Ruddlesden-Popper Hybrid Lead Iodide Perovskite 2D Homologous Semiconductors

Article abstract of DOI:10.1021/acs.chemmater.6b00847

The hybrid two-dimensional (2D) halide perovskites have recently drawn significant interest because they can serve as excellent photoabsorbers in perovskite solar cells. Here we present the large scale synthesis, crystal structure, and optical characterization of the 2D (CH₃(CH₂)₃NH₃)₂(CH₃NH₃)_n-1Pb_nI_3n+1 (n = 1, 2, 3, 4, ∞) perovskites, a family of layered compounds with tunable semiconductor characteristics. These materials consist of well-defined inorganic perovskite layers intercalated with bulky butylammonium cations that act as spacers between these fragments, adopting the crystal structure of the Ruddlesden-Popper type. We find that the perovskite thickness (n) can be synthetically controlled by adjusting the ratio between the spacer cation and the small organic cation, thus allowing the isolation of compounds in pure form and large scale. The orthorhombic crystal structures of (CH₃(CH₂)₃NH₃)₂(CH₃NH₃)Pb₂I₇ (n = 2, Cc2m; a = 8.9470(4), b = 39.347(2) ?, c = 8.8589(6)), (CH₃(CH₂)₃NH₃)₂(CH₃NH₃)₂Pb₃I₁₀ (n = 3, C2cb; a = 8.9275(6), b = 51.959(4) ?, c = 8.8777(6)), and (CH₃(CH₂)₃NH₃)₂(CH₃NH₃)₃Pb₄I₁₃ (n = 4, Cc2m; a = 8.9274(4), b = 64.383(4) ?, c = 8.8816(4)) have been solved by single-crystal X-ray diffraction and are reported here for the first time. The compounds are noncentrosymmetric, as supported by measurements of the nonlinear optical properties of the compounds and density functional theory (DFT) calculations. The band gaps of the series change progressively between 2.43 eV for the n = 1 member to 1.50 eV for the n = ∞ adopting intermediate values of 2.17 eV (n = 2), 2.03 eV (n = 3), and 1.91 eV (n = 4) for those between the two compositional extrema. DFT calculations confirm this experimental trend and predict a direct band gap for all the members of the Ruddlesden-Popper series. The estimated effective masses have values of m_h = 0.14 m₀ and m_e = 0.08 m₀ for holes and electrons, respectively, and are found to be nearly composition independent. The band gaps of higher n members indicate that these compounds can be used as efficient light absorbers in solar cells, which offer better solution processability and good environmental stability. The compounds exhibit intense room-temperature photoluminescence with emission wavelengths consistent with their energy gaps, 2.35 eV (n = 1), 2.12 eV (n = 2), 2.01 eV (n = 3), and 1.90 eV (n = 4) and point to their potential use in light-emitting diodes. In addition, owing to the low dimensionality and the difference in dielectric properties between the organic spacers and the inorganic perovskite layers, these compounds are naturally occurring multiple quantum well structures, which give rise to stable excitons at room temperature.

Full text of DOI:10.1021/acs.chemmater.6b00847

Article
pubs.acs.org/cm
Ruddlesden−Popper Hybrid Lead Iodide Perovskite 2D Homologous
Semiconductors
Constantinos C. Stoumpos, Duyen H. Cao, Daniel J. Clark, Joshua Young, James M. Rondinelli,⊥
†
,¶
†,¶
‡
,†
§,⊥
‡
†
Joon I. Jang, Joseph T. Hupp, and Mercouri G. Kanatzidis*
†
Department of Chemistry, Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208, United States
‡
Department of Physics, Applied Physics and Astronomy, Binghamton University, P.O. Box 6000, Binghamton, New York 13902,
United States
§
Department of Materials Science and Engineering, Drexel University, 3141 Chestnut Street, Philadelphia, Pennsylvania 19102,
United States
⊥
Department of Materials Science and Engineering, Northwestern University, 2220 Campus Drive, Evanston, Illinois 60208,
United States
*
S Supporting Information
ABSTRACT: The hybrid two-dimensional (2D) halide perovskites
have recently drawn significant interest because they can serve as
excellent photoabsorbers in perovskite solar cells. Here we present the
large scale synthesis, crystal structure, and optical characterization of
the 2D (CH (CH ) NH ) (CH NH ) Pb I (n = 1, 2, 3, 4, ∞)
3
2 3
3 2
3
3 n−1 n 3n+1
perovskites, a family of layered compounds with tunable semi-
conductor characteristics. These materials consist of well-defined
inorganic perovskite layers intercalated with bulky butylammonium
cations that act as spacers between these fragments, adopting the
crystal structure of the Ruddlesden−Popper type. We find that the
perovskite thickness (n) can be synthetically controlled by adjusting
the ratio between the spacer cation and the small organic cation, thus
allowing the isolation of compounds in pure form and large scale. The
orthorhombic crystal structures of (CH (CH ) NH ) (CH NH )-
3
2 3
3 2
3
3
Pb I (n = 2, Cc2m; a = 8.9470(4), b = 39.347(2) Å, c = 8.8589(6)), (CH (CH ) NH ) (CH NH ) Pb I (n = 3, C2cb; a =
2
7
3
2 3
3 2
3
3 2 3 10
8
.9275(6), b = 51.959(4) Å, c = 8.8777(6)), and (CH (CH ) NH ) (CH NH ) Pb I (n = 4, Cc2m; a = 8.9274(4), b = 64.383(4) Å,
3 2 3 3 2 3 3 3 4 13
c = 8.8816(4)) have been solved by single-crystal X-ray diffraction and are reported here for the first time. The compounds are
noncentrosymmetric, as supported by measurements of the nonlinear optical properties of the compounds and density functional
theory (DFT) calculations. The band gaps of the series change progressively between 2.43 eV for the n = 1 member to 1.50 eV for the
n = ∞ adopting intermediate values of 2.17 eV (n = 2), 2.03 eV (n = 3), and 1.91 eV (n = 4) for those between the two compositional
extrema. DFT calculations confirm this experimental trend and predict a direct band gap for all the members of the Ruddlesden−
Popper series. The estimated effective masses have values of m = 0.14 m and m = 0.08 m for holes and electrons, respectively, and
h
0
e
0
are found to be nearly composition independent. The band gaps of higher n members indicate that these compounds can be used as
efficient light absorbers in solar cells, which offer better solution processability and good environmental stability. The compounds
exhibit intense room-temperature photoluminescence with emission wavelengths consistent with their energy gaps, 2.35 eV (n = 1),
2.12 eV (n = 2), 2.01 eV (n = 3), and 1.90 eV (n = 4) and point to their potential use in light-emitting diodes. In addition, owing to
the low dimensionality and the difference in dielectric properties between the organic spacers and the inorganic perovskite layers,
these compounds are naturally occurring multiple quantum well structures, which give rise to stable excitons at room temperature.
INTRODUCTION
and CsSnI as an efficient hole transporter in dye-sensitized
3
■
10,11
solar cells (DSCs).
These were followed by the utilization
The class of halide perovskite compounds of the chemical
+
+
3
+
2 2
2+
formula AMX (A = Cs , CH NH , or HC(NH ) ; M = Ge ,
of CH NH PbI as a light absorbing material producing solid-
3
3
3
3
3
2
+
2+
−
−
−
12−14
Sn , Pb ; X = Cl , Br , I ) has witnessed a spectacular surge
in scientific interest in the last five years and has enabled
revolutionary achievements in the field of solid-state photo-
state solar cells with remarkable efficiency in late 2012.
1
−9
voltaics.
The recent strong scientific activity in the halide
Received: February 28, 2016
Revised: March 29, 2016
perovskites was triggered by the successful demonstration of
1
−9
CH NH PbX (X = Br, I) as an efficient light absorber
3
3
3
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
3
1
Despite the astonishing photovoltaic performance, reach-
single layer. The exploration of this exciting phenomenon drew
the attention of many researchers who devised novel experi-
48,49
15−17
ing ∼20% conversion efficiency in 2015,
some drawbacks
of CH NH PbI have become evident, with major concerns
42−47
ments
and performed extensive theoretical analyses
3
3
3
focusing on the insufficient long-term stability of the devices
incorporating these materials and the toxicity of lead.
Although the latter problem could be potentially resolved by
aiming to understand and control the origins of the unique
properties of the 2D perovskites. The most remarkable finding
was that the stability of the exciton does not arise from the
dimensional confinement alone but that the organic material
plays an important role by modulating the dielectric properties of
18,19
replacing Pb with the more environmentally friendly Sn metal
20−25
(a technology still in its infancy),
the former has not yet been
50,51
addressed in a satisfactory manner. One solution to the long-term
stability problem has been recently proposed in the form of the
layered iodide perovskites, which are two-dimensional (2D)
derivatives of the three-dimensional (3D) perovskite formed
the material.
The stability of the exciton in the n = 1 2D
perovskites gives rise to an intense photoluminescence (PL),
which persists even at room temperature. As a result, the 2D
perovskites are frequently employed in optical and electronic
26,27
34
by “slicing” the 3D frameworks into well-defined 2D slabs.
devices such as field-effect transistors (FETs), light-emitting
31
52
The 2D derivatives in hybrid lead iodides have not been studied
extensively; however, they offer far more tunability and flexibility
in terms of being able to control the physical properties.
diodes (LEDs), and hard radiation detectors. The higher n
members of the homologous series, however, remain less
53−55
explored.
They are now of intense interest because of
Further motivation for examining the 2D hybrid iodide
perovskites in much greater detail comes from recent processing-
dependent optical responses observed in 3D perovskite films. In
particular, we previously observed the formation of red colored
films upon spin-coating CH NH PbI solutions onto meso-
their tunable optical properties, which have led to the fabrication
26,27
of promising solar cells,
and studies of their photophysical
56−58
properties.
The lack of extensive synthetic and structural
information on the higher 2D perovskites further prompted us to
investigate the Ruddlesden−Popper homologous series.
In this article, we report the scalable synthesis, crystal
structure, and optical properties of the homologous 2D series of
3
3
3
porous TiO films prior to the formation of the expected black
2
CH NH PbI film, in addition to slightly varying absorption
3
3
3
28
spectra of CH NH PbI films with different PbI fraction.
lead iodide perovskites (CH (CH ) NH ) (CH NH )
3 2 3 3 2 3 3 n−1
30,46,53,56
3
3
3
2
These observations led us to the hypothesis that some inter-
mediate 2D compounds may be formed during the film fabrication
process and that these optical responses could be due to quantum
confinement effects in either nanoscale variants of perovskite
CH NH PbI or the intrinsic properties of 2D metastable deri-
Pb I
.
We find that the (CH (CH ) NH ) (CH NH )-
3 2 3 3 2 3 3
n 3n+1
Pb I (n = 2), (CH (CH ) NH ) (CH NH ) Pb I (n = 3),
2
7
3
2 3
3 2
3
3 2 3 10
and (CH (CH ) NH ) (CH NH ) Pb I (n = 4) members
3
2 3
3 2
3
3 3
4 13
combine the structural features of the simple 2D (n = 1) and 3D
(n = ∞) perovskite end members. We further find that these
intermediate phases exhibit tunable band gaps resulting from
quantum confinement due to the dimensional reduction of the
perovskite spacer layers. Furthermore, we show that the n = 2−4
compounds display optoelectronic properties that are different
from either the n = 1 or n = ∞ end-members. Thus, they define a
novel class of naturally forming superlattice semiconductors with
thickness-dependent optical properties reminiscent of the
3
3
3
vative phases. Such confinement effects studied through pair
29
distribution function (PDF) analysis on CH NH PbI films
3
3
3
arise from a structurally disordered perovskite dimensionally
confined within the nanosized TiO pores and are responsible
2
for a significant blue shift (∼50 meV) of the optical absorption
edge. Thus, we set out to study in detail the 2D perovskites
as a well-defined example of naturally forming semiconductor
materials with quantum confinement.
59,60
complex quantum-wells in AlGaAs/GaAs heterostructures.
The 2D perovskites have the generic chemical formula of
(
RNH ) (A)n−1MX
(n is an integer), where RNH is a
3
2
3n+1 3
EXPERIMENTAL SECTION
■
primary aliphatic or aromatic alkylammonium cation acting as a
spacer between the perovskite layers, and the A and M cations
and X anions form the perovskite framework. The 2D network
consists of inorganic perovskite layers of corner-sharing [MX6]
octahedra confined between interdigitating bilayers of long chain
alkylammonium cations. The unit layers are held together by a
combination of Coulombic and hydrophobic forces, which maintain
the structural integrity. The existence of the 2D homologous
M = Pb series of halide perovskites (RNH ) (A)n−1MX3n+1 has
been known for more than 25 years
demonstrated for the M = Sn series.
Starting Materials. All chemicals were purchased from Sigma-
Aldrich and used as received. Methylammonium iodide (MAI) was
synthesized by neutralizing equimolar amounts of a 57% w/w aqueous
hydriodic acid (HI) and 40% w/w aqueous methylamine (CH NH )
4−
3
2
(pH ≈ 7). The white precipitate was collected by evaporation of the
30
solvent using rotary evaporation at 60 °C under reduced pressure.
For convenience, we denote the 3D CH
NH
3
PbI
perovskite as
family
3
3
3
MAPbI and the 2D (n-CH (CH ) NH ) (CH NH ) Pb I
3
3
2
3
3
2
3
n−1
n 3n+1
as (BA) (MA) Pb I (n = 4, 3, 2, 1) throughout.
2
n−1
n 3n+1
3
2
30,31
Syntheses. MAPbI (n = ∞). PbO powder (2232 mg, 10 mmol) was
3
and was subsequently
Although these 2D
32−34
dissolved in a mixture of 57% w/w aqueous HI solution (10.0 mL,
7
6 mmol) and 50% aqueous H PO (1.7 mL, 15.5 mmol) by heating to
3 2
halide perovskites are analogues to the oxide perovskites
described by Ruddlesden and Popper,
studied compounds in the class of 2D halide perovskites are the
boiling under constant magnetic stirring for about 5 min, which formed
a bright yellow solution. Subsequent addition of solid CH NH Cl
35−37
the most well-
3
3
(675 mg, 10 mmol) to the hot yellow solution initially caused the
precipitation of a black powder, which rapidly redissolved under stirring
to afford a clear bright yellow solution. The stirring was then
discontinued, and the solution was left to cool to room temperature
and left to stand overnight to afford black polyhedral crystals. The
crystals were collected by suction filtration and dried under reduced
pressure. Yield 3.8 g (60%). Diffuse reflectance infrared Fourier
38
(
RNH ) MI (n = 1) family, where M is a group 14 or
3 2 4
39
lanthanide metal, and little is known for the higher members.
The initial discovery of the 2D perovskites immediately drew
attention because these systems can be regarded as natural
multiple-quantum-wells in which the semiconducting inorganic
layers act as potential “wells” and the insulating organic layers act
−1
transformed (DRIFT) spectrum, (KBr, cm ): 3180br, 2823w, 2711w,
4
0,41
as potential “barriers”.
The electronic confinement in the
2
485w, 2383w, 1820w, 1581s, 1467s, 1248m, 960s, 910s, 490m.
BA) (MA) Pb I (n = 4). PbO powder (2232 mg, 10 mmol) was
perovskite 2D semiconductors in subnanometer layers induces
the generation of stable excitons with unusually high binding
energy and a Bohr radius that extends beyond the limits of a
(
2
3
4 13
dissolved in a mixture of 57% w/w aqueous HI solution (10.0 mL,
76 mmol) and 50% aqueous H PO (1.7 mL, 15.5 mmol) by heating to
3
2
B
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
boiling under constant magnetic stirring for about 5 min, which formed
a bright yellow solution. Subsequent addition of solid CH NH Cl
solution initially produced a black precipitate, which was subsequently
dissolved under heating the combined solution to boiling. The stirring
was then discontinued, and the solution was left to cool to room
temperature during which time orange rectangular-shaped plates started
to crystallize. The precipitation was deemed to be complete after ∼2 h.
The crystals were isolated by suction filtration and thoroughly dried
under reduced pressure. Yield 3.5 g (49% based on total Pb content).
DRIFT spectrum, (KBr, cm ): 3025br, 2960w, 2929w, 2880w, 2739w,
2472m, 2389w, 1830w, 1572s, 1475s, 1385w, 1390m, 1255w, 1159m,
1081m, 1047m, 1004m, 968m, 920s, 787w, 737m, 474m.
3
3
(
507 mg, 7.5 mmol) to the hot yellow solution initially caused
the precipitation of a black powder, which rapidly redissolved under
stirring to afford a clear bright yellow solution. In a separate beaker,
n-CH (CH ) NH (248 μL, 2.5 mmol)was neutralized with HI 57% w/w
3
2
3
2
(
5 mL, 38 mmol) in an ice bath resulting in a clear pale yellow solution.
−1
Addition of the n-CH (CH ) NH I solution to the PbI solution initially
3
2
3
3
2
produced a black precipitate, which subsequently dissolved under
heating the combined solution to boiling. The stirring was then
discontinued, and the solution was left to cool to room temperature
during which time black rectangular-shaped plates started to crystallize.
The precipitation was deemed to be complete after ∼2 h. The crystals
were isolated by suction filtration and thoroughly dried under reduced
pressure. Yield 2.1 g (33% based on total Pb content). DRIFT spectrum,
Characterization. Single-crystal X-ray diffraction data were
collected using an image plate STOE IPDS II diffractometer using
Mo Kα radiation (λ = 0.71073 Å), operating at 50 kV and 40 mA. Data
reduction and numerical absorption corrections were performed using
the X-AREA suite. Single-crystals of the (BA) (MA) Pb I
2
n−1
n 3n+1
−1
(
KBr, cm ): 3174br, 2962w, 2929w, 2725m, 2485w, 2382w, 1815w,
compounds were mounted on the end of the glass tip directly from
the mother liquor, and the use of glue or dispersion oil was avoided due
to the clear deterioration of the crystals. All structures were solved by
1
7
577s, 1468s, 1250m, 1149m, 1072m, 1034m, 960m, 912s, 785w,
35w, 476m.
2
(
BA) (MA) Pb I (n = 3). PbO powder (2232 mg, 10 mmol) was
direct methods and refined by full-matrix least-squares on F using the
2
2
3 10
6
1
62
dissolved in a mixture of 57% w/w aqueous HI solution (10.0 mL,
6 mmol) and 50% aqueous H PO (1.7 mL, 15.5 mmol) by heating to
SHELXTL-2013 program package. The PLATON functions
63
7
operating within the WinGX platform were used to ensure the
validity of the refined crystal structures.
3
2
boiling under constant magnetic stirring for about 5 min, which formed
a bright yellow solution. Subsequent addition of solid CH NH Cl
The structures were refined based on the following considerations.
The perovskite lattice {PbnI3n+1} was refined anisotropically without any
constraints on the Pb and I atoms. The highly anisotropic shape of the
thermal ellipsoids of coordinated iodide ions appears due to a high
thermal motion relative to the Pb−I bond vector. This is typical for
the halide perovskites structures at room temperature, and therefore
3
3
(
450 mg, 6.67 mmol) to the hot yellow solution initially caused
the precipitation of a black powder, which rapidly redissolved under
stirring to afford a clear bright yellow solution. In a separate beaker,
n-CH (CH ) NH (327 μL, 3.33 mmol) was neutralized with HI 57%
3
2
3
2
w/w (5 mL, 38 mmol) in an ice bath resulting in a clear pale yellow
6
4
solution. Addition of the n-CH (CH ) NH I solution to the PbI
no special treatment was used. All organic atoms were refined
isotropically. Restraints were applied to the C−C and C−N bond
lengths. Each cation was treated as having equivalent thermal parameters.
Significant disorder exists in the interlayer cations, particularly for the
CH CH − tail of butylammonium (the NH CH CH − head is
3
2
3
3
2
solution initially produced a black precipitate, which was subsequently
dissolved under heating the combined solution to boiling. The stirring
was then discontinued, and the solution was left to cool to room
temperature during which time deep-red/purple rectangular-shaped
plates started to crystallize. The precipitation was deemed to be
complete after ∼2 h. The crystals were isolated by suction filtration and
thoroughly dried under reduced pressure. Yield 2.5 g (36% based on
3
2
3
2
2
relatively stable), causing the atoms to move and destabilize the
refinement. The bond length restraints cannot account for the motion of
the interlayer cations. Therefore, the cations were modeled manually on
idealized positions based on reasonable bond length and bond angle
parameters. Thus, the whole butylammonium cations were constructed
on top of the mirror planes imposed by the space group by slightly
modifying the positions of the off-plane Q-peaks found in the electron
density map. The mirror planes have the (x, y, 1/4)/(x, y, 3/4) and (x, y,
1/2) coordinates for the Ccmm (and Cc2m) and Acam space groups,
respectively, whereas in the case of C2cb where no mirror plane is
crystallographically imposed, the cations were modeled based on the initial
−1
total Pb content). DRIFT spectrum, (KBr, cm ): 3174br, 2962w,
2
1
7
929w, 2873w, 2713m, 2463w, 2382w, 1810w, 1573s, 1468s, 1389w,
335w, 1255w, 1149m, 1072m, 1017m, 1001m, 964m, 914s, 785w,
48w, 735w, 484m.
(
BA) (MA)Pb I (n = 2). PbO powder (2232 mg, 10 mmol) was
2 2 7
dissolved in a mixture of 57% w/w aqueous HI solution (10.0 mL,
6 mmol) and 50% aqueous H PO (1.7 mL, 15.5 mmol) by heating to
7
3
2
boiling under constant magnetic stirring for about 5 min, which formed
a bright yellow solution. Subsequent addition of solid CH NH Cl
Q-peaks. In the final steps of the refinement, the {(CH NH ) Pb I }
3
3
3
3 n−1 n 3n+1
(
338 mg, 5 mmol) to the hot yellow solution initially caused the
structure segment was refined without restraints, while the butylammonium
cations were fixed. The purpose of this treatment was to generate
chemically reasonable models without disorder. The modeled cations
increase the R-values slightly compared to the unrestrained or dis-
ordered refinement and make the e.s.u.’s unrealistically small. Because of
this, a BLOC instruction was used to finalize the refinement, which fixed
the positions of the butylammonium cations and refined the rest of the
atoms separately, to maintain the proper e.s.u.’s of the perovskite layers.
Powder and film X-ray diffraction patterns were collected using a
Rigaku MiniFlex 600 X-ray diffractometer (Cu Kα, 1.5406 Å) operating
at 40 kV and 15 mA.
precipitation of a black powder, which rapidly redissolved under stirring
to afford a clear bright yellow solution. In a separate beaker,
n-CH (CH ) NH (694 μL, 7 mmol) was neutralized with HI 57%
w/w (5 mL, 38 mmol) in an ice bath resulting in a clear pale yellow
solution. Addition of the n-CH (CH ) NH I solution to the PbI
solution initially produced a black precipitate, which was subsequently
dissolved under heating the combined solution to boiling. The stirring
was then discontinued, and the solution was left to cool to room
temperature during which time cherry red rectangular-shaped plates
started to crystallize. The precipitation was deemed to be complete after
3
2
3
2
3
2
3
3
2
∼
2 h. The crystals were isolated by suction filtration and thoroughly
Scanning electron microscope (SEM) images were acquired at an
accelerating voltage of 10−20 kV using either a Hitachi SU8030 or a
Hitachi SU3400 instruments equipped with Oxford X-max 80 SDD EDS
detectors.
Optical diffuse-reflectance spectra were collected at room tempera-
ture using a Shimadzu UV-3600 PC double-beam, double-mono-
dried under reduced pressure. Yield 3.0 g (41% based on total Pb
−1
content). DRIFT spectrum, (KBr, cm ): 3174br, 2962w, 2929w,
2
1
7
880w, 2721w, 2474w, 2389w, 1808w, 1574s, 1468s, 1385w, 1335w,
255w, 1149m, 1070m, 1017m, 1002m, 968m, 914s, 785w, 748m,
37m, 476m.
(
BA) PbI (n = 1). PbO powder (2232 mg, 10 mmol) was dissolved in
chromator spectrophotometer on powdered samples using BaSO as a
100% reflectance reference. The samples were irradiated with a halogen
2
4
4
a mixture of 57% w/w aqueous HI solution (10.0 mL, 76 mmol) and
0% aqueous H PO (1.7 mL, 15.5 mmol) by heating to boiling under
5
(NIR/Visible, 50W, 2000H) and a D lamp (UV, 2000H), and the
3
2
2
constant magnetic stirring for about 5 min, which formed a bright yellow
solution. Subsequent addition of liquid. In a separate beaker,
n-CH (CH ) NH (924 μL, 10 mmol) was neutralized with HI 57%
spectra were recorded using a PbS photoconductive element (NIR) and
a PMT (R928) detectors from 2500−200 nm. Band gaps were
6
5,66
determined as described elsewhere.
3
2
3
2
w/w (5 mL, 38 mmol) in an ice bath resulting in a clear pale yellow
DRIFT spectra were recorded on a Nicolet 6700 IR spectrometer in
−1
solution. Addition of the n-CH (CH ) NH I solution to the PbI
the 400−4000 cm spectral region with KBr beam splitter.
3
2
3
3
2
C
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
a
Table 1. Crystal Data and Structure Refinement for (BA) (MA)n−1Pb I
(n = 2−4) at 293(2) K
2
n 3n+1
empirical formula
formula weight
(BA) (MA)Pb I (n = 2)
(BA) (MA) Pb I (n = 3)
(BA) (MA) Pb I (n = 4)
2
2 7
2
2
3 10
2
3
4 13
1483.04
293(2) K
0.71073 Å
2103.00
2722.95
temperature
wavelength
refinement method
cryst syst
full-matrix least-squares on F2
orthorhombic
Cc2m
orthorhombic
C2cb
orthorhombic
Cc2m
space group
color
red
dark red
8.9275(6)
51.959(4)
8.8777(6)
90°
black
unit cell dimensions
a, Å
8.9470(4)
39.347(2)
8.8589(6)
90°
8.9274(4)
64.383(3)
8.8816(4)
90°
b, Å
c, Å
α = β = γ, deg
volume, Å3
Z
3118.7(3)
4
4118.0(5)
5104.9(4)
4
4
3
density (calc), g/cm
absorption coeff., mm−
3.159
3.392
3.543
21.026
4640
1
17.712
19.739
F(000)
2560
3600
3
cryst size (mm )
0.140 × 0.081 × 0.025
2.521−25.000°
99.7%
0.098 × 0.092 × 0.052
1.568−24.996°
99.9%
0.079 × 0.073 × 0.009
2.471−24.999°
99.4%
θ range
completeness to θ
index ranges
−10 ≤ h ≤ 10
−10 ≤ h ≤ 10
−61 ≤ k ≤ 61
−10 ≤ l ≤ 10
12858
−10 ≤ h ≤ 10
−76 ≤ k ≤ 76
−10 ≤ l ≤ 10
13717
−
−
46 ≤ k ≤ 46
10 ≤ l ≤ 10
reflns collected
9452
independent reflns
2732 [Rint = 0.0365]
2732/3/79
3633 [Rint = 0.0722]
3633/3/84
1.103
4671 [Rint = 0.0506]
4671/7/135
0.986
data/restraints/parameters
GOF
1.145
final R indices
Robs = 0.0529
wRobs = 0.1394
Rall = 0.0638
wRall = 0.1478
0.933 and −1.687
Robs = 0.0781
wRobs = 0.1748
Rall = 0.1234
wRall = 0.1997
3.289 and −1.719
Robs = 0.0431
wRobs = 0.1123
Rall = 0.0864
wRall = 0.1387
[
I > 2σ(I)]
R indices [all data]
largest diff. peak and hole (e Å−3)
1.746 and −1.470
a
2
2
2
4
1/2
2
2
2
2
2
R = Σ∥F | − |F ∥/Σ|F |, wR = {Σ[w(|F | − |F | ) ]/Σ[w(|F | )]} and n = 2, w = 1/[σ (Fo ) + (0.0768P) + 22.2041P]; n = 3, w = 1/[σ (Fo ) +
o
c
o
o
c
o
2
2
2
2
2
2
(
0.0916P) + 15.8934P]; n = 4, w = 1/[σ (Fo ) + (0.0736P) ], where P = (Fo + 2Fc )/3.
Raman spectra were recorded on a DeltaNu Advantage NIR energy on the other hand was above the band gap (except (BA) PbI )
2
4
spectrometer equipped with a CW diode laser (785 nm, 60 mW)
and a CCD camera detector in a backscattered geometry. The
powdered samples were packed in standard melting point capillaries
(ℏωTHG = 2.07 eV), and therefore THG was significantly under-
estimated because of the self-absorption processes in the material.
However, this absorption effect was properly taken into account for the
comparison between SHG and THG using the absorbance data.
Computational Methods. We investigated three of the syn-
(0.8 mm ID).
Photoluminescence spectra were collected on oriented rectangular
crystals of (BA) (MA)n−1PbnI3n+1 (n = 4, 3, 2, 1) and rhombic
thesized compounds using density functional theory: (BA)
(BA) (MA) Pb 10, and (BA) (MA) Pb 13. All calculations were
performed with the Vienna ab initio Simulation Package (VASP)
using projector augmented-wave (PAW) potentials within the PBEsol
2 4
PbI ,
2
dodecahedral crystals of MAPbI using Horiba LabRam Evolution high-
2
2
3
I
2
3
I
4
3
69,70
resolution confocal Raman microscope spectrometer (600 g/mm
diffraction grating) equipped with a diode CW laser (473 nm,
7
1,72
2
5 mW) and a Synapse CCD camera. The incident laser beam was
exchange-correlation functional.
We initially generated a centro-
parallel to the (010) direction of the crystals and focused at ∼1 μm spot
size. Unless stated otherwise, the maximum power output of the laser
source was filtered to 0.1% of the maximum power output.
symmetric and a noncentrosymmetric structure for each of the three
compounds through manipulation of the organic cations, which was
inferred by the symmetries identified in the experimental struc-
ture refinements. Subsequently, we fixed the lattice parameters and cell
volume to those determined in our experiments (Table 1) and fully
relaxed the internal atomic positions until the forces were less than
Second harmonic generation (SHG) and third harmonic generation
(
THG) spectra were recorded on powdered (BA) (MA) Pb I
2 n−1 n 3n+1
perovskite samples and a powdered α-SiO reference for a particle size
2
−1
distribution from 90 μm−125 μm placed in standard melting point
1 meV Å using a 500 eV plane-wave cutoff and 5 × 1 × 5 Monkhorst−
73
capillaries (0.8 mm ID) using reflection geometry with a fiber-optic
Pack mesh to sample the Brillouin zone. We computed the density
of states and band structures in the noncentrosymmetric crystal
structures, which were found to be the lowest energy phases in each case
using a 5 × 1 × 5 Monkhorst−Pack mesh and increased 550 eV plane-
wave cutoff.
67,68
bundle as described previously.
The bundle was coupled to a Horiba
iHR320 spectrometer (600 g/mm diffraction grating) equipped with a
Synapse CCD camera. To prevent absorption effects during the SHG
measurement, the fundamental beam from an optical parametric
oscillator, pumped by a Nd:YAG laser, was tuned to λ = 1800 nm so that
the SHG photon energy (ℏωSHG = 1.38 eV) was below the band gap of
all the materials. The incident pulse energy was tuned to ∼30 μJ using
a Glan-Thompson polarizer with a beam spot size of ∼0.5 mm.
All detector exposure times were scaled to 60 s. The THG photon
RESULTS AND DISCUSSION
■
Synthesis. The 2D (CH (CH ) NH ) (CH NH )
3 2 3 3 2 3 3 n−1
Pb I
family of perovskite compounds (n = 1−4) was
n 3n+1
D
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
Figure 1. SEM images (top) and photographs of the (BA) (MA)n−1PbnI3n+1 perovskite crystals (bottom) (scale bars = 200 μm).
2
synthesized from a stoichiometric reaction between PbI2,
The above procedure consists of a scalable and efficient
method of preparing these layered compounds in pure form. The
reaction scheme described above can be conveniently used to
prepare the 2D perovskite members (n = 2, 3, and 4), which
contradicts a proposed axiom that Ruddlesden−Popper halide
CH NH I (MAI), and n-butylamine (BA). A homogeneous
3
3
solution of concentrated, I -free hydroiodic acid containing
2
stoichiometric amounts of PbI and CH NH I (MA), according
2
3
3
to the desired composition, was allowed to react with half the
stoichiometric amount of BA by addition of the neutralized base
into the boiling acid solution under vigorous stirring. This highly
exothermic reaction resulted in the formation of a clear, bright
yellow solution, which upon cooling to ambient conditions
precipitates into the layered perovskite compounds (n = 2−4) in
the form of colorful rectangular plates with the spectral range
spanning from red to black (Figure 1). We find that the use of
BA as the reaction limiting reagent is essential in obtaining the
compounds in pure form; at the same time, it is detrimental to
the reaction yield, which is limited to ∼50% based on the total
Pb content due to the high solubility of CH NH PbI in the
76
perovskites with n > 3 cannot be isolated in pure form. Note
that when we attempted to isolate higher members of the 2D
2+
+
series by adjusting the Pb /MA ratio, the reactions always
yielded the n = 4 member. Nevertheless, we believe that n > 4
members of the 2D series may be isolated by varying either
the nature of the “spacer” (the long chain amine) or the
“perovskitizer” (the small organic cation)a hypothesis
whose testing lies outside the scope of the present work. To
the best of our knowledge, this is the only method by which
the individual compounds with n > 2 can be isolated on a gram
scale. The availability of pure bulk materials with n > 2 facilitates
in-depth studies of their photophysical properties and the
deployment of the title compounds in future solid-state solar
3
3
3
HI/H PO solvent medium. The employment of this concept
3
2
takes advantage of the different solubility of the 2D perovskite
members in the solvent medium, which increases as the number
of perovskite layers increases up to a maximum solubility
obtained for the CH NH PbI end-member (∼0.6 M at the
27,77
cells.
Crystal Structure Description. Basic Structural Charac-
teristics. The crystal structures of the (BA) (MA) Pb I
n 3n+1
3
3
3
2
n−1
boiling point). Interestingly, the solubility trend is inverted in
intermediate polarity solvents such as acetone or acetonitrile.
In these solvents, the 2D compounds are relatively soluble,
where the member with the largest organic fraction becomes
the most soluble (i.e., n = 1). Using HI/H PO as a solvent
(n = 1−4) compounds are shown in Figure 2, and selected
crystallographic information is tabulated in Tables 1 and 2.
Detailed crystallographic data are provided in the Supporting
Information (Section S2). All compounds crystallize in
orthorhombic space groups; (BA) PbI4 crystallizes in the
3
2
2
medium, we find that if BA is used stoichiometrically or MA is
primitive centrosymmetric Pbca space group, as previously
2
6,54,74,75
38,78
used in excess, as has been previously recommended,
reported,
whereas (BA) (MA)Pb I , (BA) (MA) Pb I ,
2 2 7 2 2 3 10
then the end product is contaminated either with the (n−1)
member in the former case, which acts as a kinetic barrier, or
with the (n + 1) member in the latter case, which acts as a
thermodynamic sink during synthesis of the compounds. Thus,
the driving force to guide the reaction into a single n-member
product comes from careful control of the stoichiometry based
on the “limiting reagent” principle.
and (BA) (MA) Pb I , which have been determined in the
2 3 4 13
present study, crystallize in the polar (C ), base-centered Cc2m,
2v
C2cb, and Cc2m, space groups, respectively. The noncentrosym-
metric configurations represent the case where all MA cations
are oriented in the same direction or nearly so, which results
in an unquenched net dipole moment within the unit cell.
The respective centrosymmetric (D ) space groups Ccmm,
2
h
E
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
Figure 2. Crystal structures of the 2D lead iodide perovskites, (BA) (MA)n−1Pb I , extending from n = 1 to n = ∞. The L-value denotes the thickness
2
n 3n+1
of the inorganic layer in each compound. The numerical values refer to the distance between the terminal iodide ions of each layer and were determined
directly from the refined crystal structures.
3
0
Acam, and Ccmm, which provide an equally good structural
solution, were considered to be less likely on the basis of the well-
known classical crystallographic Hamilton statistical significance
BA cations. Similar to the parent compound, the layers consist
4
−
of tilted, corner-sharing [PbI6] octahedra that propagate in
two directions (the ac plane), whereas in the third dimension
(the b-axis), the octahedral sheets are physically disconnected by
the intercalated organic bilayers. The separation of the inorganic
layers results in an increase of the lattice constants as the
incorporation of the BA spacers requires a gap of ∼7.8 Å between
the perovskite layers; the simplest case of n = 1 has a thickness
of ∼6.4 Å, which roughly corresponds to the sum of the two
Pb−I bond lengths.
As the 2D perovskite layers grow thicker by introducing MA
cations in the crystal structure, the unit cell incrementally
expands by addition of a single perovskite layer at a time. These
changes in the unit cell can be monitored by X-ray diffraction
(XRD), which characteristically reveals an additional low angle
reflection for each added perovskite layer (Figure 3a,b). Namely,
the n = 2 analogue shows two evenly spaced reflections, the n = 3
analogue three, and the n = 4 analogue four reflections below
∼2θ = 14°. The 2θ cut off for numbering is based on the fact that
at this angle the d-spacing matches the distance between the
discrete perovskite layers in both the 2D [(111) reflection] and
3D perovskites [(110) reflection]. This simple numerical
estimation of the crystal structure could be a useful analytical
7
9,80
criteria,
whereas SHG measurements and DFT calculations,
independently, bear out this assignment. Nevertheless, the
structural solutions in the centrosymmetric space groups are also
provided as a supplement in this work (see Supporting
Information) in aid of the quest for addressing the ferroelectric
behavior of these compounds by other researchers, as will be
discussed in detail in the SHG properties section. We consider
the centrosymmetric crystal structures to represent the average
structure of the bulk materials, whereas the polar crystal
structures may be considered as domains of a polycrystal and
may account for the presence of ferroelectric domains where
local polarization can be observed. This is supported by our
0
K DFT calculations, presented below, which find the
noncentrosymmetric structures to be lower in energy than the
centrosymmetric phases.
The (BA) (MA) Pb I family joins a handful of other
n 3n+1
2
n−1
structurally characterized multilayer hybrid halide perovskites
n > 1) with a group 14 metal in the octahedral site. These
(
3
0
include the (n = 2) (PEA) (MA)Pb I (in P1
recently reported (n = 3) (PEA) (MA) Pb I (in P1)
̅
)
and the very
2
2 7
26
2
2
3 10
perovskites (PEA is the phenylethylammonium cation), although
the crystal structures have been only tentatively assigned. For the
tool for further study of the (BA) (MA) Pb I
or related
2
n−1
n 3n+1
Ruddlesden−Popper systems.
33
n = 3 Sn compound (BA) (MA) Sn I (in Cmca) and the n = 2
The reported Ruddlesden−Popper perovskites have the same
2
2
3 10
Pb compounds, (MBA) (MA)Pb I (MBA is the 4-methyl-
gross symmetry determined by the positions of the heavy atoms
2
2 7
81
−(n+1)
benzylammonium cation), (TMA) (MA)Pb I (TMA is the
-thienylmethylammonium cation), and (ABA) (MA)Pb I
ABA is the 4-ammoniobutanoic acid cation), the Pbcn,
(i.e., the [PbnI3n+1
]
layers), but the space groups differ
2
2 7
8
2
2
(
because the nature of the organic spacer indirectly dictates the
details of crystal symmetry of the final layered compound. This
occurs because the organic spacer component influences the
relative rotation and tilting of the [PbI6]4 octahedra rather than
2
2 7
8
3
Aba2 (≡ C2cb) and C2cb space groups, respectively, have
been rigorously assigned. The unit cell in all compositions
contains two inorganic layers and two organic bilayers
−
the MA cations in the cages or the “MAn−1Pb I ” layers
n 3n+1
(
4 f.u./cell) with the inorganic layer defining the ac plane
themselves. A similar point has been previously made for the case
78,84,85
(ab plane for n = 1) and the organic bilayers intercalating along
of the (CH (CH ) NH ) PbI (n = 1, m = 3−17) series,
3
2 m
3 2
4
the b-axis (c-axis for n = 1).
where it was shown that the compounds undergo a series of
temperature-induced, structural phase transitions, with the
critical temperature of the transition displaying a strong depen-
dence on the length of the alkyl chain (m-value). Interestingly,
the set of phase transitions in the 2D (RNH ) PbI perovskites
Each 2D inorganic layer relates to the tetragonal parent 3D
compound (n = ∞) by slicing the perovskite along the (110)
plane such that some of the oriented MA cations are partially
(n = 2, 3, 4, ...) or fully (n = 1) substituted by the terminal
3
2
4
F
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
Table 2. Selected Bond Lengths [Å] and Bond Angles [°] for (BA) (MA)n−1Pb I
at 293(2) K with Estimated Standard
2
n 3n+1
Deviations in Parentheses
a
b
(
BA) PbI (n = 1)
(BA) (MA)Pb I (n = 2)
(BA) (MA) Pb I (n = 3)
(BA) (MA) Pb I (n = 4)
MAPbI (n = ∞)
3
2
4
2
2 7
2
2
3 10
2
3
4 13
label
distances
label
distances
label
distances
label
distances
label
distances
Pb−I(1)
Pb−I(2)
Pb−I(2)′
3.205(3)
Pb(1)−I(7)
Pb(1)−I(1)
Pb(1)−I(2)
Pb(1)−I(5)
Pb(2)−I(6)
Pb(2)−I(3)
Pb(2)−I(4)
Pb(2)−I(5)
3.081(7)
Pb(1)−I(4)
Pb(1)−I(1)
3.065(8)
3.157(3)
3.249(8)
3.047(4)
3.152(6)
3.160(6)
3.170(6)
3.179(6)
Pb(1)−I(4)
Pb(1)−I(3)
Pb(1)−I(6)
Pb(1)−I(2)
Pb(2)−I(5)
Pb(2)−I(13)
Pb(2)−I(1)
Pb(2)−I(4)
Pb(3)−I(10)
Pb(3)−I(8)
Pb(3)−I(12)
Pb(3)−I(3)
Pb(4)−I(11)
Pb(4)−I(9)
Pb(4)−I(7)
Pb(4)−I(5)
3.12(2)
Pb−I(1)
Pb−I(1)′
Pb−I(2)
3.125(8)
3.196(8)
3.1613(8)
3.1554(18)
3.2072(18)
3.1561(14)
3.1749(16)
3.249(8)
3.14(2)
Pb(1)−I(4)′
Pb(2)−I(5)
Pb(2)−I(2)
Pb(2)−I(3)
Pb(2)−I(2)′
Pb(2)−I(3)′
3.154(8)
3.164(8)
3.122(17)
3.156(8)
3.170(8)
3.26(2)
3.075(6)
3.169(2)
3.1715(19)
3.278(8)
3.06(2)
3.161(8)
3.172(9)
3.27(2)
3.01(2)
3.164(8)
3.175(9)
3.330(15)
a
b
(
BA) PbI (n = 1)
(BA) (MA)Pb I (n = 2)
(BA) (MA) Pb I (n = 3)
(BA) (MA) Pb I (n = 4)
MAPbI (n = ∞)
3
2
4
2
2 7
2
2
3 10
2
3
4 13
label
angles
label
angles
label
angles
label
angles
166.1(12) Pb−I(1)-Pb′ 180
Pb(2)−I(2)−Pb(2)′ 172.28(17) Pb(1)−I(2)−Pb(1)′ 168.8(14) Pb−I(2)-Pb′ 163.55(6)
label
angles
Pb−I(2)-Pb′ 155.38(8) Pb(1)′-I(1)−Pb(1) 173.1(3)
Pb(1)−I(2)−Pb(1)′ 164.3(3)
Pb(1)−I(1)−Pb(2) 169.52(12) Pb(2)′-I(1)−Pb(2)
Pb(2)−I(3)−Pb(2)′ 165.1(4)
Pb(2)′-I(4)−Pb(2) 167.6(3)
Pb(1)−I(5)−Pb(2) 165.62(11)
Pb(2)−I(3)−Pb(2)′ 164.57(15) Pb(1)−I(3)−Pb(3)
167.6(12)
167.2(6)
165.0(9)
Pb(1)−I(4)−Pb(1)′ 171.3(3)
Pb(1)−I(4)−Pb(2)
Pb(2)−I(5)−Pb(4)
Pb(1)−I(6)−Pb(1)′ 172.4(12)
Pb(4)−I(7)−Pb(4)′ 162.6(14)
Pb(3)′-I(8)−Pb(3)
Pb(4)′-I(9)−Pb(4)
170.4(14)
172.9(14)
Pb(3)−I(12)−Pb(3)′ 165.5(13)
Pb(2)−I(13)−Pb(2)′ 172.7(14)
a
38
b
64
Centrosymmetric reference. Noncentrosymmetric reference.
strongly resembles the properties of the 3D APbI counterparts,
as a guiding principle in predicting the crystal structures of the
3
+
where small changes in the A cation (from Cs to CH NH to
higher n-members of the (BA) (MA) Pb I
family.
3
3
2
n−1
n 3n+1
HC(NH ) ) result in similar structural changes in the inorganic
Comparison between the 2D and 3D Perovskites. To
elaborate on the structural differences among the compounds, let
us first consider the fundamental monolayer consisting of
2
2
−
{
PbI } frameworks, where α-, β-, and γ-phases evolve as a
3
64
function of temperature. It is very likely that similar phase
transitions exist for the higher members of the 2D perovskite
series as well, but because of their complexity, we limit our
discussion to the room temperature structures of these com-
pounds in the present manuscript. Such considerations allow us
to suggest that the role of the A cations in the 3D perovskites is
assumed by the spacer cations in the 2D perovskites.
18‑
“[Pb I ] ” units shaded in yellow in Figure 4. In the simplest
9
36
case of the (BA) PbI (n = 1) compound (Figure 4, left column),
2
4
a single slab is isolated between the BA ions, which align
perpendicular to it as viewed along the [110] direction; the
+
3
-NH “head” of the ammonium cation points toward the center
of the rhombic cavity generated by four adjacent octahedra, while
Given the complexity in assigning the proper space group in
Ruddlesden−Popper perovskites, as discussed above, we
generated the simulated precession images of reciprocal space
from the raw experimental single-crystal diffraction data for the
the −CH3 “tail” points toward the interlayer space. The
+
accommodation of the -NH3 “head” in the near vicinity of the
corner-connected octahedra (presumably by electrostatic
attraction and aided by NH −H···I−Pb hydrogen bonding)
2
(
BA) (MA) Pb I family (Figure 3c,d). The data confirm
forming the perovskite slab forces the octahedra to tilt and
2
n−1
n 3n+1
the assigned space groups (Cc2m or Ccmm for n = 2, C2cb or
Acam for n = 3, and Cc2m or Ccmm for n = 4) based on
systematically absent reflection conditions; moreover, they
uncover an underlying trend that compounds with an odd
number of layers (n = 1, 3) systematically adopt a higher
symmetry configuration (in terms of increasing space group
number) in comparison with the structures consisting of an even
number of layers (n = 2, 4). Such a distinction between crystal
structures of the odd and even layers, already established for
induces a 155.20° Pb−I−Pb angle in (BA) PbI . It is important
2
4
to point out that the corner-connected octahedra in a single slab
always need to rotate in an opposite sense (out-of-phase) to
maintain their structural integrity. Interestingly, the other end-
member, MAPbI (n = ∞), exhibits similar behavior with respect
3
to octahedral tilting although in a more complex fashion as the
octahedral tilting also occurs in the third direction. To better
1
8‑
illustrate this point, we define the different “[Pb I ] ” units
9
36
expanding along the c-axis shaded in blue, green, or yellow for
86,87
18‑
oxide perovskites of the Ruddlesden−Popper type,
can serve
clarity purposes (Figure 4, right). The individual “[Pb I ] ”
9
36
G
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
Figure 3. (a, b) X-ray diffraction patterns (Cu Kα) of the (BA) (MA)n−1PbnI3n+1 perovskites: (a) Bulk powder diffraction and (b) close-up views of the
2
characteristic regions between 2θ = 2−15° (left) and 2θ = 25−30° (right) where some pronounced differences of the 2D versus 3D perovskites occur.
Miller hkl indices are included to describe the strong preferred orientation of the 2D materials. (c, d) Precession images of the BA MA Pb I
2
n−1
n 3n+1
simulated from single-crystal X-ray diffraction (Mo Kα). (c) Images along the (001) direction emphasizing on the characteristic allowed reflections and
reflection conditions for the assigned space groups. Note that for n = 1, the corresponding direction is (010). (d) Images along the (101) direction
̅
̅
emphasizing on the ordering of the 2D perovskites with increasing number of layers and with respect to the parent compounds n = 1 and n = ∞ (top).
The green circles correspond to the marked reflections in panel b, whereas the red circles highlight the number of distinct, extra reflections for the layered
̅
perovskites with respect to the 3D perovskite. Note that for n = 1 and n = ∞, the corresponding directions is (110), whereas for n = 4, the (101) direction
is shown for clarity.
Figure 4. Views of the (BA) PbI single-layer (n = 1), (BA) (MA) Pb I three-layer (n = 3) 2D, and MAPbI (n = ∞) 3D lead iodide perovskite crystal
2
4
2
2
3 10
3
18‑
structures, highlighting their three-dimensional distortions. The defined “[Pb I ] ” slabs represent a 3 × 3 building block of the perovskite in two
9
36
dimensions and are used as a basic model to visualize the distortion. A single slab is composed of nine corner-connected octahedra (yellow) that distort
out-of-phase along the perovskite plane. The tilting scheme in the n = 1 2D perovskite can be solely described by a single slab. As the perovskite thickness
18‑
27‑
increases, the “[Pb I ] ” slabs connect to form “[Pb I ] ” boxes in three dimensions, which are the representative building units for the three-layer
9
36
27 81
structure in the n = 3 2D perovskite and a fragment of the infinite layers in the n = ∞ 3D perovskite. This figure illustrates the difference in the distortion
modes of the n = 3 perovskite emphasizing the in-plane and out-of-plane views of the perovskite slabs. Slabs belonging in different layers have been
drawn in different colors (yellow, blue, green) to project the connectivity of the slabs in the third dimension.
H
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
Figure 5. Optical properties of the (BA) (MA)n−1PbnI3n+1 perovskites (for n = 1, 2, 3 4, ∞). (a) Optical absorption of polycrystalline samples obtained
2
2
65,66
from diffuse reflectance measurements converted using the Kubelka−Munk function (α/S = (1 − R) /2R).
crystals with the wide planar facets oriented perpendicular to the laser beam (λ = 473 nm) with the illumination vector pointing toward the (010) axis
(b) Photoluminescence of oriented
exc
for n = 2, 3, 4, (001) for n = 1 and (100) for n = ∞. Plots a and b are scaled to match; the dashed lines in panel a correspond to the photoluminescence
maximum in panel b and are used as a guide to the eye to illustrate the absorption features that correspond to the emission maxima. (c) Plot of the band
2
gap and photoluminescence versus layer thickness with the latter expressed in the form of 1/n . Note that in panels a−c the energy is plotted in a
reciprocal scale and the wavelength in a linear scale. (d) Plot of the exciton energy, defined by convention as the difference in energy of the absorption
2
edge to the emission maximum (see Table 3) versus layer thickness expressed in the form of 1/n .
2
7‑
slabs fuse along the [001] direction to form a “[Pb I ] ” box
observe that the configuration of the perovskite slab as well as
the perovskite box in BA (MA) Pb I differs from both of the
2
7 81
and 3D octahedral network within which the MA cations are
2
2
3 10
contained. The MA cations align along the crystallographic c-axis
parent compounds. The most obvious difference arises from
the fact that the out-of-phase rotation axis now coincides with the
orientation of BA cations’ (the [010] direction), whereas the
plane parallel to the MA cation (the [101] direction) remains
relatively undistorted. The term “relatively” is used because the
decrease in symmetry from tetragonal (n = ∞) to orthorhombic
(n = 3), which precludes a 180° Pb−I−Pb angle connectivity
between the octahedra, introduces a small relative rotation.
An examination of the placement of the -NH3+ “heads” reveals
that the heads of the BA cations appear optimally “centered” in
(BA) PbI with respect to the perovskite slab, whereas the MA
heads lose their ideal alignment by becoming slightly “off-
centered” with respect to the perovskite slabs (unlike β-MAPbI3).
In that sense, neither BA nor MA is in its preferred configuration,
with BA trying to distort the perovskite along the b-axis and MA
trying to distort the ac-plane.
+
with the -NH groups again, as in the case of BA cations, pointing
3
at the center of the I rectangle generated by four adjacent
4
octahedra, causing a significant degree of distortion. As a result of
this distortion, the Pb−I−Pb angle in MAPbI becomes 163.55°.
3
The distinct structural features of MAPbI , however, become
3
evident when the perovskite begins to self-assemble into
2
7‑
“
[Pb I ] boxes”, as shown in Figure 4. The boxes consist of
27 81
1
8‑
three stacked “[Pb I ] ” slabs, which themselves display an out-
9
36
of-phase octahedral distortion and connect along the direction
defined by the C−N bond while displaying no rotation in the
2
4
0
0 −
plane perpendicular to the C−N bond direction (the a a c tilt
88
pattern in Glazer notation, Figure 4, right column). This
pattern is characteristic of the 3D CH NH MX perovskites in
3
3
3
1
0
64
3
general and is opposite to the CsMX3 and HC(NH ) MX
counterparts, which adopt an in-phase tilting (a a c in Glazer
notation) when the “[Pb I ] ” slabs connect along the unique
2 2
0
0 +
1
8‑
The net result of this competition is an overall distortion in the
intermediate 2D perovskites, which occurs across two different
directions (as opposed to (BA) PbI and MAPbI ), but it is
9
36
crystallographic direction. On the basis of the above character-
istics, it becomes evident that the nature and orientation of the
ammonium cation influence the tilting pattern in the hybrid
perovskites and strongly impact the degree of distortion of the
inorganic component.
2
4
3
less severe compared to the parent compounds due to the
competing interactions between MA and BA. For example, in
(BA) (MA) Pb I , the distortion angles are (Pb−I−Pb)
=
2
2
3 10
010
A unique situation arises in the (BA) (MA) Pb I
169.52°, (Pb−I−Pb) = 164.58°, and (Pb−I−Pb) = 172.26°.
2
n−1
n 3n+1
101
101
multilayered compounds where both MA and BA cations are
present, owing to the competition between the two cations trying
to satisfy their stereochemical demands. Because the situation is
similar for n = 2, 3, and 4, we will limit our discussion to the n = 3
compound for simplicity (Figure 4, center column). Continuing
The mixed cation configuration as well as the competition
between the constituting cations renders the multilayered
perovskites fundamentally different from both the simple 2D
single-layer compounds as well as the 3D perovskites and
provides them with some unique properties, which will be
discussed in detail below.
1
8‑
with the “[Pb I ] slab” concept, see Figure 4, we can clearly
9
36
I
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
Table 3. Optical Parameters of the (BA) (MA)n−1Pb I
Perovskites
2
n 3n+1
band gap
Eg (eV)
excitonic
absorption (eV)
photoluminescence
PL (eV)
(E − PL)
electron effective
hole effective
b
g
a
b
compound
(meV)
mass (me,xz
)
mass (mh,xz)
(
(
(
(
BA) PbI4
2.43
2.17
2.03
1.91
1.50
2.35
2.08
1.96
1.85
1.59
2.35
2.12
2.01
1.90
80
50
0.082
0.144
2
BA) (MA)Pb I
2
2 7
BA) (MA) Pb I
3 10
20
0.097
0.094
0.141
0.153
2
2
BA) (MA) Pb I
4 13
10
2
3
MAPbI3
1.60
n/a
a
b
Position of the excitonic peak in the diffuse reflectance spectra. Calculated at the DFT-PBEsol level in this work. mxz is reported as the average of
the effective mass along the x and z directions and is in units of the bare electron mass.
Band Gaps, Electronic Structure, and Photolumines-
cence. Figure 5 displays the room-temperature absorption
and photoluminescence spectra of bulk samples from the
(
BA) (MA) Pb I
series. The optical absorption edges are
2
n−1
n 3n+1
remarkably sharp and increase in energy with decreasing n value,
from 1.50eV (n = ∞) to 2.43eV (n = 1), a property arising from
the dimensional reduction of the 3D perovskite lattice into the
89−91
lower dimensionality homologous structures (Figure 5a).
From the experimental spectra, we assign the higher energy
absorption edge to the band gap of the materials (Table 3). In all
compounds, the onset of the optical absorption consists of sharp
edges, although occasionally an absorption tail can be observed
in samples of the n = 3 and n = 4 compounds possibly arising
from small impurities of intergrown higher order homologous
members. The sharp nature of the absorption edges points
toward a direct band gap in all compounds. A similar band gap
trend can be also observed from thin-films of these materials
deposited by spin coating dimethylformamide (DMF) precursor
Figure 6. Electronic band structure of the polar configurations of
selected (BA) (MA) Pb I perovskites. (a) (BA) PbI (n = 1), (b)
2
n−1
n 3n+1
2
4
(BA) (MA) Pb I (n = 3), and (c) (BA) (MA) Pb I (n = 4) along
2
2
3 10
2
3
4 13
the Γ(0,0,0)-X/S(1/2,0,0)-U/R(1/2,0,1/2)-Z(0,0,1/2)-Γ(0,0,0) path
(solid vertical lines) throughout the Brillouin zone. The Fermi level is
set to 0 eV and indicated by the horizontal broken red line.
27
solutions as we reported previously, confirming that thin films
of pure 2D compounds can be made by direct deposition onto
the substrates.
Using DFT, we computed the electronic band structure
and density of states for the n = 1, 3, and 4 structures (Figure 6).
All compounds are semiconducting, with the valence band
almost exclusively consisting of I 5p states with a small amount of
Pb 6s character, while the bottom of the conduction band
primarily consists of Pb 6p states (Figure 7). The three studied
compounds also each exhibit a clear direct band gap. The band
gaps are 1.99 eV for (BA) PbI , 1.78 eV for (BA) (MA) Pb I ,
2
4
2
2
3 10
and 0.96 eV for (BA) (MA) Pb I . Although consistently
2
3
4 13
underestimated within DFT at the PBEsol level, we find the
same general trend of decreasing band gap with increasing
number of perovskite layers to be in agreement with our experi-
mental results (Table 3).
92
As with other Pb-based metal−organic halide perovskites,
the bands near the valence and conduction edges are highly
dispersive in reciprocal space along the corresponding real-space
axes in which the PbI6 octahedra are corner-connected,
indicating small hole and electron effective masses (m and m ,
Figure 7. Density of states of the polar configurations of selected
h
e
respectively) along the x and z directions. By applying the
parabolic band approximation to the computed electronic
structure, we find that m and m are approximately 0.14 m
0
(BA)
2
(MA)n−1Pb
I3n+1 perovskites. (a) (BA) PbI (n = 1), (b)
n 2 4
(
BA) (MA) Pb I (n = 3), and (c) (BA) (MA) Pb I (n = 4). Iodine
2
2
3 10
2
3
4 13
5
p (red) and lead 6s (blue) states make up the top of the valence band,
h
e
while Pb 6p states (green) form the bottom of the conduction band.
and 0.08 m along both x and z, where m is the rest electron
0
0
mass, and are nearly independent of the number of perovskite
and spacing layers (Table 3). Because in the Ruddlesden−
Popper structure the inorganic semiconducting network is
disconnected along the y axis (Figure 1), it has a nearly flat bands
along this direction, indicating that the effective masses along this
direction are extremely large, which would tend to restrict the
charge transport to the xz plane. When spin−orbit inter-
actions (SOIs) are included in our calculations, for example,
in the n = 1 phase, we find the band gap decreases to 1.25 eV, and
there are changes to the band degeneracies, but the band
curvature remains the same; furthermore, the noncentrosym-
metric configuration discussed above remains lower in energy
than the centrosymmetric. Unfortunately, the large size of the
unit cells prevented us from including SOI in the higher order
compounds, but we expect the same trends to persist.
J
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
In addition to the primary absorption edge, we consistently
observe another peak below this region appearing in the 2D
perovskites. For clarity, the raw reflectance data clearly
illustrating the presence of the second peak are also provided
in the Supporting Information (Figure S1). The intensity of the
second peak is strongest for the n = 1 compound and
progressively subsides as the thickness of the inorganic slabs
increases reaching a minimum for n = ∞ where the peak sits
nearly on top of the absorption edge (Figure 5a). The presence of
the secondary absorption in the optical absorption peak spectra
of these compounds, which is similar in nature with that observed
Unlike the absorption spectra, which are characterized by an
absorption edge and an excitonic peak, the photoluminescence
spectra consist of one single emission peak corresponding to the
energy value of the excitonic peak in the absorption spectra. This
is because the relaxation pathway is dictated by direct radiative
recombination of an exciton, which is typical for an excitonic
material. As predicted from the above 2D effects, the energy
differences between the band gap and the photoluminescence
emission values (or the excitonic absorption values) decrease
with increased n (80 meV for n = 1 to almost 10 meV for n = 4) as
shown in Table 3; here, the emphasis is given to the correct trend
in this energy difference as a function of n, but the actual exciton
binding energies should be more precisely determined at low
temperatures.
93
for the 3D perovskites CsPbCl and CsPbBr , suggests the
3
3
presence of stable excitons even at room temperature. Because of
the presence of excitons, the band gap of the material was esti-
mated by extrapolating the high energy edge (the fundamental
edge) to the hypothetical line parallel to the energy axis where
the fundamental edge is interrupted by the low energy absorption
For the 3D perovskite, it was challenging to extract the correct
bandgap energy due to the spectral proximity between the
primary absorption edge and the excitonic peak. In fact, the
primary absorption onset at ∼1.50 eV in Figure 5 (panel a)
should be interpreted as the “low-energy tail” of the excitonic
(
the exciton).
MAPbI in itself was reported to have a relatively large exciton
3
9
4,95
transition, not the fundamental band edge (E ) that is located
binding energy (E ) of ∼50 meV,
the high-frequency regime due to charge screening from
resonant phonons. This value has been recently questioned
in favor of a negligible binding energy model assuming an exciton
Bohr radius of ∼20 nm. On the other hand, E ≈ 400 meV or
even 500 meV for exfoliated nanosheets has been reported for
the n = 1 2D perovskite homologue, whereas the estimated
energies from the multiple-layer 2D systems tend to float
which becomes smaller in
g
b
slightly higher in energy due to the contribution of the exciton
binding energy. This picture is indeed in line with the spectral
matching of the photoluminescence maximum (∼1.6 eV) and
the absorption saturation (∼1.59 eV). Another remarkable
feature of the photoluminescence spectra is the line shape of
the photoluminescence peaks, which is symmetric for the 3D
perovskites but not for the 2D analogues. Such a behavior has
been recently assigned to trap states in the perovskite framework
characteristic of both 2D and 3D perovskites, but the number of
77
96
40
b
5
8
46,54
between these two values.
The effect of higher exciton
binding energies for lower dimensional perovskites is typical and
can be understood in terms of reduced dielectric screening and
56,57
trap states has been shown to be higher in the 2D compounds.
50,51
Electronic Structure Calculations and Phase Stability.
To obtain further insight into the structural character and
electronic structure of these 2D perovskite systems, we per-
formed several investigations using DFT calculations. We first
computed the relative energies of (BA) PbI (n = 1),
stronger quantum confinement effects.
A similar effect was
observed in few-layered transition metal dichalcogenide
9
7,98
systems.
These excitons are favored by the multiple-
quantum-well nature of these 2D perovskites.
Unlike the artificially constructed quantum well structures of
2
4
(
BA) (MA) Pb I (n = 3), and (BA) (MA) Pb I (n = 4) in
2 2 3 10 2 3 4 13
the classical III−V semiconductors (for example, the GaAs/
both the centrosymmetric and noncentrosymmetric crystal
99
AlAs/GaAs heterostructures) where the excitons are observ-
structures to evaluate their differences in stability. In (BA) PbI ,
2
4
able only at low temperature, the excitons in BA (MA)
2
n−1
we found that the Pbc2 and Pcab phases are energetically
58
1
Pb I
are stable at room temperature. As it has been
n 3n+1
degenerate within the precision of our calculations, with the
4
1
suggested, the stable excitons in the 2D perovskites are due to
Pbc2 structure negligibly lower in energy by 0.58 meV/f.u.
1
the Coulombic charge screening effect, which results from the
Although this could support the classification of (BA) PbI into
2
4
2
−
charged nature of the perovskite {MAn−1PbnI3n+1
}
slabs as
the polar Pbc2 symmetry, it more accurately indicates that
1
opposed to the neutral covalent framework of GaAs-based
quantum well structures. The consequence of this difference is
that upon generation, the exciton is strongly confined within the
built-in electric field between the positive organic spacers and the
negative perovskite slabs, which acts upon the photogenerated
intergrowths of the two phases is likely, granting local variations
in noncentrosymmetry. Remarkably, when the number of
perovskite layers is increased to n = 3 and n = 4, the non-
centrosymmetric phase becomes significantly more stable than
the centrosymmetric structure. In (BA) (MA) Pb I , we find
2
2
3 10
−
+
e /h pairs inhibiting exciton splitting into free carriers. This in
turn causes strong photoluminescence at room temperature
that the C2cb structure is lower in energy than the centric (Acam)
phase by 527.9 meV/f.u., while in (BA) (MA) Pb I , the
2
3
4 13
(Figure 5b).
formation energy of the centric structure (Ccmm) lies much
higher than that of the acentric (Cc2m) space group, which is
strongly favored. The results are in good agreement with similar
We chose to perform the photoluminescence studies using a
confocal microscope setup with a steady-state excitation source
473 nm) on oriented crystals (excitation along (010) direction
for n = 2, 3, 4, (001) direction for n = 1 and (100) direction
for n = ∞ of the (BA) (MA) Pb I compounds. Previous
(
calculations on MAPbI , which predict the acentric (I4cm) struc-
3
ture to lay slightly lower in energy compared to the centric
100
2
n−1
n 3n+1
(I4/mcm) structure by 100 meV/f.u., which is very close to the
studies have shown both 2D (BA) PbI and 3D MAPbI
experimentally determined value of 10.4 kJ/mol (= 108 meV/f.u.),
2
4
3
1
perovskite compounds display photoluminescence at room
obtained earlier from temperature-dependent H NMR experi-
3
1,64
temperature,
although the photoluminescence intensity
ments for the isotropic reorientation of MA ions at room
temperature. These large energy differences show that the
noncentrosymmetric structures in the 2D perovskite with larger
numbers of layers are favorable, despite the weak SHG signal
obtained from the compounds experimentally. These initial
and wavelength can vary widely based on the preparation
conditions. The photoluminescence emission energy of all the
2D perovskites described here decreases with increasing slab
thickness similar to the band gap trend (Figure 5).
K
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
results reveal that there is a complex interrelationship between
the perovskite layer thickness (n) and a yet to be understood
correlation length scale required for cooperative lifting of
inversion symmetry. Detailed calculations for the structural
interrelationships will be explored in a future study.
cannot be expressed simultaneously in centrosymmetric
68
compounds, as described previously. Although insightful, this
comparison is difficult to achieve in the case of the lead iodide
perovskites because of the very strong fluorescence, which masks
some of the Raman-active vibrational modes and lowers the
quality of the spectra. Therefore, to probe if inversion symmetry
is present in the structure, we performed SHG and THG
measurements on polycrystalline samples of the perovskites.
All of the 2D compounds and the 3D perovskites displayed
weak SHG response as well as THG response. Typically, the
SHG signal vanishes for a centrosymmetric compound within the
dipole approximation, whereas THG signal is insensitive to the
crystal symmetry. The criterion for classifying a material as
noncentrosymmetric according to the Kurz-Perry method is
satisfied if the SHG ratio between the sample and the known
Our DFT calculations reveal a contrasting trend in the stability
of the polar structure in the (BA) (MA)n−1PbnI3n+1 perovskite
2
relative to the experimental SHG results (see Figure 8a below).
(
2)
α-quartz reference (SiO : χ = 0.6 pm/V)), defined as I(SHG)/
q
2
−2
106,107
I (SHG), is larger than 10 ,
as indicated by the red dashed
line in Figure 8, panel a. The measured SHG ratios for the 2D and
D perovskite samples are rather close to the borderline for
3
considering them to be noncentrosymmetric (red dashed line),
with the n = 1 and n = 2 perovskites exhibiting SHG signals
slightly above the red dashed line and the n = 3, n = 4 and n = ∞
lying slightly below. We have confirmed from repeated, averaged
out measurements that there is a general trend of an increasing
SHG response with a decreasing thickness of perovskite layers
Figure 8. (a) I(SHG)/I (SHG) (solid bars) for powdered samples of
q
(
BA) (MA) Pb I and MAPbI . The red dashed line is the criterion
2 n−1 n 3n+1 3
for the noncentrosymmetry assignment based on the SiO method.
2
(
b) SHG/THG (solid bars) for the powdered samples of (BA) (MA)
2 n−1
Pb I
n 3n+1
and MAPbI3.
(that is, SHG(n = 1) > SHG(n = 2) > SHG(n = 3) > SHG (n = 4) >
SHG(n = ∞)), which is unusual since the SHG trend opposes
the expected behavior, which dictates that the SHG response
should increase following the reduction of the band gap. Since in
some instances a centrosymmetric medium may yield a small,
nonzero SHG response due to secondary effects such as
The DFT calculations show the n = 1 compound to be less
susceptible to symmetry-breaking, whereas the compound
experimentally exhibits the relatively strongest SHG response.
Along the same lines, when the number of layers increases and
the material acquires 3D-like characteristics, the SHG response
declines reaching a minimum (although nonzero) value for the
n = ∞ perovskite despite the DFT calculation results predicting
an opposite trend. This behavior could possibly be rationalized
by the ability of the perovskite to accommodate defects that
could generate individual domains with opposing polarity and
lead to the expression of residual local polarization. This is
reasonable because the MA ions are small enough to realign
inside the perovskite cages at a small energy cost; however, the
BA ions are relatively rigid and in that sense they “lock” the
crystal structure in a preferred configuration. In other words,
106
quadrupole transitions and surface-induced processes,
the
Kurz-Perry method is not an absolute measure but rather a good
guideline for the space group assessment.
Because of the uncertainty inherent to the Kurz-Perry powder
method, we proceeded in evaluating the crystal symmetry of the
(BA)
(MA)n−1Pb I3n+1 perovskites by employing an alternative
2
n
method based on the comparison of the relative intensities
108
between the SHG and the THG signals. This method is more
reliable because any effects arising from extrinsic sources would
cancel out in the “ratio” of the two harmonic generation
intensities. In principle, SHG as the lower order NLO term
should produce a much stronger response than THG in a
noncentrosymmetric medium. Figure 8, panel b shows the
experimental SHG/THG ratios of the samples and the reference
MAPbI and the 2D slabs favors an acentric configuration that
3
cannot be kinetically maintained because of the MA
reorientations, whereas (BA) PbI favors a centric configuration
2
4
that once perturbed cannot be kinetically restored. These
statements can be further supported by a recent report for the
chloride analogue of the n = 1 compound where clear
material α-SiO . To compensate for the THG absorption effect,
2
which produces photons with higher energy than the band gap,
we employed a proper correction factor to each material based on
its relative absorbance values at the SHG and THG wavelengths,
900 and 600 nm, respectively (discussed above in the context of
101
ferroelectric behavior was observed, whereas the ferroelectric
102
properties of MAPbI are still under debate. An important
3
aspect of the ferroelectricity discussion that cannot be neglected
103,104
is the light-enhanced polarization observed in perovskites,
Figure 5). As expected, the response for α-SiO
over THG for α-SiO thus confirming that α-SiO
noncentrosymmetric. In contrast, all the (BA) (MA) Pb I
n 3n+1
2
SHG dominates
an observation that suggests that the generation and motion of
domains in the halide perovskites is in fact light-induced, and
therefore the polarization of the halide perovskites should be
2
2
is indeed
2
n−1
perovskites exhibit a THG response, which dominates over SHG
suggesting that the perovskite samples are more likely to be
centrosymmetric and that the SHG response occurs mainly
through secondary processes.
Nevertheless, it is still remarkable that the SHG response from
the bulk material, albeit weak, is not negligible. This finite SHG
response seems to arise from local symmetry-breaking effects
within the highly polarizable structure of the perovskites. Our
analysis therefore implies that the perovskite samples can exhibit
large local polar domains with differing polarities, but in the
105
better described in the context of photoferroics.
Nonlinear Optical Properties of the 2D Perovskites.
Although the compounds presented here were assigned as polar,
based on crystallographic criteria and DFT calculation, some
ambiguity in the assignment of the space group remains. Indirect
evidence of the acentricity of the compounds is also provided by
means of Raman and IR spectroscopy through the mutual
exclusion principle (see Supporting Information, Figure S2),
where the vibrational modes that are both IR and Raman active
L
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
overall bulk material “matrix” averages out to centrosymmetric.
Similar symmetry-breaking phenomena have also been observed
for the noncentrosymmetric artificial GaAs/AlAs quantum-
wells, which could be regarded as the III−IV semiconductor
equivalents of the natural quantum-well halide Ruddlesden−
Popper systems.
In view of the NLO results, the 2D perovskites may be best
described as noncentrosymmetric (Cc2m for even n and C2cb for
odd n) reflecting the presence of local domains generated by local
symmetry breaking, whereas the centrosymmetric models
semiconducting properties of the perovskites to the observed
systematically evolving optoelectronic properties.
1
09
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the ACS
Publications website at DOI: 10.1021/acs.chemmater.6b00847.
■
*
S
Vibrational spectra, detailed crystallographic tables for
(
BA) (MA) Pb I
(n = 2−4) for the noncentrosym-
2
n−1
n 3n+1
metric (proper assignment) and centrosymmetric (alter-
native model) refinements (PDF)
Crystallographic files (CIF)
(Ccmm for even n and Acam for odd n) refer to the averaged
crystal structure resulting from the random orientation of the
local polar domains. A similar description may also apply to the
Crystallographic files (CIF)
MAPbI n = ∞ homologue where the local crystal structure is
3
polar and can be described using the noncentrosymmetric
110,111
AUTHOR INFORMATION
Corresponding Author
E-mail: m-kanatizdis@northwestern.edu.
(
ferroelectric) model (space group I4cm),
apparent crystal structure of bulk samples is centrosymmetric
space group I4/mcm) due to the canceling out of the
whereas the
■
*
(
polarization from the randomly oriented local polar domains.
Author Contributions
These authors contributed equally to this work.
¶
CONCLUDING REMARKS
Notes
■
The authors declare no competing financial interest.
We have demonstrated the facile, large-scale synthesis of the 2D
(
BA) (MA) Pb I (n = 1, 2, 3, 4, ∞) perovskites that belong
2
n−1
n 3n+1
ACKNOWLEDGMENTS
to the Ruddlesden−Popper family. The convenient isolation of
the pure materials allowed for the accurate characterization of the
crystal structure of the compounds, which are reported here for
the first time for n = 2, 3, and 4. The 2D perovskites crystallize in
polar space groups reflecting the alignment of the MA cations to
generate an uncompensated dipole moment. All 2D homologous
compounds have sharp absorption edges in the visible spectral
range suggesting a direct band gap, which asymptotically
■
This work was supported by Grant No. SC0012541 from the U.S.
Department of Energy, Office of Science. D.H.C. acknowledges
support from the Link Foundation through the Link Foundation
Energy Fellowship Program. J.Y. and J.M.R. were supported by
the National Science Foundation (NSF) through the Pennsylvania
State University MRSEC under Award No. DMR-1420620. DFT
calculations were performed on the Extreme Science and
Engineering Discovery Environment (XSEDE), which is
supported by NSF Grant No. ACI-1053575, and the QUEST
high performance computing facility at Northwestern University,
which is jointly supported by the Office of the Provost, the Office
for Research, and Northwestern University Information
Technology. Electron microscopy was done at the Electron
Probe Instrumentation Center (EPIC) at Northwestern
University. Raman spectroscopy was performed at the Integrated
Molecular Structure Education and Research Center (IMSERC)
at Northwestern University. Confocal microscope studies were
performed at the SPID facility (NUANCE Center-Northwestern
University), which has received support from the State of Illinois
through the International Institute for Nanotechnology.
approaches the band gap of the 3D MAPbI perovskite. DFT
3
calculations confirm that the 2D perovskites are indeed direct
band gap semiconductors with large bandwidths and small
effective masses for both electrons and holes. The absorption of
the 2D perovskites is accompanied by a strong photo-
luminescence emission at room temperature, which is character-
istically red-shifted with respect to the absorption edge. The
optical physics of the 2D perovskites is characterized by the
presence of stable excitons, which act like the active chro-
mophores at room temperature, owing to their natural quantum
well electronic structure, which is analogous to that of the
artificial III−V semiconductor quantum wells.
In view of their exciting properties, the 2D perovskites can
indeed be a new source of functional, tunable semiconductors
expanding on the properties of the 3D species. For example, the
intense room-temperature photoluminescence possible from the
REFERENCES
■
(
1) Burschka, J.; Pellet, N.; Moon, S. J.; Humphry-Baker, R.; Gao, P.;
Nazeeruddin, M. K.; Gratzel, M. Sequential Deposition as a Route to
2
D perovskites points to their potential utility in light-emitting
High-Performance Perovskite-Sensitized Solar Cells. Nature 2013, 499,
diodes and lasers. On the other hand, the optimal band gaps of
higher n members indicate that these compounds can be used as
efficient light absorbers for solution-processed solar cells, which
offers better solution processability in addition to their superior
environmental stability. More complicated concepts like electro-
optical modulators can also be envisaged, taking advantage of the
quantum well tunable architecture of the 2D perovskites. The
great advantage of the 2D perovskites over the 3D ones is that
the complex electronic structure of the 2D materials can be
chemically tuned through subtle modification of the crystal
structure, adjusting the layer-to-layer spacing or the spacer-
perovskite interactions. Work in progress is directed in synthesiz-
ing and structurally characterizing more homologous series to be
able to correlate the effects of the chemical composition in the
3
(
16−319.
2) Jeon, N. J.; Noh, J. H.; Kim, Y. C.; Yang, W. S.; Ryu, S.; Il Seol, S.
Solvent Engineering for High-Performance Inorganic-Organic Hybrid
Perovskite Solar Cells. Nat. Mater. 2014, 13, 897−903.
(3) Boix, P. P.; Agarwala, S.; Koh, T. M.; Mathews, N.; Mhaisalkar, S. G.
Perovskite Solar Cells: Beyond Methylammonium Lead Iodide. J. Phys.
Chem. Lett. 2015, 6, 898−907.
(
4) Gao, P.; Gratzel, M.; Nazeeruddin, M. K. Organohalide Lead
Perovskites for Photovoltaic Applications. Energy Environ. Sci. 2014, 7,
448−2463.
5) De Angelis, F. Modeling Materials and Processes in Hybrid/
2
(
Organic Photovoltaics: From Dye-Sensitized to Perovskite Solar Cells.
Acc. Chem. Res. 2014, 47, 3349−3360.
(6) Green, M. A.; Ho-Baillie, A.; Snaith, H. J. The Emergence of
Perovskite Solar Cells. Nat. Photonics 2014, 8, 506−514.
M
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
(
7) Miyasaka, T. Perovskite Photovoltaics: Rare Functions of Organo
Lead Halide in Solar Cells and Optoelectronic Devices. Chem. Lett.
015, 44, 720−729.
8) Stoumpos, C. C.; Kanatzidis, M. G. The Renaissance of Halide
Enhanced Moisture Stability. Angew. Chem., Int. Ed. 2014, 53, 11232−
11235.
(27) Cao, D. H.; Stoumpos, C. C.; Farha, O. K.; Hupp, J. T.; Kanatzidis,
M. G. 2D Homologous Perovskites as Light-Absorbing Materials for
Solar Cell Applications. J. Am. Chem. Soc. 2015, 137, 7843−7850.
(28) Cao, D. H.; Stoumpos, C. C.; Malliakas, C. D.; Katz, M. J.; Farha,
2
(
Perovskites and Their Evolution as Emerging Semiconductors. Acc.
Chem. Res. 2015, 48, 2791−802.
(
Devices. Small 2015, 11, 10−25.
(
9) Jung, H. S.; Park, N.-G. Perovskite Solar Cells: From Materials to
O. K.; Hupp, J. T.; Kanatzidis, M. G. Remnant PbI , an Unforeseen
2
Necessity in High-Efficiency Hybrid Perovskite-Based Solar Cells? APL
Mater. 2014, 2, 091101.
10) Chung, I.; Song, J.-H.; Im, J.; Androulakis, J.; Malliakas, C. D.; Li,
H.; Freeman, A. J.; Kenney, J. T.; Kanatzidis, M. G. CsSnI3:
Semiconductor or Metal? High Electrical Conductivity and Strong
Near-Infrared Photoluminescence from a Single Material. High Hole
Mobility and Phase-Transitions. J. Am. Chem. Soc. 2012, 134, 8579−
(29) Choi, J. J.; Yang, X.; Norman, Z. M.; Billinge, S. J.; Owen, J. S.
Structure of Methylammonium Lead Iodide Within Mesoporous
Titanium Dioxide: Active Material in High-Performance Perovskite
Solar Cells. Nano Lett. 2014, 14, 127−133.
8587.
(30) Calabrese, J.; Jones, N. L.; Harlow, R. L.; Herron, N.; Thorn, D.
(
11) Kojima, A.; Teshima, K.; Shirai, Y.; Miyasaka, T. Organometal
L.; Wang, Y. Preparation and Characterization of Layered Lead Halide
Halide Perovskites as Visible-Light Sensitizers for Photovoltaic Cells. J.
Am. Chem. Soc. 2009, 131, 6050−6051.
Compounds. J. Am. Chem. Soc. 1991, 113, 2328−2330.
(31) Ishihara, T.; Takahashi, J.; Goto, T. Exciton-State in Two-
(
12) Etgar, L.; Gao, P.; Xue, Z.; Peng, Q.; Chandiran, A. K.; Liu, B.;
Dimensional Perovskite Semiconductor (C H NH ) PbI . Solid State
10
21
3
2
4
Nazeeruddin, M. K.; Grat
Heterojunction Solar Cells. J. Am. Chem. Soc. 2012, 134, 17396−17399.
13) Lee, M. M.; Teuscher, J.; Miyasaka, T.; Murakami, T. N.; Snaith,
̈
zel, M. Mesoscopic CH NH PbI /TiO
3 3 3 2
Commun. 1989, 69, 933−936.
(32) Mitzi, D. B.; Wang, S.; Feild, C. A.; Chess, C. A.; Guloy, A. M.
(
Conducting Layered Organic-Inorganic Halides Containing < 110>-
H. J. Efficient Hybrid Solar Cells Based on Meso-Superstructured
Oriented Perovskite Sheets. Science 1995, 267, 1473−1476.
Organometal Halide Perovskites. Science 2012, 338, 643−647.
(33) Mitzi, D. B.; Feild, C. A.; Harrison, W. T. A.; Guloy, A. M.
(
14) Kim, H.-S.; Lee, C.-R.; Im, J.-H.; Lee, K.-B.; Moehl, T.; Marchioro,
Conducting Tin Halides with a Layered Organic-Based Perovskite
̈
A.; Moon, S.-J.; Humphry-Baker, R.; Yum, J.-H.; Moser, J. E.; Gratzel,
M.; Park, N.-G. Lead Iodide Perovskite Sensitized All-Solid-State
Submicron Thin Film Mesoscopic Solar Cell with Efficiency Exceeding
Structure. Nature 1994, 369, 467−469.
(34) Kagan, C. R.; Mitzi, D. B.; Dimitrakopoulos, C. D. Organic-
Inorganic Hybrid Materials as Semiconducting Channels in Thin-Film
9%. Sci. Rep. 2012, 2, 591.
Field-Effect Transistors. Science 1999, 286, 945−947.
(
15) Zhou, H. P.; Chen, Q.; Li, G.; Luo, S.; Song, T. B.; Duan, H. S.;
(
35) Ruddlesden, S. N.; Popper, P. New Compounds of the K NiF
2 4
Hong, Z. R.; You, J. B.; Liu, Y. S.; Yang, Y. Interface Engineering of
Type. Acta Crystallogr. 1957, 10, 538−539.
Highly Efficient Perovskite Solar Cells. Science 2014, 345, 542−546.
(
36) Ruddlesden, S. N.; Popper, P. The Compound Sr Ti O and its
3 2 7
(
16) Jeon, N. J.; Noh, J. H.; Yang, W. S.; Kim, Y. C.; Ryu, S.; Seo, J.;
Structure. Acta Crystallogr. 1958, 11, 54−55.
Seok, S. I. Compositional Engineering of Perovskite Materials for High-
(37) For a recent review, see: Schaak, R. E.; Mallouk, T. E. Perovskites
Performance Solar Cells. Nature 2015, 517, 476−480.
by Design: A Toolbox of Solid-State Reactions. Chem. Mater. 2002, 14,
455−1471.
38) Mitzi, D. B. Synthesis, Crystal Structure, and Optical and Thermal
Properties of (C H NH ) MI (M = Ge, Sn, Pb). Chem. Mater. 1996, 8,
(
17) Yang, W. S.; Noh, J. H.; Jeon, N. J.; Kim, Y. C.; Ryu, S.; Seo, J.;
1
(
Seok, S. I. High-Performance Photovoltaic Perovskite Layers Fabricated
Through Intramolecular Exchange. Science 2015, 348, 1234−1237.
4
9
3
2
4
(
18) Fabini, D. Quantifying the Potential for Lead Pollution from
7
(
(
91−800.
39) Mitzi, D. B.; Liang, K. Preparation and Properties of
C H NH ) EuI : A Luminescent Organic−Inorganic Perovskite with
Halide Perovskite Photovoltaics. J. Phys. Chem. Lett. 2015, 6, 3546−
548.
19) Niu, G.; Guo, X.; Wang, L. Review of Recent Progress in Chemical
Stability of Perovskite Solar Cells. J. Mater. Chem. A 2015, 3, 8970−
980.
20) Hao, F.; Stoumpos, C. C.; Cao, D. H.; Chang, R. P. H.; Kanatzidis,
3
(
4
9
3
2
4
a Divalent Rare-Earth Metal Halide Framework. Chem. Mater. 1997, 9,
990−2995.
40) Kataoka, T.; Kondo, T.; Ito, R.; Sasaki, S.; Uchida, K.; Miura, N.
Magneto-Optical Study on Excitonic Spectra in (C H NH ) PbI .
2
(
8
(
6
13
3
2
4
M. G. Lead-Free Solid-State Organic-Inorganic Halide Perovskite Solar
Phys. Rev. B: Condens. Matter Mater. Phys. 1993, 47, 2010−2018.
Cells. Nat. Photonics 2014, 8, 489−494.
(41) Muljarov, E. A.; Tikhodeev, S. G.; Gippius, N. A.; Ishihara, T.
(
21) Noel, N. K.; Stranks, S. D.; Abate, A.; Wehrenfennig, C.;
Excitons in Self-Organized Semiconductor-Insulator Superlattices - PbI-
Based Perovskite Compounds. Phys. Rev. B: Condens. Matter Mater. Phys.
Guarnera, S.; Haghighirad, A.; Sadhanala, A.; Eperon, G. E.; Pathak, S.
K.; Johnston, M. B.; Petrozza, A. M.; Herz, L. M.; Snaith, H. J. Lead-Free
Organic-Inorganic Tin Halide Perovskites for Photovoltaic Applica-
tions. Energy Environ. Sci. 2014, 7, 3061−3068.
1
(
995, 51, 14370−14378.
42) Ishihara, T.; Takahashi, J.; Goto, T. Optical-Properties Due to
Electronic-Transitions in 2-Dimensional Semiconductors
(
22) Ogomi, Y.; Morita, A.; Tsukamoto, S.; Saitho, T.; Fujikawa, N.;
(
C H NH ) PbI . Phys. Rev. B: Condens. Matter Mater. Phys. 1990,
Shen, Q.; Toyoda, T.; Yoshino, K.; Pandey, S. S.; Ma, T.; Hayase, S.
n 2n+1 3 2 4
4
(
2, 11099−11107.
43) Hong, X.; Ishihara, T.; Nurmikko, A. V. Dielectric Confinement
CH NH Sn Pb I Perovskite Solar Cells Covering up to 1060 nm. J.
3
3
x
(1‑x) 3
Phys. Chem. Lett. 2014, 5, 1004−1011.
Effect on Excitons in PbI -Based Layered Semiconductors. Phys. Rev. B:
(
23) Zuo, F.; Williams, S. T.; Liang, P.-W.; Chueh, C.-C.; Liao, C.-Y.;
4
Condens. Matter Mater. Phys. 1992, 45, 6961−6964.
Jen, A. K. Y. Binary-Metal Perovskites Toward High-Performance
(44) Ishihara, T.; Hong, X.; Ding, J.; Nurmikko, A. V. Dielectric
Planar-Heterojunction Hybrid Solar Cells. Adv. Mater. 2014, 26, 6454−
Confinement Effect for Exciton and Biexciton States in PbI -Based 2-
6460.
4
(
24) Hao, F.; Stoumpos, C. C.; Chang, R. P.; Kanatzidis, M. G.
Dimensional Semiconductor Structures. Surf. Sci. 1992, 267, 323−326.
Anomalous Band Gap Behavior in Mixed Sn and Pb Perovskites Enables
Broadening of Absorption Spectrum in Solar Cells. J. Am. Chem. Soc.
(45) Hirasawa, M.; Ishihara, T.; Goto, T. Exciton Features in 0-
4
‑
Dimensional, 2-Dimensional, and 3-Dimensional Networks of [PbI ]
4
2
(
014, 136, 8094−8099.
Octahedra. J. Phys. Soc. Jpn. 1994, 63, 3870−3879.
25) Hao, F.; Stoumpos, C. C.; Guo, P.; Zhou, N.; Marks, T. J.; Chang,
(46) Koutselas, I. B.; Ducasse, L.; Papavassiliou, G. C. Electronic
Properties of Three- and Low-Dimensional Semiconducting Materials
with Pb Halide and Sn Halide Units. J. Phys.: Condens. Matter 1996, 8,
1217−1227.
R. P. H.; Kanatzidis, M. G. Solvent-Mediated Crystallization of
CH NH SnI Films for Heterojunction Depleted Perovskite Solar
3
3
3
Cells. J. Am. Chem. Soc. 2015, 137, 11445−11452.
26) Smith, I. C.; Hoke, E. T.; Solis-Ibarra, D.; McGehee, M. D.;
Karunadasa, H. I. A Layered Hybrid Perovskite Solar-Cell Absorber with
(
(47) Kondo, T.; Iwamoto, S.; Hayase, S.; Tanaka, K.; Ishi, J.; Mizuno,
M.; Ema, K.; Ito, R. Resonant Third-Order Optical Nonlinearity in the
N
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
Layered Perovskite-Type Material (C H NH ) PbI . Solid State
(68) Stoumpos, C. C.; Frazer, L.; Clark, D. J.; Kim, Y. S.; Rhim, S. H.;
Freeman, A. J.; Ketterson, J. B.; Jang, J. I.; Kanatzidis, M. G. Hybrid
Germanium Iodide Perovskite Semiconductors: Active Lone Pairs,
Structural Distortions, Direct and Indirect Energy Gaps, and Strong
Nonlinear Optical Properties. J. Am. Chem. Soc. 2015, 137, 6804−6819.
6
13
3
2
4
Commun. 1998, 105, 503−506.
(
48) Knutson, J. L.; Martin, J. D.; Mitzi, D. B. Tuning the Band Gap in
Hybrid Tin Iodide Perovskite Semiconductors Using Structural
Templating. Inorg. Chem. 2005, 44, 4699−4705.
(
49) Umebayashi, T.; Asai, K.; Kondo, T.; Nakao, A. Electronic
̈
(69) Kresse, G.; Furthmuller, J. Efficiency of ab-initio Total Energy
Structures of Lead Iodide Based Low-Dimensional Crystals. Phys. Rev. B:
Condens. Matter Mater. Phys. 2003, 67, 155405.
Calculations for Metals and Semiconductors Using a Plane-Wave Basis
Set. Comput. Mater. Sci. 1996, 6, 15−50.
(
50) Shimizu, M.; Fujisawa, J.-I.; Ishi-Hayase, J. Influence of Dielectric
(70) Kresse, G.; Hafner, J. Ab initio Molecular Dynamics for Liquid
Confinement on Excitonic Nonlinearity in Inorganic-Organic Layered
Semiconductors. Phys. Rev. B: Condens. Matter Mater. Phys. 2005, 71,
Metals. Phys. Rev. B: Condens. Matter Mater. Phys. 1993, 47, 558−561.
71) Blochl, P. E. Projector Augmented-Wave Method. Phys. Rev. B:
Condens. Matter Mater. Phys. 1994, 50, 17953−17979.
72) Perdew, J. P.; Ruzsinszky, A.; Csonka, G. I.; Vydrov, O. A.;
(
̈
205306.
(
51) Tanaka, K.; Takahashi, T.; Kondo, T.; Umebayashi, T.; Asai, K.;
(
Ema, K. Image Charge Effect on Two-Dimensional Excitons in an
Inorganic-Organic Quantum-Well Crystal. Phys. Rev. B: Condens. Matter
Mater. Phys. 2005, 71, 045312.
Scuseria, G. E.; Constantin, L. A.; Zhou, X.; Burke, K. Restoring the
Density-Gradient Expansion for Exchange in Solids and Surfaces. Phys.
Rev. Lett. 2008, 100, 136406.
(
52) Shibuya, K.; Koshimizu, M.; Takeoka, Y.; Asai, K. Scintillation
(
73) Monkhorst, H. J.; Pack, J. D. Special Points for Brillouin-Zone
Integrations. Phys. Rev. B 1976, 13, 5188−5192.
74) Tabuchi, Y.; Asai, K.; Rikukawa, M.; Sanui, K.; Ishigure, K.
Properties of (C H NH ) PbI : Exciton Luminescence of an Organic/
6
13
3
2
4
Inorganic Multiple Quantum Well Structure Compound Induced by 2.0
(
MeV Protons. Nucl. Instrum. Methods Phys. Res., Sect. B 2002, 194, 207−
Preparation and Characterization of Natural Lower Dimensional
Layered Perovskite-Type Compounds. J. Phys. Chem. Solids 2000, 61,
212.
(
53) Papavassiliou, G. C.; Koutselas, I. B. Structural, Optical and
8
(
37−845.
75) Elleuch, S.; Dammak, T.; Abid, Y.; Mlayah, A.; Boughzala, H.
Synthesis, Structural and Optical Properties of a Novel Bilayered
Organic−Inorganic Perovskite C Pb I . J. Lumin. 2010, 130, 531−535.
Related Properties of Some Natural Three- and Lower-Dimensional
Semiconductor Systems. Synth. Met. 1995, 71, 1713−1714.
(
54) Tanaka, K.; Kondo, T. Bandgap and Exciton Binding Energies in
Lead-Iodide-Based Natural Quantum-Well Crystals. Sci. Technol. Adv.
Mater. 2003, 4, 599−604.
5
2 5
(
76) Papavassiliou, G. C. Three- and Low-Dimensional Inorganic
Semiconductors. Prog. Solid State Chem. 1997, 25, 125−270.
77) Valverde-Chavez, D. A.; Ponseca, C. S.; Stoumpos, C. C.; Yartsev,
(
55) Ishihara, T. Optical-Properties of PbI-Based Perovskite
(
Structures. J. Lumin. 1994, 60−61, 269−274.
A.; Kanatzidis, M. G.; Sundstrom, V.; Cooke, D. G. Intrinsic
Femtosecond Charge Generation Dynamics in Single Crystal
(
56) Wu, X.; Trinh, M. T.; Niesner, D.; Zhu, H.; Norman, Z.; Owen, J.
S.; Yaffe, O.; Kudisch, B. J.; Zhu, X. Y. Trap States in Lead Iodide
Perovskites. J. Am. Chem. Soc. 2015, 137, 2089−2096.
CH NH PbI . Energy Environ. Sci. 2015, 8, 3700−3707.
3 3 3
(
57) Wu, X.; Trinh, M. T.; Zhu, X. Y. Excitonic Many-Body
(78) Billing, D. G.; Lemmerer, A. Synthesis, Characterization and
Phase Transitions of the Inorganic-Organic Layered Perovskite-Type
Hybrids [(C H NH ) PbI ], n = 4, 5 and 6. Acta Crystallogr., Sect. B:
Interactions in Two-Dimensional Lead Iodide Perovskite Quantum
Wells. J. Phys. Chem. C 2015, 119, 14714−14721.
n
2n+1
3
2
4
(
58) Yaffe, O.; Chernikov, A.; Norman, Z. M.; Zhong, Y.;
Struct. Sci. 2007, 63, 735−747.
Velauthapillai, A.; van der Zande, A.; Owen, J. S.; Heinz, T. F. Excitons
in Ultrathin Organic-Inorganic Perovskite Crystals. Phys. Rev. B:
Condens. Matter Mater. Phys. 2015, 92, 045414.
(79) Hamilton, W. Significance Tests on the Crystallographic R Factor.
Acta Crystallogr. 1965, 18, 502−510.
(80) Rogers, D. On the Application of Hamilton’s Ratio Test to the
Assignment of Absolute Configuration and an Alternative Test. Acta
Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr. 1981, 37,
(
59) Bugajski, M.; Reginski, K. Optical Properties of Semiconductor
Quantum Wells. Opto-Electr. Rev. 1996, 4, 83−100.
60) Miller, D. A. B. Optical Physics of Quantum Wells. In Quantum
(
7
(
34−741.
Dynamics of Simple Systems; Oppo, G. -L., Barnett, S. M., Riis, E.,
Wilkinson, M., Eds.; Institute of Physics: London, 1996; pp 239−266.
81) Papavassiliou, G. C.; Mousdis, G. A.; Raptopoulou, C. P.; Terzis,
A. Some New Luminescent Compounds Based on 4-MethyIbenzyl-
(
61) Sheldrick, G. A Short History of SHELX. Acta Crystallogr., Sect. A:
amine and Lead Halides. Z. Naturforsch., B: J. Chem. Sci. 2000, 55b, 536−
Found. Crystallogr. 2008, 64, 112−122.
62) Spek, A. L. Structure Validation in Chemical Crystallography.
Acta Crystallogr., Sect. D: Biol. Crystallogr. 2009, 65, 148−155.
63) Farrugia, L. WinGX and ORTEP for Windows: an Update. J. Appl.
Crystallogr. 2012, 45, 849−854.
64) Stoumpos, C. C.; Malliakas, C. D.; Kanatzidis, M. G.
5
40.
(
(82) Zhu, X.-H.; Mercier, N.; Riou, A.; Blanchard, P.; Frere, P.
(
C H SCH NH ) (CH NH )Pb I : Non-Centrosymmetrical Crystal
4
3
2
3
2
3
3
2 7
(
Structure of a Bilayer Hybrid Perovskite. Chem. Commun. 2002, 2160−
161.
83) Mercier, N. (HO C(CH ) NH ) (CH NH )Pb I : a Predicted
2
(
(
2
2
3
3
2
3
3
2 7
Semiconducting Tin and Lead Iodide Perovskites with Organic Cations:
Phase Transitions, High Mobilities, and Near-Infrared Photolumines-
cent Properties. Inorg. Chem. 2013, 52, 9019−9038.
Non-Centrosymmetrical Structure Built up from Carboxylic Acid
Supramolecular Synthons and Bilayer Perovskite Sheets. CrystEngComm
2
(
005, 7, 429−432.
(
65) Chondroudis, K.; McCarthy, T. J.; Kanatzidis, M. G. Chemistry in
84) Lemmerer, A.; Billing, D. G. Synthesis, Characterization and
Molten Alkali Metal Polyselenophosphate Fluxes. Influence of Flux
Composition on Dimensionality. Layers and Chains in APbPSe4,
A Pb(PSe ) (A = Rb, Cs), and K Eu(PSe ) . Inorg. Chem. 1996, 35,
Phase Transitions of the Inorganic-Organic Layered Perovskite-Type
Hybrids [(C H NH ) PbI ], n = 7, 8, 9 and 10. Dalton Trans. 2012,
n
2n+1
3
2
4
4
4
2
4
4 2
4
1, 1146−1157.
85) Billing, D. G.; Lemmerer, A. Synthesis, Characterization and
Phase Transitions of the Inorganic-Organic Layered Perovskite-Type
Hybrids [(C H NH ) PbI ] (n = 12, 14, 16 and 18). New J. Chem.
8
(
40−844.
(
66) McCarthy, T. J.; Kanatzidis, M. G. Synthesis in Molten Alkali
Metal Polyselenophosphate Fluxes: A New Family of Transition Metal
Selenophosphate Compounds, A MP Se (A = K, Rb, Cs; M = Mn, Fe)
n
2n+1
3
2
4
2
2
26
2
(
008, 32, 1736−1746.
and A M’ P Se (A = K, Cs; M’ = Cu, Ag). Inorg. Chem. 1995, 34, 1257−
2
2
2
6
86) Aleksandrov, K. S.; Beznosikov, B. V.; Misyul, S. V. Successive
1267.
(
67) Haynes, A. S.; Saouma, F. O.; Otieno, C. O.; Clark, D. J.;
Phase Transitions in Crystals of K MgF -Type Structure. Phys. Status
2 4
Shoemaker, D. P.; Jang, J. I.; Kanatzidis, M. G. Phase-Change Behavior
and Nonlinear Optical Second and Third Harmonic Generation of the
One-Dimensional K(1‑x)Cs PSe and Metastable β-CsPSe . Chem. Mater.
Sol. A 1987, 104, 529−543.
(87) Aleksandrov, K. S.; Bartolome, J. Octahedral Tilt Phases in
Perovskite-Like Crystals with Slabs Containing an Even Number of
Octahedral Layers. J. Phys.: Condens. Matter 1994, 6, 8219−8235.
x
6
6
2
015, 27, 1837−1846.
O
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials
Article
(
88) Glazer, A. The Classification of Tilted Octahedra in Perovskites.
Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem. 1972, 28, 3384−
392.
89) Axtell, E. A.; Liao, J. H.; Pikramenou, Z.; Kanatzidis, M. G.
Dimensional Reduction in II-VI Materials: A Cd Q (A = K, Q = S, Se,
(106) Hollis, D. B. Review of Hyper-Rayleigh and Second-Harmonic
Scattering in Minerals and Other Inorganic Solids. Am. Mineral. 1988,
73, 701−706.
3
(
(107) Kurtz, S. K.; Perry, T. T. A Powder Technique for the Evaluation
of Nonlinear Optical Materials. J. Appl. Phys. 1968, 39, 3798−3813.
2
3
4
(
́
108) Baledent, V.; Chattopadhyay, S.; Fertey, P.; Lepetit, M. B.;
Te; A = Rb, Q = S, Se), Novel Ternary Low-Dimensional Cadmium
Chalcogenides Produced by Incorporation of A Q in CdQ. Chem. - Eur.
J. 1996, 2, 656−666.
(
Greenblatt, M.; Wanklyn, B.; Saouma, F. O.; Jang, J. I.; Foury-Leylekian,
P. Evidence for Room Temperature Electric Polarization in RMn O
5
2
2
Multiferroics. Phys. Rev. Lett. 2015, 114, 117601.
109) Tronc, P.; Kitaev, Y. E.; Panfilov, A. G.; Limonov, M. F.; Wang,
90) Axtell, E. A.; Park, Y.; Chondroudis, K.; Kanatzidis, M. G.
(
Incorporation of A Q into HgQ and Dimensional Reduction to
2
G.; Smirnov, V. P. Bound-state Symmetries and Optical Transitions in
GaAs/AlAs Quantum Wells and Superlattices with Impurities and
Defects. Phys. Rev. B: Condens. Matter Mater. Phys. 2000, 61, 1999−2007.
A Hg Q and A Hg Q (A = K, Rb, Cs; Q = S, Se). Access of Li Ions in
2
3
4
2
6
7
A Hg Q Through Topotactic Ion-Exchange. J. Am. Chem. Soc. 1998,
2
6
7
1
(
20, 124−136.
(110) Bertoluzzi, L.; Sanchez, R. S.; Liu, L.; Lee, J.-W.; Mas-Marza, E.;
91) Androulakis, J.; Peter, S. C.; Li, H.; Malliakas, C. D.; Peters, J. A.;
Han, H.; Park, N.-G.; Mora-Sero, I.; Bisquert, J. Cooperative Kinetics of
Liu, Z. F.; Wessels, B. W.; Song, J. H.; Jin, H.; Freeman, A. J.; Kanatzidis,
M. G. Dimensional Reduction: A Design Tool for New Radiation
Detection Materials. Adv. Mater. 2011, 23, 4163−4167.
Depolarization in CH NH PbI Perovskite Solar Cells. Energy Environ.
3
3
3
Sci. 2015, 8, 910−915.
(111) Kim, H.-S.; Kim, S. K.; Kim, B. J.; Shin, K.-S.; Gupta, M. K.; Jung,
(
92) Umari, P.; Mosconi, E.; De Angelis, F. Relativistic GW
H. S.; Kim, S.-W.; Park, N.-G. Ferroelectric Polarization in CH NH PbI
3
3
3
Calculations on CH NH PbI and CH NH SnI Perovskites for Solar
3
3
3
3
3
3
Perovskite. J. Phys. Chem. Lett. 2015, 6, 1729−1735.
Cell Applications. Sci. Rep. 2014, 4, 4467.
93) Sebastian, M.; Peters, J. A.; Stoumpos, C. C.; Im, J.; Kostina, S. S.;
(
Liu, Z.; Kanatzidis, M. G.; Freeman, A. J.; Wessels, B. W. Excitonic
Emissions and Above-Band-Gap Luminescence in the Single-Crystal
Perovskite Semiconductors CsPbBr3 and CsPbCl . Phys. Rev. B:
3
Condens. Matter Mater. Phys. 2015, 92, 235210.
(
94) D’Innocenzo, V.; Grancini, G.; Alcocer, M. J. P.; Kandada, A. R. S.;
Stranks, S. D.; Lee, M. M.; Lanzani, G.; Snaith, H. J.; Petrozza, A.
Excitons Versus Free Charges in Organo-Lead Tri-Halide Perovskites.
Nat. Commun. 2014, 5, 3586.
(
95) Miyata, A.; Mitioglu, A.; Plochocka, P.; Portugall, O.; Wang, J. T.
W.; Stranks, S. D.; Snaith, H. J.; Nicholas, R. J. Direct Measurement of
the Exciton Binding Energy and Effective Masses for Charge Carriers in
Organic-Inorganic Tri-Halide Perovskites. Nat. Phys. 2015, 11, 582−
U94.
(
96) Lin, Q.; Armin, A.; Nagiri, R. C. R.; Burn, P. L.; Meredith, P.
Electro-Optics of Perovskite Solar Cells. Nat. Photonics 2014, 9, 106−
12.
97) Mak, K. F.; Lee, C.; Hone, J.; Shan, J.; Heinz, T. F. Atomically
Thin MoS : A New Direct-Gap Semiconductor. Phys. Rev. Lett. 2010,
1
(
2
105, 136805.
(
98) Ramasubramaniam, A. Large Excitonic Effects in Monolayers of
Molybdenum and Tungsten Dichalcogenides. Phys. Rev. B: Condens.
Matter Mater. Phys. 2012, 86, 115409.
(
99) Jones, E. D.; Drummond, T. J.; Hjalmarson, H. P.; Schirber, J. E.
Photoluminescence Studies of GaAs/AlAs Short Period Superlattices.
Superlattices Microstruct. 1988, 4, 233−236.
(
100) Quarti, C.; Mosconi, E.; De Angelis, F. Interplay of Orientational
Order and Electronic Structure in Methylammonium Lead Iodide:
Implications for Solar Cell Operation. Chem. Mater. 2014, 26, 6557−
6569.
(
101) Liao, W.-Q.; Zhang, Y.; Hu, C.-L.; Mao, J.-G.; Ye, H.-Y.; Li, P.-F.;
Huang, S. D.; Xiong, R.-G. A Lead-Halide Perovskite Molecular
Ferroelectric Semiconductor. Nat. Commun. 2015, 6, 7338.
(
102) Fan, Z.; Xiao, J.; Sun, K.; Chen, L.; Hu, Y.; Ouyang, J.; Ong, K. P.;
Zeng, K.; Wang, J. Ferroelectricity of CH NH PbI Perovskite. J. Phys.
3
3
3
Chem. Lett. 2015, 6, 1155−1161.
103) Xiao, Z.; Yuan, Y.; Shao, Y.; Wang, Q.; Dong, Q.; Bi, C.; Sharma,
(
P.; Gruverman, A.; Huang, J. Giant Switchable Photovoltaic Effect in
Organometal Trihalide Perovskite Devices. Nat. Mater. 2014, 14, 193−
198.
(
104) Coll, M.; Gomez, A.; Mas-Marza, E.; Almora, O.; Garcia-
Belmonte, G.; Campoy-Quiles, M.; Bisquert, J. Polarization Switching
and Light-Enhanced Piezoelectricity in Lead Halide Perovskites. J. Phys.
Chem. Lett. 2015, 6, 1408−1413.
(
105) Butler, K. T.; Frost, J. M.; Walsh, A. Ferroelectric Materials for
Solar Energy Conversion: Photoferroics Revisited. Energy Environ. Sci.
015, 8, 838−848.
2
P
DOI: 10.1021/acs.chemmater.6b00847
Chem. Mater. XXXX, XXX, XXX−XXX