The mixed lead halide perovskites are promising candidates for solar cells, but the intrinsic instabilities due to phase segregation, halide migration, and hydration hinder the practical applications. Currently, less attention is paid to the humidity induced degradation. This important issue of mixed halide perovskites is now investigated by Fourier transform infrared (FTIR) spectroscopy. After hydration of CH3NH3Pb(I1−xBrx)3, a broad band at around 3200 cm−1 of –OH vibration appears in the FTIR spectra and the bands in the range of 1200–1700 cm−1 become much stronger. The dominant FTIR mode at 1660 cm−1 due to the intermediate complex is sensitive on hydration, implying it could be used as a characteristic peak to evaluate the hydration degree of mixed halide perovskites. These information are of interest to optimize the crystal structure of this material to improve the stability so that to promote the photovoltaic applications.

The metal halide perovskites with a chemical structure of ABX3 [A is typically methylammonium (MA = CH3NH3+), formamidinium (FA = HC(NH2)2+), or Cs+, B is Pb2+ or Sn2+, and X is I, Br, or Cl] have recently emerged as highly promising photovoltaic materials. The certified power conversion efficiencies (η) of single junction perovskite solar cells have increased rapidly during the past years and now exceed 22.7%.1–3 Especially, the mixed organic–inorganic hybrid lead halide perovskites, of which bandgaps can be tuned by halide compositions, have attracted enormous attention for flexible optoelectronic applications, such as light emitting devices, photodetectors, and lasers.4–11 The composition-dependent electronic structures allow for precise engineering of their optical response between ∼1.5 and 2.4 eV, which also enable the harvest of full solar spectrum.12–14 Thus, the mixed halide systems are ideal for use in forming multijunction solar cells with silicon or copper indium gallium (di) selenide.15,16 A tandem cell made with a state-of-the-art silicon component and perovskites could surpass 30% efficiency by further engineering.16,17

However, the main drawback for the mixed perovskite photovoltaics is their long-term stability.18–35 Different to pure phase perovskites (e.g., MAPbI3), one particular concern on the mixed halide perovskites is the phase segregation.36–40 Large halide ionic displacements have been noted at room temperature in perovskite films.36–40 The migration of halide ions within the perovskite crystallites, resulting in a phase segregation, directly changes the optical and electronic properties of mixed halide perovskites. The ionic migration will have significant implications in the performance of photovoltaic devices.36–40 It has been proposed that under light soaking halide ions migrate within the perovskite layer forming I-rich minority and Br-rich majority phases.28 

In addition to the phase segregation, it is still lack of sufficient attention on the humidity induced instability of the mixed halide perovskites.33–35 Regarding the interactions with water, various degradation pathways were proposed counting on the specific conditions applied, involving the formation of PbI2, I2, H2, methylamine, and hydrated forms of MAPbI3.23,24,33–35 For example, by carefully controlling the relative humidity (RH) of an environmental chamber, in situ absorption spectroscopy and in situ grazing incidence X-ray diffraction were used to monitor phase changes in perovskite degradation process.33 Yang et al. demonstrated the formation of a hydrated intermediate containing isolated PbI64 octahedra as the first step of the degradation of MAPbI3.33 To date, research mainly focuses on pure phase perovskites.33–35 

Nevertheless, the moisture induced degradation of mixed perovskites containing two halide elements still deserves further study. First, mixed halide perovskites are more important than pure phase perovskites since the bandgaps of mixed lead halide perovskites can be tuned by halide compositions, which are more suitable for flexible optoelectronic applications. It is necessary and significant to conduct a direct study on mixed perovskites. Second, mixed halide perovskites are not simple mixtures of two pure phases. The separate contributions of two different halides are still unknown. Moreover, due to the coexistence of two different halogens, there is a possible phase segregation. Despite intensive research on light induced phase segregation, moisture induced phase segregation is still unclear. Third, there is still a lack of simple and fast technique to characterize the hydration degree. In this work, MAPb(I1−xBrx)3 is used as a model system to investigate the humidity-stability of the mixed halide perovskites. The dominant Fourier-transform infrared (FTIR) mode at 1660 cm−1 due to the intermediate complex is sensitive on hydration, implying it could be used as a characteristic peak to evaluate the hydration degree of mixed halide perovskites.

A 2 ml solution consisting of MABr (Dyesol, 0.15M, 34 mg), MAI (Dyesol, 0.15M, 48 mg), PbBr2 (Alfa Aesar, 0.15M, 110 mg), and PbI2 (Alfa Aesar, 0.15M, 138 mg) in N,N-dimethylformamide (DMF, Sigma-Aldrich) was used as a precursor solution. A 0.2 µm polytetrafluoroethylene filter was used to remove any particulates. A 400 µl portion of the precursor solution was drop-casted on a cover slide. Then 100 °C heating treatment was conducted to remove the solvent. In order to observe the evident structural degradation, no protective layer was added.

The microscopic images were captured by using an inverted metallurgical microscope equipped with a CCD camera (Nikon Eclipse MA100). Photoluminescence (PL) measurements were conducted on a PerkinElmer LS55 fluorescence spectrometer. A 400 nm monochromatic light was used as excitation. The UV-visible diffuse reflectance spectra of the solid film were recorded by an integrating sphere equipped on a Perkin Elmer Lambda1050 spectrometer. The absorption spectra were then converted from the diffuse reflectance spectra. FTIR measurements were performed on Bruker VERTEX 70 with a Hyperion scanning microscope.

Figure 1 shows the color evolution of the mixed halide perovskite film under ambient condition with a relative humidity (RH) of 54%. The initial precursor solution is light yellow, mainly reflecting the color of PbI2 reactants. After heating treatment, the as-grown mixed halide perovskite film becomes black (stage b). It fleetly turns to dark red (stage c) and then becomes red subsequently (stage e) after cooling. Stage e can be recovered to the initial state immediately by re-heating at 100 °C. Similar color evolution can be observed again (from stage i to l), indicating this process is reversible. If the perovskite film is stored in the ambient condition for 4 h or longer, the color turns to orange and some black spots appear. After storing for 12 h in ambient, the color recovers to dark red with large black spots. The color evolution of the separated MAPbI3 and MAPbBr3 films is also recorded for comparison (Fig. S1 of the supplementary material). The as-grown MAPbI3 film is black but rapidly goes to light yellow. Finally, the film becomes a mixture of black and yellow spots. However, the MAPbBr3 film constantly remains red. Thus, it is reasonable to propose that the observed color variation of the mixed perovskite film is predominately caused by the transformation of iodine related composition.

FIG. 1.

Color evolution of the mixed halide perovskite film under ambient condition with RH of 54%.

FIG. 1.

Color evolution of the mixed halide perovskite film under ambient condition with RH of 54%.

Close modal

Optical microscopic images of the mixed halide perovskite film at stage b [Fig. 2(a)] and stage e [Fig. 2(b)] are shown in Fig. 2. There are many crystalline particles in the as-grown perovskite film. After the color changes, many dendritic, fibril-like structures appear. Some of the dendritic crystals are sufficiently large to be viewed with the naked eyes. Similar morphological transformation was observed during the decomposition of MAPbI3.33 It is reported that the occurrence of dendritic crystals is accompanied by the formation of PbI2 crystals.33 

FIG. 2.

Optical microscopic images of the mixed halide perovskite film at stage b (a) and stage e (b).

FIG. 2.

Optical microscopic images of the mixed halide perovskite film at stage b (a) and stage e (b).

Close modal

Optical characterizations including PL and absorption spectra are recorded to study the structural transformation. The comparison of PL spectra from stage b and stage e displays three main differences. First, the peak at 525 nm, dropped remarkably (i). Second, the chief peak blueshifts from 590 to 570 nm (ii). Third, a new peak appears at 720 nm (iii). The PL peaks are assigned based on the PL spectra of the separated MAPbI3 and MAPbBr3 films (Fig. S2 of the supplementary material). Figure S2 suggests that the PL peak of the MAPbBr3 film is located at 550 nm. The peak at 525 nm should be fluorescence of PbBr2 rather than the emission of MAPbBr3. In the as-prepared mixed halide film, the PbBr2 precipitates due to the limited solid solubility, which frequently happens when x > 0.2 for MAPb(I1−xBrx)3.15 The PL spectrum of MAPbI3 exhibits a strong peak at 740 nm with a shoulder band at 720 nm, which are caused by the emission of MAPbI3 and PbI2, respectively. Therefore, the occurrence of 720 nm-PL band indicates the formation of PbI2 at stage e, which is consistent with the morphological transformation observed by using the microscope. The appearance of PbI2 implies that the iodine element is isolated from the mixed perovskite; consequently, the rest of the main part becomes Br-rich [MAPb(I1−xBrx)3, x > 0.5] phase, explaining the blueshift of the chief band according to Vegard’s law.41 Similar PL variation has been observed during the phase segregation. Sadhanala and co-workers reported that the PL peak position of MAPbBr1.2I1.8 films redshifts from 1.68 to 1.94 eV after 2 weeks under inert and dark conditions,7 which is explained by the gradual and spontaneous room-temperature phase segregation of an as-prepared halide composition within the miscibility gap.28 The deduction of iodine content enables more fraction of Br in the majority composite so that the 520 nm-PL peak of isolated PbBr2 disappears at stage e.

The transformation of electronic state is further confirmed by the absorption spectra (Fig. 3). The absorption edge is located at 740 nm for stage b, while it blueshifts to 600 nm for stage e, in good agreement with the color variation from black to red. From stage e to f, the absorption edge slightly redshifts. And after long-term storage, the absorption edge largely recovers to 710 nm, suggesting the reversible structural transformation, in agreement with the color recovery. For regular semiconductors, it is unusual that the photon energy of the absorption band edge is lower than the PL peak. However, here the PL peak is ascribed to bandgap emission, while the absorption edge is affected by bandtails. Due to the large Urbach tail in the absorption spectra, the PL peak was frequently observed at energy above the absorption edge in different perovskites.41,42 Note that the minority PbI2 crystals do not contribute substantially to the absorption spectra.35,37 Thus, the increment of bandgap from stage b to e is also consistent with the PL peak blueshift (Fig. 3).

FIG. 3.

PL (a) and absorption spectra (b) of the mixed halide perovskite film acquired at different stages.

FIG. 3.

PL (a) and absorption spectra (b) of the mixed halide perovskite film acquired at different stages.

Close modal

PL and absorption spectra explain the color evolution successfully and reveal the transformation of electronic states; however, the direct insight into the structure is insufficient, which is a common drawback for most of the current studies.33 Fourier-transform infrared spectroscopy (FTIR) is a fast, noncontact, and nondamaging optical tool to identify molecules and study chemical bonding. Here, the conformational change in mixed perovskites during the color evolution is comprehensively investigated by the FTIR for the first time. A continuous FTIR test of the mixed halide perovskite film that began at stage b is shown in Fig. 4(a). Each test lasts for 30 s. In order to display dividedly, the FTIR spectra are vertically translated according to the test order. All the FTIR spectra in Fig. 4(a) show almost the identical bands with N–H vibrational and stretching modes at 3225 and 1586 cm−1, C–H vibrational and stretching modes at 2935 and 1485 cm−1, and C–N stretching modes at 1258 cm−1.43–46 Weak bands ascribing to the in-plane bends of C–OH at 1350 cm−1 and an unknown band at 1660 cm−1 are also observable. A similar FTIR band at 1631 cm−1 is ascribed to PbI2–MAI–DMF complex by Guo et al.46 In fact, the existence of intermediate phases of PbI2–MAI–solvent in perovskites by solution methods is commonly accepted.46,47

FIG. 4.

[(a) and (b)] Continuous FTIR test of the mixed halide perovskite film that began at stage b (a) and stage e (b). (c) The 1660 cm−1 FTIR band during the continuous test. (d) The 1660 cm−1 band intensity versus test order.

FIG. 4.

[(a) and (b)] Continuous FTIR test of the mixed halide perovskite film that began at stage b (a) and stage e (b). (c) The 1660 cm−1 FTIR band during the continuous test. (d) The 1660 cm−1 band intensity versus test order.

Close modal

However, after the color turns red, the FTIR spectra contain more bands with higher intensities [Fig. 4(b)]. Especially, the bands at 1660 cm−1 become very prominent in all FTIR spectra obtained from stage e and later. In addition, the band at around 3200 cm−1 becomes broad and remarkable, as highlighted by the shadow area. In addition to the N–H at 3225 cm−1 and C–H at 2935 cm−1, the emerged −ν(OH) modes of C–OH (3400 cm−1) and intercalated H2O (3225 cm−1) after color evolution lead the band to be broad and striking. These two characters evidently affirm that the observed color evolution is caused by the hydration of the mixed halide perovskites according to the following reaction:

in which x < y. The iodine content decreases in the majority explaining the blueshift of the chief PL band and the absorption band edge, as well as the color evolution. Moreover, the less iodine enables more bromide in the mixed halide film, which relieves the precipitation of PbBr2. Thus, the PL band at 520 nm disappears at stage e. The appearance of PbI2 alongside the hydration is already confirmed by the microscopic images and PL peak at 720 nm. Most importantly, the 1660 cm−1 FTIR band of the PbI2–MAI–DMF complex sensitively changes, implying that this band can be developed as an indicator for the hydration of halide perovskites.

All the FTIR bands at 1660 cm−1 in Figs. 4(a) and 4(b) are collected in Fig. 4(c) by setting the background to zero. The band intensity versus the test order is plotted in Fig. 4(d). As the measurement continues, the 1660 cm−1 FTIR band gradually increases and attains a very high value at stage e. After that the FTIR band intensity partially drops due to the reversible reaction. During the cooling (from stage b to e), the mixed perovskite film adsorbs excessive water and then is hydrated overly. After stage e, the dehydration reaction dominates until the reversible reaction reaches balance (stage h). The dehydration reaction can be accelerated by heating treatment so that the color recovers immediately at stage i. Figure S3 (supplementary material) shows that the 1660 cm−1 FTIR band almost vanishes after heating. In order to further confirm, the 1660 cm−1 FTIR band is caused by hydration rather than phase separation or thorough decomposition. FTIR spectra of the separated MAPbI3 and MAPbBr3 films, as well as MAI, PbI2, MABr, and PbBr2 are also obtained and shown in Figs. S4 and S5 of the supplementary material. The sharp 1660 cm−1 FTIR band cannot be found in any reactant and MAPbBr3 film. But a similar band at 1660 cm−1 can be found in the MAPbI3 film after storing for 5 min according to the hydration,

which was verified previously.33 Moreover, the as-grown mixed halide perovskite films are stored in dark and dry conditions for comparison. The color evolutions are displayed in Fig. S5 of the supplementary material. In the dark, the color red keeps covering a large area in 5 min, similar to the ambient condition, indicating that light irradiation is not the main external stimulations for the observed color variation. However, when the film was stored in a desiccator, the color changed very slowly, further confirming that the humidity plays a very important role for the observed effect.

In order to provide a robust explanation for the experimental observation, we calculated the infrared (IR) spectrum of (MA)4PbI6 · 2H2O from first principles as implemented in the Vienna ab initio simulation package (VASP).48,49 The exchange and correlation function are approximated by generalized gradient approximation (GGA) with Perdew-Burke-Ernzerhof (PBE) functional.51 A kinetic energy cutoff of 500 eV and a 4 × 4 × 4 Γ-centered k-point mesh are used in the structural relaxation, where the Hellman-Feynman forces are minimized to 0.01 eV/Å. The IR spectrum is calculated from the vibrational frequencies which are performed only at the Γ point within the density-functional perturbation theory (DFPT).50–52 As shown in the inset of Fig. 5(a), the monoclinic crystal with a = 10.39 Å, b = 11.31 Å, c = 10.55 Å, and β = 91.2° is chosen for the (CH3NH3)4PbI6 · 2H2O system.53 The calculated IR spectrum plotted in Fig. 5 exhibits mainly two groups of bands, one is in the range of 3000–3500 cm−1 due to vibrations of –OH (highlighted by the shadow) and the other is in the range of 1200–1800 cm−1. The calculated IR spectrum matches well with the two characters observed in the experimental results. An enlarged IR spectrum in the range of 1000–2000 cm−1 is shown in Fig. 5(b). Two dominant vibrational frequencies at 1503 and 1598 cm−1 can be found. Density functional theory (DFT) calculations reveal that the peak around 1503 cm−1 is mainly caused by the bending mode of the N–H bond and the stretching mode of O–H and C–H bonds, while the IR peak around 1598 cm−1 originates from the bending mode of the N–H and O–H bonds. The details of the vibrations at frequencies of 1503 and 1598 cm−1 are presented in videos in the supplementary material. Interestingly, there is no band at 1660 cm−1 in the calculated spectrum, further confirming that it is caused by the PbI2–MAI–DMF complex.

FIG. 5.

(a) Calculated IR spectrum of (MA)4PbI6 · 2H2O broadened by the Gaussian resolution function with a width of 30 cm−1. Inset: the optimized monoclinic crystal structure of (MA)4PbI6 · 2H2O. (b) Calculated IR spectrum in the region between 1000 and 2000 cm−1.

FIG. 5.

(a) Calculated IR spectrum of (MA)4PbI6 · 2H2O broadened by the Gaussian resolution function with a width of 30 cm−1. Inset: the optimized monoclinic crystal structure of (MA)4PbI6 · 2H2O. (b) Calculated IR spectrum in the region between 1000 and 2000 cm−1.

Close modal

In summary, mixed halide perovskites were fabricated by a common solution method. The DMF solution of stoichiometric Pb2+ and halide (I and Br) is drop-casted on a substrate followed by annealing. In the ambient condition, the as-grown black film becomes red rapidly. The appearance of PbI2 and enlarged bandgap of the majority during the color evolution are verified by PL and absorption spectra. And in the FTIR spectra, a broad band at around 3200 cm−1 and a sharp band at 1660 cm−1 of the PbI2–MAI–DMF complex appear for the red film, revealing the hydration of mixed halide perovskites. And the hydration reaction is partially reversible. These conclusions are important messages in view of further developments of the long-term stable photovoltaics based on the mixed perovskites. Even though in practical applications the solar cells are encapsulated, the current techniques cannot totally avoid hydration. Therefore, the comprehensive understanding on hydration induced degradation will be beneficial for the promotion of perovskites with better intrinsic stability from the crystal structure, eventually to innovate the perovskite based photovoltaic technology. In this work, we find that hydration is accompanied by ionic migration and phase segregation. In the past years, great effort has already been made to suppress the ionic migration.25–35,54 For example, potassium passivation could significantly eliminate ionic migration,54 which may simultaneously improve the stability resisting to moisture.

See supplementary material for color evolution, PL spectra, and FTIR spectra of separated MAPbI3 and MAPbBr3 films (PDF); FTIR spectra of the mixed halide perovskite film at different stages (PDF); color evolution of the mixed halide perovskite films stored in dark and dry conditions (PDF); and vibrations of (MA)4PbI6·2H2O at different frequencies (Video).

This work was jointly supported by the National Natural Science Foundation of China (No. 11604155), the China Postdoctoral Science Foundation (Grant Nos. 2016M600428 and 2017T100386), and Jiangsu Planned Projects for Postdoctoral Research Funds (No. 1601023A).

1.
A.
Kojima
,
K.
Teshima
,
Y.
Shirai
, and
T.
Miyasaka
,
J. Am. Chem. Soc.
131
,
6050
(
2009
).
2.
See https://www.nrel.gov/pv/assets/images/efficiency-chart.png for National Renewable Energy Labs Efficiency Chart (2017); accessed
26 June 2017
.
3.
M.
Saliba
,
T.
Matsui
,
J.-Y.
Seo
,
K.
Domanski
,
J.-P.
Correa-Baena
,
M. K.
Nazeeruddin
,
S. M.
Zakeeruddin
,
W.
Tress
,
A.
Abate
,
A.
Hagfeldt
, and
M.
Grätzel
,
Energy Environ. Sci.
9
,
1989
(
2016
).
4.
Y.
Cao
,
N.
Wang
,
H.
Tian
,
J.
Guo
,
Y.
Wei
,
H.
Chen
,
Y.
Miao
,
W.
Zou
,
K.
Pan
,
Y.
He
,
H.
Cao
,
Y.
Ke
,
M.
Xu
,
Y.
Wang
,
M.
Yang
,
K.
Du
,
Z.
Fu
,
D.
Kong
,
D.
Dai
,
Y.
Jin
,
G.
Li
,
H.
Li
,
Q.
Peng
,
J.
Wang
, and
W.
Huang
,
Nature
562
,
249
(
2018
).
5.
K.
Lin
,
J.
Xing
,
L. N.
Quan
,
F. P. G.
de Arquer
,
X.
Gong
,
J.
Lu
,
L.
Xie
,
W.
Zhao
,
D.
Zhang
,
C.
Yan
,
W.
Li
,
X.
Liu
,
Y.
Lu
,
J.
Kirman
,
E. H.
Sargent
,
Q.
Xiong
, and
Z.
Wei
,
Nature
562
,
245
(
2018
).
6.
N. K.
Kumawat
,
A.
Dey
,
A.
Kumar
,
S. P.
Gopinathan
,
K. L.
Narasimhan
, and
D.
Kabra
,
ACS Appl. Mater. Interfaces
7
,
13119
(
2015
).
7.
A.
Sadhanala
,
F.
Deschler
,
T. H.
Thomas
,
S. E.
Dutton
,
K. C.
Goedel
,
F. C.
Hanusch
,
M. L.
Lai
,
U.
Steiner
,
T.
Bein
,
P.
Docampo
,
D.
Cahen
, and
R. H.
Friend
,
J. Phys. Chem. Lett.
5
,
2501
(
2014
).
8.
J.
Song
,
J.
Li
,
X.
Li
,
L.
Xu
,
Y.
Dong
, and
H.
Zeng
,
Adv. Mater.
27
,
7162
(
2015
).
9.
J.
Lu
,
X.
Sheng
,
G.
Tong
,
Z.
Yu
,
X.
Sun
,
L.
Yu
,
X.
Xu
,
J.
Wang
,
J.
Xu
,
Y.
Shi
, and
K.
Chen
,
Adv. Mater.
29
,
1700400
(
2017
).
10.
B.
Tang
,
H.
Dong
,
L.
Sun
,
W.
Zheng
,
Q.
Wang
,
F.
Sun
,
X.
Jiang
,
A.
Pan
, and
L.
Zhang
,
ACS Nano
11
,
10681
(
2017
).
11.
Y.
Wang
,
X.
Li
,
J.
Song
,
L.
Xiao
,
H.
Zeng
, and
H.
Sun
,
Adv. Mater.
27
,
7101
(
2015
).
12.
D. P.
McMeekin
,
G.
Sadoughi
,
W.
Rehman
,
G. E.
Eperon
,
M.
Saliba
,
M. T.
Horantner
,
A.
Haghighirad
,
N.
Sakai
,
L.
Korte
,
B.
Rech
,
M. B.
Johnston
,
L. M.
Herz
, and
H. J.
Snaith
,
Science
351
,
151
(
2016
).
13.
R. E.
Beal
,
D. J.
Slotcavage
,
T.
Leijtens
,
A. R.
Bowring
,
R. A.
Belisle
,
W. H.
Nguyen
,
G.
Burkhard
,
E. T.
Hoke
, and
M. D.
McGehee
,
J. Phys. Chem. Lett.
7
,
746
(
2016
).
14.
B. A.
Rosales
,
M. P.
Hanrahan
,
B. W.
Boote
,
A. J.
Rossini
,
E. A.
Smith
, and
J.
Vela
,
ACS Energy Lett.
2
,
906
(
2017
).
15.
Y.
Zhou
,
F.
Wang
,
Y.
Cao
,
J.-P.
Wang
,
H.-H.
Fang
,
M. A.
Loi
,
N.
Zhao
, and
C.-P.
Wong
,
Adv. Energy Mater.
7
,
1701048
(
2017
).
16.
V.
Sivaram
,
S. D.
Stranks
, and
H. J.
Snaith
,
Prog. Photovoltaics: Res. Appl.
23
,
805
(
2015
).
17.
H.
Yuan
,
E.
Debroye
,
K.
Janssen
,
H.
Naiki
,
C.
Steuwe
,
G.
Lu
,
M.
Moris
,
E.
Orgiu
,
H.
Uji-i
,
F.
De Schryver
,
P.
Samr
,
J.
Hoens
, and
M. J.
Roeffaers
,
Phys. Chem. Lett.
7
,
561
(
2016
).
18.
P.
Docampo
and
T. A.
Bein
,
Acc. Chem. Res.
49
,
339
(
2016
).
19.
J. S.
Manser
,
M. I.
Saidaminov
,
J. A.
Christians
,
O. M.
Bakr
, and
P. V.
Kamat
,
Acc. Chem. Res.
49
,
330
(
2016
).
20.
R.
Gottesman
,
A.
Zaban
,
R.
Gottesman
, and
A.
Zaban
,
Acc. Chem. Res.
49
,
320
(
2016
).
21.
F.
Matsumoto
,
S. M.
Vorpahl
,
J. Q.
Banks
,
E.
Sengupta
, and
D. S.
Ginger
,
J. Phys. Chem. C
119
,
20810
(
2015
).
22.
J. M.
Frost
,
K. T.
Butler
,
F.
Brivio
,
C. H.
Hendon
,
M.
van Schilfgaarde
, and
A.
Walsh
,
Nano Lett.
14
,
2584
(
2014
).
23.
G.
Niu
,
X. D.
Guo
, and
L.
Wang
,
J. Mater. Chem. A
3
,
8970
(
2015
).
24.
Y.
Wang
,
X.
Li
,
S.
Sreejith
,
F.
Cao
,
Z.
Wang
,
M. C.
Stuparu
,
H.
Zeng
, and
H.
Sun
,
Adv. Mater.
28
,
10637
(
2016
).
25.
D. W.
de Quilettes
,
W.
Zhang
,
V. M.
Burlakov
,
D. J.
Graham
,
T.
Leijtens
,
A.
Osherov
,
V.
Bulovic
,
H. J.
Snaith
,
D. S.
Ginger
, and
S. D.
Stranks
,
Nat. Commun.
7
,
11683
(
2016
).
26.
X.
Deng
,
X.
Wen
,
J.
Zheng
,
T.
Young
,
C. F. J.
Lau
,
J.
Kim
,
M.
Green
,
S.
Huang
, and
A.
Ho-Bailliea
,
Nano Energy
46
,
356
(
2018
).
27.
E. T.
Hoke
,
D. J.
Slotcavage
,
E. R.
Dohner
,
A. R.
Bowring
,
H. I.
Karunadasa
, and
M. D.
McGehee
,
Chem. Sci.
6
,
613
(
2015
).
28.
F.
Brivio
,
C.
Caetano
, and
A.
Walsh
,
J. Phys. Chem. Lett.
7
,
1083
(
2016
).
29.
R.
Nishikubo
,
N.
Tohnai
,
I.
Hisaki
, and
A.
Saeki
,
Adv. Mater.
29
,
1700047
(
2017
).
30.
B.
Conings
,
J.
Drijkoningen
,
N.
Gauquelin
,
A.
Babayigit
,
J.
D’Haen
,
L.
D’Olieslaeger
,
A.
Ethirajan
,
J.
Verbeeck
,
J.
Manca
,
E.
Mosconi
,
F.
De Angelis
, and
H. G.
Boyen
,
Adv. Energy Mater.
5
,
1500477
(
2015
).
31.
A.
Latini
,
G.
Gigli
, and
A.
Ciccioli
,
Sustainable Energy Fuels
1
,
1351
(
2017
).
32.
R.
Yang
,
L.
Zhang
,
Y.
Cao
,
Y.
Miao
,
Y.
Ke
,
Y.
Wei
,
Q.
Guo
,
Y.
Wang
,
Z.
Rong
,
N.
Wang
,
R.
Li
,
J.
Wang
,
W.
Huang
, and
F.
Gao
,
Appl. Phys. Lett.
111
,
073302
(
2017
).
33.
J.
Yang
,
B. D.
Siempelkamp
,
D.
Liu
, and
T. L.
Kelly
,
ACS Nano
9
,
1955
(
2015
).
34.
A. M. A.
Leguy
,
Y.
Hu
,
M.
Campoy-Quiles
,
M. I.
Alonso
,
O. J.
Weber
,
P.
Azarhoosh
,
M.
van Schilfgaarde
,
M. T.
Weller
,
T.
Bein
, and
J.
Nelson
,
Chem. Mater.
27
,
3397
(
2015
).
35.
Z.
Xiao
,
L.
Zhao
,
N. L.
Tran
,
Y. L.
Lin
,
S. H.
Silver
,
R. A.
Kerner
,
N.
Yao
,
A.
Kahn
,
G. D.
Scholes
, and
B. P.
Rand
,
Nano Lett.
17
,
6863
(
2017
).
36.
D. A.
Egger
,
A. M.
Rappe
, and
L.
Kronik
,
Acc. Chem. Res.
49
,
573
(
2016
).
37.
S. J.
Yoon
,
S.
Draguta
,
J. S.
Manser
,
O.
Sharia
,
W. F.
Schneider
,
M.
Kuno
, and
P. V.
Kamat
,
ACS Energy Lett.
1
,
290
(
2016
).
38.
Y.
Yuan
and
J.
Huang
,
Acc. Chem. Res.
49
,
286
(
2016
).
39.
O.
Hentz
,
Z.
Zhao
, and
S.
Gradecak
,
Nano Lett.
16
,
1485
(
2016
).
40.
Y.
Fang
,
Q.
Dong
,
Y.
Shao
,
Y.
Yuan
, and
J.
Huang
,
Nat. Photonics
9
,
679
(
2015
).
41.
L. M.
Pazos-Outón
,
M.
Szumilo
,
R.
Lamboll
,
J. M.
Richter
,
M.
Crespo-Quesada
,
M.
Abdi-Jalebi
,
H. J.
Beeson
,
M.
Vrućinić
,
M.
Alsari
,
H. J.
Snaith
,
B.
Ehrler
,
R. H.
Friend
, and
F.
Deschler
,
Science
351
,
1430
(
2015
).
42.
X. J.
She
,
J. J.
Wu
,
J.
Zhong
,
H.
Xu
,
Y. C.
Yang
,
R.
Vajtai
,
J.
Lou
,
Y.
Liu
,
D. L.
Du
,
H. M.
Li
, and
P. M.
Ajayan
,
Nano Energy
27
,
138
(
2016
).
43.
C. D.
Zangmeister
,
Chem. Mater.
22
,
5625
(
2010
).
44.
T.
Yamanaka
,
K.
Masumori
,
R.
Ishikawa
,
K.
Ueno
, and
H.
Shirai
,
J. Phys. Chem. C
120
,
25371
(
2016
).
45.
A. E.
Williams
,
P. J.
Holliman
,
M. J.
Carnie
,
M. L.
Davies
,
D. A.
Worsley
, and
T. M.
Watson
,
J. Mater. Chem. A
2
,
19338
(
2014
).
46.
X.
Guo
,
C.
McCleese
,
C.
Kolodzie
,
A. C. S.
Samia
,
Y.
Zhao
, and
C.
Burda
,
Dalton Trans.
45
,
3806
(
2016
).
47.
Y.
Li
,
J.
Wang
,
Y.
Yuan
,
X.
Dong
, and
P.
Wang
,
Sustainable Energy Fuels
1
,
1041
(
2017
).
48.
G.
Kresse
and
J.
Hafner
,
Phys. Rev. B
47
,
558
(
1993
).
49.
G.
Kresse
and
J.
Furthmller
,
Comput. Mater. Sci.
6
,
15
(
1996
).
50.
J. P.
Perdew
,
K.
Burke
, and
M.
Ernzerhof
,
Phys. Rev. Lett.
77
,
3865
(
1996
).
51.
S.
Baroni
,
S.
de Gironcoli
,
A.
Dal Corso
, and
P.
Giannozz
,
Rev. Mod. Phys.
73
,
515
(
2001
).
52.
D.
Karhánek
,
T.
Bučko
, and
J.
Hafner
,
J. Phys.: Condens. Matter
22
,
265006
(
2010
).
53.
A.
Wakamiya
,
M.
Endo
,
T.
Sasamori
, and
N.
Tokitoh
,
Chem. Lett.
43
,
711
(
2014
).
54.
M.
Abdi-Jalebi
,
Z.
Andaji-Garmaroudi
,
S.
Cacovich
,
C.
Stavrakas
,
B.
Philippe
,
J. M.
Richter
,
M.
Alsari
,
E. P.
Booker
,
E. M.
Hutter
,
A. J.
Pearson
,
S.
Lilliu
,
T. J.
Savenije
,
H.
Rensmo
,
G.
Divitini
,
C.
Ducati
,
R. H.
Friend
, and
S. D.
Stranks
,
Nature
555
,
497
(
2018
).

Supplementary Material