A highly efficient mesoscopic solar cell based on CH3NH3PbI3-xClx fabricated via sequential solution deposition

Article abstract of DOI:10.1039/c4cc01962h

A mixed halide perovskite of CH₃NH₃PbI_3-xCl_x is synthesized via two-step sequential solution deposition by using a mixture of PbCl₂ and PbI₂ as the precursor to overcome the low solubility of pure PbCl₂ with easy morphology control. 11.7% power conversion efficiency is achieved for the mesoscopic cell, much higher than the cell constructed via a spin-coating process. This journal is

Full text of DOI:10.1039/c4cc01962h

ChemComm
View Article Online
View Journal
COMMUNICATION
A highly efficient mesoscopic solar cell based on
CH3NH3PbI3ÀxClx fabricated via sequential
solution deposition†
Yingzhuang Ma,a Lingling Zheng,a Yao-Hsien Chung,a Saisai Chu,a Lixin Xiao,*ab
Zhijian Chen,ab Shufeng Wang,ab Bo Qu,ab Qihuang Gong,a Zhaoxin Wu*c and
Xun Houc
Cite this:DOI: 10.1039/c4cc01962h
Received 17th March 2014,
Accepted 12th May 2014
DOI: 10.1039/c4cc01962h
www.rsc.org/chemcomm
A mixed halide perovskite of CH3NH3PbI3ÀxClx is synthesized via used for the fabrication of solution-based CH3NH3PbI3ÀxClx cells,
two-step sequential solution deposition by using a mixture of PbCl2 but only 7.6% PCE for the CH3NH3PbI3ÀxClx cell containing the
and PbI2 as the precursor to overcome the low solubility of pure PbCl2 mesoporous TiO2 layer.4 While CH3NH3PbI3 devices with shorter
with easy morphology control. 11.7% power conversion efficiency diffusion lengths achieved the efficiencies over 15% by two-step
is achieved for the mesoscopic cell, much higher than the cell sequential solution deposition, with easier morphology control.7,22
constructed via a spin-coating process.
The morphology of the resultant perovskite film is very sensitive to
the conditions during the spin-coating procedure, e.g., thickness
+
Hybrid perovskite materials in the form of AMX3 (A = CH3NH3 ; of the compact TiO2 layer, annealing time, annealing tempreture
M = Pb2+, or Sn2+; and X = ClÀ, BrÀ, or IÀ) were first discovered and the thickness of the perovskite layer.15 The morphology of
by Weber and developed by Mitzi et al., and then CH3NH3PbI3 the perovskite is crucial to the device performance, because the
or CH3NH3PbBr3 was used as the dye-absorbing layer in a liquid defects in perovskite crystallites and the interfaces may be the trap
dye-sensitized solar cell (DSC) reported by Miyasaka et al.1–3 of charges, preventing them reaching the electrodes. Therefore, it is
Following the pioneering work, the perovskites were used in necessary to improve the morphology of the CH3NH3PbI3ÀxClx cell
solid-state solar cells to avoid the dissolution of perovskite in the through the solution-based method.
presence of a liquid electrolyte. As a result, much higher power
To further improve the photovoltaic performance of mixed
conversion efficiencies (PCEs) were achieved than those of liquid halide perovskite of CH3NH3PbI3ÀxClx via the solution process, in
cells.4–22 Recently, Snaith et al. found that the electron–hole diffu- this work, we successfully figure out the feasible compositions of
sion lengths exceeded 1 mm for the mixed halide perovskite of the precursor for the CH3NH3PbI3ÀxClx perovskite by two-step
CH3NH3PbI3ÀxClx, while it was only 100 nm for the triiodide sequential deposition, permitting much easier morphology con-
perovskite of CH3NH3PbI3, showing the promising prospects of trol. A mixture precursor of PbI2 and PbCl2 is more promising for
CH3NH3PbI3ÀxClx.8 The mixed halide perovskite CH3NH3PbI3ÀxClx the preparation of CH3NH3PbI3ÀxClx than pure PbCl2, due to the
achieved 15.4% PCE through mix-vapor vacuum deposition to poor solubility of pure PbCl2. The device performance reaches
give a uniform perovskite film with a planar heterojunction.10 11.7%, which is the highest efficiency for CH3NH3PbI3ÀxClx
However, vacuum deposition will greatly increase the cost of via the solution process containing a mesoporous TiO2 layer
large-scale fabrication compared to the cost-effective solution based on our knowledge, much higher than 4.8% for the device
process. Up to now, spin-coating deposition has been widely fabricated via spin-coating deposition.
The CH3NH3PbI3ÀxClx perovskite was first reported by Snaith
et al. by using a mixture of CH3NH3I and PbCl2 with a molar ratio
a State Key Laboratory for Mesoscopic Physics and Department of Physics,
of 3 : 1 via spin-coating deposition (Scheme 1b).4,5,8 Following this
Peking University, Beijing 100871, PR China. E-mail: xiao66@pku.edu.cn
b New Display Device and System Integration Collaborative Innovation Center of the
reaction, we first tried to use saturated (0.5 M) PbCl2 as the
precursor in the two-step sequential deposition, and spin-coated it
West Coast of the Taiwan Strait, Fuzhou 350002, China
c Key Laboratory of Photonics Technology for Information, Key Laboratory for
Physical Electronics and Devices of the Ministry of Education,
Department of Electronic Science and technology, School of Electronic and
Information Engineering, Xi’an Jiaotong University, Xi’an, 710049, PR China.
E-mail: zhaoxinwu@mail.xjtu.edu.cn
on the mesoporous TiO2 film, and then dipped into CH3NH3I
† Electronic supplementary information (ESI) available: Material synthesis, EDX
results, device preparation procedures, and device characterization. See DOI:
10.1039/c4cc01962h
Scheme 1 The reactions for CH3NH3PbI3ÀxClx.
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun.

View Article Online
Communication
ChemComm
solution to form perovskite on site. Unfortunately, the absorption confirmed by the energy-dispersive X-ray (EDX) spectrum as shown
of the resultant film was very weak. It might be due to the poorer in Fig. S1 (ESI†). The elemental mapping analysis of the surface
solubility of PbCl2 than that of PbI2 in dimethylformamide (DMF), confirms that the resultant perovskite of Film a is a mixed halide
because high concentration (at least 1 M) of the precursor is very CH3NH3PbI3ÀxClx with approximately x E 0.26. The images of Pb, I
important to obtain high loading in the mesoporous TiO2 film for and Cl also show that the perovskites are well distributed. The
high device performance.7 However, in the presence of additives, absorption spectra of the two perovskite films are shown in Fig. 2b.
e.g., PbI2 or CH3NH3I, the solubility of PbCl2 can be signifi- Film a has strong broadband absorption in the visible region from
cantly increased due to the common ion effect.15,31 In addition, 400 nm to 800 nm, on the contrary, Film b has rather poor
the higher formation energy of chlorine incorporation into the absorption, indicating insufficient perovskite filling in the porous
perovskite matrix than that of iodine might also be the reason.26 TiO2 layer. These results indicate that using the mixture of PbI2 and
Therefore, we changed the precursor to a mixture of 0.5 M PbCl2 and PbCl2 as the precursor is an effective way to form the mixed halide
0.5 M PbI2 (molar ratio 1 : 1) according to the reaction of Scheme 1a, perovskite of CH3NH3PbI3ÀxClx, following Scheme 1a.
and followed the two-step sequential deposition to give Film a with
To further evaluate the differences in the resultant CH3NH3-
the color changing from chartreuse to dark brown, indicating the PbI3ÀxClx perovskite layers between two-step sequential deposi-
formation of the perovskite. From the reaction formula given below, tion and spin-coating deposition, we also fabricated another
it could also be inferred that Scheme 1a tends to give less byproducts perovskite layer via spin-coating the mixture of CH3NH3I, PbCl2
compared to Scheme 1b. The excess methylammonium iodide and and PbI2 (molar ratio 2 : 1 : 1) also following Scheme 1a for
chloride are assumed to be lost via evaporation during the annealing comparison (hereafter referred to as Film c). The morphology of
process.4 Less byproducts may leave fewer pin-holes in the perovs- Film a, Film b and Film c was measured by both scanning
kite layer during the annealing process.15 As comparison, the electron microscopy (SEM) and atomic force microscopy (AFM)
resultant film formed by using PbCl2 as the precursor (Scheme 1b) as shown in Fig. 3.
is referred to as Film b. The illustrations of different fabrication
procedures for CH3NH3PbI3ÀxClx layers are shown in Fig. 1.
The structures of the resultant perovskite films were investigated
via X-ray diffraction (XRD) as shown in Fig. 2a. The structure of the
synthesized perovskite is consistent with tetragonal phase structures
reported in the previous reports.23–25 Strong peaks at 14.031 and
27.521, corresponding to the (110) and (220) planes, confirm the
formation of a tetragonal perovskite structure with lattice parameters
a = b = 8.84 Å and c = 12.57 Å. The signals of Film b are much weaker
than those of Film a, indicating that more perovskite crystallites
formed in mesoporous TiO2 by the mixture of PbI2 and PbCl2 than
pure PbCl2 as the precursor. The composition of Film a was
Fig. 1 The illustrations of fabrication procedures of the provskite layer
fabricated via spin-coating deposition and two-step sequential deposition.
Fig. 3 The SEM and AFM images of (a) Film a, (b) Film b and (c) Film c. (d) The
cross-sectional SEM view of Device A. (e) TEM images of CH3NH3PbI3ÀxClx
Fig. 2 (a) XRD spectra of Film a and b. (b) Absorbance of Film a and b.
deposited in the mesoporous TiO2 film.
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2014

View Article Online
ChemComm
By using the mixture of PbI2 and PbCl2 as the precursor
Communication
Table 1 Photovoltaic performances of the hybrid perovskite solar cells
(Film a), the perovskite via two-step sequential deposition did
not only penetrate deeply into the mesoporous TiO2, but also
grew into crystallites to cover the TiO2 layer with an in-plain
grain size of 300–350 nm as shown in Fig. 3a, and the grain size
has been proved to be suitable with a strong light scattering
effect to further enhance the light absorption.30 On the con-
trary, there were almost no more perovskite crystallites on top
of the mesoporous TiO2 for Film b fabricated via two-step
Device
Process
Voc [V]
Jsc [mA cmÀ2
]
FF
PCE [%]
A
B
Sequential
Spin-coating
Vacuum
1.04
0.86
1.07
0.80
17.2
9.4
21.5
17.8
0.65
0.59
0.67
0.53
11.7
4.8
15.4
7.6
Ref. 10
Ref. 4
Spin-coating
Compared with Device B, the IPCE of Device A is over 60%
sequential deposition by using PbCl2 as the precursor as shown between 400 to 700 nm, and the highest value is nearly 80% at
in Fig. 3b, indicating insufficient pore filling in the mesoporous around 500 nm, matching the stronger absorption region of the
TiO2 layer. This might be one of the possible reasons for the perovskite layer. The PCE of Device A is up to 11.7%, with a Jsc
absorption difference by using different precursors. For Film c, of 17.2 mA cmÀ2, a Voc of 1.04 V, and a FF of 0.65, which is the
the SEM images (Fig. 3c) showed that no continuous perovskite highest efficiency for CH3NH3PbI3ÀxClx via the solution process
capping layer was formed on top of the mesoporous TiO2, containing a mesoporous TiO2 layer based on our knowledge,
although there was penetration into the mesoporous TiO2 layer. much higher than that of 7.6% obtained via the spin-coating
The above results indicate that the morphology of the perovs- process with the mesoporous TiO2 layer as reported listed in
kite layer fabricated via two-step sequential deposition is easier Table 1.4 The corresponding Device B showed a Jsc of 9.4 mA cmÀ2
,
to control compared with via spin-coating deposition. The a Voc of 0.86 V, a FF of 0.59 and 4.8% PCE (Fig. 4a), a much lower
surface morphology of the three films was also investigated PCE than that of the Device A fabricated via two-step sequential
via AFM images. As a result, the crystalline arrangement in Film deposition. The improvement of photovoltaic performances may be
a formed a continuous microstructure consistent with the SEM mainly ascribed to the different perovskite morphology of the two
image, the absorption of incident light was enhanced due to devices. Compared with Device A, there is 0.18 V decrease in the Voc
the presence of the crystalline perovskite capping layer and the of Device B because of the direct contact between the uncovered
light scattering effect, which have also been proved to be an TiO2 and the p-type HTM due to the absence of the perovskite
effective way in DSCs.22,27 These results could be proved by the capping layer, resulting in the charge recombination between the
cross-section image (Fig. 3d) with elemental mapping (Fig. S2, TiO2 and HTM.15,28,29 Besides, the absence of the perovskite
ESI†) of Film a. It could be clearly identified that the perovskite capping layer might lead to lower absorption, resulting in lower
not only formed a continuous capping layer, but also pene- Jsc. It should be noted here that the lower PCE than 15% for
trated well into the mesoporous TiO2 film. The perovskite pore- CH3NH3PbI3 via two-step sequential deposition might be due to the
filling should improve the device performance because higher lower solubility of PbCl2 than that of PbI2 in the solvent of DMF.
electron densities can be sustained in the TiO2 layer, increasing High concentration should be crucial to the growth of perovskite
electron transporting rates as previously reported.28 To further crystallites as reported in the case of CH3NH3PbI3 via two-step
evaluate the interfacial properties of TiO2/perovskite, the trans- sequential deposition.7 This could be also supported by the lower
mission electron microscopy (TEM) image (Fig. 3e) was obtained PCEs for CH3NH3PbI3ÀxClx via the solution process than that via
from the scratch from Film a. The image reveals the formation of vacuum deposition (sufficient chlorine source). Compared to the
the CH3NH3PbI3ÀxClx perovskite thin film on the TiO2 surface, CH3NH3PbI3ÀxClx cell with 15.4% PCE fabricated via vacuum
which prohibits the direct contact between TiO2 and hole deposition, the Voc and the FF of the spin-coated CH3NH3PbI3ÀxClx
transporting material (HTM).
cells are significantly lower, due to the difficulty in controlling the
The experimental details of device fabrication are given in morphology of the perovskite crystallites formed via a spin-coating
the ESI.† Devices A and B correspond to Film a and Film c, process, leading to the lower PCE of 7.6%. While Device A
respectively. Devices were measured under a standard AM 1.5 fabricated via two-step sequential deposition is an effective method
illumination of 100 mW cmÀ2. The J–V curves are shown in to reduce the Voc and FF losses caused by the morphology defects of
Fig. 4a. The incident photon to current conversion efficiency CH3NH3PbI3ÀxClx perovskite. Further work should be focused on
(IPCE) was measured as shown in Fig. 4b.
optimizing the conditions to control the morphology of the per-
ovskite capping layer to further increase the lower Jsc caused by the
low solubility of PbCl2 as mentioned above.
In conclusion, we introduce two-step sequential solution
deposition for the fabrication of the CH3NH3PbI3ÀxClx meso-
scopic solar cell. The mixture of PbI2 and PbCl2 is used rather
than pure PbCl2 as the precursor to ensure high loading in the
mesoporous TiO2 film, with easy morphology control, which
results in a great enhancement of the light absorbance for the
device and suppressing the electron hole recombination by
separating the bare TiO2 and the HTM layer. As a result, 11.7%
Fig. 4 (a) The photovoltaic performances of Devices A and B. (b) Absor-
bance and IPCE of the perovskite films in Devices A and B.
PCE for the device has been achieved with a Jsc of 17.2 mA cmÀ2
,
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun.

View Article Online
Communication
ChemComm
11 L. Etgar, P. Gao, Z. Xue, Q. Peng, A. K. Chandiran, B. Liu and
a Voc of 1.04 V of, and a FF of 0.6, which is the highest efficiency of
CH3NH3PbI3ÀxClx fabricated via a solution process containing a
mesoporous TiO2 layer based on our knowledge. This is much
higher than 4.8% for the device fabricated via a spin-coating process.
This work suggests a promising way for the solution-based prepara-
tion of the mixed halide perovskite of CH3NH3PbI3ÀxClx for highly
efficient solar cells.
This study was partly financially supported by the National
Natural Science Foundation of China (10934001, 61177020,
11074016, 11121091, 61275034 and 61106123) and the National
Basic Research Program of China (2013CB328700). The authors
are thankful to Mr Xiao Yu and Prof. Dechun Zou in the College
of Chemistry and Molecular Engineering of Peking University
for their kind help with the measurements of SEM images.
¨
M. Gratzel, J. Am. Chem. Soc., 2012, 134, 17396.
12 D. Bi, L. Yang, G. Boschloo, A. Hagfeldt and E. M. Johansson,
J. Phys. Chem. Lett., 2013, 4, 1532.
13 D. Sabba, H. M. Kumar, N. Yantara, T. T. T. Pham, N. G. Park,
¨
M. Gratzel and P. P. Boix, Nanoscale, 2014, 6, 1675.
¨
14 M. H. Kumar, N. Yantara, S. Dharani, M. Gratzel, S. Mhaisalkar,
P. P. Boix and N. Mathews, Chem. Commun., 2013, 49, 11089.
15 G. E. Eperon, V. M. Burlakov, P. Docampo, A. Goriely and H. J. Snaith,
Adv. Funct. Mater., 2014, 24, 151.
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 J. H. Heo, S. H. Im, J. H. Noh, T. N. Mandal, C. S. Lim, J. A. Chang
and S. I. Seok, Nat. Photonics, 2013, 7, 486.
18 H. S. Kim, J. W. Lee, N. Yantara, P. P. Boix, S. A. Kulkarni,
S. G. Mhaisalkar and N. G. Park, Nano Lett., 2013, 13, 2412.
19 A. Abrusci, S. D. Stranks, P. Docampo, H. L. Yip, A. K. Y. Jen and
H. J. Snaith, Nano Lett., 2013, 13, 3124.
20 N. G. Park, J. Phys. Chem. Lett., 2013, 4, 2423.
21 J. Burschka, A. Dualeh, F. Kessler, E. Baranoff, N. L. Cevey-Ha, C. Yi
¨
and M. Gratzel, J. Am. Chem. Soc., 2011, 133, 18042.
Notes and references
22 D. Liu and T. L. Kelly, Nat. Photonics, 2014, 8, 133.
1 D. Weber, Z. Naturforsch., B: Anorg. Chem. Org. Chem., 1978, 23 T. Baikie, Y. Fang, J. M. Kadro, M. Schreyer, F. Wei, S. G. Mhaisalkar
33, 1443. and T. J. White, J. Mater. Chem. A, 2013, 1, 5628.
2 D. B. Mitzi, C. A. Field, W. T. A. Harrison and A. M. Guloy, Nature, 24 C. C. Stoumpos, C. D. Malliakas and M. G. Kanatzidis, Inorg. Chem.,
1994, 369, 467. 2013, 52, 9019.
3 A. Kojima, K. Teshima, Y. Shirai and T. Miyasaka, J. Am. Chem. Soc., 25 S. Colella, E. Mosconi, P. Fedeli, A. Listorti, F. Gazza, F. Orlandi,
2009, 131, 6050.
P. Ferro, T. Besagni, A. Rizzo, G. Calestani, G. Gigli, F. D. Angelis and
4 M. M. Lee, J. Teuscher, T. Miyasak, T. N. Murakami and H. J. Snaith,
Science, 2012, 338, 643.
R. Mosca, Chem. Mater., 2013, 25, 4613.
¨
26 E. Mosconi, A. Amat, M. K. Nazeeruddin, M. Gratzel and
5 J. M. Ball, M. M. Lee, A. Hey and H. J. Snaith, Energy Environ. Sci.,
2013, 6, 1739.
6 H. S. Kim, C. R. Lee, J. H. Im, K. B. Lee, T. Moehl, A. Marchioro,
F. D. Angelis, J. Phys. Chem. C, 2013, 117, 13902.
27 S. Hore, P. Nitz, C. Vetter, C. Prahl, M. Niggemann and R. Kern,
Chem. Commun., 2005, 2011.
¨
S. J. Moon, R. H. Baker, J. H. Yum, J. E. Moser, M. Gratzel and 28 T. Leijtens, B. Lauber, G. E. Eperon, S. D. Stranks and H. J. Snaith,
N. G. Park, Sci. Rep., 2012, 2, 591. J. Phys. Chem. Lett., 2014, 5, 1096.
7 J. Burschka, N. Pellet, S. J. Moon, R. Humphry-Baker, P. Gao, 29 H. J. Snaith, N. C. Greenham and R. H. Friend, Adv. Mater., 2004,
¨
M. K. Nazeeruddin and M. Gratzel, Nature, 2013, 499, 316.
8 S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou, M. J. Alcocer, 30 Q. Chen, H. Zhou, Z. Hong, S. Luo, H.-S. Duan, H.-H. Wang, Y. Liu,
T. Leijtens and H. J. Snaith, Science, 2013, 342, 341. G. Li and Y. Yang, J. Am. Chem. Soc., 2013, 136, 622.
9 G. Xing, N. Mathews, S. Sun, S. S. Lim, Y. M. Lam, M. Gratzel and 31 J. Mendham, R. C. Denney, J. D. Barnes and M. J. K. Thomas, Vogel’s
16, 1641.
¨
T. C. Sum, Science, 2013, 342, 344.
10 M. Liu, M. B. Johnston and H. J. Snaith, Nature, 2013, 501, 395.
Quantitative Chemical Analysis, Prentice Hall, New York, 6th edn,
2000, ISBN 0-582-22628-7.
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2014