Reproducible fabrication of efficient perovskite-based solar cells: X-ray crystallographic studies on the formation of CH3NH 3PbI3 layers

Article abstract of DOI:10.1246/cl.140074

Lead halide complexes formed during the fabrication of perovskite (CH ₃NH₃PbI₃)-based solar cells were examined by single-crystal X-ray diffraction analysis. Based on the observed results, the fabrication protocol using a sequential deposition method was optimized to reproducibly provide highly efficient solar cells with power conversion efficiencies of more than 10%.

Full text of DOI:10.1246/cl.140074

Received: January 31, 2014 | Accepted: February 4, 2014 | Web Released: February 6, 2014
CL-140074
Reproducible Fabrication of Efficient Perovskite-based Solar Cells: X-ray Crystallographic
Studies on the Formation of CH NH PbI Layers
3
3
3
Atsushi Wakamiya,*1,2 Masaru Endo, Takahiro Sasamori, Norihiro Tokitoh,
1
1
1
3
3
1
Yuhei Ogomi, Shuzi Hayase, and Yasujiro Murata
Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011
1
2
PRESTO, Japan Science and Technology Agency, 4-1-8 Honcho, Kawaguchi, Saitama 332-0012
Graduate School of Life Sciences and Systems Engineering, Kyushu Institute of Technology,
3
2-4 Hibikino, Wakamatsu-ku, Kitakyushu, Fukuoka 808-0196
(
E-mail: wakamiya@scl.kyoto-u.ac.jp)
Lead halide complexes formed during the fabrication of
perovskite (CH3NH3PbI3)-based solar cells were examined by
single-crystal X-ray diffraction analysis. Based on the observed
results, the fabrication protocol using a sequential deposition
method was optimized to reproducibly provide highly efficient
solar cells with power conversion efficiencies of more than 10%.
spin-coating of an ethanol suspension of a TiO2 paste (PST-
18NR, TiO2 paste:ethanol = 1:3.5 wt ratio) resulted in the
deposition of ca. 200 nm-thick-mesoporous films of TiO2
nanoparticles (particle diameter: ca. 20 nm). In an Ar-filled
glovebox, PbI2 was then introduced into the TiO2 nanopores by
spin-coating a solution of PbI2 in DMF (1.0­1.1 M) at 70 °C.
Upon drying (70 °C, 1 h), the films turned to deep yellow.
Afterwards, the films were dipped for 20 s in a 0.06 M solution
of CH3NH3I in 2-propanol. During this process, the color of the
films turned to red-black from the corner of the square substrates
1
Following the pioneering work of Miyasaka et al., recent
reports on solid-state solar cells with high power conversion
2­7
16
efficiencies (PCEs) instigated investigations into perovskite-
based solar cells with CH3NH3PbX3 (X = halogen) as light
(see movie in the SI ). This may be attributed to the slightly
nonhomogeneous distributions of the PbI2 loadings on the
substrate surface during the spin-coating process. The films were
quickly rinsed with 2-propanol and dried (70 °C, 30 min). The
hole-transporting layer was deposited on the perovskite layer by
spin-coating a solution of 2,2¤,7,7¤-tetrakis(N,N-di-p-methoxy-
phenylamine)-9,9¤-spirobifluorene (spiro-OMeTAD) in chloro-
benzene (0.058 M) containing 4-tert-butylpyridine (0.19 M)
and lithium bis(trifluoromethylsulfonyl)imide (0.031 M) as
well as tris[2-(1H-pyrazol-1-yl)-4-tert-butylpyridine]cobalt(III)
8­12
absorbers.
This new promising type of cost-effective solar
cells has attracted much attention, mostly because perovskite
light absorbers can be easily processed in solution, similar to
those in organic photovoltaics13 and dye-sensitized solar cells.
High PCEs of over 15% have been reported for some perovskite-
14
5,7
based solar cells in literature, but the cells often suffer from
2
large PCE variations, which severely hamper systematic studies
on further improvements of materials and device structures,
as well as operating mechanisms. It is, therefore, important to
develop and establish a reproducible fabrication method for
highly efficient solar cells. Several methods, e.g., solution
¹
3
17
tris[bis(trifluoromethylsulfonyl)imide] (5.6 © 10 M) as do-
pants. For the back contact, a gold layer (80 nm) was thermally
deposited on top of the device. The performance of the solar
cells was measured using a 2 mm square mask in air without
sealing.
1
­5
6
(
single-step spin-coating or sequential deposition ) and vapor
7
deposition processes, have been reported for the formation of
perovskite layers. In this work, we used a sequential deposition
method and examined the influence of the purity of the starting
materials on the formation of the resulting perovskite layers in
solar cells with high PCEs. Moreover, we report the results
of single-crystal X-ray diffraction analyses conducted on lead
halide complexes, which are potentially formed during the
production process. Based on the results obtained, the fabrica-
tion protocol for the solar cells was optimized. We found that for
the reproducible fabrication of highly efficient perovskite solar
cells, the water content of lead iodide (PbI2) as the starting
material is crucial. Our optimized method uses dehydrated PbI2
and thus enables the highly reproducible fabrication of solar
cells with PCEs of more than 10%.
During the preparation of 1.0 M PbI2 solutions in DMF, we
noticed that the PbI (99.999% trace metal basis) we purchased
2
did not dissolve completely, even after stirring overnight at
70 °C. An analysis of the water content by the Karl Fischer
method indicated that our batch of PbI contained nonnegligible
2
amounts of water (ca. 2000 ppm). The use of low-concentrated
PbI2 solutions in DMF (<1.0 M) from this starting material
resulted in a cell with a low PCE of 5.23% (J
10.7 mA cm , VOC = 0.82 V, FF = 0.60, Figure 1). The water
content was easily reduced to ca. 100 ppm, by simply heating the
=
SC
¹2
PbI for 10­15 min under vacuum (405­450 °C, 0.2 mmHg). In
2
contrast, this dehydrated PbI2 dissolved easily in DMF. The use
of a 1.0 M PbI2 solution prepared from this material significantly
improved both JSC and VOC values to afford a cell with a PCE
To minimize the influence of water and oxygen, we prepared
the perovskite and hole-transporting layers in an inert glovebox
¹
2
of 10.2% (JSC = 17.2 mA cm , VOC = 0.96 V, FF = 0.62,
Figure 1). A PbI2 solution with a higher concentration (1.1 M)
further increased the JSC to afford a cell with a PCE of 11.0%
(Ar; H2O and O2, <0.1 ppm) using dehydrated (H2O, <8 ppm)
N,N-dimethylformamide (DMF), 2-propanol, and chloroben-
zene. Otherwise, the solar cells were fabricated according to
¹
2
(JSC = 18.9 mA cm , VOC = 0.92 V, FF = 0.63, Figure 1).
These results suggest that the solution process for the
formation of the CH3NH3PbI3 layers is very important in order
to obtain highly efficient solar cells. To gain a better under-
standing of this process, we conducted single-crystal X-ray
6
literature procedures. Patterned transparent conducting oxide
substrates (FTO, 25 mm © 25 mm) were covered with a compact
TiO layer by spray pyrolysis of bis(acetylacetonato)titanium
2
6,15
diisopropoxide in ethanol (0.05 M) at 450 °C.
Subsequent
Chem. Lett. 2014, 43, 711–713 | doi:10.1246/cl.140074
© 2014 The Chemical Society of Japan | 711

Figure 3. Thermogravimetric analysis of PbI2¢DMF (N2;
¹
1
heating rate: 5 °C min ). Photos of lead iodide crystals before
(
left) and after (right) heating.
Thermogravimetric analysis of PbI2¢DMF (N2; heating rate:
¹
1
5
°C min ) showed that the release of coordinated DMF started
around 70 °C and was completed at 112.6 °C, concomitant with
a macroscopic color change of the sample from pale to deep
yellow (Figure 3). We concluded that PbI2 in DMF could
potentially be deposited initially onto the mesoporous TiO2
Figure 1. (a) Schematic cross section (left) and photographic
top view of the device (right). (b) Solar cell performance:
photocurrent density­voltage curves of perovskite-based solar
¹
2
surface as PbI ¢DMF, and coordinated DMF could be released
2
cells measured under 100 mW cm simulated AM 1.5 sun light
solid lines) and in the dark (dashed lines). Measured condition:
during the drying process. Considering this notion, we also
examined the effect of the drying temperature on the perform-
ance of the fabricated solar cell. Heating the cells to temper-
atures of 100 or 130 °C instead of 70 °C changed the color of
the PbI2 layer more rapidly and resulted in decreased VOC and
FF values (70 °C: VOC = 0.93­0.95 V, FF = 0.61­0.66; 100 °C:
VOC = 0.86­0.92 V, FF = 0.53­0.61; 130 °C: VOC = 0.83­
(
delay time, 1000 ms; step, 50 mV; forward scan (from negative
to positive voltage). Black lines: untreated PbI2 (1.0 M);
blue lines: dehydrated PbI2 (1.0 M); red lines: dehydrated PbI2
(1.1 M).
0
.91 V, FF = 0.52­0.60), accompanied by an increased variation
of these values relative to their position in the cells.
Details about the sequential deposition process, i.e., how
the perovskite structure of CH3NH3PbI3 is formed during the
dipping process of microcrystalline PbI2 in a CH3NH3I solution,
should also be interesting and important for the fabrication of
highly efficient solar cells. The substitution of a DMF molecule
in PbI2¢DMF by CH3NH3I leads to the perovskite CH3NH3PbI3,
the formation of which was evident from the observed color
change of the yellow PbI2¢DMF crystals to black immediately
after they were dipped into a solution of CH3NH3I in 2-propanol
Figure 2. X-ray structure of PbI2¢DMF. Selected bond lengths
¡): Pb(1)­I(1), 3.1563(4); Pb(1)­I(1)#1, 3.1662(4); Pb(1)­
I(2)#2, 3.2516(4); Pb(1)­I(2), 3.2746(4); Pb(1)­I(2)#3,
.3205(4); Pb(1)­O(1), 2.431(4); O(1)­C(1), 1.251(7).
(
1
6
(
Figure 4 and movie in the SI ). However, on cutting these
3
black crystals, we found the inside of these crystals to be still
yellow (Figure 4), indicating that the formation of CH3NH3PbI3
took place only on the surface of the crystals. During the
optimization of the cell fabrication process, we found the
optimum dipping period of the PbI2 substrates to be 20 s.
Prolonging the duration of the dipping (t > 20 s) also led to the
formation of black films, but the resulting cells often exhibited
low VOC values (t = 90 s, VOC = 0.87 V; t = 30 min, VOC =
0.61 V).
diffraction studies on several lead halide complexes, possibly
formed during the solution process discussed above.
After cooling a 1.0 M solution of PbI2 in DMF from 70 °C
to room temperature, needle-like, pale yellow crystals were
obtained. Interestingly, single-crystal X-ray diffraction analysis
of these crystals revealed that one DMF molecule was
coordinated to Pb to form one-dimensional structures of
18
PbI ¢DMF along the a axis (Figure 2). The Pb­I and
In spite of several trials, dropping crystals of PbI ¢DMF
2
2
Pb­O bond lengths observed were 3.1563(4)­3.3205(4) and
into a solution of CH3NH3I in 2-propanol did not produce single
crystals of CH3NH3PbI3 suitable for X-ray diffraction analysis.
However, pale yellow crystals were obtained from these
solutions during the collection of crystals under a microscope
2.431(4) ¡, respectively. The formation of a Pb­O coordination
bond implies the possibility of attractive interactions between Pb
and oxygen on the surface of the mesoporous TiO2.
712 | Chem. Lett. 2014, 43, 711–713 | doi:10.1246/cl.140074
© 2014 The Chemical Society of Japan

In summary, we examined the influence of the water content
in the lead halide starting material on the fabrication process of
perovskite-based solar cells. Further improvements of materials
and device structures are currently in progress in our laboratories
and will be reported in due course.
This study was supported by the PRESTO project of the
Japan Science and Technology Agency (JST). We would like
to thank Prof. Y. Tachibana at RMIT University for fruitful
discussions on spray pyrolysis and Prof S. Yoshikawa at Kyoto
University for the use of vacuum deposition equipment.
Figure 4. Pictures of PbI2¢DMF crystals before and after
dipping into a solution of CH3NH3I in 2-propanol (t = 5 s).
(a)
References and Notes
1
2
3
A. Kojima, K. Teshima, Y. Shirai, T. Miyasaka, J. Am. Chem.
Soc. 2009, 131, 6050.
M. M. Lee, J. Teuscher, T. Miyasaka, T. N. Murakami, H. J.
Snaith, Science 2012, 338, 643.
H.-S. Kim, C.-R. Lee, J.-H. Im, K.-B. Lee, T. Moehl, A.
Marchioro, S.-J. Moon, R. Humphry-Baker, J.-H. Yum, J. E.
Moser, M. Grätzel, N.-G. Park, Sci. Rep. 2012, 2, 591.
J. H. Noh, S. H. Im, J. H. Heo, T. N. Mandal, S. I. Seok, Nano
Lett. 2013, 13, 1764.
J. H. Heo, S. H. Im, J. H. Noh, T. N. Mandal, C.-S. Lim, J. A.
Chang, Y. H. Lee, H.-j. Kim, A. Sarkar, M. K. Nazeeruddin,
M. Grätzel, S. I. Seok, Nat. Photonics 2013, 7, 486.
J. Burschka, N. Pellet, S.-J. Moon, R. Humphry-Baker, P. Gao,
M. K. Nazeeruddin, M. Grätzel, Nature 2013, 499, 316.
M. Liu, M. B. Johnston, H. J. Snaith, Nature 2013, 501, 395.
G. Xing, N. Mathews, S. Sun, S. S. Lim, Y. M. Lam, M.
Gräzel, S. Mhaisalkar, T. C. Sum, Science 2013, 342, 344.
S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou, M. J. P.
Alcocer, T. Leijtens, L. M. Herz, A. Petrozza, H. J. Snaith,
Science 2013, 342, 341.
4
5
(b)
6
7
8
9
1
0 W. A. Laban, L. Etgar, Energy Environ. Sci. 2013, 6, 3249.
Figure 5. Distribution of performance parameters for perov-
skite-based solar cells (optimized fabrication protocol).
11 Z. Ku, Y. Rong, M. Xu, T. Liu, H. Han, Sci. Rep. 2013, 3,
132.
3
1
2 P. Docampo, J. M. Ball, M. Darwich, G. E. Eperon, H. J.
Snaith, Nat. Commun. 2013, 4, 2761.
in air. X-ray structure analysis revealed that these crystals were
II
composed of (CH3NH3)4PbI6¢2H2O, containing isolated Pb I6
13 G. Li, R. Zhu, Y. Yang, Nat. Photonics 2012, 6, 153.
14 A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo, H. Pettersson,
Chem. Rev. 2010, 110, 6595.
1
9
octahedra (SI). These crystals have previously been reported
to be obtained from aqueous solutions of Pb(NO3)2 and
2
0
CH3NH3I. The water molecules observed in our crystals were
most likely introduced by exposing the sample to air. This result
again supports the importance of the use of dehydrated solvents
and starting materials for the preparation of the perovskite
layers, as the procedure to obtain perovskite crystals from
the reaction of an excess amount of CH3NH3I with PbI2 is
comparable to that on the surface of our films during the dipping
15 Y. Tachibana, K. Umekita, Y. Otsuka, S. Kuwabata, J. Phys. D:
Appl. Phys. 2008, 41, 102002.
16 Supporting Information is available electronically on the CSJ-
Journal Web site, http://www.csj.jp/journals/chem-lett/index.
html.
17 J. H. Noh, N. J. Jeon, Y. C. Choi, M. K. Nazeeruddin, M.
Grätzel, S. I. Seok, J. Mater. Chem. A 2013, 1, 11842.
18 X-ray crystallographic data for PbI2¢DMF; monoclinic, P21/n,
a = 4.53800(10) ¡, b = 18.2704(6) ¡, c = 11.5897(4) ¡, ¢ =
process. In the presence of nonnegligible amounts of H O and
excess amounts of CH3NH3I, lead halide complexes containing
isolated Pb I6 octahedra, such as (CH3NH3)4PbI6¢2H2O, could
2
II
3
¹3
92.690(2)°, V = 959.86(5) ¡ , Z = 4, Dcalcd = 3.696 g cm ,
potentially be formed on the surface of the perovskite films. This
observation also underlines the importance of rinsing the films
with 2-propanol after the dipping process.
Based on these findings, we optimized the fabrication
protocol as described above, which allowed the reproducible
production of highly efficient solar cells with PCEs more than
R1 = 0.0205, wR2 = 0.0467 [I > 2·(I)], GOF = 1.092
(CCDC-982210).
19 X-ray crystallographic data for (CH3NH3)4PbI6¢2H2O;
monoclinic, P21/n, a = 10.3937(3) ¡, b = 11.3055(2) ¡, c =
3
10.5519(2) ¡, ¢ = 91.2981(13)°, V = 1239.59(5) ¡ , Z = 2,
¹3
Dcalcd = 3.035 g cm , R1 = 0.0168, wR2 = 0.0348 [I > 2·(I)],
GOF = 1.022 (CCDC-982211).
10% (Figure 5). We also confirmed that these cells retained their
performance without sealing for at least 3 months, when they
were stored under an inert atmosphere (SI).
20 B. R. Vincent, K. N. Robertson, T. S. Cameron, O. Knop, Can.
J. Chem. 1987, 65, 1042.
Chem. Lett. 2014, 43, 711–713 | doi:10.1246/cl.140074
© 2014 The Chemical Society of Japan | 713