Unravelling the potential for dithienopyrrole sensitizers in dye-sensitized solar cells

Article abstract of DOI:10.1021/cm401144j

Two D-π-A dyes based on the dithieno[3,2-b:2′,3′-d]pyrrole π-bridge (DTP) were synthesized, characterized using UV-vis absorption spectroscopy and electrochemistry, modeled using quantum chemical calculations, and used as sensitizers in dye-sensitized solar cells (DSCs). The photoelectrochemical properties and DSC performance are thoroughly compared with their cyclopenta[1,2-b:5,4-b′]dithiophene (CPDT) analogues. The use of DTP results in a small increase in the zero-zero transition energy reflecting the higher lying lowest unoccupied molecular orbital that is commonly reported for DTP relative to CPDT systems. This increased optical gap manifests in slightly blue-shifted incident photon-to-collected electron conversion efficiency (IPCE) responses; however, increased open-circuit photovoltage values and improved charge-transfer kinetics relative to the CPDT systems result in comparable power conversion efficiencies. The present report highlights the potential of DTP for the development of tailored sensitizers employing stronger acceptors.

Full text of DOI:10.1021/cm401144j

Article
pubs.acs.org/cm
Unravelling the Potential for Dithienopyrrole Sensitizers in Dye-
Sensitized Solar Cells
Lauren E. Polander,*,† Aswani Yella,† Joel Teuscher,† Robin Humphry-Baker,† Basile F. E. Curchod,‡
̈
Negar Ashari Astani,‡ Peng Gao,† Jacques-E. Moser,§ Ivano Tavernelli,‡ Ursula Rothlisberger,‡
Michael Gratzel,† Md. Khaja Nazeeruddin,*,† and Julien Frey†
̈
†Laboratory of Photonics and Interfaces, ‡Laboratory of Computational Chemistry and Biochemistry, and §Group for Photochemical
́
Dynamics, Institute of Chemical Sciences and Engineering, School of Basic Sciences, Ecole Polytechnique Fed
CH-1015 Lausanne, Switzerland
́ ́
erale de Lausanne,
S
* Supporting Information
ABSTRACT: Two D−π−A dyes based on the dithieno[3,2-b:2′,3′-d]pyrrole
π-bridge (DTP) were synthesized, characterized using UV−vis absorption
spectroscopy and electrochemistry, modeled using quantum chemical
calculations, and used as sensitizers in dye-sensitized solar cells (DSCs).
The photoelectrochemical properties and DSC performance are thoroughly
compared with their cyclopenta[1,2-b:5,4-b′]dithiophene (CPDT) analogues.
The use of DTP results in a small increase in the zero−zero transition energy
reflecting the higher lying lowest unoccupied molecular orbital that is
commonly reported for DTP relative to CPDT systems. This increased
optical gap manifests in slightly blue-shifted incident photon-to-collected
electron conversion efficiency (IPCE) responses; however, increased open-
circuit photovoltage values and improved charge-transfer kinetics relative to the CPDT systems result in comparable power
conversion efficiencies. The present report highlights the potential of DTP for the development of tailored sensitizers employing
stronger acceptors.
KEYWORDS: dye-sensitized solar cells, D−π−A, dithienopyrrole
(DTP) is more accessible (up to 80% yield in one step) as
showcased by Rasmussen et al.11,12 Notably, N-substituted
DTP possesses similar electronic properties with respect to
CPDT13 but also has the potential to incorporate a variety of
N-substituents. As a result, this position offers greater versatility
toward the control of dye aggregation, which makes it
particularly attractive for DSC applications.14,15
The only direct comparison between CPDT and DTP π-
bridges in sensitizers for DSCs concluded that the latter
exhibits (i) faster recombination of electrons injected into TiO2
to the electrolyte and (ii) slower regeneration of the oxidized
dye by the electrolyte.16 These issues can be addressed through
the insulation of the TiO2 surface and by choosing the
appropriate cobalt(II/III) redox mediator, respectively.17
Herein, we demonstrate the full potential for DTP as an
effective π-bridge using the two principles outlined above.
Although a variety of primary amines are compatible with the
DTP synthesis, a branched N-alkyl substitution was selected in
this study to mimic the steric properties of the CPDT bridge.
The LP225 and LP227 sensitizers were modeled after Y123
and JF419; they incorporate dialkyoxy-substituted biphenyl
INTRODUCTION
■
Over the past two decades, considerable research efforts have
been devoted to dye-sensitized solar cells (DSCs) due to their
potential as a low-cost photovoltaic technology.1 The basic
working principle, which relies on a delicate energetic balance
between the anode, sensitizer, and electrolyte, is comprehen-
sively described in a review by Hagfeldt et al.2 The most
efficient cells are based on relatively expensive or synthetically
challenging ruthenium(II) and porphyrin sensitizers that have
provided efficiencies up to 11.7% and 11.9%, respectively.3,4
More accessible organic dyes take the form of a D−π−A
structure, which is reminiscent of push−pull chromophores
used in nonlinear optics.5 In this context, oligothiophenes and
their fused derivatives are intensively exploited as π-bridges.6
The two highest performing organic dyes, Y123 and JF419
(Chart 1), incorporate cyclopenta[1,2-b:5,4-b′]dithiophene
(CPDT), reaching power conversion efficiencies of up to
10.3%.7,8 The success of CPDT is attributed to (i) the
advantageous steric effect of the alkyl groups placed in the 4,4-
position, which helps to prevent aggregation of the dyes on the
surface, and (ii) the effective electronic communication
between the donor (D) and acceptor (A), which enables
efficient charge transfer.
Received: April 10, 2013
Revised: June 1, 2013
Contrary to the CPDT bridge, which is synthetically costly
(35% yield over five steps),9,10 dithieno[3,2-b:2′,3′-d]pyrrole
© XXXX American Chemical Society
A
dx.doi.org/10.1021/cm401144j | Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
Chart 1. Structures of DTP-Based Dyes, LP225 and LP227, along with Their CPDT-Based Analogues, Y123 and JF419
(D1)6,17 and fluorene (D2)8 donors that are known to reduce
and matrix-assisted laser desorption ionization (MALDI) mass
́
spectrometry data were collected at the Ecole Polytechnique Fed
́ ́
erale
charge recombination in the device by preventing the oxidized
redox mediator to access the TiO2 surface (Chart 1).
Additionally, the dyes were tested in DSCs using a mixture of
tris(2,2′-bipyridine)cobalt(II) and tris(2,2′-bipyridine)cobalt-
(III) complexes ([CoII/III(bpy)3]2+/3+, see Experimental Details
for full electrolyte composition), which provides ample driving
force for oxidized dye regeneration, resulting in quantitative
regeneration yields and efficiencies up to 8.7%.18
de Lausanne (EPFL).
Compound 2. A solution of 1 (0.250 g, 0.771 mmol), flame-dried
sodium tert-butoxide (0.222 g, 2.31 mmol), tris-
(dibenzylideneacetone)dipalladium (Pd2(dba)3, 0.018 g, 0.019
mmol), and 2,2′-bis(diphenylphosphino)-1,1′-binaphthyl (BINAP,
0.048 g, 0.077 mmol) in dry toluene (1.5 mL) was purged with
nitrogen. After 20 min, tridecan-7-amine (0.154 g, 0.771 mmol) was
added, and the mixture was stirred overnight at 120 °C. Upon
completion by TLC (18 h), the reaction mixture was diluted with
hexanes and filtered through silica gel eluting with hexanes followed by
2% ethylacetate in hexanes. The solvent was removed under reduced
pressure, and the crude product was purified by flash chromatography
(silica gel, 2% ethylacetate in hexanes) to yield slightly yellow solid
(0.206 g, 0.571 mmol, 74%). 1H NMR (400 MHz, CDCl3) δ 7.08 (d, J
= 5.6 Hz, 1H), 7.00 (d, J = 5.2 Hz, 1H), 4.19 (tt, J = 10.0, 5.0 Hz, 1H),
2.08−1.92 (m, 2H), 1.88−1.72 (m, 2H), 1.30−1.06 (m, 14H), 1.06−
0.92 (m, 2H), 0.79 (t, J = 7.0 Hz, 6H). 13C{1H} NMR (100 MHz,
CDCl3) δ 144.1, 122.3, 114.7, 111.8, 59.8, 35.1, 31.5, 28.9, 26.5, 22.5,
14.0. HRMS (EI): m/z calcd for C21H31NS2 (M+): 361.1898. Found:
361.1896. Anal. Calcd. for C21H31NS2: C, 69.75; H, 8.64; N, 3.87.
Found: C, 69.75; H, 8.71; N, 3.83.
EXPERIMENTAL DETAILS
■
Materials. Starting materials were reagent grade and were used
without further purification. Solvents were obtained as anhydrous
grade from Acros Organics (toluene, THF, DMF). Compound 1, 3,3′-
dibromo-2,2′-bithiophene, was purchased from Sigma-Aldrich. The
4,4,5,5-tetramethyl-1,3,2-dioxaborolane derivatives of D119 and D2,8
referred to as D1-bpin and D2-bpin, were synthesized according to
reported procedures. Compound numbering is based on Scheme 1.
a
Scheme 1. Synthesis of LP225 and LP227
Compound 3. A solution of 2 (0.800 g, 2.21 mmol) in dry THF
(22 mL) was cooled to −78 °C under nitrogen. After 20 min, freshly
prepared lithium diisopropylamine (LDA, 0.237 g, 2.21 mmol) was
added dropwise, and the mixture was stirred for 2 h at −78 °C.
Dimethylformamide (0.162 g, 2.21 mmol) was added dropwise, and
the mixture was allowed to warm to room temperature for 1 h. The
reaction mixture was then poured into water, and the organic portions
were extracted with dichloromethane, dried over MgSO4, and filtered.
The solvent was removed under reduced pressure, and the crude
product was purified by flash chromatography (silica gel, 1:1 hexanes/
dichloromethane). Two bands were collected, which corresponded to
unreacted starting material 2 (0.156 g, 0.431 mmol, 20% recovered
starting material) and the desired product as an orange solid (0.620 g,
1
1.59 mmol, 72%). H NMR (400 MHz, CDCl3) δ 9.86 (s, 1H), 7.65
(s, 1H), 7.34 (d, J = 5.6 Hz, 1H), 7.01 (d, J = 5.6 Hz, 1H), 4.22 (tt, J =
10.0, 5.0 Hz, 1H), 2.08−1.93 (m, 2H), 1.92−1.78 (m, 2H), 1.30−1.06
(m, 14H), 1.06−0.92 (m, 2H), 0.79 (t, J = 7.0 Hz, 6H). 13C{1H}
NMR (100 MHz, CDCl3) δ 183.1, 148.5, 143.4, 139.9, 128.3, 123.5,
120.3, 115.1, 111.7, 60.3, 35.1, 31.6, 28.9, 26.6, 22.5, 14.0. HRMS
(ESI): m/z calcd for C22H32NOS2 ([M + H]+): 390.1925. Found:
390.1920. Anal. Calcd. for C22H31NOS2: C, 67.82; H, 8.02; N, 3.59.
Found: C, 67.86; H, 8.22; N, 3.53.
Compound 4. A solution of 3 (0.500 g, 1.28 mmol) in THF (25
mL) was covered with foil and cooled to −78 °C in a dry ice/acetone
bath. N-Bromosuccinimide (NBS, 0.228 g, 1.28 mmol) was added, and
the mixture was stirred at −78 °C. After 30 min, the mixture was
slowly warmed to room temperature, and the solvent was removed
under reduced pressure. The crude product was purified by passing
through a plug of silica gel eluting with dichloromethane to yield the
product as an orange solid (0.505 g, 1.08 mmol, 84%). 1H NMR (400
a
(a) Pd2dba3, BINAP, NaOtBu, toluene, reflux; 74%. (b) i. LDA, THF,
−78 °C; ii. DMF, −78 °C to rt; 72%. (c) NBS, THF, −78 °C to rt;
84%. (d) Dioxaborolane derivative of D1 or D2, Pd(PPh3)4, K2CO3,
THF, reflux; 93%. (e) Cyanoacetic acid, piperidine, CHCl3, reflux;
90−93%.
Chromatographic separations were performed according to standard
flash column chromatography methods using silica (Merck: silica gel
60, 230−400 mesh ASTM). Analytical thin-layer chromatography
(TLC) was performed on aluminum-backed sheets precoated with
silica 60 F254 adsorbent (0.25 mm thick; Merck, Germany) and
visualized under UV light (254 and 365 nm).
Characterization. 1H and 13C NMR were recorded at 400 and 100
MHz, respectively, on a Bruker AV 400 instrument at ambient
temperature. Elemental analysis (EA), electrospray ionization (ESI),
B
dx.doi.org/10.1021/cm401144j | Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
MHz, CDCl3) δ 9.85 (s, 1H), 7.63 (s, 1H), 7.07 (s, 1H), 4.15 (tt, J =
10.0, 5.0 Hz, 1H), 2.02−1.90 (m, 2H), 1.90−1.78 (m, 2H), 1.30−1.06
(m, 14H), 1.06−0.92 (m, 2H), 0.80 (t, J = 7.0 Hz, 6H). 13C{1H}
NMR (100 MHz, CDCl3) δ 183.0, 145.9, 142.4, 140.3, 123.1, 120.1,
115.4, 115.2, 115.0, 60.5, 35.1, 31.5, 28.9, 26.6, 22.5, 14.0. HRMS
(ESI): m/z calcd for C22H31BrNOS2 ([M + H]+): 468.1030. Found:
468.1042. Anal. Calcd. for C22H30BrNOS2: C, 56.40; H, 6.45; N, 2.99.
Found: C, 56.40; H, 6.69; N, 2.97.
red band. The collected fraction was washed with water to remove
traces of acetic acid, and the solvent was removed under reduced
pressure to yield a shiny purple solid (0.081 g, 0.065 mmol, 90%). 1H
NMR (400 MHz, (CD3)2CO/C5D5N) δ 8.50 (br. s, 1H), 8.12 (br. s,
1H), 7.74−7.65 (m, 3H), 7.53 (d, J = 8.2 Hz, 4H), 7.27 (d, J = 8.2 Hz,
2H), 7.21−7.09 (m, 6H), 6.65 (s, 2H), 6.58 (d, J = 8.0 Hz, 2H), 4.55
(tt, J = 10.0, 5.0 Hz, 1H), 4.02 (t, J = 6.4 Hz, 8H), 2.27−2.12 (m, 2H),
2.12−2.00 (m, 4H), 2.00−1.90 (m, 2H), 1.84−1.66 (m, 8H), 1.54−
1.40 (m, 8H), 1.38−1.12 (m, 28H), 0.91 (t, J = 7.0 Hz, 6H), 0.86 (t, J
= 7.0 Hz, 6H), 0.77 (t, J = 6.8 Hz, 6H). 13C{1H} NMR (100 MHz,
(CD3)2CO/C5D5N) δ 165.5, 160.7, 157.9, 150.9, 148.9, 148.6, 148.2,
146.1, 144.2, 134.9, 133.8, 131.5, 131.2, 129.3, 127.3, 125.5, 124.8,
123.8, 123.3, 123.1, 118.1, 114.3, 108.1, 106.5, 101.0, 95.6, 68.9, 68.6,
60.8, 35.5, 32.4, 32.3, 32.2, 27.2, 26.6, 26.5, 23.3, 23.2, 14.4, 14.34,
14.27. HRMS (ESI): m/z calcd for C79H101N3O6S2 ([M]+):
1251.7132. Found: 1251.7131. Anal. Calcd. for C79H101N3O6S2·1/
2H2O: C, 75.20; H, 8.15; N, 3.33. Found: C, 75.17; H, 8.32; N, 3.28.
LP227. To a solution of 5b (0.078 g, 0.062 mmol) and cyanoacetic
acid (0.016 g, 0.185 mmol, 3.0 equiv) in chloroform (1.5 mL) was
added piperidine (0.037 g, 0.431 mmol, 7.0 equiv). After stirring at 80
°C overnight, the reaction mixture was diluted with dichloromethane
(20 mL) and acidified with 2 M aqueous HCl (20 mL). After 5 min of
vigorous stirring, the reaction mixture changed from bright red to dark
purple. The organic portions were extracted carefully with dichloro-
methane and washed with brine, and the solvent was removed under
reduced pressure. The crude product was purified by flash
chromatography (silica gel) eluting sequentially with dichloromethane,
a solution of 5% methanol in dichloromethane, and a solution of 5%
methanol and 0.1% acetic acid in dichloromethane to collect a bright
red band. The collected fraction was washed with water to remove
traces of acetic acid, and the solvent was removed under reduced
pressure to yield a shiny purple solid (0.076 g, 0.057 mmol, 93%). 1H
NMR (400 MHz, (CD3)2CO/C5D5N) δ 8.49 (br. s, 1H), 8.13 (br. s,
1H), 7.93 (d, J = 8.2 Hz, 2H), 7.74 (s, 1H), 7.70 (d, J = 8.3 Hz, 2H),
7.31 (s, 2H), 7.15 (d, J = 8.2 Hz, 2H), 7.09 (dd, J = 8.1, 1.3 Hz, 2H),
6.72 (s, 2H), 6.55 (s, 2H), 4.58 (tt, J = 10.0, 5.0 Hz, 1H), 4.17 (t, J =
6.0 Hz, 4H), 4.06 (t, J = 6.4 Hz, 4H), 2.26−2.14 (m, 4H), 2.00−1.87
(m, 8H), 1.80 (quint., J = 7.2 Hz, 4H), 1.60 (quint., J = 7.6 Hz, 4H),
1.50 (quint., J = 7.4 Hz, 4H), 1.42−1.02 (m, 40H), 0.88−0.0.82 (m,
12H), 0.78 (t, J = 7.0 Hz, 6H). 13C{1H} NMR (100 MHz,
(CD3)2CO/C5D5N) δ 165.5, 161.3, 157.4, 156.5, 155.1, 149.5,
148.8, 148.2, 145.7, 144.2, 135.5, 133.7, 128.6, 127.2, 125.5, 124.5,
123.9, 123.0, 120.2, 119.7, 118.8, 118.1, 114.1, 107.9, 101.2, 98.7, 95.4,
68.8, 68.6, 60.8, 48.0, 35.5, 32.4, 32.3, 27.4, 27.2, 26.7, 26.5, 23.30,
23.27, 23.2, 14.33, 14.27. HRMS (ESI): m/z calcd for C85H109N3O6S2
([M]+): 1331.7758. Found: 1331.7742. Anal. Calcd. for
C85H109N3O6S2·1/2H2O: C, 76.08; H, 8.26; N, 3.13. Found: C,
76.15; H, 8.48; N, 3.08.
Optical and Electrochemical Properties. Electronic spectro-
scopic data were collected in solution using a Hewlett-Packard Diode
Array spectrophotometer. Emission spectra were recorded with a
Fluorolog Horiba Jobin Yvon model FL-1065. Voltammetric measure-
ments employed a PC controlled AutoLab PSTAT 10 electrochemical
workstation and were carried out under anaerobic conditions (argon-
filled glove, oxygen, and water <1 ppm). Measurements were carried
out with 0.1 M Bu4NPF6 in dichloromethane using a set of glassy
carbon, Pt plate, and Pt wire as working, counter, and pseudoreference
electrodes, respectively. Ferrocene was used as an internal standard
(E° = +0.69 V vs NHE).20
Theoretical Calculations. Full geometry optimizations of
compounds LP225 and LP227 in their singlet ground state were
performed with density functional theory (DFT) using the M06
functional21 with the DGDZVP basis set22,23 for all the atoms (double-
ζ and polarizations), an ultrafine integration grid, and tight geometrical
convergence criteria with the Gaussian 09 package.24 All calculations
were carried out including implicit solvent effects. Linear-response
time-dependent density functional theory (LR-TDDFT) was used in
combination with the BMK functional25 for the calculations of
excitation energies, while ionization energies were computed at the
(U)DFT/M06 level. Graphical representation of the molecules and
Compound 5a. A solution of 4 (0.048 g, 0.103 mmol), D1-bpin19
(0.100 g, 0.108 mmol), and Pd(PPh3)4 (0.012 g, 0.010 mmol) in THF
(5 mL) was degassed with nitrogen. A 2 M aqueous solution of K2CO3
(0.5 mL) was added, and the mixture was heated to 70 °C overnight.
The reaction mixture was quenched with water, and the organic
portions were extracted with dichloromethane, washed with brine, and
dried over MgSO4. The drying agent was removed via filtration, and
the solvent was removed under reduced pressure. The crude product
was purified by flash chromatography (silica gel, 1:1 dichloromethane/
1
hexanes) to yield an orange solid (0.114 g, 0.096 mmol, 93%). H
NMR (400 MHz, CDCl3) δ 9.83 (s, 1H), 7.61 (s, 1H,), 7.51 (d, J = 8.7
Hz, 2H), 7.45 (d, J = 8.7 Hz, 4H), 7.25 (d, J = 8.7 Hz, 2H), 7.17 (m,
6H), 7.12 (s, 1H), 6.57−6.48 (m, 4H), 4.22 (tt, J = 10.0, 5.0 Hz, 1H),
3.97 (t, J = 6.7 Hz, 4H), 3.95 (t, J = 6.6 Hz, 4H), 2.09−1.96 (m, 2H),
1.92−1.83 (m, 2H), 1.83−1.68 (m, 8H), 1.50−1.43 (m, 4H), 1.43−
1.37 (m, 4H), 1.37−1.32 (m, 8H), 1.32−1.25 (m, 10H), 1.25−1.09
(m, 14H), 0.90 (t, J = 7.1 Hz, 6H), 0.86 (t, J = 7.0 Hz, 6H), 0.79 (t, J =
6.9 Hz, 6H). 13C{1H} NMR (100 MHz, CDCl3) δ 182.8, 159.6, 157.0,
148.1, 147.5, 145.4, 139.5, 133.6, 130.9, 130.3, 128.2, 126.5, 124.1,
123.2, 122.9, 113.7, 106.2, 105.3, 100.4, 68.4, 68.2, 60.3, 35.1, 31.7,
31.6, 31.5, 29.5, 29.4, 29.1, 29.0, 26.6, 25.8, 22.73, 22.66, 22.6, 22.5,
14.1, 14.0 (four aromatic resonances not observed due to overlap).
HRMS (ESI): m/z calcd for C76H101N2O5S2 ([M + H]+): 1185.7152.
Found: 1185.7114. Anal. Calcd. for C76H100N2O5S2: C, 76.98; H, 8.50;
N, 2.36. Found: C, 77.27; H, 8.82; N, 2.25.
Compound 5b. A solution of 4 (0.044 g, 0.095 mmol), D2-bpin8
(0.100 g, 0.100 mmol), and Pd(PPh3)4 (0.012 g, 0.009 mmol) in THF
(5 mL) was degassed with nitrogen. A 2 M aqueous solution of K2CO3
(0.5 mL) was added, and the mixture was heated to 70 °C overnight.
The reaction mixture was quenched with water, and the organic
portions were extracted with dichloromethane, washed with brine, and
dried over MgSO4. The drying agent was removed via filtration, and
the solvent was removed under reduced pressure. The crude product
was purified by flash chromatography (silica gel, 3:2 dichloromethane/
1
hexanes) to yield an orange solid (0.112 g, 0.088 mmol, 93%). H
NMR (400 MHz, CDCl3) δ 9.84 (s, 1H), 7.85 (d, J = 8.1 Hz, 2H),
7.61 (s, 1H), 7.50 (d, J = 8.2 Hz, 2H), 7.20−7.11 (m, 5H), 7.09 (d, J =
8.2 Hz, 2H), 6.52 (s, 2H), 6.41 (s, 2H), 4.22 (tt, J = 10.0, 5.0 Hz, 1H),
4.07 (t, J = 7.2 Hz, 4H), 4.00 (t, J = 7.2 Hz, 4H), 2.10−1.95 (m, 2H),
1.95−1.85 (m, 6H), 1.85−1.75 (m, 4H), 1.60−1.52 (m, 6H), 1.52−
1.43 (m, 4H), 1.43−1.00 (m, 50H), 0.93−0.86 (m, 12H), 0.80 (t, J =
7.0 Hz, 6H). 13C{1H} NMR (100 MHz, CDCl3) δ 182.8, 160.0, 156.6,
155.6, 154.2, 149.2, 148.6, 147.7, 145.0, 142.8, 139.4, 134.3, 131.9,
127.6, 126.4, 124.1, 123.7, 123.0, 122.5, 119.8, 118.8, 113.6, 106.0,
100.0, 97.9, 68.4, 68.0, 60.3, 47.3, 35.1, 31.7, 31.64, 31.58, 29.4, 29.0,
27.2, 26.9, 26.6, 26.0, 25.8, 22.7, 22.6, 22.6, 14.1, 14.0 (one aromatic
resonance not observed due to overlap). HRMS (ESI): m/z calcd for
C82H109N2O5S2 ([M + H]+): 1265.7778. Found: 1265.7753. Anal.
Calcd. for C82H108N2O5S2·1/2H2O: C, 77.25; H, 8.62; N, 2.20. Found:
C, 77.06; H, 8.70; N, 2.14.
LP225. To a solution of 5a (0.085 g, 0.072 mmol) and cyanoacetic
acid (0.018 g, 0.215 mmol, 3.0 equiv) in chloroform (1.5 mL) was
added piperidine (0.043 g, 0.502 mmol, 7.0 equiv). After stirring at 80
°C overnight, the reaction mixture was diluted with dichloromethane
(20 mL) and acidified with 2 M aqueous HCl (20 mL). After 5 min of
vigorous stirring, the reaction mixture changed from bright red to dark
purple. The organic portions were extracted carefully with dichloro-
methane and washed with brine, and the solvent was removed under
reduced pressure. The crude product was purified by flash
chromatography (silica gel) eluting sequentially with dichloromethane,
a solution of 5% methanol in dichloromethane, and a solution of 5%
methanol and 0.1% acetic acid in dichloromethane to collect a bright
C
dx.doi.org/10.1021/cm401144j | Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
their orbitals was obtained with the software VMD v.1.9.0.26 For full
details concerning the quantum chemical calculations, please see the
Supporting Information.
one electron is injected per TiO2 nanoparticle per pulse. The probe
light consists of a xenon arc lamp passed through a 665 nm cutoff filter
and a water filter, focused onto the sample and collected in a
monochromator at 750 nm. The detector is a fast photomultiplier tube
connected to a digital oscilloscope. Typical data are averaged over
1000 laser shots and smoothed using a Savitzky−Golay filter.
Electrolyte Composition. 4-tert-Butylpyridine (TBP), lithium
bis(trifluoromethanesulfonyl)imide (LiTFSI), and the solvents were
reagent grade and were used without further purification. Cobalt
complexes [Co(bpy)3](TFSI)2 and [Co(bpy)3](TFSI)3 (bpy = 2,2′-
bipyridine) were synthesized according to reported procedures.17
Their bis(trifluoromethane) sulfonimide (TFSI) salt was chosen to
improve solubility in acetonitrile. The cobalt electrolyte consists of
0.25 M [Co(bpy)3](TFSI)2, 0.05 M [Co(bpy)3](TFSI)3, 0.25 M TBP,
and 0.1 M LiTFSI in acetonitrile.
RESULTS AND DISCUSSION
■
The synthesis of LP225 and LP227 is extremely straightfor-
ward, which highlights the appeal of DTP (Scheme 1). The
DTP core 2 was prepared from commercially available 3,3′-
dibromo-2,2′-bithiophene via Buchwald−Hartwig ring closure
with tridecan-7-amine.28 The key intermediate 4 was
synthesized via selective formylation of the DTP core using 1
equiv of lithium diisopropylamine (LDA) in THF at −78 °C;
the resulting lithiated species was converted to 3 through the
addition of DMF. This compound was then brominated using 1
equiv of NBS in THF at −78 °C. Finally, LP225 and LP227
were obtained in two steps from compound 4 via Suzuki
coupling with the 4,4,5,5-tetramethyl-1,3,2-dioxaborolane de-
rivative of D119 or D2,8 respectively, followed by Knoevenagel
condensation with cyanoacetic acid.
Solar Cells Preparation. Nanocrystalline TiO2 pastes were
prepared using a previously reported procedure.27 The TiO2
transparent electrodes were prepared by screen-printing onto
fluorine-doped tin oxide (FTO, Solar 4 mm thickness, 10 Ω sq−1,
Nippon Sheet Glass) conducting glass. The film thickness was adjusted
to ca. 4.0 μm by the number of screen-printing cycles. The screen-
printing paste was composed of ∼20 nm diameter anatase particles
that gave a mesoporous layer with ∼32 nm pores. To render high
power conversion efficiency, an ca. 4 μm scattering layer (400 nm
diameter, Catalysts & Chemicals Ind. Co. Ltd. (CCIC), HPW-400)
was deposited on the transparent layer. The 8.0 μm thick TiO2
electrodes were immersed into a tetrahydrofuran/ethanol solution of
the corresponding dye (1:4 ratio, 0.2 mM) with 2 equiv of 3α,7α-
dihydroxy-5β-cholic acid (chenodeoxycholic acid, CDCA) and kept for
6 h at room temperature. The dye-adsorbed TiO2 electrode and
thermally platinized counter electrode were assembled into a sealed
sandwich-type cell with a gap of a hot-melt ionomer film, Surlyn (25
μm, Du Pont). Devices were completed by filling the electrolyte
through a predrilled hole in the counter electrode. A self-adhesive,
antireflecting, ultraviolet cutoff film (λ < 380 nm, ARKTOP, ASAHI
Glass) was attached to the top of the active area to decrease light
reflection loss. A black mask (0.159 mm2) that is smaller than the
active area of the cells was used in subsequent photovoltaic studies.
Solar Cells Characterization. For photovoltaic measurements of
the DSCs, the irradiation source originated from a 450 W xenon light
source (Osram XBO 450) with a filter (Schott 113), whose power was
regulated to the AM 1.5 G solar standard by using a reference Si
photodiode equipped with a color-matched filter (KG-3, Schott) to
reduce the mismatch in the region of 350−750 nm between the
simulated light and AM 1.5 G to less than 4%. The measurement-
settling time between applying a voltage and measuring a current for
the J−V characterization of DSCs was fixed to 80 ms. The incident
photon-to-collected electron conversion efficiency (IPCE) measure-
ment was plotted as a function of wavelength by using the light from a
300 W xenon lamp (ILC Technology), which was focused through a
Gemini-180 double monochromator (Jobin Yvon) onto the photo-
voltaic cell under test. A computer-controlled monochromator 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 4 s. To reduce scattered light from the edge
of the glass electrodes of the dyed TiO2 layer, a light-shading mask was
used on the DSCs, so the active area of DSCs was fixed to 0.2 cm2.
Photoinduced Absporption Spectroscopy (PIA). PIA spectra
were recorded over 500−1100 nm on transparent devices containing
either [CoII/III(bpy)3]2+/3+ electrolyte or pure acetonitrile solvent in a 8
μm TiO2 film loaded with dyes LP225 or LP227. The samples were
excited with a 470 nm LED, modulated at 9 Hz (square wave), and
driven by the detection lock-in amplifier. White light from a halogen
bulb was used as a probe. These spectra allowed identification of the
position of oxidized dye species.
The optical properties of LP225 and LP227 are strongly
solvent and pH-dependent (Figure S1 and Table S1,
Supporting Information) due to changes in the protonation
state of the cyanoacrylic acceptor. The absorption spectra in
dichloromethane solution very closely resemble that of the
protonated form of the dyes and exhibit maxima at 526 and 541
nm, respectively (Figure 1, Table 1, and Table S2, Supporting
Information). These values are similar in trend to that observed
for Y123 and JF419 with a slight hypsochromic shift.
Linear-response time-dependent density functional theory
(LR-TDDFT/BMK/SMD(CH2Cl2)), based on DFT/M06
geometries of LP225 and LP227, provides absorption spectra
in close agreement with the experimental evidence (see
Figure 1. (a) Steady-state absorption spectra measured in dichloro-
methane solutions and (b) absorption and normalized emission (λex
=
460 nm) spectra of 3.3 μm thick TiO2 films sensitized with LP225,
LP227, Y123, and JF419.
Table 1. Summary of Optical and Electrochemical Data
a
b
c
d
dye
λmax/nm (ε/104 M−1 cm−1
)
E0−0 /eV ES+/S /V ES+/S* /V
LP225
LP227
526 (5.77)
541 (5.26)
542 (5.05)
548 (4.75)
2.22
2.15
2.17
2.10
+0.99
+0.84
+1.01
+0.85
−1.23
−1.31
−1.16
−1.25
e
Y123
Transient Absorption Spectroscopy (TA). Nanosecond laser
flash photolysis was applied to samples loaded with dyes LP225 or
LP227 (identical to PIA samples). The samples were excited by 7 ns
(fwmh) pulsed laser light produced at a repetition rate of 20 Hz by an
optical parametric oscillator pumped by a frequency-tripled Q-
switched Nd:YAG laser. The output excitation wavelength was
tuned at 520 nm, and the laser fluence on the sample was kept at a
low level (50 μJ cm−2 per pulse) to ensure that, on average, less than
e
JF419
a
Derived from the absorption in dichloromethane solutions.
b
Estimated from the intersection point of absorption and normalized
c
emission spectra on TiO2. Half-wave oxidation potentials measured
using cyclic voltammetry and reported vs NHE. Estimated according
to: ES+/S* = ES+/S − E0−0. From ref 8.
d
e
D
dx.doi.org/10.1021/cm401144j | Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
Supporting Information for details concerning quantum
chemical calculations). In particular, the lowest energy
transition (S0→S1) of LP225 is blue-shifted by less than 0.1
eV compared to the first vertical transition of LP227. The
overall hypsochromic shift observed for the DTP relative to the
CPDT dyes is also mirrored in the quantum chemical
calculations.8 The dominant occupied-to-virtual orbital con-
tribution to the S0→S1 transition is HOMO to LUMO in both
cases (82.2% for LP225 and 81.4% for LP227), which confirms
the intramolecular charge transfer (ICT) character of the main
transition (Figure 2).
In functional DSCs, a favorable balance between charge
recombination and dye regeneration is essential to high
efficiency. Their respective time constants τrec and τreg can be
quantified using transient absorption (TA) decay measure-
ments (Figure 3). Following excitation at 470 nm, photo-
induced absorption signals that appear at ca. 600 and 700−800
nm are characteristic of ICT band bleaching and triphenyl-
amine radicals, respectively. Time constants are derived from
TA decays of the corresponding radical. The absorption traces
can be fitted to an exponential decay function ΔA(t) ∝ A0
exp[−(t/τ)], where A0 is the pre-exponential factor and τ is the
characteristic time.29 The mean times of charge recombination
(τrec) and oxidized dye regeneration (τreg) were derived using
cells containing an inert or a [CoII/III(bpy)3]2+/3+ electrolyte,
respectively. Data are reported in Table 2.
Figure 2. Illustrations of the Kohn−Sham (KS) HOMO and LUMO
representing the dominant components of the first vertical transition
of LP225 and LP227, computed with LR-TDDFT/BMK/SMD-
(CH2Cl2). Illustrations of the KS HOMO and LUMO of Y123 and
JF419 are virtually identical to LP225 and LP227. The isosurface value
in all cases is set to 0.03 au.
Figure 3. (a, b) PIA following excitation at 470 nm and (c, d) TA
decay traces of 8.0 μm thick TiO2 films sensitized with LP225 (blue)
and LP227 (red). Decays in inert (i) and [CoII/III(bpy)3]2+/3+ (e)
electrolytes. Laser pulse of 50 μJ cm−2 (λexc = 530 nm), probe at 750
nm.
Table 2. Kinetic Parameters for Dye Regeneration
To more accurately estimate the situation in the device, the
absorption and emission were acquired on sensitizers anchored
to the TiO2 semiconductor.2 The absorption maxima are
similar for those dyes that share the same donor moiety, with
slightly different band shapes. The emission spectra show
marked bathochromic shifts associated with changing the
bridge from DTP to CPDT and the donor from D1 to D2.
Consequently, the zero−zero transition energy (E0−0) values
for the DTP derivatives are slightly larger relative to those
estimated for their CPDT analogues. This observation is
consistent with the trends observed for the optical gaps in
solution and with an increase in the energy gap for DTP
relative to CPDT estimated by DFT calculations.13
The energy levels of the sensitizer with respect to the
electrochemical potential of the electrolyte and conduction
band of TiO2 (−0.5 V vs NHE)2 also play an important role in
the resulting DSC performance. The ground state oxidation
potentials (ES+/S) of LP225 and LP227 are measured at +0.99
and +0.84 V vs NHE,20 very similar to those of their CPDT
counterparts. These values are in good agreement with the
trends in ionization energies computed at the (U)DFT/M06
level of theory (Table S2, Supporting Information) and provide
ample driving force for regeneration (−ΔG°) from a
[CoII/III(bpy)3]2+/3+ electrolyte (+0.56 V vs NHE). Using the
ES+/S of the sensitizers and the estimated E0−0 value, the
excited-state oxidation potentials (ES+/S*) for the DTP dyes
were estimated at −1.23 and −1.31 V, respectively. The driving
force for injection of the photoexcited electron into the TiO2
conduction band is slightly larger than their CPDT analogues,
thus ensuring quantitative injection quantum yields into TiO2.8
a
dye
−ΔG°/eV
τrec/μs
τreg/μs
τreg/τrec
Φ
reg/%
LP225
LP227
0.43
0.28
0.45
0.29
1700
1900
550
7.7
3.4
6.8
7.8
220
560
81
>99
>99
99
b
Y123
b
JF419
330
42
98
a
b
Φreg = kreg/(kreg + krec), τi = 1/ki. From ref 8.
The τrec were estimated at 1.7 and 1.9 ms for LP225 and
LP227, respectively. Notably, these times constants are
significantly longer than for the CPDT derivatives, which
indicates that electrons injected into the TiO2 recombine
slower with the oxidized form of the DTP dyes. Adding the
active cobalt(II/III) species to the electrolyte gives accelerated
TA decays, with τreg values of 7.7 and 3.4 μs for LP225 and
LP227, respectively. In both cases, the τrec/τreg ratio is large,
suggesting extremely favorable charge-transfer kinetics com-
pared to previous systems,16 which result in quantitative
regeneration quantum yields (Φreg). The use of DTP instead of
CPDT also appears more favorable for the regeneration of the
oxidized dye according to this ratio despite −ΔG° similarities.
The incident photon-to-current efficiency (IPCE) spectra
were measured on films coated with LP225 and LP227 with an
optimized [CoII/III(bpy)3]2+/3+ electrolyte composition. The
IPCEs exhibit a similar shape for all the dyes with intensity
maxima around 80−85% (Figure 4a). However, we observe up
to a 25 nm blue-shift in the onset for a given DTP dye
compared to its CPDT analogue. The decrease in photon
collection is consistent with the estimated E0−0 values. The
E
dx.doi.org/10.1021/cm401144j | Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
a
Table 3. Detailed Photovoltaic Parameters
dye
JSC/mA cm−2
VOC/mV
FF
η/%
LP225
LP227
13.4 (13.2 0.2)
14.1 (13.9 0.2)
14.1
901 (890 16)
811 (810 5)
876
0.74 (0.74 0.00)
0.77 (0.76 0.01)
0.78
8.86 (8.63 0.21)
8.72 (8.62 0.09)
9.77
b
Y123
b
JF419
16.2
840
0.76
10.3
a
b
Simulated AM 1.5 G conditions. Average values are based on three replicate measurements. From ref 8.
conduction band edge. This effect provides the DTP dyes with
higher VOC relative to CPDT, as exemplified with LP225 and
Y123 (the VOC of LP227 is limited by the electron lifetime).
CONCLUSION
■
The conduction band shift observed with LP225 and LP227
should prove particularly advantageous when closing the optical
gap of DTP dyes through the use of stronger electron
acceptors. Indeed, large VOC (>900 mV) values are usually
obtained with high oxidation redox mediators, which is
detrimental to dye regeneration and current densities.16,31 As
evidenced in this report, the DTP bridge can naturally
overcome this issue if favorable charge-transfer kinetics are
created upon insulation of the TiO2 surface. Further work is
currently underway to improve the spectral response of such
DTP dyes.
Figure 4. (a) Photocurrent action spectra of mesoscopic solar cells
with [CoII/III(bpy)3]2+/3+ electrolyte and (b) J−V characteristics
measured under simulated AM 1.5 G full sun illumination (100 mW
cm−2).
solar-to-electricity conversion efficiencies of these DSCs were
evaluated by recording the J−V characteristics under simulated
AM 1.5 G illumination (Figure 4b). As expected from
integration of the IPCE, the photocurrent densities (JSC)
measured for the DTP dyes are ca. 1−2 mA cm−2 lower than
for their CPDT analogues. On the other hand, marked
differences in VOC are observed in Table 3, where dyes
assembled with the D1 donor tend to exhibit higher values.
These variations may be related to fluctuation in the electron
density and/or shifts in the conduction band edge of the
TiO2.30 Using transient photocurrent decay measurements, we
studied the electron lifetime and capacitance as a function of
voltage (Figure 5). The lifetime of electrons injected in the
ASSOCIATED CONTENT
■
S
* Supporting Information
Characterizing data for experimentally studied compounds; full
details of quantum chemical calculations; supplemental figures
discussed in the text (PDF). This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: lauren.polander@epfl.ch; mdkhaja.nazeeruddin@epfl.
ch.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We acknowledge the European Community’s Seventh Frame-
work Programme (FP7/2007-2013) under grant agreement no.
281063 of the Powerweave project for financial support and
SSSTC (Sino-Swiss Science and Technology Cooperation).
MKN thanks the World Class University program, Korea
University funded by the National Research Foundation of
Korea (Grant R31-2008-000-10035-0). Support from the Swiss
National Science Foundation (Grant No. 200020-130082) and
the NCCR-MUST interdisciplinary research program is also
gratefully acknowledged.
Figure 5. (a) Electron lifetime and (b) capacitance versus open-circuit
photovoltage of cells made with TiO2 films coated with LP225,
LP227, Y123, and JF419.
TiO2 (τn) is shorter for LP227 compared to LP225. A similar
trend is observed for the CPDT dyes, which indicates that
shorter τn are linked to intrinsic limitations of the D2 donor.
However, as expected for LP225 and LP227, the use of a bulky
donor prevents the cobalt(III) species from accessing the TiO2
surface, which drastically limits recombination compared to the
literature.16 Furthermore, decreased capacitance at the same
potential bias was observed for the DTP dyes relative to CPDT.
This is attributed to a 60−100 mV upward shift of the TiO2
REFERENCES
■
(1) O’Regan, B.; Gratzel, M. Nature 1991, 353, 737−740.
̈
(2) Hagfeldt, A.; Boschloo, G.; Sun, L.; Kloo, L.; Pettersson, H.
Chem. Rev. 2010, 110, 6595−6663.
(3) Yu, Q.; Wang, Y.; Yi, Z.; Zu, N.; Zhang, J.; Zhang, M.; Wang, P.
ACS Nano 2010, 4, 6032−6038.
(4) Yella, A.; Lee, H. W.; Tsao, H. N.; Yi, C.; Chandiran, A. K.;
Nazeeruddin, M. K.; Diau, E. W. G.; Yeh, C. Y.; Zakeeruddin, S. M.;
Gratzel, M. Science 2011, 334, 629−634.
̈
F
dx.doi.org/10.1021/cm401144j | Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
(5) Materials for Nonlinear Optics: Chemical Perspectives; Marder, S.
R., Sohn, J. E., Stucky, G. D., Eds.; ACS Symp. Ser.; American
Chemical Society: Washington, DC, 1991; Vol. 455.
(31) Yum, J.-H.; Baranoff, E.; Kessler, F.; Moehl, T.; Ahmad, S.;
Bessho, T.; Marchioro, A.; Ghadiri, E.; Moser, J.-E.; Yi, C.;
Nazeeruddin, M. K.; Gratzel, M. Nat. Commun. 2012, 3, 631.
̈
(6) Mishra, A.; Fischer, M. K. R.; Bauerle, P. Angew. Chem., Int. Ed.
̈
2009, 48, 2474−2499.
(7) Tsao, H. N.; Yi, C.; Moehl, T.; Yum, J.-H.; Zakeeruddin, S. M.;
Nazeeruddin, M. K.; Gratzel, M. ChemSusChem 2011, 4, 591−594.
̈
(8) Yella, A.; Humphry-Baker, R.; Curchod, B. F. E.; Astani, N. A.;
Teuscher, J.; Polander, L. E.; Mathew, S.; Tavernelli, I.; Rothlisberger,
U.; Moser, J.-E.; Gratzel, M.; Nazeeruddin, M. K.; Frey, J. 2013,
̈
10.1021/cm401593b.
(9) Brzezinski, J. Z.; Reynolds, J. R. Synthesis 2002, 1053−1056.
(10) Coppo, P.; Cupertino, D. C.; Yeates, S. G.; Turner, M. L. J.
Mater. Chem. 2002, 12, 2597−2599.
(11) Ogawa, K.; Rasmussen, S. C. J. Org. Chem. 2003, 68, 2921−
2928.
(12) Ogawa, K.; Rasmussen, S. C. Macromolecules 2006, 39, 1771−
1778.
(13) Odom, S. A.; Lancaster, K.; Beverina, L.; Lefler, K. M.;
́
Thompson, N. J.; Coropceanu, V.; Bredas, J.-L.; Marder, S. R.; Barlow,
S. Chem.Eur. J. 2007, 13, 9637−9646.
(14) Cai, N.; Zhang, J.; Xu, M.; Zhang, M.; Wang, P. Adv. Funct.
Mater. 2013, DOI: 10.1002/adfm.201203348.
(15) Mo, H.; Radke, K. R.; Ogawa, K.; Heth, C. L.; Erpelding, B. T.;
Rasmussen, S. C. Phys. Chem. Chem. Phys. 2010, 12, 14585.
(16) Xu, M.; Zhang, M.; Pastore, M.; Li, R.; De Angelis, F.; Wang, P.
Chem. Sci. 2012, 3, 976−983.
(17) Feldt, S. M.; Gibson, E. A.; Gabrielsson, E.; Sun, L.; Boschloo,
G.; Hagfeldt, A. J. Am. Chem. Soc. 2010, 132, 16714−16724.
(18) Feldt, S. M.; Wang, G.; Boschloo, G.; Hagfeldt, A. J. Phys. Chem.
C 2011, 115, 21500−21507.
(19) Gao, P.; Kim, Y. J.; Yum, J.-H.; Holcombe, T. W.; Nazeeruddin,
M. K.; Gratzel, M. J. Mater. Chem. A 2013, 1, 5535−5544.
̈
(20) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877−910.
(21) Zhao, Y.; Truhlar, D. G. Theor. Chim. Acta 2008, 120, 215−241.
(22) Godbout, N.; Salahub, D. R.; Andzelm, J.; Wimmer, E. Can. J.
Chem. 1992, 70, 560−571.
(23) Sosa, C.; Andzelm, J.; Elkin, B. C.; Wimmer, E.; Dobbs, K. D.;
Dixon, D. A. J. Phys. Chem. 1992, 96, 6630−6636.
(24) Frisch, M. J. T.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson,
G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov,
A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.;
Toyota, K.; Fukuda, R.; Hasegawa, 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.; Zakrezewski, V. G.; Voth, G. A.; Salvador, P.;
Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman, J.
B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09; Gaussian, Inc.:
Wallingford CT, 2009.
(25) Boese, A. D.; Martin, J. M. L. J. Chem. Phys. 2004, 121, 3405−
3416.
(26) Humphrey, W.; Dalke, A.; Schulten, K. J. Mol. Graph. 1996, 14,
33−38.
(27) Ito, S.; Murakami, T. N.; Comte, P.; Liska, P.; Gratzel, C.;
̈
Nazeeruddin, M. K.; Gratzel, M. Thin Solid Films 2008, 516, 4613−
̈
4619.
(28) Koeckelberghs, G.; De Cremer, L.; Vanormelingen, W.; (null);
Verbiest, T.; Persoons, A.; Samyn, C. Tetrahedron 2005, 61, 687−691.
(29) Anderson, A. Y.; Barnes, P. R.; Durrant, J. R.; O’Regan, B. C. J.
Phys. Chem. C 2011, 115, 2439−2447.
(30) Barnes, P. R. F.; Miettunen, K.; Li, X.; Anderson, A. Y.; Bessho,
T.; Gratzel, M.; O’Regan, B. C. Adv. Mater. 2013, 25, 1881−1922.
̈
G
dx.doi.org/10.1021/cm401144j | Chem. Mater. XXXX, XXX, XXX−XXX