Journal of

Materials Chemistry A
View Article Online

Published on 25 July 2014. Downloaded by Sungkyunkwan University on 10/11/2014 03:08:56.

PAPER

Cite this: J. Mater. Chem. A, 2014, 2,

15897

View Journal | View Issue

Planar heterojunction perovskite/PC71BM solar

cells with enhanced open-circuit voltage via a
(2/1)-step spin-coating process
Chien-Hung Chiang,b Zong-Liang Tsengb and Chun-Guey Wu*ab

Received 17th July 2014

Accepted 23rd July 2014

A two-step, spin-coating process to synthesize a high quality perovskite lm at room temperature at

ambient atmosphere is reported. Combining an optimized PEDOT:PSS hole-transport layer and a
PC71BM acceptor, the device achieved a power conversion eciency of 16.31% with a remarkably high

DOI: 10.1039/c4ta03674c
www.rsc.org/MaterialsA

Voc of 1.05 V and FF of 0.78. The eciency reproducibility of the devices is also high: 80% of devices
achieve eciency higher than 15%.

Perovskite, named aer Russian mineralogist Lev Perovski, is a

type of mineral that was discovered two centuries ago. It was
known as a group of material with interesting magnetic and
electrical properties. In 2009, Miyasaka et al.1 discovered the
photovoltaic performance of organic-perovskites (such as
methylammonium lead halide, CH3NH3PbI3) when they used
very small CH3NH3PbI3 nanoparticles to sensitize a mesoporous
TiO2 anode. Combined with a liquid electrolyte, the resulting
device has a PCE of 34%. Four years later (2013), based on a
similar cell architecture with a solid electrolyte, Gr
atzel et al.2
and Snaith et al.3 achieved the remarkable eciency of above
15%. In addition to the phenomenal photovoltaic eciency,
perovskite materials also have the advantages of low cost, high
absorption coecient, excellent carrier transport, tunable
composition and structure as well as the ability to be fabricated
by various processing methods.4,5 Several reports also suggest
that solar cells with eciencies of up to 20% are realistically
achievable.6,7 The solution-processable perovskite photovoltaic
devices should be able to compete with conventional thin-lm
solar cells both in terms of cost and eciency. Therefore, aer
Gr
atzel's report,2 research on organic-perovskite based solar cell
has increased.
Organicinorganic perovskites were rst implemented in
mesoscopic dye-sensitized solar cells (DSCs).1 However, mesoscopic DSC fabrication requires a processing temperature of up
to 500
C to sinter the TiO2 support, which renders them
incompatible with exible substrates. Later studies814 also

Department of Chemistry, National Central University and Research Center for New
Generation Photovoltaics, National Central University, Jhong-Li, Taiwan 32001,
ROC. E-mail: t610002@cc.ncu.edu.tw; Fax: +886-3-4227664

Research Center for New Generation Photovoltaics, National Central University,

Jhong-Li, Taiwan 32001, ROC. Fax: +886-3-4227664
Electronic supplementary information (ESI) available: Supplementary data are
provided in the electronic supporting information. This material is available
online. See DOI: 10.1039/c4ta03674c

This journal is The Royal Society of Chemistry 2014

found that perovskite lm can also be applied to planar heterojunction solar cells, which are adaptable to low temperature
solution processes. Several lm deposition methods were used
to fabricate perovskite lms in planar heterojunction solar cells
such as co-evaporation of two precursors3 (PbCl2 and CH3NH3I
(MAI)), one-step solution process14 (mixing lead halide and
CH3NH3I in a solvent before deposition), vapor-assisted solution process,15 or two-step dipping process.2
The application of perovskites in organic (small molecule or
polymer) solar cells (which professor Snaith named as an
inverted cell architecture) started in 2013. Jeng et al.16 used
perovskite as a donor and PC61BM as an acceptor to construct a
perovskite/fullerene planar heterojunction hybrid solar cell
with the eciency of 3.9%. The low photovoltaic performance
may be due to the fact that the perovskite phase was not pure, as
revealed by the XRD pattern provided in the paper. Using a
similar cell architecture but changing the ratio and concentration of PbI2 and MAI, Lam et al.17 pushed the eciency of the
device up to 7.4%. Snaith et al.18 introduced a thin compact TiOx
lm between PC61BM and Al electrode and used CH3NH3PbI3xClx as an absorber to make a device with an eciency of
up to 9.8%, in 2014. The same year, Yang et al.19 used the same
absorber CH3NH3PbI3xClx to fabricate a device with 11.5%
eciency and a very high ll factor of 0.72. Gr
atzel et al.20 used a
well-designed perovskite cell architecture to achieve an eciency of 12.04%. They co-evaporated PbI2 and MAI to prepare
the perovskite lm, inserted an electron blocking (polyTPD)
layer in between the perovskite and PEDOT:PSS, and used Au as
an electrode. Note that the resultant cell had an impressive high
open-circuit voltage of 1.05 V. Nevertheless, when the active area
of the device was enlarged from 0.09 cm2 to 0.89 cm2, the eciency decreased from 12.04% to 8.27% due to a signicant
decrease in the FF value (from 0.67 to 0.52). Hence, improving
the FF in perovskite-based cells is very critical for obtaining
high eciency cells when the active area is enlarged.

J. Mater. Chem. A, 2014, 2, 1589715903 | 15897

View Article Online

Published on 25 July 2014. Downloaded by Sungkyunkwan University on 10/11/2014 03:08:56.

Journal of Materials Chemistry A

Planar heterojunction perovskite solar cells have the

advantage of a simple architecture and thus can be fabricated
easily.21 All organic-perovskite lms used in OPV-typed (inverted) solar cells were fabricated either using a one-step solution
process17,18 or the co-evaporation method.20 Very recently, Xiao
et al.22 used an interdiusion method to prepare organicperovskite lms, resulting in high eciency and high device
yields. In this paper, we present a very similar method called a
two-step spin coating process that is related to Xiao's method
(actually our paper was nearly ready when we noticed Xiao's
paper) to fabricate highly crystalline CH3NH3PbI3 lms.
Furthermore, we optimized the quality of the PEDOT:PSS layer
such that a highly crystalline CH3NH3PbI3 lm formed at room
temperature. In contrast, in Xiao's paper, they had heated the
PbI2 and MAI layers to form a perovskite by the interdiusion
between two components. Using our method, the reaction
between PbI2 and CH3NH3I was well-controlled by the thickness
of PbI2 and the process of adding CH3NH3I. Therefore the
preparation of the perovskite lm can be carried out in an
ambient atmosphere at room temperature. This method has
great potential to be used in assembling cost-eective, exible
solar cells. The optimized device under AM 1.5 (100 mW cm2)
radiation achieved a remarkably high conversion eciency of
16.31% with a remarkably high Voc of 1.05 V. Device fabrication
is also very reproducible: ca. 80% of devices have eciency
higher than 15% (see Fig. S1 of ESI).
The fabrication steps and architecture of the inverted-type
perovskite solar cell reported in this paper are shown in Fig. 1.
Detailed preparation procedures and materials used can be
found in the Experimental section. The cell is the simplest
bilayer planar hetero-junction structure generally used in
organic small molecule solar cells (SMSCs). In this device,
instead of organic small molecules, the perovskite is used as the
light absorber and hole transport layer, whereas PCBM is the
electron transport layer. However, there are four inorganic/

Fig. 1

Paper

organic interphases in this simple inverted perovskite cell

(vs. two interphases in a typical SMSC). Each interphase should
be carefully engineered to obtain a highly ecient device. For
example, we found that a high quality PEDOT:PSS dense lm
can be obtained when the ITO substrate is preheated to 120
C
for 5 min. On the other hand, PEDOT:PSS lm deposited on
cold ITO is rough and has some defects (see Fig. S2, ESI). A
smooth PEDOT:PSS lm is essential for fabricating highly
crystalline/good adhesion CH3NH3PbI3 lm without defects at
room temperature, as revealed by the SEM micrographs also
displayed in Fig. S2, ESI. Therefore, all devices discussed in
this article were fabricated by depositing PEDOT:PSS lms on
preheated ITO substrates via spin-coating.
In the two-step spin-coating process, various thicknesses of
PbI2 lms were rst deposited on the PEDOT:PSS surface using
a PbI2/DMF solution and by varying the spin rate. Nevertheless,
when the spin rate is reduced to 1000 rpm, the morphology of
the resulting 250 nm PbI2 lm is very rough and also cannot
adhere well to the substrate. Therefore, high quality PbI2 lms
with a thickness of 165 nm (spin rate 1500 rpm) were used for
the next fabrication step. To prepare the CH3NH3PbI3 lm, a
MAI/isopropanol (IPA) solution with a concentration range from
10 mg mL1 to 50 mg mL1 was spin-coated on top of the PbI2
lm at a xed spin rate of 1500 rpm for 30 seconds, and then a
layer PCBM was deposited via spin coating at ambient atmosphere. One important advantage of fabricating perovskite lms
using a two-step spin-coating method is that the fabrication of
PEDOT:PSS/perovskite/PC61BM can be carried out in an
ambient atmosphere at room temperature. Aer employing Ca/
Al electrodes prepared using vacuum thermal evaporation, the
photovoltaic parameters of the resulting cells were measured,
and the results are summarized in Table 1.
Data in Table 1 reveal that the eciency of perovskite solar
cell is very sensitive to the concentration of the MAI/IPA solution. The concentration of MAI not only aects the perovskite

Preparation steps and architecture of inverted perovskite/PCBM solar cells (SC: Spin-coating, TE: thermal evaporation).

15898 | J. Mater. Chem. A, 2014, 2, 1589715903

This journal is The Royal Society of Chemistry 2014

View Article Online

Paper

Journal of Materials Chemistry A

Published on 25 July 2014. Downloaded by Sungkyunkwan University on 10/11/2014 03:08:56.

Table 1 Device parameters for solar cells based on perovskite lms

prepared with varying MAI concentrationsa

MAI Conc. (mg mL1)

Jsc (mA cm2)

Voc (V)

FF

h (%)

20
30
32
34
36
38
40
50

8.60
8.37
9.94
11.89
13.56
14.59
13.65
9.45

0.27
0.51
0.80
0.85
0.90
0.91
0.90
0.78

0.27
0.53
0.70
0.70
0.72
0.74
0.71
0.66

0.81
2.34
5.65
7.12
8.90
9.92
8.86
5.02

a
Device
architecture:
ITO/PEDOT:PSS/perovskite/PC61BM/Ca/Al.
Thicknesses of PbI2, PEDOT:PSS, and PC61BM layers are 165 nm,
40 nm, and 140 nm, respectively.

formation rate and lm homogeneity but also the purity and

crystallinity of the resulting CH3NH3PbI3 lm, which determines the eciency of the corresponding solar cell. However, in
a one-step process reported previously, PbI2 and MAI with
dierent stoichiometries were mixed rst and then spin-coated
on the substrate. The formation rate and purity of CH3NH3PbI3
depends on the mole ratio of PbI2 and MAI in the solution. It is
dicult to control the formation rate (which aects the lm
morphology) and stoichiometry of the perovskite separately.
Moreover, the ratio of PbI2 and MAI and the reaction rate
between the two species can be manipulated individually in a
two-step process reported in this paper. This is another very
important advantage of the two-step process. Furthermore, the
two-step method (spin-coating of PbI2 lm followed by dipping
the lm in a MAI solution) has been used to fabricate regular
DSC type perovskite solar cells.2 However, the dipping process
may damage the PEDOT:PSS/PbI2 interface as the resulting cell
was found to have a low ll factor (unpublished result).
To understand why the concentration of MAI/IPA solution
aects the eciency of the resulting perovskite solar cells, X-ray
diraction patterns of the perovskite lms on PEDOT:PSS were
obtained and are shown in Fig. 2 (the original 2D data are
provided in Fig. S3 of ESI). Using a lower (36 mg mL1) MAI

X-ray diraction patterns of perovskite lms prepared using

varying MAI concentrations and related compounds (MAI and PbI2).

Fig. 2

This journal is The Royal Society of Chemistry 2014

concentration, the reaction between PbI2 and MAI is not

complete, as unreacted PbI2 was detected. When using a higher
(40 mg mL1) MAI concentration, although no PbI2 is observed,
a new phase with a diraction peak at 2q of 10.9
is present. We
do not know exactly the structure and components of this new
phase; perhaps it is a CH3NH3PbI3-intercalated compound. The
pure CH3NH3PbI3 phase was observed only when the concentration of MAI/IPA solution is ca. 38 mg mL1. The peaks
appearing at 2q of 13.7
, 19.5
, 24.0
, 27.9, 31.4
, 40.2
, 42.7
,
and 49.9
indicate the formation of the perovskite structure.23,24
Interestingly, most reported XRD patterns of CH3NH3PbI3 lms
only exhibited three low angle (<35
) diraction peaks at 2q of
13.9
, 28.3
and 31.7
, corresponding to the (110), (220) and
(310) planes of the tetragonal perovskite structure. Nevertheless, the XRD pattern of CH3NH3PbI3 prepared using our twostep method exhibited two extra peaks at 19.5
(210) and 24.0

(202). These two diraction peaks were observed in the simulated perovskite XRD pattern reported by Gr
atzel et al.20 A higher
number of diraction peaks indicates that the CH3NH3PbI3
lms prepared are highly crystalline. Furthermore, the
concentration of MAI also aects the morphology of the
perovskite lm as revealed by the SEM micrographs in Fig. S4 of
ESI. A smooth, dense perovskite lm was observed when the
concentration of MAI is close to 38 mg mL1 with lower or
higher MAI concentrations resulting in a rough lm with some
defects.
PC71BM (has good electron mobility25,26) was used to replace
PC61BM, which is the most commonly used acceptor in perovskite solar cells, as an electron transporting material, and the
eciency of the resulting device increases from 9.92% to
13.29%, mainly due to an increase in the short-circuit current
(Jsc, see Table 2). A signicant increase in the Jsc without an
obvious change in the open-circuit voltage (Voc) and ll factor
(FF) suggests that the electron transport rate of the acceptor
layer has a tremendous eect on the Jsc and therefore the
conversion eciency (h) of the inverted perovskite solar cell. To
further optimize the device, the PC71BM layer was solvent
annealed (the detailed process was described in the Experimental section) at room temperature and the photovoltaic data
of the resulting device are also provided in Table 2. Upon
solvent annealing of the PC71BM layer for 24 h, the eciency of
the perovskite cell increases further to 16.31%, the highest
eciency for an OPV-type perovskite solar cell reported today
with a signicant improvement in both the photocurrent (from
18.13 mA cm2 to 19.98 mA cm2) and Voc (from 0.93 to 1.05 V).
Extending the solvent annealing time did not further improve
the eciency.
The IV and IPCE curves of the champion (highest eciency)
cell (h 16.31%) are displayed in Fig. 3. The IPCE curve shows a
strong spectral response in the range from 350 nm to 750 nm
with an eciency of over 80% for all the wavelengths. The
integrated current (19.20 mA cm2) from the IPCE curve is very
close to the Jsc (19.98 mA cm2) obtained from the IV curve. It is
notable that the champion cell has a good FF of 0.78 and
remarkably high Voc of 1.05 V, which is the highest among all
perovskite solar cells reported in the literature.27,28 The Voc of the
champion device is much higher than that of Xiao's device,

J. Mater. Chem. A, 2014, 2, 1589715903 | 15899

View Article Online

Journal of Materials Chemistry A

Table 2

Paper

Photovoltaic parameters for solar cells based on varying acceptors and acceptor solvent annealing timea

Acceptor

Solvent annealing time

(hour)

Jsc (mA cm2)

Voc (V)

FF

h (%)

Rsb (U cm2)

PC61BM
PC71BM
PC71BM

0
0
24

14.59
18.13
19.98

0.91
0.95
1.05

0.74
0.76
0.78

9.92
13.29
16.31

5.60
5.41
4.41

Published on 25 July 2014. Downloaded by Sungkyunkwan University on 10/11/2014 03:08:56.

Device architecture: ITO/PEDOT:PSS/perovskite/PCBM. Thickness of perovskite: 360 nm; thickness of PEDOT:PSS: 40 nm; thickness of PCBM:
140 nm. b Rs: series resistance calculated from the corresponding IV curves.

SEM diagram of the cross-section of the champion cell without

the Al electrode.

Fig. 4

Fig. 3 IV and IPCE curves of the champion perovskite/PCBM solar

cell reported in this paper.

although the FF is slightly lower. We believe that a high Voc and

FF, which depend on the series resistance of the device, are the
key parameters for achieving such high eciency in inverted
perovskite solar cells. To explore why the cell prepared using the
two-step spin-coating method exhibited such a high Voc and FF,
the layer stacking of the device (or the morphology of the
perovskite lm in the vertical direction) was investigated.
The SEM of a cross section of the champion cell (without the
Ca/Al electrode) is shown in Fig. 4. The SEM image revealed a
well-dened, layer-by-layer structure with clear interfaces
between each layer. The thickness of the PEDOT:PSS, perovskite
and the PC71BM layers estimated from the SEM pictures is

15900 | J. Mater. Chem. A, 2014, 2, 1589715903

40 nm, 360 nm (close to the value measured a depth-prole

meter), and 140 nm, respectively. The thickness of the
CH3NH3PbI3 and PC71BM layers is similar to that reported by
Yang et al.,19 but the PC71BM layer is much thicker than that of
the cell fabricated by Gr
atzel et al.20 Furthermore, although
perovskite lms appeared to be composed of aggregated particles, most particles have a grain size close to the lm thickness.
Therefore, excitons can diuse within a single particle (crystallite) to the perovskite/PCBM interphase to undergo charge
separation, while the holes can transport to the PEDOT:PSS/ITO
anode via a single perovskite granule under very low internal
resistance. Each layer is in close contact with one another,
which makes the interphase resistance also very low. The series
resistance calculated from the IV curves of the corresponding
devices is provided in Table 2. Perovskite solar cells with a
higher Voc and FF also have a lower series resistance. The
champion cell reported in this article has a series resistance of
as low as 4.41 U cm2. The structure of the cross-section provides
a reason to explain why the cell with a thick perovskite lm and
4 organic/inorganic interphases still has a very high Voc, FF and
therefore high power conversion eciency.
We found that to achieve a conversion eciency of higher
than 15%, high quality (crystalline and pure) perovskite lms
and solvent annealing of the device are necessary for inverted
perovskite solar cells. To learn more about the eect of solvent
annealing on the photovoltaic performance of the solar cell, UV/
Vis absorption and PL spectra of the device (without Ca/Al

This journal is The Royal Society of Chemistry 2014

View Article Online

Published on 25 July 2014. Downloaded by Sungkyunkwan University on 10/11/2014 03:08:56.

Paper

electrode) with and without solvent annealing were obtained

and are displayed in Fig. 5. The perovskite lm has an absorption prole that extends to the NIR (800 nm) with an absorption
coecient (3, absorbance divided by the lm thickness) of 4.1 
104 cm1 at 550 nm. High 3 indicates that the perovskite lm
prepared via the two-step method is very dense.20,26 Solvent
annealing of the PC71BM layer did not change the absorption
prole and intensity of perovskite lm; however, the intensity of
PL, due to the recombination of perovskite excitons, decreased
signicantly. The decrease in the PL intensity without changing
the absorption prole suggests that the increase in the Jsc may
not due to a structural change of the perovskite lm. Instead, it
suggests that electron transfer from the perovskite to PC71BM is
more ecient, probably due to better perovskite/PC71BM
interfacial contact from the solvent annealing. The PL intensity
of the PEDOT:PSS/perovskite/PC71BM lm is almost zero aer it
was solvent annealed for 1 h, which indicates that exciton
separation and charge transfer are very ecient when the
perovskite and PC71BM have good contact. However, the device

Fig. 5 UV/Vis absorption (a) and PL (b) spectra of PEDOT:PSS/

perovskite/PCBM lms before and after solvent annealing. (The PL
spectra of glass/perovskite, PEDOT:PSS/perovskite and perovskite/
PCBM lms are also shown for reference.)

This journal is The Royal Society of Chemistry 2014

Journal of Materials Chemistry A

with the highest eciency (16.31%) was observed when the

perovskite/PC71BM lm was solvent annealed for 24 h (see Table
S2). Therefore, extending the solvent annealing time may also
increase the ordering of the PC71BM lm, which increases the
electron transport rate. Furthermore, the perovskite is known as
a bipolar absorber, indicating that it can be a hole and electron
transporter. The PL intensity of the PEDOT:PSS/perovskite and
perovskite/PCBM lms are much weaker than that of the glass/
perovskite lm, which suggests that the charge separation of
the exciton can occur in PEDOT:PSS/perovskite and perovskite/
PCBM interfaces. Moreover, the PL intensity of perovskite/
PCBM lm is weaker than that of PEDOT:PSS/ perovskite lm,
suggesting that perovskite/PC71BM interphase is a better place
for excitons to split.
Compared to the photovoltaic parameters of the inverted
perovskite cell recently reported by Xiao et al.22 (Jsc: 19.6 mA
cm2, Voc: 0.99 V, FF: 0.793 and 15.4% eciency) or the
champion regular perovskite cell reported by Kelly et al.28 (Jsc:
20.4 mA cm2, Voc: 1.03 V, FF: 0.749 and 15.7% eciency), our
device showed the highest open circuit voltage, as well as a
comparable ll factor and short current, suggesting the
importance of the cell interface engineering. CH3NH3PbI3 has a
low exciton binding energy (Eb) (19  3 meV calculated by Lam
et al.9), much lower than that of conjugated molecules,29 and a
high conductivity (103 S cm2, measured by Gr
atzel et al.30).
We believe that the thickness of the perovskite layer can be
further increased or other derivatives can be used that exhibit
high electric conductivity (such as CH3NH3PbI3xClx) to
increase the photocurrent without a loss in the Voc and FF of the
cell. With careful engineering of each layer, inverted (or OSCtyped) planar hetero-junction perovskite solar cell with higher
eciency can be expected in the future.
It is known31 that some perovskite solar cells show photocurrent hysteresis at certain voltage scanning rate (or sweep
delay time) and scan direction. The photocurrent hysteresis
may be due to either charge traps in low quality perovskite lms
or ferroelectric properties of the perovskite material and/or the
electromigration of ion in perovskites. The eciency reported
for a device with photocurrent hysteresis may not truly represent the performance of the device. We also used dierent
sweep delay times and two opposite scan directions to measure
the IV curves and the results are displayed in Fig. 6. Our
perovskite solar cell shows no photocurrent hysteresis at all
sweep delay times and scan directions. The results also suggested that the hysteresis in the photocurrent is more likely due
to the defects or charge traps in the perovskite lm, which can
be solved by the lm fabrication process. The device fabricated
with our method has a negligible amount of charge traps, and
therefore a high open-circuit voltage and power conversion
eciency.
In conclusion, we report a two-step spin coating method for
the synthesis of highly crystalline and dense perovskite lms
with better control of the lm stoichiometry and formation rate,
which aects the lm morphology and charge trap defects of
the perovskite. The ITO/PEDOT:PSS/perovskite/PCBM lm can
be fabricated in an ambient atmosphere using a solution
process at a low temperature. The champion cell has a power

J. Mater. Chem. A, 2014, 2, 1589715903 | 15901

View Article Online

Published on 25 July 2014. Downloaded by Sungkyunkwan University on 10/11/2014 03:08:56.

Journal of Materials Chemistry A

Paper

in literature.30 UV/Vis absorption and PL spectra were recorded

on a Hitachi U-4100 and F-7000 spectrometer, respectively. The
thickness of the lms was measured using a depth-prole meter
(Veeco Dektak 150, USA). Five lines on a 1 cm  1 cm lm were
made by carefully scratching with a tip and the average height
between the hills and valleys was used to determine the lm
thickness. GIXRD data were collected in the 2q range of 550

on a Brucker powder diractometer (D8 Discover) equipped
with a 2D detector using Cu Ka1 radiation. Scanning electron
microscopy (SEM) was performed on a Hitachi S-800 microscope at 15 kV. Samples (the surface and cross-section of a lm
on substrates) for SEM imaging were mounted on a metal stub
with a piece of conducting tape, and then coated with a thin
layer of gold lm to avoid charging.

Device fabrication and photovoltaic performance

measurement

(a) IV curves measured with dierent delay times between

each measurement. (b) IV curves measured at dierent sweep
directions.

Fig. 6

conversion eciency of 16.31% with a remarkably high Voc of

1.05 V and FF of 0.78. The eciency reproducibility of the
devices is also high: ca. 80% of devices achieved a power
conversion eciency of higher than 15%. The results demonstrate that fabricating highly ecient inverted perovskite solar
cells in a more controllable way is possible via a new perovskite
lm preparation method and careful interface engineering.

Experimental

PEDOT:PSS was spin-coated on a heat pretreated patterned ITO

under 5000 rpm for 50 s and then annealed at 120
C for
15 min. For depositing the perovskite layer, rst a layer of PbI2
was spin-coated on top of the PEDOT:PSS-coated ITO substrate
from a 0.5 M DMF solution. Aer the PbI2 layer was dried in
air, CH3NH3I was spin-coated on top of PbI2 lm from an
isopropanol solution of various concentrations at spin rates of
1500 rpm to form the perovskite structure. Specically, the
highest eciency device was observed using a concentration of
CH3NH3I/IPA of ca. 38 mg mL1 at a spin rate of 1500 rpm.
Aer the perovskite lm was formed, 4 wt% PCBM in chlorobenzene was spin-coated onto the perovskite layer at 1000 rpm
for 30 s. Solvent annealing of PCBM layer was performed by
covering the perovskite/PCBM lm with a Petri dish for 24
hours before placing in a vacuum evaporator. All the fabrication procedures were performed in an ambient atmosphere at
room temperature. Finally, the PEDOT:PSS/perovskite/PCBM
lm was transferred to a vacuum chamber for coating with the
Ca/Al electrode (20 nm/100 nm). The device area was 0.5 cm 
0.2 cm. The IV characteristics of the solar cells were determined using a Keithley 4200 source measuring unit under a
simulated AM 1.5G light (Wacom solar simulator) at 100 mW
cm2 and the light intensity was calibrated by KG-5 Si diode.
The external quantum eciency (EQE) or incident photo-tocurrent conversion eciency (IPCE) was measured in air. A
chopper and lock-in amplier were used for the phase sensitive
detection with a QE-R3011 measurement system (Enlitech Inc.,
Taiwan).

Materials and physicochemical studies

An aqueous dispersion of PEDOT:PSS (1.31.7 wt%, from H.C.
Stark Baytron P AI 4083) was obtained from Heraeus Co.
Fullerene derivatives (PC61BM (99.8%) and PC71BM (99.8%))
were purchased from Solenne B. V., Netherlands. PbI2
(99.999%) was purchased from Aldrich Co. All the above materials were used as received. ITO-covered glass substrates
purchased from Merck Co. were photolithographically
patterned in our laboratory with HCl(aq). CH3NH3I (MAI,
99.99%) was synthesized using the same method as published

15902 | J. Mater. Chem. A, 2014, 2, 1589715903

Acknowledgements
Financial support from the National Science Council (NSC),
Taiwan, ROC (grand number: NSC101-2113-M-008-008-MY3) is
acknowledged. The devices fabrication was carried out at the
Advanced Laboratory of Accommodation and Research for
Organic Photovoltaics, Ministry of Science and Technology,
Taiwan, ROC.

This journal is The Royal Society of Chemistry 2014

View Article Online

Paper

Published on 25 July 2014. Downloaded by Sungkyunkwan University on 10/11/2014 03:08:56.

References
1 A. Kojima, K. Teshima, Y. Shirai and T. Miyasaka, J. Am.
Chem. Soc., 2009, 131, 6050.
2 J. Burschka, N. Pellet, S.-J. Moon, R. Humphry-Baker, P. Gao,
M. K. Nazeeruddin and M. Gr
atzel, Nature, 2013, 499, 316.
3 M. Liu, M. B. Johnston and H. J. Snaith, Nature, 2013, 501,
395.
4 C. Kagan, D. Mitzi and C. A. Dimitrakopoulos, Science, 1999,
286, 945.
5 D. B. Mitzi, Prog. Inorg. Chem., 2007, 48, 1121.
6 N.-G. Park, J. Phys. Chem. Lett., 2013, 4, 2423.
7 M. D. McGehee, Nature, 2013, 50, 323.
8 S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou,
M. J. P. Alcocer, T. Leijtens, L. M. Herz, A. Petrozza and
H. J. Snaith, Science, 2013, 342, 341.
9 G. Xing, N. Mathews, S. Sun, S. S. Lim, Y. M. Lam, M. Gr
atzel,
S. Mhaisalkar and T. C. Sum, Science, 2013, 342, 344.
10 J. M. Ball, M. M. Lee, A. Hey and H. J. Snaith, Energy Environ.
Sci., 2013, 6, 1739.
11 W. A. Laban and L. Etgar, Energy Environ. Sci., 2013, 6, 3249.
12 Y. Zhao and K. Zhu, J. Phys. Chem. Lett., 2013, 4, 2880.
13 L. Etgar, P. Gao, Z. Xue, Q. Peng, A. K. Chandiran, B. Liu,
M. K. Nazeeruddin and M. Gr
atzel, J. Am. Chem. Soc., 2012,
134, 17396.
14 G. E. Eperon, V. M. Burlakov, P. Docampo, A. Goriely and
H. J. Snaith, Adv. Funct. Mater., 2014, 24, 151.
15 Q. Chen, H. Zhou, Z. Hong, S. Luo, H.-S. Duan, H.-H. Wang,
Y. Liu, G. Li and Y. Yang, J. Am. Chem. Soc., 2014, 136, 622.
16 J.-Y. Jeng, Y.-F. Chiang, M.-H. Lee, S.-R. Peng, T.-F. Guo,
P. Chen and T.-C. Wen, Adv. Mater., 2013, 25, 3727.
17 S. Sun, T. Salim, N. Mathews, M. Duchamp, C. Boothroyd,
G. Xing, T. C. Sumbce and Y. M. Lam, Energy Environ. Sci.,
2014, 7, 399.

This journal is The Royal Society of Chemistry 2014

Journal of Materials Chemistry A

18 P. Docampo, J. M. Ball, M. Darwich, G. E. Eperon and

H. J. Snaith, Nat. Commun., 2014, 4, 2761.
19 J. You, Z. Hong, Y. (M. ) Yang, Q. Chen, M. Cai, T.-B. Song,
C.-C. Chen, S. Lu, Y. Liu, H. Zhou and Y. Yang, ACS Nano,
2014, 8, 1674.
20 O. Malinkiewicz, A. Yella, Y. H. Lee, G. M. Espallargas,
M. Gr
atzel, M. K. Nazeeruddin and H. J. Bolink, Nat.
Photonics, 2014, 8, 128.
21 G. E. Eperon, V. M. Burlakov, P. Docampo, A. Goriely and
H. J. Snaith, Adv. Funct. Mater., 2014, 24, 151.
22 Z. Xiao, C. Bi, C. Shao, Q. Dong, Q. Wang, Y. Yuan, Y. Wang,
Y. Gao and Y. G. Huang, Energy Environ. Sci., DOI: 10.1039/
c4ee01138d.
23 A. Kojima, K. Teshima, Y. Shirai and T. Miyasaka, J. Am.
Chem. Soc., 2009, 131, 6050.
24 J.-H. Im, C.-R. Lee, J.-W. Lee, S.-W. Park and N.-G. Park,
Nanoscale, 2011, 3, 4088.
25 T. B. Singh, N. Marjanovi
c, G. J. Matt, S. G
unes, N. S. Saricici,
A. M. Ramil, A. Andreev, H. Sitter, R. Schw
odiauer and
S. Bauer, Org. Electron., 2005, 6, 105.
26 S. Kazim, M. K. Nazeeruddin, M. Gr
atzel and S. Ahmad,
Angew. Chem., Int. Ed., 2014, 53, 2812.
27 P.-W. Liang, C.-Y. Liao, C.-C. Chueh, F. Zuo, S. T. Williams,
X.-K. Xin, J. Lin and A. K. Y. Jen, Adv. Mater., 2014, 26,
37483754.
28 D. Liu and T. L. Kelly, Nat. Photonics, 2014, 8, 133.
29 V. I. Arkhipov and H. B
assler, Phys. Status Solidi A, 2004, 201,
1152.
30 L. Etgar, P. Gao, Z. Xue, Q. Peng, A. K. Chandiran, B. Liu,
M. K. Nazeeruddin and M. Gr
atzel, J. Am. Chem. Soc., 2012,
134, 17396.
31 H. J. Snaith, A. Abate, J. M. Ball, G. E. Eperon, T. Leijtens,
N. K. Noel, S. D. Stranks, J. T.-W. Wang, K. Wojciechowski
and W. Zhang, J. Phys. Chem. Lett., 2014, 5, 1511.

J. Mater. Chem. A, 2014, 2, 1589715903 | 15903