Systematic studies on chain lengths, halide species, and well thicknesses for lead halide layered perovskite thin films

Article abstract of DOI:10.1246/bcsj.79.1607

Two-dimensional layered perovskite compounds, (C_nH _2n+1NH₃)₂(CH₃NH₃) _m-1Pb_mX_3m+1 (n = 2, 3, 4, 6, and 10; X = Cl, Br, and I; m = 1, 2, and 3) were systematically prepared. The influences of the barrier-size, halide species, and well thickness of the perovskite thin films on the quantum confinement structures were investigated. The layered perovskite films showed a strong and clear absorption peak due to excitons confined in inorganic quantum-wells. The exciton peak shifted to lower energy as the halide species was changed from Cl to Br and I. Furthermore, fine multilayer perovskite compound films were prepared by varying the spin-coating conditions.

Full text of DOI:10.1246/bcsj.79.1607

ꢀ 2006 The Chemical Society of Japan
Bull. Chem. Soc. Jpn. Vol. 79, No. 10, 1607–1613 (2006) 1607
Systematic Studies on Chain Lengths, Halide Species, and Well
Thicknesses for Lead Halide Layered Perovskite Thin Films
ꢀ1;2
Yuko Takeoka,
Keisuke Asai,3 Masahiro Rikukawa,1 and Kohei Sanui1
1Department of Chemistry, Sophia University, 7-1 Kioi-cho, Chiyoda-ku, Tokyo 102-8554
2Department of Applied Chemistry, Tohoku University, Aoba, Aramaki, Aoba-ku, Sendai 980-8579
3PRESTO, Japan Science and Technology Agency (JST), 4-1-8 Hon-cho, Kawaguchi 332-0012
Received March 10, 2006; E-mail: y-tabuch@sophia.ac.jp
Two-dimensional layered perovskite compounds, (CnH2nþ1NH3)2(CH3NH3)mꢁ1PbmX3mþ1 (n ¼ 2, 3, 4, 6, and 10;
X ¼ Cl, Br, and I; m ¼ 1, 2, and 3) were systematically prepared. The influences of the barrier-size, halide species, and
well thickness of the perovskite thin films on the quantum confinement structures were investigated. The layered
perovskite films showed a strong and clear absorption peak due to excitons confined in inorganic quantum-wells. The
exciton peak shifted to lower energy as the halide species was changed from Cl to Br and I. Furthermore, fine multilayer
perovskite compound films were prepared by varying the spin-coating conditions.
Low-dimensional systems can be readily constructed by a
simple method involving organic–inorganic hybrid materials
without using the large-scale pieces of apparatus needed by
conventional techniques.1,2 Organic materials with specific
functionality are combined on a molecular scale with an inor-
ganic matrix with other target properties, creating an organic–
inorganic hybrid with either a combination of useful properties
or new phenomena arising from the interaction between organ-
ic and inorganic components. Organic–inorganic perovskite-
type compounds are one of the versatile hybrid materials that
are currently studied by many groups due to their potential for
unique electrical, magnetic, and optical properties.3,4
The most commonly studied system with the general formu-
la (CnH2nþ1NH3)2PbX4 (X: halide) naturally forms a layered
structure consisting of two-dimensional sheets of [PbX6]4ꢁ
octahedra and organic ammonium (CnH2nþ1NH3þ) layers.5,6
The layered structure is self-organized via hydrogen bonds be-
tween the inorganic sheet halogens and the organic ammonium
groups and also via van der Waals interactions between adja-
cent organic tails. Since the bandgap of the organic layers is
much larger than that of the inorganic layer based on [PbX6]4ꢁ
octahedra (by at least 3 eV), the organic layers in this system
act as insulating barrier layers, and the inorganic layers behave
as semiconductor layers. The interfaces between the inorganic
and organic layers are intrinsically flat, and it is thought that
excitons are confined to behave as an ideal two-dimensional
system. In fact, the existence of very stable excitons with ex-
tremely large exciton binding energy (320 meV) and oscilla-
tor strength (ꢂ0:7) was reported for (C6H13NH3)2PbI4.7 The
stable excitons provide excellent optical properties, such as
strong photoluminescence and high optical nonlinearity, which
makes the family of organic–inorganic layered perovskites a
promising and growing research field.8–11
dimensional (3-D) structure is constructed by using smaller
þ
þ
15,16
alkylamines such as NH4 and CH3NH3
.
With this
approach, it is possible to create new materials with different
dimensional structures that vary from zero to three, all using
the same simple combination of organic amines and metal
halides.17 In addition to these one- to three-dimensional struc-
tures, the existence of intermediate dimensionality structures
with variable well thicknesses, (CnH2nþ1NH3)2(CH3NH3)mꢁ1
-
MmX3mþ1, has been reported.18 In these structures, methylam-
monium (CH3NH3þ) fits into the center of eight MX6 corner-
þ
shared octahedra, while a long-chain alkylammonium RNH3
fits only into the periphery of a set of four MX6 octahedra. To
maintain charge neutrality, methylammonium ions incorporate
into the inorganic framework and serve as counter-ions to
make the two layers cohere. The multilayer CH3NH3MX3 per-
ovskite sheets are sandwiched between the [CnH2nþ1NH3]þ
layers.19,20 Mitzi et al. also discovered that there was a metal–
semiconductor transition in (CnH2nþ1NH3)2(CH3NH3)mꢁ1Snm-
X3mþ1, dependent on the thickness of the inorganic perovskite
sheets.21,22 These intermediate dimensionality structures form
natural multi-quantum-well structures with variable well thick-
ness (controlled by m) and even well depth (controlled by the
halogen), providing an ideal model for tailoring the optical and
electrical properties.
To explore the effects of well-size, barrier-size, and well-
thickness (multilayer) on the structural, thermal, and optical
properties using this series of layered perovskite compounds,
we have performed extended systematical studies which are
shown in Scheme 1. Although several studies involving this
series of compounds have been reported, most of them focused
on one scientific matter, and it was difficult to investigate
the relationships between the properties and the structures
throughout the entire compound and to show the correlations
among the three different phases. Extremely pure compounds,
(CnH2nþ1NH3)2(CH3NH3)2Pb3X10 (X ¼ Cl, Br, and I), with
multi-layer inorganic sheets (m ꢃ 3) have not yet been
Another unique feature of these materials is the tunability of
their nano structures.12–14 For example, although two-dimen-
sional structures are found in (CnH2nþ1NH3)2PbX4, a three-
Published on the web October 10, 2006; doi:10.1246/bcsj.79.1607

1608 Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006)
Lead Halide Layered Perovskite Thin Films
powder that found was collected by filtration. CnPbI4 (n ¼ 2
and 3) cannot be collected as powder samples. CnPbBr4 and
CnPbCl4 were prepared in a similar manner to CnPbI4, but the sol-
vents were different. N,N-Dimethylformamide (DMF) was used as
the reaction solvent for CnPbX4 (X ¼ Br and Cl), and powder
samples were obtained by pouring the reaction solutions into ace-
tone. The yield of CnPbX4 was ca. 70%. Anal. Calcd for C12H32-
N2PbBr4: C, 19.71; H, 4.41; N, 3.83%. Found: C, 19.80; H, 4.47;
N, 3.80%.
2D Normal
4-
Well layer (PbX6
Barrier layer
)
)
4-
Well layer (PbX6
Bilayer Compounds, (C6H13NH3)2(CH3NH3)Pb2X7. The
bilayer (m ¼ 2) perovskite, (C6H13NH3)2(CH3NH3)Pb2I7 (C6-
Pb2I7), was prepared by reacting stoichiometric amounts of
C6H13NH3I and CH3NH3I with PbI2. CH3NH3I (2.5 mmol) and
C6H13NH3I (5.0 mmol) were dissolved in DMF (20 mL) at
35 ꢄC in advance. After the two amines were dissolved completely
in DMF, PbI2 (5.0 mmol) was added to the solutions and reacted
for 1 h under nitrogen. Then, the reaction solutions were poured
into nitromethane (X ¼ I) to precipitate the product. Centrifugal
separation was used because the bilayer compounds were too fine
to separate by filtration. The analogous halide compounds,
C6Pb2X7 (X ¼ Br and Cl) were synthesized similarly, except that
the reaction solutions were poured into acetone to precipitate the
product. The yield of C6Pb2X7 was approximately 60%. Anal.
Calcd for C9H27N3Pb2Br7: C, 9.37; H, 2.62; N, 3.64%. Found:
C, 9.36; H, 2.59; N, 3.67%.
+
Halogen
(X = I, Cl, Br)
CnH2n+1NH3
+
CH3NH3
(n = 2, 3, 4, 6, 10)
Barrier Thickness
Well Size
Well Thickness
Trilayer Compounds, (C6H13NH3)2(CH3NH3)2Pb3X10. Pure
trilayer compound, (C6H13NH3)2(CH3NH3)2Pb3I10 (C6Pb3I10),
was obtained only when the reaction was carried out at 35 ꢄC
with a feed ratio of C6H13NH3I:CH3NH3I:PbI2 = 2:3:3. CH3-
NH3I (7.5 mmol) and C6H13NH3I (5.0 mmol) were at first dis-
solved in DMF (30 mL) at 35 ꢄC. After the two amines were com-
pletely dissolved, PbI2 (7.5 mmol) was added to the mixed solu-
tions. The reaction solutions were poured into nitromethane.
C6Pb3I10 was obtained as a powder by centrifugation in a proce-
dure similar to that described above. The compounds C6Pb3X10
(X ¼ Br and Cl) were synthesized similarly except that acetone
was used instead of nitromethane. The yield of C6Pb3X10 was
approximately 60%.
Three-Dimensional Compounds CH3NH3PbX3. The three-
dimensional perovskite (m ¼ 1) CH3NH3PbI3 was synthesized
by reacting a stoichiometric amount of PbI2 with CH3NH3I. A
mixed solution of CH3NH3I (5.0 mmol) and PbI2 (5.0 mmol) in
DMF (10 mL) was prepared at room temperature. When the
reaction solution was poured into nitromethane, microcrystalline
CH3NH3PbI3 precipitated. The compounds CH3NH3PbX3 (X ¼
Br and Cl) were synthesized similarly except for the use of
acetone for the reprecipitation. The yields of CH3NH3PbX3 were
approximately 90%. Anal. Calcd for CH6NPbBr3: C, 2.51; H,
1.25; N, 2.92%. Found: C, 2.41; H, 1.34; N, 2.81%.
Scheme 1. Schematic view of this study with layered perov-
skites which explores the effects of barrier-thickness, well-
size, and well-thickness on their structural, thermal, and
optical properties.
obtained due to the difficulty in stoichiometrically controlling
the inorganic and organic components. We, herein, attempt-
ed the fabrication of (CnH2nþ1NH3)2(CH3NH3)mꢁ1PbmX3mþ1
(X ¼ Cl, Br, and I), which has mono (m ¼ 1), bi (m ¼ 2),
and tri (m ¼ 3) layer structures of inorganic two-dimensional
sheets. Furthermore, a systematic study was undertaken to
determine whether desired crystalline structures could be pre-
pared and to establish the optimum conditions for the prepara-
tion of multi-layered compounds.
Experimental
Materials. n-Alkylamines and hydrohalogenic acids (HCl
35 wt %, HBr 48 wt %, and HI 57 wt %) were obtained from Wako
Pure Chemical Industries, Ltd. Lead halides (PbX2, X ¼ I, Br, and
Cl, purity ꢃ 99.99%) were purchased from Kojundo Chemical
Laboratory Co., Ltd. and were used as received. A series of alkyl-
ammonium salts were prepared by reacting n-alkylamine with a
stoichiometric amount of hydrohalogenic acid aqueous solution
for 3 h at room temperature, and the crude product was recrystal-
lized from methanol. The layered perovskite-type compounds
were prepared as follows.
Film Preparation. DMF solutions of crystalline powder were
prepared and were filtered with a 0.5 mm filter prior to use. Films
were fabricated on hydrophilic substrates by spin-coating at ap-
proximately 6000 r.p.m. using a MIKASA 1H-D7. The substrates
were heated at ca. 60 ꢄC during the spin-coating process to obtain
high quality films due to the high boiling point of DMF. C6Pb3X10
spin-coated films could not be obtained from solutions that were
prepared from the powder sample. The C6Pb3X10 films were
obtained directly using the reaction solution.
Monolayer Compounds, (CnH2nþ1NH3)2PbX4. (CnH2nþ1
-
NH3)2PbI4:
The PbI-based monolayer (m ¼ 1) perovskites,
(CnH2nþ1NH3)2PbI4 (CnPbI4, n ¼ 2, 3, 4, 6, and 10), were pre-
pared by treating the corresponding CnH2nþ1NH3I with a stoichio-
metric amount of PbI2. CnH2nþ1NH3I (5.0 mmol) was dissolved in
acetone (10 mL) at room temperature. PbI2 (1.15 g, 2.5 mmol) was
then added to this solution under nitrogen. The reaction solutions
were stirred for 1 h after the PbI2 was completely dissolved. The
solution was poured into nitromethane, and the microcrystalline
Structural Analysis.
Optical absorption spectra of the
spin-coated films were obtained on a U-3300 spectrophotometer
(Hitachi) at room temperature. Fluorescence spectra were record-
ed on an F-3000 spectrophotometer (Hitachi) at different excita-

Y. Takeoka et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006) 1609
Table 1. Interlayer Spacing Obtained from X-ray Dif-
fraction and Absorption Band of (CnH2nþ1NH3)2PbX4
(X ¼ Cl, Br, and I)
n = 2
˚
d-Spacing/A
Exciton peaks/nm
Br
n = 3
n = 4
n = 6
n
Cl
Br
I
Cl
I
2
3
4
6
10
11.3
12.6
14.0
18.4
25.2
11.6
12.6
14.0
18.0
23.9
11.9
13.0
13.8
16.1
21.5
329
331
330
330
329
392, 420
402
403
395
388
—
505
509
512
509
n = 10
Energy (eV)
4.03.5 3.0 2.5 2.0
5
10 15 20 25 30 35 40
2θ / degrees
(a) X = Cl
Fig. 1. X-ray diffraction patterns of (CnH2nþ1NH3)2PbBr4
(n ¼ 2, 3, 4, 6, and 10) spin-coated films at room tem-
perature.
(b) X = Br
(c) X = I
tion wavelengths (300 nm for X ¼ Br and Cl, 310 nm for X ¼ I).
X-ray diffraction measurements were performed over a 2ꢀ range
of 1.5 to 40ꢄ using a Rigaku RINT2000 diffractometer with a
˚
Ni-filtered copper Kꢁ target (ꢂ ¼ 1:5418 A) at 40 kV and 25–
40 mA. Simultaneous thermogravimetric and differential thermal
analyses (TG-DTA) was conducted for the freshly prepared
powders on TG-DTA200 (Seiko) under a nitrogen flow of 200
mL minꢁ1. Differential scanning calorimetry (DSC) thermograms
were also collected on a DSC200 (Seiko) under a nitrogen flow of
40 mL minꢁ1. The topology and microstructure of CnPbmX3mþ1
spin-coated films were examined by atomic force microscopy
(AFM). AFM measurements were performed on a Nanoscope IIIa
(Digital Instruments) operating in the tapping mode.
300
400
500
600
700
800
Wavelength/ nm
Fig. 2. Optical absorption spectra of (C6H13NH3)2PbX4
(X ¼ Cl, Br, and I) spin-coated films at room temperature.
˚
mal hydrocarbon chain is considered to increase by 2.5 A per
carbon atom, assuming that the alkyl chain is oriented with
the lowest energy. The increment of the layer spacing value
Results and Discussion
˚
b is much smaller than 2.5 A each case. As described above,
Synthesis and Structural Analysis of Monolayer Perov-
Layered perov-
the order of the b values is CnPbCl4 > CnPbBr4 > CnPbI4
within the n range examined in this study. Also, note that the
a value, which corresponds to the layer spacing of the three-
dimensional compounds, is larger than that of the correspond-
ing cubic perovskite compounds, and its order is CnPbCl4 <
CnPbBr4 < CnPbI4. These results suggest that the larger halo-
gen atom in the [PbX6]4ꢁ layer provides more space parallel
to the organic layer surface, which leads to the tilting and/or
interdigitating of the alkyl chains.
The optical absorption spectra of the monolayer (m ¼ 1)
compounds, C6PbX4, were measured at room temperature to
examine the effect of changing halogen atom on the optical
properties of these perovskite compounds. The monolayer
films have a strong and sharp absorption at 330 (X ¼ Cl),
395 (X ¼ Br), and 512 nm (X ¼ I), even at room temperature,
as shown in Fig. 2. Because neither PbX2 nor C6H13NH3X
have absorption peaks in the vicinity of the peaks exhibited
by C6PbX4, the characteristic peaks in the monolayer films
are attributed to an exciton in the [PbX6]4ꢁ inorganic layer,
which is sandwiched between organic barrier layers. The fact
that the excitons can be observed even at room temperature
is demonstrative of the high stability of the exciton states in
these monolayer films. The exciton peak red-shifted as the
halogen atom was changed from Cl to Br to I. This red shift
is attributed to the change of the energy-gap for the corre-
skite Compounds (CnH2nþ1NH3)2PbX4.
skite-type compounds, (CnH2nþ1NH3)2PbX4 (CnPbX4), were
systematically synthesized with various CnH2nþ1NH3X and
PbX2 materials. The powders that were obtained were white
(X ¼ Cl and Br) and orange (X ¼ I), and polycrystalline thin
films of the compounds were easily fabricated by spin-coating.
At first, to examine the effects of alkyl chain length on the lay-
ered structures, we prepared monolayer CnPbX4 films with
various alkylamines (n ¼ 2, 3, 4, 6, and 10) in the organic lay-
ers. Figure 1 shows the X-ray diffraction patterns obtained
from the CnPbBr4 spin-coated films that have PbBr4 inorganic
sheets. A series of (00l) diffraction peaks corresponding to the
interlayer spacing was clearly observed, indicating the forma-
tion of a two-dimensional layered structure for each sample.
The PbCl and PbI-based systems also exhibited similar X-
ray diffraction patterns, suggesting primarily c-axis orienta-
tion. Table 1 shows the interlayer spacings for CnPbX4 calcu-
lated from the X-ray diffraction patterns using the Bragg equa-
tion. Since the layer spacings increase linearly with increasing
the alkyl chain length in all these monolayer perovskite sam-
˚
ples, the interlayer spacing l can be expressed as: l (A) =
a þ bn, where n is the number of carbon atoms in the alkyl-
amine; a ¼ 7:32, 8.06, and 9.27; b ¼ 1:79, 1.59, and 1.21
for X ¼ Cl, Br, and I, respectively. The chain length of a nor-

1610 Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006)
Lead Halide Layered Perovskite Thin Films
sponding inorganic perovskites, indicating that the optical
properties can be tailored by changing the halogen species.
The effect of the distance between the inorganic layers of
two-dimensional materials on the electronic structures was
investigated by comparing the exciton peaks in the optical ab-
sorption spectra. The exciton absorption for CnPbX4 films var-
ied according to the changes in the halogen species, whereas
no significant change was observed in the exciton absorption
with any of the alkylamines examined. As summarized in
Table 1, the CnPbX4 films with n ¼ 3, 4, 6, and 10 had a single
clear exciton peak at around 510 nm for X ¼ I, at 395 nm for
X ¼ Br, and at 330 nm for X ¼ Cl. Even if the distance
between the adjacent lead halide layers decreases to n ¼ 3,
the qualitative features of the absorption spectra are essentially
the same. This demonstrates that the distance between the
inorganic layers with C3H7NH2 as the organic components is
enough to confine the exciton within the inorganic layers, in
other words, the interaction between the inorganic layers is too
weak to overlap with each other for CnPbX4 with three or more
alkyl chains. In contrast, the C2PbX4 film with C2H5NH2
exhibits a different phenomenon, i.e., the exciton absorption
is dependent on the halogen type. Although the C2PbCl4 film
had a single sharp peak attributable similar to a two-dimen-
sional structure, the C2PbBr4 film had two absorption peaks
at 390 and 402 nm that associated with the X-ray diffrac-
tion results, and the C2PbI4 has a very broad peak at around
520 nm (Fig. 3). This phenomenon can be explained by the
following. The distances between the inorganic quantum-wells
in C2PbBr4 and C2PbI4 films are not large enough to prevent
percolation of the wave function, which is due to the interac-
tion with the neighboring inorganic layers; therefore, these
films did not show a clear exciton absorption. This is caused
by the difference in the bandgap energy and electron affinity
between the halogens. Electron affinity (Ea) and ionic potential
(Ip) decrease as the halogen atom changes from Cl to Br to I.
Accordingly, the electronegativity ꢃ (¼ðIp þ EaÞ=2) decreases,
which means that the covalent bond character becomes strong-
er than the ionic character. Based on this reason, the ability
of fabricating a two-dimensional confinement structure for a
series of CnPbX4 follows Cl > Br > I. For the C2PbCl4 film
with a large bandgap energy, the quantum-well made up of
[PbCl6]4ꢁ is deep enough to prevent the percolation of the
wave function and to maintain the two-dimensional exciton
features.
Thermal Analysis of Monolayer Perovskite Compounds.
The thermal decomposition behavior of the monolayer perov-
skite compounds was investigated by simultaneous thermogra-
vimetric analysis (TGA) and differential thermal analysis
(DTA). The decomposition temperatures of the standard
layered compounds, C4PbX4, were 232, 276, and 270 ꢄC for
X ¼ Cl, Br, and I, respectively. The X ¼ I and Br compounds
with butylammonium cations tend to be more thermally stable
than that of X ¼ Cl. We observed weight loss (20%) upon
heating the C4PbBr4 to 310 ꢄC and a somewhat larger weight
loss above 310 ꢄC. If the observed weight loss is due to the
decomposition of the organic components (C4H9NH3þ), the
estimated weight loss of C4PbBr4 is 22%. Several thermal
studies on the layered perovskites with various alkylamines
were undertaken in which it was demonstrated that all of these
layered perovskite compounds are stable to 200 ꢄC. Also, note
that the thermal stability depends on the orientation of the
alkyl chains in the organic sheet. The Br and I compounds with
the larger tilt angles (smaller layer spacing) have the higher
thermal stability. The phase transition behavior of the layered
perovskite compounds based on PbX2 was also investigated.
Table 2 summarizes the thermal transition peaks observed in
the DSC thermograms of monolayer perovskites, CnPbX4.
The transition temperatures are dependent on the chain length
of the organic components and halogen atoms. C4PbX4
(X ¼ Cl, Br, and I) exhibited no structural phase transitions
below room temperature on the cooling scans. This is an
important feature for practical applications in optical devices
because the exciton peak does not skip at low temperatures.
Owing to the differences in the PbX2 matrices, the phase-tran-
sition behavior is not constant for a given chain length. Some
trends for the transition temperatures can be found in these
results. The CnPbX4 with relatively shorter or longer alkyl
chains exhibit plural and lower transition peaks. In other
words, there is a critical point in the transition temperatures
Table 2. Thermal Transitions of (CnH2nþ1NH3)2PbX4 from
DSC ThermogramsaÞ
Energy (eV)
CnPbX4
n
Transition point/ꢄC
4 3.5
3
2.5
2
Heating Scan
Cooling
(a) X ¼ Cl
3
4
6
5.8, 67, 96
53
ꢁ57, 4.5, 31
14, 55
61, 91
47
ꢁ20, 27
8.4, 49
(a) X = Cl
10
(b) X = Br
(b) X ¼ Br
(c) X ¼ I
3
4
6
10
ꢁ102, ꢁ5:3
110
ꢁ87, ꢁ74, ꢁ48, 24
ꢁ7:0, 54
ꢁ110, 3.6
107
ꢁ54, 22
ꢁ14, 45
(c) X = I
700
3
4
6
10
—
1.1
10, 83, 102
14, 55
—
ꢁ23
300
400
500
600
800
Wavelength (nm)
76, 98
8.4, 49
Fig. 3. Optical absorption spectra of (C2H5NH3)2PbX4
(X ¼ Cl, Br, and I) spin-coated films at room temperature.
a) (C3H7NH3)2PbI4 cannot be collected as a powder.

Y. Takeoka et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006) 1611
(001)
m = 1
(a) X = Cl
m = 1
(b) X = Br
(c) X = I
(002)
(003)
(16.4 Å)
(18.0 Å)
m = 1
(18.4 Å)
(004) (005)
m = 2
m = 2
m = 2
(23.2 Å)
(22.6 Å)
(24.5 Å)
m = 3
m = 3
m = 3
(28.0 Å)
(28.5 Å)
(29.4 Å)
5
10 15 20 25 30 35 40
5
10 15 20 25 30 35 40
/ degrees
5
10 15 20 25 30 35 40
2θ
Fig. 4. X-ray diffraction patterns of (C6H13NH3)2(CH3NH3)mꢁ1PbmX3mþ1 (m ¼ 1, 2, and 3) spin-coated films with (a) X ¼ Cl,
(b) Br, and (c) I at room temperature.
for the series of perovskites with different alkyl chains. As
Table 3. Interlayer Spacing Obtained from X-ray Diffrac-
noted above, the C4PbX4 had a single and relatively high tran-
tion and Absorption Band of (C6H13NH3)2(CH3NH3)mꢁ1
-
sition point on the both scans. This phenomenon can be ex-
plained by assuming that the spatial organization of the organ-
ic layers is determined by the size of halogen atoms and the
length of alkyl chains. When a larger halogen atom is used
for the fabrication of the perovskite, the space for the organic
layers in the organic–inorganic frameworks necessarily enlarg-
es. Shorter alkyl chains are loosely packed; consequently, plu-
ral and lower transition temperatures are observed as described
above. In contrast, longer alkyl chains readily tilt and/or inter-
digitate, and a part of them are also loosely packed. Therefore,
they have plural and lower transition points. Thus, the use of
alkyl ammoniums with the intermediate chain lengths and
adequate halogen atoms has somewhat tighter packing than
for the other chain lengths, resulting in thermally stable perov-
skite compounds with single and higher transition points.
Synthesis and Structural Analysis of Multilayered Per-
ovskite Compounds. Multilayered perovskite compounds
with different inorganic layer thicknesses, (CnH2nþ1NH3)2-
(CH3NH3)mꢁ1PbmX3mþ1 (Abbreviated to CnPbmX3mþ1, m ¼
1{3), were fabricated using two alkylammonium materials
having different alkyl chain lengths, (CnH2nþ1NH2 and
CH3NH2), with a controlled ratio between them. Monolayer
(m ¼ 1) and bilayer (m ¼ 2) perovskite compounds were pre-
pared by the stoichiometric reaction (CnH2nþ1NH2:CH3NH2 =
2:0 and 2:1), whereas pure trilayer (m ¼ 3) compounds can not
be obtained by the stoichiometric reaction. The optimum ex-
perimental conditions for preparing the trilayer films were in-
vestigated by reacting CnH2nþ1NH3X, CH3NH3X, and PbX2 at
various molar ratios and directly spin-coating the reacting so-
lutions. Note that the CnH2nþ1NH3X/PbX2 molar ratio was
fixed at 1.5 and the CnH2nþ1NH3X:CH3NH3X:PbX2 molar ra-
tios were changed within 2:2 to 3.5:3. Small variations in the
molar ratio had significant effects on the structure of multilayer
compounds. It was found from the optical analysis and struc-
tural results that the spin-coated films fabricated from a mixed
solution of CnH2nþ1NH3X:CH3NH3X:PbX2 in a molar ratio
at 2:3:3 had the purest trilayer structure. Moreover, similar re-
sults were obtained for all other halogen analogues (X ¼ Cl,
Br, and I). Each solution of CnH2nþ1NH3X and CH3NH3X
was maintained at a temperature of 35 ꢄC before adding the
PbX2 powder to obtain high quality trilayer, CnPb3X10, films.
PbmX3mþ1 (X ¼ Cl, Br, and I)
˚
d-Spacing/A
Cl Br
Exciton peaks/nm
Sample
I
Cl
Br
I
Monolayer (m ¼ 1) 18.4 18.0 16.4 330
395
431
450
528
512
569
604
740
Bilayer (m ¼ 2)
Trilayer (m ¼ 3)
Cubic (m ¼ 1)
25.2 23.2 22.6 349
29.4 28.5 28.0 356
5.68 5.80 6.28 397
Otherwise, other multilayer compounds CnPbmX3mþ1 (m ¼ 1,
2, and ꢃ 4) are easily produced as contaminants. Figure 4
shows the X-ray diffraction patterns of C6PbmX3mþ1 spin-coat-
ed films with m ¼ 1, 2, and 3 fabricated using optimum condi-
tions. A series of diffraction patterns of the (00l) plane is clear-
ly observed in all films. This observation strongly supports the
facts that these films are highly oriented with the c-axis per-
pendicular to the substrate and that the perovskite films are
well-crystallized. In the cases of m ¼ 2 and 3, no diffraction
peaks corresponding to the lower order or the higher order
structure were detected by the X-ray diffraction measurements.
The interlayer d-spacing values calculated from the X-ray dif-
˚
fraction results were 18.0, 23.2, and 28.5 A for C6PbmBr3mþ1
(m ¼ 1, 2, and 3), respectively. The increments in the layer
˚
spacings observed between each system were 5.2 A for the
˚
mono-bilayer and 5.3 A for the bi-trilayer. The thickness of
one [PbBr6]4ꢁ octahedra layer was calculated to be 5.9 A based
˚
on the lattice constant of the cubic perovskite CH3NH3PbBr3.23
Therefore, these X-ray diffraction results clearly demonstrate
that the obtained spin-coated films consist of multilayered
structure with mono, bi, or trilayers of inorganic sheets. Table 3
lists the d-spacing values of these multilayer films as calculat-
ed from the X-ray diffraction measurements. Although a sim-
ilar increment in the d-spacing was observed in the X-ray dif-
fraction patterns, the Cl and I systems showed longer and
shorter increments than the thickness of [PbX6]4ꢁ calculated
from the cubic perovskite compounds.
Morphology of Two-Dimensional Thin Films with Multi-
layer Quantum-Wells. Figure 5 shows the AFM topology
images of C6PbmBr3mþ1 (m ¼ 1, 2, 3, and 1). In each spin-
coated film, polycrystalline grains covered the substrate, while
grain morphology was quite different and was dependent on

1612 Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006)
Lead Halide Layered Perovskite Thin Films
between [PbX6]4ꢁ octahedra becomes larger. The red shifts
of exciton peaks from monolayer to bilayer compounds can
be explained by this model. However, it is impossible to apply
this model to the trilayer compounds because only the interac-
tion between adjacent [PbX6]4ꢁ octahedra is taken into consid-
eration in this model. The [PbX6]4ꢁ octahedra in the trilayer
compounds have been divided into two types: Octahedra
in the exterior layers is contact with five [PbX6]4ꢁ units and
the ones in the center layers is contact with six units. The
transfer energy of the former type is similar to the bilayer com-
pounds, while the latter type is similar to the 3D compounds. If
we take this model into consideration, the trilayer compounds
are thought to have two absorption bands. The trilayer com-
pounds actually have a single absorption peak, which does
not originate from the bilayer or cubic compounds. Therefore,
an alternative picture based on delocalized excitations is need-
ed in the compounds with m ꢃ 3. In other words, a unified
picture combining the localized excitations for m ꢅ 2 and
the delocalized ones for m ꢃ 3 should be constructed. A shift
in the exciton peak was observed as the halogen was changed
from Cl to Br to I, which was identical to those described in
Fig. 2. The excitonic bands in the chloride compounds arise
at smaller energies than those of the corresponding bromide
and iodide compounds. This is due to the decrease in the band-
gap as the halogen changes from Cl to Br and I. While the
exciton peaks of the chlorides are sharper than those of the
bromides, the exciton peaks assigned to impurities (the lower
multilayer compound) are rarely observed in the m ¼ 2 and
3 chloride compounds.
The multilayer bromide compounds, C6PbmBr3mþ1, films
exhibited photoluminescence strong enough to detect with a
naked eye even at room temperature. A strong and sharp peak
of monolayer C6PbBr4 was observed at 406 nm with a full
width of 120.1 meV at 295 K. This PL band is consistent with
free exciton emission, because the Stokes shift of this film is as
narrow as 40 meV. With respect to C6PbmBr3mþ1 (m ¼ 2 and
3) compounds, photoluminescence exhibited a pronounced
spectral peak in the visible range with the maximum wave-
lengths at 437 and 459 nm, respectively. The emission of mul-
tilayer compounds was red-shifted by about 33 and 55 nm rel-
ative to that of the monolayer C6PbBr4. Although the absolute
efficiency was not determined, the photoluminescence from
the m ¼ 2 and 3 spin-coated films was significantly weaker
than that of the C6PbBr4 films. The exciton state in the multi-
layered perovskite compounds with thicker inorganic layers is
less stable, and the emission associated with this state was cor-
respondingly red-shifted because the bandgap energy and the
exciton binding energy decrease with increasing m as a result
of quantum confinement or dimensionality effects.
(a), (b)
(a)
(b)
400 nm
0 nm
(c)
(d)
(c), (d)
1000 nm
0 nm
75 µm × 75 µm
Fig. 5. Tapping mode AFM topology images of (C6H13-
NH3)2(CH3NH3)mꢁ1PbmBr3mþ1 spin-coated films with
(a) m ¼ 1, (b) m ¼ 2, (c) m ¼ 3, and (d) m ¼ 1.
the compositions. The monolayer compounds form plate-like
crystals with an average long side length of about 10 mm
(Fig. 5a), highlighting the fact that a layered structure can be
formed with the perovskite sheets approximately parallel to
the substrate. This is consistent with the results that the X-
ray diffraction patterns in which only (00l) reflections were ob-
served in Fig. 4. In the m ¼ 1 compounds (Fig. 5d), square-
like crystals were observed with an average dimension of
3–5 mm. The morphology of the layered perovskites with
multi-layer quantum wells was affected dramatically by the in-
organic sheet width m in the layered structure. For the bilayer
compounds, the surface was covered by fine grains like a her-
ringbone, which proves the fact that the bilayer compounds are
too fine to collect by filtration, as described in Experimental
Section. By contrast, plate-like grains were also observed for
the trilayer films. The morphology of trilayer films appeared
to be smaller and to distribute more randomly compared to
the monolayer films. Since excess CH3NH3Br may exist in
the trilayer films, it would prevent the crystallization and
growth of the trilayer crystals. This result was consistent with
the fact that the half-width of the X-ray diffraction peak for the
trilayer film is slightly wider than that of the mono or bilayer
films, as shown in Fig. 4.
Optical Properties of Two-Dimensional Thin Films with
Multilayer Quantum-Wells. The optical absorption spectra
of the C6PbmX3mþ1 (m ¼ 1, 2, 3, and 1; X ¼ Cl, Br, and I)
spin-coated films were measured at room temperature. The
exciton absorption peaks of these films are listed in Table 3.
Each spin-coated film has a strong absorption peak corre-
sponding to the lowest exciton states; the form of the peaks
is due to the inorganic semiconductor wells confined by elec-
tronic potential and dielectric confinement. In all of the halo-
gen systems, the absorption peaks of C6PbmX3mþ1 red-shifted
as the inorganic layer thickness, m, increased. There are two
possible reasons for the red shift. The first reason is a decrease
in the quantum confinement effect due to the expansion of the
exciton Bohr radius, resulting in a decrease in exciton binding
energy. The second reason is a decrease in the bandgap.24 As
the inorganic sheet thickness increases, the conduction and
valence bands become wider because the transfer energy
Conclusion
It has been demonstrated that the two-dimensional structure
of lead halide perovskite compounds can be systematically
varied over a wide range by the simple combination of organic
alkylammonium and lead halide. The thin films of these perov-
skite compounds were fabricated by the conventional spin-coat
method. Through this approach, it is possible to provide poten-
tial thin films for optical and electrical devices and to vary the
key characteristics such as the distance between the inorganic

Y. Takeoka et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006) 1613
Phys. Lett. 1995, 66, 1169.
sheets, the thermal properties and stabilities, and the level of
quantum confinement effect. There have been numerous re-
ports related to these perovskite compounds. However, a com-
prehensive study of the simple two-dimensional layered per-
ovskite compounds, (CnH2nþ1NH3)2(CH3NH3)mꢁ1PbmX3mþ1
(n ¼ 2, 3, 4, 6, and 10; X ¼ Cl, Br, and I; m ¼ 1, 2, 3, and
1), involving the structural study, thermal properties and sta-
bility, fundamental optical properties, and quantum confine-
ment effect, has not yet been published. Although little success
in the use of these materials in devices has been achieved thus
far, much effort should be devoted to try to understand the
structural, optical, and electrical properties in order to materi-
alize the potential applications. This fundamental and other
systematic studies have revealed many new scientific facts.
The size of the halogen atom in the layered perovskite com-
pounds affects not only the optical properties but also the
formation of molecular cavity consisting of the organic com-
ponent, therefore resulting in the different organization of
alkyl chains corresponding to the cavity size. It was thought
that the quantum confinement effect was independent on the
depth of organic barrier layer with a certain of alkyl length.
However, wave function percolation was observed with the
(C2H5NH3)2PbBr4 films even though the film formed a layered
structure. Fine (C2H5NH3)2PbX4 (X ¼ Cl and I) films could
not to be fabricated regardless of the sizes. Furthermore, fine
multilayer perovskite compound films could be prepared by
varying the spin-coating conditions. The ability to manipulate
the multilayer structure in these compounds will make it pos-
sible to fine-tune the optical properties such as absorption and
photoluminescence.
3
48, 1.
4
125.
5
For a recent review: D. B. Mitzi, Prog. Inorg. Chem. 1999,
G. C. Papavassiliou, Prog. Solid State Chem. 1997, 25,
Y. I. Dolchenko, T. Inabe, Y. Maruyama, Bull. Chem. Soc.
Jpn. 1986, 59, 563.
6
S. S. Nagapetyan, Y. I. Dolchenko, E. R. Arakelova, V. M.
Koshkin, Y. T. Strchkov, V. E. Schklover, Russ. J. Inorg. Chem.
1988, 33, 1614.
7
1989, 69, 933.
T. Ishihara, J. Takahashi, T. Goto, Solid State Commun.
8
9
D. B. Mitzi, Chem. Mater. 1996, 8, 791.
C.-Q. Xu, T. Kondo, H. Sakakura, K. Kumata, Y.
Takahashi, R. Ito, Solid State Commun. 1991, 79, 245.
10 M. Era, S. Morimoto, T. Tsutsui, S. Saito, Appl. Phys. Lett.
1994, 65, 676.
11 K. Shibuya, M. Koshimizu, Y. Takeoka, K. Asai, Nucl.
Instrum. Methods Phys. Res., Sect. B 2002, 194, 207.
12 Y. Takeoka, K. Asai, M. Rikukawa, K. Sanui, Chem.
Commun. 2001, 2592.
13 T. Matsui, A. Yamaguchi, Y. Takeoka, M. Rikukawa,
K. Sanui, Chem. Commun. 2002, 1094.
14 Y. Takeoka, M. Fukasawa, T. Matsui, M. Rikukawa,
K. Sanui, Chem. Commun. 2005, 378.
15 D. Weber, Z. Naturforsch. B: Chem. Sci. 1978, 33, 1443.
16 A. Oglish, D. Weber, J. Chem. Phys. 1987, 87, 6373.
17 Y. Takeoka, K. Asai, M. Rikukawa, K. Sanui, Chem. Lett.
2005, 34, 602.
18 J. Calabrese, N. L. Jones, R. L. Harlow, N. Herron, D. L.
Thorn, Y. Wang, J. Am. Chem. Soc. 1991, 113, 2328.
19 G. C. Papavassiliou, A. O. Patsis, D. J. Lagouvardos, I. B.
Koustselas, Synth. Met. 1993, 55–57, 3889.
20 G. C. Papavassiliou, I. B. Koutselas, Synth. Met. 1995, 71,
1713.
21 D. B. Mitzi, C. A. Field, W. T. A. Harrison, A. M. Guloy,
Nature 1994, 369, 467.
This work was supported by the Precursory Research for
Embryonic Science and Technology (PRESTO) and Core
Research for Evolutional Science and Technology (CREST)
of the Japan Science and Technology Agency (JST). The
authors gratefully acknowledge Dr. F. S. Howell for his help
with English corrections.
22 D. B. Mitzi, S. Wang, C. A. Field, C. A. Chess, A. M.
Guloy, Science 1995, 267, 1473.
23 A. Poglitsh, D. Weber, J. Chem. Phys. 1987, 87, 6376.
24 T. Ishihara, in Optical Properties of Low-Dimensional
Materials, ed. by T. Ogawa, Y. Kanematsu, World Scientific,
Singapore, 1995, p. 288.
References
1
C. B. Murray, C. R. Kagan, M. G. Bawendi, Science 1995,
270, 1335.
Y. Sorek, R. Reisfeld, I. Finkelstein, S. Ruschin, Appl.
2