The Role of Surface Termination in Halide Perovskites for Efficient Photocatalytic Synthesis

Article abstract of DOI:10.1002/anie.202002939

Halide perovskites have received attention in the field of photocatalysis owing to their excellent optoelectronic properties. However, the semiconductor properties of halide perovskite surfaces and the influence on photocatalytic performance have not been systematically clarified. Now, the conversion of triose (such as 1,3-dihydroxyacetone (DHA)) is employed as a model reaction to explore the surface termination of MAPbI₃. By rational design of the surface termination for MAPbI₃, the production rate of butyl lactate is substantially improved to 7719 μg g^?1 cat. h^?1 under visible-light illumination. The MAI-terminated MAPbI₃ surface governs the photocatalytic performance. Specially, MAI-terminated surface is susceptible to iodide oxidation, which thus promotes the exposure of Pb^II as active sites for this photocatalysis process. Moreover, MAI-termination induces a p-doping effect near the surface for MAPbI₃, which facilitates carrier transport and thus photosynthesis.

Full text of DOI:10.1002/anie.202002939

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: The Role of Surface Termination in Halide Perovskite for Efficient
Photocatalytic Synthesis
Authors: Qi Chen, Yuanyuan Dong, Kailin Li, Wenjia Luo, Cheng Zhu,
Haoliang Guan, Hao Wang, Lanning Wang, Kailin Deng,
Huanping Zhou, Haipeng Xie, Yang Bai, and Yujing Li
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202002939
Link to VoR: https://doi.org/10.1002/anie.202002939

10.1002/anie.202002939
Angewandte Chemie International Edition
RESEARCH ARTICLE
The Role of Surface Termination in Halide Perovskite for Efficient
Photocatalytic Synthesis
Yuanyuan Dong,[a] Kailin Li,[a] Wenjia Luo,[b] Cheng Zhu,[a] Haoliang Guan,[a] Hao Wang,[a] Lanning
Wang,[a] Kailin Deng,[c] Huanping Zhou,[d] Haipeng Xie,[e] Yang Bai,[a] Yujing Li*[a] and Qi Chen*[a]
[a]
Dr Y. Dong, K. Li, Ch. Zhu, H. Guan, H. Wang, L. Wang, Dr Y. Bai, Prof. Y. Li and Prof. Q. Chen.
Experimental Center of Advanced Materials, Beijing Key Laboratory of Nanophotonics and Ultrafine Optoelectronic Systems, School of Materials Science
and Engineering. Beijing Institute of Technology. Beijing 100081, P.R. China.
E-mail: qic@bit.edu.cn , yjli@bit.edu.cn
[b]
[c]
Prof. W. Luo, School of Chemistry and Chemical Engineering, Southwest Petroleum University, Chengdu, 610500, P.R. China.
Dr K. Deng. Key Laboratory of Agro-products Quality and Safety Control in Storage and Transport Process, Ministry of Agriculture and Rural
Affairs/Institute of Food Science and Technology, Chinese Academy of Agricultural Sciences, Beijing 100193, P.R. China.
Prof. H. Zhou. Department of Materials Science and Engineering, College of Engineering, Peking University, Beijing 100871, P. R. China
Prof. H. Xie. School of Physics and Electronics, Central South University, Changsha, Hunan 410083, P.R. China.
[d]
[e]
Supporting information for this article is given via a link at the end of the document. ((Please delete this text if not appropriate))
Abstract: Halide perovskites have received extensive attention in the
field of photocatalysis owing to their excellent optoelectronic
properties. However, the semiconductor properties of halide
perovskite surface and the influences on the photocatalytic
performances have not been systematically clarified. Here we employ
the conversion of triose (e.g. 1,3-dihydroxyacetone (DHA)) as a model
reaction to explore the surface termination of MAPbI3. By rational
design of the surface termination for MAPbI3, we substantially improve
the production rate of butyl lactate to 7719 μg/g cat./h under visible
light illumination. It reveals that MAI-terminated MAPbI3 surface
governs the photocatalytic performance. Specially, MAI-terminated
surface is susceptible to iodide oxidation, which thus promotes the
exposure of Pb(II) as active sites for this photocatalysis process.
Moreover, MAI-termination induces p-doping effect near the surface
for MAPbI3, which facilitates the carrier transport and thus
photosynthesis. The understanding of the surface termination in
MAPbI3 photocatalyst suggests a new avenue to engineer halide
perovskite semiconductor materials for future photocatalysis.
the APbBr3 (A = Cs or MA) type perovskite colloids can selectively
catalyze the formation of C-C bond via a new mode of small
molecule
activation.[6e]
The
as-prepared
perovskite
photocatalysts NCs are effective with a TON of over 52,000 for α-
alkylation, 3 orders of magnitude higher than conventional Ir- or
Ru-based catalysts. Moreover, the perovskite photocatalysts are
much more economical than conventional organometallic
compound (2-order cost lower). The efficient photocatalytic
performances and low-cost render halide perovskite as new
promising candidates for broad application in photocatalytic
chemical synthesis.
In the aspect of photocatalysis, controlling the semiconducting
properties of photocatalysts is the primary concern, because they
determine how much photoexcitation occurs in a semiconductor
under solar illumination and how many photoexcited carriers
reach the surface where the photocatalytic reaction takes
place.[10] Given a specific semiconductor material, surface
modifications are important not only to activate the semiconductor
but also to facilitate charge separation under photoexcitation in
photocatalytic process. It is reported that the electronic
conductivity of bulk MAPbI3 can be tuned between p-type and n-
type by controlling growth conditions.[11] However, the
semiconductor properties of the halide perovskite materials
surface have not been systematically reported, especially in the
field of photocatalysis. Specially, MAPbI3 has two major surface
termination (MAI and PbI2).[12] MAI-terminated surface are easily
prone to expose the Pb(II) sites caused by the oxidation of iodide,
while the PbI2-terminated surface are relative robust.[13] The
difference in surface termination of MAPbI3 will affect the
activation of reactant molecules and the carrier migration in the
photocatalytic process to some extent. Therefore, it is highly
desirable to study the termination, clarify the surface species and
the surface semiconductor properties of MAPbI3 perovskite.
Herein, we employ the conversion of triose (e.g. DHA) as a
model reaction, commonly using Pb (II) as a catalyst,[14] to explore
the surface termination of MAPbI3, and further investigate its
effects on photocatalytic performance. MAPbI3 photocatalysts are
readily prepared via the coevaporation method under the certain
relative humidity in the customized golove box (Figure 1a and
Scheme S1). The photocatalytic synthesis reaction of butyl lactate
from DHA and butanol is shown in Figure 1b. We demonstrate the
yield of butyl lactate is 77 mg/L over MAPbI3 photocatalyst.
Introduction
The hybrid ABX3 perovskite semiconductors have
revolutionized photovoltaics, enabling solution-processable solar
cells to reach the power conversion efficiency of 25.2%.[1] They
have also been demonstrated to show good performance in high-
gain photodetectors,[2] light-emitting diodes,[3] and lasers[4]. The
excellent optoelectronic properties are attributed to strong light
absorption, long charge-carrier lifetimes and diffusion lengths.[5]
Given the above beneficial optoelectronic properties, the halide
perovskite materials have also received extensive attention in the
field of photocatalysis.[6] Under visible light irradiation, halide
perovskite photocatalysts could catalyze the dye degradation,[7]
[6a]
water splitting,[6f] the reduction of CO2
and the oxidation of
benzylic alcohols.[8] To achieve good separation/migration of
photoexcited holes and electrons, many composite/heterojunction
photocatalytic
semiconductors
have
been
developed.
MAPbI3/rGO, CsPbBr3 NC/BZNW/MRGO, α-Fe2O3/Amine-
RGO/CsPbBr3, CPB-PCN, CsPbBr3/Cs4PbBr6 and FAPbBr3/TiO2
hybrid photocatalyst are prepared and exhibit the improved
photocatalytic performance.[6c, 9] Recently, Yan et al. discover that
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202002939
Angewandte Chemie International Edition
RESEARCH ARTICLE
According to our investigation, this is the first report of
photocatalytic synthesis of butyl lactate from triose under visible
(GCMS) and confirms the formation of target product butyl lactate
(Figure S1). Figure 1f shows that MAPbI3 exhibits good
photocatalytic performances with the production rate of 5837 μg/g
cat./h and the yield of 58.4 mg/L, respectively. Other perovskite
materials have also been tested for this reaction (Figure 1f and
Table S1). It can be seen that MAPbI3 and FAPbI3 could
effectively convert DHA into butyl lactate. Although the conversion
of DHA reaches 15%, the Br-based perovskite materials are
inactive for the synthesis of butyl lactate. It implies that some other
products are produced in this photocatalytic process (Figure S1).
Moreover, the MAPbBrI-x (x=Br/I molar ratio) could promote the
photocatalytic synthesis of butyl lactate (Figure S2). The yield of
butyl lactate is positively related with the amount of I.
light illumination at room temperature.
A
series of
characterizations indicate that the desirable MAI-terminated
MAPbI3 perovskite surface facilitates the effective photocatalytic
synthesis of butyl lactate. MAI-terminated surface are easily
prone to expose the Pb(II) sites caused by the oxidation of iodide,
which are considered as the active sites for this photocatalytic
synthesis. Furthermore, MAI-terminated induces p-doping effect
near the surface for MAPbI3. A photocatalytic mechanism is
hence proposed. Upon solar light illumination, electrons and holes
are generated in MAPbI3. The photo-generated holes are
consumed to oxidize DHA into pyruvaldehyde, whereas the
exposed Pb(II) sites on MAI-terminated MAPbI3 surface
contribute to promote the pyruvaldehyde and butanol substrates
towards the final product.
Given the reported work on the conversion of triose catalyzed
by Pb(II) ions at the relative high temperature[14], it is supposed
that the naked Pb(II) sites on the surface of MAPbI3 play a crucial
role in this photocatalytic reaction. Due to the chemical instability,
MAPbI3 will decompose to HI, CH3NH2 and PbI2, etc. in polar
solvent.[16] PbI2 is confirmed to be inactive towards this
photocatalytic reaction (Table S2). Additionally, the amount of Pb
dissolved in the reaction supernatant is only 7.8 μg/mL
(determined by ICP) and hence can be considered to be negligible
compared with the amount of MAPbI3 catalyst (3.3 mg/mL in
terms of Pb). We further conduct a contrast experiment (Table S2)
to exclude the possible influence of the leaking Pb(II) ions. The
oleophilic TiO2 is added into the reactant solution with free Pb(II)
ions and irradiated under the same condition. No butyl lactate can
be detected. Therefore, we reason that the leaking Pb(II) ions in
solution are inactive. The naked Pb(II) sites in inorganic-
framework on the surface of MAPbI3 are considerd as the catalytic
sites for the photocatalytic synthesis of butyl lactate. Owing to the
lower oxidation potential, iodide is easily oxidized to form iodide
vacancies, leading to the exposed Pb(II) species on the surface
of MAPbI3.[13b] To confirm this speculation, it is important to
understand the surface termination of MAPbI3. In view of the
crystal structure of this ionic compound, there are two major
termination: 1) MA+ cations hydrogen bonded with inorganic
Results and Discussion
2-
scaffold, and 2) partially covalent I- in PbI4 octahedron.
Unfortunately, these two species are most likely catalytically inert
based on control experiment (Table S2). Considering the
difference in ionic radius between halogen ions and the formation
energy of halide perovskites[17], the surface defects may be
responsible for the different photocatalytic results of halide
perovskites (Figure 1f and Figure S2). It is reported that the
formation energy of iodine vacancies (VI) is low, which favors the
formation of defects and the corresponding naked Pb(II) sites,
especially at the surface.[13a] We hence propose that the positively
charged Pb resulting from the iodine vacancies/defects, serving
as the central sites, are responsible for the catalytic process.
To identify the role of Pb, methimazole (MMI) is selected to
passivate the exposed Pb(II) sites based on the Lewis acid-base
theory (Figure 2a),[18] wherein the PbI2·MMI/MAI·PbI2·MMI adduct
is formed via the Pb-S coordination reaction. The diffraction peak
of 9.85° in XRD patterns (Figure 2b) confirms the formation of the
adduct, which is consistent with our previous work.[18a] The
Fourier-transform infrared spectra (FTIR) in Figure 2c also
provide detailed evidence about the successful passivation. The
fingerprint peaks associated to the S=C stretching for pristine MMI
at 738 cm−1 shifts to 735 cm−1 after MAPbI3 treated with MMI
(denoted as MAPbI3-S). The down-shift of the S=C vibration in
MMI is indicative of the interaction between Pb and S, leading to
Figure 1. (a) Scheme of the preparation of MAPbI3 via co-evaporation method.
(b) The schematic diagram of the conversion of DHA and butanol into butyl
lactate over catalyst. (c) XRD pattern, (d) SEM image and (e) UV-vis absorption
spectra of MAPbI3. (f) The photocatalytic activities of MAPbI3, FAPbI3 and
MAPbBr3 photocatalysts.
MAPbI3 powder is prepared using a simple co-evaporation
method with the customized glove box (Scheme S1) under a
relative humidity of 35%±5% (Figure 1a), with further procedural
details described in the Experimental Section. The powder X-ray
diffraction (XRD) patterns (Figure 1c) verify the tetragonal phase
of MAPbI3.[15] The scanning electron microscopy (SEM) image
(Figure 1d) shows that the average size of MAPbI3 is around 730
nm. The optical absorption edge of MAPbI3 is determined from
UV-Vis absorption spectrum (Figure 1e) with the typical band gap
of 1.50 eV (827 nm).[6f] With the verified structure, MAPbI3 is thus
employed to photocatalyze the reaction between DHA and
butanol as described in Figure 1b. The photocatalytic synthesis of
butyl lactate is conducted in a side-irradiation sealed quartz
vessel with a Xe-lamp (CEL-S500) and a 420 nm cut-off filter. The
product is analyzed by gas chromatograph-mass spectrometer
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202002939
Angewandte Chemie International Edition
RESEARCH ARTICLE
the weakened S=C bond strength. The electronic states of Pb and
S determined by X-ray photoelectron spectroscopy (XPS) further
confirm the interaction between Pb and S. As Figure 2d shows,
the peak at 138.1 eV is assigned to the Pb 4f 7/2 of MAPbI3. After
treatment with MMI, the existence of S at the surface is evidenced
by the observed apparent signals of S 2p. Furthermore, the Pb 4f
peak of MAPbI3-S significantly shifts to the higher binding energy.
It thus verifies that the exposed Pb(II) sites are successfully
passivated by MMI molecule at the perovskite surface via the Pb–
S coordination.
catalyze the reaction alone. Additionally, Table S2 shows that the
reaction cannot proceed with solely the visible light or the MAPbI3
at dark. Based on the above investigation, it is found that the
MAPbI3 with the exposed Pb(II) sites could effectively convert
DHA and butanol into butyl lactate at room temperature with
visible light.
Although the naked Pb(II) site has been confirmed as a terminal
species, it is the result of the iodide vacancies resulting from the
oxidation of surface iodine. In view of the two major termination
structure (MAI or PbI2), it is still necessary to further explore the
surface termination of MAPbI3. The preparation process (details
in SI) hence be investigated in detail because the surface of
MAPbI3 will restructure during the moisture annealing process.[19]
In a moisture annealing process, highly hygroscopic MA cations
pull moisture from the environment. Water molecules adsorb on
the surface vacancies or edges/boundaries, leading to the surface
solvation/dissolution of perovskite crystals, and then facilitate the
merging of initial primary crystals into larger grains.[20] SEM
images of MAPbI3-x (Figure S3) agree with the above reported
phenomenon. Therefore, the ambient humidity is an important
factor affecting the surface reconstruction. We thus prepare a
series of MAPbI3-x materials via moisture annealing technique in
the customized glove box (Scheme S1), where x represents the
relative humidity of 15%, 25%, 35%, 45% and 55% with the
fluctuation within 5%.
Table 1. The surface composition of moisture-annealed MAPbI3 determined by
XPS.
Catalyst
Pb
I
N
I/Pb
2.97
N/Pb
0.55
MAPbI3-
15%
17.69%
52.6%
9.69%
MAPbI3-
25%
17.07%
15.43%
16.02%
16.33%
53.07%
53.27%
54.58%
55.36%
9.72%
10.38%
9.83%
9.64%
3.11
3.45
3.41
3.39
0.57
0.67
0.61
0.59
MAPbI3-
35%
Figure 2. (a) The schematic diagram of the passivation of MAPbI3. (b) XRD
patterns of original and MAPbI3-S photocatalysts. (c) FTIR spectra of MMI,
original and passivated MAPbI3. (d) The Pb 4f and S 2p signal of MAPbI3 and
MAPbI3-S. (e) Steady-state PL spectra and (f) photocatalytic activities of MAPbI3
and MAPbI3-S photocatalysts.
MAPbI3-
45%
MAPbI3-
55%
Upon the effective passivation, the exposed Pb(II) sites are
covered by MMI, which implies that the iodine vacancies in
MAPbI3 are efficiently repaired. The crystal structure with fewer
defects is evidenced by the increased intensity of XRD diffraction
peak of MAPbI3-S sample (Figure 2b). Photoluminescence (PL)
spectra further verifies the well-passivated surface of MAPbI3-S
(Figure 2e). Compared with MAPbI3 reference, the PL intensity of
MAPbI3-S significantly enhanced, which further confirms the
defects are effectively healed. Namely, the exposed Pb(II) sites at
the MAPbI3 surface are well poisoned by MMI. The photocatalytic
activities of MAPbI3-S should decrease because of the sheltered
Pb(II) sites. As Figure 2f shows, after treatment with 5wt.% MMI,
the yield declines to only 10% of the untreated one. By further
increasing the amount of poisoning species to 10 wt.%, the target
product butyl lactate cannot be detected. The results serve as
solid evidences that the naked Pb(II) sites on MAPbI3 surface play
an essential role in this reaction. Moreover, the control
experiments (Table S2) are carried out under the same reaction
conditions. Neither PbI2, MAI and the leaking Pb(II) ions can
The surface composition is determined by XPS in Figure 3a and
Table 1. Under the lower relative humidity (≤ 35%±5%), the
amount of I and N gradually increase with the humidity, while the
amount of Pb decreases. According to the moisture annealing
process, the hydrophilic MA cations pull moisture from the
environment, facilitating the diffusion of MA cations.[21] Therefore,
it is deduced that the MAI gradually enriches on the surface with
the increasing humidity, and reaches a maximum at the humidity
of 35%. Namely, MAPbI3 surface is terminated by MAI under the
lower humidity (≤ 35%±5%), resulting in a more p-type feature on
the surface. When the humidity further increases, the amount of
N decreases on the MAPbI3-x surface, accompanied with the
increment of the Pb amount (Table 1). It implies that MAPbI3
surface is terminated by PbI2 under the relative high humidity
(›35%±5%), leading to a more n-type feature on the surface.
These observations can be attributed to the self-doping effects
during the moisture annealing process.[22]
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202002939
Angewandte Chemie International Edition
RESEARCH ARTICLE
The self-doping effects are further confirmed by the valence
band spectra in Figure 3b. Under the low relative humidity (≤
35%±5%), the Fermi level of MAPbI3 moves from 1.27 to 0.31 eV
from the valance band maximum (VBM) with the increase of
humidity. Given the band gap of 1.5 eV (Figure 1e) for the
perovskite, the shift puts the Fermi level very close to the VBM. It
suggests that the surface of MAPbI3 tends to be a more p-type
feature. At the high humidity (›35%±5%), the Fermi level from
VBM increases from 0.31 to 1.29 eV, which implies that Fermi
level moves close to the conduction band minimum (CBM). The
surface of MAPbI3 tends to be a more n-type feature at this point.
Furthermore, we have employed density functional theory
(DFT) to clarify the band structure of the as-designed MAPbI3.
The details about computation methods and structure models are
described in the SI. Figure 3c and Figure S4 show the band
structure of MAI-terminated and PbI2-terminated MAPbI3,
respectively. The band structure is significantly different between
them. After the vacuum energy level alignment, the schematic
diagram of the two energy band structures is described in Figure
3d. Compared with PbI2-terminated MAPbI3, the VBM of MAI-
terminated MAPbI3 slightly shifts up and is closer to the vacuum
energy level. It suggests that MAI-terminated MAPbI3 tends to be
a more p-type feature on the surface, which is consistent with our
experimental results.
curves
by
the
biexponential
equation:
Y=A1exp(−t/τ1)+A2exp(−t/τ2), where the short lifetime τ1 stands for
the surface nonradiative recombination and the long lifetime τ2 for
the bulk recombination. As the humidity increases, the surface
nonradiative recombination lifetime τ1 displays the similar trend to
that of microstrain (Table S3). It is deduced that the microstrain
stemming from the reconstruction could lead to the formation of
surface defects (e.g. VI), resulting in the exposure of Pb(II) sites.
The trend of the photocatalytic activity described in Figure S10
further proves it. However, MAPbI3-15% seems to be an
exception, which may originate from the excessive defects
caused by an extreme microstrain.
The photocatalytic performances of the MAPbI3-x materials are
shown in Figure 3a. Clearly, the yield of butyl lactate depends on
the humidity, reaching the maximum of 58.4 mg/L with the
humidity of 35%. The production rate of target product also follows
the similar trend (Figure S5). Combined with the changes of the
surface
N amount with humidity, it is inferred that the
photocatalytic activities could be improved with the enrichment of
MAI on MAPbI3 surface. Namely, MAI-terminated surface could
facilitate the photocatalytic reaction. This should be attributed to
the fact that the iodine species on the MAI-terminated surface are
easily oxidized, leading to the exposed Pb(II) as the terminating
species in MAPbI3.[13b, 23] Whereas, the augmentation of PbI2 on
the MAPbI3 surface would decrease the photocatalytic activities.
Because the PbI2-terminated surface is robust owing to the
stronger (shorter) Pb-I bonds, it is not easy to form iodide
vacancies and the resulting naked Pb(II) sites.[12] DFT is further
employed to calculate the formation energies of iodine vacancies
between MAI-terminated and PbI2-terminated MAPbI3. The top
view of the MAI- and PbI2-terminated MAPbI3 (110) surfaces
without and with iodine vacancy is shown in Figure S6. The
calculation suggests that the formation energy of iodine defects is
slightly higher on PbI2-terminated (ΔG=1.80 eV) than on MAI-
terminated (ΔG=1.74 eV) MAPbI3, which is consistent with the
experimental results.
Figure 3. (a) The relationship plots among the amount of N and I determined
by XPS, the yield of butyl lactate and the relative humidity. (b) the valence band
spectra of the MAPbI3-x materials. (c) Band structure of MAI-terminated MAPbI3
with iodine vacancies. (d) The schematic diagram of the energy band structure
of MAI-terminated and PbI2-terminated MAPbI3 upon the vacuum energy level
alignment. (e) high resolution Pb 4f XPS spectra of the MAPbI3-x materials. (f)
The schematic diagram of the photocatalytic mechanism for the synthesis of
butyl lactate over MAPbI3.
Additionally, the surface reconstruction, stemming from the
destruction and reconstruction of the bond between MA+ and the
Pb-I octahedron,[24] leads to the crystal lattice distortion and the
corresponding microstrain.[20b] The resulted microstrain can
mainly be decerned in the FWHM of XRD diffraction peaks of
MAPbI3-x (Figure S7). Figure S8 shows the calculated microstrain
of MAPbI3-x based on the modified Williamson-Hall (W-H) method
(Eq. S1).[25] It is found that the microstrain of MAPbI3-x basically
displays a “volcano-shaped” dependence with the humidity. The
microstrain resulting from the surface reconstruction possibly
induce the formation of defects including the surface and bulk
defects. The carrier lifetime determined by the TRPL (Figure S9)
can distinguish these defects to some extent. We fit the decay
Furthermore, the destruction and reconstruction of the bond
between MA+ and the Pb-I octahedron would also lead to changes
in the electronic properties of elements.[26] Specifically, the
binding energy of Pb 4f first shifts to lower binding energy and
then shifts to higher binding energy (Figure 3e). Not only for Pb,
the shifts of the binding energy of other elements show the similar
trend (Figure S11). According to the previous results, the shift to
low binding energy indicating the elongated bond of Pb-I, which
suggests that the Fermi energy level moves downward in terms
of band structure.[27] It suggests the surface of semiconductor
tends to be a more p-type feature, which is consistent with the
results of valance band spectra (Figure 3b). On the contrary, the
shift to high binding energy indicates a more n-type nature of the
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202002939
Angewandte Chemie International Edition
RESEARCH ARTICLE
surface. The influences of surface self-doping on the
photocatalytic activities are described in Figure S12. With the
addition of hole scavenger, the production rate of target product
drops to 4%, which verifies the crucial role of photo-generated
hole in this photocatalysis system.
DHA requires additional energy to adsorb on the surface of
MAPbI3 photocatalyst without iodine vacancies. Obviously, the
exposure of Pb(II) sites is very important for the adsorption and
activation of DHA on MAPbI3 photocatalyst, whereas it is the first
prerequisite step of the reaction.
Based on the above analyses, it can be inferred that MAI-
terminated MAPbI3 could facilitate the photocatalytic synthesis of
butyl lactate. Thereinto, MAI-terminated surface induces the p-
type self-doping of MAPbI3, which promote the oxidation of DHA
to pyruvaldehyde.[28] The GC spectrum of Figure S13 confirms the
pyruvaldehyde in the reaction product. Moreover, the
pyruvaldehyde further reacts with butanol to produce butyl lactate
via the path of K-E (consecutive keto–enol tautomerization) or
1,2-H (1,2-hydride shift), where the exposed Pb(II) on MAI-
terminated MAPbI3 surface decreases the activation barrier.[14, 29]
Hereby, we propose a photocatalytic mechanism as illustrated in
Figure 3f. Upon the visible light illumination, electrons and holes
are generated in MAPbI3. The photogenerated holes are
consumed to oxidize the reactant DHA into pyruvaldehyde (Figure
S13), whereas the photogenerated electrons are utilized by the
exposed Pb(II) to promote the pyruvaldehyde and butanol
substrates to produce the final product butyl lactate.
To further improve the photocatalytic activities, it is an effective
strategy to expose more Pb(II) sites on the surface of MAPbI3 as
previously mentioned. It has been reported that iodide is easy to
be oxidized even in ambient processing due to its low oxidation
potential.[13b] We thus conducted ageing process to create more
surface defects or vacancies (e.g. VI, VMA) to regulate the
exposure of Pb(II) sites. The aged sample is designated as
MAPbI3-A1 and MAPbI3-A2 with the different ageing time
(Experiment section in supporting information). The fresh MAPbI3
and aged counterpart show the similar XRD patterns (Figure 5a),
indicating that the crystal structure of MAPbI3 did not change
significantly. However, the full width at half maximum (FWHM) of
the (220) crystal plane gradually increases from 0.147 to 0.164,
which may result from the microstrain between perovskite crystals
during ageing. The broadening of the strongest peak suggests a
reduced crystallinity of the materials.[30] It implies that the surface
defects of MAPbI3 may increase after ageing. Figure 5b depicts
the steady-state PL spectra of the fresh and aged MAPbI3. The
decreased PL intensity indicate the enhanced non-radiative
recombination, which results from the augmented surface defects
of the aged sample (MAPbI3-A1 and MAPbI3-A2). The broadening
of the emission spectra of the aged samples comes from the
hydration and/or hydrogen bonding interactions between water
molecules, organic cation (MA+) and halide ions,[31] leading to the
formation of new surface defects. A blue shift of 16 nm for PL peak
of MAPbI3-A1 also suggests the formation of additional defects
after ageing.[32]
Figure 4. (a) Proposed reaction mechanism based on reference 14 for the
conversion of DHA to butyl lactate. (b) Free energy diagram for the conversion
of DHA into butyl lactate over MAI-terminated MAPbI3 photocatalyst with iodine
vacancies under irradiated condition.
To further confirm the crucial role of the exposed Pb(II) sites
caused by the iodine vacancies, we simulate the adsorption of
reactants, intermediate and products on the surface of MAI-
terminated MAPbI3 photocatalyst with and without iodine
vacancies based on this reaction mechanism (Figure 4a). Top and
side view of the slab for MAI-terminated MAPbI3 with and without
iodine vacancy are shown in S14. The geometries of the adsorbed
intermediates are listed in Table S4. The crucial role of the
exposed Pb(II) sites caused by the iodine vacancy is obvious.
When an iodine vacancy is present, one of the octahedrons on
the surface is missing a corner and the adjacent Pb site is
exposed. As an adsorbate molecule approaches the surface, one
oxygen atom from the adsorbate bonds with the exposed Pb atom
(Top and side view of state 2 in the first two columns of Table S4).
The resulting adsorption free energy of DHA (Δ퐺1→2) is -0.69 eV
(Figure 4b), which implies that DHA can spontaneously adsorb on
the surface of MAPbI3 photocatalyst with iodine vacancies. The
activation of DHA and subsequent reaction can be continued.
However, the cleans surface of MAPbI3 without iodine vacancies
is chemically saturated, which attracts the adsorbates mainly
through dispersion effects (Top and side view of state 2 in the last
two columns of Table S4), leading to weak adsorption and
difficulty in the activation of reactant. The resulting adsorption free
energy of DHA (Δ퐺1→2) is 0.02 eV (Figure S15), which implies that
Figure 5. (a) XRD patterns, (b) steady-state PL spectra, (c) the photocatalytic
activities and (d) UV-vis absorption spectra of fresh and aged MAPbI3.
The fresh and aged MAPbI3 are hence evaluated for the photo-
catalytic tests (Figure 5c). Compared with MAPbI3, the aged
sample MAPbI3-A1 exhibits an enhanced performance in
conversion, yield and production rate of butyl lactate from triose.
According to the abovementioned investigations, the improved
photocatalytic activities should be attributed to the more exposed
Pb(II) sites caused by the augmented surface defects. However,
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202002939
Angewandte Chemie International Edition
RESEARCH ARTICLE
[1]
[2]
[3]
[4]
[5]
Best Research-Cell Efficiency Chart 2019, https://www.nrel.gov/pv/cell-
efficiency.html.
J. Wang, J. Li, S. Lan, C. Fang, H. Shen, Q. Xiong, D. Li, ACS Nano
2019, 13, 5473-5484.
L. Zhao, K. M. Lee, K. Roh, S. U. Z. Khan, B. P. Rand, Adv. Mater.
2019, 31, 1805836.
N. Zhang, Y. Fan, K. Wang, Z. Gu, Y. Wang, L. Ge, S. Xiao, Q. Song,
Nat. Commun. 2019, 10, 1770.
a) J. Peng, Y. Chen, K. Zheng, T. Pullerits, Z. Liang, Chem. Soc. Rev.
2017, 46, 5714-5729; b) Y. Dong, Y. Zhao, S. Zhang, Y. Dai, L. Liu, Y.
Li, Q. Chen, J. Mater. Chem. A 2018, 6, 21729-21746.
a) Y.-F. Xu, M.-Z. Yang, B.-X. Chen, X.-D. Wang, H.-Y. Chen, D.-B.
Kuang, C.-Y. Su, J. Am. Chem. Soc. 2017, 139, 5660-5663; b) K. Chen,
X. Deng, G. Dodekatos, H. Tüysüz, J. Am. Chem. Soc. 2017, 139,
12267-12273; c) H. Huang, H. Yuan, K. P. F. Janssen, G. Solís-
Fernández, Y. Wang, C. Y. X. Tan, D. Jonckheere, E. Debroye, J.
Long, J. Hendrix, J. Hofkens, J. A. Steele, M. B. J. Roeffaers, ACS
Energy Lett. 2018, 3, 755-759; d) Z. Zhang, Y. Liang, H. Huang, X. Liu,
Q. Li, L. Chen, D. Xu, Angew. Chem. Int. Ed. 2019, 58, 7263-7267; e)
X. Zhu, Y. Lin, Y. Sun, M. C. Beard, Y. Yan, J. Am. Chem. Soc. 2019,
141, 733-738; f) S. Park, W. J. Chang, C. W. Lee, S. Park, H.-Y. Ahn,
K. T. Nam, Nat. Energy 2016, 2, 16185; g) Z. Hong, W. K. Chong, A. Y.
R. Ng, M. Li, R. Ganguly, T. C. Sum, H. S. Soo, Angewandte Chemie
2019, 58, 3456-3460.
G. Gao, Q. Xi, H. Zhou, Y. Zhao, C. Wu, L. Wang, P. Guo, J. Xu,
Nanoscale 2017, 9, 12032-12038.
S. Schünemann, M. van Gastel, H. Tüysüz, ChemSusChem 2018, 11,
2057-2061.
a) Y. Wu, P. Wang, X. Zhu, Q. Zhang, Z. Wang, Y. Liu, G. Zou, Y. Dai,
M.-H. Whangbo, B. Huang, Adv. Mater. 2018, 30, 1704342; b) Y. Jiang,
J.-F. Liao, Y.-F. Xu, H.-Y. Chen, X.-D. Wang, D.-B. Kuang, J. Mater.
Chem. A 2019, 7, 13762-13769; c) Y. Jiang, J.-F. Liao, H.-Y. Chen, H.-
H. Zhang, J.-Y. Li, X.-D. Wang, D.-B. Kuang, Chem 2020; d) Angew
Chem Int Ed 2018; e) P. Qiu, Q. Wang, Y. Zhao, Y. Dai, Y. Dong, C.
Chen, Q. Chen, Y. Li, Frontiers in Chemistry 2020, 8.
the photocatalytic activities of MAPbI3-A2 obviously decreases
(Figure 5c). The significantly reduced PL and XRD peak intensity
(Figure 5a and 5b) indicates the further augmented defects of
MAPbI3-A2. Namely, the excessive ageing induces too much
surface defects and hence decreases the photocatalytic activities,
which may be attributed to the decline in the light absorption of
MAPbI3 (Figure 5d).
Furthermore, the varies in catalytic activities of MAPbI3 with
reaction time are shown in Figure S16. As the reaction going on,
the yield of butyl lactate increases almost linearly. The catalytic
stability of MAPbI3 is also investigated by recycling it in a
[6]
refreshed reaction atmosphere every
2 h of continuous
photocatalytic reaction (Figure S17). After cycling 3 times, the
photocatalytic activities of MAPbI3 decreases. The decreased
intensity of diffraction peaks and the appearance of PbI2 in XRD
patterns indicates that the crystal structure of MAPbI3 are partly
destroyed after 3 cycles (Figure S18). Some protection strategies
should be carried out to improve the photocatalytic stability of
MAPbI3 in the future. Additionally, Table S5 lists the catalytic
performance comparison between MAPbI3 and other
homogeneous/heterogeneous catalysts. Compared with the
condition with high temperature and high pressure, the visible light
irradiation at ambient temperature used in this work generates a
very competent performance.
[7]
[8]
[9]
[10] T. Hisatomi, J. Kubota, K. Domen, Chem. Soc. Rev. 2014, 43, 7520-
7535.
[11] Q. Wang, Y. Shao, H. Xie, L. Lyu, X. Liu, Y. Gao, J. Huang, Appl. Phys.
Lett. 2014, 105, 163508.
Conclusion
[12] E. Mosconi, J. M. Azpiroz, F. De Angelis, Chem. Mater. 2015, 27, 4885-
4892.
[13] a) C. Ran, J. Xu, W. Gao, C. Huang, S. Dou, Chem. Soc. Rev. 2018,
47, 4581-4610; b) Z. Liu, J. Hu, H. Jiao, L. Li, G. Zheng, Y. Chen, Y.
Huang, Q. Zhang, C. Shen, Q. Chen, H. Zhou, Adv. Mater. 2017, 29,
1606774.
[14] Y. Wang, W. Deng, B. Wang, Q. Zhang, X. Wan, Z. Tang, Y. Wang, C.
Zhu, Z. Cao, G. Wang, H. Wan, Nat. Commun. 2013, 4, 2141.
[15] H. Zhang, Z. Yang, W. Yu, H. Wang, W. Ma, X. Zong, C. Li, Adv.
Energy Mater. 2018, 8, 1800795.
[16] K. O. Brinkmann, J. Zhao, N. Pourdavoud, T. Becker, T. Hu, S. Olthof,
K. Meerholz, L. Hoffmann, T. Gahlmann, R. Heiderhoff, M. F. Oszajca,
N. A. Luechinger, D. Rogalla, Y. Chen, B. Cheng, T. Riedl, Nature
Communications 2017, 8, 13938.
[17] a) Y. Liu, K. Palotas, X. Yuan, T. Hou, H. Lin, Y. Li, S.-T. Lee, ACS
Nano 2017, 11, 2060-2065; b) W.-J. Yin, T. Shi, Y. Yan, Appl. Phys.
Lett. 2014, 104, 063903; c) A. Zohar, I. Levine, S. Gupta, O. Davidson,
D. Azulay, O. Millo, I. Balberg, G. Hodes, D. Cahen, ACS Energy Lett.
2017, 2, 2408-2414.
[18] a) L. Liu, S. Huang, Y. Lu, P. Liu, Y. Zhao, C. Shi, S. Zhang, J. Wu, H.
Zhong, M. Sui, H. Zhou, H. Jin, Y. Li, Q. Chen, Adv. Mater. 2018, 30,
1800544; b) P. D. Akrivos, Coordination Chemistry Reviews 2001, 213,
181-210; c) J.-W. Lee, H.-S. Kim, N.-G. Park, Accounts of Chemical
Research 2016, 49, 311-319.
[19] a) H. Zhou, Q. Chen, G. Li, S. Luo, T.-b. Song, H.-S. Duan, Z. Hong, J.
You, Y. Liu, Y. Yang, Science 2014, 345, 542-546; b) J. You, Y. Yang,
Z. Hong, T.-B. Song, L. Meng, Y. Liu, C. Jiang, H. Zhou, W.-H. Chang,
G. Li, Y. Yang, Appl. Phys. Lett. 2014, 105, 183902.
[20] a) Q.-Q. Ge, J. Ding, J. Liu, J.-Y. Ma, Y.-X. Chen, X.-X. Gao, L.-J. Wan,
J.-S. Hu, J. Mater. Chem. A 2016, 4, 13458-13467; b) S. R. Raga, M.-
C. Jung, M. V. Lee, M. R. Leyden, Y. Kato, Y. Qi, Chem. Mater. 2015,
27, 1597-1603.
In conclusion, we employ the conversion of triose (DHA) as a
probe reaction to investigate the surface termination in MAPbI3
and further explore its role on the photocatalytic process.
Specially, MAI-terminated MAPbI3 exhibits good photocatalytic
activities with the yield of 77 mg/L and the production rate of 7719
μg/g cat./h, respectively. By surface passivation experiment and
ageing experiments, we further confirm that the exposed Pb(II)
sites, caused by the oxidation of iodide on the MAI-terminated
MAPbI3 surface, play a crucial role in this photocatalytic synthesis
reaction. Moreover, MAI-termination induces a p-type doping on
MAPbI3 surface, facilitating the photocatalytic synthesis of butyl
lactate. The understanding on the surface regulation of MAPbI3
may bring in new insights in the design of halide perovskite
semiconductor materials for photocatalysis.
Acknowledgements
This work was supported by the National Natural Science
Foundation of China (21706080, 21975028, 51673025), China
Postdoctoral Science Foundation (2019T120054, 2018M631359),
National Key Research and Development Program of China
Grant (2016YFB0700700), the start-up funding of BIT and Young
Talent Thousand Program.
[21] G. E. Eperon, S. N. Habisreutinger, T. Leijtens, B. J. Bruijnaers, J. J.
van Franeker, D. W. deQuilettes, S. Pathak, R. J. Sutton, G. Grancini,
D. S. Ginger, R. A. J. Janssen, A. Petrozza, H. J. Snaith, ACS Nano
2015, 9, 9380-9393.
[22] a) S. Berweger, G. A. MacDonald, M. Yang, K. J. Coakley, J. J. Berry,
K. Zhu, F. W. DelRio, T. M. Wallis, P. Kabos, Nano Letters 2017, 17,
1796-1801; b) J. Huang, S. Tan, P. D. Lund, H. Zhou, Energy &
Environmental Science 2017, 10, 2284-2311.
[23] Y. Chen, N. Li, L. Wang, L. Li, Z. Xu, H. Jiao, P. Liu, C. Zhu, H. Zai, M.
Sun, W. Zou, S. Zhang, G. Xing, X. Liu, J. Wang, D. Li, B. Huang, Q.
Chen, H. Zhou, Nat. Commun. 2019, 10, 1112.
[24] G. Grancini, V. D'Innocenzo, E. R. Dohner, N. Martino, A. R. Srimath
Kandada, E. Mosconi, F. De Angelis, H. I. Karunadasa, E. T. Hoke, A.
Petrozza, Chem. Sci. 2015, 6, 7305-7310.
The authors also acknowledge Qin Du, Dr. Sai Ma, Dr. Lang Liu,
Dr. Changli Chen, Dr. Yizhou Zhao, Dr. Na Liu, Dr. Xiuxiu Niu and
Dr. Nengxu Li for helpful discussions.
Keywords: halide perovskite • surface termination • naked Pb(II)
sites • photocatalysis • triose
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202002939
Angewandte Chemie International Edition
RESEARCH ARTICLE
[25] Y. T. Prabhu, K. V. Rao, V. S. S. Kumar, B. S. Kumari, World Journal of
Nano Science and Engineering 2014, 4, 21-28.
[26] a) V. M. Pereira, A. H. Castro Neto, Phys. Rev. Lett. 2009, 103,
046801; b) S.-M. Choi, S.-H. Jhi, Y.-W. Son, Phys. Rev. B 2010, 81,
081407; c) L. Wang, Z. Zeng, W. Gao, T. Maxson, D. Raciti, M. Giroux,
X. Pan, C. Wang, J. Greeley, Science 2019, 363, 870-874.
[27] a) Y. Li, X. Xu, C. Wang, C. Wang, F. Xie, J. Yang, Y. Gao, J. Phys.
Chem. C 2015, 119, 23996-24002; b) Y. Wang, W.-Y. Rho, H.-Y. Yang,
T. Mahmoudi, S. Seo, D.-H. Lee, Y.-B. Hahn, Nano Energy 2016, 27,
535-544.
[28] Y. Hayashi, Y. Sasaki, Chem. Commun. 2005, 2716-2718.
[29] M. S. Holm, S. Saravanamurugan, E. Taarning, Science 2010, 328,
602-605.
[30] a) M. Bellardita, A. Di Paola, B. Megna, L. Palmisano, Appl. Catal. B-
Environ. 2017, 201, 150-158; b) M. Toyoda, Y. Nanbu, Y. Nakazawa,
M. Hirano, M. Inagaki, Appl. Catal. B-Environ. 2004, 49, 227-232.
[31] B. Murali, S. Dey, A. L. Abdelhady, W. Peng, E. Alarousu, A. R.
Kirmani, N. Cho, S. P. Sarmah, M. R. Parida, M. I. Saidaminov, A. A.
Zhumekenov, J. Sun, M. S. Alias, E. Yengel, B. S. Ooi, A. Amassian, O.
M. Bakr, O. F. Mohammed, ACS Energy Lett. 2016, 1, 1119-1126.
[32] A. Mehta, J. Im, B. H. Kim, H. Min, R. Nie, S. I. Seok, ACS nano 2018,
12, 12129-12139.
7
This article is protected by copyright. All rights reserved.

10.1002/anie.202002939
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
MAI-terminated MAPbI3 could convert DHA into butyl lactate with the production rate of 7719 μg/g cat./h under visible light illumination.
Specially, MAI-termination induces p-doping effect near the surface, leading to the oxidation of DHA into pyruvaldehyde. Moreover,
MAI-termination is susceptible to iodide oxidation, resulting in the Pb(II) sites exposed, which further promote the reaction of
pyruvaldehyde and butanol to produce butyl lactate.
Institute and/or researcher Twitter usernames: ((optional))
8
This article is protected by copyright. All rights reserved.