D-π-A sensitizers for dye-sensitized solar cells: Linear vs branched oligothiophenes

Article abstract of DOI:10.1021/cm903542v

Two donor-π-acceptor (D-π-A) dyes, coded as L-3T-DPA 1 and B-5T-DPA 2, were synthesized for application in dye-sensitized solar cells. These D-π-A sensitizers use diphenylamine as donor, an oligothiophene as π-bridge, and cyanoacrylic acid as an acceptor group that can be anchored to the surface of TiO₂. While the two dyes comprise the same donor and acceptor units, the bridging oligothiophene is linear in one case and branched in the other case. Photophysical and electro-chemical properties of the dyes were investigated by UV-vis spectrometry and cyclic voltammetry. The dyes were subsequently implemented as sensitizers in dye-sensitized solar cells. Photovoltaic devices with dye 1 showed a maximum monochromatic incident photon to current efficiency (IPCE) of 80% and an overall conversion efficiency of 6.8% under full sunlight (AM 1.5G, 100 mW cm^-2) irradiation. The photovoltaic performance of branched dye 2 was lower because of less dye loading on the TiO₂ surface. The dyes were also tested in ionic liquid and solid-state devices and showed good efficiencies. Long-term stability measurements were performed over 1000 h at full sunlight and at 60°C in ionic liquid devices. Branched dye 2 thereby showed, excellent stability retaining 96% of its initial efficiency, while linear dye 1 retained 73% after 1.000 h of irradiation.

Full text of DOI:10.1021/cm903542v

1836 Chem. Mater. 2010, 22, 1836–1845
DOI:10.1021/cm903542v
D-π-A Sensitizers for Dye-Sensitized Solar Cells: Linear vs Branched
Oligothiophenes
Markus K. R. Fischer,†,§ Sophie Wenger,‡,§ Mingkui Wang,‡ Amaresh Mishra,*,†
,‡
Shaik M. Zakeeruddin,* Michael Gratzel,* and Peter Bauerle*
,‡
,†
€
€
†Institute of Organic Chemistry II and Advanced Materials, University of Ulm, Albert-Einstein-Allee 11,
D-89081 Ulm, Germany, and ‡Laboratory of Photonics and Interfaces, Institute of Chemical Sciences
and Engineering, Ecole Polytechnique Feꢀdeꢀrale de Lausanne (EPFL), Station 6, CH-1015 Lausanne,
Switzerland. § These authors contributed equally to this work.
Received November 23, 2009. Revised Manuscript Received January 11, 2010
Two donor-π-acceptor (D-π-A) dyes, coded as L-3T-DPA 1 and B-5T-DPA 2, were synthesized
for application in dye-sensitized solar cells. These D-π-A sensitizers use diphenylamine as donor, an
oligothiophene as π-bridge, and cyanoacrylic acid as an acceptor group that can be anchored to the
surface of TiO2. While the two dyes comprise the same donor and acceptor units, the bridging
oligothiophene is linear in one case and branched in the other case. Photophysical and electro-
chemical properties of the dyes were investigated by UV-vis spectrometry and cyclic voltammetry.
The dyes were subsequently implemented as sensitizers in dye-sensitized solar cells. Photovoltaic
devices with dye 1 showed a maximum monochromatic incident photon to current efficiency (IPCE)
of 80% and an overall conversion efficiency of 6.8% under full sunlight (AM 1.5G, 100 mW cm-2
)
irradiation. The photovoltaic performance of branched dye 2 was lower because of less dye loading
on the TiO2 surface. The dyes were also tested in ionic liquid and solid-state devices and showed good
efficiencies. Long-term stability measurements were performed over 1000 h at full sunlight and at
60 °C in ionic liquid devices. Branched dye 2 thereby showed excellent stability retaining 96% of its
initial efficiency, while linear dye 1 retained 73% after 1000 h of irradiation.
Introduction
using either RuII-polypyridyl complexes or metal-free
organic dyes, and efficiencies over 11% have been reported
using ruthenium sensitizers.1,3,4 In recent years, metal-free
organic dyes have attracted increased interest as an alter-
native to ruthenium complexes because of low material
costs, ease of synthesis, high molar extinction coefficients,
and environmental friendly materials.5 Although remark-
able advances have been made with metal-free organic dyes
as sensitizers in DSSCs and efficiencies exceeding 9% have
been reached, there is still a need to optimize their chemical
and physical properties for further improvement of the
device performance.6,7
The conversion of solar energy to electricity is attract-
ing considerable attention as a form of clean renewable
energy resources. In this context, dye-sensitized solar cells
(DSSCs) are considered as a credible alternative to con-
ventional inorganic silicon-based solar cells.1,2 Compared
to silicon solar cells, DSSCs offer major advantages such
as low material cost, tunability of the absorption spec-
trum, and encouraging efficiency even under diffuse light.
DSSCs efficiently convert solar energy into electricity
The properties of D-π-A dyes can be easily tuned by
varying donor, spacer, and acceptor moieties. Our inten-
tion was to use electron rich oligothiophenes as π-con-
jugated bridge as they are known to be ideal building
blocks for DSSCs.5,8 Upon photoexcitation, electron
transfer from the donor to the acceptor is a fast process
*Corresponding author: amaresh.mishra@uni-ulm.de (A.M.), peter.
baeuerle@uni-ulm.de (P.B.), shaik.zakeer@epfl.ch (S.M.K.), michael.
graetzel@epfl.ch (M.G.).
€
€
(1) (a) Gratzel, M. Nature 2001, 414, 338. (b) O'Regan, B.; Gratzel, M.
Nature 1991, 353, 737.
€
(2) (a) Hagfeldt, A.; Gratzel, M. Acc. Chem. Res. 2000, 33, 269. (b)
€
Gratzel, M. Acc. Chem. Res. 2009, 42, 1788. (c) Yanagida, S.; Yu, Y.;
Manseki, K. Acc. Chem. Res. 2009, 42, 1827.
(3) Nazeeruddin, M. K.; Pechy, P.; Renouard, T.; Zakeeruddin, S. M.;
Humphry-Baker, R.; Comte, P.; Liska, P.; Cevey, L.; Costa, E.;
€
Shklover, V.; Spiccia, L.; Deacon, G. B.; Bignozzi, C. A.; Gratzel,
€
(5) Mishra, A.; Fischer, M. K. R.; Bauerle, P. Angew. Chem., Int. Ed.
M. J. Am. Chem. Soc. 2001, 123, 1613.
2009, 48, 2474.
(4) (a) Chiba, Y.; Islam, A.; Watanabe, Y.; Komiya, R.; Koide, N.;
Han, L. Jpn. J. Appl. Phys. 2006, 45, L638. (b) Cao, Y.; Bai, Y.; Yu,
Q.; Cheng, Y.; Liu, S.; Shi, D.; Gao, F.; Wang, P. J. Phys. Chem. C
2009, 113, 6290. (c) Gao, F.; Wang, Y.; Shi, D.; Zhang, J.; Wang, M.;
(6) (a) Ferrere, S.; Zaban, A.; Gregg, B. A. J. Phys. Chem. B 1997, 101,
4490. (b) Tian, H.; Liu, P.-H.; Zhu, W.; Gao, E.; Wu, D.-J.; Cai, S.
J. Mater. Chem. 2000, 10, 2708. (c) Ferrere, S.; Gregg, B. A. New J.
Chem. 2002, 26, 1155.
€
Jing, X.; Humphry-Baker, R.; Wang, P.; Zakeeruddin, S. M.; Gratzel,
(7) (a) Ito, S.; Zakeeruddin, S. M.; Humphry-Baker, R.; Liska, P.;
M. J. Am. Chem. Soc. 2008, 130, 10720. (d) Cao, Y.; Bai, Y.; Yu, Q.;
Cheng, Y.; Liu, S.; Shi, D.; Gao, F.; Wang, P. J. Phys. Chem. C 2009,
113, 6290. (e) Chen, C.-Y.; Wang, M.; Li, J.-Y.; Pootrakulchote, N.;
ꢀ
Charvet, R.; Comte, P.; Nazeeruddin, M. K.; Pechy, P.; Takata,
€
M.; Miura, H.; Uchida, S.; Gratzel, M. Adv. Mater. 2006, 18, 1202.
(b) Zhang, G.; Bala, H.; Cheng, Y.; Shi, D.; Lv, X.; Yu, Q.; Wang, P.
Chem. Commun. 2009, 2198.
€
Alibabaei, L.; Ngoc-le, C.-h.; Decoppet, J.-D.; Tsai, J.-H.; Gratzel, C.;
€
Wu, C.-G.; Zakeeruddin, S. M.; Gratzel, M. ACS Nano 2009, 3, 3103.
(8) Ooyama, Y.; Harima, Y. Eur. J. Org. Chem. 2009, 2903.
r
2010 American Chemical Society
pubs.acs.org/cm
Published on Web 01/27/2010

Article
Chem. Mater., Vol. 22, No. 5, 2010 1837
˚
to generate charge-separated species. In this respect,
triphenylamine or substituted derivatives have been suc-
cessfully used as donors for sensitizers in dye-sensitized
solar cells.9,10 It has also been observed that the structure
of the π-conjugated bridging group played a decisive role
in device performance.11 The bridging group can serve
both, as a part of the light absorbing chromophore and as
a pathway for transporting charges. Herein, we present
synthesis and characterization of a linear and a related
branched dye comprising diphenylamine oligothiophene
and cyanoacrylic acid. These dyes were charecterized in
DSSCs using volatile solvent-based electrolyte, ionic
liquid electrolyte, and a solid organic hole-conductor
(spiro-MeOTAD) as redox mediator.
was dried over molar sieve (4 A) prior to use, Pd2(dba)3 CHCl3,
3
HP(tBu)3BF4, n-BuLi (1.6 mol L-1 in hexane) were purchased
from Acros, diphenylamine and NaOtBu were purchased from
Fluka, cyanoacetic acid was purchased from ABCR, and
ammonium acetate was purchased from Merck. The potassium
phosphate solution was prepared by dissolving potassium phos-
€
phate monohydrate (Riedel-de Haen) in deionized water and
was degassed prior to use.
5-Bromo-30,40-diethyl-2,20:50,200-terthiophene (4). To a solu-
tion of terthiophene 3 (1.0 g, 3.29 mmol) in dry dimethylforma-
mide (30 mL) was added dropwise at 0 °CN-bromosuccinimide
(585 mg, 3.29 mmol, dissolved in 30 mL of dry dimethyl-
formamide). The reaction mixture was allowed to warm to room
temperature and was stirred for another 3 h. The reaction mixture
was poured into water and extracted three times with diethyl ether.
The combined organic extracts were washed with water (5 ꢀ
200 mL), brine (2 ꢀ 100 mL), filtered, and dried over Na2SO4. The
solvent was removed by rotary evaporation. After column chro-
matography (SiO2, hexane) terthiophene 4 (880 mg, 2.3 mmol,
70%) was obtained as a colorless viscous oil. 1H NMR (CDCl3,
400 MHz) δ ppm: 1.22 (t, 3J = 7.55 Hz, 3H, CH3), 1.22 (t, 3J =
7.55 Hz, 3H, CH3), 2.73 (q, 3J = 7.56 Hz, 2H, CH2), 2.75 (q, 3J =
7.56 Hz, 2H, CH2), 6.89 (d, 3J = 3.86 Hz, 1H), 7.02 (d, 3J = 3.86
Hz, 1H), 7.07 (dd, 3J = 5.16, 3.60 Hz, 1H), 7.15 (dd, 3J = 3.60,
Experimental Section
General Remarks. NMR spectra were recorded on a Bruker
AMX 500 (1H NMR: 500 MHz, 13C NMR: 125 MHz) or an
Avance 400 spectrometer (1H NMR: 400 MHz, 13C NMR: 100
MHz), normally at 25 °C. Chemical shift values (δ) are ex-
pressed in parts per million using residual solvent protons (1H
NMR, δH = 7.26 for CDCl3 and δH = 3.57 for THF-d8; 13C
NMR, δC = 77.0 for CDCl3 and δC = 67.2 for THF-d8) as
internal standard. The splitting patterns are designated as
follows: s (singlet), d (doublet), t (triplet), m (multiplet), and q
(quartet). The assignments are PhH (phenyl protons), CHO
(aldehyde proton), and ThH (thiophene protons). Melting
4J = 1.15 Hz, 1H), 7.32 (dd, 3J = 5.16, 4J = 1.15 Hz, 1H). 13
C
NMR (CDCl3, 100 MHz) δ ppm: 15.29, 15.37, 20.99, 21.03,
111.78, 125.53, 126.06, 126.11, 127.40, 128.80, 130.19, 130.39,
135.77, 137.60, 141.13, 141.64. EI-mass 382 [Mþ] (calcd. for
C16H15BrS3 381.95). Elemental analysis: calc. (%) for C16H15BrS3
C: 50.12; H: 3.94; S: 25.09; found: C: 50.19; H: 3.96; S: 25.04.
N,N-Diphenyl-30,40-diethyl-2,20:50,200-terthiophene-5-amine (5).
5-Bromoterthiophene 4 (500 mg, 1.3 mmol, 1.0 equiv), diphe-
nylamine (220 mg, 1.3 mmol, 1.0 equiv), and NaOtBu (138 mg,
1.44 mmol, 1.1 equiv) were added to a Schlenk tube. Then dry
degassed toluene (10 mL) was added to the mixture. Pd2-
€
points were determined using a Buchi B-545 apparatus and were
not corrected. Elemental analyses were performed on an Ele-
mentar Vario EL. Thin layer chromatography was carried out
on aluminum plates, precoated with silica gel, Merck Si60 F254
.
Preparative column chromatography was performed on glass
columns packed with silica gel, Merck Silica 60, particle size
40-43 μm. CI mass spectra were recorded on a Finnigan MAT
SSQ-7000, EI-mass were recorded on a Varian Saturn 2000
GC-MS, and MALDI-TOF mass spectra on a Bruker Daltonics
Reflex III.
(dba)3 CHCl3 (13.5 mg, 13 μmol, 0.01 equiv) and HP(tBu)3BF4
3
(7.6 mg, 26.1 μmol, 0.02 equiv) were suspended in 2 mL of dry
degassed toluene and stirred for 10 min. The catalyst suspension
was then added to the reactants. The reaction mixture was
stirred for 12 h at 100 °C. The solvent was subsequently removed
under reduced pressure. The mixture was redissolved in dichloro-
methane, and the resulting solution was washed with water and
brine and dried over Na2SO4. After column chromatography
(SiO2, hexane:DCM = 3:1) amine 5 (505 mg, 1.07 mmol, 82%)
Materials. Tetrahydrofurane (Merck) was dried under reflux
over sodium/benzophenone (Merck). Dimethylformamide
(Merck) was first refluxed over P4O10 (Merck) and distilled,
€
then refluxed over BaO (Riedel-de Haen), and distilled again.
The dry dimethylformamide was stored over molar sieve (4 A).
˚
All synthetic steps were carried out under argon atmosphere
(except Knoevenagel condensation). 2-Isopropoxy-4,4,5,5-tet-
ramethyl-1,3,2-dioxaborolane was purchased from Aldrich and
1
was obtained as a slightly yellow semisolid. H NMR (CDCl3,
400 MHz) δ ppm: 1.22 (t, 3J = 7.53 Hz, 3H, CH3), 1.22 (t, 3J =
7.53 Hz, 3H, CH3), 2.76 (q, J = 7.52 Hz, 2H, CH2), 2.77 (q,
3
3J = 7.52 Hz, 2H, CH2), 6.65 (d, 3J = 3.87 Hz, 1H, ThH), 6.94
(d, 3J = 3.88 Hz, 1H, ThH), 7.02-7.10 (m, 3H, PhH þ ThH),
7.15 (dd, 3J = 3.59, 4J = 1.12 Hz, 1H, ThH), 7.18-7.22 (m, 4H,
PhH), 7.26-7.32 (m, 5H, PhH þ ThH). 13C NMR (CDCl3, 100
MHz) δ ppm: 15.31, 15.33, 20.99, 21.09, 121.15, 122.59, 123.06,
124.18, 125.20, 125.70, 127.36, 129.17, 129.28, 130.20, 130.36,
136.13, 140.60, 141.10, 147.65, 151.11. EI-mass 472 [MþH]þ
(calcd. for C28H25NS3 471.11). Elemental analysis: calc. (%) for
C28H25NS3 C:71.30; H: 5.34; N: 2.97; found: C:71.34; H: 5.33; N:
2.86.
5-Diphenylamino-30,40-diethyl-2,20:50,200-terthiophene-500-carbal-
dehyde (6). n-BuLi (0.44 mL, 0.70 mmol, 1.6 mol L-1, 1.1 equiv)
was added dropwise at -78 °C to a solution of 5 (300 mg, 0.64
mmol, 1.0 equiv) in 5 mL of dry THF, and the mixture was kept at
this temperature and stirred for 0.5 h. Dry dimethylformamide
(92 μL, 1.27 mmol, 2.0 equiv) was added. The reaction mixture
was allowed to warm toroom temperature and stirred for another
(9) (a) Velusamy, M.; JustinThomas, K. R.; Lin, J. T.; Hsu, Y. C.; Ho,
K. C. Org. Lett. 2005, 7, 1899. (b) Hagberg, D. P.; Edvinsson, T.;
Marinado, T.; Boschloo, G.; Hagfeldt, A.; Sun, L. Chem. Commun.
2006, 2245. (c) Justin Thomas, K. R.; Hsu, Y.-C.; Lin, J. T.; Lee, K.-M.;
Ho, K.-C.; Lai, C.-H.; Cheng, Y.-M.; Chou, P.-T. Chem. Mater. 2008,
20, 1830. (d) Ning, Z.; Tian, H. Chem. Commun. 2009, 5483.
(10) (a) Bonhote, P.; Moser, J.-E.; Humphry-Baker, R.; Vlachopoulos,
€
N.; Zakeeruddin, S. M.; Walder, L.; Gratzel, M. J. Am. Chem. Soc.
1999, 121, 1324. (b) Hagberg, D. P.; Marinado, T.; Karlsson, K. M.;
Nonomura, K.; Qin, P.; Boschloo, G.; Brinck, T.; Hagfeldt, A.; Sun, L.
J. Org. Chem. 2007, 72, 9550. (c) Qin, P.; Zhu, H.; Edvinsson, T.;
Boschloo, G.; Hagfeldt, A.; Sun, L. J. Am. Chem. Soc. 2008, 130,
8570. (d) Qin, H.; Wenger, S.; Xu, M.; Gao, F.; Jing, X.; Wang, P.;
€
Zakeeruddin, S. M.; Gratzel, M. J. Am. Chem. Soc. 2008, 130, 9202.
(e) Tian, H.; Yang, X.; Chen, R.; Zhang, R.; Hagfeldt, A.; Sun, L.
J. Phys. Chem. C 2008, 112, 11023. (f) Ning, Z.; Zhang, Q.; Wu, W.;
Pei, H.; Liu, B.; Tian, H. J. Org. Chem. 2008, 73, 3791. (g) Choi, H.;
€
Baik, C.; Kang, S. O.; Ko, J.; Kang, M.-S.; Nazeeruddin, M. K.; Gratzel,
M. Angew. Chem., Int. Ed. 2008, 47, 327.
(11) Wang, Z.-S.; Koumura, N.; Cui, Y.; Takahashi, M.; Sekiguchi, H.;
Mori, A.; Kubo, T.; Furube, A.; Hara, K. Chem. Mater. 2008, 20, 3993.

1838 Chem. Mater., Vol. 22, No. 5, 2010
Fischer et al.
2 h. HCl (3 mL, 6 mol L-1) was added for hydrolysis, and the
mixture was stirred for 0.5 h. The mixture was neutralized with a
NaHCO3-solution and extracted with diethyl ether (3 ꢀ 20 mL).
The combined organic phases were washed with water and brine
and dried over Na2SO4. The solvent was removed by rotary
evaporation, and the residue was purified by column chromato-
graphy (SiO2, hexane:dichloromethane = 1:2) to provide carbal-
dehyde 6 (248 mg, 0.50 mmol, 78%) as a red oil. 1H NMR
(CDCl3, 400 MHz) δ ppm: 1.22 (t, 3J = 7.50 Hz, 3H, CH3), 1.25
(t, 3J = 7.50 Hz, 3H, CH3), 2.76 (q, 3J = 7.50 Hz, 2H, CH2), 2.84
(q, 3J = 7.50 Hz, 2H, CH2), 6.64 (d, 3J = 3.88 Hz, 1H, ThH), 6.97
(d, 3J = 3.92 Hz, 1H, ThH), 7.03-7.10 (m, 2H, PhH), 7.15-7.23
(m, 4H, PhH), 7.23 (d, 3J = 4.08 Hz, 1H, ThH), 7.27-7.32 (m,
4H, PhH), 7.69 (d, 3J = 3.99 Hz, 1H, ThH), 9.87, (s, 1H, CHO).
13C NMR (CDCl3, 100 MHz) δ ppm: 14.82, 15.21, 20.99, 21.26,
120.55, 122.75, 123.28, 124.84, 125.72, 128.04, 128.82, 129.19,
132.83, 136.88, 141.03, 141.81, 143.57, 146.33, 147.47, 152.04,
182.55. CI-mass 499.9 [MþH]þ (calcd. for C29H25NOS3 499.11)
Elemental analysis: calc. (%) for C29H25NOS3 C: 69.70; H: 5.04;
N: 2.80; found: C: 69.84; H: 5.23; N: 2.67.
2-Cyano-3-[5-diphenylamino-30,40-diethyl-2,20:50,200-terthien-500-
yl]acrylic Acid (1). A mixture of carbaldehyde 6 (200 mg, 0.4
mmol, 1.0 equiv) and cyanoacetic acid (102 mg, 1.2 mmol, 3 equiv)
was vacuum-dried, and acetonitrile (20 mL) was added together
with piperidine (16 μL, 0.2 mmol). The mixture was refluxed for 6 h.
After cooling, the solvent was removed in vacuum. The residue
wasthendissolvedindichloromethaneandwashedwithwaterand
brine and finally dried over Na2SO4. The solvent was removed by
rotary evaporation, and the crude product was purified by column
chromatography (SiO2, dichloromethane:methanol=3:1) to pro-
vide dye 1 (165 mg, 0.29 mmol, 73%) as a red solid. Mp: 245-
247 °C. 1H NMR (CDCl3, 500 MHz) δ ppm: 1.10 (t, 3J = 7.43,
3H), 1.12 (t, 3J = 7.43, 3H), 2.67 (q, 3J = 7.54 Hz, 2H), 2.73 (d,
3J = 7.54 Hz, 2H), 6.55 (d, 3J = 3.90 Hz, 1H, ThH), 6.89 (d, 3J =
3.87 Hz, 1H, ThH), 6.96-7.01 (m, 2H, PhH), 7.08 (d, 3J = 3.77
Hz, 1H, ThH), 7.10-7.14 (m, 4H, PhH), 7.19-7.24 (m, 4H, PhH),
7.61 (d, 3J = 3.95 Hz, 1H, ThH), 8.34 (s, 1H, C=CH). 13C NMR
(CDCl3, 125 MHz) δ ppm: 14.22, 14.52, 20.66, 20.94, 117.66,
120.64, 122.59, 123.06, 124.63, 125.47, 125.77, 128.68, 128.99,
129.22, 131.96, 136.05, 140.70, 142.53, 142.89, 147.66, 151.77,
164.01. CI-mass 567.1 [MþH]þ (calcd. for C32H26N2O2S3
566.12). Elemental analysis: calc. (%) for C32H26N2O2S3 C:
67.81; H: 4.62; N: 4.94; found: C: 67.69; H: 4.65; N: 4.85.
thiophene-2-amine 7 (1.50 g, 5.97 mmol, 1.0 equiv) was cooled to
-78 °C. n-BuLi (4.48 mL, 7.16 mmol, 1.60 mol L-1, 1.2 equiv)
was added dropwise to the above solution and stirred for 0.5 h
while keeping the temperature at around -78 °C. 2-Isopropy-
loxy-4,4,5,5-tetramethyl-1,3,2-dioxaborolane (1.39 g, 7.46 mmol,
1.25 equiv) was added dropwise at -78 °C. The reaction mixture
was allowed to warm to roomtemperature and was stirred for 1 h.
The reaction was quenched with a saturated NH4Cl-solution
(50 mL) and extracted three times with 50 mL of diethyl ether.
The combined organic extracts were washed with brine and dried
over Na2SO4. The solvent was removed after filtration under
reduced pressure. Boronic ester 8 (2.42 g, 5.96 mmol, 100%, 93%
GC-purity) was obtained as a greenish viscous oil, which solidi-
fied upon cooling and was used for further reactions without any
purification. 1H NMR (CDCl3, 400 MHz) δ ppm: 1.32 (s, CH3),
6.66 (d, 3J= 3.74 Hz, 1H), 7.03-7.09 (m, 2H, PhH), 7.15-7.20
(m, 4H, PhH), 7.24-7.29 (m, 4H, PhH), 7.41 (d, 3J= 3.75 Hz,
1H). 13C NMR (CDCl3, 100 MHz) δ ppm: 24.73, 83.82, 118.86,
122.29, 123.59, 123.66, 129.08, 129.24, 136.80, 147.70, 159.07. EI-
mass (m/z) 378 [MþH]þ, 377 ([Mþ] (calcd. for C22H24BNO2S
377.16).
N,N,N0000,N0000-Tetraphenyl-2,20:50,200:300,2000:5000,20000-quinquethio-
phene-5,50000-diamine (10). Branched terthiophene 9 (0.500 g, 1.0
mmol, 1.0 equiv), boronic ester 8 (0.830 g, 2.2 mmol, 2.2 equiv),
Pd2(dba)3 CHCl3 (21 mg, 0.02 mmol, 0.02 equiv), and HP-
3
(tBu)3BF4 (12 mg, 0.04 mmol, 0.04 equiv) were dissolved in
15 mL of well-degassed THF. A well-degassed aqueous K3PO4
solution (6.0 mL, 6.0 mmol, 1 mol L-1, 6.0 equiv) was added slowly.
The reaction mixture was stirred at room temperature for 5 h. Water
(20 mL) with a few drops of HCl (2 mol L-1) was added to the
mixture. The organic layer was separated, and the aqueous layer was
extracted three times with 10 mL of diethyl ether/ethyl acetate (1:1).
The combined organic extracts were washed with brine and dried
over Na2SO4. The solvents were removed under reduced pressure,
and the residue was purified by column chromatography (crude
product on silica gel, hexane:EtOAc = 19:1). The resulting pre-
cleaned product was finally purified by preparative HPLC (hexane:
DCM = 80:20) to provide diamine 10 (0.225 g, 3.01 mmol, 75%) as
a yellow-orange solid. Mp 79 °C. 1H NMR ([D8] THF, 400 MHz) δ
ppm: 6.57 (d, 3J = 3.90 Hz, 2H, ThH), 6.96 (d, 3J = 3.90 Hz, 1H,
ThH) 6.98-7.04 (m, 9H, PhH þ ThH), 7.11-7.14 (m, 8H, PhH þ
ThH), 7.19 (d, 3J = 5.30 Hz, 1H, ThH), 7.22-7.26 (m, 8H, PhH þ
ThH), 7.43 (d, 3J = 5.30 Hz, 1H, ThH). 13C NMR ([D8] THF, 100
MHz) δ ppm: 121.8, 122.1, 122.9, 123.3, 123.4, 123.4, 123.6, 123.6,
123.8, 123.9, 126.0, 128.1, 129.5, 129.8, 129.8, 130.3, 131.4, 131.5,
132.0, 132.7, 133.5, 136.3, 138.6, 139.9, 148.4, 148.5, 151.4, 151.8.
MS (MALDI-TOF): m/z 745.9 [M]þ (calc. for C44H30N2S5: 746.1).
Elemental analysis: calc. (%) for C44H30N2S5 C: 70.74; H: 4.05; N:
3.75; found: C: 70.92; H: 4.05; N: 3.64.
N,N-Diphenylthiophene-2-amine (7). The compound was pre-
pared according to the literature procedure.12,13 2-Bromothio-
phene (489 mg, 3.0 mmol, 1.0 equiv), diphenylamine (508 mg,
3.0 mmol, 1.0 equiv), and NaOtBu (317 mg, 3.3 mmol, 1.1 equiv)
were weighed into a round-bottom flask, and dry degassed
toluene (3 mL) was added. Pd2(dba)3 CHCl3 (31.1 mg, 30
3
5,50000-Bis(diphenylamino)-2,20:50,200:300,2000:5000,20000-quinquethio-
phene-500-carbaldehyde (11). N-BuLi (0.19 mL, 0.30 mmol, 1.6mol
L-1, 1.5 equiv) was added dropwise at -78 °C to a solution of
diamine 10 (150.0 mg, 0.20 mmol, 1.0 equiv) in 5 mL of dry THF,
and the mixture was stirred for 0.5 h at this temperature. Dry
dimethylformamide (94 μL, 1.20 mmol, 6.0 equiv) was added
dropwise via syringe. The reaction mixture was allowed to warm
to room temperature. Then HCl (3 mL, 6 mol L-1) was added for
hydrolysis, and the mixture was stirred for 0.5 h. The mixture was
neutralized with a NaHCO3-solution and then extracted with
diethyl ether (3 ꢀ 20 mL). The combined organic phases were
washed with brine and dried over Na2SO4. The solvent was
removed by rotary evaporation, and the residue was purified by
column chromatography (compound on silica, hexane:dichloro-
methane=1:1) to provide carbaldehyde 11 as a red oil. The pro-
μmol, 0.01 equiv) and HP(tBu)3BF4 (17.4 mg, 60 μmol, 0.02
equiv) were suspended in 3 mL of dry degassed toluene and
stirred for 10 min. The catalyst suspension was added to the
reactants to give a purple reaction mixture. The reaction was
stirred for 16 h at 105 °C. The solvent was removed under
reduced pressure. The product (415 mg, 1.7 mmol, 55%) was
obtained after column chromatography (neutral alumina,
hexane) as a colorless solid. The analytical data were the same
as in the literature.
N,N-Diphenyl-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-
thiophene-2-amine (8). A THF (15 mL) solution of N,N-diphenyl-
(12) Watanabe, M.; Yamamoto, T.; Nishiyama, M. Chem. Commun.
2000, 133.
(13) Hooper, M. W.; Utsunomiya, M.; Hartwig, J. F. J. Org. Chem.
2003, 68, 2861.

Article
Chem. Mater., Vol. 22, No. 5, 2010 1839
duct was redissolved in a small amount of THF and precipitated
into methanol. After concentration under reduced pressure and
filtration, the product (111 mg, 0.14 mmol, 74%) was obtained as
a red powder. Mp 85 °C. 1H NMR ([D8] THF, 400 MHz) δ ppm:
6.56 (d, 3J = 3.92 Hz, 1H, ThH), 6.58 (d, 3J = 3.89 Hz, 1H, ThH),
6.99-7.06 (m, 8H, PhH), 7.09 (d, 3J = 3.76 Hz, 1H, ThH),
7.11-7.15 (m, 8H, PhH þ ThH), 7.22-7.27 (m, 9H, PhHþ ThH),
7.90 (s, 1H ThH), 9.84 (s, 1H, CHO). 13C NMR ([D8] THF, 100
MHz) δ ppm: 121.3, 121.8, 123.3, 123.4, 123.5, 123.6, 123.7, 123.9,
124.1, 124.1, 129.4, 129.8, 129.8, 130.3, 130.5, 131.3, 132.2, 132.6,
134.4, 139.4, 139.7, 141.0, 141.5, 142.1, 148.3, 148.4, 151.8, 152.6,
182.6. MS (MALDI-TOF): m/z 774.2 [M]þ (calc. for
C45H30N2OS5: 774.1). Elemental analysis: calc. (%) for C45H30-
N2OS5 C: 69.73; H: 3.90; N: 3.61; found: C: 69.60; H: 3.96; N: 3.59.
dye and 10 mM chenodeoxycholic acid (cheno) as coadsorbant
in dichloromethane. Cells were sealed with a platinized FTO
counter electrode using a hot-melt (Surlyn, DuPont). Cells were
then filled with an electrolyte through a hole in the counter
electrode. The hole was then sealed with a Surlyn disk and a thin
glass to avoid leakage of the electrolyte. We compared the
photovoltaic performance using a volatile solvent-based elec-
trolyte (coded as Z960) and an ionic liquid electrolyte (coded as
Z952). The composition of the electrolytes was as follows: Z960:
1.0 M 1,3-dimethylimidazolium iodide, 0.03 M iodine, 0.1 M
guanidinium thiocyanate, 0.5 M tert-butylpyridine, 0.05 M LiI
in a mixture of acetonitrile and valeronitrile (85/15, v/v). Z952:
1,3-dimethylimidazoliumiodide/1-ethyl-3-methylimidazoliu-
miodide/1-ethyl-3methylimidazolium tetracyanoborate/iodine/
N-butylbenzoimidazole/guanidinium thiocyanate (molar ratio
12/12/16/1.67/3.33/0.67). The photoanode was covered with a
UV-cutoff/antireflecting polymer.
2-Cyano-3-[5,50000-bis(diphenylamino)-2,20:50,200:300,2000:5000,20000
-
quinquethien-500-yl]acrylic Acid (2). Aldehyde 11 (50.0 mg, 64.5
μmol, 1.0 equiv), cyanoacetic acid (16.5 mg, 194 μmol, 3.0 equiv),
and ammonium acetate (2.49 mg, 32.3 μmol, 0.5 equiv) were
suspendedinglacialacetic acid(5 mL) andstirred underrefluxfor
5 h. The mixture was cooled and extracted three times with
dichloromethane. The organic phase was washed with a NaH-
CO3-solution and then with brine and finally dried over Na2SO4.
The solvent was removed after filtration by rotary evaporation,
and the residue was purified by column chromatography (silica,
dichloromethane, and 5% glacial acetic acid) to provide 2 as a
dark red solid. The solid was redissolved in 2 mL of tetrahydro-
furane and precipitated into 15 mL of methanol under vigorous
stirring. The precipitate was sonicated for 1 h, filtered off, and
washed with a large amount water and subsequently with metha-
nol (50 mL). After drying under high vacuum product 2 (43.0 mg,
51.1 μmol, 79%) was obtained as a darkred solid. Mp: 249 °C. 1H
NMR ([D8]THF, 500 MHz) δ ppm: 6.48-6.60 (m, 2H, ThH),
6.90-7.07 (m, 9H, PhH þ ThH), 7.09-7.17 (m, 8H, PhH þ
ThH), 7.20-7.30 (m, 9H, PhH þ ThH), 7.83 (s, 1H, ThH), 8.32
(s, 1H, C = CH). 13C NMR ([D8]THF, 125 MHz) δ ppm: 100.5,
117.1, 121.2, 121.8, 123.3, 123.4, 123.4, 123.6, 123.8, 123.9, 124.1,
129.5, 129.8, 129.8, 130.5, 130.5, 131.3, 131.4, 131.4, 132.1, 132.2,
132.3, 141.6, 145.5, 148.3, 148.4, 148.4, 148.4, 163.9. MS
(MALDI-TOF): m/z 841.5 [M]þ (calc. for C48H31N3O2S5:
841.1). Elemental analysis: calc. (%) for C48H31N3O2S5 C:
68.46; H: 3.71; N: 4.99; found: C: 68.17; H: 3.90; N: 4.75.
For solid-state device fabrication, a spray pyrolysis technique
was used to coat the FTO conducting glass substrates (LOF
Industries, TEC 15 Ω/square, 2.2 mm thickness) with a thin
compact layer of TiO2 in order to prevent electron-hole recombi-
nation arising from direct contact between the hole-conductor
(spiro-OMeTAD) and the highly doped FTO layer. A 1.8 μm
mesoporous layer of the TiO2 nanoparticles with a typical diameter
of 20 nm was deposited by doctor-blading on top of this compact
layer. The TiO2 electrodes were stained by dipping in a dye solution
for 5 h consisting of 0.3 mM dye and 10 mM cheno in dichlor-
omethane. The spiro-OMeTAD solution (137 mM in chloro-
benzene) contained final concentrations of 112 mM tert-butylpyr-
idine and 21 mM Li[CF3SO2]2N (added from highly concentrated
acetonitrile solutions). Finally, a gold contact (100 nm) was
deposited on the organic semiconductor film by evaporation
(EDWARDS AUTO 500 Magnetron Sputtering System).
Photovoltaic Characterization and Stability Tests. A 450 W
xenon light source (Oriel, USA) was used to give an irradiance of
100 mW cm-2 (the equivalent of one sun at air mass global, AM
1.5G, at the surface of solar cells). The spectral output of the
lamp was matched in the region of 350-750 nm with the aid of a
€
Schott K113 Tempax sunlight filter (Prazisions Glas & Optik
GmbH, Germany) to reduce the mismatch between the simu-
lated and true solar spectra to less than 2%. The current-
voltage characteristics of the cells were obtained by applying an
external potential bias to the cell and measuring the generated
photocurrent with a Keithley model 2400 digital source meter
(Keithley, USA). The effective area of the devices was defined
with the use of a metal mask to be 0.159 cm2.
Optical and Voltammetric Measurements. Optical measure-
ments were carried out in 1 cm cuvettes with Merck Uvasol
grade solvents, and absorption spectra were recorded on a
Perkin-Elmer Lambda 19 spectrometer. Cyclic voltammetry
experiments were performed with
a computer-controlled
EG&G PAR 273 potentiostat in a three-electrode single-com-
partment cell with a platinum working electrode, a platinum
wire counter electrode, and an Ag/AgCl reference electrode. All
potentials were internally referenced to the ferrocene/ferroce-
nium couple. The concentration of molecules adsorbed on the
TiO2 film (“dye loading”) was calculated using the Lambert-
Beer law and the molar extinction coefficient of the dye in
solution.
A similar data acquisition system was used to control the
incident photon-to-current conversion efficiency (IPCE) mea-
surement. Under computer control, light from a 300 W xenon
lamp (ILC Technology, USA) was focused through a Gemini-
180 double monochromator (Jobin Yvon Ltd., UK) onto the
photovoltaic cell under test. A computer-controlled mono-
chromator was incremented through the spectral range (300-
900 nm) to generate a photocurrent action spectrum with a
sampling interval of 10 nm and a current sampling time of 5 s.
For stability tests, devices were irradiated at open-circuit under
a Suntest CPS plus lamp (ATLAS GmBH, 100 mW cm-2) in
ambient air at 60 °C.
Transient Photovoltage and Photocurrent Decay Measure-
ments. Transient decays were measured with different white
light steady-state biases and a superimposed red light perturba-
tion pulse (0.05 s square pulse width, 100 ns rise and fall time),
incident on the photoanode side of the test device. The white and
red lights were supplied by light-emitting diodes. The voltage
Device Fabrication. Dyes were tested in photovoltaic devices
using standard mesoporous TiO2 films with a 8 μm thick
transparent layer of 20 nm sized particles and a 5 μm thick
scattering layer of 400 nm sized particles on top of fluorine-
doped tin oxide (FTO) coated glass. The detailed fabrication
procedures for the TiO2 pastes and film have been described
elsewhere.14 Films were immersed for 5 h in a solution of 0.3 mM
(14) Wang, P.; Zakeeruddin, S. M.; Comte, P.; Charvet, R.; Humphry-
€
Baker, R.; Gratzel, M. J. Phys. Chem. B 2003, 107, 14336.

1840 Chem. Mater., Vol. 22, No. 5, 2010
Fischer et al.
Scheme 1. Synthesis of Dye 1a
a a) NBS, DMF, 0 °C to RT, 70%. b) Diphenylamine, [Pd2(dba)3] CHCl3, HP(tBu)3BF4, NaOtBu, toluene, 100 °C, 82%. c) 1. n-BuLi, -78 °C, THF,
3
1 h, 2. DMF, 2 h, 3. HCl, 78%. d) Cyanoacetic acid, piperidine, acetonitrile, 80 °C, 4 h, 73%.
Scheme 2. Synthesis of Dye 2a
a a) 1. [Pd2(dba)3] CHCl3, HP(tBu)3BF4, NaOtBu, toluene, 16 h, 105 °C, 55%. b) 1. n-BuLi, -78 °C, THF, 1 h, 2. 2-Isopropyloxy-4,4,5,5-tetramethyl-
3
1,3,2-dioxaborolane, 1 h, 93%. c) [Pd2(dba)3] CHCl3, HP(tBu)3BF4, THF, 5 h, 25 °C. d) 1. n-BuLi, -78 °C, THF, 1 h, 2. DMF, 2 h, 3. HCl, 74%.
e) Cyanoacetic acid, NH4OAc, glacial acetic acid, reflux, 5 h, 79%.
3
dynamics were recorded at open-circuit on a PC-interfaced
Keithley 2400 source meter with a 500 μs response time. The
perturbation light source was set to a suitably low level for the
voltage decay kinetics to be monoexponential. By varying the
white light bias intensity, the recombination rate constant could
be estimated at open-circuit potentials of the device (corres-
ponding to different free charge densities in the TiO2 film). The
chemical capacitance (C) at the TiO2/electrolyte interface was
calculated from the amount of injected charge by the red pulse
(ΔQ) and the associated increase in photovoltage (ΔV) using
C=ΔQ/ΔV. The injected charge was obtained from integration
of the short-circuit photocurrent transient. The density of trap
states (DOS) was calculated from DOS = C/[ed(1-p)], where e is
the elementary charge, C is the capacitance in F cm-2, d is the
film thickness, and p is the porosity of the film (68%).
imide in dimethylformamide afforded bromoterthiophene 4
in 70% yield (Scheme 1). Cross-coupling of 5-bromo-
terthiophene 4 and diphenylamine by Pd-catalyzed Hart-
wig-Buchwald reaction12,13 gave amine 5 in 82% yield.
Deprotonation of the free R-position of 5 using n-butyl
lithium (n-BuLi) and subsequent reaction of the lithium
species with dimethylformamide furnished corresponding
aldehyde 6 in 78% yield. Finally, linear dye 1 was prepared
in 73% yield by a Knoevenagel condensation of aldehyde 6
with cyanoacetic acid using piperidine as catalyst.
Branched dye 2 was synthesized starting from 2-bro-
mothiophene and diphenylamine, which were converted
to N,N-diphenyl-thiophene-2-amine 7 via a Hartwig-
Buchwald reaction. Deprotonation of the free R-position
at the thiophene ring of 7 with n-BuLi and subsequent
reaction with 2-isopropoxy-4,4,5,5-tetramethyl-1,3,2-di-
oxaborolane as electrophile afforded boronic ester 8
in 93% yield (Scheme 2). A 2-fold Suzuki-type cross-
coupling reaction of branched diiodinated terthiophene
Results and Discussion
Synthesis. Linear and branched dyes L-3T-DPA 1 and
B-5T-DPA 2 were synthesized as follows (Schemes 1 and 2).
Monobromination of terthiophene 3using N-bromosuccin-

Article
Chem. Mater., Vol. 22, No. 5, 2010 1841
915 with boronic ester 8 gave peripheral-functionalized
dendritic oligothiophene 10 in 75% yield. Aldehyde 11
was prepared in 74% yield by lithiation of 10 at the core
and quenching of the metalated species with anhydrous
dimethylformamide. Knoevenagel condensation of alde-
hyde 11 and cyanoacetic acid under acidic conditions
furnished corresponding dye 2 in 79% yield.
Photophysical Properties. Absorption and emission
spectra of the two dyes in dichloromethane solution are
displayed in Figure 1, and the corresponding data are
presented in Table 1. Dye 1 showed an absorption maxi-
mum in the visible region at 450 nm with a molar
extinction coefficient of 19600 M-1 cm-1, arising from
a charge transfer (CT) transition. On the other hand, dye
2 showed a CT transition at 503 nm, which compared to
dye 1 is 53 nm red-shifted but somewhat weaker (16600
M-1 cm-1). The molar extinction coefficients of the
transitions at higher energy (390 nm, 300 nm) are appro-
ximately doubled for branched dye 2 compared to linear
dye 1 (386 nm, 295 nm). The shape of the absorption
spectra of 1 and 2 are comparable to similar dyes where
the thiophenes adjacent to the diphenylamine moieties
are replaced by phenyl rings.9c Emission maxima of dyes 1
and 2 can be found at 694 and 791 nm, respectively, when
excited at their respective CT bands. A large Stoke’s shift
was observed for both dyes 1 (7813 cm-1) and 2 (7238
cm-1), which is an indication of significant structural
changes between the ground and the excited state
and, accordingly, specify a photoinduced charge transfer
process.
A negative solvatochromism was observed for the
charge transfer band by increasing polarity of the solvent
(Figure 2a,b), indicating that the dyes possess smaller
dipole moments in the excited state than in the ground
state. This is due to the fact that in polar solvents the
electron-withdrawing power of the carboxylic acid de-
creases because of its partial deprotonation in the excited
state. This information was further supported by the blue-
shift (about 20 nm) of the CT band upon addition of
triethylamine to dye solutions in THF or 1,4-dioxane. A
similar blue-shift of the MLCT (metal-to-ligand charge-
transfer) transition from 600 to 570 nm has been observed
for bis(tetrabutylammonium)trithiocyanato(4,40,400-tri-
carboxy-2,20:60,200-terpyridine)ruthenium(II) (known as
black dye) by increasing the pH of the solution and is
ascribed to deprotonation of the carboxylic acid group.3
Figure 2c shows absorption spectra of both dyes ad-
sorbed on the surface of transparent 5 μm thick TiO2
films. Absorption of the blank TiO2 film was subtracted
from the curve. Absorption spectra of both dyes were
∼40-50 nm blue-shifted in the onset region when an-
chored at the TiO2 surface compared to solution spectra.
The stronger blue-shift of the CT-band of both dyes on
TiO2 films is attributed to π-stacking interactions or
formation of aggregates on the TiO2 surface. A blue-shift
is commonly observed in the spectral response of organic
Figure 1. UV-vis and normalized emissionspectra ofdye 1 (redtriangle)
and 2 (blue circle) in dichloromethane. The emission spectra are normal-
ized to their respective CT bands.
dyes adsorbed on the surface of TiO2 compared to the
solution, which is due to the electronic interactions of the
dye to the semiconductor surface.17 The concentration of
the branched dye (0.11 M) adsorbed on the film is about
half the concentration of the linear dye (0.23 M), which
we attribute to impaired surface packing due to the large
size of the branched molecule.
The observed CT character in solution was corrobo-
rated by semiempirical calculations. The quantum-che-
mical Austin Model 1 (AM1) method under restricted
Harthree Fock conditions was used to analyze the elec-
tron distribution of the frontier orbitals of both dyes
(Figure 3). The electron density distribution of the highest
occupied molecular orbitals (HOMO) of both molecules
is mainly located at the oligothiophene and diphenyl-
amine moieties, whereas electron density of the lowest
unoccupiedmolecularorbital (LUMO) is primarily located
at the cyanoacrylic acid acceptor and the neighboring
thiophene ring. Hence, the presence of strong electron
density relocation between HOMO and LUMO is present
and supports the occurrence of an intramolecular charge
transfer transition (ICT) in the UV-vis spectra. Further-
more, the calculated electron distribution shows favorable
directionality for injection of photoexcited electrons from
the donor group to the TiO2 film via the bridge and
anchoring group.
Redox Properties. The energetic alignment of HOMO
and LUMO energy levels is crucial for an efficient opera-
tion of the sensitizer in a DSSC. To ensure efficient
electron injection from the excited dye into the conduc-
tion band of TiO2 films, the LUMO level must be higher
in energy than the TiO2 conduction band edge. The
HOMO level of the dye must be lower in energy than
the redox potential of the I-/I3 redox couple for effi-
-
cient regeneration of the dye cation after photoinduced
electron injection into the TiO2 film. The electrochemical
(16) Johansson, T.; Mammo, W.; Svensson, M.; Andersson, M. R.;
€
Inganas, O. J. Mater. Chem. 2003, 13, 1316.
(17) (a) Xu, M.; Wenger, S.; Bala, H.; Shi, D.; Li, R.; Zhou, Y.;
€
Zakeeruddin, S. M.; Gratzel, M.; Wang, P. J. Phys. Chem. C
2009, 113, 2966. (b) Zhang, G.; Bai, Y.; Li, R.; Shi, D.; Wenger, S.;
€
(15) Mishra, A.; Ma, C.-Q.; Janssen, R. A. J.; Bauerle, P. Chem. Eur. J.
€
Zakeeruddin, S. M.; Gratzel, M.; Wang, P. Energy Environ. Sci. 2009,
2009, 15, 13521.
2, 92.

1842 Chem. Mater., Vol. 22, No. 5, 2010
Table 1. Optical and Electrochemical Data of Sensitizers L-3T-DPA 1 and B-5T-DPA 2 in Dichloromethane
Fischer et al.
dyes
λab [nm] (ε [L mol-1 cm-1])
λemmax [nm]
Egopt [eV]
E0ox1 [V] b,c
E0ox2 [V]b,c
HOMO [eV]d
LUMO [eV]e
L-3T-DPA (1)
295 (13700)
386 (13200)a
450 (19600)
300 (25800)
390 (31200)
503 (16000)
694
737 (sh)
2.14
0.18
0.56
-5.19
-3.05
B-5T-DPA (2)
680 (sh)
791
1.94
0.18
0.65
-5.19
-3.25
a Only visible as shoulder. b Measured in DCM/TBAHFP (0.1 M), c ≈ 1 10-3 mol L-1, 295 K, scan rate = 100 mV s-1, vs Fcþ/Fc. c Values
3
determined by cyclic voltammetry measurement. d Calculated from Eonsetox1 set EHOMO(Fcþ/Fc) = -5.1 eV vs vacuum.16 e Determined from the optical
band gap.
The first oxidation potential (E0ox1) for both com-
pounds was found at 0.18 V (vs Fcþ/Fc) and is assigned
to the oxidation of the diphenylthienyl-amine moiety.18
The HOMO energy levels were calculated from the onset
of the first oxidation wave (Eonsetox1 = 0.09 V vs Fc/Fcþ)
and resulted for both dyes to -5.19 eV. The second
oxidation potential was observed at 0.56 and 0.65 V for
1 and 2, respectively, and is assigned to the oxidation of
the oligothiophene units. The LUMO levels of both
sensitizers were calculated from the difference between
HOMO energies and UV-vis absorption onset values. It
is interesting to note that changing the donor from
triphenylamine to diphenylthienyl-amine decreases the
oxidation potential by about 290 mV, as the electron
density is increased from phenyl to the more electron-rich
thiophene.9c
The HOMO-LUMO levels of both dyes are suitable
for DSSCs. The LUMO energy levels (1: -3.05 eV; 2:
-3.25 eV) are sufficiently higher than the conduction
band edge of TiO2 (-4.0 eV). Hence, an effective electron
transfer from the excited dye to the TiO2 is ensured. The
HOMO energy levels of -5.19 eV for both dyes are about
360 meV lower in energy than the standard potential of
the I3-/I- redox couple (-4.83 eV vs vacuum),7b and thus
a sufficient driving force for the regeneration of the
oxidized dye is available.
Photovoltaic Performance and Device Stability. Photo-
voltaic tests were conducted to evaluate the potential of
the linear and branched sensitizers in dye-sensitized solar
cells (see the Experimental Section for detailed cell
compositions). The incident photon-to-current conver-
sion efficiencies (IPCE) of devices with dye 1 and 2 using
an acetonitrile-based electrolyte are shown in Figure 5a.
The IPCE of linear dye 1 exceeds 80% in the spectral
region between 500 and 550 nm. Considering the light
absorption and scattering loss by the conducting glass,
the absorbed photon-to-current conversion efficiency
(APCE) is close to unity in this region. In contrast, the
IPCE of branched dye 2 has a maximum of 67% at 510 nm.
The estimated maximum IPCE with dye 2, assuming an
APCE of unity, Lambert-Beer-type absorbance in the TiO2
Figure 2. Normalized absorption spectra of (a) linear (1) and (b)
branched (2) dye in different solvents. (c) Absorption spectra on trans-
film, and accounting for reflectance and absorbance in the
conducting glass, is about 10% higher than the measured
IPCE. This loss is probably caused by reduced charge
collection efficiency: The lower dye loading favors increased
parent 5 μm thick TiO2 films (blank TiO2 film is subtracted from
measurement).
behavior of the dyes was investigated by cyclic voltam-
metry, and the data are listed in Table 1. Reversible redox
waves were obtained for both dyes in the oxidative region
(Figure 4).
(18) (a) Rohde, D.; Dunsch, L.; Tabet, A.; Hartmann, H.; Fabian, J.
J. Phys. Chem. B 2006, 110, 8223. (b) Su, Y. Z.; Lin, J. T.; Tao, Y.-T.;
Ko, C.-W.; Lin, S.-C.; Sun, S.-S. Chem. Mater. 2002, 14, 1884.

Article
Chem. Mater., Vol. 22, No. 5, 2010 1843
Figure 3. Frontier orbitals (AM1-calculation) of L-3T-DPA 1 (linear) and B-5T-DPA 2 (branched) (carbons in black, nitrogens in blue, oxygens in red,
sulfurs in yellow, and hydrogens are omitted for clarity reasons).
24 h of light soaking, we obtained stabilized values of
5.9% for dye 1 and 2.6% for dye 2. The markedly lower
photocurrent of branched dye 2 with ionic liquid electro-
lyte compared to volatile electrolyte is probably due to the
large recombination current at the uncovered TiO2 sur-
face, since the ionic liquid electrolyte contains a high
concentration of triiodide compared to the volatile elec-
trolyte (about 0.24 M vs 0.03 M).19 In solid-state devices
with spiro-MeOTAD as a hole transporting material we
measured efficiencies of 3.0% with dye 1 and 2.6% with
dye 2 (Figure 6). Since it is difficult to fill the pores with
the hole conductor, only thin films of TiO2 (here 1.8 μm)
can be used, which limit Jsc accordingly. However, Voc is
Figure 4. Cyclic voltammograms of 1 (red triangle) and 2 (blue circle);
DCM/TBAHFP (0.1 M), c ≈ 1 ꢀ 10-3 mol L-1, 295 K, scan rate =
enhanced by about 200 mV since the redox level of the
hole conductor is situated at a more negative energy
(about -4.9 eV) than the redox level of the iodide/tri-
iodide couple (about -4.83 eV). Jsc of the device with
linear dye 1 is around 1.7 mA cm-2 higher than that with
branched dye 2. This result is consistent with the results of
the liquid devices, where the adsorbed amount of dye 2 on
the surface of TiO2 is lower than for dye 1 resulting in
lower Jsc values.
The stability of devices with ionic liquid electrolyte was
tested over 1000 h at full sunlight and 60 °C. The evolu-
tion of the photovoltaic parameters is shown in Figure 7.
The parameters were normalized to stabilized values after
24 h of light soaking, which allowed for optimal reorga-
nization of the dye molecules on the TiO2 surface. The cell
with the linear dye 1 showed more rapid degradation of
performance, retaining only 73% of its initial efficiency
after 1000 h. We attribute this degradation to desorp-
tion of dye molecules, which induces a loss of Jsc and
100 mV s-1, vs Fcþ/Fc.
charge recombination between free electrons in the TiO2
conduction band and oxidized species (tri-iodide) in the
electrolyte at the uncovered TiO2 surface. This hypothesis
was confirmed with photovoltage decay experiments (see the
section on Charge Recombination).
The photovoltaic parameters, i.e. conversion efficiency
(η), short-circuit photocurrent density (Jsc), open-circuit
photovoltage (Voc), and fill factor (FF) under full sunlight
(AM 1.5G, 100 mW cm-2) of devices with volatile, ionic
liquid electrolyte, and a solid hole conductor are sum-
marized in Table 2. The photocurrent-voltage curves of
the devices with volatile electrolyte are shown in
Figure 5b. With the volatile electrolyte we measured a
conversion efficiency of 6.8% for dye 1 and 5.0% for dye
2. The lower Jsc of the device with branched dye 2 is
consistent with the lower IPCE. The 40 mV lower Voc can
be explained with increased charge recombination via the
exposed TiO2 surface not covered with dye as discussed
in more detail later. With the ionic liquid electrolyte, after
(19) Bai, Y.; Cao, Y.; Zhang, J.; Wang, M.; Li, R.; Wang, P.; Zakeeruddin,
€
S. M.; Gratzel, M. Nat. Mater. 2008, 7, 626.

1844 Chem. Mater., Vol. 22, No. 5, 2010
Fischer et al.
Figure 5. Incident photon-to-current conversion efficiency (a) and current density-voltage curve under full sunlight (AM 1.5G, 100 mW cm-2) and in the
dark (b) of devices sensitized with L-3T-DPA 1 (red triangle) and B-5T-DPA 2 (blue circle) using Z960 electrolyte.
Table 2. Photovoltaic Parameters of Devices with Sensitizers L-3T-DPA 1
and B-5T-DPA 2 in Liquid, Ionic Liquid, and Solid State DSCs at Full
Sunlight (AM 1.5G, 100 mW cm-2
)
Voc
Jsc
η
dye
redox mediator
(mV) (mA cm-2) FF (%)
1 L-3T-DPA volatile electrolyte (Z960) 659
619
-13.5
-11.0
0.76 6.8
0.74 5.0
2 B-5T-DPA
1 L-3T-DPA ionic liquid (Z952)
2 B-5T-DPA
659
570
-11.9
-6.1
0.75 5.9
0.75 2.6
1 L-3T-DPA solid hole conductor
2 B-5T-DPA
834
827
-6.7
-5.0
0.54 3.0
0.62 2.6
associated Voc. In contrast, under similar conditions, the
cell with branched dye 2 showed excellent photochemical
and thermal stability, retaining 96% of its initial efficiency
after 1000h. The higherstabilityof the brancheddyecanbe
ascribedtotwo reasons:1. thebulkystructureofdye, which
probably protect the TiO2 surface and hinders desorption,
and 2. suppression of electron transfer from TiO2 to the
electrolyte (i.e., dark current). This result means that the
initial disadvantage of branched dye 2 is nearly ruled out.
Charge Recombination. We measured photovoltage
decay transients of devices with the volatile electrolyte
at various white light bias intensities in order to study
charge recombination rates between electrons in the TiO2
conduction band and tri-iodide ions in the electrolyte.20
The recombination rates are plotted in Figure 8a versus
the open-circuit potential (Voc) of the device adjusted by
varying the white light bias. Since Voc in the device is given
by the difference between the redox level of the electrolyte
and the quasi-Fermi level of TiO2, which is determined by
the concentration of free charge carriers, this plot allows
us to compare the recombination rate at equal charge
density concentrations in the TiO2 films. The recombina-
tion rate is known to increase exponentially with increas-
ing bias light due to a filling of intraband trap states in the
TiO2, which allows a faster detrapping of electrons to the
conduction band and subsequent recombination with tri-
iodide.21 We measured an approximate 4-fold increase in
recombination rate for the device with branched dye 2
Figure 6. Current density-voltage curve under full sunlight (AM 1.5G,
100 mW cm-2) and in the dark of devices sensitized with L-3T-DPA 1 and
B-5T-DPA 2 using a solid hole conductor. TiO2 film thickness 1.8 μm,
overnight dipping, 0.4 cm2 active surface area.
compared to linear dye 1, which we attribute to an
increased recombination current via the larger fraction
of uncovered TiO2 surface. The chemical capacitance of
the TiO2/electrolyte interface in devices with dye 1 or 2 is
plotted vs Voc in Figure 8b. The chemical capacitance is
proportional to the density of trap states (DOS) at the
quasi-Fermi level corresponding to that Voc and com-
monly shows an exponential distribution.22 We found a
lateral shift to the left of about 25 mV of the DOS curve of
the device with dye 2. This shift is most likely due to a
change in the surface charge on the uncovered TiO2
surface, i.e. a positive charging by free protons in the
electrolyte, which induces a downward shift of the con-
duction band of the TiO2 toward the iodide/tri-iodide
potential. Both the higher recombination rate and the
downward shifted conduction band in the device with dye
2 explain the lower Voc measured in the current-voltage
curves. Additionally, the higher charge recombination
rate reduces the charge collection efficiency and conse-
quently the Jsc and IPCE.
Conclusion
We have synthesized two D-π-A organic dyes compris-
ing linear and branched oligothiophenes for application
(20) Duffy, N. W.; Peter, L. M.; Wijayantha, K. G. U. Electrochem.
Commun. 2000, 2, 262.
(21) Bisquert, J.; Vikhrenko, V. S. J. Phys. Chem. B 2004, 108, 2313.
(22) Bisquert, J. Phys. Chem. Chem. Phys. 2003, 5, 5360.

Article
Chem. Mater., Vol. 22, No. 5, 2010 1845
Figure 7. Photovoltaic parameters of devices using the linear dye 1 (a) and the branched dye 2 (b) during a 1000 h accelerated stability test (full sunlight,
60 °C). Parameters are normalized to stabilized values after 24 h of light soaking.
Figure 8. Comparison of (a) charge recombination kinetics and (b) chemical capacitance and density of trap states (DOS) of dye 1 (red) and 2 (blue) in
devices with the branched and linear dye.
in dye-sensitized solar cells. In a nonvolatile electrolyte,
IPCE of the linear dye 1 exceeding 80% in the spectral
region between 500 and 550 nm was observed with an
overall conversion efficiency of 6.8%. The lower effi-
ciency in the case of branched dye 2 is ascribed to the
reduction of the photocurrent due to less dye loading on
the TiO2 surface. A moderate decrease in the efficiency
was observed for the linear dye in ionic liquid devices,
whereas for the branched dye the efficiency was reduced
to half. The performance of both dyes in solid-state
devices is nearly similar. An approximate 4-fold increase
in the recombination rate was measured for the device
with branched dye 2 compared to linear dye 1 in volatile
electrolyte, which was attributed to an increased recom-
bination current via the larger fraction of uncovered TiO2
surface. Long-term stability measurements of devices
with ionic liquid electrolyte were performed over 1000 h
at full sunlight and 60 °C showing higher stability for the
cell comprising branched dye. This result shows that the
initial disadvantage of branched dye 2 is nearly ruled out.
The higher durability of the branched dye makes this
molecular design an interesting option in further devel-
opment of new molecules, where compromise on the dye
loading should be minimized.
Acknowledgment. We thank Pascal Comte for TiO2 film
€
preparation and Carole Gratzel for fruitful discussions.
S.W., M.W., S.M.Z., and M.G. thank the Swiss National
Science Foundation for financial support and A.M, M.K.R.F,
and P.B thank the German Ministry of Education and
Research (BMBF) for financial support in the frame of project
OPEG 2010.