Squaraine-arylamine sensitizers for highly efficient p-type dye-sensitized solar cells

Article abstract of DOI:10.1021/ol301860w

New near-IR (NIR) squaraine dyes (p-SQ1 and p-SQ2) containing one and two anchoring groups were synthesized and used as the sensitizers of p-type DSSCs. The dye (0.113%) with two anchoring groups (p-SQ2) shows better performance than the dye (0.053%) with only one anchoring group (p-SQ1). Cosensitized p-type DSSCs using two dyes with complementary absorption were tested. They have broadened IPCE spectra and good cell performance.

Full text of DOI:10.1021/ol301860w

ORGANIC
LETTERS
2012
Vol. 14, No. 18
4726–4729
Squaraine-Arylamine Sensitizers for
Highly Efficient p‑Type Dye-Sensitized
Solar Cells
Chia-Hao Chang, Yung-Chung Chen, Chih-Yu Hsu, Hsien-Hsin Chou, and Jiann T. Lin*
Institute of Chemistry, Academia Sinica, Nankang, Taipei 11529, Taiwan
jtlin@gate.sinica.edu.tw
Received July 6, 2012
ABSTRACT
New near-IR (NIR) squaraine dyes (p-SQ1 and p-SQ2) containing one and two anchoring groups were synthesized and used as the sensitizers of p-type
DSSCs. The dye (0.113%) with two anchoring groups (p-SQ2) shows better performance than the dye (0.053%) with only one anchoring group (p-SQ1).
Cosensitized p-type DSSCs using two dyes with complementary absorption were tested. They have broadened IPCE spectra and good cell performance.
Dye-sensitized solar cells (DSSCs) have attracted con-
are adsorbed on the TiO2 of n-type DSSCs.5 An alternative
approach first reported by Lindquist et al. includes using
tandem cells which combine an n-type semiconductor
as the photoanode and a p-type semiconductor such as
NiO6 as the photocathode.7 With such a device design,
dyes absorbing in different spectral regions are adsorbed
on different electrodes. Theoretical calculation predicts
that an efficiency up to 43% is possible for tandem
cells.8 Compared with n-type DSSCs, research on p-type
DSSCs is still in its infancy.9 In 2008, Sun et al. developed
triphenylamine-based dyes with different acceptors for
p-type DSSCs, and the PCEs ranged from 0.03% to
€
siderable attention since the first report by Gratzel et al. in
1991.1 Although the power conversion efficiency (PCE) of
n-type DSSCs based on metal-containing dyes such as
ruthenium2 and Zn-porphyrin complexes3 have achieved
11% and 12.3% under AM 1.5 illumination, these dyes
are either expensive or complex in synthetic procedures.
Therefore, metal-free sensitizers are promising alternatives
due to their low cost, high molar absorption coefficients,
and design flexibility. So far, the best PCE of organic dyes
has achieved 10%.4
To improve the PCE of DSSCs, the sensitizers should
havebroader and intense absorption spectra. A convenient
approach is using cosensitized systems in which different
dyes which absorb in the complementary spectral region
(5) Yum, J.-H.; Baranoff, E.; Wenger, S.; Nazeeruddin, M. K.;
€
Gratzel, M. Energy Environ. Sci. 2011, 4, 842.
(6) (a) Gibson, E. A.; Smeigh, A. L.; Pleux, L. L.; Fortage, J.;
Boschloo, G.; Blart, E.; Pellegrin, Y.; Odobel, F.; Hagfeldt, A.;
€
€
(1) O’Regan, B.; Gratzel, M. Nature 1991, 353, 737.
(2) (a) Gao, F.; Wang, Y.; Shi, D.; Zhang, J.; Wang, M.; Jing, X.;
Hammarstrom, L. Angew. Chem., Int. Ed. 2009, 48, 4402. (b) Odobel,
F.; Pleux, L. L.; Pellegrin, Y.; Blart, E. Acc. Chem. Res. 2010, 43, 1063.
€
Humphry-Baker, R.; Wang, P.; Zakeeruddin, S. M.; Gratzel, M. J. Am.
€
(7) (a) He, J.; Lindstrom, H.; Hagfeldt, A.; Lindquist, S.-E. J. Phys.
€
Chem. Soc. 2008, 130, 10720. (b) Chen, C.-Y.; Wang, M.; Li, J.-Y.;
Pootrakulchote, N.; Albabaei, L.; Ngoc-le, C.-h.; Decoppet, J.-D.; Tsai,
Chem. B 1999, 103 (42), 8940. (b) He, J.; Lindstrom, H.; Hagfeldt, A.;
Lindquist, S.-E. Sol. Energy Mater. Sol. Cells 2002, 62, 265.
(8) Green, M. A. Third Generation Photovoltaics: Advanced Solar
Energy Conversion; Springer-Verlag: Berlin, Heidelberg, 2003.
(9) Hagfeldt, A.; Boschloo, G.; Sun, L.; Kloo, L.; Peterssen, H.
Chem. Rev. 2011, 110, 6595.
(10) (a) Qin, P.; Zhu, H.; Edvinsson, T.; Boschloo, G.; Hagfeldt, A.;
Sun, L. J. Am. Chem. Soc. 2008, 130, 8570. (b) Qin, P.; Wiberg, J.;
Gibson, E. A.; Linder, M.; Li, L.; Brinck, T.; Hagfeldt, A.; Albinsson,
B.; Sun., L. J. Phys. Chem. C 2010, 114, 4738.
€
€
J.-H.; Gratzel, C.; Wu, C.-G.; Zakeeruddin, S. M.; Gratzel, M. ACS
Nano 2009, 3, 3103.
(3) (a) Bessho, T.; Zakeeruddin, S. M.; Yeh, C.-Y.; Diau, E. W.-G.;
Gratzel, M. Angew. Chem., Int. Ed. 2010, 49, 6646. (b) 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.
(4) Zheng, W.; Cao, Y.; Bai, Y.; Wang, Y.; Shi, Y.; Zhang, M.; Wang,
€
F.; Pan, C.; Wang, P. Chem. Mater. 2010, 22, 1915.
r
10.1021/ol301860w
Published on Web 09/13/2012
2012 American Chemical Society

Scheme 1. Synthesis of the Dyes
Figure 1. Absorption spectra of the squaraine dyes in EtOH
(10ꢀ5 mol/L). Insert: absorption spectra on NiO films (∼0.5 μm).
0.15%.10 Later Bach11 reported arylamine dyes containing
perylene monoimide (PMI) acceptors and oligothiophene
spacers with PCEs ranging from 0.09% to 0.41%. Re-
cently, we also reported arylamine dyes for p-type DSSCs
and found that dyes with two anchoring groups more
effectively injected holes and suppressed dark currents
compared to dyes with only one anchoring group.12 So
far the best efficiency of p-type DSSCs is about one order
lower than that of n-type DSSCs due to the slow hole
mobility of the NiO electrode, fast charge recombination,
and nonideal nature of the redox mediator, iodide. As
∼50% of solar light is in the red to near-IR (NIR) region,
dyes capable of absorbing light intensely in this region
should be beneficial for light harvesting. Therefore, we set
out to develop squaraine dyes for p-type DSSCs because
they normally exhibit strong absorption in the redꢀ
NIR region (700ꢀ850 nm). This characteristic renders
squaraine dyes useful in many applications, e.g., organic
field-effect trainsistors,13 chemosensors,14 and bioimaging.15
Though there have been examples of bulk-heterojunction
solar cells16 and n-type DSSCs17 based on squaraine dyes,
there is no report of squaric acid dyes for p-type DSSCs.
Herein we report new squaraine dyes for p-type DSSCs with
one or two anchoring groups. Cosensitized p-type DSSC
cells based on one of the new dyes will also be reported.
The prototype molecular structure of p-type dyes has a
donor-spacer-acceptor skeleton, and the anchoring groups
for adsorbing on the photocathode (e.g., NiO) surface are
tethered at the donor and farther away from the acceptor.
The new squaraine dyes and their synthetic protocol of the
compounds are illustrated in Schemes 1, S1, and S2 (see
Supporting Information (SI)). Compound 1 was prepared
according to the published procedures.17g A Pd-catalyzed
Suzuki coupling18 reaction between 1and appropriate Suzuki
reagents, prepared in situ from triphenylamine derivatives
with carboxylic ester as the protection group of carboxylic
acid, afforded 2 and 3. Treatment of 2 and 3 with trifluoro-
acetic acid provided the desired products, p-SQ1 and p-SQ2.
The absorption spectra of the dyes in EtOH are shown in
Figure 1, and the relevant data are summarized in Table 1.
Both dyes show a very intense band at 646 nm with the
molar extinction coefficient exceeding 200 000 Mꢀ1cmꢀ1
.
This band is ascribed to a squaraine-centered πꢀπ* elec-
tron transition with minimal charge transfer character
from the arylamine to the squaraine (see computation,
vide infra). Small Stokes shifts (∼30 nm) of the emission
spectra (Figure S1) for the two dyes is ascribed to the
dominance of the squaraine-centered πꢀπ* transition
in the S0 fS1 transition (vide infra). The broadened absorp-
tion spectra feature upon adsorption on NiO (Figure 1) is
attributed to the aggregation of the dye molecules.19
The results of density functional calculations at the
B3LYP/6-31G* level of the dyes are included in Table S1
(see SI). The dihedral angles betweensuccessiveunits ofthe
molecules are shown in Figure S2. The twist angle between
the planar bis(indoline) squaraine segment and the aryl-
amine is ∼36° for both dyes. The computed frontier orbitals
of the compounds and their corresponding energy states
are included in Figures S3 and S4 (see SI), respectively.
(11) Nattestad, A.; Mozer, A. J.; Fischer, M. K. R.; Cheng, Y.-B.;
€
Mishra, A.; Bauerle, P.; Bach, U. Nat. Mater. 2010, 9, 31.
(12) Yen, Y.-S.; Chen, W.-T.; Hsu, C.-Y.; Chou, H.-H.; Lin, J. T.;
Yeh., M.-C. Org. Lett. 2011, 13, 4930.
(13) Smits, E. C. P.; Setayesh, S.; Anthopoulos, T. D.; Buechel, M.;
Nijssen, W.; Coehoorn, R.; Blom, P. W. M.; de Boer, B.; de Leeuw,
D. M. Adv. Mater. 2007, 19, 734.
(14) Yan, Z. Q.; Guang, S. Y.; Xu, H. Y.; Liu, X. Y. Analyst 2011,
136, 1916.
(15) Gassensmith, J. J.; Braumes, J. M.; Smith, B. D. Chem. Commun.
2009, 6329.
(16) Bagnis, D.; Beverina, L.; Huang, H.; Silvestri, F.; Yao, Y.; Yan,
H.; Pagani, G. A.; Marks, T. J.; Facchetti, A. J. Am. Chem. Soc. 2010,
132, 4074.
(17) (a) Takechi, K.; Sudeep, P. K.; Kamat, P. V. J. Phys. Chem. B
€
2006, 110, 16169. (b) Burke, A.; Schmidt-Mende, L.; Ito, S.; Gratzel, M.
Chem. Commun. 2007, 234. (c) Choi, H.; Kim, J.-J.; Song, K.; Ko, J.;
€
Nazeeruddin, M. K.; Gratzel, M. J. Mater. Chem. 2010, 20, 3280. (d)
Paek, S.; Choi, H.; Kim, C.; Cho, N.; So, S.; Song, K.; Nazeeruddin,
M. K.; Ko, J. Chem. Commun. 2011, 47, 2874. (e) Li, J.-Y.; Chen, C.-Y.;
Lee, C.-P.; Chen, S.-C.; Lin, T.-H.; Tsai, H.-H.; K.-C.; Ho, K.-C.; Wu,
C.-G. Org. Lett. 2010, 12, 5454. (f) Yan, Z.; Guang, S.; Su, X.; Xu, H.
J. Phys. Chem. C 2012, 116, 8894. (g) Shi, Y.; Hill, R. B. M.; Yum, J.-H.;
€
Dualeh, A.; Barlow, S.; Gratzel, M.; Marder, S. R.; Nazeeruddin, M. K.
(18) Suzuki, A. J. Organomet. Chem. 1999, 576, 147.
ꢀ
(19) Cornil, J.; Beljonee, D.; Calbert, J.-P.; Bredas, J.-L. Adv. Mater.
2001, 13, 1053.
Angew. Chem., Int. Ed. 2011, 50, 6619. (h) Funabiki, K.; Mase, H.; Saito,
Y.; Otsuka, A.; Hibino, A.; Tanaka, N.; Miura, H.; Himori, Y.;
Yoshida, T.; Kubota, Y.; Matsui, M. Org. Lett. 2012, 14, 1246.
Org. Lett., Vol. 14, No. 18, 2012
4727

Table 1. Electrooptical and Electrochemical Parameters of the
Dyes
Table 2. DSSC Performance Parameters and Dye Loading of
the Dyes
λ
M
abs (ε ꢁ 10ꢀ5
HOMO/
JSC
(mA cmꢀ2
VOC
η
dye loading
a
b
c
d
e
ꢀ1 cmꢀ1 a
)
λem E0‑0 λNiO Eox
nm eV nm mV
LUMO E0ꢀ0*
eV
dyes
)
(V)
FF
(%) (10ꢀ7 mol/cm2)
dyes
nm
V
P1 (THF)
2.31
2.88
1.22
1.92
0.132 0.331 0.101
0.140 0.330 0.133
0.117 0.371 0.053
0.140 0.420 0.113
1.29
1.58
1.07
1.98
p-SQ1 324 (0.31), 646 (2.40) 675 1.87 662 336 5.44/3.57 ꢀ0.83
p-SQ2 328 (0.31), 646 (2.32) 679 1.87 663 350 5.45/3.58 ꢀ0.82
P1 (EtOH)
p-SQ1 (EtOH)
p-SQ2 (EtOH)
a Recorded in EtOH at 298 K. b E0‑0 = 1240/λopt c Recorded in NiO
.
film dependent on EtOH at 298 K. d Oxidation potential reported is
referenced to the potential of ferrocene (Eox = 212 mV vs Ag/AgNO3)
which was used as an internal reference. Scan rate: 100 mV/s. Eox
1/2(Epa þ Epc). e vs NHE.
=
I2 (0.1 M)/LiI (1.0 M)/tert-butylpyridine (TBP, 0.5 M) in
CH3CN. The device performance statistics under AM 1.5
illumination are collected in Table 2. Due to the lower
solubility of p-SQ dyes in THF, ethanol was used as the
dipping solvent for the cell fabrication. Figures 2 and 3 show
the photocurrentꢀvoltage (JꢀV), dark currents, and the
incident photon-to-current conversion efficiencies (IPCEs)
of the cells. The short-circuit photocurrent density (JSC),
open circuit voltage (VOC), and fill factor (FF) of the device
are in the range of 1.22ꢀ1.92 mA cmꢀ2, 0.117ꢀ0.140 V, and
0.371ꢀ0.420, respectively, corresponding to an overall con-
version efficiency of 0.053ꢀ0.113%. The performance is
comparable to that of the dye reported by Sun (P1) (Figure
S7) (η = 0.101% in THF, η = 0.135% in ethanol) fab-
ricated and measured under similar conditions. It is impor-
tant to note that the DSSCs in ethanol conditions show
better PCEs than in THF, likely due to the higher dye load-
ing density (Table 2). In view of the unfavorable hole
injection of the dyes (vide supra), we speculate that hole
injection may take place also via hopping, in which two
anchoring groups may have an advantage.
Both the HOMO (highest occupied molecular orbital)
and LUMO (lowest unoccupied molecular orbital) of the
dyes are mainly composed of the bis(indoline) squaraine
moiety, and the HOMO has a minor contribution from
the arylamine moiety (vide infra). Consequently, p-SQ1
and p-SQ2 have nearly the same HOMO (p-SQ1: ꢀ4.50 eV;
p-SQ2: ꢀ4.53 eV) and LUMO (p-SQ1: ꢀ2.28 eV; p-SQ2:
ꢀ2.31 eV) levels. Accordingly, they also have very high
oscillator strength (f): p-SQ1, 1.83; p-SQ2: 1.82. This and
the small changes of Mulliken charge (Figure S5 and
Table S1) of arylamine in the S0 f S1 transition (p-SQ1,
0.03; p-SQ2: 0.01) imply unfavorable hole injection from
the excited dye to the NiO. Thus, the squaraine dyes in
this study are different from the dyes we reported
earlier,12 in which the HOMOs of the dyes have a
significant contribution from the arylamine.
The cyclic voltammograms and the electrochemical data
of the dyes are presented in Figure S6 and Table 1,
respectively. Only one quasi-reversible redox wave attri-
butable to the oxidation of the bis(indoline) squaraine
moiety (vide supra) was observed for the dyes. The two
dyes have very similar oxidations (ΔE = 14 mV) because
they have similar HOMO features (vide supra). In com-
parison, the dye with two carboxylic acid groups is oxi-
dized ata significantly higherpotential (ΔE > 60mV) than
that with only one carboxylic acid group in our earlier
report.12 The HOMO energy levels of the dyes were
calculated from the oxidation potential and by comparison
with ferrocene (5.1 eV).20 The HOMO/LUMO gap for
both dyes (1.87 eV) was estimated from the intersection of
the normalized absorption and emission spectra. The data
werethenused toobtainthe LUMO(LUMO= HOMOꢀ
The cell of p-SQ2 exhibited a higher short-circuit photo-
current density (JSC), open-circuit voltage (VOC), and fill
factor (FF) than that of p-SQ1. Consequently, the former
had a higher conversion efficiency than the latter (p-SQ1:
0.053%; p-SQ2: 0.113%). The higher dye loading density
of p-SQ2 (1.98 ꢁ 10ꢀ7 mol/cm2) than p-SQ1 (1.07 ꢁ 10ꢀ7
mol/cm2) isbelieved tohaveimportant contributionsto the
significantly higher current density of p-SQ2 (1.92 mA cmꢀ2
)
than p-SQ1 (1.22 mA cmꢀ2). Better dark current suppression
E0ꢀ0) energy levels (Table 1). No discernible reduction
wave due to the reduction of the dyes prevents us from
estimating the ability of hole injection6b from the excited
dye to the NiO. The LUMO levels of dyes (p-SQ1: ꢀ0.83 V;
p-SQ2: ꢀ0.82 V) are higher than that of the Iꢀ/I3 redox
ꢀ
system (0.44 V vs NHE), suggesting that the reduced dyes
can efficiently reduce the oxidized electrolyte. p-Type
DSSCs were fabricated with the use of p-SQ dyes and
nanostructured NiO. The cells had an effective area of
0.25 cm2, and the electrolyte used was composed of
Figure 2. Current densityꢀvoltage curves of P1, p-SQ1, p-SQ2
and dark-current characteristics of p-SQ1, p-SQ2 dissolved
in EtOH under AM 1.5 solar simulator of 100 mW cmꢀ2. Cell
area = 0.25 cm2.
(20) Cardona, C. M.; Li, W.; Kaifer, A. E.; Stockdale, D.; Bazan,
G. C. Adv. Mater. 2011, 23, 2367.
4728
Org. Lett., Vol. 14, No. 18, 2012

Table 3. Cosensitized System Performance Parameters of P1
and p-SQ2
dyes in vol. ratio
JSC (mA cmꢀ2
)
VOC (V)
FF
η (%)
P1:p-SQ2 = 10:0
P1:p-SQ2 = 8:2
P1:p-SQ2 = 6:4
P1:p-SQ2 = 4:6
P1:p-SQ2 = 2:8
P1:p-SQ2 = 0:10
2.88
2.78
2.72
2.90
2.48
1.92
0.140
0.142
0.147
0.154
0.145
0.140
0.330
0.324
0.305
0.302
0.327
0.420
0.133
0.128
0.122
0.135
0.118
0.113
Figure 3. IPCE plots of P1, p-SQ1, and p-SQ2.
by p-SQ2 may also be important (vide infra). The two
carboxylic acid groups in p-SQ2 may be favorable for more
upright anchoring and therefore more compact packing of
the molecules on the NiO surface. Similar to our previous
observations,12 the dye with two anchoring groups exhibited
a higher VOC: p-SQ1, 0.117 V; p-SQ2, 0.140 V. This outcome
may be attributed to more effective coverage of the naked
NiO surface by dye molecules from exposure to the elec-
trolyte, i.e., suppression of dark current. The smaller dark
current in the DSSC of p-SQ2 wasfurther supportedby the
dark current measurements (Figure 2).
Figure 4. IPCE plots of cosensitizer system based on P1 and
p-SQ2.
Electrochemical impedance was measured under a
bias of ꢀ0.15 V in the dark, and the Nyquist plots
of DSSCs with the dyes are shown in Figure S8a.
The second semicircle can be used to derive the
charge recombination resistance on the NiO surface.
The higher resistance in the DSSC of p-SQ2 (p-SQ1:
136 Ω; p-SQ2: 188 Ω) provides further support for its
lower dark current. The electrochemical impedance
measurement under illumination was also tested, and
the Nyquist plots of DSSCs are shown in Figure S8b.
Upon illumination of 100 mW cmꢀ2 light under open
circuit conditions, the radius of the intermediate
frequency semicircle in the Nyquist plot represents
the electron transport resistance. No significant dif-
ference in the electron transport resistance (Rct) was
found between the cells of p-SQ1 (156 Ω) and p-SQ2
(148 Ω).
Similar to other squaraine dyes, the dyes in this study
have somewhat narrower absorption bands which limit
light harvesting efficiency. A dye (P1) developed by Sun10a
has complementary absorption with p-SQ2. Therefore,
cosensitized p-type DSSCs using p-SQ2 with P1 were
tested. The performance data are presented in Table 3.
The UV absorption spectra on NiO and IPCE plots are
shown in Figures S9 and 4, respectively. Though we found
no apparent improvement in the cosensitized systems, the
IPCE spectra clearly indicated that both dyes had a
significant contribution to the current density at various
dye ratios, which is consistent with the UV absorption
spectra of the dye film. It is encouraging that the efficiency
was retained upon introduction of the second sensitizer in
view of the fact that thin NiO film (∼1.5 μm) was used.
In summary, we have synthesized two NIR squaraine
dyes (SQ) with high molar extinction coefficients. Theycan
be used as the sensitizers of p-type DSSCs. The dye with
two anchoring groups has better cell performance than
that with only one anchoring group due to its more
compact packing on the NiO surface and more effective
suppression of the dark current. High open-circuit volt-
age (0.140 V) and conversion efficiency (0.113%) were
achieved with one of the cells using ethanol as the dipping
solvent. Cosensitized p-type DSSCs were reported for the
first time. The cell exhibited a broader IPCE spectrum and
high cell performance with the use of a second dye having
an absorption spectrum complementary with the SQ dye.
Development of sensitizers with improved absorption at a
shorter wavelength region and improved hole injection is
currently ongoing.
Acknowledgment. We acknowledge the support of the
Academia Sinica (AC) and NSC (Taiwan) and the Instru-
mental Center of Institute of Chemistry (AC).
Supporting Information Available. Synthetic proce-
dures for new compounds and other physical properties.
These materials are available free of charge via the
Internet at http://pubs.acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 18, 2012
4729