Tunable light-harvesting polymers containing embedded dipolar chromophores for polymer solar cell applications

Article abstract of DOI:10.1002/pola.25902

A series of light-harvesting conjugated polymers were designed and synthesized for polymer solar cells. These newly designed polymers comprise an unusual two-dimensional conjugated structure with an electron-rich thiophene-triphenylamine backbone and stable planar indacenodithiophene π-bridges terminated with tunable electron acceptors. It was found that the electron-withdrawing strength of the acceptor could be used to manipulate the energy level of the lowest unoccupied molecular orbital and bandgap (as much as 0.3 eV), generating derivatives with complementary absorbance in the visible spectrum. This approach provides great flexibility in fine tuning the electronic and optical properties of the resultant polymers and facilitates the investigation of how these chemical modifications alter the subsequent photovoltaic properties of these materials.

Full text of DOI:10.1002/pola.25902

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
Tunable Light-Harvesting Polymers Containing Embedded Dipolar
Chromophores for Polymer Solar Cell Applications
David F. Zeigler,1 Kung-Shih Chen,2 Hin-Lap Yip,2 Yong Zhang,2 Alex K.-Y. Jen1,2
1Department of Chemistry, University of Washington, Seattle, Washington 98195-1700
2Department of Materials Science and Engineering, University of Washington, Seattle, Washington 98195-2120
Correspondence to: A. K.-Y. Jen (E-mail: ajen@u.washington.edu)
Received 17 October 2011; accepted 14 December 2011; published online 4 January 2012
DOI: 10.1002/pola.25902
ABSTRACT: A series of light-harvesting conjugated polymers
were designed and synthesized for polymer solar cells. These
newly designed polymers comprise an unusual two-dimen-
sional conjugated structure with an electron-rich thiophene–
triphenylamine backbone and stable planar indacenodithio-
phene p-bridges terminated with tunable electron acceptors.
It was found that the electron-withdrawing strength of the
acceptor could be used to manipulate the energy level of the
lowest unoccupied molecular orbital and bandgap (as much
as 0.3 eV), generating derivatives with complementary ab-
sorbance in the visible spectrum. This approach provides
great flexibility in fine tuning the electronic and optical prop-
erties of the resultant polymers and facilitates the investiga-
tion of how these chemical modifications alter the
C
V
subsequent photovoltaic properties of these materials.
2012
Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 50:
1362–1373, 2012
KEYWORDS: bulk-heterojunction solar cells; charge transfer;
conjugated polymers; D-p-A chromophore; low bandgap poly-
mers; molecular modeling; organic solar cells; NLO; postpoly-
merization modification
INTRODUCTION Solar energy is the largest source of renew-
able energy in the world.1 Solar cells can convert this energy
to usable electricity, but commonly used inorganic solar cells
are prohibitively expensive for widespread commercialization.2
Thus, over the past decade, polymer solar cells (PSCs) based
on the bulk heterojunction (BHJ) structure have emerged as a
promising alternative to inorganic cells because of their poten-
tial for low-cost solution processability, light weight, flexibility,
and large-scale printing.3,4 A significant amount of research
has focused on the molecular engineering of linear donor–
acceptor (D–A) polymers to tune their bandgap (Eg) and
energy levels to enhance PSC performance. Typically, these
polymers consist of alternating electron-rich and electron-poor
monomers. The characteristics of the monomers dictate the
energy levels of the highest occupied molecular orbital
(HOMO) and the lowest unoccupied molecular orbital (LUMO),
and the bandgap of the polymer.5 Although this method can
theoretically be used to optimize the electronic attributes of
light-harvesting polymers, it is difficult to systematically fine
tune their energy levels because it requires parallel syntheses
of modified donor and/or acceptor monomers and polymer-
ization reactions. This leads to inconsistent electronic proper-
ties across different polymer batches.6,7
cations.7–14 The D–p-bridge–A motif is common to many sec-
ond-order nonlinear optical (NLO) chromophores. This
architecture exploits well-established knowledge of the
structure/property relationships and charge transfer in NLO
dyes, which can greatly decrease their bandgap.15–17 Unlike
common D–A-based semiconducting polymers, the acceptors
of these polymers terminate the electron-poor side chains,
which are connected through a p-bridge to an electron-rich
conjugated backbone.7–14
Such tunable D–p-bridge–A polymers generally consist of an
alternating D1–D2 backbone, with one of the donors con-
nected to the electron acceptor through a p-bridge to form a
push–pull chromophore incorporated into the polymer back-
bone. The p-bridge is terminated with a synthetic handle
that can subsequently be functionalized with a range of
acceptors after polymerization.7–14 Postfunctionalization rep-
resents a flexible and generally applicable method to easily
tune the electronic and optical properties of the polymers
and to probe structure/property relationships of the poly-
mers. To date, most D–p-bridge–A polymers have utilized tri-
phenylamine–styrylthiophene or thienylene–vinylene-based
chromophores using indenofluorene,7 fluorene,8 silafluor-
ene,9 cyclopentadithiophene,10 and carbazole11–14 donors,
though these materials have very similar properties and
weaknesses.
To address this issue, several groups have explored the de-
velopment of tunable D–p-bridge–A polymers for PSC appli-
C
V
2012 Wiley Periodicals, Inc.
1362
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1362–1373

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
p-bridge. We theorized that inclusion of thiophene in the
polymer backbone would minimize steric repulsion, thereby
increasing the planarity of the polymer backbone and the
relative intensity of the ICT absorbance. In addition, recently
reported conjugated copolymers based on the IDT donor
have exhibited high PCE and mobilities (10ꢀ1 to 10ꢀ2 cm2
Vꢀ1 ꢀ1) despite its bulky hexylbenzyl side chains, which
s
hamper intermolecular p–p stacking.24–28 This suggests that
the high hole mobilities of IDT-based D–A copolymers is the
result of efficient intramolecular hole transport along their
rigid, fused backbone. Thus, we hypothesized that incorpo-
rating the fused IDT unit as the polymer p-bridge could pro-
vide adequate carrier mobility and charge transfer proper-
ties, while also enhancing the polymer solubility through the
peripheral hexylbenzyl groups.
FIGURE 1 Structure of PThTPA-IDT polymer functionalized
with DCN, TBA, and DCNIO acceptors.
As predicted, the PThTPA-IDT polymers have a more intense
ICT absorbance peak than p–p* transition peak and have
higher absorbance coefficients than those reported for other
D–p-bridge–A polymers. Also, IDT p-bridge imbued the poly-
mers with enhanced solubility and processability. However,
when relatively weak acceptors were used, the PThTPA-IDT
polymers showed weaker ICT characteristics compared with
polymers utilizing a vinylene p-bridge. This weaker ICT is
due to the large stabilization energy of the fused IDT p-
bridge, which hinders effective charge transfer in the side
chain. Interestingly, stronger acceptors facilitated efficient
charge transfer and led to better ICT characteristics. Despite
the high-hole mobility of IDT and TPA units, these copoly-
mers showed moderate mobilities (10ꢀ3 to 10ꢀ4 cm2 Vꢀ1
Importantly, these polymers consistently show weak internal
charge transfer (ICT) transition absorbance compared with
linear D–A polymers, whose absorbance profile is typically
dominated by their long wavelength ICT peak. In general,
this is the result of poor spatial overlap between the HOMO
orbitals on the polymer backbone and the LUMO orbitals on
the side-chain acceptor. Limited electronic communication
between the electron-rich backbone and pendant acceptors
leads to weak ICT oscillator strength, thereby limiting the in-
tensity of the ICT absorbance.18 This poor electronic commu-
nication could potentially be due to interrupted conjugation
along the polymer backbone or between the backbone and
side chain. For example, the phenyl–phenyl backbone link-
ages of such polymers are inherently problematic because
these bonds experience severe twisting due to the steric hin-
drance between phenyl protons.19 In addition, benzene has a
relatively high-aromatic stabilization energy and hence, has
poor effective conjugation with its chemical environment.20
Conversely, in the case of thienylene–vinylene side-chain
polymers, it is possible that steric repulsion between the
vinylene protons and backbone aromatic protons push the
side chain out of plane with the backbone.21 In both cases,
the electron-rich backbone is electronically isolated from
the electron-poor side chain, which leads to a low-wave-
length p–p* transition absorbance that is significantly more
intense than the longer wavelength ICT peak.7–14 The result-
ant absorbance profile overlaps poorly with the solar spec-
trum, which decreases PSC performance.22,23 Moreover, most
reported side-chain polymers for PSCs suffer from poor solu-
bility in common organic solvents due to inadequate solubi-
lizing groups on fluorene and carbazole moieties.7 This poor
solubility complicates device processing and results in thin
active layers with low optical densities. One recent report of
an indenofluorene-based side-chain polymer exhibited
improved processability, but still displayed a very intense
low wavelength p–p* transition.7
s
ꢀ1) as measured by the organic field effect transistor
(OFET) method. In addition, the BHJ active layer exhibited
severe phase segregation for two of the three polymers. The
resultant PSCs exhibited a depressed Jsc and FF, likely due to
increased recombination rates caused by poor morphology
and mobility. Although the PSCs have limited efficiencies,
this approach provides a good system to study the effects
of the fine tuning of the electronic and optical properties of
D–p-bridge–A polymers on PSC performance.
EXPERIMENTAL
Materials
Unless otherwise specified, all chemicals were purchased
from Aldrich or TCI and used without further purification.
3-Bromo-4-iodothiophene,29,30 N,N-diphenyl-4-(trimethylstan-
nyl)aniline,31,32 indacenodithiophene,33 and 2,5-bis(trimethyl-
stannyl)thiophene34 were prepared as previously reported.
Solvents for synthesis were purified by distillation. All chem-
ical reactions were carried out in a nitrogen atmosphere.
Synthesis of Compound 1
IDT (1.51 g, 1.7 mmol) was dissolved in anhydrous DMF (40
ꢁ
mL) and the solution was cooled to 0 C. POCl3 (0.292 g, 1.9
mmol) was added to this dropwise. The solution was gradu-
ally warmed to room temperature, heated to 50 ꢁC, and
allowed to stir for 8 h. Then, saturated NaC2H3O2 (60 mL)
was added and the solution was stirred for a further 30 min.
The mixture was extracted with dichloromethane and the
combined organic layers were washed repeatedly with
To further investigate the structure–property relationships of
D–p-bridge–A polymers, a new class of D–p-bridge–A poly-
mer [PThTPA-indacenoditihophene (IDT)] was synthesized
and functionalized with three acceptors (Fig. 1). These poly-
mers consist of a thiophene–TPA backbone and a novel IDT
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1362–1373
1363

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
saturated NaCl solution. Then, the organic fraction was dried
with Na2SO4, concentrated in vacuo, and the crude product
was purified by column chromatography (1:1 dichlorometha-
ne:hexane) to yield compound 1, a yellow solid (0.99 g).
Yield: 64%.
calcd for C83H87NOS2: C, 84.57; H, 7.44. Found: C, 84.52; H,
7.40.
Synthesis of Compound 6
Compound 5 (0.810 g, 0.7 mmol) was dissolved in THF (50
ꢁ
mL) and cooled to 0 C. NBS (0.269 g, 1.5 mmol) was added
1H NMR (300 MHz, CDCl3, ppm): 9.93 (s, 1H), 7.77 (s, 1H),
7.69 (s, 1H), 7.59 (s, 1H), 7.43 (d, J ¼ 6 Hz, 1H), 7.29 (m,
11H), 7.21 (d, J ¼ 3 Hz, 2H), 7.18 (d, J ¼ 3 Hz, 2H), 7.14 (d,
J ¼ 6 Hz, 2H), 2.69 (t, J ¼ 6 Hz, 8H), 1.70 (m, 8H), 1.42 (m,
25H), 0.99 (m, 12H). 13C NMR (125 MHz, CDCl3, ppm):
183.02, 158.13, 154.81, 142.79, 142.15, 141.79, 135.91,
128.31, 128.25, 127.28, 123.90, 118.03, 62.87, 35.49, 31.78,
31.59, 29.50, 22.88, 14.33. HRMS calcd for C65H74OS2 [M þ
H]þ: 935.4192; found, 935.4190. Anal. calcd for C65H74OS2:
C, 83.53; H, 7.98. Found: C, 83.49; H, 7.96.
in one portion, the solution was allowed to gradually warm
to room temperature and stirred overnight. The solution was
poured into H2O and extracted with ethyl acetate. The com-
bined organic layers were dried with Na2SO4 and concen-
trated in vacuo to yield compound 6, a bright yellow solid
(0.872 g). Yield: 95%.
1H NMR (300 MHz, CDCl3, ppm): 9.82 (s, 1H), 7.67 (s, 1H),
7.58 (s, 1H), 7.47 (m, 2H), 7.38 (d, J ¼ 9 Hz, 3H), 7.22–7.09
(m, 20H), 7.04 (d, J ¼ 9 Hz, 2 H), 6.97 (d, J ¼ 9 Hz, 3 H),
2.60 (m, 8H), 1.61 (m, 8H), 1.33 (m, 25H), 0.90 (m, 12H).13
C
NMR (125 MHz, CDCl3, ppm): 182.72, 158.11, 155.99,
154.90, 153.56, 151.61, 148.39, 147.32, 147.33, 145.12,
141.94, 141.76, 141.50, 141.07, 138.89, 138.20, 133.11,
129.37, 128.71, 128.69, 128.77, 127.86, 127.74, 126.20,
124.58, 123.60, 123.21, 118.84, 118.23, 117.28, 63.14, 62.88,
35.63, 31.79, 31.33, 29.19, 22.60, 14.13. Anal. calcd for
Synthesis of Compound 2
Compound 1 (0.95 g, 1.0 mmol) was dissolved in tetrahydro-
furan (THF; 50 mL) and cooled to 0 ꢁC. NBS (0.199 g, 1.1
mmol) was added in one portion, the solution was gradually
warmed to room temperature and stirred overnight. The so-
lution was poured into H2O and extracted with ethyl acetate.
The combined organic layers were dried with Na2SO4 and
concentrated in vacuo to yield compound 2, a yellow solid
(0.957 g). Yield: 93%.
C83H85Br2NOS2: C, 74.76; H, 6.43. Found: C, 84.36; H, 6.41.
Synthesis of PThTPA-IDT-CHO
Compound 12 (0.351 g, 0.26 mmol) and 2,5-bis(trimethyl-
stannyl)thiophene (0.108 g, 0.26 mmol) were added to an
oven-dried flask and degassed three times. Pd(PPh3)4 (7 mg)
was added to the flask and subsequently anhydrous toluene
(4 mL) and DMF (0.5 mL) were added. The solution was
1H NMR (300 MHz, CDCl3, ppm): 9.93 (s, 1H), 7.77 (s, 1H),
7.69 (s, 1H), 7.59 (s, 1H), 7.29–7.13 (m, 17H), 2.69 (t, J ¼ 6
Hz, 8H), 1.70 (m, 8H), 1.42 (m, 24H), 0.99 (t, J ¼ 3 Hz,
12H).13C NMR (125 MHz, CDCl3, ppm): 183.10, 158.08,
154.83, 142.84, 142.19, 141.80, 135.91, 128.29, 128.20,
127.29, 123.87, 117.99, 62.88, 35.49, 31.78, 31.59, 29.50,
22.88, 14.33. Anal. calcd for C65H73BrOS2: C, 77.11; H, 7.26.
Found: C, 77.08; H, 7.25.
ꢁ
stirred at 110 C for 36 h. Bromobenzene (0.2 mL, 2 mmol)
was added and the solution was stirred at 110 ꢁC for 12 h.
Trimethyl(phenyl)tin (0.35 mL, 2 mmol) was added and the
ꢁ
solution was stirred at 110 C for another 12 h. The solution
was then cooled to rt, poured into MeOH (100 mL), the pre-
cipitate filtered through a Soxhlet thimble and purified by
Soxhlet extraction for 12 h with acetone, 12 h with metha-
nol, 12 h with hexanes and collected with CHCl3. The CHCl3
solution was concentrated and precipitated into MeOH. Fil-
tration yielded PThTPA-IDT-CHO, a red solid (228 mg). Yield:
69%.
Synthesis of Compound 5
Compound 2 (0.911 g, 0.9 mmol) was added to an oven-
dried flask and degassed three times. Pd2(dba)3 (4 mg) and
P(o-tol)3 (5 mg) were added to the flask and subsequently 4
(0.441 g, 0.1 mmol) dissolved in toluene (5 mL) was added,
ꢁ
the solution was cooled to ꢀ78 C and degassed three more
ꢁ
times. The solution was stirred at 100 C for 24 h. Then, the
1H NMR (300 MHz, CDCl3, ppm): 9.79 (s, 1H), 7.63 (s, 1H),
7.52–7.37 (br, 10 H), 7.22–7.06 (br, 23H), 2.56 (m, 8H), 1.59
(m, 8H), 1.28 (m, 24H), 0.86 (m, 12H). GPC (THF, polysty-
rene standard): Mw ¼ 43.1 ꢂ 103 g/mol; PDI ¼ 1.94.
solution was concentrated, dissolved in dichloromethane and
extracted with brine. The combined organic layers were
dried with Na2SO4 and concentrated in vacuo. The crude
product was purified by column chromatography (1:1
dichloromethane:hexane) to yield compound 5, a bright yel-
low solid (0.826 g). Yield: 78%.
Synthesis of PThTPA-IDT-DCN
PThTPA-IDT-CHO (50 mg) was dissolved in THF (20 mL) and
ethanol was added until the initial appearance of precipitate.
Then, malononitrile (90 mg) was added. After 10 min, pyri-
dine (0.05 mL) was added. The solution was warmed to
1H NMR (300 MHz, CDCl3, ppm): 9.82 (s, 1H), 7.67 (s, 1H),
7.58 (s, 1H), 7.47 (d, J ¼ 6 Hz, 3H), 7.31–7.06 (m, 29H), 2.58
(t, J ¼ 9 Hz, 8H), 1.62 (m, 8H), 1.33 (m, 25H), 0.90 (m, 12H).
13C NMR (125 MHz, CDCl3, ppm): 182.75, 158.01, 155.96,
154.93, 153.59, 151.64, 148.33, 147.40, 147.39, 145.10,
141.94, 141.72, 141.48, 141.08, 138.82, 138.17, 133.13,
129.37, 128.71, 128.65, 128.74, 127.88, 127.70, 126.21,
124.55, 123.58, 123.20, 118.82, 118.24, 117.27, 63.11, 62.88,
35.63, 31.79, 31.33, 29.19, 22.60, 14.13. HRMS calcd for
ꢁ
50 C and stirred at that temperature for 16 h. The solution
was concentrated in vacuo, dissolved in a small volume of
dichloromethane, precipitated in methanol, and filtered to
yield PThTPA-IDT-dicyanovinyl (DCN) as a deep red solid
(47 mg). Yield: 91%.
1H NMR (300 MHz, CDCl3, ppm): 7.63–7.41 (br, 11H), 7.22–
7.06 (br, 24H), 2.55 (m, 8H), 1.28 (m, 25H), 0.86 (m, 12H).
C
83H87NOS2 [M þ H]þ: 1178.7233; found, 1178.7231. Anal.
1364
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1362–1373

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
GPC (THF, polystyrene standard): Mw ¼ 42.8 ꢂ 103 g/mol;
PDI ¼ 1.90.
AgCl, and Pt mesh were used as working electrode, reference
electrode, and counter electrode, respectively. The differential
scanning calorimetry (DSC) was performed using DSC2010
(TA instruments) under a heating rate of 10 C min and a
nitrogen flow of 50 mL minꢀ1. The AFM images under tap-
ping mode were taken from the actual devices fabricated for
photovoltaic measurement on a Veeco multimode AFM with
a Nanoscope III controller. All spectra were plotted using
OriginPro 8.
Synthesis of PThTPA-IDT-TBA
ꢀ1
ꢁ
PThTPA-IDT-CHO (50 mg) was dissolved in THF (20 mL) and
ethanol was added until the initial appearance of precipitate.
Then, 1,3-diethyl-2-thiobarbituric acid (TBA; 140 mg) was
added. After 10 min, pyridine (0.05 mL) was added. The so-
ꢁ
lution was warmed to 50 C and stirred at that temperature
for 16 h. The solution was concentrated in vacuo, dissolved
in a small volume of dichloromethane, precipitated in metha-
nol and filtered to yield PThTPA-IDT-TBA as a dark solid (53
mg). Yield: 93%.
Fabrication and Characterization of OFET
and PSC Devices
The field effect transistors were fabricated with a bottom-
gate, top-contact configuration. The heavily n-doped silicon
substrates with a 300-nm thick thermally grown SiO2 dielec-
tric (from Montco Silicon Technologies) were first cleaned by
sonication in acetone and isopropanol and exposed to air
plasma. The cleaned substrates were then treated with hex-
amethyldisilazane (HMDS) through vapor phase deposition
in a vacuum oven (200 mTorr, 100 ꢁC, 3 h). Subsequently,
the semiconductor polymer films were spin-coated in a glove
box from their 10 mg mLꢀ1 chloroform:o-dichlorobenzene
(1:1, v/v) solutions, which were stired overnight and filtered
with 0.2-lm PTFE filter. Interdigitated source and drain elec-
trodes (W ¼ 1000 lm, L ¼ 12 lm) were deposited by evap-
orating a 50-nm thick gold film and defined with a shadow
mask. The transfer and output characteristics were measured
in glove box using an Agilent 4155B semiconductor parame-
teranalyzer. The saturation field-effect mobility (l) was cal-
culated from the following equation:
1H NMR (300 MHz, CDCl3, ppm): 8.64 (s, 1H), 7.73–7.65 (m,
3H), 7.58–7.44 (br, 10H), 7.22–7.06 (br, 21H), 4.58 (m, 4H),
2.56 (m, 8H), 1.28 (m, 30H), 0.86 (m, 12H). GPC (THF, poly-
styrene standard): Mw ¼ 42.1 ꢂ 103 g/mol; PDI ¼ 1.87.
Synthesis of ThTPA-IDT-DCNIO
PThTPA-IDT-CHO (50 mg) was dissolved in THF (20 mL) and
ethanol was added until the initial appearance of precipitate.
Then, 3-dicyanomethylene-1-indanone (DCNIO; 135 mg) was
added. After 10 min, pyridine (0.05 mL) was added. The so-
ꢁ
lution was warmed to 50 C and stirred at that temperature
for 16 h. The solution was concentrated in vacuo, dissolved
in a small volume of dichloromethane, precipitated in metha-
nol, and filtered to yield PThTPA-IDT-DCNIO as a dark solid
(51 mg). Yield: 90%.
1H NMR (300 MHz, CDCl3, ppm): 8.87 (d, 1H), 8.65 (d, 1H),
7.88 (m, 2H), 7.72–7.61 (m, 4H), 7.54–7.43 (m, 9H), 7.23–
7.07 (m, 22H), 2.56 (m, 8H), 1.28 (m, 25H), 0.86 (m, 12H).
GPC (THF, polystyrene standard): Mw ¼ 42.2 ꢂ 103 g/mol;
PDI ¼ 1.91.
2
Ids ¼ lðW=2LÞCiðVgs ꢀ VthÞ
where W and L are the channel width and length, respec-
tively. Ci is the capacitance of insulating SiO2 layer per unit,
Vgs and Vth are the gate voltage and the threshold voltage,
respectively. Vth was obtained as the ꢂ intercept of the linear
section of the plot of (Ids)1/2 versus Vgs. The subthreshold
swing was estimated by taking the inverse of the slope of Ids
versus Vgs in the region of exponential current increase.
Quantum Mechanical Calculations
All density functional theory (DFT) calculations were per-
formed using Gaussian 09(A.02)35 employing the hybride
B3LYP36,37 exchange-correlation functional with a split va-
lence 6-31G*38 basis set. Alkyl substituents were replaced by
methyl groups for computational simplicity as their replace-
ment with shorter chains does not significantly affect opti-
mized geometry or predicted energy levels of the polymers.
Fabrication of Photovoltaic and Hole Only Devices
To fabricate conventional configuration solar cells, ITO-
coated glass substrates (15 X/sq.) were first cleaned with
detergent, de-ionized water, acetone, and isopropyl alcohol.
General Measurement and Characterization
UV–vis spectra were tested using a Perkin-Elmer Lambda-9
spectrophotometer. 1H NMR and 13C NMR spectra were col-
lected on a Bruker AV 300 or 500 spectrometers operating
at 300 or 125 MHz in deuterated chloroform solution with
TMS as reference. Elemental analysis was conducted using a
ThermoFisher Scientific Thermo Finnigan EA 1112 Flash Ele-
mental Analyzer. HRMS spectra were recorded on an Applied
Biosystems QTOF QStar XL Mass Spectrometer. Polymer mo-
lecular weights were measured by a Waters 1515 gel perme-
ation chromatograph (GPC) with a refractive index detector
at room temperature (THF as the eluent). Cyclic voltamme-
tries (CVs) of polymer films were conducted on a BAS CV-
50W voltammetric system with a three-electrode cell in ace-
tonitrile with 0.1 M of tetrabutylammonium hexafluorophos-
phate (Bu4NPF6) using a scan rate of 50 mV sꢀ1. ITO, Ag/
R
Subsequently, PEDOT:PSS (BaytronV PVP AI 4083, filtered at
0.45 lm) layer (ꢃ45 nm) was spin-coated (5000 rpm) on
the cꢁleaned ITO-coated glass substrates and then annealed at
120 C for 30 min under ambient conditions. After that, the
substrates were loaded into a nitrogen-filled glove box. Sub-
sequently, the active layer was spin-coated (2000 rpm) onto
the PEDOT:PSS layer from a homogeneous blending solution
of polymer:PC71BM. The solution was prepared by dissolving
the polymer and PC71BM with a particular blending weight
ratio in chloroform:o-dichlorobenzene (o-DCB) overnight and
filtered with a 0.2-lm PTFE filter. Finally, the substrates
were transferred into the evaporator with shadow masks to
define the active area of the devices (10.08 mm2) and
pumped under high vacuum (<2 ꢂ 10ꢀ7 Torr). Then calcium
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1362–1373
1365

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
SCHEME 1 Synthesis of PThTPA-IDT polymers. (i) POCl3, DMF, 0 ꢁC to rt; (ii) N-bromosuccinimide, THF, 12 h, rt; (iii) N-bromosuc-
cinimide, THF, 12 h, 0 ꢁC; (iv) (a) n-BuLi, THF, ꢀ78 ꢁC, 1 h, (b) ClSnMe3, hexane, rt, 12 h; (v) Pd2(dba)3, P(o-tol)3, THF, reflux, 16 h;
(vi) N-bromosuccinimide, THF, 12 h, rt; (vii) 2,5-bis(trimethylstannyl)thiophene, Pd(PPh3)4, toluene, DMF, 110 ꢁC, 48 h; (viii) malo-
nonitrile, CHCl3, pyridine, 50 ꢁC, 16 h; (ix) TBA, CHCl3, pyridine, 50 ꢁC, 16 h; (x) 3-dicyanomethylene-1-indanone, CHCl3, pyridine,
50 ꢁC, 16 h.
(30 nm) and aluminum (100 nm) were thermally evaporated
onto the active layer sequentially. The unencapsulated solar
cells were measured in glove box conditions using a Keithley
2400 SMU source measurement unit and an Oriel Xenon
lamp (450 W) with an AM1.5 filter as the solar simulator. A
reference silicon solar cell with a KG5 filter, which has been
previously standardized by the National Renewable Energy
Laboratory (NREL), was used to calibrate the light intensity
to 100 mW cmꢀ2. To fabricate the hole only device, the same
procedure to the photovoltaic device was followed except
that MoO3 was used to relace the calcium.
as a catalyst in toluene/DMF solution. The polymerization
was carried out at 110 C under nitrogen atmosphere for 48
ꢁ
h. The resulting polymers were collected by precipitating the
reaction solution in methanol followed by filtration. After
Soxhlet extraction with acetone, hexane and methanol for 12
h each, the final polymer was collected by extraction with
CHCl3 and the resultant solution was precipitated in metha-
nol. This polymer was then functionalized with malononitrile
(DCN), TBA, and DCNIO acceptors via a Knoevenagel conden-
sation to yield their respective polymers.
Polymer Characterization
RESULTS AND DISCUSSION
The polymers all have good solubility in a wide range of or-
ganic solvents including THF, dichloromethane, chloroben-
zene, and dichlorobenzene. The molecular weights of the
polymers were measured by GPC with polystyrene as stand-
ard and THF as eluent. The number-average molecular
weights (Mn) of PThTPA-IDT-DCN, PThTPA-IDT-TBA, and
PThTPA-IDT-DCNIO are 22.0, 21.7, and 22.1 kDa with poly-
dispersity indices (PDI) of 1.90, 1.87, and 1.91, respectively.
Because of the improved solubility of PThTPA-IDT, Mn is
higher than those reported for many reported D–p-bridge–A
polymers.7–14 This improved solubility is attributed to the
solubilizing IDT unit.
Synthesis
General synthesis of the TPA-IDT monomer and PThTPA-IDT
are shown in Scheme 1. The IDT unit was prepared as
reported previously.33 The IDT molecule was then monofor-
mylated in moderate yield via the Vilsmeier–Haack reaction
and then brominated with NBS to give 2. Concurrently, tri-
phenylamine was monobrominated with NBS to yield 3,
which was lithiated and quenched with trimethyltin chloride
to generate 4. Then, 2 and 4 were coupled via a Stille reac-
tion to yield 5, which was then dibrominated with NBS. Co-
polymer PThTPA-IDT-CHO was synthesized by Stille cross-
coupling between and 2,5-bis(trimethylstannyl)thiophene,
using tetrakis(triphenylphosphine)palladium(0) (Pd(PPh3)4)
The thermal properties of these polymers were evaluated
using DSC. All three PThTPA-IDT polymers exhibit two
1366
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1362–1373

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
the longer-wavelength absorption band is attributed to ICT
from the electron-rich backbone through the p-bridge to the
electron-deficient side-chain acceptors.8 Similar to reported
D–p-bridge–A polymers, stronger electron acceptors decrease
Eg and redshift the ICT absorption peak by increasing the
polymer LUMO level.7–14
Interestingly, PThTPA-IDT-DCN exhibits a very weak ICT
shoulder on the main p–p* transition peak instead of the full
ICT peak as displayed by PThTPA-IDT-TBA and PThTPA-IDT-
DCNIO. This unusual phenomenon suggests that the DCN
acceptor is not strong enough to pull electron density away
from the polymer backbone through the p-bridge and form a
stable ICT state. This is attributed to the high-aromatic stabi-
lization energy of the IDT unit, making the formation of an
ICT state with a weak acceptor energetically unfavorable.
The absorbance profile suggests that there is a threshold
acceptor strength that is sufficient to funnel charge through
the IDT p-bridge by stabilizing the ICT state; the DCN
acceptor does not possess this threshold electron-withdraw-
ing strength. In contrast, both TBA and DCNIO have suffi-
cient electron withdrawing strength to stabilize the ICT state
of the IDT p-bridge. In addition, both the TBA and DCNIO
acceptors blueshift the p–p* transition peak by ꢃ25 nm,
while the DCN acceptor does not cause a similar hypsochro-
mic shift. This p–p* peak shift also does not occur in any
existing D–p-bridge–A polymers. Although this shift is not
fully understood, it is attributed to the fact that these accept-
ors pull electron density away from the electron-rich back-
bone via the IDT p-bridge to form a CT state. This CT state
stabilizes the polymer HOMO, effectively increasing the p–p*
transition energy. If this is accurate, the shift is not seen in
other D–p-bridge–A polymers because there is very poor
electronic communication between the polymer backbone
and the side chain and thus, regardless of its electron-with-
drawing strength, the acceptor is unable to pull electrons
away from the electron-rich backbone.
FIGURE 2 DSC spectra for PThTPA-IDT-DCN (solid line),
PThTPA-IDT-TBA (dotted line), and PThTPA-IDT-DCN (dashed
line).
identical glass transition temperatures (Tg) around 70 and
100 ꢁC (Fig. 2), which correspond to Tg for the solubilizing
alkyl chains and the polymer backbone, respectively. This
amorphous behavior is expected based on the propellor-like
twist of the triphenylamine moiety, which inhibits intermo-
lecular p–p packing.5
The UV–vis absorption spectra of the polymers in CHCl3 sol-
utions and thin films are shown in Figure 3 and the sum-
marized data is listed in Table 1. The absorption maxima in
chloroform solution were observed at 445, 581, and 639 nm
for PThTPA-IDT-DCN, PThTPA-IDT-TBA, and PThTPA-IDT-
DCNIO, respectively. The peaks of the thin film absorption
are slightly broadened compared with the solution peaks,
but otherwise are identical. As expected for amorphous
materials, the spectra show no evidence of strong intermo-
lecular packing. The absorption coefficients were calculated
from the solid state absorbance and are 1.11 ꢂ 105 cmꢀ1
,
0.87 ꢂ 105 cmꢀ1, and 0.86 ꢂ 105 cmꢀ1 for PThTPA-IDT-
DCN, PThTPA-IDT-TBA, and PThTPA-IDT-DCNIO, respectively.
Two obvious absorption bands are observed for the poly-
mers in films and solutions. The shorter wavelength absorb-
ance can be assigned to the backbone p–p* transition while
Encouragingly, the ratio of the low-energy p–p* ICT peak
intensity to the high-energy backbone p–p* transition peak
intensity of PThTPA-IDT-TBA and PThTPA-IDT-DCNIO
FIGURE 3 (a) Solution absorbance (CHCl3) and (b) thin film absorbance for PThTPA-IDT-CHO (dotted-dashed line), PThTPA-IDT-
DCN (solid line), and PThTPA-IDT-TBA (dotted line) and PThTPA-IDT-DCN (dashed line).
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1362–1373
1367

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
TABLE 1 PThTPA-IDT Characteristics
CHCl3 Solution
Thin Film
Mw
Mn
kmax
Eg
kmax
Eg
HOMOb
(ev)
LUMOc
(eV)
opt
opt
Polymer
(kDa)
(kDa)
PDI
(nm)
(eV)
(nm)
(eV)
aa
PThTPA-IDT-DCN
PThTPA-IDT-TBA
PThTPA-IDT-DCNIO
a
22.0
21.7
22.1
41.8
40.5
42.2
1.90
1.87
1.91
439
581
639
2.05
1.86
1.69
c
439
581
639
2.05
1.86
1.69
1.11
0.87
0.86
ꢀ5.18
ꢀ5.14
ꢀ5.13
ꢀ3.13
ꢀ3.28
ꢀ3.44
Absorption coefficient of thin films at kmax (ꢂ105 cmꢀ1).
LUMO calculated from HOMO and optical bandgap.
b
HOMO calculated from the oxidation onset of the CV curve.
(ꢃ1.2:1) is much larger than that of previous D–p-bridge–A
polymers (ꢃ0.5:1),7–14 which leads to better overlap with
the solar spectrum. In addition, the absorption coefficients of
thin films of the three polymers are much higher than those
reported for indenofluorene-TPA side-chain polymers (7 ꢂ
104 cmꢀ1 at p–p* peak; 3 ꢂ 104 cmꢀ1 at ICT peak)7 and car-
bazole thienylene–vinylene side-chain polymers (ꢃ5 ꢂ 104
cmꢀ1 at kmax).14 This increase in relative ICT intensity and
optical density is attributed to the inclusion of thiophene in
the polymer backbone. It is known that phenyl–thiophene
linkages are more planar than phenyl–phenyl bonds due to
the minimized steric hindrance between adjacent aromatic
protons.39 In addition, thiophene has a lower aromatic stabi-
lization energy than benzene,19 which should improve the
electron delocalization along the backbone and the electronic
communication between the polymer backbone and side-
chain acceptor. The increase in ICT intensity could also be
due to the inclusion of the IDT p-bridge. However, due to the
high resonance stabilization energy and resultant poor
charge transfer character of the IDT unit compared with vi-
nylene-based p-bridges, we think it is unlikely that the IDT
p-bridge is enhancing the ICT absorbance.
This hypothesis was corroborated by DFT35,40,41 calculations
at the B3LYP/6-31G* level (Fig. 4).36–38 Although DFT often
overestimates theoretical energy levels, the B3LYP/6-31G*
method has been found to be an accurate formalism for pre-
dicting the optical and geometrical properties of conjugated
polymers.42 The optimized geometry shows that the dihedral
angle between the thiophene monomer and the TPA phenyl
rings is ꢃ22ꢁ. Conversely, the phenyl–phenyl linkages of
related fluorene-TPA side-chain polymers exhibit a dihedral
angle of ꢃ36ꢁ. It is known that twists in the backbone of
conjugated polymers breaks their p-conjugation and shortens
the delocalization length of the polymer.43 As a result, DFT
models predict that fluorene-TPA polymers have a relatively
short effective conjugation length. This is seen in the HOMO
wave function, which is interrupted by the fluorene unit. In
contrast, since the thiophene–TPA linkage is relatively planar,
the HOMO wave function delocalizes along the polymer back-
bone and enhances the effective polymer conjugation length.
Contrary to the differences in the polymer HOMO wave func-
tions, the LUMO wave functions are quite similar. In both
cases, the LUMO is almost entirely localized on the side
FIGURE 4 Density functional theory models of HOMO and LUMO orbitals for PThTPA-IDT-DCN and PFTPA-DCN.
1368
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1362–1373

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
opt
where Eg
denotes the optical bandgaps of the polymers.
PThTPA-IDT-DCN, PThTPA-IDT-TBA, and PThTPA-IDT-DCNIO
have HOMO levels of ꢀ5.18, ꢀ5.14, and ꢀ5.13 eV and LUMO
levels of ꢀ3.13, ꢀ3.28, and ꢀ3.44 eV, respectively. The
strength of the acceptor has minimal effect on polymer
HOMO, but a significant effect on the LUMO and Eg, in agree-
ment with the DFT predictions. Stronger acceptors stabilize
the LUMO to a greater extent, which thereby lowers Eg.
Field Effect Transistor and Photovoltaic Properties
Finally, the PThTPA-IDT behavior and device performance
were evaluated. The device parameters are summarized in
Table 2. Top contact OFET devices were fabricated to test
the lateral hole mobility of the PThTPA-IDT polymers. Poly-
mer thin films were spin-coated on Si/SiO2 substrates coated
with an HMDS self-assembled monolayer (SAM) and gold
was used for the source and drain electrodes. Figure 6
depicts the resultant I–V transfer and output OFET character-
istics obtained by sweeping Vgs from ꢀ100 to 10 V under a
Vds of ꢀ100 V. The OFET devices displayed typical p-channel
characteristics. The hole mobilities of PThTPA-IDT-DCN,
FIGURE 5 CV curves for PThTPA-IDT-DCN (solid line), PThTPA-
IDT-TBA (dotted line), and PThTPA-IDT-DCN (dashed line).
chain and acceptor. There is negligible overlap of the HOMO
and LUMO wave functions. This poor wave function overlap
in the PThTPA-IDT polymers indicates that any change in the
electronic nature of the acceptor will have a large effect on
the polymer LUMO and a very minimal effect on the polymer
HOMO. This poor spatial overlap also likely leads to a weak
ICT oscillator strength and could explain why the ICT transi-
tion is less intense compared with typical linear D–A
polymers.18
PThTPA-IDT-TBA, and PThTPA-IDT-DCNIO were 2.63 ꢂ 10ꢀ4
,
1.28 ꢂ 10ꢀ3, and 5.30 ꢂ 10ꢀ4 cm2 Vꢀ1
s
ꢀ1, respectively.
The space charge limited current (SCLC) model was also
employed to investigate the vertical hole mobilities. The
mobilities were extracted by modeling the dark current in
the SCLC region. The calculated vertical hole mobility of
PThTPA-IDT-DCN, PThTPA-IDT-TBA, and PThTPA-IDT-DCNIO
BHJ films were 6.76 ꢂ 10ꢀ5, 1.15 ꢂ 10ꢀ4, and 1.66 ꢂ 10ꢀ4
The HOMO and LUMO levels of the PThTPA-IDT polymers
were investigated using CV of polymer films on indium tin
oxide (ITO) substrates in a 0.1 M Bu4NPF6 acetonitrile solu-
tion at a scan rate of 50 mV sꢀ1 using ITO, Ag/AgCl and Pt
mesh as the working electrode, reference electrode and
counter electrode, respectively. The CV curves are shown in
Figure 5 and the data is summarized in Table 1. All three
polymers exhibit two quasi-reversible oxidation processes.
The first is attributed to the oxidation of the triphenylamine
unit and the second may be due to the oxidation of the fused
IDT moiety, which can delocalize and stabilize the resultant
charge. The HOMO was calculated from the equation
cm2 Vꢀ1 ꢀ1, respectively. The vertical mobilities are roughly
s
one order of magnitude lower than the horizontal mobilities.
This small discrepancy suggests that these polymers exhibit
nearly isotropic charge transport characteristics, which can
be attributed to the hyperbranched amorphous nature of the
polymers.
The hole mobilities were largely insensitive to annealing, as
expected for amorphous polymers. These mobility values are
very similar to the previously reported hole mobilities of
other D–p-bridge–A polymers.7–14 Although these mobilities
are reasonable for PSC applications, they are lower than
those reported for IDT-containing linear D–A polymers,24–28
which might lead to increased charge recombination and
inefficient charge transfer. It is likely that the mobilities are
limited by poor intermolecular charge transport due to weak
intermolecular interactions.
EHOMOðeVÞ ¼ ꢀ eðEox ꢀ E1=2ðferroceneÞ þ 4:8VÞ
where Eox is the onset oxidation potential of the polymers
versus Ag/Agþ. Conversely, the reductive curves were
entirely irreversible and changed drastically upon repeated
cycles. Therefore, the LUMO was calculated from the optical
bandedge and HOMO energy level from the equation
The photovoltaic properties of the PThTPA-IDT polymers
were tested using PC71BM as the acceptor in the conven-
tional device configuration of ITO/PEDOT:PSS (40 nm)/poly-
mer:PC71BM/Ca (30 nm)/Al (100 nm). PC71BM, which
ELUMO ¼ EHOMO þ Egopt
TABLE 2 PThTPA-IDT Device Characteristics
Polymer
lFET (cm2 Vꢀ1 sꢀ1
)
lSCLC (cm2 Vꢀ1 sꢀ1
)
Voc (V)
JSC (mA cmꢀ2
)
FF
PCE (%)
PThTPA-IDT-DCN
PThTPA-IDT-TBA
PThTPA-IDT-DCNIO
2.63 ꢂ 10ꢀ4
1.28 ꢂ 10ꢀ3
5.30 ꢂ 10ꢀ4
6.76 ꢂ 10ꢀ5
1.15 ꢂ 10ꢀ4
1.66 ꢂ 10ꢀ4
0.83
0.87
0.86
3.42
5.19
5.17
0.43
0.44
0.45
1.23
1.97
1.98
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1362–1373
1369

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
FIGURE 6 Output (a,c,e) and transfer (b,d,f) characteristics of PThTPA-IDT-DCN (a,b), PThTPA-IDT-TBA (c,d) and PThTPA-IDT-
DCNIO (e,f) OFETs on HMDS-modified SiO2 as a function of Vgs (W ¼ 1000 lm and L ¼ 12 lm).
absorbs more visible light than PC61BM,44 was utilized as
the n-type acceptor to efficiently harvest solar output. The
BHJ active layer was prepared by spin coating a chloroform:
chlorobenzene (1:1, v/v) solution of polymer:PC71BM on the
PEDOT:PSS layer. All the devices were heated at 150 ꢁC for
10 min prior to electrode deposition. The device parameters
are summarized in Table 2. Thermal annealing improved the
device performance marginally compared with unannealed
devices (data not shown). This insensitivity to thermal
annealing is expected for noncrystalline polymers. Solvent
annealing with dichlorobenzene vapor was also explored, but
resulted in unfavorable morphologies and poor device per-
formance. In addition, different polymer:PCBM ratios were
tested, but the best device performances were obtained with
a 1:4 blending ratio. This unfavorable optimal blending ratio
is likely due to the hyperbranched nature of the polymers,
1370
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1362–1373

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
FIGURE 8 External quantum efficiency plot for PThTPA-IDT-
DCN (solid line), PThTPA-IDT-TBA (dotted line), and PThTPA-
IDT-DCN (dashed line).
FIGURE 7 Current voltage curve for PThTPA-IDT-DCN (solid
line), PThTPA-IDT-TBA (dotted line), and PThTPA-IDT-DCN
(dashed line).
which introduces large interstitial spaces around the poly-
mer chain. PCBM molecules initially intercalate into these
voids before PCBM forms continuous pure-phase domains in
the BHJ and thereby generates balanced electron-hole charge
transport and reduced charge recombination.45
Photovoltaic devices fabricated from PThTPA-IDT-DCN
showed a PCE up to 1.23% with a Voc of 0.83 V, a FF of 0.43,
and a Jsc of 3.42 mA cmꢀ2. Comparatively, PThTPA-IDT-TBA
and PThTPA-IDT-DCNIO showed nearly identical FF (0.44
and 0.45), Voc (0.87 and 0.86 V), Jsc (5.19 and 5.17 mA
FIGURE 9 Tapping mode atomic force microscopy topography (a–c) and phase images (d–f) of polymer:PC71BM (1:4) blends of
(a,d) PThTPA-IDT-DCN, (b,e) PThTPA-IDT-TBA, and (c,f) PThTPA-IDT-DCNIO.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1362–1373
1371

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
cmꢀ2), and PCE (1.97 and 1.98%). Representative current–
voltage curves for the three polymers measured under
standard illumination conditions (100 mW cmꢀ2, AM 1.5G)
are plotted in Figure 7. Since the Voc is largely governed by
the difference between the polymer HOMO level and PCBM
LUMO and the polymers have nearly identical HOMO levels
as measured by CV, the Voc values are expected to be very
similar for all the PThTPA-IDT polymers. These Voc values
are reasonably high for PSC devices. The FF values of the
three polymers are nearly identical as well, suggesting simi-
lar loss mechanisms for each polymer. However, despite the
improved absorbance profile compared with existing D–p-
bridge–A polymers, each of the PThTPA-IDT devices has com-
paratively low Jsc values, which results in low PCE values.
We speculate that the low Jsc and FF values are due to the
limited external quantum efficiency (EQE) of the polymers
(Fig. 8). All three polymers have EQE values below 30%,
which is lower than many high-efficiency D–A polymers.
explains why the device parameters are nearly identical for
these two polymers, despite their inherent chemical, elec-
tronic, and morphological differences.
CONCLUSIONS
In summary, a new class of tunable light-harvesting polymer
(PThTPA-IDT) with dipolar chromophores embedded in the
conjugated backbone was designed and synthesized. This
precursor polymer was functionalized with three electron-
accepting moieties via a Knoevenagel condensation reaction.
The thiophene–triphenylamine backbone contributed to
enhanced conjugation length and improved ICT characteris-
tics, as seen in the absorbance spectra. The photovoltaic
properties of these polymers were investigated and the high-
est achieved PCE for PThTPA-IDT-DCN, PThTPA-IDT-TBA, and
PThTPA-IDT-DCNIO were 1.23, 1.97, and 1.98%, respectively.
The PCE was limited by the low Jsc and FF of this polymer
system, which are attributed to poor morphology and rela-
tively low hole mobilities. The undesirable morphology and
low mobilities could contribute to decreased charge separa-
tion efficiency and increased recombination rates in the
active layer.
These limited EQE values could be due to several factors.
First, the polymers exhibit relatively small absorption coeffi-
cients compared with D–A polymers. Second, based on the
limited hole mobilities, geminate recombination is likely a
major loss pathway. This is supported by the relatively thin
optimized BHJ thicknesses (70–80 nm). Finally, higher
recombination rates could be due to the undesirable mor-
phology of the polymers (Fig. 9). Both PThTPA-IDT-DCN and
PThTPA-IDT-DCNIO exhibit severe phase segregation with
PC71BM, which could contribute to exciton recombination
before the excitons are able to diffuse to polymer/fullerene
interfaces. Conversely, PThTPA-IDT-TBA shows no significant
polymer or fullerene domains, which may contribute to
increased geminate recombination. The reasons for these
striking morphological differences between the three poly-
mers are not known and are currently under investigation.
This work was supported by the National Science Foundation’s
NSF-STC program under Grant No. DMR-0120967, the AFOSR
(FA9550-09-1-0426), the Office of Naval Research (N00014-
11-1-0300), and the World Class University (WCU) program
through the National Research Foundation of Korea under the
Ministry of Education, Science and Technology (R31-
21410035). A.K.-Y.J. thanks the Boeing-Johnson Foundation for
financial support. D.F.Z. thanks Dr. Joshua Davies for his invalu-
able guidance and advice.
REFERENCES AND NOTES
Interestingly, the device parameters of PThTPA-IDT-TBA and
PThTPA-IDT-DCNIO are nearly identical despite the signifi-
cant differences in their electronic properties and film mor-
phologies. The EQE for PThTPA-IDT-TBA approaches 30%
from 400 to 625 nm, while the EQE for PThTPA-IDT-DCNIO
for the same range never surpasses 25%. The maximum
EQE for PThTPA-IDT-TBA is slightly higher than that of
PThTPA-IDT-DCNIO probably because PThTPA-IDT-TBA pos-
sesses a marginally higher hole mobility. However, the photo-
response of PThTPA-IDT-DCNIO is extended to longer wave-
lengths because of its smaller bandgap. As a result, both
polymers have very similar aggregate EQEs over the visible
spectrum. Moreover, the morphology of PThTPA-IDT-TBA/
PCBM films shows no uninterrupted polymer or PCBM net-
works, which likely contributes to a high rate of geminate
recombination. Conversely, the PThTPA-IDT-DCNIO/PCBM
films exhibit severe phase segregation, suggesting that exci-
tonic recombination is the major loss pathway. Thus, the
higher mobility and geminate recombination rates of
PThTPA-IDT-TBA are very nearly balanced with the red-
shifted absorbance and excitonic recombination rates of
PThTPA-IDT-DCNIO. Since Jsc and FF are intimately related to
absorbance, mobility, recombination rates and EQE, this
1 Morton, O. Nature 2006, 443, 19–22.
2 Boukai, A.; Haney, P.; Katzenmeyer, A.; Gallatin, G. M.; Talin,
A. A.; Yang, P. Chem. Phys. Lett. 2011, 501, 153–158.
3 Yu, G.; Gao, J.; Hummelen, J. C.; Wudl, F.; Heeger, A. J. Sci-
ence 1995, 270, 1789–1791.
4 Dennler, G.; Scharber, M. C.; Brabec, C. J. Adv. Mater. 2009,
21, 1323–1338.
5 Oberhumer, P. M.; Huang, Y.-S.; Massip, S.; James, D. T.; Tu,
G.; Albert-Seifriend, S.; Beljonne, D.; Cornil, J.; Kim J.-S.; Huck,
W. T. S.; Greenham, N. C.; Hodgkiss, J. M.; Friend, R. H. J.
Chem. Phys. 2011, 134, 114901.
6 Dufresne, S.; Guarin, S. A. P.; Bolduc, A.; Bourque, A. N.;
Skene, W. G. Photochem. Photobiol. Sci. 2009, 8, 796–804.
7 Duan, C.; Cai, W.; Zhong, C.; Li, Y.; Wang, X.; Huang, F.; Cao,
Y. J. Polym. Sci. Part A: Polym. Chem. 2011, 49, 4406–4415.
8 Huang, F.; Chen, K.-S.; Yip, H.-L.; Hau, S. K.; Acton, O.;
Zhang, Y.; Luo, J.; Jen, A. K.-Y. J. Am. Chem. Soc. 2009, 131,
13886–13887.
9 Duan, C.; Cai, W.; Huang, F.; Zhang, J.; Wang, M.; Yang, T.;
Zhong, C.; Gong, X.; Cao, Y. Macromolecules 2010, 43,
5262–5268.
10 Cheng, Y. J.; Hung, L. C.; Cao, F. Y.; Kao, W. S.; Chang, C.
Y.; Hsu, C. S. J. Polym. Sci. Part A: Polym. Chem. 2011, 49,
1791–1801.
1372
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1362–1373

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
11 Hsu, S.-L.; Chen, C.-M.; Wei, K.-H. J. Polym. Sci. Part A:
Polym. Chem. 2010, 48, 5128–5134.
31 Li, Z. H.; Wong, M. S. Org. Lett. 2006, 8, 1499.
32 Qian, G.; Dai, B.; Luo, M.; Yu, D.; Zhan, J.; Zhang, Z.; Ma,
D.; Wang, Z. Y. Chem. Mater. 2008, 20, 6208–6216.
12 Duan, C.; Chen, K.-S.; Huang, F.; Yip, H.-L.; Liu, S.; Zhang,
J.; Jen, A. K.-Y.; Cao, Y. Chem. Mater. 2010, 22, 6444–6452.
33 Zhang, Y.; Zou, J.; Yip, H.-L.; Chen, K.-S.; Zeigler, D. F.; Sun,
Y.; Jen A. K.-Y. Chem. Mater. 2011, 23, 2289–2291.
13 Sahu, D.; Padhy, H.; Patra, D.; Huang, J.-H.; Chu, C.-W.; Lin,
H.-C. J. Polym. Sci. Part A: Polym. Chem. 2010, 48, 5812–5823.
34 van Pham, C.; Macomber, R. S.; Mark, H. B., Jr.; Zimmer, H.
J. Org. Chem. 1984, 49, 5250–5253.
14 Zhang, Z.-G.; Fn, H.; Min, J.; Zhang, S.; Zhang, J.; Zhang,
M.; Guo, X.; Zhan, X.; Li, Y. Polym. Chem. 2011, 2, 1678–1687.
35 Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Men-
nucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.;
Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnen-
berg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hase-
gawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai,
H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.;
Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov,
V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell,
A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.;
Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.;
Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J.
W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.;
Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.;
15 Marder, S. R.; Cheng, L.-T.; Tiemann, B. G.; Friedli, A. C.;
Blanchard-Desce, M.; Perry, J. W.; Skindhøj, J. Science 1994,
263, 511–514.
16 Liu, S.; Haller, M. A.; Ma, H.; Dalton, L. R.; Jang, S.-H.; Jen,
A. K.-Y. Adv. Mater. 2003, 15, 603.
17 Cheng, Y.-J.; Luo, J.; Hau, S.; Bale, D. H.; Kim, T.-D.; Shi, Z.;
Lao, D. B.; Tucker, N. M.; Tian, Y.; Dalton, L. R.; Reid, P. J.; Jen,
A. K.-Y. Chem. Mater. 2007, 19, 1154–1163.
18 Hirani, B.; Li, J.; Djurovich, P. I.; Yousufuddin, M.; Oxgaard,
J.; Persson, P.; Wilson, S. R.; Bau, R.; Goddard, W. A.; Thomp-
son, M. E. Inorg. Chem. 2007, 46, 3865–3875.
19 Li, Z.; Zhao, Y.; Zhou, J.; Shen, Y. Eur. Polym. J. 2000, 36,
2417–2421.
¨
20 Sancho-Garcia, J. C.; Foden, C. L.; Grizzi, I.; Greczynski, G.;
Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D.
ꢀ
de Jong, M. P.; Salaneck, W. R.; Bredas, J.-L.; Cornil, J. J.
J. Gaussian 09, Revision A.02; Gaussian, Inc.: Wallingford, CT,
Phys. Chem. B 2004, 108, 5594–5599.
2009.
21 Meng, K.; Ding, Q.; Wang, S.; He, Y.; Li, Y.; Gong, Q. J.
Phys. Chem. B 2010, 114, 2602–2606.
36 Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B: Condens. Matter
1988, 37, 785–789.
22 Schulz, G. L.; Chen, X.; Holdcroft, S. Appl. Phys. Lett. 2009,
94, 023302.
37 Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
38 Ditchfield, R.; Hehre, W. J.; Pople, J. A. J. Chem. Phys.
1971, 54, 724–728.
23 Bundgaard, E.; Krebs, F. C. Macromolecules 2006, 39, 2823–2931.
24 Chen, C.-P.; Chan, S.-H.; Chao, T.-C.; Ting, C.; Ko, B.-T. J.
Am. Chem. Soc. 2008, 130, 12828–12833.
39 Poolmee, P.; Ehara, M.; Hannongbua, S.; Nakatsuji, H. Poly-
mer 2005, 46, 6474–6481.
25 Chen, Y.-C.; Yu, C.-Y.; Fan, Y.-L.; Huang, L.-I.; Chen, C.-P.;
Ting, C. Chem. Commun. 2010, 46, 6503–6505.
40 Hohenberg, P.; Kohn, W. Phys. Rev. 1964, 136, B864–B871.
41 Kohn, W.; Sham, L. Phys. Rev. 1965, 140, A1133–A1138.
26 Chen, C.-P.; Chan, S.-H.; Chao, T.-C.; Ting, C.; Lin, C.-S.; Ko,
B.-T. Macromolecules 2008, 41, 5519–5526.
42 Blouin, N.; Michaud, A.; Gendron, D.; Wakim, S.; Blair, E.;
Neagu-Plesu, R.; Belleteˆte, M.; Durocher, G.; Tao, Y.; Leclerc,
M. J. Am. Chem. Soc. 2008, 130, 732–742.
27 Yu, C.-Y.; Chen, C.-P.,Chan, S.-H.; Hwang, G.-W.; Ting, C.
Macromolecules 2009, 21, 3262–3269.
43 Nguyen, T.-Q.; Wu, J.; Doan, V.; Schwartz, B. J.; Tolbert, S.
H. Science 2000, 288, 652–656.
28 Zhang, W. M.; Smith, J.; Watkins, S. E.; Gysel, R.; McGehee,
M.; Salleo, A.; Kirkpatrick, J.; Ashraf, S.; Anthopoulos, T.; Hee-
ney, M.; McCulloch, I. J. Am. Chem. Soc. 2010, 132, 11437–11439.
44 Boland, P.; Sunkavalli, S. S.; Chennuri, S.; Foe, K.; Abdel-
Fattah, T.; Namkoong, G. Thin Solid Films 2010, 518,
1728–1731.
29 D’Auria, M.; de Mico, A.; D’Onofrio, F.; Piancatelli, G. J.
Chem. Soc. Perkin Trans. I 1987, 1777–1780.
45 Mayer, A. C.; Toney, M. F.; Scully, S. R.; Rivnay, J.; Brabec,
C. J.; Scharber, M.; Koppe, M.; Heeney, M.; McCulloch, I.;
McGehee, M. D. Adv. Funct. Mater. 2009, 19, 1173–1179.
30 Spagnolo, P.; Zanirato, P.; Gronowitz, S. J. Org. Chem.
1982, 47, 3177–3180.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1362–1373
1373