Molecular dipole, dye structure and electron lifetime relationship in efficient dye sensitized solar cells based on donor-π-acceptor organic sensitizers

Article abstract of DOI:10.1016/j.orgel.2014.08.058

In this work we report the synthesis and characterization of two new push-pull organic dyes (LC95 and LC107) to be employed as sensitizers in solar cells. Both molecules contain the bis(4-hexyloxyphenylamino)styril unit as the donor group, the cyanoacrylate acid as the acceptor, and the selenophene-thiophene (LC95 dye) and thiophene-selenophene (LC107 dye) moieties as the conjugated linkers. Dye sensitized solar cells employing these two photosensitizers and the cobalt(II/III) redox electrolyte exhibit good solar to energy conversion efficiencies of 6.3% and 6.5% measured under the 100 mW cm^-2simulated AM1.5 sunlight. The efficiencies are slightly lower with the iodine/iodide electrolyte. The performance of these two dyes has been discussed and compared to three closely related sensitizers, i.e. C214, C216 and C218, by means of experimental measurements and quantum chemistry computations, with special attention to differences on their geometries, molecular dipoles and electron recombination lifetimes.

Full text of DOI:10.1016/j.orgel.2014.08.058

Organic Electronics 15 (2014) 3162–3172
Contents lists available at ScienceDirect
Organic Electronics
journal homepage: www.elsevier.com/locate/orgel
Molecular dipole, dye structure and electron lifetime
relationship in efficient dye sensitized solar cells based on
donor–p–acceptor organic sensitizers
Clàudia Climent a, Lydia Cabau b, David Casanova c,d, , Peng Wang e, , Emilio Palomares b,f,
⇑
⇑
⇑
a Departament de Química Física and Institut de Química Teòrica i Computacional (IQTCUB), Universitat de Barcelona, Martí i Franquès, 1-11,
08028 Barcelona, Spain
b Institute of Chemical Research of Catalonia (ICIQ), Avda. Països Catalans 16, 43007 Tarragona, Spain
c Kimika Fakultatea, Euskal Herriko Unibertsitatea (UPV/EHU), Donostia International Physics Center (DIPC), P.K. 1072, 20080 Donostia, Spain
d IKERBASQUE, Basque Foundation for Science, 48013 Bilbao, Spain
e State Key Laboratory of Polymer Physics and Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, China
f ICREA, Avda. Lluís Companys 28, Barcelona E-08030, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 30 June 2014
Received in revised form 25 August 2014
Accepted 30 August 2014
Available online 19 September 2014
In this work we report the synthesis and characterization of two new push–pull organic
dyes (LC95 and LC107) to be employed as sensitizers in solar cells. Both molecules contain
the bis(4-hexyloxyphenylamino)styril unit as the donor group, the cyanoacrylate acid as
the acceptor, and the selenophene–thiophene (LC95 dye) and thiophene–selenophene
(LC107 dye) moieties as the conjugated linkers. Dye sensitized solar cells employing these
two photosensitizers and the cobalt(II/III) redox electrolyte exhibit good solar to energy
conversion efficiencies of 6.3% and 6.5% measured under the 100 mW cmÀ2 simulated
AM1.5 sunlight. The efficiencies are slightly lower with the iodine/iodide electrolyte. The
performance of these two dyes has been discussed and compared to three closely related
sensitizers, i.e. C214, C216 and C218, by means of experimental measurements and quan-
tum chemistry computations, with special attention to differences on their geometries,
molecular dipoles and electron recombination lifetimes.
Keywords:
Organic dyes
Dye sensitized solar cells
Thiophene
Selenophene
Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction
photovoltaic devices [1], has emerged as one of the front-
runners with certified efficiencies overpassing 12% under
The search of low carbon emission and efficient energy
alternatives to replace the conventional fossil fuels has
become an extremely intense field of research and one of
the main challenges for scientists. In the last two decades,
dye sensitized solar cells (DSSC), among other molecular
standard measured conditions [2]. As it has been previ-
ously reported, the molecular sensitizer in DSSC plays a
crucial role, not only on the photo-generation of the device
electrical current, but in many cases, also minimizing the
losses due to undesired back-electron transfer reactions.
Since the paradigmatic ruthenium compounds [3], such
as N719 [4] or N3 [5], were firstly employed as sensitizers
great, progress has been achieved in the design of novel
ruthenium complexes and organic dyes.
In particular, metal-free organic molecules [6] present
several advantages over coordination compounds, such as
easy upscaling and fabrication cost, relatively large molar
extinction coefficients and abundance of raw materials.
⇑
Corresponding authors at: Institute of Chemical Research of Catalonia
(ICIQ), Avda. Països Catalans 16, 43007 Tarragona, Spain (E. Palomares).
State Key Laboratory of Polymer Physics and Chemistry, Changchun
Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun
130022, China (P. Wang). Kimika Fakultatea, Euskal Herriko Unibertsita-
tea (UPV/EHU), Donostia International Physics Center (DIPC), P.K. 1072,
20080 Donostia, Spain (D. Casanova).
http://dx.doi.org/10.1016/j.orgel.2014.08.058
1566-1199/Ó 2014 Elsevier B.V. All rights reserved.

C. Climent et al. / Organic Electronics 15 (2014) 3162–3172
3163
Although the most common push–pull molecular structure
following the donor-bridge-acceptor pattern (D– –A) rep-
ril unit as the donor moiety, the cyanoacrylate acid as the
acceptor, and the selenophene-thiophene (LC95 dye) and
p
resents a very promising scheme for efficient molecular
sensitizers, the photoelectric conversion yields of DSSC
remain below the competitive cost-effective rate limit. In
thiophene-selenophene (LC107 dye) moieties as the p-
bridges (Scheme 1). We have characterized their photo-
physical properties in solution and explored the charge
photogeneration and photovoltage of DSSCs using two dif-
ferent redox shuttles. We also compare the performance of
LC95 and LC107 to three sensitizers with the same donor
the organic D–p–A type dyes, triarylamine derivatives are
among the most common donor groups, while the cyano-
acrylate acid is in many cases chosen as the acceptor and
anchoring group on an n-type semiconductor surface, such
and acceptor groups and different
p-bridge, that is, bithi-
as TiO2. The
p
-linker, typically containing one or more con-
ophene in C214, [9] biselenophene in C216 [9] and cyclo-
pentadienyl (CPDT) in C218 [10]. We perform quantum
calculations in order to understand the differences
between these dyes and to rationalize the effects of varying
jugated rings, establishes the interaction and the spatial
separation between the two molecular ends, and it is used
to tune the photophysical properties of the sensitizer.
Variations on the linker unit are a common strategy to
the position of chemical moieties on the p-bridge over the
enhance the optical response of the D–
to increase the efficiencies of DSSCs. In particular, the
extension of the -conjugation system is typically used
p
–A dye in order
final device performance. We aim, finally, to determine
whether the presence of molecular dipoles on the TiO2 sur-
face, the chemical nature of the p-bridge, or both factors
p
to red shift and to broaden the absorption spectra [7].
Alternatively, the decrease of the molecular energy gap
can be attained via rigidification of the bridge. Suppression
of the rotational disorder of molecules enhances the delo-
are key to enhance the charge recombination reactions
that ultimately limit the device efficiency under working
conditions.
calization capacity of
[8]. Nevertheless, the role of the linker (
p
-electrons, leading to smaller gaps
-bridge) on the
p
2. Experimental section
back-electron transfer between the photo-injected elec-
trons at the TiO2 surface and the oxidized electrolyte is
not yet fully understood and is still objective of many
discussions.
In the present study we perform the complete synthesis
of two new dyes with the bis(4-hexyloxyphenylamino)sty-
2.1. Materials
All the reagents were purchased from Sigma–Aldrich or
Alfa Aesar and they were used as received without further
purification unless otherwise stated.
Scheme 1. Molecular structures of the dyes employed in our study.

3164
C. Climent et al. / Organic Electronics 15 (2014) 3162–3172
2.2. Synthesis
was removed. Finally, the crude was purified by column
chromatography (SiO2, hexane/CH2Cl2 8:2) to obtain yel-
low oil as
product. (2.9 g, 42% yield).). 1H NMR
2.2.1. General synthesis of 3
a
The preparation of the (4-(bis(4-(pentyloxy)phenyl)-
amino)phenyl)boronic acid (Scheme 2) is the common
starting point towards the synthesis of the LC95 and
LC107 dyes.
(400 MHz, CDCl3) dH: 7.21 (d, J = 9.1 Hz, 2H); 7.00 (d,
J = 9.1 Hz, 4H); 6.79 (d, J = 9.1 Hz, 2H) 3.90 (t, J = 6.7 Hz,
4H); 1.75 (m, 4H); 1.39 (m, 12H); 0.91 (t, J = 6.6 Hz, 6H).
(4-(Bis(4-(pentyloxy)phenyl)amino)phenyl)boronic acid
(3). First of all, it is always necessary to normalize the nBuLi
with diphenilacetic acid. In a Schlenk flask, 2 (2.9 g,
5.53 mmol) and 50 ml of dry THF were added and were
stirred under argon atmosphere at À78 °C. nBuLi 1.6 M in
hexane (5.58 ml, 8.85 mmol) was added dropwise during
one hour and the mixture was stirred for 1 h at À78 °C. Fol-
lowing this, B(OMe)3 (731.42 ml, 276.5 mmol) was added
dropwise during 1 h and the reaction was stirred again
for 1 h at À78 °C. After heating to room temperature, the
solution was extracted with Et2O and the mixture was then
washed with H2O and HCl. The organic layer was dried
over Na2SO4 and the solvent removed. The crude was puri-
fied adding hexane to obtain a white solid. (1.62 g, 60%
yield).
4-(Hexyloxy)iodobenzene (1). In a round bottom flask
with 550 ml of acetone, 4-iodophenol (10 g, 45.46 mmol),
potassium carbonate (8.8 g, 63.6 mmol) and 18-crown-6
(2.25 g, 8.5 mmol) were added. The mixture was stirred
and then bromohexane (9 ml, 63.6 mmol) was added. The
reaction under argon was refluxed overnight. After cooling
at room temperature the crude was filtered off and the sol-
vent removed. The crude was extracted with Et2O and
washed with brine and H2O. The solution was dried over
Na2SO4 and the solvent was removed. The crude was puri-
fied by column chromatography (SiO2, hexane/CH2Cl2 9:1)
to obtain finally colorless oil. (12.7 g, 93% yield). 1H NMR
(400 MHz, CDCl3) dH: 7.53 (d, J = 8.8 Hz, 2H); 6.66 (d,
J = 8.8 Hz, 2H); 3.89 (t, J = 6.6 Hz, 2H); 1.76 (m, 2H); 1.39
(m, 6H); 0.91 (t, J = 6.6 Hz, 3H).
4-Bromo-N,N-bis(4-(hexyloxy)phenyl)aniline (2). 4-Bro-
moaniline (2.93 g, 17.03 mmol), KOH (15.32 g, 27.2 mmol),
2.2.2. Synthesis of LC95
2,5-Dibromoselenophene (4). In a Schlenk flask, selen-
ophene (1 g, 7.63 mmol), N-bromosuccinimide (3.39 g,
19.08 mmol) and 20 ml of dimethylformamide were added
and the mixture was stirred overnight at room tempera-
ture and in the dark (Scheme 3). Then a solution of Na2S2-
O3 was added to remove excess of bromine. After that, the
solution was extracted with Et2O and washed with brine
and H2O. Then the organic layer was dried over MgSO4
CuI
(0.51 g,
2.72 mmol),
phenanthroline
(0.49 g
2.72 mmol) were added in a round flask. Then 1 (13 g
40 mmol) and 25 ml of toluene were added. The reaction
was stirred and heated up to 165 °C under argon overnight.
After cooling to room temperature the solution was
extracted with CH2Cl2 and washed with brine and H2O.
The organic layer was dried over Na2SO4 and the solvent
OC6H13
OC6H13
NH2
OH
OC6H13
Br
nBuLi / B(OMe)3
THF
Br
N
2
CuI / KOH/ Phenantroline
Toluene
K2CO3
18-Crown-6
Acetone
N
3
I
I
1
OH
C6H13
O
Br
C6H13
O
B
OH
Scheme 2. Synthetic route of (4-(bis(4-(pentyloxy)phenyl)amino)phenyl)boronic acid (3).
OC6H13
Br
OHC
Br
CHO
Se
PdII(dppf)2Cl2 / K2CO3 (1M)
DME
Se
OH
OH
NBS
DMF
S
Se
S
Br
B
N
3
5
OH
4
C6H13
O
B
OC6H13
OC6H13
OH
cianoacetic acid / piperidine
CHCl3
PdII(dppf)2Cl2 / K2CO3 (1M)
DME
COOH
NC
N
6
N
7
CHO
S
S
Se
Se
LC95
C6H13
O
C6H13O
Scheme 3. Synthetic route of the LC95 dye.

C. Climent et al. / Organic Electronics 15 (2014) 3162–3172
3165
OHC
OHC
PdII(dppf)2Cl2 / K2CO3 (1M)
DME
Se
S
Se
Se
Se
OH
OH
S
POCl3 DMF
Cl Cl
NBS
THF
Br
Br
B
10
8
9
OC6H13
OC6H13
PdII(dppf)2Cl2 / K2CO3 (1M)
DME
OHC
Br
CHO
Se
S
S
Se
NBS
DMF
N
3
N
11
10
OH
B
CHO
C6H13
O
Se
S
C6H13
O
12
OH
OC6H13
OC6H13
COOH
N
NC
Se
N
CHO
Se
S
cianoacetic acid / piperidine
CHCl3
C6H13
O
LC107
S
12
C6H13
O
13
Scheme 4. Synthetic route of the LC107 dye.
and concentrated. Finally, the crude was purified by col-
umn chromatography using hexane as eluent to obtain a
colorless liquid as a product (1.9 g, 86% yield). 1H NMR
(400 MHz, CDCl3) dH: 6.98 (s, 2H).
J = 4.3 1H); 7.05 (d, J = 9.1 Hz, 4H); 6.89 (d, J = 8.7 Hz, 2H);
6.83 (d, J = 8.7 Hz, 4H); 3.92 (t, J = 6.7 Hz, 4H); 1.75 (m, 4H);
1.39 (m, 12H); 0.91 (t, J = 6.6 Hz, 6H) 13C NMR (100 MHz,
CDCl3) dC: 184.1; 156.4; 153.5; 149.6; 147.5; 147.2;
141.1; 140.6; 136.5; 128.5; 127.4; 126.9; 126.0; 125.5;
123.2; 120.3; 115.9; 68.8; 32.1; 30.2; 29.8; 26.3; 23.13;
21.6; 18.0; 14.6.
5-(5-Bromoselenophen-2-yl)thiophene-2-carbaldehyde (5).
In a Schlenk flask, 4 (261 mg, 0.91 mmol), (5-formylthio-
phen-2-yl) boronic acid (143 mg, 0.91 mmol), PdII(dppf)Cl2
(30 mg, 0.041 mmol) and 28 ml of dimethoxyethane were
added and the mixture was degassed. Then the solution
was stirred at room temperature for 30 min. After this time,
1 ml of K2CO3 2 M was added and the mixture was degassed
again. Then the mixture was heated up to 90 °C for 2 h. After
cooling at room temperature, H2O was added and the solu-
tion was extracted with Et2O and washed with brine. Then
the organic layer was dried over MgSO4 and concentrated.
Finally, the crude was purified by column chromatography
(SiO2, hexane/ethyl acetate 8:2) to obtain a yellow solid as
a product (31 mg, 20% yield). 1H NMR (400 MHz, CDCl3)
dH: 9.84 (s, 1H); 7.62 (d, J = 4 HZ, 1H); 7.22 (d, J = 4 Hz,
1H); 7.10 (d, J = 4 Hz, 1H). 13C NMR (100 MHz, CDCl3) dC:
183.0; 148.6; 143.2; 142.6; 137.7; 134.6; 128.7; 125.5;
118.3.
5-(5-(4-(Bis(4-(hexyloxy)phenyl)methyl)phenyl)thiophen-
2-yl)selenophene-2-carbaldehyde (6). In a Schlenk flask, 5
(61 mg. 0.19 mmol), 3 (135 mg, 0.27 mmol), PdII(dppf)Cl2
(6.3 mg, 0.0086 mmol) and 10 ml of dimethoxyethane
were added and the mixture was degassed. Then the solu-
tion was stirred at room temperature for 30 min. After this
time 1 mL of K2CO3 2 M was added and the mixture was
degassed again. Then the mixture was heated up to 90 °C
for 2 h. After cooling at room temperature, H2O was added
and the solution was extracted with Et2O and washed with
brine. Then the organic layer was dried over MgSO4 and
concentrated. The crude was purified by column chroma-
tography (hexane/ethyl acetate 8:2) to obtain a red solid
as a product. (98 mg, 75% yield). 1H NMR (400 MHz, CDCl3)
dH: 9.69 (s, 1H); 7.84 (d, J = 4.3 Hz, 1H); 7.37 (d, J = 9.1 Hz,
2H); 7.34 (d, J = 4.3 Hz, 1H) 7.21 (d, J = 4.3 Hz, 1H); 7.08 (d,
LC95 sensitizer (7). In
a Schlenk flask, 6 (78 mg,
0.11 mmol), cyanoacetic acid (29 mg, 0.33 mmol), piperi-
dine (197 mg 2.3 mmol) and 20 ml of dry chloroform were
added and was refluxed overnight. Then water (20 ml) was
added. The solution was acidified with 20% aqueous HCl
and extracted with CHCl3. The organic layer was dried over
anhydrous Na2SO4 and concentrated. The crude was puri-
fied by column chromatography (CHCl3/methanol 9:1) to
obtain a red solid as a product. The final product was dis-
solved in CHCl3 and washed with a 2 M HCl aqueous solu-
tion and water. The removal of solvent under a reduced
pressure gave a violet solid as a product (40 mg, 48% yield).
1H NMR (400 MHz, DMSO) dH: 8.41 (s, 1H); 7.92 (d,
J = 4.3 Hz, 1H); 7.68 (d, J = 4.3 Hz, 2H); 7.51 (d, J = 4.3 Hz,
2H) 7.46 (d, J = 9.1 Hz, 2H); 7.05 (d, J = 9.1 Hz 4H); 6.92
(d, J = 9.0 Hz, 4H); 6.72 (d, J = 9.1 Hz, 2H); 3.94 (t,
J = 6.7 Hz, 4H); 1.70 (m, 4H); 1.39 (m, 12H); 0.88 (t,
J = 6.7 Hz, 6H) 13C NMR (100 MHz, DMSO) dC: 164.1;
156.4; 153.5; 149.6; 147.5; 147.2; 141.1; 140.6; 136.5;
128.5; 127.4; 126.9; 126.0; 125.5; 123.2; 120.3; 115.9;
68.8; 30.99; 28.69; 25.19; 21.6; 22.06; 13.91.
2.2.4. Synthesis of LC107
2-Bromoselenophene (8).
A solution of selenophene
(0.6 g 4.61 mmol) and N-bromosuccinimide (NBS) (0.9 g
5.07 mmol) in THF (25 ml) was stirred under argon at room
temperature overnight (Scheme 4). After that, the solvent
was evaporated and distilled to obtain colorless oil
(0.37 g, 38% yield). 1H NMR (400 MHz, DMSO) dH: 8.10
(dd, J = 6.0 Hz, 1.4 Hz, 1H); 7.33 (dd, J = 4.0 Hz, 1.4 Hz,
1H); 7.09 (dd, J = 6.0 Hz, 4 Hz, 1H).

3166
C. Climent et al. / Organic Electronics 15 (2014) 3162–3172
5-Bromoselenophene-2-carbaldehyde (9). A solution of 8
CDCl3) dC: 184.1; 156.4; 153.5; 149.6; 147.5; 147.2;
141.1; 140.6; 136.5; 128.5; 127.4; 126.9; 126.0; 125.5;
123.2; 120.3; 115.9; 68.8; 32.1; 30.2; 29.8; 26.3; 23.13;
21.6; 18.0; 14.6.
(0.37 g, 1.75 mmol), DMF (17 ml) and POCl3 (0.81 g,
5.26 mmol) was stirred at room temperature for 30 min.
After that, the mixture was refluxed for 4 h. After cooling
at room temperature, H2O was added and the solution
was extracted with Et2O and washed with brine. Then
the organic layer was dried over MgSO4 and concentrated.
The crude was purified by column chromatography (petro-
leum ether/dichloromethane 8:2) to obtain a red oil as a
product. (0.25 g, 60% yield) 1H NMR (400 MHz, CDCl3) dH:
9.66 (s, 1H); 7.74 (d, J = 4.4 Hz, 1H); 7.20 (d, J = 4.4 Hz, 1H).
5-(Thiophen-2-yl)selenophene-2-carbaldehyde (10). In a
Schlenk flask, 5 (191 mg, 0.70 mmol), thiophen-2-ylboron-
ic acid (148 mg, 1.16 mmol), PdII(dppf)2Cl2 (26.43 mg,
0.036 mmol) and 20 ml of dimethoxyethane were added
and the mixture was degassed. Then the solution was stir-
red at room temperature for 30 min. After this time, 3 ml of
K2CO3 2 M were added and the mixture was degassed
again. Then the mixture was heated up to 90 °C for 2 h.
After cooling at room temperature, H2O was added and
the solution was extracted with Et2O and washed with
brine. Then the organic layer was dried over MgSO4 and
concentrated. The crude was purified by column chroma-
tography (hexane/ethyl acetate 8:2) to obtain a violet solid
as a product. (128 mg, 67% yield). 1H NMR (400 MHz,
DMSO) dH: 9.77 (s, 1H); 8.19 (d, J = 4.3 Hz, 1H); 7.70 (dd,
J = 5.1 Hz, 1.2 Hz 1H); 7.64 (d, J = 4.3 Hz, 1H) 7.57 (dd,
J = 3.7 Hz, 1.1 Hz 1H); 7.15 (dd, J = 5.1 Hz, 3.7 Hz 1H). 13C
NMR (100 MHz, CDCl3) dC: 184.1; 152.8; 147.7; 140.6;
138.8; 128.6; 127.5; 127.0; 126.5.
5-(5-Bromothiophen-2-yl)selenophene-2-carbaldehyde
(11). In a Schlenk flask, 10 (90 mg, 0.37 mmol) and N-
bromosuccinimide (73 mg, 0.41 mmol) were dissolved
in CHCl3. The solution was stirred at room temperature
and in the dark for 2 days. H2O was then added (7 ml),
the crude was extracted in CH2Cl2, and the organic layer
was dried over Na2SO4 and concentrated to obtain a
green solid. (78 mg, 66% yield). 1H NMR (400 MHz,
CDCl3) dH: 9.73 (s, 1H); 7.86 (d, J = 4.3 Hz, 1H); 7.33 (d,
J = 4.3 Hz 1H); 7.64 (d, J = 4.3 Hz, 1H) 7.02 (dd,
J = 14 Hz, 4.3 Hz 1H); 13C NMR (100 MHz, CDCl3) dC:
184.0; 140.3; 131.4; 127.0; 126.5.
LC107 sensitizer (13). In a Schlenk flask, 8 (100 mg,
0.14 mmol), cyanoacetic acid (37.2 mg, 0.42 mmol), piper-
idine (252.6 mg 2.94 mmol) and 30 ml of dry chloroform
were added and was refluxed for 10 h. Then water
(25 ml) was added. The solution was acidified with 20%
aqueous HCl and extracted with chloroform. The organic
phase was dried over anhydrous sodium sulfate. The sol-
vent was removed and the crude was purified by column
chromatography (chloroform/methanol 9:1) to obtain a
red solid as a product. The final product was dissolved in
chloroform and washed with a 2 M HCl aqueous solution
and water. The removal of solvent under a reduced pres-
sure gave a violet solid as a product (64 mg, 62% yield).
1H NMR (400 MHz, DMSO) dH: 8.45 (s, 1H); 8.12 (d,
J = 4.3 Hz, 1H); 7.37 (d, J = 4.3 Hz, 2H); 7.52 (d, J = 9 Hz,
2H) 7.39 (d, J = 4.3 Hz, 1H); 7.05 (d, J = 9.0 Hz 4H); 6.92
(d, J = 9.0 Hz, 4H); 6.75 (d, J = 9.0 Hz, 2H); 3.94 (t,
J = 6.7 Hz, 4H); 1.70 (m, 4H); 1.39 (m, 12H); 0.90 (t,
J = 6.6 Hz, 6H) 13C NMR (100 MHz, DMSO) dC: 166.2;
156.4; 153.5; 149.6; 147.5; 147.2; 141.1; 140.6; 136.5;
128.5; 127.4; 126.9; 126.0; 125.5; 123.2; 120.3; 115.9;
68.8; 30.99; 28.69; 25.19; 21.6; 22.06; 13.91.
2.3. Device fabrication
A screen-printed double layer film of interconnected
TiO2 particles was used as the mesoporous negative elec-
trode. A 2.4 lm thick transparent layer of 25 nm sized tita-
nia particles were first printed on the fluorine-doped SnO2
(FTO) conducting glass electrode and further coated by a
5.5 lm thick scattering layer of 400 nm sized titania parti-
cles. The resulting electrodes were gradually heated under
airflow at 325 °C for 5 min, 375 °C for 5 min, 450 °C for
15 min, and 500 °C for 15 min. Electrodes as prepared were
submerged into a 40 Á 10À3 M aqueous TiCl4 solution at
70 °C for 24 min and then washed with ethanol. Electrodes
were heated again at 500 °C for 30 min and cooled to 50 °C
before immersing them into a dye solution containing
5-(5-(4-(Bis(4-(hexyloxy)phenyl)methyl)phenyl)thiophen-
2-yl)selenophene-2-carbaldehyde (12):. In a Schlenk flask,
11 (61 mg. 0.19 mmol), 3 (135 mg, 0.27 mmol), PdII(dppf)2-
Cl2 (6.3 mg, 0.0086 mmol) and 10 ml of dimethoxyethane
were added and the mixture was degassed. Then the solu-
tion was stirred at room temperature for 30 min. After this
time, 1 ml of K2CO3 2 M was added and the mixture was
degassed again. Then the mixture was heated up to 90 °C
for 2 h. After cooling at room temperature, H2O was added
and the solution was extracted with Et2O and washed with
brine. Then the organic layer was dried over MgSO4 and
concentrated. The crude was purified by column chroma-
tography (hexane/ethyl acetate 8:2) to obtain a red solid
as a product. (98 mg, 75% yield). 1H NMR (400 MHz, CDCl3)
dH: 9.69 (s, 1H); 7.84 (d, J = 4.3 Hz, 1H); 7.37 (d, J = 9.1 Hz,
2H); 7.34 (d, J = 4.3 Hz, 1H) 7.21 (d, J = 4.3 Hz, 1H); 7.08 (d,
J = 4.3 1H); 7.05 (d, J = 9.1 Hz, 4H); 6.89 (d, J = 8.7 Hz, 2H);
6.83 (d, J = 8.7 Hz, 4H); 3.92 (t, J = 6.7 Hz, 4H); 1.75 (m, 4H);
1.39 (m, 12H); 0.91 (t, J = 6.6 Hz, 6H) 13C NMR (100 MHz,
LC95 and LC107 sensitizer (150 lM) in acetonitrile/tert-
butanol (v/v, 1/1) overnight. Then the films were washed
with acetonitrile and dried by air flow. The platinized
counter electrode was carried out by applying a drop of
5 Á 10À3 M H2 PtCl6 in ethanol dry solution and spreading
onto the conducting glass substrate (FTO). The coated glass
was heated under airflow at 390 °C for 15 min. Finally, the
dye-sensitized working electrode and the counter elec-
trode were assembled in a sandwich form using a thermo-
plastic frame that melts around 100 °C. The counter
electrode has an internal space, which was filled with a
liquid electrolyte using a vacuum backfilling system. After
that, the hole was sealed with a Bynel sheet and a thin
glass cover by heating. Both cobalt (Co-phen) and iodine
(BY01) electrolytes were used in this study. The Co-phen
electrolyte is composed of 0.25 M tris(1,10-phenanthro-
line)cobalt(II)
di[bis(trifluoromethanesulfonyl)imide],
0.05 M tris(1,10-phenanthroline)cobalt(III) tris[bis(trifluo-
romethanesulfonyl)imide], 0.5 M 4-tert-butylpyridine and

C. Climent et al. / Organic Electronics 15 (2014) 3162–3172
3167
0.1 M lithium bis(trifluoromethanesulfonyl)imide in aceto-
nitrile. The BY01 electrolyte solution consist in 0.03 M of
iodine, 0.05 M of lithium iodide, 1.0 M of 4-tert-butylpyri-
dine, 0.1 M of guanidinium thiocyanate and 1.0 M of 1,3-
dimethylimidazolium iodide in acetonitrile.
2.4. Characterization measurements
UV–visible and fluorescence spectra were recorded in
1 cm path length quartz cell on a Shimadzu UV spectro-
photometer 1700 and Aminco-Bowman Series 2 lumines-
cence spectrometer, respectively. 1H NMR experiments
were measured on a Bruker 500 MHz spectrometer using
tetramethylsilane (TMS) as reference compound.
1H and 13C NMR spectra were recorded on both a Bruker
Avance 400 (400 MHz for 1H and 100 MHz for 13C) and a
Mercury Varian 300 (300 MHz for 1H and 75 MHz for 13C)
spectrometer. The deuterated solvents used are indicated;
chemical shifts, d, are given in ppm, referenced to the sol-
vent residual signal (1H, 13C). Coupling constants, J, are
given in Hz.
Fig. 1. Electronic absorption spectra of LC95 and LC107 sensitizers
dissolved in chloroform. Vertical bars correspond to the lowest singlet-to-
singlet computed electronic transitions and their strength (in arbitrary
units). Computational values have been shifted to match the absorption
maxima of the main peak.
2.5. Device characterization
The photovoltaic measurements were carried out with a
150 W solar simulator from ABET Technologies with the
appropriate set of filters for the correct simulation of the
1.5 AM G solar spectrum. The incident light power was
measured to be 1000 W m2 with a calibrated silicon photo-
diode. The applied potential and cell current were mea-
sured with a Keithley 2400 digital source meter. Current
versus voltage (IV) curves were measured automatically
with home-built Labview software. Transient photoelectri-
cal experiments were measured with an Autolab-
PGSTAT302N electrochemical workstation. Details on
charge extraction and photovoltage decay measurements
can be found in our previous publication [8b].
between different conformations can be found as Support-
ing Information.
3. Results and discussion
The light-harvesting capacity of DSSCs is strongly
dependent on the spectral response of the molecular sensi-
tizer and the capability of the electrolyte to regenerate effi-
ciently the oxidized dye. Therefore, we initially evaluated
the absorption spectra of LC95 and LC107 dyes in chloro-
form (Fig. 1). The absorption profiles for the two dyes are
very similar, with absorption maximum molar extinction
coefficients of 31.2 Â 103 MÀ1 cmÀ1 at 516 nm (LC95) and
32.5 Â 103 MÀ1 cmÀ1 at 525 nm (LC107). These measured
parameters are close to the values reported for the bithi-
ophene (C214) and biselenophene (C216) analogues
(Table 1). The absorption band at lower energies corre-
sponds to the intramolecular charge transfer transition
2.6. Computational details
Molecular structures have been optimized within the
density functional theory (DFT) [11] employing the LRC-
x
PBEh long range corrected functional [12] with the 6–
31 + G(d) basis set. Transition energies were computed with
the time-dependent version of DFT (TDDFT) [13] and the
same functional and the larger 6-311 + G(2d,2p) basis. The
effect of the solvent (chloroform) has been taken into con-
siderationin allcalculationswith thepolarizable continuum
model (PCM) [14]. Comparison to optimized geometries,
excitation energies and oscillator strengths in acetonitrile
solution are presented as Supporting Information. All calcu-
lations were performed with the Q-Chem package [15].
Molecular structures of the studied dyes might present
different conformations regarding the relative orientation
of the two five member rings in the linker and the disposi-
between the donor and the acceptor units. The p–p* tran-
sition and intermolecular charge transfer (ICT) absorption
bands for LC107 show a small shift, but it is unlikely that
this variation can be due to different intermolecular
stacking interactions.
p-p
Optical maximum absorption transitions of LC97,
LC107, C214 and C216 dyes extend across a narrow energy
range. Although computed vertical transition energies
appear systematically blue shifted with respect to experi-
mental absorption peaks, the computed relative values
are in good agreement with the experiment (Table 1).
The energy ordering and the almost 10 nm of separation
between the LC95 and LC107 maxima (ꢀ0.04 eV) is well
recovered, and the oscillator strengths show a similar
trend to the extinction coefficients.
tion of the p-bridge with the cyanoacrylic acid group. Com-
putational results discussed in the manuscript correspond
to the ground state most stable forms represented in Fig. 1.
The fused form of the linker in C218 fixes the cis disposi-
tion of the two thiophene units, while in all other cases
the two rings are preferably in the trans form. Comparison
As discussed by Wang and collaborators [16], the
molecular rigidity imposed by the cyclopentadienyl linker
(CPDT) in C218 dye has a major impact on the absorption

3168
C. Climent et al. / Organic Electronics 15 (2014) 3162–3172
Table 1
Experimental maximum absorption wavelengths kmax and extinction coefficients, and computed vertical excitation energies Ev, oscillator strengths f and
ground state dipole moments of the C214, LC107, LC95, C216 and C218 dyes in chloroform.
l
Dye
kmax (nm)
Ev (nm)
emax (103 MÀ1 cmÀ1
)
f
l (Debye)
C214
525a,523b
410
33.8a,30.5b
1.83
7.80
7.70
7.63
7.89
9.98
LC107
LC95
C216
C218
525
417
413
421
417
32.5
1.86
1.85
1.87
2.03
516
31.2
549b
555a
35.1b
62.7a
a
Ref. [10].
Ref. [9].
b
spectrum. As a result, C218 presents slightly lower transi-
tion energy and a much higher molecular extinction coef-
ficient than the non-CPDT counterparts. Again, this
behavior is also captured by computational results. Our
DFT calculations predict a fairly noticeable loss of planarity
at the conjugated bridge of C214, LC107, LC95 and C216
dyes. The torsion angles of linkers with at least one sele-
nium atom are of the order of 15° and a bit smaller (ꢀ8°)
for the bithiophene unit in C214 (Table S1). These results
are very similar to previous calculations performed for
the bithophene and biselenophene cases [9]. It is also
worth mentioning that steric hindrance prevents the
D–p–A structure provides the proper hole and particle sep-
aration for a fast interfacial electron injection from the
excited state to the titania conduction band and dye regen-
eration. The HOMOs of C214, LC107, LC95 and C216 are
virtually degenerate. On the other hand, there are notice-
able differences in the energies of the first unoccupied
orbital (Table S5). The selenophene ring has a larger stabil-
coplanarity between the donor and p-bridge moieties. This
effect is present in the entire set of dyes of Table S1, with
dihedral angles between the phenyl of the triarylamine
group and the linker being practically the same in all cases
(26–28°).
Decomposition of the TDDFT excitation indicates that
the main contribution to the optical state corresponds to
the HOMO-to-LUMO electronic promotion with sizeable
participation of HOMO-1 and LUMO + 1 orbitals (Table S3
in Supporting Information). In all cases, the HOMO is
mostly localized over the triarylamine donor group, while
the LUMO is centered at the anchoring moiety with some
Fig. 3. Schematic representation of the HOMO and LUMO energy diagram
of LC95, LC107, C214, C216. The two rings in the conjugated linkers in the
delocalization towards the
est conjugated ring of the
p
-linker, mainly over the clos-
D-X1-X2-A dyes are indicated as (X1–X2), where
thiophene and selenophene, respectively.
T and Se refer to
p-bridge (Fig. 2). This common
Fig. 2. Isodensity representation of the frontier molecular orbitals of LC107 dye involved in the S0 to S1 electronic transition. LC95 orbitals show very similar
distributions.

C. Climent et al. / Organic Electronics 15 (2014) 3162–3172
3169
ization effect on the LUMO of the cyanoacrylic acid. As a
result, LC107 and C216 present LUMOs with lower ener-
gies than LC95 and C214 dyes (Fig. 3). The stabilization
effect due to the second five-member conjugated ring (clo-
ser to the donor moiety) is less important. Thus, the small
differences between excitation energies of these dyes
(Fig. 1 and Table 1) are controlled by the energies of their
LUMOs, i.e. C214 > LC95 > LC107 > C216, which can be eas-
ily tuned and understood by the presence and disposition
of thiophene and/or selenophene rings in the separator.
The short-circuit photocurrent density (jsc) is directly
linked to the light harvesting ability of the molecular dyes,
that is, the probability of the optical transition. This is
reflected by the rather good correlation between the com-
puted oscillator strengths in Table 1 and the measured jsc.
Moreover, effective electron injection occurs for all studied
dyes as the LUMO energy level is well above the TiO2 con-
duction band. Thus, the differences in the IPCE (Incident
Photon to Current Efficiency conversion) (Fig. 4) can be
attributed to a better packing of the C218 dye, which
allows for a better TiO2 surface coverage. Interestingly,
devices made using iodine or the Co-phen electrolyte show
similar IPCE values for LC107 and LC95, while there is an
improvement for devices made with C214 when using
the iodine electrolyte implying that the dye ground state
regeneration by the electrolyte occurs more efficiently
with iodine electrolyte in the case of C214. A perfect match
between the measured IPCE and the photocurrent mea-
sured under standard conditions (1 sun at 1.5 AMG solar
spectrum) can be seen for the two dyes in Figs. 4 and 5.
On the other hand, the open circuit photovoltage (Voc
)
in DSSC devices is established by the quasi-Fermi level of
the dye-coated titania film and the red/ox potential of
the electrolyte. Hence, the Voc is the result of two main
contributions: (i) a thermodynamic factor determined by
the position of the conduction band edge, and (ii) a kinetic
contribution linked to the density of conduction band elec-
trons in the semiconductor that may recombine either
with the oxidized dye or the oxidized electrolyte. More-
over, the dipole moment of the adsorbed sensitizer can
induce a shift in the TiO2 conduction band, and the high
Voc observed in several organic dyes has been attributed
to a positive dipole moment orientation towards the sur-
face [17]. In fact, it has been shown that, for certain dyes,
the Voc of DSSCs linearly increases with the sensitizer’s
dipole moment component perpendicular to the semicon-
ductor surface [18].
Our results show a good relationship between mea-
sured Voc of the cell and the computed dipole moments
of the organic dyes (Table 1). Noticeably, C218 presents a
much larger dipole moment than the rest of dyes, which
can be attributed to structural factors, i.e. larger molecular
Fig. 4. IPCE measurements for all devices containing as electrolytes (A) Co-phen and (B) BY01 electrolytes.
Fig. 5. jÀV characteristics measures at irradiation of 100 mW cmÀ2 AM1.5 sunlight of cells with Co-phen (A) and BY01 (B) electrolytes.

3170
C. Climent et al. / Organic Electronics 15 (2014) 3162–3172
linearity and planarity. However, this variation by itself
cannot explain the much larger Voc measured for the CPDT
linker compared to the more flexible bridges in C214,
LC107, LC95 and C216. We would also like to emphasize
that the very small differences between LC95 and LC107
are well reproduced by their dipole moments.
On the other hand, it is sometimes argued that the
geometry of the sensitizer can have a crucial impact on
the measured photovoltages [19]. The molecular planarity
favors ordered close packing of adsorbed dyes onto the
semiconductor surface. As discussed above, in the set of
studied organic dyes there are basically no differences in
the molecular planarity of LC95, LC107, C214 and C216
dyes, thus, we do not expect major differences regarding
their coverage of the semiconductor surface. The planarity
at the CPDT linker in C218, however, has effects on the sur-
face coverage leading to higher photocurrent as it can be
seen in Figs. 4 and 5. All the relevant physical parameters
Table 2
Photovoltaic parameters measured at the AM1.5 full sunlight conditions of C214, LC107, LC95, C216 and C218.a
Electrolyte
Dye
Voc (V)
jsc (mA/cm2)
FF
Efficiency (%)
Co(II)phen/Co(III)phen
C214
LC107
LC95
C216
C218
0.873
0.817
0.813
0.832
0.929
10.03
11.00
10.17
10.72
14.24
0.70
0.72
0.71
0.71
0.73
6.1
6.5
6.3
6.3
9.6
Iodine/iodide
C214
LC107
LC95
C216
C218
0.757
0.704
0.700
0.700
0.806
11.13
10.27
9.48
10.66
12.87
0.70
0.74
0.76
0.72
0.72
5.9
5.4
5.0
5.4
7.5
a
The spectral distribution of our light source simulates de AM1.5G solar emission (ASTM G173-03). The measured photocurrent matches the overall
integral of the measured IPCE spectrum with an error less than 5%.
Fig. 6. Charge stored in the dye-grafted titania film (Q) as a function of open-circuit photovoltage (Voc) with Co-phen (A) and BY01 (B) electrolytes.
Fig. 7. Plots of lifetime of photoinjected electrons in titania (s) as a function of charge with Co-phen (A) and BY01 (B) electrolytes.

C. Climent et al. / Organic Electronics 15 (2014) 3162–3172
3171
for the IV curves represented in Fig. 5 are gathered in
Table 2.
position. Nevertheless, the larger effect over the measured
valued of Voc is, unequivocally, due to the reduction of the
electron recombination dynamics between the electrons at
the TiO2 surface and the oxidized electrolyte.
We carried out charge extraction and transient photo-
voltage decay measurements, as reported before, to ana-
lyze if the different dipoles promoted by the dye surface
coverage affect the position of the TiO2 conduction band
edge and, furthermore, if this shift can be correlated to
slower electron lifetimes. As it can be seen in Fig. 6, the
measured charge (the total charge at the device at a given
photovoltage) versus cell photovoltage has similar slope
and the shift is within 25 mV between, for example, the
C214 and the LC95 dye when using cobalt or iodine red/
ox electrolytes. Hence, we can conclude that the different
dyes, when absorbed onto the TiO2 surface, induce mild
variations on the energy level of the TiO2 conduction band
edge, however, the relative shift cannot be responsible for
the 100 mV difference in voltage observed for the dye
C218. The much larger value of Voc measured for the
C218 dye is also a consequence of the slower recombina-
tion dynamics between the photo-injected electrons and
the oxidized electrolyte (Cobalt(III)phen or iodide). The lar-
ger electron lifetime value for the C218 devices when com-
pared with the other dyes (Fig. 7) is in good agreement
with better molecular packing at the TiO2 surface blocking
the back-electron transfer of the TiO2 electrons to the
electrolyte.
Acknowledgements
Financial support for this work was provided by the
Spanish Ministerio de Economía y Competitividad through
Projects CTQ2011-23862-C02-02 and CTQ2010-18859.
D.C. gratefully acknowledges the IKERBASQUE, Basque
Foundation for Science for financial support. E.P. would like
to thank ICIQ and ICREA for financial support for this pro-
ject. P.W. thanks the financial support from National 973
Program (No. 2011CBA00702).
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.orgel.2014.08.058.
References
[1] B. O’Regan, M. Grätzel, Nature 353 (1991) 737–740.
[2] A. Yella, H.-W. Lee, H.N. Tsao, C. Yi, A.K. Chandiran, M.K.
Nazeeruddin, E.W.-G. Diau, C.-Y. Yeh, S.M. Zakeeruddin, M. Grätzel,
Science 334 (2011) 629–634.
[3] A. Reynal, E. Palomares, Eur. J. Inorg. Chem. 2011 (2011) 4509–4526.
[4] M.K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E. Mueller, P.
Liska, N. Vlachopoulos, M. Graetzel, J. Am. Chem. Soc. 115 (1993)
6382–6390.
[5] (a) M.K. Nazeeruddin, S.M. Zakeeruddin, R. Humphry-Baker, M.
Jirousek, P. Liska, N. Vlachopoulos, V. Shklover, C.-H. Fischer, M.
Grätzel, Inorg. Chem. 38 (1999) 6298–6305;
4. Conclusions
We have designed and synthesized two new organic D–
p–A dyes with a p-bridge that combines the selenophene
and the thiophene moieties and we have compared their
efficiency in DSSC with other paradigmatic organic dyes.
The measured parameters for IV curves, short-circuit pho-
tocurrent (jsc), open-circuit voltage (Voc) and efficiency
have been rationalized considering the formation of
dipoles at the semiconductor surface, due to the dye’s
structural properties and their effect on the TiO2 conduc-
tion band edge and the electron lifetime. We have shown
that the device photocurrent was higher for the dye that
exhibited better molecular packing and was independent
of the molecular dipole and its possible influence over
charge injection from the excited state of the dye. More-
over, the different dyes shifted different energy position
of the TiO2 conduction band edge as measured using the
charge extraction techniques. A shift of about 25 mV is
observed between the different set of dyes. However, these
small differences cannot explain by themselves the large
difference in Voc observed for C218 at 1 sun either when
using cobalt or iodine electrolytes. In fact, transient photo-
voltage decay measurements demonstrate that for C218
based devices the electron lifetime was longer too. Thus,
the larger Voc observed for C218 can be explained not only
by the TiO2 conduction band edge shift due to the molecu-
lar dipole at the C218 dyes, but also due to the better
molecular packing that retards the back-electron transfer
from the photo-injected electrons at the TiO2 and the oxi-
dized electrolytes. Based on these results we can conclude
that the molecular dipole that can be formed at the TiO2
surface may have an effect over the TiO2 conduction band
(b) M.K. Nazeeruddin, R. Splivallo, P. Liska, P. Comte, M. Gratzel,
Chem. Commun. (2003) 1456–1457.
[6] J.N. Clifford, M. Planells, E. Palomares, J. Mater. Chem. 22 (2012)
24195–24201.
[7] (a) D.P. Hagberg, T. Marinado, K.M. Karlsson, K. Nonomura, P. Qin, G.
Boschloo, T. Brinck, A. Hagfeldt, L. Sun, J. Org. Chem. 72 (2007) 9550–
9556;
(b) J. Liu, R. Li, X. Si, D. Zhou, Y. Shi, Y. Wang, X. Jing, P. Wang, Energy
Environ. Sci. 3 (2010) 1924–1928;;
(c) H. Ellis, S.K. Eriksson, S.M. Feldt, E. Gabrielsson, P.W. Lohse, R.
Lindblad, L. Sun, H. Rensmo, G. Boschloo, A. Hagfeldt, J. Phys. Chem. C
117 (2013) 21029–21036.
[8] (a) Y. Bai, J. Zhang, D. Zhou, Y. Wang, M. Zhang, P. Wang, J. Am.
Chem. Soc. 133 (2011) 11442–11445;
(b) N. Cai, R. Li, Y. Wang, M. Zhang, P. Wang, Energy Environ. Sci. 6
(2013) 139–147;
(c) J.-H. Chen, C.-H. Tsai, S.-A. Wang, Y.-Y. Lin, T.-W. Huang, S.-F.
Chiu, C.-C. Wu, K.-T. Wong, J. Org. Chem. 76 (2011) 8977–8985;
(d) K. Lim, C. Kim, J. Song, T. Yu, W. Lim, K. Song, P. Wang, N. Zu, J. Ko,
J. Phys. Chem. C 115 (2011) 22640–22646.
[9] R. Li, X. Lv, D. Shi, D. Zhou, Y. Cheng, G. Zhang, P. Wang, J. Phys. Chem.
C 113 (2009) 7469–7479.
[10] R. Li, J. Liu, N. Cai, M. Zhang, P. Wang, J. Phys. Chem. B 114 (2010)
4461–4464.
[11] (a) R.G. Parr, Y. Weitao, Density-Functional Theory of Atoms and
Molecules, Oxford University Press, New York, 1989;
(b) T. Ziegler, Chem. Rev. 91 (1991) 651–667.
[12] M.A. Rohrdanz, K.M. Martins, J.M. Herbert, J. Chem. Phys. 130 (2009)
054112.
[13] (a) M.E Casida, in: D.P. Chong (Ed.), Recent Advances in Density
Functional Methods, Part I, vol. 1, World Scientific, Singapore, 1995;
(b) E. Runge, E.K.U. Gross, Phys. Rev. Lett. 52 (1984) 997–1000.
[14] T.N. Truong, E.V. Stefanovich, Chem. Phys. Lett. 240 (1995) 253–260.
[15] Y. Shao, L.F. Molnar, Y. Jung, J. Kussmann, C. Ochsenfeld, S.T. Brown,
A.T.B. Gilbert, L.V. Slipchenko, S.V. Levchenko, D.P. O’Neill, R.A.
DiStasio Jr., R.C. Lochan, T. Wang, G.J.O. Beran, N.A. Besley, J.M.
Herbert, C.Y. Lin, T.V. Voorhis, S.H. Chien, A. Sodt, R.P. Steele, V.A.

3172
C. Climent et al. / Organic Electronics 15 (2014) 3162–3172
Rassolov, P.E. Maslen, P.P. Korambath, R.D. Adamson, B. Austin, J.
[18] S. Rühle, M. Greenshtein, S.G. Chen, A. Merson, H. Pizem, C.S.
Sukenik, D. Cahen, A. Zaban, J. Phys. Chem. B 109 (2005) 18907–
18913.
[19] (a) M. Xu, R. Li, N. Pootrakulchote, D. Shi, J. Guo, Z. Yi, S.M.
Zakeeruddin, M. Grätzel, P. Wang, J. Phys. Chem. C 112 (2008)
19770–19776;
Baker, E.F.C. Byrd, H. Dachsel, R.J. Doerksen, A. Dreuw, B.D. Dunietz,
A.D. Dutoi, T.R. Furlani, S.R. Gwaltney, A. Heyden, S. Hirata, C.-P. Hsu,
G. Kedziora, R.Z. Khalliulin, P. Klunzinger, A.M. Lee, M.S. Lee, W.
Liang, I. Lotan, N. Nair, B. Peters, E.I. Proynov, P.A. Pieniazek, Y.M.
Rhee, J. Ritchie, E. Rosta, C.D. Sherrill, A.C. Simmonett, J.E. Subotnik,
H.L.W. Iii, W. Zhang, A.T. Bell, A.K. Chakraborty, Phys. Chem. Chem.
Phys. 8 (2006) 3172–3191.
(b) S. Haid, M. Marszalek, A. Mishra, M. Wielopolski, J. Teuscher, J.-E.
Moser, R. Humphry-Baker, S.M. Zakeeruddin, M. Grätzel, P. Bäuerle,
Adv. Funct. Mater. 22 (2012) 1291–1302;
(c) Q. Chai, W. Li, S. Zhu, Q. Zhang, W. Zhu, ACS Sustain. Chem. Eng. 2
(2013) 239–247;
[16] M. Xu, M. Zhang, M. Pastore, R. Li, F. De Angelis, P. Wang, Chem. Sci. 3
(2012) 976–983.
[17] P. Chen, J.H. Yum, F.D. Angelis, E. Mosconi, S. Fantacci, S.-J. Moon,
R.H. Baker, J. Ko, M.K. Nazeeruddin, M. Grätzel, Nano Lett. 9 (2009)
2487–2492.
(d) A. El-Shafei, M. Hussain, A. Islam, L. Han, J. Mater. Chem. A 1
(2013) 13679–13686.