Efficient dye-sensitized photovoltaic wires based on an organic redox electrolyte

Article abstract of DOI:10.1021/ja405012w

An organic thiolate/disulfide redox couple with low absorption in the visible region was developed for use in fabricating novel dye-sensitized photovoltaic wires with an aligned carbon nanotube (CNT) fiber as the counter electrode. These flexible wire devices achieved a maximal energy conversion efficiency of 7.33%, much higher than the value of 5.97% for the conventional I^-/I₃^- redox couple. In addition, the aligned CNT fiber also greatly outperforms the conventional Pt counter electrode with a maximal efficiency of 2.06% based on the thiolate/disulfide redox couple.

Full text of DOI:10.1021/ja405012w

Communication
pubs.acs.org/JACS
Efficient Dye-Sensitized Photovoltaic Wires Based on an Organic
Redox Electrolyte
Shaowu Pan,†,‡ Zhibin Yang, Houpu Li, Longbin Qiu, Hao Sun, and Huisheng Peng *
†
†
†
†
†,‡,
†State Key Laboratory of Molecular Engineering of Polymers, Department of Macromolecular Science, and Laboratory of Advanced
Materials, Fudan University, Shanghai 200438, China
‡
School of Materials Science and Engineering, Tongji University, 4800 Caoan Road, Shanghai 201804, China
*
S Supporting Information
fabrication of flexible dye-sensitized photovoltaic wires. An
aligned carbon nanotube (CNT) fiber with high flexibility,
tensile strength, and electrical conductivity serves as a counter
electrode, and a Ti wire with perpendicularly aligned titania
nanotubes functions as the working electrode. The two
electrodes are intertwined together to form the wire cell (Figure
ABSTRACT: An organic thiolate/disulfide redox couple
with low absorption in the visible region was developed for
use in fabricating novel dye-sensitized photovoltaic wires
with an aligned carbon nanotube (CNT) fiber as the
counter electrode. These flexible wire devices achieved a
maximal energy conversion efficiency of 7.33%, much
1
). The design and use of an aligned CNT fiber as the counter
−
−
higher than the value of 5.97% for the conventional I /I3
redox couple. In addition, the aligned CNT fiber also
greatly outperforms the conventional Pt counter electrode
with a maximal efficiency of 2.06% based on the thiolate/
disulfide redox couple.
ye-sensitized solar cells (DSCs) have been widely studied
as a promising new type of photovoltaic devices because of
D
their easy fabrication, low cost, and flexibility compared with the
1
conventional, rigid silicon wafer-based technology. To further
improve their energy conversion efficiencies, much effort has
2
−6
been devoted to optimization of components and creative
7
−9
novel structures of DSCs.
The electrolyte that functions to
transfer electrons from the counter electrode to the oxidized dye
−
−
is a key component in the DSC, and the I /I redox couple has
3
been generally used. Although this conventional redox couple
works efficiently, disadvantages such as corrosion of the current
collector (typically based on Ag or Cu) and partial absorption in
the visible region has largely limited the applicable production of
DSCs. Therefore, a lot of attention has been paid to discovering
alternative redox couples for the future development of
Figure 1. Schematic illustration of a photovoltaic wire with a CNT fiber
as the counter electrode, a Ti wire impregnated with titania nanotubes as
the working electrode, and the thiolate/disulfide redox couple as the
electrolyte.
1
0−13
DSCs.
Among them, organic redox couples show promising
advantages such as high efficiency and safety.
However, as another key component, the counter electrode
electrode effectively catalyzes the thiolate/disulfide redox couple
with an energy conversion efficiency of 7.33%, which is much
higher than the efficiency of 2.06% based on the Pt counter
electrode under the same conditions. The use of the organic
redox couple also proved to be critical in the development of the
(typically composed of Pt) exhibits very low catalytic activity
toward organic redox couples such as the thiolate/disulfide redox
1
4,15
couple.
The low fill factor is produced from a large reduction
resistance of the disulfide on the standard platinized counter
electrode, leading to a low energy conversion efficiency. This
phenomenon indicates that Pt may be not suitable as a counter
electrode to catalyze the organic redox couple. Therefore, highly
catalytic counter electrode materials are urgently required. Some
−
−
dye-sensitized photovoltaic wires. Corrosion due to the I /I
3
redox couple decreases the energy conversion efficiency and
lifetime in the photovoltaic wire more than in a conventional
planar device, and also requires more complex fabrication. In
addition, the novel dye-sensitized photovoltaic wire constructed
from the organic redox couple shows much higher efficiency than
16−19
attempts have been made to explore carbon materials
and
20,21
organic polymers as counter electrodes.
The DSCs appeared
−
−
the I /I3 redox couple under the same conditions with the
in a rigid, planar format, although organic redox couples were
proposed to be particularly promising for making flexible DSCs.
In this work, a new organic redox couple showing negligible
absorption in the visible region was developed for use in the
Received: May 18, 2013
©
XXXX American Chemical Society
A
dx.doi.org/10.1021/ja405012w | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
aligned CNT fiber as the counter electrode. This photovoltaic
wire is flexible, lightweight, and weaveable.
by electrochemical anodization and had a length of 30 μm and
26
diameters of 70−100 nm (Figure S5). The modified Ti wire
was immersed into a solution of N719 dye, and an aligned CNT
fiber was twisted around the dye-absorbed working electrode.
The photovoltaic wire could be sealed in a glass capillary tube or a
flexible fluorinated ethylenepropylene tube with subsequent
infiltration of the electrolyte by capillary force. Figure 2c shows a
typical SEM image of a photovoltaic wire with pitch distance of 1
mm. The two fiber electrodes are in close contact with each other
−
For this organic redox couple, the thiolate (T ) and disulfide
T ) components, whose structures are shown in Scheme 1, were
2
(
−
Scheme 1. Structures of the Thiolate (T ) and Disulfide (T )
2
Components of the Redox Couple
(
Figure 2d), and the twist structure was well-maintained during
bending, indicating the high stability and flexibility of the wire cell
(
Figure S6).
Figure 3 compares current density−voltage (J−V) curves of
dye-sensitized photovoltaic wires having aligned CNT fibers with
synthesized by neutralization of 5-mercapto-1-methyltetrazole
with NMe OH and oxidation of 5-mercapto-1-methyltetrazole,
4
respectively [for synthetic details, see the Experimental Section
and Figures S1 and S2 in the Supporting Information (SI)]. The
−
−
−
UV−vis spectra of the T /T and I /I redox couples (Figure
2
3
S3) show that the organic redox couple has less absorption in the
visible region, which is widely known to be important for highly
efficient DSCs.
To prepare the aligned CNT fiber, a CNT array was first
synthesized by chemical vapor deposition, and continuous CNT
sheets with a thickness of ∼20 nm were then pulled out of the
2
2
array. Many layers of CNT sheets were stacked and further
spun into a CNT fiber, typically using a microprobe rotation
speed of 2000 rpm. The diameters of the CNT fibers were
controlled from 25 to 100 μm by increasing the number of sheet
layers from 2 to 25. Figure S4 shows a micrograph of the spinning
process used to produce an aligned CNT fiber. Figure 2a shows a
Figure 3. J−V curves of photovoltaic wires using CNT fibers with
diameters of 25, 40, 60, and 100 μm or a Pt wire with diameter of 25 μm
as counter electrodes, measured under AM 1.5 illumination using the
−
same T /T redox couple.
2
−
increasing diameters as counter electrodes and the T /T redox
2
couple as the electrolyte measured under AM 1.5 illumination. In
this work, the surface of the top half of the photovoltaic wire was
appropriately irradiated. The photovoltaic parameters [open-
circuit voltage (V ), short-circuit current density (J ), fill factor
oc
sc
(
FF), and energy conversion efficiency (η)] are summarized at
Table 1. The V values were almost the same for the CNT fibers
oc
Table 1. Photovoltaic Parameters in Figure 3
counter electrode
Voc (mV)
Jsc (mA cm−2)
FF
η (%)
2
4
6
1
5 μm fiber
0 μm fiber
0 μm fiber
00 μm fiber
682
678
680
676
669
14.05
14.46
15.69
13.80
12.29
0.43
0.61
0.66
0.65
0.25
4.13
5.95
7.01
6.10
2.06
Figure 2. SEM images of (a, b) the CNT fiber and (c, d) part of the
twisted photovoltaic wire at (a, c) low and (b, d) high magnification.
Pt wire
scanning electron microscopy (SEM) image of an aligned CNT
fiber with a uniform diameter of 60 μm. A higher-magnification
image (Figure 2b) further indicates that the CNTs are highly
aligned along the twisting direction in the fiber. The resulting
CNT fibers are lightweight, flexible, strong (tensile strength of
with different diameters. However, as the diameter increased
from 25 to 60 μm, the average J and FF values continuously
sc
2
increased from 14.05 to 15.69 mA/cm and 0.43 to 0.66,
respectively. As a result, η improved from 4.13% to 7.01%. With a
further increase in the diameter to 100 μm, η decreased to 6.10%.
The optimal diameter for the CNT fiber was 60 μm as a result of
2
3
2
1
0 −10 MPa), and conductive (electrical conductivity of 10 −
3 −1
27
1
0 S cm ). They represent a new family of electrode materials
for various optoelectronic and electronic microdevices.
two opposite factors. On one hand, larger CNT fibers have
23−25
more catalytic sites. On the other hand, larger CNT fibers
become less flexible, and the twisted structures are less uniform.
In addition, a larger fiber shades more dyes from illumination. As
a comparison, a Pt wire with a diameter of 25 μm was used as the
Here their use as counter electrodes in DSCs was investigated.
A Ti wire with titania nanotubes perpendicularly aligned on
the surface was used as the working electrode to fabricate the dye-
sensitized photovoltaic wire. The titania nanotubes were grown
counter electrode under the same conditions. The V was similar
oc
B
dx.doi.org/10.1021/ja405012w | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
to those of the CNT fibers, but both J and FF sharply decreased,
energy conversion efficiencies than the Pt counter electrode in
sc
2
9
producing a lower η of 2.06%. Notably, the photovoltaic
parameters varied by less than 3% (e.g., η = 7.18 ± 0.17% at a
CNT fiber diameter of 60 μm).
planar DSCs. Unexpectedly, these aligned CNT fibers
−
exhibited even higher performances with the organic T /T
2
−
−
couple than the I /I3 couple. Dye-sensitized photovoltaic wires
were compared using the same aligned CNT fiber with diameter
of 60 μm as the counter electrode but different electrolytes
(Figure 5 and Table 2). For the same composition of the two
The catalytic activities of aligned CNT fibers and Pt wire with
−
the T /T2 redox couple were further studied by cyclic
voltammetry (Figure 4). A pair of redox peaks should be
−
−
−
Figure 5. J−V curves of photovoltaic wires using T /T or I /I as the
2
3
redox couple with the same 60 μm diameter CNT fiber as the counter
Figure 4. Cyclic voltammograms of CNT fibers with diameters of 25, 40,
−
electrode, measured under AM 1.5 illumination.
6
0, and 100 μm and a Pt wire with diameter of 25 μm using the T /T
2
−1
electrolyte at a scan rate of 50 mV s . The inset graph shows the peak-
to-peak voltage separation (V ) for different CNT fiber diameters.
Table 2. Photovoltaic Parameters in Figure 5
pp
electrolyte
Voc (mV)
Jsc (mA cm−2)
FF
η (%)
−
−
−
0.4 M T /0.4 M T
683
628
742
15.49
1.25
0.69
0.63
0.60
7.33
0.49
5.97
2
−
observed for the half-reaction 2T ⇄ T + 2e . In the case of the
2
−
0
0
.4 M I /0.4 M I
3
Pt wire, however, the redox peaks could not be well-identified
because of the poor catalytic activity. In contrast, the two redox
peaks of the CNT fibers were clearly observed. In cyclic
voltammetry, a higher reduction peak current (I ) and lower
peak-to-peak voltage separation (V ) indicate better catalytic
−
−
.6 M I /0.05 M I3
13.36
−
−
−
components (0.4 M for each) in the I /I and T /T couples,
the wire cell based on I /I showed much lower η of 0.49%
compared with 7.33% for T /T . The lower η was mainly derived
from the lower J , which may be explained by the fact that the I /
I3 electrolyte absorbs more visible light than the T /T
electrolyte (Figure S3). To decrease the absorbing effect, the
composition of the I /I3 electrolyte was systematically
optimized to finally increase the efficiency to 5.97%, which is
p
3
2
−
−
pp
3
28
−
activity for the electrodes. Obviously, I rises with increasing
p
2
−
diameter of the CNT fiber from 25 to 100 μm becausde of the
increasing surface area. Vpp for the two redox peaks decreased
with increasing diameter from 25 to 100 μm. In addition, no
obvious decreases in current density or shifts in peak location
were found for any of the CNT fibers after 10 cycles (Figures
S7−S10).
To further understand the different catalytic activities of
aligned CNT fibers with different diameters, electrochemical
impedance spectroscopy (EIS) was used to study the dye-
sensitized photovoltaic wire. As shown in Figure S11, two
semicircles were generally observed in all of the plots. The left
one corresponds to the interface charge-transfer resistance
between the counter electrode and the electrolyte, while the right
one represents the sum of the electrolyte resistance and the
interface charge-transfer resistance between the working
electrode and the electrolyte. The size of the first semicircle
decreased with increasing diameter of the CNT fiber from 25 to
sc
−
−
2
−
−
−
still significantly lower than the value of 7.33% for the T /T
2
electrolyte. The above conclusion was further verified by EIS
(Figures S13 and S14). The charge-transfer resistance at the
counter electrode−electrolyte interface was much lower for the
−
−
−
T /T electrolyte than for the I /I electrolyte.
2
3
These photovoltaic wires exhibit high flexibility. They can be
easily bent (Figure 6a,b) or even made into a knot (Figure S15)
without breaking. Figure 6c compares the J−V curves for a
photovoltaic wire made from a 60 μm diameter CNT fiber before
and after bending, which overlap well. The high flexibility of the
photovoltaic wire was further quantitatively studied by tracing
the photovoltaic parameters (Figure 6d,e). The η values
decreased slightly over the first 20 cycles and then were well-
maintained in the following 80 cycles.
In summary, an organic thiolate/disulfide redox couple with
low absorption in the visible region was developed for use in
fabricating novel dye-sensitized photovoltaic wires. These
flexible wire devices achieved high energy conversion efficiencies
with a maximal value of 7.33%. This work also opens a new
avenue in the development of highly efficient optoelectronic and
electronic devices by designing matchable materials in different
parts and making effective structures.
1
00 μm, indicating a decrease of the charge-transfer resistance. In
other words, the CNT fibers with diameters of 60 and 100 μm
result in lower resistances with higher catalytic activities. This
conclusion agrees with the cyclic voltammetry results. As
expected, the wire cell based on the 25 μm diameter Pt wire as
the counter electrode showed a much higher resistance (Figure
S12).
CNT materials were previously found to work efficiently as
−
−
counter electrodes for the I /I redox couple. For instance,
3
aligned CNT sheets or composite films produced even higher
C
dx.doi.org/10.1021/ja405012w | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
(12) Daeneke, T.; Kwon, T. H.; Holmes, A. B.; Duffy, N. W.; Bach, U.;
Spiccia, L. Nat. Chem. 2011, 3, 211.
13) Yanagida, S.; Yu, Y. H.; Manseki, K. Acc. Chem. Res. 2009, 42,
827.
14) Li, D. M.; Li, H.; Luo, Y. H.; Li, K. X.; Meng, Q. B.; Armand, M.;
Chen, L. Q. Adv. Funct. Mater. 2010, 20, 3358.
15) Tian, H. N.; Jiang, X. A.; Yu, Z.; Kloo, L.; Hagfeldt, A.; Sun, L. C.
Angew. Chem., Int. Ed. 2010, 49, 7328.
16) Wu, M. X.; Lin, X.; Wang, Y. D.; Wang, L.; Guo, W.; Qu, D. D.;
Peng, X. J.; Hagfeldt, A.; Gratzel, M.; Ma, T. L. J. Am. Chem. Soc. 2012,
34, 3419.
17) Liu, G. H.; Li, X.; Wang, H.; Rong, Y. G.; Ku, Z. L.; Xu, M.; Liu, L.
F.; Hu, M.; Yang, Y.; Han, H. W. Carbon 2013, 53, 11.
18) Hao, F.; Wang, Z.; Luo, Q.; Lou, J.; Li, J. B.; Wang, J. P.; Fan, S. S.;
Jiang, K. L.; Lin, H. J. Mater. Chem. 2012, 22, 22756.
19) Hao, F.; Dong, P.; Zhang, J.; Zhang, Y. C.; Loya, P. E.; Hauge, R.
H.; Li, J. B.; Lou, J.; Lin, H. Sci. Rep. 2012, 2, 368.
20) Tian, H. N.; Yu, Z.; Hagfeldt, A.; Kloo, L.; Sun, L. J. Am. Chem. Soc.
011, 133, 9413.
21) Burschka, J.; Brault, V.; Ahmad, S.; Breau, L.; Nazeeruddin, M. K.;
Marsan, B.; Zakeeruddin, S. M.; Gratzel, M. Energy Environ. Sci. 2012, 5,
089.
22) Chen, T.; Wang, S. T.; Yang, Z. B.; Feng, Q. Y.; Sun, X. M.; Li, L.;
Wang, Z. S.; Peng, H. S. Angew. Chem., Int. Ed. 2011, 50, 1815.
23) Baughman, R. H.; Zakhidov, A. A.; de Heer, W. A. Science 2002,
297, 787.
24) Avouris, P.; Chen, Z. H.; Perebeinos, V. Nat. Nanotechnol. 2007, 2,
05.
25) Avouris, P.; Freitag, M.; Perebeinos, V. Nat. Photonics 2008, 2,
41.
26) Chen, T.; Qiu, L. B.; Kia, H. G.; Yang, Z. B.; Peng, H. S. Adv.
(
1
(
(
(
̈
1
(
(
(
(
2
(
Figure 6. (a, b) Photographs of a photovoltaic wire (a) before and (b)
after bending. (c) J−V curves of the photovoltaic wires before and after
bending. (d, e) Photovoltaic parameters before and after different
numbers of bending cycles.
̈
6
(
(
ASSOCIATED CONTENT
■
*
S
Supporting Information
(
6
(
3
Synthesis of the organic electrolyte and structure and property
characterizations of devices. This material is available free of
charge via the Internet at http://pubs.acs.org.
(
Mater. 2012, 24, 4623.
AUTHOR INFORMATION
Corresponding Author
penghs@fudan.edu.cn
Notes
■
(27) Zhang, S.; Ji, C. Y.; Bian, Z. Q.; Yu, P. R.; Zhang, L. H.; Liu, D. Y.;
Shi, E. Z.; Shang, Y. Y.; Peng, H. T.; Cheng, Q.; Wang, D.; Huang, C. H.;
Cao, A. Y. ACS Nano 2012, 6, 7191.
(
28) Gong, F.; Wang, H.; Xu, X.; Zhou, G.; Wang, Z. S. J. Am. Chem.
Soc. 2012, 134, 10953.
29) Huang, S.; Yang, Z.; Zhang, L.; He, R.; Chen, T.; Cai, Z.; Luo, Y.;
The authors declare no competing financial interest.
(
Lin, H.; Cao, H.; Zhu, X.; Peng, H. J. Mater. Chem. 2012, 22, 16833.
ACKNOWLEDGMENTS
This work was supported by NSFC (91027025, 21225417),
MOST (2011CB932503, 2011DFA51330), STCSM
■
(
11520701400, 12nm0503200), the Fok Ying Tong Education
Foundation, and The Program for Professor of Special
Appointment at Shanghai Institutions of Higher Learning.
REFERENCES
1) O’Regan, B.; Grat
2) Zhu, K.; Neale, N. R.; Miedaner, A.; Frank, A. J. Nano Lett. 2007, 7,
■
(
(
̈
zel, M. Nature 1991, 353, 737.
6
(
2
9.
3) Varghese, O. K.; Paulose, M.; Grimes, C. A. Nat. Nanotechnol.
009, 4, 592.
(
4) Hamann, T. W.; Ondersma, J. W. Energy Environ. Sci. 2011, 4, 370.
(
5) Zhang, Q. F.; Dandeneau, C. S.; Zhou, X. Y.; Cao, G. Z. Adv. Mater.
2
(
009, 21, 4087.
6) Jennings, J. R.; Ghicov, A.; Peter, L. M.; Schmuki, P.; Walker, A. B. J.
Am. Chem. Soc. 2008, 130, 13364.
7) Martinson, A. B. F.; Hamann, T. W.; Pellin, M. J.; Hupp, J. T.
Chem.Eur. J. 2008, 14, 4458.
8) Chen, T.; Qiu, L.; Yang, Z.; Peng, H. Chem. Soc. Rev. 2013, 42,
031.
9) Weintraub, B.; Wei, Y. G.; Wang, Z. L. Angew. Chem., Int. Ed. 2009,
8, 8981.
10) Wang, M. K.; Chamberland, N.; Breau, L.; Moser, J. E.; Humphry-
Baker, R.; Marsan, B.; Zakeeruddin, S. M.; Gratzel, M. Nat. Chem. 2010,
, 385.
11) Yella, A.; Lee, H. W.; Tsao, H. N.; Yi, C. Y.; Chandiran, A. K.;
Nazeeruddin, M. K.; Diau, E. W. G.; Yeh, C. Y.; Zakeeruddin, S. M.;
Gratzel, M. Science 2011, 334, 629.
(
(
5
(
4
(
̈
2
(
̈
D
dx.doi.org/10.1021/ja405012w | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX