Polymer solar cells with enhanced open-circuit
voltage and efficiency
Hsiang-Yu Chen
1,2
, Jianhui Hou
1
*
, Shaoqing Zhang
1
,YongyeLiang
3
, Guanwen Yang
2
, Yang Yang
2
,
Luping Yu
3
,YueWu
1
*
and Gang Li
1
Following the development of the bulk heterojunction
1
structure,
recent years have seen a dramatic improvement in the efficiency
of polymer solar cells. Maximizing the open-circuit voltage in a
low-bandgap polymer is one of the critical factors towards
enabling high-efficiency solar cells. Study of the relation
between open-circuit voltage and the energy levels of the
donor/acceptor
2
in bulk heterojunction polymer solar cells has
stimulated interest in modifying the open-circuit voltage by
tuning the energy levels of polymers
3
. Here, we show that the
open-circuit voltage of polymer solar cells constructed based on
the structure of a low-bandgap polymer, PBDTTT
4
, can be tuned,
step by step, using different functional groups, to achieve values
as high as 0.76 V. This increased open-circuit voltage combined
with a high short-circuit current density results in a polymer
solar cell with a power conversion efficiency as high as 6.77%,
as certified by the National Renewable Energy Laboratory.
Polymer solar cells (PSCs) have attracted much attention due to
their potential in low-cost solar energy harvesting, as well as appli-
cations in flexible, light-weight, colourful and large-area devices.
With the discovery of efficient photo-induced electron transfer
from a conjugated polymer to fullerene
1
, the bulk heterojunction
(BHJ) PSC has become one of the most successful device structures
developed in the field to date. By simply blending polymers (electron
donors) with fullerene (electron acceptor) in organic solvents, a self-
assembling interpenetrating network can be obtained using various
coating technologies ranging from laboratory-scale spin coating or
spray coating to large-scale fabrication technologies such as inkjet
printing
5,6
,doctorblading
2
,gravure
7
, slot-die coating
8
and flexo-
graphic printing
9
. In the last few years, several effective methods
have been developed to optimize the interpenetrating network
formed by the electron donor and acceptor, including solvent
annealing (or slow-growth)
10
, thermal annealing
11–13
and mor-
phology control using mixed solvent mixtures
14
or additives
15
in
the solutions of donor/acceptor blends. Poly(3-hexylthiophene)
(P3HT) in particular has been subject to increasing interest in the
polymer research community, but significant progress has also been
made in developing new active-layer polymer materials
4,15–22
.Since
around 2008, the efficiency of PSCs has risen to 6% using new con-
jugated polymers as electron donors
19
. Although progress has been
impressive, there is still much to do before the realization of practical
applications of PSCs. Many factors need to be taken into account in
efficiently converting sunlight into electricity. The absorption range,
the photon–electron conversion rate and the carrier mobilities of
the light-harvesting polymers are among the crucial parameters for
achieving high-efficiency solar cells. Furthermore, fabricating large-
area devices without significantly losing efficiency while maintaining
long device lifetimes remains challenging
23
.
In principle, the strategies used to improve BHJ solar cell effi-
ciency include (i) reducing the bandgap of polymers so as to
harvest more sunlight, which leads to higher short-circuit current
density (J
sc
) and (ii) lowering the highest occupied molecular
orbital (HOMO) of the polymers, which increases the open-
circuit voltage (V
oc
). With the rise in interest in using low-
bandgap polymers to harvest more sunlight from longer wave-
lengths, much effort has been made recently in reducing the
bandgap of polymers. Others
4,15,21
have reported PSCs with power
conversion efficiencies (PCE) of over 5% using different low-
bandgap polymers. The extended absorption of sunlight at longer
wavelengths directly reflects on the value of J
sc
, and a current
density of up to 16 mA cm
22
has been achieved
15
. On the other
hand, PSCs with high V
oc
have been realized by other
groups
19,20,22
by using polymers that absorb at shorter wavelengths.
To push the PCE of PSCs towards the predicted theoretical limit-
ation
24
, however, achieving both a high J
sc
and a high V
oc
is critical,
indeed essential. To match the energy level of the commonly used
electron a cceptor [6,6]-phenyl-C
61
-butyric acid methyl ester (PCBM),
both the HOMO and the lowest unoccupied molecular orbital
(LUMO) of the polymer need to be considered while tuning the
bandgap of polymers. It is known that the energy difference
between the LUMOs of donor and acceptor should be larger than
0.3 eV for efficient charge separation
24
, which directly relates to
the J
sc
of solar cells. However, the V
oc
of PSCs is limited by the
difference between the HOMO of the donor and the LUMO of
the acceptor
2
. As a result,narrowingthebandgapofpolymerswithout
sacrificing efficient charge separation as well as high V
oc
becomes a
major hurdle in achieving high-efficiency PSCs. In this work, we
attempt to alter the HOMO of poly[4,8-bis-substituted-benzo
[1,2-b:4,5-b
0
]dithiophene-2,6-diyl-alt-4-substituted-thieno[3, 4-b]thio-
phene-2,6-diyl] (PBDTTT)-derived polymers
4
by adding different
electron-withdrawing functional groups, step by step. We have
shown that the addition of more than one electron-withdrawing
group is effective in further lowering the HOMO of PBDTTT.
As reported, PSCs based on the copolymer of benzo[1,2-b:4,5-
b
0
]dithiophene and thieno[3,4-b]thiophene (ref. 4) (hereafter
referred to as PBDTTT–E, Fig. 1) can yield a J
sc
greater than
15 mA cm
22
with a V
oc
of 0.6 V. We chose this polymer system
and tried to increase PSC performance by increasing the V
oc
through molecular design. Previous studies on thiophene-based
polymers have shown that the alkyloxy chain has a much stronger
electron-donating effect than an alkyl chain
25
. As a result, the
HOMO of poly(3-alkoxythiophene) is higher than that of poly(3-
alkylthiophene). Based on this knowledge, we replaced the alkyloxy
group on the carbonyl of the thieno[3,4-b]thiophene unit with an
alkyl side chain (hereafter referred to as PBDTTT–C). Both
1
Solarmer Energy Inc., El Monte, California 91731, USA,
2
Department of Materials Science and Engineering, University of California, Los Angeles,
Los Angeles, California 90095, USA,
3
Department of Chemistry and the James Franck Institute, The University of Chicago, 929 East 57th Street, Chicago,
Illinois 60637, USA.
*
e-mail: jianhuih@solarmer.com; yuew@solarmer.com
LETTERS
PUBLISHED ONLINE: 25 OCTOBER 2009 | DOI: 10.1038/NPHOTON.2009.192
NATURE PHOTONICS | VOL 3 | NOVEMBER 2009 | www.nature.com/naturephotonics 649
© 2009 Macmillan Publishers Limited. All rights reserved.
PBDTTT–E and PBDTTT–C (Fig. 1a) were thus synthesized and
their HOMO and LUMO levels measured by means of electroche-
mical cyclic voltammetry (CV). Lower HOMO and LUMO levels
were observed for PBDTTT–C than for PBDTTT–E, as shown in
Fig. 1b. Interestingly, the alkyl group not only results in a lower
HOMO but also a lower LUMO; both the HOMO and the
LUMO of PBDTTT–C are 0.1 eV lower than those of
PBDTTT–E. Therefore, the bandgap of PBDTTT–C is approxi-
mately the same as that of PBDTTT–E, which is supported by
their absorption spectra (Fig. 1c).
In addition to grafting side chains, substitution of the carbon
atom in selected locations also affects the energy levels of a
polymer. In the recent work
26
, higher values of V
oc
are observed
when fluorine, an atom of high electron affinity, is introduced to
the thieno[3,4-b]thiophene unit, a PCE of 6.1% having been
demonstrated
26
. To this end, PBDTTT–C was modified with a flu-
orine atom to lower its HOMO level. The structure of the designed
and synthesized PBDTTT–CF is shown in Fig. 1a. The HOMO and
LUMO of PBDTTT–CF were measured and compared with
PBDTTT–C (Fig. 1b). As was observed with the addition of the
alkyl group, the introduction of fluorine further lowered both the
HOMO and LUMO levels. It is rather interesting that the HOMO
and LUMO of this polymer system have this peculiar property. It
is known that the application of a functional group to a polymer
backbone causes either the HOMO or LUMO level to shift.
However, to our knowledge, a change in both the HOMO and
LUMO levels with just one synthetic modification to the polymer
backbone has not been reported in any polymer system. The
reason for this simultaneous change is not clear and is still under
investigation. The absorption spectra of polymer films based on
these three polymers show similar absorption edges (at
770 nm), suggesting similar bandgaps. This is consistent with
the CV results.
PSC devices based on these three polymers were fabricated and
tested under simulated 100 mW cm
22
AM1.5G illumination (see
Methods). The optimized weight ratios of polymer to PC
70
BM for
PBDTTT–E, PBDTTT–C and PBDTTT–CF are 1:1, 1:1.5 and
1:1.5, respectively. Device current density/voltage (J–V) character-
istics are shown in Fig. 2a and the parameters listed in Table 1.
More than 200 devices were fabricated and their highest efficiencies
and average values compared to provide numerical data. From the
S
S
O
O
S
O
S
O
n
S
S
O
O
S
O
S
n
S
S
O
O
S
O
S
n
F
PBDTTT–E PBDTTT–CF
−7.0
−6.5
−6.0
−5.5
−5.0
−4.5
−4.0
−3.5
−3.0
−2.5
−2.0
a
b
c
PBDTTT−E
PBDTTT−CF
PBDTTT−C
PCBM
−6.10
−4.30
−5.22
−3.45
−5.12
−3.35
−5.01
−3.24
Energy level (eV)
300 400 500 600 700 800 900
0.0
0.2
0.4
0.6
0.8
1.0
PBDTTT−E
PBDTTT−C
PBDTTT−CF
Normalized absorption (a.u.)
Wavelength (nm)
PBDTTT–C
Figure 1 | Comparison of three different polymers. a–c, Chemical structure
(a) , energy levels (b) and absorption spectra (c) of PBDTTT–E, PBDTTT–C
and PBDTTT–CF. Different HOMOs and LUMOs were obtained when
different functional groups were attached to the PBDTTT backbone.
Nevertheless, the absorption edges of these three polymers are about
the same.
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
−20
−16
−12
−8
−4
0
4
a
b
PBDTTT−E
PBDTTT−C
PBDTTT−CF
Current density (mA cm
−2
)
Bias (V)
Voltage (V)
−0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
−0.1
Current (mA)
V
oc
= 0.7631 V
I
sc
= 0.63240 mA
J
sc
= 13.364 mA cm
−2
Fill factor = 66.39%
I
max
= 0.51800 mA
V
max
= 0.6185 V
P
max
= 0.32030 mW
Eciency = 6.77%
Figure 2 | Characterization of devices based on PBDTTT–E, PBDTTT–C and
PBD TTT–CF. a, Current density v ersus voltage (J–V) curves obtained from
our laboratory . b, J–V curve of a device based on PBDTTT–CF certified by the
NREL. A significant increase in the open-circuit voltage (V
oc
) is clearly seen
between the PBDTTT–E and PBDTTT–CF. A V
oc
of up to 0.76 V was
observed in devices based on PBDTTT–CF.
LETTERS
NATURE PHOTONICS DOI: 10.1038/NPHOTON.2009.192
NATURE PHOTONICS | VOL 3 | NOVEMBER 2009 | www.nature.com/naturephotonics650
© 2009 Macmillan Publishers Limited. All rights reserved.
J–V curves, a significant increase in V
oc
is clearly observed from
PBDTTT–E to PBDTTT–CF. A V
oc
as high as 0.76 V was observed
in devices based on PBDTTT–CF. Combined with its high J
sc
and fill
factor (FF), a high PCE of 7.38+0.4% (a 5% device variation),
measured in more than 75 devices, was achieved in the PBDTTT–
CF system, the highest measured PCE being 7.73%. Devices were
then encapsulated and sent to the National Renewable Energy
Laboratory (NREL) for certification. A certified efficiency of
6.77% (the highest efficiency achieved so far for organic solar
cells) was rated, which differs by 8% from the average obtained
efficiency of 7.38%. The current/voltage curve and corresponding
parameters are shown in Fig. 2b. Among the parameters, the drop
in efficiency results predominantly from the lower J
sc
, most prob-
ably due to degradation of the device
27
. Although the devices were
encapsulated with UV epoxy before shipping to NREL, this non-
ideal encapsulation process has been observed to cause an efficiency
drop of 3% over a 10-day storage period in a nitrogen-filled glove-
box (,0.1 ppm O
2
and H
2
O). Furthermore, part of the device metal
finger (non-active area) was exposed to air during wiring for
measurement purposes. Oxidation of the calcium/aluminium elec-
trodes when exposed to air could also contribute to this difference.
To further confirm the accuracy of the measurements, the external
quantum efficiency (EQE) of the devices based on the three polymers
was measured using an EQE system (Model QEX7) purchased from
PV Measurements. The EQE curves are shown in Fig. 3a. All the
devices show rather efficient photoconversion efficiency in the
range 400–700 nm, with EQE values of 50–70%. In the PBDTTT–
CF device, the highest EQE value was 68.7% at 630 nm, which, to
the best of our knowledge, is the highest achieved in a low-
bandgap PSCs system. The J
sc
values were then calculated by
integrating the EQE data with an AM1.5G reference spectrum. The
calculated J
sc
values were 213.0, 214.1 and 215.0 mA cm
22
for
devices based on PBDTTT–E, PBDTTT–C and PBDTTT–CF,
respectively. These values are rather consistent with those (within
4% error, see Table 1) obtained from the J–V measurement.
In fact, the J
sc
values for the PBDTTT–CF solar cell obtained using
these two methods are rather close (215.0 and 215.2 mA cm
22
,
1% difference). The absorption curves of polymer:PC
70
BM blend
films are shown in Fig. 3b. Significant absorption increases are
observed in the absorption range 300–600 nm after blending with
PC
70
BM. According to the absorption spectrum, the polymer films
still have 50% transparency (optical density, OD 0.3) under opti-
mized device conditions, suggesting their potential as a material for
tandem or stackable solar cells
28
. The internal quantum efficiency
(IQE) of the device based on PBDTTT–CF was obtained from its
absorption spectrum and EQE curve. As shown in Fig. 3c, the
average IQE was higher than 90% in the range 400–700 nm, indicat-
ing a highly efficient overall photoconversion process in the cell. To
reach such a high IQE, it is necessary to simultaneously achieve effi-
cient light absorption, exciton diffusion, charge separation and
carrier collection. Physically, this indicates a close to ideal polymer
active layer morphology, with a bicontinuous interpenetrating poly-
mer/fullerene network. Combined with the high V
oc
and FF, a solar
cell with efficiency greater than 7% was achieved. The carrier mobi-
lities of these three devices were measured using the space charge
Ta b le 1 | Solar cell parameters of devices based on the three different polymers.
LUMO HOMO V
oc
(V) J
sc
(mA cm
22
)FF(%) PCE(%)
Best Ave
PBDTTT–E 23.24 25.01 0.62 213.2 63 5.15 4.8
PBDTTT–C 23.35 25.12 0.7 214.7 64.1 6.58 6.3
PBDTTT–CF 23.45 25.22 0.76 215.2 66.9 7.73 7.4
LUMO, lowest unoccupied molecular orbital; HOMO, highest occupied molecular orbital; FF, fill factor, PCE, power conversion efficiency.
300 400 500 600 700 800 900
0
10
20
30
40
50
60
70
80
90
a
b
c
PBDTTT−E
PBDTTT−C
PBDTTT−CF
EQE (%)
Wavelength (nm)
400 500 600 700 800
0.0
0.1
0.2
0.3
0.4
OD
Wavelength (nm)
400 500 600 700 800
Wavelength (nm)
PBDTTT−E
PBDTTT−C
PBDTTT−CF
0.0
0.2
0.4
0.6
0.8
1.0
0
20
40
60
80
100
OD
Quantum eciency (%)
Absorbance
EQE
IQE
PBDTTT−CF
Figure 3 | Photoconversion efficiency study. a–c, External quantum
efficiency (EQE) (a) and absorption (OD, optical density) (b) for devices
based on PBDTTT–E, PBDTTT–C and PBDTTT–CF, and internal quantum
efficiency (IQE) for device based on PBDTTT–CF (c). The average IQE is
greater than 90% in the range 400–700 nm. Near 100% IQE was obtained
for the device based on PBDTTT–CF, indicating a highly efficient o v er all
photoconversion process in the cell.
NATURE PHOTONICS DOI: 10.1038/NPHOTON.2009.192
LETTERS
NATURE PHOTONICS | VOL 3 | NOVEMBER 2009 | www.nature.com/naturephotonics 651
© 2009 Macmillan Publishers Limited. All rights reserved.
limited current (SCLC) method reported previously
29
. The hole
mobilities obtained using this method were 4 10
24
,2 10
24
and 7 10
24
cm
2
V
21
s
21
for PBDTTT–E, PBDTTT–C and
PBDTTT–CF devices, respectively. The observed variation in the
hole mobilities of these three polymers may be the reason for the
difference in thickness existing in the optimized devices (790,
660 and 970 Å, respectively).
The nanoscale morphologies of the polymer/PC
70
BM films were
studied using tapping-mode atomic force microscopy (AFM).
Surface topography (left) and phase images (right) were taken for
each film and are shown in Fig. 4. Surface roughness values
measured from the topography images were 0.96, 0.92 and
0.84 nm for PBDTTT–E (Fig. 4a), PBDTTT–C (Fig. 4b) and
PBDTTT–CF (Fig. 4c) films, respectively. Very different mor-
phologies were observed for these three polymers in their phase
images (Fig. 4, right panels). As shown in Fig. 4, fibrillar features
are clearly seen in PBDTTT–E (Fig. 4a) and PBDTTT–C (Fig. 4b)
films, whereas domains with different shapes are observed in the
PBDTTT–CF (Fig. 4c) film. Fibrillar features can be clearly seen
even in the topography image of the PBDTTT–C film. The very
different morphology of these three polymer films suggests that
the interactions between the molecules of each polymer may be
different. The fact that the polymer system (PBDTTT–CF) with see-
mingly lowest organization shows the highest performance differs
20.0 nm
10.0 nm
20.0 nm
20.0°
20.0°
20.0°
0 Height 500 nm 0 Phase 500 nm
0 Height 500 nm
0
Phase 500 nm
0 Height 500 nm 0 Phase 500 nm
a
b
c
Figure 4 | Tapping-mode atomic force microscopy images of the three films used in making the devices (under optimized device conditions).
a–c, The topography of ea ch film is sho wn in the left panels, and the corr esponding phase images in the right panels. Very different surface morphologies
were obtained for the different polymer films. Fibrillar features are clearly seen in PBDTTT–E (a) and PBDTTT–C (b) films, whereas domains with differ ent
shapes are observed in the PBD TTT–CF (c) film. The fibrillar features can be clearly seen even in the topography image of the PBD TTT–C film.
LETTERS
NATURE PHOTONICS DOI: 10.1038/NPHOTON.2009.192
NATURE PHOTONICS | VOL 3 | NOVEMBER 2009 | www.nature.com/naturephotonics652
© 2009 Macmillan Publishers Limited. All rights reserved.
greatly from the widely studied benchmark P3HT solar cell
system
30,31
, in which the noodle-like structure typically correlates
very well to the crystallinity
32
as well as the PCE. This indicates
that the correlation of the morphology and PCE in the new
polymer systems differs significantly from conventional under-
standing. Further investigation is under way to provide
more information.
To conclude, tuning the V
oc
of PSCs by means of molecular
design has been realized using a step-by-step approach. Applying
stronger electron-withdrawing groups to the backbone of polymers
has been found to be effective in lowering the HOMO of polymers,
which directly affects the V
oc
of PSCs. Based on the PBDTTT
polymer derivative system, PSCs with a PCE higher than 7% have
been realized by combining the advantages of a low HOMO level
in the polymer (high V
oc
) and long wavelength absorption (high
J
sc
). Regarding the LUMO level of PCBM, there is still much
capacity for increasing the V
oc
by tuning the energy levels of the
polymers. A further improvement in efficiency can be expected if
the HOMO of the polymer can be further lowered without loss of J
sc
.
Methods
All the polymers were synthesized in our laboratory and PC
70
BM was purchased
from Nano-C (used as received). Polymers and PC
70
BM were then dissolved in
chlorobenzene with 1:1 (PBDTTT–E:PC
70
BM ¼ 10 mg ml
21
:10 mg ml
21
) and
1:1.5 (10 mg ml
21
:15 mg ml
21
) weight ratio, respectively. 1,8-Diiodooctane
(purchased from Sigma Aldrich, used as received) with 3% volume ratio was then
added to the solutions and stirred before use. The solutions were spin-coated on
indium tin oxide (ITO)/glass substrates with a pre-coated PEDOT:PSS
(poly(ethylenedioxythiophene):polystyrene sulphonate) layer. The device active area
was 0.1 cm
2
for all the solar cell devices discussed in this work. Device
characterization was carried out in air after encapsulation under simulated AM1.5G
irradiation (100 mW cm
22
) using a xenon-lamp-based solar simulator.
EQE measurements of the encapsulated devices were performed in air
(PV Measurements, Model QEX7). Some oxidization of the electrodes
(calcium/aluminium) was observed when the devices were open to the air.
Received 3 August 2009; accepted 28 September 2009;
published online 25 October 2009
References
1. Sariciftci, N. S., Smilowitz, L., Heeger, A. J. & Wudl, F. Photoinduced electron-
transfer from a conducting polymer to buckminsterfullerene. Science 258,
1474–1476 (1992).
2. Brabec, C. J. et al. Origin of the open circuit voltage of plastic solar cells. Adv.
Funct. Mater. 11, 374–380 (2001).
3. Hou, J. H. et al. Bandgap and molecular energy level control of conjugated
polymer photovoltaic materials based on benzo[1,2-b:4,5-b’]dithiophene.
Macromolecules 41, 6012–6018 (2008).
4. Liang, Y. Y. et al. Development of new semiconducting polymers for high
performance solar cells. J. Am. Chem. Soc. 131, 56–57 (2009).
5. Hoth, C. N., Choulis, S. A., Schilinsky, P. & Brabec, C. J. High photovoltaic
performance of inkjet printed polymer: fullerene blends. Adv. Mater. 19,
3973–3978 (2007).
6. Aernouts, T., Aleksandrov, T., Girotto, C., Genoe, J. & Poortmans, J. Polymer
based organic solar cells using ink-jet printed active layers. Appl. Phys. Lett.
92, 033306 (2008).
7. Pudas, M., Hagberg, J. & Leppavuori, S. Gravure offset printing of polymer inks
for conductors. Progr. Org. Coatings 49, 324–335 (2004).
8. Krebs, F. C., Gevorgyan, S. A. & Alstrup, J. A roll-to-roll process to flexible
polymer solar cells: model studies, manufacture and operational stability studies.
J. Mater. Chem. 19, 5442–5451 (2009).
9. Krebs, F. C. Fabrication and processing of polymer solar cells: a review of
printing and coating techniques. Sol. Eng. Mater. Sol. Cells 93, 394–412 (2009).
10. Li, G. et al. High-efficiency solution processable polymer photovoltaic cells by
self-organization of polymer blends. Nature Mater. 4, 864–868 (2005).
11. Ma, W. L., Yang, C. Y., Gong, X., Lee, K. & Heeger, A. J. Thermally stable,
efficient polymer solar cells with nanoscale control of the interpenetrating
network morphology. Adv. Funct. Mater. 15, 1617–1622 (2005).
12. Padinger, F., Rittberger, R. S. & Sariciftci, N. S. Effects of postproduction
treatment on plastic solar cells. Adv. Funct. Mater. 13, 85–88 (2003).
13. Li, G., Shrotriya, V., Yao, Y. & Yang, Y. Investigation of annealing effects and film
thickness dependence of polymer solar cells based on poly(3-hexylthiophene).
J. Appl. Phys. 98, 043704 (2005).
14. Zhang, F. L. et al. Influence of solvent mixing on the morphology and
performance of solar cells based on polyfluorene copolymer/fullerene blends.
Adv. Funct. Mater. 16, 667–674 (2006).
15. Peet, J. et al. Efficiency enhancement in low-bandgap polymer solar cells by
processing with alkane dithiols. Nature Mater. 6, 497–500 (2007).
16. Chen, C. P., Chan, S. H., Chao, T. C., Ting, C. & Ko, B. T. Low-bandgap
poly(thiophene–phenylene–thiophene) derivatives with broaden absorption
spectra for use in high-performance bulk-heterojunction polymer solar cells.
J. Am. Chem. Soc. 130, 12828–12833 (2008).
17. Blouin, N., Michaud, A. & Leclerc, M. A low-bandgap poly(2,7-carbazole)
derivative for use in high-performance solar cells. Adv. Mater. 19,
2295–2300 (2007).
18. Gadisa, A. et al. A new donor–acceptor–donor polyfluorene copolymer
with balanced electron and hole mobility. Adv. Funct. Mater. 17,
3836–3842 (2007).
19. Park, S. H. et al. Bulk heterojunction solar cells with internal quantum efficiency
approaching 100%. Nature Photon. 3, 297–303 (2009).
20. Chen, M. H. et al. Efficient polymer solar cells with thin active layers based on
alternating polyfluorene copolymer/fullerene bulk heterojunctions. Adv. Mater.
21, 1–5 (2009).
21. Hou, J., Chen, H.-Y., Zhang, S., Li, G. & Yang, Y. Synthesis, characterization and
photovoltaic properties of a low bandgap polymer based on silole-containing
polythiophenes and benzo[c][1,2,5]thiadiazole. J. Am. Chem. Soc. 130,
16144–16145 (2008).
22. Wang, E. et al. High-performance polymer heterojunction solar cells of a
polysilafluorene derivative. Appl. Phys. Lett. 92, 033307 (2008).
23. Krebs, F. C. et al. A complete process for production of flexible large area
polymer solar cells entirely using screen printing—first public demonstration.
Sol. Eng. Mater. Sol. Cells 93, 422–441 (2009).
24. Scharber, M. C. et al. Design rules for donors in bulk-heterojunction solar cells—
towards 10% energy-conversion efficiency. Adv. Mater. 18, 789–794 (2006).
25. Shi, C. J., Yao, Y., Yang, Y. & Pei, Q. B. Regioregular copolymers of 3-
alkoxythiophene and their photovoltaic application. J. Am. Chem. Soc. 128,
8980–8986 (2006).
26. Liang, Y. Y. et al. Highly efficient solar cell polymers developed via
fine-tuning of structural and electronic properties. J. Am. Chem. Soc. 131,
7792–7799 (2009).
27. Jorgensen, M., Norrman, K. & Krebs, F. C. Stability/degradation of polymer
solar cells. Sol. Eng. Mater. Sol. Cells 92, 686–714 (2008).
28. Shrotriya, V., Wu, E. H. E., Li, G., Yao, Y. & Yang, Y. Efficient light harvesting
in multiple-device stacked structure for polymer solar cells. Appl. Phys. Lett.
88, 064104 (2006).
29. Shrotriya, V., Yao, Y., Li, G. & Yang, Y. Effect of self-organization in
polymer/fullerene bulk heterojunctions on solar cell performance. Appl. Phys.
Lett. 89, 063505 (2006).
30. Li, G. et al. ‘Solvent annealing’ effect in polymer solar cells based on
poly(3-hexylthiophene) and methanofullerenes. Adv. Funct. Mater. 17,
1636–1644 (2007).
31. Li, G., Shrotriya, V., Yao, Y., Huang, J. S. & Yang, Y. Manipulating regioregular
poly(3-hexylthiophene): [6,6]-phenyl-C-61-butyric acid methyl ester blends—
route towards high efficiency polymer solar cells. J. Mater. Chem. 17,
3126–3140 (2007).
32. Chu, C. W. et al . Control of the nanoscale crystallinity and phase separation in
polymer solar cells. Appl. Phys. Lett. 92,
103306 (2008).
Acknowledgements
The National Renewable Energy Laboratory (NREL) is thanked for conducting the
certification of devices. The authors in particular thank D.C. Olson at NREL for his help in
verifying and certifying the performances of our devices.
Author contributions
H.Y.C. conceived and performed PSC fabrication, measurements and data analysis. Y.L.
and L.Y. contributed to the design of the polymer’s main chain structure. J.H. designed the
polymer structures and synthesis routes. S.Z. synthesized the polymers. G.Y. performed the
AFM image scans. J.H. and Y.W. conceptualized and directed the research project. All
authors discussed the results and commented on the manuscript.
Additional information
Reprints and permission information is available online at http://npg.nature.com/
reprintsandpermissions/. Correspondence and requests for materials should be addressed to
J.H. and Y.W.
NATURE PHOTONICS DOI: 10.1038/NPHOTON.2009.192
LETTERS
NATURE PHOTONICS | VOL 3 | NOVEMBER 2009 | www.nature.com/naturephotonics 653
© 2009 Macmillan Publishers Limited. All rights reserved.