, Y.-L.; Rogers, J. A.; Gao, F.; Fincher, C. R. Appl. Phys. Lett. 2003 82 463. (11) Chiang, C. K.; Fincher, C. R.; Park, Y. W. ; Heeger, A. J.; Shirakawa, H.; Louis, E. J.; Gau, S. C.; MacDiarmid, A. G. Phys. Rev. Lett. 1977 39 1098. (12) Chien, J. C. W.; Karasz, F. E.; Shimamura, K. Macromolecules 1982 15 1012. (13) Roth, S.; Filzmoser, M. Adv. Mater. 1990 2 356. (14) Brdas, J.-L.; Cornil, J.; Heeger, A. J. Adv. Mater. 1996 8 447. (15) Walzer, K.; Maennig, B.; Pfeiffer, M.; Leo, K. Chem. Rev. 2007 107 1233. (16) Maennig, B.; Pfeiffer, M.; No llau, A.; Zhou, X.; Leo, K.; Simon, P. Phys. Rev. B 2001 64 195208. (17) Novikov, S. V.; Dunlap, D. H.; Kenkre V. M.; Parris, P. E.; Vannikov, A. V. Phys. Rev. Lett. 1998 81 4472. (18) Tessler, N.; Preezant, Y.; Rappaport, N.; Roichman, Y. Adv. Mater. 2009 21 2741.
248 (19) Bredas, J. L.; Street, G. B. Acc. Chem. Res. 1985 18 309. (20) Reeves, B. D.; Unur, E.; Ananthakrishnan, N.; Reynolds, J. R. Macromolecules 2007 40 5344. (21) Takahashi, A.; Fukutome, H. Solid State Commun. 1987 62 279. (22) Roth, S. In Field Theoretical Tools for Polymer and Particle Physics ; Springer Berlin 1998; Vol. 508, p 240. (23) Su, W. P.; Schrieffer, J. R.; Heeger, A. J. Phys. Rev. Lett. 1979 42 1698. (24) Gershenson, M. E.; Podzorov, V.; Morpurgo, A. F. Rev. Mod. Phys. 2006 78 973. (25) Druger, S. D.; Ratner, M. A.; Nitzan, A. Mol. Cryst. Liq. Cryst. 1990 190 171 (26) Brdas, J.-L.; Norton, J. E.; Cornil, J.; Coropceanu, V. Acc. Chem. Res. 2009 42 1691. (27) Cornil, J.; Beljonne, D.; Calbert, J.-P.; Brdas, J.-L. Adv. Mater. 2001 13 1053. (28) Tiwari, S.; Greenham, N. Opt. Quant. Electronics. 2009 41 69. (29) Yang, X.; Wang, L.; Wang, C.; Long, W.; Shuai, Z. Chem. Mater. 2008 20 3205. (30) Coehoorn, R.; Pasveer, W. F.; Bobbert, P. A.; Michels, M. A. J. Phys. Rev. B 2005 72 155206. (31) Podzorov, V.; Menard, E.; Borissov, A.; Ki ryukhin, V.; Rogers, J. A.; Gershenson, M. E. Phys. Rev. Lett. 2004 93 086602. (32) Novikov, S. V.; Vannikov, A. V. J. Phys. Chem. C 2009 113 2532. (33) Schmechel, R.; Seggern, H. v. physica status solidi (a) 2004 201 1215. (34) Sheraw, C. D.; Jackson, T. N.; Eaton, D. L.; Anthony, J. E. Adv. Mater. 2003 15 2009. (35) Katz, H. E.; Laquindanum, J. G.; Lovinger, A. J. Chem. Mater. 1998 10 633. (36) Heeney, M.; Bailey, C.; Genevicius, K.; Shkunov, M.; Sparrowe, D.; Tierney, S.; McCulloch, I. J. Am. Chem. Soc. 2005 127 1078. (37) Yamamoto, T.; Kokubo, H.; Kobashi, M.; Sakai, Y. Chem. Mater. 2004 16 4616. (38) Coppo, P.; Cupertino, D. C.; Yeates, S. G.; Turner, M. L. Macromolecules 2003 36 2705. (39) Letizia, J. A.; Salata, M. R.; Tribout, C. M.; Facchetti, A. ; Ratner, M. A.; Marks, T. J. J. Am. Chem. Soc. 2008 130 9679.
249 (40) Zhan, X.; Tan, Z. a.; Domercq, B.; An, Z.; Zhang, X.; Barlow, S.; Li, Y.; Zhu, D.; Kippelen, B.; Marder, S. R. J. Am. Chem. Soc. 2007 129 7246. (41) Bijleveld, J. C.; Zoombelt, A. P.; Mathij ssen, S. G. J.; Wienk, M. M.; Turbiez, M.; de Leeuw, D. M.; Janssen, R. A. J. J. Am. Chem. Soc. 2009 131 16616. (42) Kim, F. S.; Guo, X.; Watson, M. D.; Jenekhe, S. A. Adv. Mater. 2010 22 478. (43) Usta, H.; Risko, C.; Wang, Z.; Huang, H.; Deliomeroglu, M. K.; Zhukhovitskiy, A.; Facchetti, A.; Marks, T. J. J. Am. Chem. Soc. 2009 131 5586. (44) Steckler, T. T.; Zhang, X.; Hwang, J.; Honeya ger, R.; Ohira, S.; Zhang, X.-H.; Grant, A.; Ellinger, S.; Odom, S. A.; Sweat D.; Tanner, D. B.; Rinzler, A. G.; Barlow, S.; Brdas, J.-L.; Kippelen, B.; Marder, S. R.; Reynolds, J. R. J. Am. Chem. Soc. 2009 131 2824. (45) Schidleja, M.; Melzer, C.; Seggern, H. v. Adv. Mater. 2009 21 1172. (46) Brgi, L.; Turbiez, M.; Pfeiffer, R.; Bi enewald, F.; Kirner, H.-J.; Winnewisser, C. Adv. Mater. 2008 20 2217. (47) Zaumseil, J.; Sirringhaus, H. Chem. Rev. 2007 107 1296. (48) Cornil, J.; Brdas, J.-L.; Zaumseil, J.; Sirringhaus, H. Adv. Mater. 2007 19 1791. (49) Sirringhaus, H.; Wilson, R. J.; Friend, R. H.; Inbasekaran, M.; Wu, W.; Woo, E. P.; Grell, M.; Bradley, D. D. C. Appl. Phys. Lett. 2000 77 406. (50) Coropceanu, V.; Cornil, J.; da Silva Filho, D. A.; Olivier, Y.; Silbey, R.; Brdas, J.-L. Chem. Rev. 2007 107 926. (51) Soci, C.; Hwang, I.-W.; Moses, D.; Zhu, Z. ; Waller, D.; Gaudiana, R.; Brabec, C. J.; Heeger, A. J. Adv. Funct. Mater. 2007 17 632. (52) Roncali, J. Chem. Rev. 1997 97 173. (53) Balaban, A. T.; Oniciu, D. C.; Katritzky, A. R. Chem. Rev. 2004 104 2777. (54) Brocks, G.; Tol, A. J. Phys. Chem. 1996 100 1838. (55) Wudl, F.; Kobayashi, M.; Heeger, A. J. J. Org. Chem. 1984 49 3382. (56) Bundgaard, E.; Krebs, F. C. Macromolecules 2006 39 2823. (57) Bagnis, D.; Beverina, L.; Huang, H.; Silvestri, F.; Yao, Y.; Yan, H.; Pagani, G. A.; Marks, T. J.; Facchetti, A. J. Am. Chem. Soc. 2010 132 4074. (58) Blouin, N.; Michaud, A.; Gendron, D.; Wakim, S.; Blair, E.; Neagu-Plesu, R.; Bellette, M.; Durocher, G.; Tao, Y.; Leclerc, M. J. Am. Chem. Soc. 2007 130 732.
250 (59) Hou, J.; Chen, H.-Y.; Zhang, S.; Li, G.; Yang, Y. J. Am. Chem. Soc. 2008 130 16144. (60) Zou, Y.; Gendron, D.; Neagu-Plesu, R.; Leclerc, M. Macromolecules 2009 42 6361. (61) Zou, Y.; Najari, A.; Berrouard, P.; Beaupr, S.; Rda Ai ch, B.; Tao, Y.; Leclerc, M. J. Am. Chem. Soc. 2010 132 5330. (62) Santamouris, M.; Argiriou, A.; Vallindras, M. Sol. Energy 1994 52 371. (63) Green, M. A. Physica. E 2002 14 65. (64) King, R. R.; Law, D. C.; Edmondson, K. M.; Fetzer, C. M.; Kinsey, G. S.; Yoon, H.; Sherif, R. A.; Karam, N. H. Appl. Phys. Lett. 2007 90 183516. (65) Wu, J.; Walukiewicz, W.; Yu, K. M.; Shan, W.; Ager Iii, J. W.; Haller, E. E.; Lu, H.; Schaff, W. J.; Metzger, W. K.; Kurtz, S. J. Appl. Phys. 2003 94 6477. (66) Takamoto, T.; Ikeda, E.; Kurita, H.; Ohmori, M. Appl. Phys. Lett. 1997 70 381. (67) Tang, C. W. Appl. Phys. Lett. 1986 48 183. (68) Peumans, P.; Forrest, S. R. Appl. Phys. Lett. 2001 79 126. (69) Jenekhe, S. A.; Yi, S. Appl. Phys. Lett. 2000 77 2635. (70) Stevens, D. M.; Qin, Y.; Hillmyer, M. A.; Frisbie, C. D. J. Phys. Chem. C 2009 113 11408. (71) Lloyd, M. T.; Mayer, A. C.; Subramanian, S. ; Mourey, D. A.; Herman, D. J.; Bapat, A. V.; Anthony, J. E.; Malliaras, G. G. J. Am. Chem. Soc. 2007 129 9144. (72) Matsuo, Y.; Sato, Y.; Niinomi, T.; Soga, I.; Tanaka, H.; Nakamura, E. J. Am. Chem. Soc. 2009 131 16048. (73) Mikroyannidis, J. A.; Sharma, S. S.; Vijay, Y. K.; Sharma, G. D. ACS Appl. Mater. Interfaces 2009 2 270. (74) Roncali, J.; Frre, P.; Blanchard, P.; de Bettignies, R.; Turbiez, M.; Roquet, S.; Leriche, P.; Nicolas, Y. Thin Solid Films 2006 511-512 567. (75) Silvestri, F.; Irwin, M. D.; Beverina, L. ; Facchetti, A.; Pagani, G. A.; Marks, T. J. J. Am. Chem. Soc. 2008 130 17640. (76) Tamayo, A. B.; Tantiwiwat M.; Walker, B.; Nguyen, T.-Q. J. Phys. Chem. C 2008 112 15543. (77) Tamayo, A. B.; Walker, B.; Nguyen, T.-Q. J. Phys. Chem. C 2008 112 11545.
251 (78) Tantiwiwat, M.; Tamayo, A.; Luu, N.; Dang, X.-D.; Nguyen, T.-Q. J. Phys. Chem. C 2008 112 17402. (79) Velusamy, M.; Huang, J.-H.; Hsu, Y.-C.; Chou, H.-H.; Ho, K.-C.; Wu, P.-L.; Chang, W.H.; Lin, J. T.; Chu, C.-W. Org. Lett. 2009 11 4898. (80) Walker, B.; Tamayo, A. B.; Dang, X.-D.; Zalar, P.; Seo, J. H.; Garcia, A.; Tantiwiwat, M.; Nguyen, T.-Q. Adv. Funct. Mater. 2009 19 3063. (81) Winzenberg, K. N.; Kemppinen, P.; Fanchini, G.; Bown, M.; Collis, G. E.; Forsyth, C. M.; Hegedus, K.; Singh, T. B.; Watkins, S. E. Chem. Mater. 2009 21 5701. (82) Zhang, J.; Yang, Y.; He, C.; He, Y.; Zhao, G.; Li, Y. Macromolecules 2009 42 7619. (83) Tamayo, A. B.; Walker, B.; Nguyen, T.-Q. J. Phys. Chem. C 2008 112 11545. (84) Woo, C. H.; Holcombe, T. W.; Unruh, D. A.; Sellinger, A.; Frchet, J. M. J. Chem. Mater. 2010 22 1673. (85) Alam, M. M.; Jenekhe, S. A. Chem. Mater. 2004 16 4647. (86) Brabec, C. J.; Cravino, A.; Meissner, D.; Sa riciftci, N. S.; Fromherz, T.; Rispens, M. T.; Sanchez, L.; Hummelen, J. C. Adv. Mater. Funct 2001, 11, 374. (87) Christopher, R. M.; Agnese, A.; Jana, Z. ; Richard, W.; Mary, J. M.; Jeremy, H. B.; Jonathan, J. M. H.; Neil, C. G.; Richard, H. F. Appl. Phys. Lett. 2007 90 193506. (88) Granstrom, M.; Petritsch, K.; Arias, A. C.; Lux, A.; Andersson, M. R.; Friend, R. H. Nature 1998 395 257. (89) Guo, X.; Watson, M. D. Org. Lett. 2008 10 5333. (90) Holcombe, T. W.; Woo, C. H.; Kavulak, D. F. J.; Thompson, B. C.; Frechet, J. M. J. J. Am. Chem. Soc. 2009 131 14160. (91) Kietzke, T.; Egbe, D. A. M.; Horhold, H.-H.; Neher, D. Macromolecules 2006 39 4018. (92) Kietzke, T.; Hor hold, H.-H.; Neher, D. Chem. Mater. 2005 17 6532. (93) Kietzke, T.; Shin, R. Y. C.; Egbe, D. A. M.; Chen, Z.-K.; Sellinger, A. Macromolecules 2007 40 4424. (94) Kim, G. J.; Wichard, J. D. B.; Yuri, Z.; Arkady, Y.; Mats, A.; Tonu, P.; Villy, S.; Appl. Phys. Lett. 2004, 121, 12613. (95) Kim, J.-S.; Ho, P. K. H.; Murphy, C. E.; Friend, R. H. Macromolecules 2004 37 2861. (96) Kim, Y.; Cook, S.; Choulis, S. A.; Nels on, J.; Durrant, J. R.; Bradley, D. D. C. Chem. Mater. 2004 16 4812.
252 (97) Marc, M. K.; Jorgen, S.; Kornel, T. H.; Herman, F. M. S.; Sjoerd, C. V.; Jan, M. K.; Xiaoniu, Y.; Joachim, L. Appl. Phys. Lett. 2006 88 083504. (98) Marsh, R. A.; McNeill, C. R.; Abrusci, A.; Campbell, A. R.; Friend, R. H. Nano Lett. 2008 8 1393. (99) Martijn, M. W.; Kroon, J., M. ; Verhees, W. J. H.; Joop, K.; Hummelen, J. C.; van Hal, P., A. ; Janssen, R. A. J. Angew. Chem. Int. Ed. 2003 42 3371. (100) McNeill, C. R.; Abrusci, A.; Zaumseil, J.; Wilson, R.; McKiernan, M. J.; Burroughes, J. H.; Halls, J. J. M.; Greenham, N. C.; Friend, R. H. Appl. Phys. Lett. 2007 90 193506. (101) McNeill, C. R.; Greenham, N. C. Adv. Mater. 2009 21 3840. (102) McNeill, C. R.; Halls, J. J. M.; Wilson, R. ; Whiting, G. L.; Berkebile, S.; Ramsey, M. G.; Friend, R. H.; Greenham, N. C. Adv. Funct. Mater. 2008 18 2309. (103) McNeill, C. R.; Westenhoff, S.; Grov es, C.; Friend, R. H.; Greenham, N. C. J. Phys. Chem. C 2007 111 19153. (104) Sang, G.; Zhou, E.; Huang, Y.; Zou, Y.; Zhao, G.; Li, Y. J. Phys. Chem. C 2009 113 5879. (105) Sang, G.; Zou, Y.; Huang, Y.; Zhao, G.; Yang, Y.; Li, Y. Appl. Phys. Lett. 2009 94 193302. (106) Tan, Z. a.; Zhou, E.; Zhan, X.; Wang, X.; Li, Y.; Barlow, S.; Marder, S. R. Appl. Phys. Lett. 2008 93 073309. (107) Zhou, Y.; Pei, J.; Dong, Q.; Sun, X.; Liu, Y.; Tian, W. J. Phys. Chem. C 2009 113 7882. (108) Alivisatos, A. P. Science 1996 271 933. (109) Bai, Y.; Cao, Y.; Zhang, J.; Wang, M.; Li, R.; Wang, P.; Zakeeruddin, S. M.; Gratzel, M. Nat. Mater. 2008 7 626. (110) Black, C. T.; Murray, C. B.; Sandstrom, R. L.; Sun, S. Science 2000 290 1131. (111) Dayal, S.; Kopidakis, N.; Olson, D. C.; Ginley, D. S.; Rumbles, G. Nano Lett. 2009 10 239. (112) Greenham, N. C.; Peng, X.; Alivisatos, A. P. Phys. Rev. B 1996 54 17628. (113) Landi, B. J.; Castro, S. L.; Ruf, H. J.; Evans, C. M.; Bailey, S. G.; Raffaelle, R. P. Sol. Energy Mater. Sol. Cells 2005 87 733. (114) Lin, Y.-Y.; Chu, T.-H.; Li, S.-S.; Chuang, C.-H.; Chang, C.-H.; Su, W.-F.; Chang, C.-P.; Chu, M.-W.; Chen, C.-W. J. Am. Chem. Soc. 2009 131 3644.
253 (115) Oosterhout, S. D.; Wienk, M. M.; van Bavel, S. S.; Thiedmann, R.; Jan Anton Koster, L.; Gilot, J.; Loos, J.; Schmidt, V.; Janssen, R. A. J. Nat. Mater. 2009 8 818. (116) O'Regan, B.; Gratzel, M. Nature 1991 353 737. (117) Sun, B.; Snaith, H. J.; Dhoot, A. S.; Westenhoff, S.; Greenham, N. C. J. Appl. Phys. 2005 97 014914. (118) Wang, P.; Zakeeruddin, S. M.; Moser, J. E.; Nazeeruddin, M. K.; Sekiguchi, T.; Gratzel, M. Nat. Mater. 2003 2 402. (119) Weintraub, B.; Wei, Y.; Wang, Zhong L. Angew. Chem. Int. Ed. 2009 48 8981. (120) Zhou, Y.; Riehle, F. S.; Yuan, Y.; Schlei ermacher, H.-F.; Niggemann, M.; Urban, G. A.; Kruger, M. Appl. Phys. Lett. 2010 96 013304. (121) Zotti, G.; Vercelli, B.; Berlin, A.; Pasini M.; Nelson, T. L.; McCullough, R. D.; Virgili, T. Chem. Mater. 2010 22 1521. (122) Peumans, P.; Yakimov, A.; Forrest, S. R. J. Appl. Phys. 2003 93 3693. (123) Veldman, D.; Meskers, S. C. J.; Janssen, R. A. J. Adv. Funct. Mater. 2009 19 1939. (124) Sariciftci, N. S.; Smilowitz, L.; Heeger, A. J.; Wudl, F. Science 1992 258 1474. (125) Yu, G.; Gao, J.; Hummelen, J. C.; Wudl, F.; Heeger, A. J. Science 1995 270 1789. (126) Kim, J. Y.; Kim, S. H.; Lee, H.-H.; Lee, K.; Ma, W.; Gong, X.; Heeger, A. J. Adv. Mater. 2006 18 572. (127) Nguyen, L. H.; Hoppe, H.; Erb, T.; Gne s, S.; Gobsch, G.; Sariciftci, N. S. Adv. Funct. Mater. 2007, 17 1071. (128) Moul, A. J.; Meerholz, K. Adv. Mater. 2008 20 240. (129) Li, G.; Shrotriya, V.; Huang, J.; Yao, Y.; Moriarty, T.; Emery, K.; Yang, Y. Nat. Mater. 2005 4 864. (130) Ma, W.; Yang, C.; Gong, X.; Lee, K.; Heeger, A. J. Adv. Funct. Mater. 2005 15 1617. (131) Yao, Y.; Hou, J.; Xu, Z.; Li, G.; Yang, Y. Adv. Funct. Mater. 2008 18 1783. (132) Peet, J.; Kim, J. Y.; Coates, N. E.; Ma, W. L.; Moses, D.; Heeger, A. J.; Bazan, G. C. Nat. Mater. 2007 6 497. (133) Scharber, M. C.; Mhlbacher, D.; Koppe M.; Denk, P.; Waldauf, C.; Heeger, A. J.; Brabec, C. J. Adv. Mater. 2006 18 789. (134) Thompson, B. C.; Kim, Y.-G.; Reynolds, J. R. Macromolecules 2005 38 5359.
254 (135) Mhlbacher, D.; Scharber, M.; Morana, M. ; Zhu, Z.; Waller, D.; Gaudiana, R.; Brabec, C. Adv. Mater. 2006 18 2884. (136) Lee, J. K.; Ma, W. L.; Brabec, C. J.; Yue n, J.; Moon, J. S.; Kim, J. Y.; Lee, K.; Bazan, G. C.; Heeger, A. J. J. Am. Chem. Soc. 2008 130 3619. (137) Hoven, C. V.; Dang, X.-D.; Coffin, R. C.; Peet, J.; Nguyen, T.-Q.; Bazan, G. C. Adv. Mater. 2010 22 E63. (138) Wang, E.; Wang, L.; Lan, L.; Luo, C.; Zhuang, W.; Peng, J.; Cao, Y. Appl. Phys. Lett. 2008 92 033307. (139) Baek, N. S.; Hau, S. K.; Yip, H.-L.; Acton, O.; Chen, K.-S.; Jen, A. K. Y. Chem. Mater. 2008 20 5734. (140) Wong, W.-Y.; Wang, X.-Z.; He, Z.; Djurisic, A. B.; Yip, C.-T.; Cheung, K.-Y.; Wang, H.; Mak, C. S. K.; Chan, W.-K. Nat. Mater. 2007 6 521. (141) Gilot, J.; Wienk, M. M.; Janssen, R. A. J. Nat. Mater. 2007 6 704. (142) Huo, L.; Hou, J.; Zhang, S.; Chen, H.-Y.; Yang, Y. Angew. Chem. Int. Ed. 2010 49 1500. (143) Chen, C.-P.; Chan, S.-H.; Chao, T.-C.; Ting, C.; Ko, B.-T. J. Am. Chem. Soc. 2008 130 12828. (144) Chen, H.-Y.; Hou, J.; Zhang, S.; Liang, Y.; Yang, G.; Yang, Y.; Yu, L.; Wu, Y.; Li, G. Nat Photon 2009 3 649. (145) Liang, Y.; Wu, Y.; Feng, D.; Tsai S.-T.; Son, H.-J.; Li, G.; Yu, L. J. Am. Chem. Soc. 2008 131 56. (146) Liang, Y.; Xu, Z.; Xia, J.; Tsai, S.-T.; Wu, Y.; Li, G.; Ray, C.; Yu, L. Adv. Mater. 2010 9999 (147) Halls, J. J. M.; Walsh, C. A.; Greenham, N. C.; Marseglia, E. A.; Friend, R. H.; Moratti, S. C.; Holmes, A. B. Nature 1995 376 498. (148) Pincus, P. J. Chem. Phys. 1981 75 1996. (149) Walheim, S.; Boltau, M.; Mlyne k, J.; Krausch, G.; Steiner, U. Macromolecules 1997 30 4995. (150) Yan, H.; Chen, Z.; Zheng, Y.; Newman, C.; Qu inn, J. R.; Dotz, F.; Kastler, M.; Facchetti, A. Nature 2009 457 679. (151) Chen, Z.; Zheng, Y.; Yan, H.; Facchetti, A. J. Am. Chem. Soc. 2008 131 8.
255 (152) Sun, X.; Zhou, Y.; Wu, W.; Liu, Y.; Tia n, W.; Yu, G.; Qiu, W.; Chen, S.; Zhu, D. The J. Phys. Chem. B 2006 110 7702. (153) Xue, J.; Uchida, S.; Rand, B. P.; Forrest, S. R. Appl. Phys. Lett. 2004 85 5757. (154) Kim, J. Y.; Lee, K.; Coates, N. E.; Moses, D.; Nguyen, T.-Q.; Dante, M.; Heeger, A. J. Science 2007 317 222. (155) Sista, S.; Park, M.-H.; Hong, Z.; Wu Y.; Hou, J.; Kwan, W. L.; Li, G.; Yang, Y. Adv. Mater. 2010 22 380. (156) Dennler, G.; Scharber, M. C.; Ameri, T. ; Denk, P.; Forberich, K.; Waldauf, C.; Brabec, C. J. Adv. Mater. 2008 20 579. (157) Thompson, B. C.; Kim, Y. G.; McCarley, T. D.; Reynolds, J. R. J. Am. Chem. Soc. 2006 128, 12714. (158) Farley, R. T. Ph. D. Dissert ation, University of Florida, 2008. (159) Brabec, C. J.; Saricift ci, N. S.; Hummelen, J. C. Adv. Funct. Mater. 2001 11 15. (160) Brabec, C. J. Sol. Energy Mater. Sol. Cells 2004 83 273. (161) Scharber, M. C.; Wuhlbacher, D.; Koppe M.; Denk, P.; Waldauf, C.; Heeger, A. J.; Brabec, C. L. Adv. Mater. 2006 18 789. (162) Sariciftci, N. S.; Smilowitz, L.; Heeger, A. J.; Wudl, F. Science 1992 258 1474. (163) Brabec, C. J.; Zerza, G.; Cerullo, G.; De Silvestri, S.; Luzzati, S.; Hummelen, J. C.; Sariciftci, S. Chem. Phys. Lett. 2001 340 232. (164) Reyes-Reyes, M.; Kim, K.; Carroll, D. L. Appl. Phys. Lett. 2005 87 083506. (165) Li, G.; Shrotriya, V.; Huang, J. S.; Yao, Y.; Moriarty, T.; Emery, K.; Yang, Y. Nat. Mater. 2005 4 864. (166) Kim, Y.; Cook, S.; Tuladhar, S. M.; Choulis S. A.; Nelson, J.; Durrant, J. R.; Bradley, D. D. C.; Giles, M.; Mcculloch, I.; Ha, C. S.; Ree, M. Nat. Mater. 2006 5 197. (167) Kim, J. Y.; Kim, S. H.; Lee, H. H.; Lee, K.; Ma, W. L.; Gong, X.; Heeger, A. J. Adv. Mater. 2006 18 572. (168) Kim, K.; Liu, J.; Namboothiry, M. A. G.; Carroll, D. L. Appl. Phys. Lett. 2007 90 163511. (169) Kim, J. Y.; Lee, K.; Coates, N. E.; Moses, D.; Nguyen, T. Q.; Dante, M.; Heeger, A. J. Science 2007 317 222. (170) Green, M. A.; Emery, K.; Hishikawa, Y.; Warta, W. Prog. Photovoltaics 2008 16 61.
256 (171) Huo, L.; Hou, J.; Zhang, S.; Chen, H.-Y.; Yang, Y. Angew. Chem. Int. Ed. 2010 49 1500. (172) Padinger, F.; Rittberger, R. S.; Sariciftci, N. S. Adv. Funct. Mater. 2003 13 85. (173) Guo, F. Q.; Kim, Y. G.; Re ynolds, J. R.; Schanze, K. S. Chem. Commun. 2006 1887. (174) Guo, F.; Ogawa, K.; Kim, Y.-G.; Danil ov, E. O.; Castellano, F. N.; Reynolds, J. R.; Schanze, K. S. Phys. Chem. Chem. Phys. 2007 9 2724. (175) Shao, Y.; Yang, Y. Adv. Mater. 2005 17 2841. (176) Giebink, N. C.; Sun, Y.; Forrest, S. R. Org. Electron. 2006 7 375. (177) Holten, D.; Gouterman, M.; Parson, W. W.; Windsor, M. W.; Rockley, M. G. Photochem. Photobiol. 1976 23 415. (178) Goodenough, J. B. Phys. Rev. 1968 171 466. (179) Beljonne, D.; Wittmann, H. F.; Kohler, A.; Graham, S.; Younus, M.; Lewis, J.; Raithby, P. R.; Khan, M. S.; Friend, R. H.; Bredas, J. L. J. Chem. Phys. 1996 105 3868. (180) Wilson, J. S.; Chawdhury, N.; Al-Mandhar y, M. R. A.; Younus, M.; Khan, M. S.; Raithby, P. R.; Kohler, A.; Friend, R. H. J. Am. Chem. Soc. 2001 123 9412. (181) Kohler, A.; Wilson, J. S.; Friend, R. H.; Al-Suti, M. K.; Khan, M. S.; Gerhard, A.; Bassler, H. J. Chem. Phys. 2002 116 9457. (182) Silverman, E. E.; Cardolaccia, T.; Zhao, X. M.; Kim, K. Y.; Haskins-Glusac, K.; Schanze, K. S. Coord. Chem. Rev. 2005 249 1491. (183) Glusac, K.; Kose, M. E.; Jiang, H.; Schanze, K. S. J. Phys. Chem. B 2007 111 929. (184) Wong, W. Y. Dalton Trans. 2007 4495. (185) Kohler, A.; Wittmann, H. F.; Frie nd, R. H.; Khan, M. S.; Lewis, J. Synth. Met. 1996 77 147. (186) Zhang, Q. T.; Tour, J. M. J. Am. Chem. Soc. 1998 120 5355. (187) van Mullekom, H. A. M.; Vekema ns, J. A. J. M.; Meijer, E. W. Chem.-Eur. J. 1998 4 1235. (188) Dhanabalan, A.; van Duren, J. K. J.; van Ha l, P. A.; van Dongen, J. L. J.; Janssen, R. A. J. Adv. Funct. Mater. 2001 11 255. (189) Thomas, C. A.; Zong, K. W.; Abboud, K. A.; Steel, P. J.; Reynolds, J. R. J. Am. Chem. Soc. 2004 126 16440.
257 (190) Winder, C.; Sariciftci, N. S. J. Mater. Chem. 2004 14 1077. (191) Steckler, T. T.; Abboud, K. A.; Craps, M.; Rinzler, A. G.; Reynolds, J. R. Chem. Commun. 2007 4904. (192) Blouin, N.; Michaud, A.; Leclerc, M. Adv. Mater. 2007 19 2295. (193) Blouin, N.; Michaud, A.; Gendron, D.; Wakim, S.; Blair, E.; Neagu-Plesu, R.; Belletete, M.; Durocher, G.; Tao, Y.; Leclerc, M. J. Am. Chem. Soc. 2008 130 732. (194) Wong, W.-Y.; Wang, X.-Z.; He, Z.; Djurisic, A. B.; Yip, C.-T.; Cheung, K.-Y.; Wang, H.; Mak, C. S. K.; Chan, W.-K. Nat. Mater. 2007 6 521. (195) Wong, W. Y.; Wang, X. Z.; He, Z.; Chan, K. K.; Djurisic, A. B.; Cheung, K. Y.; Yip, C. T.; Ng, A. M. C.; Xi, Y. Y.; Mak, C. S. K.; Chan, W. K. J. Am. Chem. Soc. 2007 129 14372. (196) Wong, W. Y.; Wang, X. Z.; He, Z.; Djurisic A. B.; Yip, C. T.; Cheung, K. Y.; Wang, H.; Mak, C. S. K.; Chan, W. K. Nat. Mater. 2007 6 704. (197) Gilot, J.; Wienk, M. M.; Janssen, R. A. J. Nat. Mater. 2007 6 704. (198) Wu, P.-T.; Bull, T.; Kim, F. S.; Luscombe, C. K.; Jenekhe, S. A. Macromolecules 2009 42 671. (199) Wyatt, M. F.; Stein, B. K.; Brenton, A. G. Anal. Chem. 2006 78 199. (200) Frapper, G.; Kertesz, M. Inorg. Chem. 1993 32 732. (201) Parker, C. A.; Rees, W. T. Analyst 1960 85 587. (202) Rogers, J. E.; Cooper, T. M.; Fleitz P. A.; Glass, D. J.; McLean, D. G. J. Phys. Chem. A 2002, 106 10108. (203) Datta, K.; Banerjee, M.; Seal, B. K.; Mukherjee, A. K. J. Chem. Soc. Perkin Trans. 2 2000 531. (204) Bhattacharya, S.; Nayak, S. K.; Chatt opadhyay, S.; Banerjee, M.; Mukherjee, A. K. J. Phys. Chem. A 2001 105 9865. (205) Karikomi, M.; Kitamura, C.; Tanaka, S.; Yamashita, Y. J. Am. Chem. Soc. 1995 117 6791. (206) Blanchard, P.; Raimundo, J. M.; Roncali, J. Synth. Met. 2001 119 527. (207) Bao, Z.; Dodabalapur, A.; Lovinger, A. J. Appl. Phys. Lett. 1996 69 4108. (208) Coakley, K. M.; McGehee, M. D. Chem. Mater. 2004 16 4533.
258 (209) Valaski, R.; Canestraro, C. D.; Mi caroni, L.; Mello, R. M. Q.; Roman, L. S. Sol. Energy Mater. Sol. Cells 2007 91 684. (210) Oevering, H.; Paddonrow, M. N.; Heppener, M.; Oliver, A. M.; Co tsaris, E.; Verhoeven, J. W.; Hush, N. S. J. Am. Chem. Soc. 1987 109 3258. (211) Chen, P. Y.; Meyer, T. J. Chem. Rev. 1998 98 1439. (212) Chawdhury, N.; Kohler, A.; Friend, R. H.; Wong, W. Y.; Lewis, J.; Younus, M.; Raithby, P. R.; Corcoran, T. C.; Al-Mandhary, M. R. A.; Khan, M. S. J. Chem. Phys. 1999 110 4963. (213) Wilson, J. S.; Kohler, A.; Friend, R. H.; Al -Suti, M. K.; Al-Mandhary, M. R. A.; Khan, M. S.; Raithby, P. R. J. Chem. Phys. 2000 113 7627. (214) Ohkita, H.; Cook, S.; Astuti, Y.; Duffy, W.; Heeney, M.; Tierney, S.; McCulloch, I.; Bradley, D. D. C.; Durrant, J. R. Chem. Commun. 2006 3939. (215) Ford, T. A.; Avilov, I.; Beljonne, D.; Greenham, N. C. Phys. Rev. B 2005 71 125212. (216) Yersin, H. Top. Curr. Chem. 2004 241 1. (217) Kooistra, F. B.; Knol, J. ; Kastenberg, F.; Popescu, L. M.; Verhees, W. J. H.; Kroon, J. M.; Hummelen, J. C. Org. Lett. 2007 9 551. (218) Hoppe, H.; Arnold, N.; Sariciftci, N. S.; Meissner, D. Sol. Energy Mater. Sol. Cells 2003 80 105. (219) Baek, N. S.; Hau, S. K.; Yip, H.-L.; Acton, O.; Chen, K.-S.; Jen, A. K. Y. Chem. Mater. 2008 20 5734. (220) Bundgaard, E.; Krebs, F. C. Sol. Energy Mater. Sol. Cells 2007 91 954. (221) Kulkarni, A. P.; Zhu, Y.; Jenekhe, S. A. Macromolecules 2008 41 339. (222) Gunbas, G. E.; Durmus, A.; Toppare, L. Adv. Mater. 2008 20 691. (223) Beaujuge, P. M.; Ellinger, S.; Reynolds, J. R. Adv. Mater. 2008 20 2772. (224) Becker, H.; Spreitzer, H. ; Ibrom, K.; Kreuder, W. Macromolecules 1999 32 4925. (225) Jin, S. H.; Park, H. J.; Kim, J. Y.; Lee, K.; Lee, S. P.; Moon, D. K.; Lee, H. J.; Gal, Y. S. Macromolecules 2002 35 7532. (226) Scherf, U. In Carbon Rich Compounds II, Macrocyclic Oligoacetylenes and Other Linearly Conjugated Systems 1999, p163. (227) Loewe, R. S.; McCullough, R. D. Chem. Mater. 2000 12 3214.
259 (228) Kim, J. Y.; Qin, Y.; Stevens, D. M.; Ugurlu, O.; Kalihari, V.; Hillmyer, M. A.; Frisbie, C. D. J. Phys. Chem. C 2009 113 10790. (229) Kitamura, C.; Tanaka, S.; Yamashita, Y. Chem. Mater. 1996 8 570. (230) van Mullekom, H. A. M.; Vekema ns, J. A. J. M.; Meijer, E. W. Chem. Eur. J. 1998 4 1235. (231) Zhu, Y.; Champion, R. D.; Jenekhe, S. A. Macromolecules 2006 39 8712. (232) Feng, X.; Liu, M.; Pisula, W.; Takase, M.; Li, J.; Mllen, K. Adv. Mater. 2008 20, 2684. (233) Hou, J.; Tan, Z.; He, Y.; Yang, C.; Li, Y. Macromolecules 2006 39 4657. (234) Jung, I.; Lee, T.; Kang, S. O.; Ko, J. Synthesis 2005 986. (235) Lee, S. J.; Gray, K. C.; Paek, J. S.; Burke, M. D. J. Am. Chem. Soc. 2008 130 466. (236) Roush, W. R.; Brown, B. B. J. Am. Chem. Soc. 1993 115 2268. (237) Katayama, H.; Nagao, M.; Nishimura, T.; Matsui, Y.; Umeda, K.; Akamatsu, K.; Tsuruoka, T.; Nawafune, H.; Ozawa, F. J. Am. Chem. Soc. 2005 127 4350. (238) Katayama, H.; Nagao, M.; Nishimura, T.; Matsui, Y.; Fukuse, Y.; Wakioka, M.; Ozawa, F. Macromolecules 2006 39 2039. (239) Koch, F.; Heitz, W. Macromol. Chem. and Phys. 1997 198 1531. (240) Thompson, B. C.; Madrigal, L. G.; Pinto, M. R.; Kang, T.-S.; Schanze, K. S.; Reynolds, J. R. J. Polym. Sci. Part A: Polym. Chem 2005 43 1417. (241) Fu, D.; Xu, B.; Swager, T. M. Tetrahedron 1997 53 15487. (242) Du, J.; Fang, Q.; Chen, X.; Ren, S.; Cao, A.; Xu, B. Polymer 2005 46 11927. (243) Grisorio, R.; Mastrorilli, P.; Nobile, C. F.; Romanazzi, G.; Suranna, G. P.; Gigli, G.; Piliego, C.; Ciccarella, G.; Cosma, P.; Acierno, D.; Amendola, E. Macromolecules 2007 40 4865. (244) Beletskaya, I. P.; Cheprakov, A. V. Chem. Rev. 2000 100 3009. (245) Johansson, D. M.; Wang, X.; Johansson, T.; Inganas, O.; Yu, G.; Srdanov, G.; Andersson, M. R. Macromolecules 2002 35 4997. (246) Ishiyama, T.; Takagi, J.; Hartwig, J. F.; Miyaura, N. Angew. Chem. Int. Ed. 2002 41 3056. (247) Jaya Prakash, D.; Sujit, R. J. Org. Chem. 2002 67 7861.
260 (248) Perzon, E.; Zhang, F.; Andersson, M.; Ma mmo, W.; Ingans, O.; Andersson, M. R. Adv. Mater. 2007 19 3308. (249) Pan, M.; Bao, Z.; Yu, L. Macromolecules 1995 28 5151. (250) Jayakannan, M.; van Hal, P. A.; Janssen, R. A. J. J. Poly. Sci. A: Poly. Chem. 2002 40 251. (251) Bundgaard, E.; Krebs, F. C. Macromolecules 2006 39 2823. (252) Chynwat, V.; Frank, H. A. Chem. Phys. 1995 194 237. (253) Englman, R.; Jortner, J. Molecular Physics: An Internat ional Journal at the Interface Between Chemistry and Physics 1970 18 145 (254) Thompson, B. C.; Kim, Y. G.; McCarley, T. D.; Reynolds, J. R. J. Am. Chem. Soc. 2006, 128, 12714. (255) Wienk, M. M.; Turbiez, M. G. R.; Stru ijk, M. P.; Fonrodona, M.; Janssen, R. A. J. Appl. Phys. Lett. 2006 88 153511. (256) Wang, X.; Perzon, E.; Mammo, W.; Oswald, F. ; Admassie, S.; Persson, N.-K.; Langa, F.; Andersson, M. R.; Ingans, O. Thin Solid Films 2006 511-512 576. (257) Hagemann, O.; Jrgensen, M.; Krebs, F. C. J. Org. Chem. 2006 71 5546. (258) Roncali, J. Acc. Chem. Res. 2009 42 1719. (259) Qin, R.; Li, W.; Li, C.; Du, C.; Veit, C.; Schleiermacher, H.-F.; Andersson, M.; Bo, Z.; Liu, Z.; Inganas, O.; Wuerfel, U.; Zhang, F. J. Am. Chem. Soc. 2009 131 14612. (260) Hou, J.; Chen, H.-Y.; Zhang, S.; Chen, R. I.; Yang, Y.; Wu, Y.; Li, G. J. Am. Chem. Soc. 2009 131 15586. (261) Lunt, R. R.; Benziger, J. B.; Forrest, S. R. Adv. Mater. 2009 22 1233. (262) Yang, H.; Shin, T. J.; Yang, L.; Cho, K.; Ryu, C. Y.; Bao, Z. Adv. Funct. Mater. 2005 15 671. (263) 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. (264) An, Z.; Yu, J.; Jones, S. C.; Barlow, S.; Y oo, S.; Domercq, B.; Prins, P.; Siebbeles, L. D. A.; Kippelen, B.; Marder, S. R. Adv. Mater. 2005 17 2580. (265) Stefan, M. L.; Sven, H.; Arnaud, C.; Mukundan, T.; Georg, K. Angew. Chem. Int. Ed. 2006 45 3364. (266) Sommer, M.; Lang, Andreas S.; Thelakkat, M. Angew. Chem. Int. Ed. 2008 47 7901.
261 (267) Carrasco-Orozco, M.; Tsoi, W. C.; O'Neill, M.; Aldred, M. P.; Vlachos, P.; Kelly, S. M. Adv. Mater. 2006 18 1754. (268) Yagai, S.; Kubota, S.; Saito, H.; Unoike K.; Karatsu, T.; Kitamura, A.; Ajayaghosh, A.; Kanesato, M.; Kikkawa, Y. J. Am. Chem. Soc. 2009 131 5408. (269) Percec, V.; Imam, M. R.; Peterca, M.; Wilson, D. A.; Graf, R.; Spiess, H. W.; Balagurusamy, V. S. K.; Heiney, P. A. J. Am. Chem. Soc. 2009 131 7662. (270) Feng, X.; Pisula, W.; Kudernac, T.; Wu, D.; Zhi, L.; De Feyter, S.; Mu llen, K. J. Am. Chem. Soc. 2009 131 4439. (271) Jin, W.; Yamamoto, Y.; Fukushima, T.; Ishii, N.; Kim, J.; Kato, K.; Takata, M.; Aida, T. J. Am. Chem. Soc. 2008 130 9434. (272) Hill, J. P.; Jin, W.; Kosaka, A.; Fukushi ma, T.; Ichihara, H.; Shimomura, T.; Ito, K.; Hashizume, T.; Ishii, N.; Aida, T. Science 2004 304 1481. (273) Vijayakumar, C.; Praveen, V. K.; Ajayaghosh, A. Adv. Mater. 2009 21 2059. (274) Zhang, X.; Chen, Z.; Wrthner, F. J. Am. Chem. Soc. 2007 129 4886. (275) Moon, K.-S.; Kim, H.-J.; Lee, E.; Lee, M. Angew. Chem. Int. Ed. 2007 46 6807. (276) Li, J. L.; Kastler, M.; Pisula, W.; Roberts on, J. W. F.; Wasserfallen, D.; Grimsdale, A. C.; Wu, J. S.; Mllen, K. Adv. Funct. Mater. 2007 17 2528. (277) Yamamoto, Y.; Fukushima, T.; Suna, Y.; Ishii, N.; Saeki, A.; Seki, S.; Tagawa, S.; Taniguchi, M.; Kawai, T.; Aida, T. Science 2006 314 1761. (278) Schmidt-Mende, L.; Fechtenkotter, A.; Mullen, K.; Moons, E.; Friend, R. H.; MacKenzie, J. D. Science 2001 293 1119. (279) Li, W.-S.; Yamamoto, Y.; Fukushima, T.; S aeki, A.; Seki, S.; Tagawa, S.; Masunaga, H.; Sasaki, S.; Takata, M.; Aida, T. J. Am. Chem. Soc. 2008 130 8886. (280) Chambon, S.; Manceau, M.; Firon, M.; Cros, S.; Rivaton, A.; Gardette, J.-L. Polymer 2008 49 3288. (281) Manceau, M.; Rivaton, A.; Gardette, J.-L. Macromol. Rapid Commun. 2008 29 1823. (282) Lincke, G. Dyes Pigm. 2003 59 1. (283) Yu, C.-Y.; Chen, C.-P.; Ch an, S.-H.; Hwang, G.-W.; Ting, C. Chem. Mater. 2009 21 3262. (284) Huo, L.; Hou, J.; Chen, H.-Y.; Zhang, S.; Jiang, Y.; Chen, T. L.; Yang, Y. Macromolecules 2009 42 6564.
262 (285) Zhou, E.; Yamakawa, S.; Tajima, K.; Yang, C.; Hashimoto, K. Chem. Mater. 2009 21 4055. (286) Chan, W. K.; Chen, Y.; Peng, Z.; Yu, L. J. Am. Chem. Soc. 1993 115 11735. (287) Song, B.; Wang, Z.; Chen, S.; Zhang, X.; Fu, Y.; Smet, M.; Dehaen, W. Angew. Chem. 2005 117 4809. (288) Song, B.; Wei, H.; Wang, Z.; Zhang, X.; Smet, M.; Dehaen, W. Adv. Mater. 2007 19 416. (289) Rochat, A. C.; Cassar, L.; Iqbal, I. US 4,579,949 1986 (290) Balakrishnan, K.; Datar, A.; O itker, R.; Chen, H.; Zuo, J.; Zang, L. J. Am. Chem. Soc. 2005 127 10496. (291) Bushey, M. L.; Hwang, A.; Stephens, P. W.; Nuckolls, C. J. Am. Chem. Soc. 2001 123 8157. (292) Pisula, W.; Kastler, M.; Wasserfallen, D.; Mondeshki, M.; Piris, J. ; Schnell, I.; Mllen, K. Chem. Mater. 2006 18 3634. (293) Simmerer, J.; Glsen, B.; Paulus, W.; Ke ttner, A.; Schuhmacher, P.; Adam, D.; Etzbach, K.-H.; Siemensmeyer, K.; Wendorff, J. H.; Ringsdorf, H.; Haarer, D. Adv. Mater. 1996 8 815. (294) Marrocchi, A.; Seri, M.; Kim, C.; Facchetti, A.; Taticchi, A.; Marks, T. J. Chem. Mater. 2009 21 2592. (295) Tamayo, A. B.; Dang, X. D.; Walker B.; Seo, J.; Kent, T.; Nguyen, T. Q. Appl. Phys. Lett. 2009 94 (296) Lincker, F.; Bourgun, P.; Masson, P.; Didier P.; Guidoni, L.; Bigot, J.-Y.; Nicoud, J.-F.; Donnio, B.; Guillon, D. Org. Lett. 2005 7 1505. (297) Ishiyama, T.; Takagi, J.; Ishida, K.; Mi yaura, N.; Anastasi, N. R.; Hartwig, J. F. J. Am. Chem. Soc. 2001 124 390. (298) Reddy, M. S.; Narender, M.; Nageswar, Y. V. D.; Rao, K. R. Synthesis 2005 2005 714. (299) Aoki, K. i.; Kudo, M.; Tamaoki, N. Org. Lett. 2004 6 4009. (300) Egbe, D. A. M.; Ulbricht, C.; Orgis, T.; Carbonnier, B.; Kietzke, T.; Peip, M.; Metzner, M.; Gericke, M.; Birckner, E.; Pakul a, T.; Neher, D.; Grummt, U.-W. Chem. Mater. 2005 17 6022. (301) Wunderlich, B. J. Chem. Phys. 1958 29 1395. (302) Chen, C.; Yu, P. H. F.; Cheung, M. K. J. Appl. Polym. Sci. 2005 98 736.
263 (303) Helgesen, M.; Krebs, F. C. Macromolecules 2010 43 1253. (304) Petersen, M. H.; Gevorgyan, S. A.; Krebs, F. C. Macromolecules 2008 41 8986. (305) Zambounis, J. S.; Hao, Z.; Iqbal, A. Nature 1997 388 131. (306) Han, X.; Chen, X.; Holdcroft, S. Chem. Mater. 2009 21 4631. (307) Helgesen, M.; Gevorgyan, S. A.; Krebs, F. C.; Janssen, R. A. J. Chem. Mater. 2009 21 4669. (308) Tromholt, T.; Gevorgyan, S. A.; Jrgensen, M.; Krebs, F. C.; Sylvester-Hvid, K. O. ACS Appl. Mater. Interfaces 2009 1 2768. (309) Liu, J.; Kadnikova, E. N.; Liu, Y.; McGehee, M. D.; Frchet, J. M. J. J. Am. Chem. Soc. 2004 126 9486. (310) Kim, J. S.; Lee, Y.; Lee, J. H.; Park, J. H.; Kim, J. K.; Cho, K. Adv. Mater. 2010 22 1355. (311) Wei, G.; Wang, S.; Renshaw, K.; Thompson, M. E.; Forrest, S. R. ACS Nano 2010 4 1927. (312) Zhang, Y.; Hau, S. K.; Yip, H.-L.; Sun, Y.; Acton, O.; Jen, A. K. Y. Chem. Mater. 2010 22 2696. (313) Mei, J.; Ogawa, K.; Kim, Y.-G.; Heston, N. C.; Arenas, D. J.; Nasrollahi, Z.; McCarley, T. D.; Tanner, D. B.; Reynolds, J. R.; Schanze, K. S. ACS Appl. Mater. Interfaces 2009 1 150. (314) Silva, J. F. M. d.; Garden, S. J.; Pinto, A. C. J. Braz. Chem. Soc., 2001 12 273. (315) Bogdanov, A.; Mironov, V.; Musi n, L.; Buzykin, B.; Konovalov, A. Russ. J. Gen. Chem. 2008 78 1977. (316) Minami, T.; Matsumoto, M.; Agawa, T. J. Chem. Soc., Chem. Commun. 1976 1053. (317) Mei, J.; Graham, K. R.; Stalder, R.; Reynolds, J. R. Org. Lett. 2010 12 660. (318) Moul, A. J.; Meerholz, K. Adv. Funct. Mater. 2009 19 3028. (319) Ichino, Y.; Ni, J. P.; Ueda, Y.; Wang, D. K. Synth. Met. 2001 116 223. (320) Kooistra, F. B.; Mihailetchi, V. D.; Popescu, L. M.; Kronholm, D.; Blom, P. W. M.; Hummelen, J. C. Chem. Mater. 2006 18 3068. (321) Sirringhaus, H.; Brown, P. J.; Friend, R. H.; Nielsen, M. M.; Bechgaard, K.; LangeveldVoss, B. M. W.; Spiering, A. J. H.; Janssen, R. A. J.; Meijer, E. W.; Herwig, P.; de Leeuw, D. M. Nature 1999 401 685.
264 (322) Tang, C.; Bang, J.; E. Stein, G.; Fredrickson, G. H.; Hawker, C. J.; Kramer, E. J.; Sprung, M.; Wang, J. Macromolecules 2008 41 4328. (323) Zhang, M.; Yang, L.; Yurt, S.; Misner M. J.; Chen, J.-T.; Coughlin, E. B.; Venkataraman, D.; Russell, T. P. Adv. Mater. 2007 19 1571. (324) Kim, S. H.; Misner, M. J.; Russell, T. P. Adv. Mater. 2004 16 2119. (325) Beaujuge, P. M.; Pisula, W.; Tsao, H. N.; Ellinger, S.; Mu llen, K.; Reynolds, J. R. J. Am. Chem. Soc. 2009 131 7514. (326) Izuhara, D.; Swager, T. M. J. Am. Chem. Soc. 2009 131 17724. (327) Wen, Y.; Liu, Y. Adv. Mater. 2010 22 1331. (328) Park, S. H.; Roy, A.; Beaupre, S.; Cho, S.; Coates, N.; Moon, J. S.; Moses, D.; Leclerc, M.; Lee, K.; Heeger, A. J. Nat Photon 2009 3 297. (329) Cates, N. C.; Gysel, R.; Dahl, J. E. P.; Sellinger, A.; McGehee, M. D. Chem. Mater. 2010 (330) Guo, X.; Kim, F. S.; Jenekhe, S. A.; Watson, M. D. J. Am. Chem. Soc. 2009 131 7206. (331) Piliego, C.; Jarzab, D.; Gigli, G.; Chen, Z.; Facchetti, A.; Loi, M. A. Adv. Mater. 2009 21 1573. (332) Usta, H.; Facchetti, A.; Marks, T. J. Org. Lett. 2008 10 1385. (333) Ishiyama, T.; Murata, M.; Miyaura, N. J. Org. Chem. 1995 60 7508.
265 BIOGRAPHICAL SKETCH Jianguo Mei, a native of China, received his co llege education from the Hefei University of Technology, and graduated with a bachelors degree in engineering. He continued his postgraduate study there, worked on heterogeneous catalysis under the guidance of Professor Shaoming Yu, and obtained a masters degree in chemical engineering. After spending seven years in chemical engineering, he decided to pursue his caree r in science and engineering, attended the graduate school at the University of New Orleans, and started his practice in homogeneous catalysis under the guidance of Profe ssor Steven P. Nolan. For family reasons, he moved to the University of Florida one year late r, which coincidently saved him being a victim of Hurricane Katrina. He joined the research group of Professor John R. Reynolds shortly after his arrival, and began his adventure in the field of organic pi-c onjugated materials for optoelectronic applicatio ns. With the support from his a dvisor and colleagues, he exposed himself to many aspects of resear ch in the Reynolds group, gained expertise in design, Synthesis and characterization of solution-processable or ganic electroactive mate rials for solar cells, electrochromics, field-effect transistors and li ght-emitting diodes. He was also briefly involved into the aspect of device fabr ication through closely working w ith his collaborators. Jianguo earned his doctoral degree from the University of Florida in the summer of 2010. He will work with Professor Zhenan Bao as a postdoctoral fell ow in Department of Ch emical Engineering at Stanford University this fall.