Enhanced Power-Conversion Efficiency in Polymer Solar Cells Using an Inverted Device Structure

Abstract

Polymer-fullerene bulk heterojunction solar cells (PSCs) are currently
attracting a great deal of attention and gaining increasing importance,
having already shown great promise as renewable, lightweight and
low-cost energy sources. Recently, the power-conversion efficiency of
state-of-the-art PSCs has exceeded 8% in the scientific literature.
However, to find viable applications for this emerging photovoltaic
technology, further enhancements in the efficiency towards 10% (the
threshold for commercial applications) are urgently required. Here, we
demonstrate highly efficient PSCs with a certified efficiency of 9.2%
using an inverted structure, which simultaneously offers ohmic contact
for photogenerated charge-carrier collection and allows optimum photon
harvest in the device. Because of the ease of use and drastic boost in
efficiency provided by this device structure, this discovery could find
use in fully exploiting the potential of various material systems, and
also open up new opportunities to improve PSCs with a view to achieving
an efficiency of 10%.

Full text

Enhanced power-conversion efficiency in polymer
solar cells using an inverted device structure
Zhicai He, Chengmei Zhong, Shijian Su, Miao Xu, Hongbin Wu*and Yong Cao
Polymer-fullerene bulk heterojunction solar cells (PSCs) are
currently attracting a great deal of attention and gaining
increasing importance, having already shown great promise
as renewable, lightweight and low-cost energy sources1–4.
Recently, the power-conversion efficiency of state-of-the-art
PSCs has exceeded 8% in the scientific literature5. However,
to find viable applications for this emerging photovoltaic tech-
nology, further enhancements in the efficiency towards 10%
(the threshold for commercial applications) are urgently
required6. Here, we demonstrate highly efficient PSCs with a
certified efficiency of 9.2% using an inverted structure, which
simultaneously offers ohmic contact for photogenerated
charge-carrier collection and allows optimum photon harvest
in the device. Because of the ease of use and drastic boost in
efficiency provided by this device structure, this discovery
could find use in fully exploiting the potential of various
material systems, and also open up new opportunities to
improve PSCs with a view to achieving an efficiency of 10%.
Recently, there have been extensive investigations of polymer
solar cells (PSCs) with an inverted device structure, using modified
indium tin oxide (ITO) as the cathode (the ITO is modified by
n-type metal oxides or metal carbonates7–13, including titanium
oxide, zinc oxide and caesium carbonate, conjugated polyelectro-
lyte14–16, self-assembled crosslinked fullerene17,18 and self-assembled
polar molecules19). Compared with conventional PSCs, inverted-
type devices demonstrate better long-term ambient stability by
avoiding the need for the corrosive and hygroscopic hole-transporting
poly(3,4-ethylenedioxylenethiophene):poly(styrenesulphonic acid)
(PEDOT:PSS) and low-work-function metal cathode, both of
which are detrimental to device lifetime18,20. Moreover, inverted
PSCs can also take advantage of the vertical phase separation and
concentration gradient in the active layer20,21, which is naturally
self-encapsulated because air-stable metals are used as the top
electrode9. The inverted device structure is therefore an ideal
configuration for all types of PSCs. Despite these clear benefits,
the performances of most inverted PSCs reported to date are
inferior to those of regular devices. Indeed, the capability of inverted
PSCs to provide independent control of photon harvesting from the
Sun’s spectrum has not been widely recognized.
In this study, we demonstrate that the efficiency of PSCs can be
further enhanced to 9%, an improvement of 10% over the previous
record-high efficiency of 8.370% (ref. 5). This is obtained with an
inverted device structure, where an alcohol-/water-soluble conju-
gated polymer, poly [(9,9-bis(3′-(N,N-dimethylamino)propyl)-2,7-
fluorene)-alt-2,7-(9,9–dioctylfluorene)] (PFN)22 is used as the ITO
surface modifier, and a blend of [6,6]-phenyl C
71
-butyric acid
methyl ester (PC
71
BM) and the low-bandgap semiconducting
polymer thieno[3,4-b]thiophene/benzodithiophene (PTB7)3is
used as the photoactive layer (see Fig. 1a for device and chemical
structure). Unlike previously reported inverted PSCs, which are
typically based on n-type metal oxides, our device is solution-
processed at room temperature, enabling easy processibility over a
large area. Accordingly, the approach is fully amenable to high-
throughput roll-to-roll manufacturing techniques, may be used to
fabricate vacuum-deposition-free PSCs of large area, and find
practical applications in future mass production. Moreover, our
discovery overturns a well-accepted belief (the inferior performance
of inverted PSCs) and clearly shows that the characteristics of high
performance, improved stability and ease of use can be integrated
into a single device, as long as the devices are optimized, both opti-
cally and electrically, by means of a meticulously designed device
structure. We also anticipate that our findings will catalyse the
development of new device structures and may move the efficiency
of devices towards the goal of 10% for various material systems.
Previously, we reported that PFN can be incorporated into
polymer light-emitting devices (PLEDs) to enhance electron injection
from high-work-function metals such as aluminium (work function
w
of 4.3 eV)22,23 and has thus been used to realize high-efficiency,
air-stable PLEDs24. Furthermore, we also found that efficient
electron injection can be obtained even in the most noble metals
with extremely high work functions, such as gold (
w
¼5.2 eV), by
lowering the effective work function (for example lowering
w
in
gold by 1.0 eV), which has previously been ascribed to the for-
mation of a strong interface dipole25.
After a thin layer (10 nm) of PFN was applied on top of the ITO,
the work function of the ITO was reduced from 4.7 eV to 4.1 eV, as
demonstrated by X-ray photoelectron spectroscopy (XPS) measure-
ments (Supplementary Fig. S1). Although the microscopic origin of
this reduction in the effective work function of ITO remains uncer-
tain at this point, we speculate that the formation of an interface
dipole at the ITO surface as a result of orientation of PFN with a
permanent dipole is the major cause for the energy level alignment
at the ITO/PFN interface25,26. The formation of an interfacial dipole
layer can be confirmed by small-angle X-ray diffraction (XRD)
(Supplementary Fig. S2). On the other hand, mechanisms27 such
as electron cloud push-back effects, the integer charge-transfer
model and the induced density of states model cannot be ruled
out at this stage. We also note that a universal model for the
energy level alignment of molecules on metal oxides has been devel-
oped recently by Greiner and colleagues28, in which Fermi level-
pinning plays a critical role in tuning the energy alignment of a
metal oxide. Regardless, as a result of the abrupt shift of vacuum
level at the interface (corresponding to a lowering of the effective
work function), the modified ITO can form ohmic contact with
the photoactive layer and can therefore be used as a cathode for
inverted-structure PSCs to facilitate transport and collection of
photogenerated charge carriers (Fig. 1b,c).
Figure 2a presents the current density versus voltage (J–V)
characteristics of the best PSCs with regular and inverted structures,
under 1,000 W m
22
air mass 1.5 global (AM 1.5G) illumination.
Institute of Polymer Optoelectronic Materials and Devices, State Key Laboratory of Luminescent Materials and Devices, South China University of
Technology, Guangzhou 510640, PR China. *e-mail: hbwu@scut.edu.cn
LETTERS
PUBLISHED ONLINE: 19 AUGUST 2012 | DOI: 10.1038/NPHOTON.2012.190
NATURE PHOTONICS | VOL 6 | SEPTEMBER 2012 | www.nature.com/naturephotonics 591
© 2012 Macmillan Publishers Limited. All rights reserved.

Device parameters such as J
SC
,V
OC
, FF and PCE are deduced from
the J–Vcharacteristics (summarized in Table 1). A high average
PCE of 8.97+0.10% was achieved in inverted structure devices
(with a maximum of 9.15%), while the conventional devices
showed an average PCE of 8.06+0.14% (with a highest value of
8.24%, consistent with our previous report5). It is clear from
Table 1 that the efficiency improvement in the inverted structure
devices is mainly due to the higher J
SC
(17.2 mA cm
22
versus
15.4 mA cm
22
), because V
OC
and FF remain nearly the same.
The external quantum efficiency (EQE) curves of the devices are
presented in Fig. 2b, with photoresponses with average EQEs of 70%
and 63% (in the range 300–700 nm) for the inverted and regular
structure devices, respectively. Note that the theoretical J
SC
values
obtained by integrating the product of the EQE data in Fig. 2b
and the AM 1.5G solar spectrum are 17.1 mA cm
22
and
15.7 mA cm
22
, respectively, which are in good agreement with
the values obtained from the J–Vcharacteristics (Fig. 2a, Table 1).
After encapsulation, the devices were sent to the National Center
of Supervision and Inspection on Solar Photovoltaic Products
Quality of China (CPVT) for certification. A certified PCE of
9.214% was obtained (Fig. 2c, Table 1, Supplementary Fig. S3).
Notably, another independent testing laboratory NEWPORT
Corporation certified a PCE of 8.6+0.3% for a representative
device (Supplementary Fig. S3). This result is 7% lower than
that measured in CPVT, which can be attributed to non-ideal cell
encapsulation. The results from both external certification labora-
tories confirmed that the efficiency of our devices is among the
highest reported to date for PSCs.
The encapsulated devices were stored in air under ambient con-
ditions and their efficiency measured periodically to assess their
long-term stability (Supplementary Fig. S4). The inverted device
retained 95% of its initial efficiency (9.00%) after exposure to air
for 62 days, whereas the regular devices lost half their initial effi-
ciency after 10 days. This demonstrates that the inverted PSCs can
achieve a high efficiency of 9% and also maintain good
ambient stability.
The drastically enhanced J
SC
in the inverted PSCs may originate
from reduced bimolecular recombination29,30 or increased absorp-
tion of photons1, or indeed a combination of both. The increase
in optical absorption was confirmed by optical modelling
ITO cathode
PFN
N
N
n
S
S
S
S
F
RO
O
n
R
ROOC
R = 2-ethylhexyl
C8H17 C8H17
MoO3
Al, Ag anode
PTB7:PC71BM
+
−
ITO
anode PEDOT:
PSS
PTB7
Evac
Evac
PFN
PFN
PC71BM
−4.7 eV
ITO
cathode
−4.1 eV
−5.2 eV
−3.31 eV
−5.15 eV
PTB7
−3.31 eV
−5.4 eV
−6.1 eV
−6.1 eV
0.6 eV
−2.7 eV
−4.3 eV −4.3 eV
Ca
cathode
−4.3 eV
Al, Ag
anode
PC71BM
MoO3
−5.15 eV
a
bc
Figure 1 | Device structure and energy levels of the inverted-type PSCs. a, Schematic of the inverted-type PSCs, in which the photoactive layer is
sandwiched between a PFN-modified ITO cathode and an Al,Ag-based top anode. Insets: chemical structures of the water-/alcohol-soluble conjugated
polymer and electron donor materials used in the study. PFN, poly [(9,9-bis(3′-(N,N-dimethylamino)propyl)-2,7-fluorene)-alt-2,7-(9,9-ioctylfluorene)]; PTB7:
thieno[3,4-b]thiophene/benzodithiophene. b,c, Schematic energy levels of the conventional (b)andinverted(c) devices under flat band conditions (open-
circuit voltage). Note that the formation of a positive interface dipole moment (taking the dipole moment directed outwards to be positive) is presented in c.
LETTERS NATURE PHOTONICS DOI: 10.1038/NPHOTON.2012.190
NATURE PHOTONICS | VOL 6 | SEPTEMBER 2012 | www.nature.com/naturephotonics592
© 2012 Macmillan Publishers Limited. All rights reserved.

calculations (Fig. 3) in which a one-dimensional transfer matrix
formalism (TMF) was applied31 and experimentally measured
refractive indices and extinction coefficients were used. Figure 3a
depicts the absorbed incident photon flux density in both devices,
each with a 100-nm-thick active layer, verifying that the inverted
structure can harvest more photons from solar spectra than the
regular devices. Furthermore, it can be seen from Fig. 3b that the
inverted device always has a higher J
SC
than the regular devices,
regardless of the active layer thickness, demonstrating the advantage
of the inverted device structure.
The stronger absorption in the inverted device can also be veri-
fied experimentally by reflectance spectra measurements on real
devices (Supplementary Fig. S5); the inverted devices show a
decreased reflectance over a wide range of spectra (400–800 nm)
when compared with the regular device.
The calculated maximum J
SC
of the inverted device under AM
1.5G illumination (440–800 nm) as a function of interlayer thick-
ness is presented in Supplementary Fig. S6b, and the performances
of devices with varied PFN thickness (5, 10 and 20 nm) are shown in
Supplementary Fig. S6a and Table S1. Although the theoretical J
SC
is
not sensitive to PFN layer thickness, the overall performance criti-
cally depends on this thickness. This is understandable, because
an interlayer that is too thick will lead to a high series resistance,
while a too-thin layer could not provide an ohmic contact for
electron extraction.
Both types of devices exhibit very high FF (70%) and demon-
strate a comparable dependence on incident light intensity
(Supplementary Fig. S7), implying that any difference in charge-
carrier recombination loss in the devices can be neglected. It can
therefore be concluded that the 10% increase in J
SC
is mainly
due to enhanced optical absorption as a result of the redistribution
of the electric field intensity of the incident photons inside the
active layer31.
To gain further insight into the origin of the enhanced J
SC
in the
device, we determined the experimental photocurrent generation
rate (G) in the devices. Given that the saturation photocurrent den-
sities (J
sat
) for the regular and inverted devices are 16.7 mA cm
22
and 18.6 mA cm
22
, respectively (Supplementary Fig. S8), the
Table 1 | Best device performance/parameters from
PTB7:PC
71
BM solar cells with conventional and inverted
device structures, measured under 1,000 W m
22
AM 1.5G
illumination.
Device type PCE (%) J
SC
(mA cm
22
)FF(%)V
OC
(V)
Conventional 8.24 15.4 70.6 0.759
Inverted 9.15 17.2 72.0 0.740
Inverted, tested by CPVT 9.214 17.46 69.99 0.754
300 400 500 600 700 800
20
0
40
60
80
EQE (%)
Wavelength (nm)
−20
−15
−10
−5
0
5
−0.2 0 0.2 0.4 0.6 0.8 1
Voltage (V)
Current density (mA cm−2)
ab
−20
−15
−10
−5
0
5
−0.2 0.2 0.4 0.6 0.8 10
Voltage (V)
Current density (mA cm−2)
c
Figure 2 | Device performances of the inverted structure (ITO/PFN interlayer/active layer/MoO
3
(10 nm)/Al) PTB7:PC
71
BM PSCs and the conventional
device (ITO/PEDOT:PSS/active layer/PFN interlayer/Ca/Al). a, Current density versus voltage (J–V) characteristics under 1,000 W m
22
AM 1.5G
illumination for conventional (open symbols) and inverted (filled symbols) devices. b,EQE/IPCE spectra for conventional (open symbols) and inverted (filled
symbols) devices. c, CPVT-certified J–Vcharacteristics of an inverted structure PTB7:PC
71
BM solar cells, with a PCE of 9.214%.
0
0 20 40 60 80 100
Position in the photoactive layer (nm)
0
5
10
15
20
0 50 100 150 200
Maximum Jsc (mA cm−2)
Active layer thickness (nm)
1.5 × 1028
ab
1 × 1028
5 × 1027
Absorbed photon flux density (m−3 s−1)
Figure 3 | Comparison of calculated optical absorption profile for conventional and inverted devices. Conventional device: ITO/PEDOT:PSS
(40 nm)/PTB7:PC
71
BM (100 nm)/PFN (5 nm)/Al (100 nm); inverted device: ITO/PFN(10 nm)/PTB7:PC
71
BM (100 nm)/MoO
3
(10 nm)/Al (100 nm).
a, Distribution of absorbed AM 1.5G photon flux density inside the active layer. b, Variation of maximum J
SC
with active layer thickness for two types of
devices, both under AM 1.5G illumination (440–800 nm). Open squares and filled circles represent conventional and inverted devices, respectively.
NATURE PHOTONICS DOI: 10.1038/NPHOTON.2012.190 LETTERS
NATURE PHOTONICS | VOL 6 | SEPTEMBER 2012 | www.nature.com/naturephotonics 593
© 2012 Macmillan Publishers Limited. All rights reserved.

maximal obtainable exciton generation rates can be determined
as 1.30 ×10
28
m
23
s
21
and 1.45 ×10
28
m
23
s
21
, respectively. In
other words, the experimental absorbed incident photon flux
density in the active layer in the inverted device is indeed 10%
higher than that of the regular device, which is in good agreement
with the observed J
SC
and the reflectance spectra of the devices
shown in Supplementary Fig. S5.
For comparison, we also fabricated inverted devices using ZnO as
an electron-selective cathode following a well-established procedure
that required high-temperature annealing11. The devices based on
PFN also show superior performance over the ZnO-based devices
(9.15% versus 8.35%, with the ZnO device showing a PCE of
8.34+0.04%, with a maximum of 8.39%; Supplementary Fig. S9
and Table S2), mainly due to a more efficient photon harvest in
this device structure (Supplementary Fig. S5).
Our approach can also be used for semitransparent PSCs7,8,10.On
the basis of the abovementioned inverted PSCs, highly efficient
semitransparent PSCs were fabricated, with the highly reflective
anode MoO
3
(10 nm)/Al (100 nm) replaced by a transparent
layer of MoO
3
(10 nm)/Ag (20 nm). The J–Vcharacteristics of
the semitransparent PSCs, under front- and rear-side AM 1.5G illu-
mination, are presented in Fig. 4. The PCE values for the device
under front-, and rear-side illumination are 6.13% and 4.96%,
respectively. We anticipate that our prototype of semitransparent
PSCs will find practical applications in building windows, foldable
curtains and invisible electronic circuits10.
In conclusion, we have successfully demonstrated highly efficient
PSCs with a certified efficiency of 9.2% using a straightforward
inverted structure, in which an alcohol-/water-soluble conjugated
polymer is used to tune the work function of ITO. By using experi-
mental photocurrent measurements in combination with optical
modelling based on a one-dimensional TMF, we have demonstrated
that, as well as offering ohmic contact for photogenerated charge-
carrier collection, the inverted structure can provide optimum
photon harvest from the sun’s spectrum. Because of a remarkably
improved J
SC
of .17 mA cm
22
, the inverted PSCs exhibit a superior
overall device performance when compared to regular devices, as
well as good ambient stability. It is worth noting that, according
to our calculation, a further increase of 10% in J
SC
could be obtained
if the optical constants of the interlayer were matched with those of
the stacks of other layers. In the quest to optimize device perform-
ance of PSCs, both optically and electrically, we anticipate that our
findings will catalyse the development of new device structures and
may further pushing the efficiency of devices towards the goal of
10% with various material systems.
Methods
Fabrication of PSCs. Electron donor material PTB7 and electron acceptor PC
71
BM
were purchased from 1-material Chemscitech and Aldrich, respectively, and used as
received. The PEDOT:PSS (Clevios P AI4083) was obtained from H.C. Starck
Clevios. The inverted device structure was ITO/PFN/PTB7:PC
71
BM(1:1.5 by
weight)/MoO
3
/Al or Ag and the regular device structure was ITO/PEDOT:PSS/
PTB7:PC
71
BM/PFN/Ca/Al (ref. 5). The PFN interlayer material was dissolved in
methanol in the presence of a small amount of acetic acid (2 mlml
21
) and its
solution (concentration, 2 mg ml
21
) was spin-coated on top of the precleaned ITO
substrate, which was treated by oxygen plasma cleaning for 4 min. For regular
devices, a 40-nm-thick PEDOT:PSS anode buffer layer was spin-cast on the ITO
substrate, then dried in a vacuum oven at 80 8C overnight. The PTB7:PC
71
BM active
blend layer, with a nominal thickness of 80 nm (with a variation of 20 nm over
the entire film), was prepared byspin-coating a mixed solvent of chlorobenzene/1,8-
diiodoctane (97:3% by volume)3solution (concentration, 25 mg ml
21
)at
1,000 r.p.m. for 2 min. A 10 nm MoO
3
layer and a 100 nm Al or Ag (20 nm for
semitransparent device) layer were subsequently evaporated through a shadow mask
to define the active area of the devices (2×8mm
2
) and form a top anode.
Characterization and measurements. PCE values were determined from J–Vcurve
measurements (using a Keithley 2400 source meter) under a 1 sun, AM 1.5G
spectrum from a solar simulator (Oriel model 91192; 1,000 W m
22
). Masks were
made using laser beam cutting technology and had well-defined areas of 3.14 or
16.0 mm
2
; these were attached to define the effective area for accurate measurement.
All masked and unmasked tests gave consistent results with relative errors within
5%. Solar simulator illumination intensity was determined using a monocrystal
silicon reference cell (Hamamatsu S1133, with KG-5 visible colour filter) calibrated
by the National Renewable Energy Laboratory (NREL). Theoretical J
SC
values
obtained by integrating the product of the EQE with the AM 1.5G solar spectrum
agreed with the measured value to within 3%. Spectral mismatch factors (M) were
calculated according to a standard procedure32, and an Mvalue of 1.03 was used to
obtain the correct photocurrent and efficiency for the PTB7 devices.
For small-angle XRD analysis, a Bruker D8 advance X-ray diffractometer with
CuKaradiation was used (operating conditions: 40 kV and 40 mA). The
diffractometer was equipped with a solid-state detector (LynxExe 1D) that facilitated
high count rates and short measuring time. XPS studies were performed on a
Thermo-VG Scientific ESCALAB 250 photoelectron spectrometer using a
monochromated AlKa(1,486.6 eV) X-ray source. All recorded peaks were
corrected for electrostatic effects by setting the C−C component of the C
1s
peak
to 284.8 eV. The base pressure in the XPS analysis chamber was ,2×10
29
mbar.
Typical atomic force microscopy (AFM) topography images of the MoO
3
deposited on top of the active layer are shown in Supplementary Fig. S11.
The films are generally smooth, with a root-mean-square (r.m.s.) roughness of
0.48 nm, implying that significant scattering at the MoO
3
interface is unlikely
to occur.
The one-dimensional spatial distribution of normalized incident light intensity
(|E|
2
) inside the devices was calculated by means of an optical TMF approach31. The
spatial distribution of the absorbed photon flux density could then be calculated by
integrating single-wavelength |E|
2
with an AM 1.5G spectrum from 440 nm to
800 nm. Finally, the theoretical maximum J
sc
for a device under AM 1.5G
illumination was determined by spatially integrating the absorbed photon flux
density within the active layer, assuming 100% internal quantum efficiency for all
wavelengths. The refractive index (n) and extinction coefficient (k) spectra (440–
840 nm) for each layer in the devices were experimentally determined using a
Horiba Jobin Yvon AUTO SE ellipsometer, and most data are shown in
Supplementary Fig. S12. The reflection spectra of the devices were recorded using
an UV–vis–NIR spectrophotometer (Lambda750, PerkinElmer).
Received 25 March 2012; accepted 6 July 2012;
published online 19 August 2012
References
1. Park, S. H. et al. Bulk heterojunction solar cells with internal quantum efficiency
approaching 100%. Nature Photon. 3, 297–303 (2009).
2. Chen, H-Y. et al. Polymer solar cells with enhanced open-circuit voltage and
efficiency. Nature Photon. 3, 649–653 (2009).
3. Liang, Y. Y. et al. For the bright future-bulk heterojunction polymer solar cells
with power conversion efficiency of 7.4%. Adv. Mater. 22, E135–E138 (2010).
4. Seo, J.-H. et al. Improved high-efficiency organic solar cells via incorporation
of a conjugated polyelectrolyte interlayer. J. Am. Chem. Soc. 133,
8416–8419 (2011).
5. He, Z. C. et al. Simultaneous enhancement of open-circuit voltage, short-circuit
current density and fill factor in polymer solar cell. Adv. Mater. 23,
4636–4643 (2011).
6. Brabec, C. J. et al. Polymer–fullerene bulk-heterojunction solar cells. Adv. Mater.
22, 3839–3856 (2010).
−5
0
5
10
15
−0.8 0
Current density (mA cm−2)
Voltage (V)
−0.6 −0.4 −0.2 0.2
Figure 4 | J–Vcharacteristics for semitransparent PSCs under front- and
rear-side AM 1.5G (1,000 W m
22
) illumination. Open squares and filled
circles represent conventional and inverted devices, respectively.
LETTERS NATURE PHOTONICS DOI: 10.1038/NPHOTON.2012.190
NATURE PHOTONICS | VOL 6 | SEPTEMBER 2012 | www.nature.com/naturephotonics594
© 2012 Macmillan Publishers Limited. All rights reserved.

7. Li, G., Chu, C. W., Shrotriya, V., Huang, J. & Yang, Y. Efficient inverted polymer
solar cells. Appl. Phys. Lett. 88, 253503 (2006).
8. Waldauf, C. et al. Highly efficient inverted organic photovoltaics using solution
based titanium oxide as electron selective contact. Appl. Phys. Lett. 89,
233517 (2006).
9. Huang, J., Li, G. & Yang, Y. A semi-transparent plastic solar cell fabricated by a
lamination process. Adv. Mater. 20, 415–419 (2008).
10. Ameri, T. et al. Fabrication, optical modeling, and color characterization of
semitransparent bulk-heterojunction organic solar cells in an inverted structure.
Adv. Funct. Mater. 20, 1592–1598 (2010).
11. Yang, T. B. et al. Solution-processed zinc oxide thin film as a buffer layer for
polymer solar cells with an inverted device structure. J. Phys. Chem. C 114,
6849–6853 (2010).
12. Sun, Y. M., Seo, J. H., Takacs, C. J., Seifter, J. & Heeger, A. J. Inverted polymer
solar cells integrated with a low-temperature-annealed sol-gel-derived ZnO film
as an electron transport layer. Adv. Mater. 23, 1679–1683 (2011).
13. Small, C. E. et al. High-efficiency inverted dithienogermole–thienopyrrolodione-
based polymer solar cells. Nature Photon. 6, 115–120 (2012).
14. Na, S-I., Kim, T. S., Oh, S-H., Kim, J-Y., Kim, S-S. & Kim, D-Y. Enhanced
performance of inverted polymer solar cells with cathode interfacial tuning via
water-soluble polyfluorenes. Appl. Phys. Lett. 97, 223305 (2010).
15. Choi, H. S. et al. Combination of titanium oxide and a conjugated polyelectrolyte
for high-performance inverted-type organic optoelectronic devices. Adv. Mater.
23, 2759–2763 (2011).
16. Zhu, Y. X., Xu, X. F., Zhang, L. J., Chen, J. W. & Cao, Y. High efficiency inverted
polymeric bulk-heterojunction solar cells with hydrophilic conjugated polymers
as cathode interlayer on ITO. Sol. Ener. Mater. Sol. Cells 97, 83–88 (2012).
17. Cheng, Y.-J., Hsieh, C.-H., He, Y. J., Hsu, C.-S. & Li, Y. F. Combination of
indene–C
60
bis-adduct and cross-linked fullerene interlayer leading to
highly efficient inverted polymer solar cells. J. Am. Chem. Soc. 132,
17381–17383 (2010).
18. Hsieh, C.-H. et al. Highly efficient and stable inverted polymer solar cells
integrated with a cross-linked fullerene material as an interlayer. J. Am. Chem.
Soc. 132, 4887–4893 (2010).
19. Park, J. S. et al. Efficient hybrid organic-inorganic light emitting diodes with
self-assembled dipole molecule deposited metal oxides. Appl. Phys. Lett. 96,
243306 (2010).
20. Xu, Z. et al. Vertical phase separation in poly(3-hexylthiophene): fullerene
derivative blends and its advantage for inverted structure solar cells. Adv. Funct.
Mater. 19, 1227–1234 (2009).
21. Campoy-Quiles, M. et al. Morphology evolution via self-organization and lateral
and vertical diffusion in polymer:fullerene solar cell blends. Nature Mater. 7,
158–164 (2008).
22. Huang, F., Wu, H. B., Wang, D. L., Yang, W. & Cao, Y. Novel electroluminescent
conjugated polyelectrolytes based on polyfluorene. Chem. Mater. 16,
708–716 (2004).
23. Wu, H. B. et al. Efficient electron injection from a bilayer cathode consisting of
aluminum and alcohol-/water-soluble conjugated polymers. Adv. Mater. 16,
1826–1830 (2004).
24. Huang, F., Wu, H. B. & Cao, Y. Water/alcohol soluble conjugated polymers as
highly efficient electron injection/transporting layer in optoelectronic devices.
Chem. Soc. Rev. 39, 2500–2521 (2010).
25. Wu, H. B., Huang, F., Peng, J. B. & Cao, Y. High-efficiency electron injection
cathode of Au for polymer light-emitting devices. Org. Elect. 6, 118–128 (2005).
26. Ishii, H., Sugiyama, K., Ito, E. & Seki, K. Energy level alignment and interfacial
electronic structures at organic/metal and organic/organic interfaces. Adv.
Mater. 11, 605–625 (1999).
27. Braun, S., Salaneck, W. R. & Fahlman, M. Energy-level alignment at
organic/metal and organic/organic interfaces. Adv. Mater. 21,
1450–1472 (2009).
28. Greiner, M. T. et al. Universal energy-level alignment of molecules on metal
oxides. Nature Mater. 11, 76–81 (2012).
29. Blom, P. W. M., Mihailetchi, V. D., Koster, L. J. A. & Markov, D. E. Device
physics of polymer:fullerene bulk heterojunction solar cells. Adv. Mater. 19,
1551–1566 (2007).
30. Shuttle, C. G., Hamilton, R., O’Regan, B. C., Nelson, J. & Durrant, J. R.
Charge-density-based analysis of the current–voltage response of
polythiophene/fullerene photovoltaic devices. Proc. Natl Acad. Sci. USA 107,
16448–16452 (2010).
31. Pettersson, L. A. A., Lucimara, S. R. & Inganas, O. Modeling photocurrent action
spectra of photovoltaic devices based on organic thin films. J. Appl. Phys. 86,
487–496 (1999).
32. Shrotriya, V. et al. Accurate measurement and characterization of organic solar
cells. Adv. Funct. Mater. 16, 2016–2023 (2006).
Acknowledgements
The authors thank Min Yun of the National Center of Supervision and Inspection on
Solar Photovoltaic Products Quality for device performance verification, Fangyan Xie of
Sun Yat-San University for XPS measurements, Bin Wu of South China University of
Technology for XRD measurements, Yanhong Meng for refractive index and extinction
coefficient measurements, Xueqing Xu for reflectance measurements and Feng Liu for
critical reading of the manuscript. H.W. and Y.C. acknowledge financial support from the
National Basic Research Program of China (no. 2009CB623602), the National Nature
Science Foundation of China (nos 50990065, 51010003, 60906032 and 61177022), the
Program for New Century Excellent Talents in University (NCET-10-0400) and
Fundamental Research Funds for the Central Universities (2009ZZ0047).
Author contributions
H.W., Z.H. and C.Z. conceived the idea and designed the experiments. Z.H. fabricated and
characterized the devices. C.Z. carried out the optical calculation. S.S. synthesized the
interlayer polymer. H.W. and Y.C. coordinated and directed the study. All authors
contributed to manuscript preparation, data analysis and interpretation, and discussed
the results.
Additional information
Supplementary information is available in the online version of the paper.
Reprints and
permission information is available online at http://www.nature.com/reprints. Correspondence
and requests for materials should be addressed to H.W.
Competing financial interests
The authors declare no competing financial interests.
NATURE PHOTONICS DOI: 10.1038/NPHOTON.2012.190 LETTERS
NATURE PHOTONICS | VOL 6 | SEPTEMBER 2012 | www.nature.com/naturephotonics 595
© 2012 Macmillan Publishers Limited. All rights reserved.