Bulk heterojunction solar cells with internal
quantum efficiency approaching 100%
Sung Heum Park1,2, Anshuman Roy1, Serge Beaupre´3, Shinuk Cho1,2, Nelson Coates1, Ji Sun Moon1,2,
Daniel Moses1, Mario Leclerc3, Kwanghee Lee1,2* and Alan J. Heeger1,2*
We report the fabrication and measurement of solar cells with 6% power conversion efficiency using the alternating
co-polymer, poly[N-900-hepta-decanyl-2,7-carbazole-alt-5,5-(40,70-di-2-thienyl-20,10,30-benzothiadiazole) (PCDTBT) in bulk
heterojunction composites with the fullerene derivative [6,6]-phenyl C70-butyric acid methyl ester (PC70BM). The
PCDTBT/PC70BM solar cells exhibit the best performance of any bulk heterojunction system studied to date, with
JSC5 10.6 mA cm
22, VOC5 0.88 V, FF5 0.66 and he5 6.1% under air mass 1.5 global (AM 1.5 G) irradiation of
100 mWcm22. The internal quantum efficiency is close to 100%, implying that essentially every absorbed photon results
in a separated pair of charge carriers and that all photogenerated carriers are collected at the electrodes.
Polymer bulk heterojunction (BHJ) solar cells based on compo-sites of an electron-donating conjugated polymer and an elec-tron-accepting fullerene offer promise for the realization of a
low-cost, printable, portable and flexible renewable energy
source1–4. Although BHJ solar cell performance has steadily
improved, with power conversion efficiencies (PCE; he) approach-
ing 5%, further improvements in efficiency are required for large-
scale commercialization5–7.
Rather than using a single junction architecture, the fundamental
BHJ concept involves the self-assembly of nanoscale heterojunctions
by spontaneous phase separation of the donor (polymer) and the
acceptor (fullerene). As a result of this spontaneous phase separ-
ation, charge-separating heterojunctions are formed throughout
the bulk of the material2.
Over the past decade, research has focused on regio-regular
poly(3-hexylthiophene) (P3HT) as the standard electron-donating
material in polymer BHJ solar cells, with important progress
having been made in understanding the device science and the
associated improvements in device efficiency. Relatively high-per-
formance polymer BHJ solar cells made from a mixture of P3HT
and [6,6]-phenyl C61 butyric acid methyl ester (PCBM) have
been reported, with maximum PCEs of he¼ 4–5% (refs 5–7).
Although approaches to improving the efficiency of P3HT/PCBM
cells are still being reported, the relatively large bandgap of P3HT
(1.9 eV) limits the fraction of the solar spectrum that can be har-
vested, and the relatively small energy difference between the top of
the p-band (highest occupied molecular orbital, HOMO) of P3HT
and the lowest unoccupied molecular orbital (LUMO) of the fuller-
ene acceptor results in a low open-circuit voltage, VOC 0.6 V.
These fundamental energies defined by the electronic structure of
the semiconducting polymer (the energy gap and the HOMO
energy) must be decreased in order achieve polymer BHJ solar
cells with PCEs of 6% and higher.
Recently, several classes of low-bandgap polymers have been
developed to better harvest the solar spectrum with deeper
HOMO energies that can potentially increase VOC (refs 8–12).
These polymers are designed to make use of internal charge transfer
from an electron-rich unit to an electron-deficient moiety within the
fundamental repeat unit. Among them, alternating co-polymers
based on poly(2,7-carbazole) derivatives11,12, with a suite of elec-
tron-deficient moieties to choose from, are particularly interesting
(see Fig. 1). The different electron-deficient moieties can be used
to tune the electronic energy gap of the semiconducting polymer,
while the deep HOMO of the carbazole leads to higher values for
VOC (ref. 11). The implied flexibility in the synthesis can lead to
both a smaller bandgap that enables the harvesting of a larger frac-
tion of the solar radiation spectrum, and a deeper HOMO energy
that increases the open circuit voltage of the photovoltaic device.
In their initial report of the synthesis and device performance of
PCDTBT, Leclerc and colleagues demonstrated a PCE of 3.6%
from a BHJ cell with VOC value approaching 0.9 V (ref. 12).
We report here solar cells with 6% PCE from BHJ composites
comprising PCDTBT/[6,6]-phenyl C71 butyric acid methyl ester
(PC70BM) with short-circuit current JSC¼ 10.6 mA cm22, open
circuit voltage VOC¼ 0.88 V and fill factor FF¼ 0.66 under air
mass 1.5 global (AM 1.5 G) irradiation of 100 mW cm22. The
internal quantum efficiency (IQE) is close to 100%, implying that
essentially every absorbed photon results in a separated pair of
charge carriers and that all photogenerated carriers are collected
at the electrodes.
Titanium oxide optical spacer and hole blocking layer
Historically, a relatively low PCE has been demonstrated in polymer
solar cells made from polymers that make use of the internal charge
transfer concept, including PCDTBT11,12. This low PCE has been
limited by the relatively low photocurrent obtained from these
devices. In BHJ cells, the photocurrent generation is governed by
two main factors13,14: (i) the fractional number of absorbed
photons in the active layer (relative to the total flux of photons
from the solar spectrum) and (ii) the IQE defined by the fraction
of collected carriers per absorbed photon. In principle, one can
simply increase the thickness of the active layer to absorb more
light. However, because of the relatively low carrier mobility of
the disordered materials (cast from solution with subsequent
phase separation), increasing the thickness increases the internal
resistance of the device. Consequently, the fill factor typically
1Center for Polymers and Organic Solids, University of California at Santa Barbara, Santa Barbara, California 93106, USA, 2Heeger Center for Advanced
Materials, Gwangju Institute of Science and Technology, Gwangju 500-712, South Korea, 3Department of Chemistry, University of Laval, G1K 7P4 Quebec
City, Quebec, Canada. *e-mail: klee@gist.ac.kr; ajhe@physics.ucsb.edu
ARTICLES
PUBLISHED ONLINE: 26 APRIL 2009 | DOI: 10.1038/NPHOTON.2009.69
NATURE PHOTONICS | VOL 3 | MAY 2009 | www.nature.com/naturephotonics 297

plummets as the thickness is increased. Based on this simple
analysis, we consider the following approach towards obtaining
higher photocurrent: maximizing the photon absorption for a
fixed active layer thickness while simultaneously improving the IQE.
To increase the photocurrent while keeping the thickness fixed,
we used an optical spacer between the photo-active layer and the
top electrode; because of the optical spacer, the maximum light
intensity is redistributed to be within the active charge separating
BHJ layer. The utility of the optical spacer has been reproduced in
recent publications15. In parallel, by choosing optimal conditions
for processing, we have demonstrated a nanoscale BHJ morphology
that results in nearly 100% IQE. This dual focused approach applied
to PCDTBT/PC70BM results in PCE, ne 6%; the highest value
reported to date for polymer BHJ solar cells.
Figure 1 shows the structure of the BHJ device together with the
molecular structures and an energy level diagram of the component
materials. From the fundamental physics of the open-circuit voltage
associated with the donor–acceptor heterojunction and the empiri-
cal relationship demonstrated in refs 16–18, the relatively deep
HOMO energy of PCDTBT, 5.5 eV, should result in a higher
open-circuit voltage. Moreover, because the spherical symmetry of
the fullerene has been lifted in PC70BM (compared to PCBM), the
PCDTBT/PC70BM BHJ material has higher absorption and, conse-
quently, enhanced photocurrent19,20.
The solution processible titanium sub-oxide (TiOx) layer
15,21,22
was introduced as an optical spacer21 and as a hole blocker23 (see
Supplementary Information) between the BHJ layer and the top
metal electrode. The TiOx layer redistributes the light intensity
within the BHJ by changing the optical interference between the
incident light and the light reflected from the metal electrode24,25.
As the active layer thickness decreases, the intensity of reflected
light increases, and the optical interference effect becomes more
pronounced. Hence, we expect that the efficacy of the TiOx layer
will be higher for thinner active layers. Hole blocking by the TiOx
is also more important for thinner-film devices. In the
PCDTBT:PC70BM solar cells reported here, the thickness of the
active layer is 80 nm, and the TiOx layer thickness is 10 nm.
With this configuration, we are able to make good use of the
optical spacer by avoiding destructive interference within the
charge separating layer between the incident light and the light
reflected from the aluminium–TiOx interface. In addition, the
bottom of the conduction band of TiOx matches the LUMO of
PC70BM. Finally, the relatively high electron mobility of PC70BM
and the hole-blocking feature of TiOx enable efficient electron col-
lection without a significant increase in the series resistance21–23.
Figure 2a shows the absorption spectra of PCDTBT solar cells with
and without the TiOx layer. The total absorption by the active layer
(including the doubled path length in the BHJ layer as a result of
reflection from the aluminium electrode) was measured in reflection
geometry as illustrated in the inset of Fig. 2a. Comparing two devices
with the same active layer thickness (80 nm), a substantial enhance-
ment in absorption is observed in the device with the TiOx layer.
Consequently, as shown in Fig. 2b, the device with the TiOx layer
demonstrates higher IPCE (incident photon-to-current efficiency)
throughout the visible range compared with the device without the
TiOx layer. Because the integration of the product of the IPCE with
the AM 1.5 solar spectrum is equal to the short-circuit current, the
higher short-circuit current (see Fig. 2c) of the device with TiOx is
consistent with the higher IPCE values.
Nanoscale morphology and photocurrent generation
Note, however, that the photocurrent is determined by the product
of the total number of absorbed photons within the solar spectrum
and the IQE of the device26–28. The IQE is determined by a three-
step process26–28: (i) migration/diffusion of the photogenerated exci-
tations to the PCDTBT/PC70BM interface; (ii) exciton dissociation
and charge separation at the interface; and (iii) collection of charge
carriers at the ITO and aluminium electrodes. Because of step (i),
the nanoscale phase separation in BHJ materials must be less than
20 nm because the exciton diffusion length is generally less than
ITO
Glass
PCDTBT:PC70BM
PEDOT:PSS
Al
PCDTBT
PC70BM
Electron
Hole
TiOx
4.7 eV
8.0 eV
4.3 eV
6.0 eV
5.5 eV
3.6 eV
ITO
Al
PC70BM
PCDTBT
TiO
x
4.3 eV
PED
O
T:PSS
5.0 eV
S
S
C
8
H
17
OMe
O
N
n
C
8
H
17
S
N
N
Figure 1 | Device structure and energy level diagram of the components. a, The bulk heterojunction (BHJ) film is a phase separated blend of PCDTBTand
PC70BM. The inset shows the transfer of photogenerated electrons from PCDTBT to PC70BM. The titanium oxide (TiOx) layer is introduced as an optical
spacer on top of the BHJ layer. b, Energy level diagram of the components of the device.
ARTICLES NATURE PHOTONICS DOI: 10.1038/NPHOTON.2009.69
NATURE PHOTONICS | VOL 3 | MAY 2009 | www.nature.com/naturephotonics298

10 nm (refs 29,30). In addition, smaller-scale phase separation
creates larger-area PCDTBT:PC70BM interfaces where charge sep-
aration can take place. For step (ii), a sufficiently large energy differ-
ence between the PCDTBT LUMO and the PC70BM LUMO is
required for ultrafast photoinduced electron transfer. For step (iii),
both the PCDTBT and PC70BM phases must form percolated net-
works with few charge-trapping sites or ‘dead ends’. The ultrafast
photo-induced charge transfer (,100 fs) at the polymer–fullerene
interface ensures that the charge separation efficiency approaches
100% (refs 2,31,32). Hence, the molecular organization and mor-
phology on the nanometre scale as described by steps (i) and (ii),
and the nanoscale morphology and the interface between the
cathode and the TiOx/BHJ as described in step (iii), provide the
route to high IQE.
The nanoscale morphology of the PCDTBT:PC70BM BHJ is
strongly affected by processing parameters such as choice of sol-
vents, blend ratio of PCDTBT to PC70BM, solution concentration,
thermal annealing, and the molecular structure of the component
materials33–35. In our experiments, thermal annealing of the
PCDTBT:PC70BM system at high temperatures reduced the FF,
JSC and VOC (see Supplementary Information). Thus, the use of
thermal annealing is eliminated as a strategy for improving device
performance. However, it is well known that for polymer-based
solar cells, performance is strongly affected by both the solvent
and the blend ratio33–36. This is expected because the solvent is
known to affect the BHJ domain size, and the donor/acceptor
blend ratio determines the formation of percolated networks.
Figure 3a–c shows defocused37,38 transmission electron micro-
scope (TEM) images of PCDTBT:PC70BM (1:4 ratio) films
dissolved in chloroform (CF), chlorobenzene (CB) and dichloroben-
zene (DCB), respectively. Although large dark clusters (200 and
300 nm) are observed in the CF and CB films, clearly defined nanos-
cale phase separation is observed in the film cast from DCB. These
features are also observed in the surface phase images measured by
atomic force microscopy (AFM), as shown in the insets of the TEM
images. The higher electron density of PC70BM compared with
PCDTBT causes electrons to be scattered more efficiently by the
PC70BM from the TEM beam. Thus, the darker regions in
the TEM images are regions of phase-separated PC70BM. Because
the exciton diffusion length (,10 nm) is much smaller than the
200–300 nm features seen in Fig. 3a,b, photo-generated excitons
will often recombine before reaching the interfaces in films cast
from solution in CB or CF, causing reduced charge carrier gener-
ation at the interfaces and a concomitant loss of photocurrent.
Figure 3d shows the IPCE spectra of solar cells comprising BHJ
films cast from CF, CB, DCB and from a mixture of CB and DCB.
The cell fabricated with a BHJ film cast from DCB has a higher IPCE
over the entire excitation spectrum compared to devices comprising
films cast from either CF or CB. Processing from a mixture of DCB
and CB also increases the IPCE compared to processing from pure
CB. Increasing the amount of DCB in the CB/DCB mixture
increases the contribution from PC70BM to the IPCE, as is
evident from the pronounced peaks around 400 and 450 nm.
Figure 3a–c demonstrates that DCB results in significantly
smaller nanoscale phase separation. Therefore, the increased IPCE
shown in Fig. 3d and obtained from devices made with films cast
from pure DCB or from mixtures of CB and DCB results from
the nanoscale phase separation. The enhanced JSC and higher FF
(see Fig. 3e) imply the formation of well-connected percolated net-
works for each of the phase-separated components (donor and
acceptor). Thus, using DCB evidently also leads to better-connected
percolated networks, which, in combination with the nanoscale
phase separation, improve the device performance.
Obviously, the connectivity is sensitive to the blend ratio of
PCDTBT to PC70BM. Figure 4a–d shows TEM images of
PCDTBT:PC70BM films cast from DCB with increasing amounts
of PC70BM. As the amount of PC70BM progressively increases,
the nanoscale phase separation can be seen more clearly, with
gradual emergence of a ‘fibrillar’ PCDTBT nanostructure. This
fibrillar PCDTBT nanostructure is most pronounced in films at
the 1:4 blend ratio, implying that increasing the amount of
PC70BM causes the PCDTBT network to form longer and better
connected pathways.
80
Without TiOx
With TiOx
TiOx
AIITO
Light
Active layer
70
60
50
40
30
20
10
0
80
70
60
50
40
30
20
10
0
400 450 700650600550500
Wavelength (nm)
400 450 700650600550500
Wavelength (nm)
0.0
−12
−8
−4
0
0.2 1.00.80.60.4
Voltage (V)
A
bs
or
pt
io
n
(%
)
IP
C
E
(%
)
C
ur
re
nt
d
en
si
ty
(
m
A
c
m
−2
)
Figure 2 | The effects of TiOx layer as an optical spacer on device
performance. a, Total absorption in the active layer measured in a reflection
geometry with the TiOx layer (red symbols) and without the TiOx layer
(black symbols). The inset shows a schematic of the device structure.
b, Incident photon-to-current efficiency (IPCE) spectra with the TiOx layer
(red symbols) and without the TiOx layer (black symbols). c, Current density
versus voltage characteristics (J–V) of the device with the TiOx layer (red
symbols) and without the TiOx layer (black symbols).
NATURE PHOTONICS DOI: 10.1038/NPHOTON.2009.69 ARTICLES
NATURE PHOTONICS | VOL 3 | MAY 2009 | www.nature.com/naturephotonics 299

Internal quantum efficiency
The morphology observed in Fig. 4d, using the 1:4 ratio of
PCDTBT to PC70BM, is ideal for polymer solar cell performance,
as is made evident by the increased IQE. However, 80% of the
film in terms of mass is now made of PC70BM, which has only
weak absorption in the visible spectral range. This is evident
from the plot of the absorption coefficient shown in Fig. 4e.
The decreased absorption coefficient for the 1:4 film reduces
the number of absorbed photons in the active layer for fixed
film thickness.
To determine the optimum blend ratio, we used the IPCE and
J–V characteristics of the solar cells. Figure 4f shows the IPCE
spectra of solar cells with various blend ratios. Although the IPCE
curve obtained from the 1:1 device shows a poor photo-response,
80
CB : DCB (1:1)
CB : DCB (1:3)
CB : DCB (3:1)
DCB DCB
CF
CF
CB
CB
70
60
50
40
30
20
10
0
Wavelength (nm)
IP
C
E
(%
)
C
ur
re
nt
d
en
si
ty
(
m
A
c
m
−2
)
300 400 500 600 700 800 0.0 0.2 0.4 0.6 0.8 1.0
−12
−10
−8
−6
−4
−2
0
2
Voltage (V)
0 1 μm
500 nm CF
1 μm0
CB500 nm
1 μm0
DCB200 nm
Figure 3 | The effects of CF, CB and DCB solvents on film morphology and device performance. a–c, TEM images of PCDTBT:PC70BM films spin-cast from
CF (a), CB (b) and DCB (c) solvents. The insets show the surface phase images measured by atomic force microscopy (AFM). d,e, IPCE spectra (d) and J–V
characteristics (e) of the devices fabricated with films cast from CF, CB and DCB.
300
0.0
3.0 × 104
6.0 × 104
9.0 × 104
1.2 × 105
1.5 × 105
1.8 × 105
2.1 × 105
400 500 600 700 800
Wavelength (nm)
A
bs
or
pt
io
n
co
effi
ci
en
t (
cm
−1
)
1:1
1:2
1:3
1:4
PCDTBT:PC70BM PCDTBT:PC70BM PCDTBT:PC70BM
1:1
1:2
1:3
1:4
1:1
1:2
1:3
1:4
300
−12
−10
−8
−6
−4
−2
0
0
10
20
30
40
50
60
70
80
2
400 500 600 700 800 0.0 0.2 0.4 0.6 0.8 1.0
Voltage (V)
C
ur
re
nt
d
en
si
ty
(
m
A
c
m
− 2
)
IP
C
E
(%
)
Wavelength (nm)
200 nm 200 nm 200 nm 200 nm1:1 1:2 1:41:3
Figure 4 | The effect of blending ratio on film morphology and device performance. a–d, TEM images of the PCDTBT:PC70BM blend films spin-cast from
DCB with increasing amounts of PC70BM: blending ratios 1:1 (a), 1:2 (b), 1:3 (c) and 1:4 (d). e, Absorption coefficients of the films with blend ratios of 1:1, 1:2,
1:3 and 1:4. f, IPCE spectra for the same films as in e. g, J–V characteristics of the devices fabricated using BHJ films with blend ratios of 1:1, 1:2, 1:3 and 1:4.
ARTICLES NATURE PHOTONICS DOI: 10.1038/NPHOTON.2009.69
NATURE PHOTONICS | VOL 3 | MAY 2009 | www.nature.com/naturephotonics300

than 6%. The highest value that we obtained in our laboratory was
6.2% with no antireflective coating on the glass. One of the
PCDTBT/PC70BM solar cells with an initial efficiency of 6.1%
(without antireflective coating) was sent to NREL for certification,
after monitoring and ensuring the device performance for one
week. NREL returned the device with a certified efficiency of
5.96%, as shown in Fig. 6b. Given the significant time lag between
our sending the device to NREL and the measurement at NREL,
we naturally expect some degradation. In fact, lifetime data gathered
in our laboratory show that the device had already degraded some-
what by the time the NREL measurement was performed (see
Supplementary Information).
Conclusion
In conclusion, we have successfully demonstrated high-efficiency
BHJ solar cells comprising PCDTBT and PC70BM (1:4 ratio) as
the active charge separating layer. With PCDTBT/PC70BM,
neither thermal annealing6,36 nor the addition of processing addi-
tives9,41 are required for achieving high efficiency. The
PCDTBT/PC70BM solar cells exhibit he¼ 6% under AM 1.5
irradiation, the highest certified value reported to date. More
important, the IQE of the PCDTB/PC70BM solar cells approaches
100%, implying that every photon absorbed leads to a separated
pair of charge carriers and that every photogenerated mobile
carrier is collected at the electrodes.
Methods
Device fabrication. Solar cells were fabricated on an indium tin oxide (ITO)-coated
glass substrate with the following structure: ITO-coated glass substrate/poly(3,4-
ethylenedioxythiophene) (PEDOT:PSS)/PCDTBT:PC70BM/TiOx/Al. The ITO-
coated glass substrate was first cleaned with detergent, ultrasonicated in acetone and
isopropyl alcohol, and subsequently dried overnight in an oven. PEDOT:PSS
(Baytron PH) was spin-cast from aqueous solution to form a film of 40 nm
thickness. The substrate was dried for 10 min at 140 8C in air and then transferred
into a glove box to spin-cast the charge separation layer. A solution containing a
mixture of PCDTBT:PC70BM (1:4) in dichlorobenzene solvent with a concentration
of 7 mg/ml was then spin-cast on top of the PEDOT/PSS layer. The film was dried
for 60 min at 70 8C in the glove box. The TiOx precursor solution diluted 1:200 in
methanol was spin-cast in air on top of the PCDTBT:PC70BM layer (5,000 rpm for
40 s). The sample was heated at 80 8C for 10 min in air. Then, an aluminium
(Al, 100 nm) electrode was deposited by thermal evaporation in a vacuum of about
5 1027 torr. Current density–voltage (J–V) characteristics of the devices were
measured using a Keithley 236 Source Measure Unit. Solar cell performance used an
Air Mass 1.5 Global (AM 1.5 G) solar simulator with an irradiation intensity of
1,000 W m22. An aperture (12.7 mm2) was used on top of the cell to eliminate
extrinsic effects such as crosstalk, waveguiding, shadow effects and so on. The
spectral mismatch factor was calculated by comparison of the solar simulator
spectrum and the AM 1.5 spectrum at room temperature.
Measurement system. Our measurement system yielded data in precise agreement
with measurements made at NREL. Results for cells returned to us after NREL
measurement had expected values. Our integrated IPCE values always agreed with
the measured short-circuit current to within a few percent.
TEM microscopy. Specimens were prepared by first casting a PCDTBT:PC70BM
blend thin film on glass. The films were baked at 70 8C for 1 h, and then removed
from the nitrogen environment and scored with a diamond scribe to define the
sample size. The substrate and film were immersed in deionized water for 20 min
and sonicated to promote delamination. Resulting pieces of the film were transferred
to a PELCO copper TEM grid with a carbon/Formvar support grid. TEM specimens
were allowed to dry under low heat to remove excess water from the transfer process.
Light field imaging was performed in an FEI T20 TEM using proper defocus for
additional phase contrast from the relatively amorphous polymer material.
Received 19 December 2008; accepted 24 March 2009;
published online 26 April 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. Yu, G., Gao, J., Hemmelen, J. C., Wudl, F. & Heeger, A. J. Polymer photovoltaic
cells: enhanced efficiencies via a network of internal donor–acceptor
heterojunctions. Science 270, 1789–1791 (1995).
3. Halls, J. J. et al. Efficient photodiodes from interpenetrating polymer networks.
Nature 376, 498–500 (1995).
4. Kim, J. et al. Efficient tandem polymer solar cells fabricated by all-solution
processing. Science 317, 222–225 (2007).
5. Li, G. et al. High-efficiency solution processible polymer photovoltaic cells by
self-organization of polymer blends. Nature Mater. 4, 864–868 (2005).
6. Ma, W. Yang, C., 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).
7. Kim, Y. et al. A strong regioregularity effect in self-organizing conjugated
polymer films and high-efficiency polythiophene:fullerene solar cells. Nature
Mater. 5, 197–203 (2006).
8. Muhlbacher, D. et al. High photovoltaic performance of a low-bandgap polymer.
Adv. Mater. 18, 2884–2889 (2006).
9. Peet, J. et al. Efficiency enhancement in low-bandgap polymer solar cells by
processing with alkane dithiols. Nature Mater. 6, 497–500 (2007).
10. Wang, E. et al. High-performance polymer heterojunction solar cells of a
polysilafluorene derivative. Appl. Phys. Lett. 92, 033307–033310 (2008).
11. Blouin, N. et al. Toward a rational design of poly(2,7-carbazole) derivatives for
solar cells. J. Am. Chem. Soc. 130, 732–742 (2008).
12. 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).
13. Hoppe, H. & Sariciftci, N. S. Organic solar cells: An overview. J. Mater. Res. 19,
1924–1945 (2004).
14. Winder, C. & Sariciftci, N. S. Low bandgap polymers for photon harvesting in
bulk heterojunction solar cells. J. Mater. Chem. 14, 1077–1086 (2004).
15. Hayakawa, A., Yoshikawa, O., Fujieda, T., Uehara, K. & Yoshikawa, S. High
performance polythiophene/fullerene bulk-heterojunction solar cell with a TiOx
hole blocking layer. Appl. Phys. Lett. 90, 163517 (2007).
16. 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).
17. Brabec, C. J., Sariciftci, N. S. & Hummenlen, J. C. Plastic solar cells. Adv. Mater.
11, 15–26 (2001).
18. Brabec, C. J. et al. Origin of the open circuit voltage of plastic solar cells.
Adv. Funct. Mater. 11, 374–380 (2001).
19. Arbogast, J. W. & Foote, C. S. Photophysical properties of C70. J. Am. Chem. Soc.
113, 8886–8889 (1991).
20. Wienk, M. M. et al. Efficient methano[70]fullerene/MDMO-PPV bulk
heterojunction photovoltaic cells. Angew Chem. Int. Ed. 42, 3371–3375 (2003).
21. Kim, J. Y. et al. New architecture for high-efficiency polymer photovoltaic cells
using solution-based titanium oxide as an optical spacer. Adv. Mater. 18,
572–576 (2006).
22. Lee, K. et al. Air-stable polymer electronic devices. Adv. Mater. 19,
2445–2449 (2007).
23. Cho, S. et al. Multilayer bipolar field-effect transistors. Appl. Phys. Lett. 92,
063505 (2008).
24. Persson, N.-K. & Inganas, O. in Organic Photovoltaics Ch. 5
(Taylor & Francis, 2005).
25. Slooff, L. H. et al. Determining the internal quantum efficiency of highly
efficient polymer solar cells through optical modeling. Appl. Phys. Lett. 90,
1435061 (2007).
26. Mayer, A. C., Scully, S. R., Hardin, B. E., Rowell, M. W. & McGehee, M. D.
Polymer-based solar cells. Mater. Today 10, 28–33 (2007).
27. Schilinsky, R., Waldauf, C. & Brabec, C. J. Recombination and loss analysis in
polythiophene based bulk heterojunction photodetectors. Appl. Phys. Lett. 81,
3885–3887 (2002).
28. Dennler, G. et al. Angle dependence of external and internal quantum
efficiencies in bulk-heterojunction organic solar cells. J. Appl. Phys. 102,
0545161 (2006).
29. Vacar, D., Maniloff, E. S., McBranch, D. W. & Heeger, A. J. Charge-transfer
range for photoexcitations in conjugated polymer/fullerene bilayers and blends.
Phys. Rev. B 56, 4573–4577 (1997).
30. Shaw, P. E., Ruseckas, A. & Samuel, I. D. W. Exciton diffusion measurements in
poly(3-hexylthiophene). Adv. Mater. 20, 3516–3520 (2008).
31. Kraabel, B. et al. Subpicosecond photoinduced electron transfer from conjugated
polymers to functionalized fullerenes. J. Chem. Phys. 104, 4267–4273 (1996).
32. Brabec, C. J. et al. Transient photoinduced electron transfer process in
conjugated polymer/fullerene bulk heterojunction in real time. Chem. Phys. Lett.
340, 232–236 (2001).
33. Yao, Y. Hou, J., Xu, Z., Li, G. & Yang, Y. Effects of solvent mixtures on the
nanoscale phase separation in polymer solar cells. Adv. Funct. Mater. 18,
1783–1789 (2008).
34. Gunes, S., Neugebauer, H. & Sariciftci, N. S. Conjugated polymer-based organic
solar cells. Chem. Rev. 107, 1324–1338 (2007).
35. Zhang, F. 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).
ARTICLES NATURE PHOTONICS DOI: 10.1038/NPHOTON.2009.69
NATURE PHOTONICS | VOL 3 | MAY 2009 | www.nature.com/naturephotonics302

36. Hoppe, H. & Sariciftci, N. S. Morphology of polymer/fullerene bulk
heterojunction solar cells. J. Mater. Chem. 16, 45–61 (2006).
37. Ma, W., Yang, C. & Heeger, A. J. Spatial Fourier-transform analysis of the
morphology of bulk heterojunction materials used in plastic solar cells. Adv.
Mater. 19, 1387–1390 (2007).
38. Moon, J. S., Lee, J. K., Cho, S., Byun, J. & Heeger, A. J. Columnlike structure of
the cross-sectional morphology of bulk heterojunction materials. Nano Lett. 9,
230–234 (2009).
39. Shrotriya, V. et al. Accurate measurement and characterization of organic solar
cells. Adv. Funct. Mater. 16, 2016–2023 (2006).
40. Metzdorf, J., Winter, S. & Wittchen, T. Radiometry in photovoltaics: calibration
of reference solar cells and evaluation of reference values. Metrologia 37,
573–578 (2000).
41. Lee, J. K. et al. Processing additives for improved efficiency from bulk
heterojunction solar cells. J. Am. Chem. Soc. 130, 3619–3623 (2008).
Acknowledgements
The research leading to the results reported here was supported by the Air Force Office of
Scientific Research, the Department of Energy and by a grant from the US Army CERDC.
The TiOx development work was carried out at the Heeger Center for Advanced Materials
(Gwangju Institute of Science and Technology (GIST) and UCSB) with support from under
the Global Research Laboratory (GRL) Program sponsored by the Korean Government.
The authors thank C. Brabec and R. Gaudiana for advice and encouragement, and for
supplying the PC70BM. The measurements at NREL were carried out by P. Ciszek and
K. Emery. We thank them for their help and cooperation.
Additional information
Supplementary information accompanies this paper at www.nature.com/naturephotonics.
The authors declare competing financial interests: details accompany the full-text HTML
version of the paper at www.nature.com/naturephotonics. Reprints and permission
information is available online at http://npg.nature.com/reprintsandpermissions/.
Correspondence and requests for materials should be addressed to K.L. and A.J.H.
NATURE PHOTONICS DOI: 10.1038/NPHOTON.2009.69 ARTICLES
NATURE PHOTONICS | VOL 3 | MAY 2009 | www.nature.com/naturephotonics 303