Near-IR dye sensitization of polymer blend solar cells

Article abstract of DOI:10.1016/j.polymer.2014.04.045

We have fabricated ternary blend solar cells based on poly(3- hexylthiophene) (P3HT) as a donor material, poly{[N,N′-bis(2-octyldodecyl) -naphthalene-1,4,5,8-bis(dicarboximide)-2,6-diyl]-alt-5,5′-(2, 2′-bithiophene)} (N2200) as an acceptor material, and a silicon naphthalocyanine derivative (SiNc) as a near-IR sensitizer. In order to discuss how molecular structures impact the dye sensitization in P3HT/N2200 solar cells, we studied two SiNc molecules with different axial groups: SiNc10 with decyldimethylsilyl oxide and SiNc6 with trihexylsilyl oxide. As a result, P3HT/N2200/SiNc10 ternary solar cells exhibited a power conversion efficiency of 1.4%, which is improved by 73% compared to P3HT/N2200 binary solar cells and also by 17% compared to P3HT/N2200/SiNc6 ternary solar cells. We discuss the origin of the improvement in the device performance in terms of dye location in ternary blend films.

Full text of DOI:10.1016/j.polymer.2014.04.045

Polymer xxx (2014) 1e5
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Near-IR dye sensitization of polymer blend solar cells
,
b
,
a
a
a
Huajun Xu a, Hideo Ohkita a , Toshiaki Hirata , Hiroaki Benten , Shinzaburo Ito
*
a Department of Polymer Chemistry, Graduate School of Engineering, Kyoto University, Katsura, Nishikyo, Kyoto 615-8510, Japan
b Japan Science and Technology Agency (JST), PRESTO, 4-1-8 Honcho Kawaguchi, Saitama 332-0012, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
We have fabricated ternary blend solar cells based on poly(3-hexylthiophene) (P3HT) as a donor ma-
terial,
poly{[N,N0-bis(2-octyldodecyl)-naphthalene-1,4,5,8-bis(dicarboximide)-2,6-diyl]-alt-5,50-(2,20-
Received 19 February 2014
Received in revised form
22 April 2014
Accepted 25 April 2014
Available online xxx
bithiophene)} (N2200) as an acceptor material, and a silicon naphthalocyanine derivative (SiNc) as a
near-IR sensitizer. In order to discuss how molecular structures impact the dye sensitization in P3HT/
N2200 solar cells, we studied two SiNc molecules with different axial groups: SiNc10 with decyldime-
thylsilyl oxide and SiNc6 with trihexylsilyl oxide. As a result, P3HT/N2200/SiNc10 ternary solar cells
exhibited a power conversion efficiency of 1.4%, which is improved by 73% compared to P3HT/N2200
binary solar cells and also by 17% compared to P3HT/N2200/SiNc6 ternary solar cells. We discuss the
origin of the improvement in the device performance in terms of dye location in ternary blend films.
Ó 2014 Elsevier Ltd. All rights reserved.
Keywords:
Near-IR dye sensitization
Polymer blend
Solar cell
1. Introduction
intense absorption bands than that of fullerene derivatives. Thus, it
would be possible to collect the solar light over a wide wavelength
Polymer solar cells have made remarkable progress over the last
decade. Very recently, power conversion efficiency (PCE)
range by using donor and acceptor conjugated polymers that have
complementary absorption bands in the visible and near-IR region.
On the other hand, ternary blend solar cells have been proposed as
another approach to extending the absorption range from visible to
near-IR region in a simple way [9e22]. Previously, we demon-
strated that ternary blend solar cells incorporating dye molecules
can improve not only the light-harvesting efficiency by additional
dye absorption but also the exciton-harvesting efficiency by long-
range Förster energy transfer. The key to success is selective dye
loading into the donor/acceptor interface where excitons can be
dissociated into electronehole pairs [13]. Both approaches are
simple and versatile, and therefore can be easily applied to various
material combinations.
a
exceeding 10% has been reported by several groups [1,2]. This is
mainly because various low-bandgap polymers have been devel-
oped to harvest the solar light in the near-IR region where a photon
flux reach a maximum [3e8]. This approach, however, has a trade-
off issue that low-bandgap polymers have an absorption band at a
longer wavelength but an absorption window at a shorter wave-
length. In other words, it is generally difficult to harvest the solar
light over the wide wavelength range from visible to near-IR region
only by two donor/acceptor materials of conjugated polymer and
fullerene. This is partly due to the limited absorption bandwidth of
most conjugated polymers and partly due to small absorption of
fullerene derivatives. Thus, it is essentially required for further
improvement toward 15% toharvest much more photons in the solar
light over the wide wavelength range from visible to near-IR region.
In order to extend the light-harvesting range from visible to
near-IR region, several alternative approaches have been proposed
in recent years. Polymer/polymer blend solar cells have potential
advantage over the conventional polymer/fullerene solar cells,
because both donor/acceptor conjugated polymers have more
In this study, we have integrated these two approaches of
polymer/polymer blend solar cells and ternary blend solar cells to
extend the light-harvesting range in a simple way. In other words,
we addressed dye sensitization of polymer/polymer solar cells.
Here, we employed poly(3-hexylthiophene) (P3HT) as a donor
material,
poly{[N,N0-bis(2-octyldodecyl)-naphthalene-1,4,5,8-
bis(dicarboximide)-2,6-diyl]-alt-5,50-(2,20-bithiophene)} (N2200)
as an acceptor material, and two different silicon naphthalocyanine
derivatives as a near-IR sensitizer. In order to discuss how molec-
ular structures impact the dye sensitization in polymer/polymer
blend solar cells, we studied two silicon naphthalocyanine de-
rivatives with different axial groups: silicon naphthalocyanine
bis(tri-n-hexylsilyl oxide) (SiNc6) and silicon naphthalocyanine
bis(n-decyldimethylsilyl oxide) (SiNc10).
* Corresponding author. Department of Polymer Chemistry, Graduate School of
Engineering, Kyoto University, Katsura, Nishikyo, Kyoto 615-8510, Japan. Tel.: þ81
75 383 2613; fax: þ81 75 383 2617.
E-mail address: ohkita@photo.polym.kyoto-u.ac.jp (H. Ohkita).
http://dx.doi.org/10.1016/j.polymer.2014.04.045
0032-3861/Ó 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Xu H, et al., Near-IR dye sensitization of polymer blend solar cells, Polymer (2014), http://dx.doi.org/10.1016/
j.polymer.2014.04.045

2
H. Xu et al. / Polymer xxx (2014) 1e5
2. Experimental
Current densityevoltage (JeV) characteristics were measured
with a DC voltage and current source/monitor (Advantest R6243) in
the dark and under AM1.5G simulated solar illumination at
100 mW cmꢀ2. The light intensity was corrected with a calibrated
silicon photodiode reference cell (Bunkoh-Keiki BS-520). The
external quantum efficiency (EQE) spectra were measured with a
digital electrometer (Advantest R8252) under monochromatic light
illumination from a 500 W xenon lamp (Thermo Oriel Model
66921) with optical cut filters and a monochromator (Thermo Oriel
Cornerstone). The illumination was carried out from the ITO side in
an N2 atmosphere at room temperature. At least 10 devices were
measured to ensure the reproducibility of the device performance.
2.1. Materials
Silicon naphthalocyanine bis(tri-n-hexylsilyl oxide) (SiNc6;
Aldrich) was employed without further purification. Silicon naph-
thalocyanine bis(n-decyldimethylsilyl oxide) (SiNc10) was synthe-
sized as follows. A mixture of silicon naphthalocyanine dihydroxide
SiNc(OH)2 (100 mg), n-decyldimethyl chlorosilane (142 mL), and dry
pyridine (10 mL) was refluxed for 6 h. After the solution obtained
had been allowed to cool, the solvent was evaporated and chloro-
form was added to the residue. The solution was washed with
saturated NaCl solution, and then dried over MgSO4. After the
evaporation of the solvent, the residue was purified by silica gel
column chromatography (toluene/hexane ¼ 1/1 (v/v) as eluent) to
3. Results and discussion
afford SiNc[OSi(CH3)2(n-C10H21)]2 (78 mg) as
(yield ¼ 51%).
a
green solid
The HOMO level was evaluated by the photoelectron yield
spectroscopy measurements of P3HT, N2200, SiNc6, and SiNc10
neat films. From the threshold energy in the cubic root of the
photoelectron yield plotted against the incident photon energy (see
the Supplementary Material), the ionization energy was estimated
to be 4.7 eV (P3HT), 5.9 eV (N2200), 5.5 eV (SiNc6), and 5.5 eV
(SiNc10). The LUMO level was evaluated from the optical bandgap
and the HOMO level. From the absorption spectrum and the pho-
toluminescence (PL) spectra of P3HT, SiNc6, SiNc10, and N2200
neat films (see the Supplementary Material), the optical bandgap
was estimated to be 2 eV (P3HT), 1.6 eV (N2200), 1.5 eV (SiNc6), and
1.5 eV (SiNc10). Therefore, the LUMO level was evaluated to be
2.7 eV (P3HT), 4.3 eV (N2200), 4.0 eV (SiNc6), and 4.0 eV (SiNc10).
As shown in Fig. 1, ternary blends of P3HT, N2200, and SiNc dyes
exhibit cascaded energy structures both in the HOMO and LUMO
levels when dye molecules are located at the donor/acceptor
interface of P3HT and N2200. The offset energy is more than 0.3 eV
both in HOMOeHOMO and LUMOeLUMO gap, which would be
large enough for efficient charge generation at the interface. This
energetic condition is an absolute requirement for efficient dye
sensitization in a longer wavelength region. Indeed, we reported
dye sensitization in P3HT/N2200/SiNc6 ternary solar cells previ-
ously [22]. As shown in the figure, SiNc10 has the same optoelec-
tronic properties as SiNc6 and therefore equivalent potential for
dye sensitization of P3HT/N2200 blend solar cells if blend
morphology is the same as well.
ꢀ
ꢁ
UVꢀvisibleðtolueneÞ:lmax ¼ 773nm ε ¼ 5:8ꢁ105 Mꢀ1 cmꢀ1
:
1H NMR (400 MHz, CDC13):
d
¼ 10.06 (s, 5,36-Nc, 8H), 8.59 (m,
1,4-Nc, 8H), 7.85 (m, 1,4-Nc, 8H),1.20 (m, 9-CH2, 4H),1.05 (m, 8-CH2,
4H), 0.95 (m, 7-CH2, 4H), 0.85 (t, 10-CH3, 6H), 0.65 (m, 6-CH2, 4H),
0.50 (m, 5-CH2, 4H), 0.15 (m, 4-CH2, 4H), ꢀ0.15 (m, 3-CH2,
4H), ꢀ1.10 (m, 2-CH2, 4H), ꢀ1.95 (m, 1-CH2, 4H), ꢀ2.60 (s, SiCH3,
12H).
2.2. Sample fabrication
The quartz, glass, or ITO-coated substrates were cleaned by
ultrasonication in toluene, acetone, and ethanol each for 15 min,
dried with N2, and cleaned with a UVeO3 cleaner for 30 min. For
photoluminescence (PL) quenching measurements, sample films
were spincoated on the cleaned quartz substrate. For photovoltaic
measurements,
poly
(3,4-ethylenedioxythiophene):poly(4-
styrenesulfonate) (PEDOT:PSS; H.C. Starck PH500) was spincoated
onto the cleaned ITO substrate at 3000 rpm and baked at 140 ꢂC for
10 min in air. Subsequently, a ternary blend film was spincoated
from a p-xylene solution of poly(3-hexylthiophene) (P3HT; Plex-
tronics Plexcore OS2100), poly{[N,N0-bis(2-octyldodecyl)-naph-
thalene-1,4,5,8-bis(dicarboximide)-2,6-diyl]-alt-5,50-(2,20-bithio-
phene)} (N2200; Polyera), and SiNc dyes (SiNc6 or SiNc10) onto the
PEDOT:PSS-coated ITO substrate. The thickness of each layer was
w40 nm (PEDOT:PSS) and w110 nm (the ternary blend active
layer). The ternary blend film was annealed at 140 ꢂC for 10 min in
an N2-filled glove box. Finally, a metal electrode of Ca/Al layer (15/
80 nm) was deposited on top of the active layer in sequence at
2.5 ꢁ 10ꢀ4 Pa. The effective device area was 0.07 cm2.
Fig. 2a and b show JeV characteristics of ternary blend solar cells
incorporating SiNc6 or SiNc10 with various dye loading
2.3. Measurements
The ionization potential of P3HT, N2200, SiNc6, and SiNc10 was
measured with a photoelectron yield spectrometer (Riken Keiki,
AC-3). All the neat films (ca. 60 nm) were fabricated by spin-coating
from a chlorobenzene solution on PEDOT:PSS-coated ITO sub-
strates. The threshold energy for the photoelectron emission was
estimated on the basis of the cubic root of the photoelectron yield
plotted against the incident photon energy (see the Supplementary
Material) as reported previously [23,24].
Absorption and PL spectra of the neat and blend films were
measured at room temperature with a spectrophotometer (Hitachi
UV-3500) and a fluorescence spectrophotometer (Hitachi F-4500),
respectively.
Fig. 1. Chemical structures and energy diagram of materials employed in this study:
P3HT, SiNc6, SiNc10, and N2200 from left to right. The figures represent the HOMO
(lower) and LUMO (upper) energy in electron volts.
Please cite this article in press as: Xu H, et al., Near-IR dye sensitization of polymer blend solar cells, Polymer (2014), http://dx.doi.org/10.1016/
j.polymer.2014.04.045

H. Xu et al. / Polymer xxx (2014) 1e5
3
Table 1
concentrations. As shown in the figure, P3HT/N2200/SiNc6 solar
The device parameters of P3HT/N2200/SiNc ternary blend solar cells.
cells exhibit modest improvements in the photocurrent at limited
dye concentrations while P3HT/N2200/SiNc10 solar cells exhibit
efficient improvements over the all dye loading concentrations up
to 50 wt%. The device parameters are summarized in Table 1. This
difference is, as shown in Fig. 2c and d, mainly ascribed to the
difference in the photocurrent generation efficiency at the dye
absorption band. However, as shown in Fig. 2e and f, there is no
substantial difference in the photon absorption efficiency between
them. This suggests that SiNc6 and SiNc10 are differently distrib-
uted in P3HT/N2200/dye ternary blend films.
In order to address the origin of the different photocurrent
generation in details, we evaluated the contribution of P3HT, N2200,
and SiNc dyes to the photocurrent generation from the integration
of the product of the EQE and the solar spectrum from 360 to 650 nm
for P3HT, from 650 to 750 nm for N2200, and from 750 to 850 nm for
SiNc dye. For SiNc6, as shown in Fig. 3, the integrated photocurrent
due to P3HT and dye bands showed a maximum at 15 wt% and then
decreased at >15 wt% while the integrated photocurrent due to
N2200 remained constant at w1 mA cmꢀ2. For SiNc10, on the other
hand, all the integrated photocurrent monotonically increased up to
50 wt%. In summary, SiNc6 does not contribute to the photocurrent
generation at high dye loading concentrations >15 wt% while
SiNc10 does contribute even at high dye loading concentrations up
to 50 wt%. These findings suggest that the dye distribution in ternary
blends should be different.
Dye
wt%
JSC/mA cmꢀ2
VOC/V
FF
PCEa/%
SiNc10
5
10
15
20
30
50
5
10
15
20
30
50
0
4.6
4.9
5.0
5.4
5.7
6.3
3.9
3.7
4.5
3.6
3.4
3.1
3.5
0.50
0.50
0.50
0.50
0.50
0.51
0.48
0.48
0.50
0.49
0.49
0.48
0.45
0.52
0.49
0.50
0.48
0.50
0.39
0.50
0.52
0.54
0.53
0.50
0.58
0.52
1.2 ꢃ 0.05
1.2 ꢃ 0.09
1.2 ꢃ 0.1
1.3 ꢃ 0.07
1.4 ꢃ 0.09
1.3 ꢃ 0.04
0.92 ꢃ 0.1
0.92 ꢃ 0.06
1.2 ꢃ 0.09
0.94 ꢃ 0.08
0.84 ꢃ 0.08
0.86 ꢃ 0.03
0.82 ꢃ 0.07
SiNc6
No
a
The PCE values are listed with their associated standard deviations.
which is dependent upon the local concentration of dye molecules
as reported previously [12]. For comparison, we also measured that
in binary blends of P3HT/SiNc and N2200/SiNc (see the
Supplementary Material). For SiNc6, as shown in Fig. 4a, the peak
wavelength showed a similar trend for all the ternary and binary
blend films. It was 772 nm at 5 wt%, which is the same as that in
solution, suggesting that SiNc6 molecules are isolated in polymer
matrices at a low dye loading. Thereafter, it steeply jumped to
795 nm at w20 wt% and then approached to 800 nm, which is the
same as that of SiNc6 neat films. This steep increase suggests that
To discuss the dye location in P3HT/N2200/SiNc, we measured
the absorption peak wavelength of dyes in the ternary blend film,
Fig. 2. a), b) JeV characteristics, c), d) EQE spectra, and e), f) photon absorption effi-
ciency of P3HT/N2200/dye blend solar cells: a), c), e) dye ¼ SiNc6 and b), d), f)
dye ¼ SiNc10 at 0 (gray lines), 5 (broken lines), 15 (dashed dotted lines), 30 (dashed
double-dotted lines), and 50 wt% (solid lines). The photon absorption efficiency is
estimated from twice the absorbance of the films assuming the reflection at the metal
electrode.
Fig. 3. Integrated photocurrent of P3HT/N2200/dye ternary solar cells with different
dye loading concentrations, which is calculated from the EQE and the solar spectrum
integrated a) from 360 to 650 nm (photocurrent from P3HT), b) from 650 to 750 nm
(photocurrent from N2200), c) from 750 to 850 nm (photocurrent from dye), and d)
from 360 to 850 nm (total photocurrent z JSC): SiNc10 (circles) and SiNc6 (triangles).
Please cite this article in press as: Xu H, et al., Near-IR dye sensitization of polymer blend solar cells, Polymer (2014), http://dx.doi.org/10.1016/
j.polymer.2014.04.045

4
H. Xu et al. / Polymer xxx (2014) 1e5
HOMO and LUMO levels. For comparison, two SiNc dye molecules
were employed: one is silicon naphthalocyanine bis(tri-n-hexylsilyl
oxide) (SiNc6) and the other is silicon naphthalocyanine bis(n-
decyldimethylsilyl oxide) (SiNc10). For SiNc6, the device perfor-
mance showed a maximum efficiency of 1.2% at 15 wt% and then
rapidly degraded at higher dye loading. For SiNc10, the power
conversion efficiency increased with dye loading concentrations
and reached the maximum (1.4%) at 30 wt%. This difference is
mainly due to the different dye distribution in ternary blend films.
On the basis of the absorption spectral analysis, we found that both
dyes are likely to be segregated from P3HT and N2200 and hence
located at the interface. In terms of the surface energy of each
material, SiNc10 is considered to be more mixed with P3HT than
SiNc6. In other words, SiNc6 would form large aggregates at the
interface and hence hinder charge transport, resulting in degraded
device performance at higher dye loading concentrations. On the
other hand, SiNc10 would be well mixed with P3HT at the interface
even at high dye loading concentrations and remain efficient
charge generation and charge transport, resulting in the improved
device performance at 30 wt%. In conclusion, dye distribution in
ternary blends, which have critical impact on the device perfor-
mance, can be controlled by a careful molecular design.
Fig. 4. Peak wavelength of the dye absorption band plotted against dye loading con-
centrations in blend films: P3HT/dye (triangles), N2200/dye (inverted triangles), and
P3HT/N2200/dye (circles). a) dye ¼ SiNc6 and b) dye ¼ SiNc10.
Table 2
The surface energy of each material employed in this study.
Materials
Surface energy/mJ mꢀ2
Acknowledgments
P3HT
18.4
21.9
23.0
21.3
N2200
SiNc6
SiNc10
This work was partly supported by the FIRST program (Devel-
opment of Organic Photovoltaics toward a Low-Carbon Society:
Pioneering Next Generation Solar Cell Technologies and Industries
via Multi-manufacturer Cooperation) and the JST PRESTO program
(Photoenergy Conversion Systems and Materials for the Next
Generation Solar Cells).
SiNc6 molecules are likely to be segregated from either P3HT or
N2200 domains. For SiNc10, on the other hand, the peak wave-
length showed a similar trend for P3HT/SiNc10 and P3HT/N2200/
SiNc10 but different for N2200/SiNc10, as shown in Fig. 4b. This
suggests that SiNc10 molecules are primarily located at P3HT do-
mains in P3HT/N2200/SiNc10 ternary blend films. Similarly to
SiNc6, the peak wavelength was 775 nm at the lowest concentra-
tion and finally approached to 818 nm with increasing dye loading
concentrations. In contrast to SiNc6, the peak wavelength was
gradually shifted from 775 to 818 nm, suggesting that SiNc10 is not
likely to be segregated but rather well mixed with P3HT. This dif-
ference is consistent with the efficient device performance of P3HT/
N2200/SiNc10 with high dye loading concentrations.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.polymer.2014.04.045.
References
[1] Green MA, Emery K, Hishikawa Y, Warta W, Dunlop ED. Prog Photovolt Res
Appl 2014;22:1e9.
[2] You J, Dou L, Yoshimura K, Kato T, Ohya K, Moriarty T, et al. Nat Commun
2013;4:1446.
To further understand the dye distribution in terms of materials
properties, we measured the surface energy of each material. As
summarized in Table 2, the order of the surface energy is as follows:
P3HT < SiNc10 < N2200 < SiNc6. Considering the surface energy
alone, SiNc10 and SiNc6 would be predicted to be located at the
P3HT/N2200 interface and at N2200 domains, respectively (details
are described in the Supplementary Material). However, this is not
the case as mentioned above: both dyes are not likely to be well
mixed with N2200. This is probably because N2200 is likely to form
aggregates in solid films and hence expels dye molecules from the
aggregates [25e27]. The same is true for highly crystalline P3HT
domains. These indicate that the majority of SiNc dye molecules are
primarily located at the P3HT/N2200 interface in ternary blend
films. In terms of the surface energy, SiNc10 is considered to be
more mixed with P3HT than SiNc6, which is consistent with the
absorption change we observed. We therefore conclude that such
mixed distribution at the interface is beneficial for the dye sensi-
tization rather than formation of large dye aggregates.
[3] Bundgaard E, Krebs FC. Sol Energy Mater Sol Cells 2007;91:954e85.
[4] Liang Y, Yu L. Acc Chem Res 2010;43:1227e36.
[5] Beaujuge PM, Fréchet JMJ. J Am Chem Soc 2011;133:20009e29.
[6] Li G, Zhu R, Yang Y. Nat Phot 2012;6:153e61.
[7] Zhou W, Zhang ZG, Ma L, Li Y, Zhan X. Sol Energy Mater Sol Cells 2013;112:
13e9.
[8] Li YF. Acc Chem Res 2012;45:723e33.
[9] Peet J, Tamayo AB, Dang XD, Seo JH, Nguyen TQ. Appl Phys Lett 2008;93:
163306.
[10] Honda S, Nogami T, Ohkita H, Benten H, Ito S. ACS Appl Mater Interfaces
2009;1:804e10.
[11] Honda S, Ohkita H, Benten H, Ito S. Chem Commun 2010;46:6596e8.
[12] Honda S, Ohkita H, Benten H, Ito S. Adv Energy Mater 2011;1:588e98.
[13] Honda S, Yokoya S, Ohkita H, Benten H, Ito S. J Phys Chem C 2011;115:
11306e17.
[14] Kubo Y, Watanabe K, Nishiyabu R, Hata R, Murakami A, Shoda T, et al. Org Lett
2011;13:4574e7.
[15] Ameri T, Min J, Li N, Machui F, Baran D, Forster M, et al. Adv Energy Mater
2012;2:1198e202.
[16] Yamamoto S, Kimura M. ACS Appl Mater Interface 2013;5:4367e73.
[17] Xu H, Wada T, Ohkita H, Benten H, Ito S. Electrochim Acta 2013;100:214e9.
[18] Ameri T, Khoram P, Min J, Brabec CJ. Adv Mater 2013;25:4245e67.
[19] Chen YC, Hsu CY, Lin RYY, Ho KC, Lin JT. ChemSusChem 2013;6:20e35.
[20] Liu F, Gu Y, Shen XB, Ferdous S, Wang HW, Russell TP. Prog Polym Sci
2013;38:1990e2052.
[21] Huang JS, Goh T, Li X, Sfeir MY, Bielinski EA, Tomasulo S, et al. Nat Phot
2013;7:479e85.
[22] Ito S, Hirata T, Mori D, Benten H, Lee LT, Ohkita H. J Photopolym Sci Technol
2013;26:175e80.
4. Conclusion
We fabricated ternary blend solar cells based on P3HT, N2200,
and SiNc dye, which have cascaded energy structures both in
Please cite this article in press as: Xu H, et al., Near-IR dye sensitization of polymer blend solar cells, Polymer (2014), http://dx.doi.org/10.1016/
j.polymer.2014.04.045

H. Xu et al. / Polymer xxx (2014) 1e5
5
[23] Kochi M, Harada Y, Hirooka T, Inokuchi H. Bull Chem Soc Jpn 1970;43:
2690e702.
[26] Yan H, Collins BA, Gann E, Wang C, Ade H, McNeill CR. ACS Nano 2012;6:
677e88.
[24] Seki K, Kanai K. Mol Cryst Liq Cryst 2006;455:145e81.
[25] Schubert M, Dolfen D, Frisch J, Roland S, Steyrleuthner R, Stiller B, et al. Adv
Energy Mater 2012;2:369e80.
[27] Steyrleuthner R, Schubert M, Howard I, Klaumünzer B, Schilling K, Chen Z,
et al. J Am Chem Soc 2012;134:18303e17.
Please cite this article in press as: Xu H, et al., Near-IR dye sensitization of polymer blend solar cells, Polymer (2014), http://dx.doi.org/10.1016/
j.polymer.2014.04.045