Hybrid Perovskites with Larger Organic Cations Reveal Autocatalytic Degradation Kinetics and Increased Stability under Light

Christie L. C. Ellis, Hamza Javaid, Emily C. Smith, and D. Venkataraman*

Article

Hybrid Perovskites with Larger Organic Cations Reveal Autocatalytic

Degradation Kinetics and Increased Stability under Light

†

Cite This: https://dx.doi.org/10.1021/acs.inorgchem.0c01133

Read Online

ACCESS

Metrics & More

Article Recommendations

sı Supporting Information

ABSTRACT: Hybrid organic−inorganic perovskites have shown

incredible promise as active materials for photovoltaic devices, but

their instability to light remains a signiﬁcant roadblock in realizing

these applications. Changing the organic cation has been shown to

aﬀect light-induced degradation. As a strategy for increasing the

stability of these materials, we replaced varying percentages of

methylammonium ion in the archetypical methylammonium lead

iodide (MAPbI ) hybrid organic−inorganic perovskite with three

signiﬁcantly larger organic ammonium cations: imidazolium,

dimethylammonium, and guanidinium. We were able to synthesize

hybrid organic−inorganic perovskites with the same 3D perovskite

structure as MAPbI with substitution of the larger ions as high as

0−30%. These substituted hybrid organic−inorganic perovskites

retained similar optoelectronic properties. We discovered that the light-induced degradation in MAPbI and its substituted

derivatives is autocatalytic, and we calculated rate coeﬃcients for the degradation. All of the substituted hybrid organic−inorganic

perovskites showed light-induced degradation slower than that of MAPbI , up to a 62% decrease in degradation rate coeﬃcient, at all

substitution percentages up to 20%. This work provides evidence that a high percentage of a variety of large ammonium cations can

be substituted into the hybrid organic−inorganic perovskite lattice without compromising its desirable optoelectronic properties.

Insight into the autocatalytic mechanism of light-induced degradation will be valuable for designing additional strategies to improve

the stability of hybrid organic−inorganic perovskites. We also oﬀer insights into how factors other than size, such as hydrogen

bonding, inﬂuence the stability of the materials. Overall, we have shown that substitution of methylammonium ion for the much

larger imidazolium, dimethylammonium, and guanidinium cations in MAPbI is a valid strategy for creating stable hybrid organic−

inorganic perovskite derivatives by slowing the rate of light-induced degradation.

INTRODUCTION

deﬁnition required for photovoltaic devices. Understanding

how hybrid organic−inorganic perovskites degrade under light

■

Hybrid organic−inorganic perovskites are a promising class of

2,13

has not been straightforward.

Many experimental

materials for a plethora of electronic and optoelectronic

−4

techniques used to probe their structure and properties cannot

be performed under light. Even when a technique allows for

light exposure, it has become apparent that multiple structural

and optoelectronic changes take place at a variety of time

scales under light, causing samples to change over the course of

applications.

Photovoltaics is one of the most prominent

applications, as the power conversion eﬃciencies of hybrid

organic−inorganic perovskite photovoltaics have skyrocketed

to match or surpass those of current commercial technologies

in just a few years. The current record eﬃciency for hybrid

4,15

organic−inorganic perovskites photovoltaics is 25.2%, and

eﬃciencies of 29.1% have been reached using tandem hybrid

organic−inorganic perovskites−silicon tandem photovol-

the measurement.

As a result, there are major incon-

sistencies in the long-term stability data.

At present, there are many proposed hypotheses on how

light causes hybrid organic−inorganic perovskite degradation,

resulting in widely varied and often contradictory conclusions.

taics.

Hybrid organic−inorganic perovskite photovoltaic

devices can be made thin, lightweight, ﬂexible, and semi-

transparent using low-temperature, solution-based, scalable

fabrication methods, all while maintaining their high

eﬃciencies.

Received: April 17, 2020

One of the signiﬁcant roadblocks to commercializing hybrid

organic−inorganic perovskites photovoltaics is their instability

to operational conditions, such as heat, humidity, applied

−11

electrical bias, and light.

Stability under light is a crucial

problem that we need to solve, as light exposure is by

XXXX American Chemical Society

Inorg. Chem. XXXX, XXX, XXX−XXX

Figure 1. Left: Ideal 3D cubic perovskite structure with ABX stoichiometry, where A is an organic cation (blue−black ion in center of connected

octahedra), B is a divalent metal cation (green ion in center of octahedra), and X is a halide anion (purple spheres on corners of octahedra). The

archetypical hybrid organic−inorganic perovskite photovoltaic material methylammonium lead iodide (MAPbI ) was used as a starting material for

this work. MAPbI has a slightly distorted tetragonal structure similar to the cubic structure shown here. Right: Organic ammonium cations used in

this work. The larger imidazolium (IA), dimethylammonium (DMA), and guanidinium (GA) ions were used to replace the MA ion as a strategy for

increasing stability of these hybrid organic−inorganic perovskite materials under light.

For example, the formation of electronic trap states under

illumination has been proposed via a variety of mecha-

ions slowed the degradation, the decrease in rate did not

correlate directly with size, suggesting that other factors such as

hydrogen bonding may contribute to the increased stability.

Our results validate the substitution of all three large ions as a

promising strategy for signiﬁcantly increasing the stability of

hybrid organic−inorganic perovskites under light. Direct

comparison of these three diﬀerent ions as a function of

both size and substitution percentage allows us to identify

trends that may point toward the mechanism of this increased

stability. They also provide insight into the autocatalytic

mechanism of this degradation reaction, which will aid in

design strategies to overcome the crucial stability roadblock.

7,18

nisms.

Another proposed hypothesis is that degradation

may occur via photochemical reactions, either within the

hybrid organic−inorganic perovskite active material itself via

deprotonation or decomposition of the cation, or with

19−23

interfacial materials.

It has been proposed that light can

activate the transport of halide ions and/or organic

4−28

cations.

Explanations of the eﬀects of this ion migration

on degradation are also varied, from the creation of deep trap

states at ionic vacancy sites or ionic double layers at interfaces

29−31

to structural destabilization.

For example, in mixed halide

systems, ion migration has been shown to give rise to

photoinduced anionic phase segregation, which alters the

EXPERIMENTAL SECTION

■

32,33

stability of the system.

Some studies invoke the necessity

Lead(II) iodide (PbI ) (99.9985%) was purchased from Alfa Aesar.

of a combination of factors for the degradation to occur, such

as heat plus light, water plus light, or oxygen plus

Methylammonium iodide (MAI), imidazolium iodide, dimethylam-

monium iodide, and guanidinium iodide were purchased from

GreatCell Solar. Poly(3,4-ethylenedioxythiophene)−polystyrene sul-

fonate (PEDOT:PSS, Al4083) was purchased from Ossilla. All

chemicals were used as purchased.

Glass substrates were cleaned by sonicating 20 min each in

deionized (DI) water with detergent, DI water, acetone, and

isopropanol, then drying at 150 °C for at least 2 h. The PEDOT:PSS

underlayer was coated on glass substrates by treating substrates with

8,24,27,34−37

light.

In several studies, hybrid organic−inorganic

perovskites with the formamidinium cation substituted for the

methylammonium (MA) ion in the A-site of the lattice have

increased light stability.

24,38−40

However, there is a similar

debate around the exact mechanism by which the formami-

dinium ion increases the stability. It is larger than MA ion, has

a much smaller (near zero) dipole moment, and has two amine

groups capable of hydrogen bonding as opposed to the single

amine group present on the MA ion. These factors aﬀect the

UV−O

through a 0.45 μm polytetraﬂuoroethylene ﬁlter onto the substrate at

500 rpm for 40 s. PEDOT:PSS was then annealed at 150 °C for 30

min.

Solutions (1.5 M) of 1:1 (MAI)/PbI and 1:1 LAI/PbI (LA =

for 20 min, then spin-coating PEDOT:PSS solution ﬁltered

structure and electronics of mixed MA FA PbI perovskites

1−x

and have also been proposed to slow or stop light-induced ion

large ammonium ion) in 7:3 gamma-butyrolactone (GBL)/dimethyl

sulfoxide (DMSO) were prepared in a N -ﬁlled glovebox and

dissolved by stirring overnight in the dark at 60 °C. After the salts

completely dissolved, these solutions were mixed in varying ratios to

transport. Overall, there is a clear need for systematic studies

investigating the complex interaction between light and hybrid

organic−inorganic perovskite materials.

In this work, we show that the light-induced degradation for

MAPbI₃and A-site cation substituted derivatives are

autocatalytic. We calculated rate coeﬃcients for the degrada-

tion and found that partial substitution of MA cation with

imidazolium, dimethylammonium, and guanidinium ions slows

the light-induced degradation. The maximum decrease in rate

was observed in the 5% guanidinium-substituted sample, which

make mixed MA1−xLA PbI solutions.

x 3

Spin-coating was performed in an N -ﬁlled glovebox. First, 100 μL

of MA₁₋_xLA PbI solution was coated onto PEDOT:PSS-coated glass

substrates at 2000 rpm for 80 s, dripping toluene antisolvent at 40 s.

The resulting ﬁlms were annealed at 100 °C in the dark for 5 min.

UV−vis absorption spectra were measured on thin ﬁlms from 250−

100 nm on a Shimadzu UV-2401PC.

The eﬀective ionic radii of organic molecules were obtained from

degraded 63% slower than MAPbI . Although all three large

the literature. Dipole moments were obtained from the literature or

Inorg. Chem. XXXX, XXX, XXX−XXX

Figure 2. (a) Shift of the (110) peak in MAPbI (top) and hybrid organic−inorganic perovskites substituted with 20% imidazolium (IA),

guanidinium (GA), and dimethylammonium (DMA). The peak shift to lower angles indicates an increase in lattice size, which generally trends with

ionic size. Dimethylammonium (DMA) and guanidinium (GA) are switched as guanidinium (GA) is slightly larger, but the diﬀerence in

dimethylammonium (DMA) and guanidinium (GA) ionic radii (0.06 Å) is much smaller than the gap between these two ions and

methylammonium (MA) or imidazolium (IA) ions. (b) (110) planes in MAPbI . (c) d-spacing between the (110) planes versus increasing

substitution percentages. Spacing between these planes increases along with substitution percentage to a point, after which it levels out. For

imidazolium (IA) and dimethylammonium (DMA), this happens at ∼15−20%, with the larger dimethylammonium (DMA) settling at a larger d-

spacing value than the smaller imidazolium. Guanidinium (GA), the largest ion, starts out with a relatively small d-spacing compared to that of

dimethylammonium (DMA) or imidazolium (IA) but eventually increases to approximately equal to/slightly larger than dimethylammonium

(DMA) at ∼30%. (d) Increase in unit cell volume with increasing substitution percentage.

calculated using WebMO with a Gaussian Hartree−Fock 6-31G(d)

basis set.

(2.78 Å) (Figure 1). We focused on the eﬀect of size, although

these ions do diﬀer in other properties such as hydrogen

bonding and dipole moment (see Figure S3).

In samples prepared from precursor solutions containing

two A-site organic cations, PXRD showed peaks that are

characteristic of a three-dimensional (3D) tetragonal perov-

Samples were exposed to 5000 K LED light with ∼5 Sun intensity

for up to 24 h or until complete degradation of 3D perovskite phase

occurred (see Figure S1). The spectrum of the LED light and the

variations in the intensity and temperature are included in Figure S2.

This high intensity was chosen to accelerate complete light-induced

degradation of samples to take place over a 24 h period. Samples were

kept in ambient air at 35 °C and 35−45% relative humidity (RH).

Dark control samples were kept in the same ambient conditions.

All powder X-ray diﬀraction (PXRD) measurements were taken on

a SmartLab SE powder diﬀractometer. Data were analyzed using the

SmartLab Studio II software to remove background, identify peaks,

calculate peak area, assign peaks to tetragonal structure, and calculate

lattice parameters and unit cell volume. Lattice parameter reﬁnement

was done by the whole powder pattern ﬁtting (WPPF) method.

Degradation kinetics data (PXRD (110) peak area vs time exposed

to light) were ﬁt to the logistical equation described in the text using

Origin 2019 to obtain a composite rate coeﬃcient k.

skite structure that is homeotypic with 3D MAPbI tetragonal

structure, with slight peak shifts to lower 2θ values (see Figures

and S4). This peak shift is consistent with a lattice expansion

and is expected with the incorporation of ion larger than the

MA ion at the A-site (see Table S1 ). We also considered the

possibility of the formation of a “hollow perovskite” phase, as

has been observed on the addition of large ammonium

42,43

dications such as ethylenediammonium.

Indeed, powder

X-ray diﬀraction patterns for “hollow perovskite” are reported

to be nearly identical to their parent analogs, making them

diﬃcult to distinguish by this technique. However, the

introduction of localized lattice disruptions in “hollow

perovskites” result in signiﬁcant metal and halide vacancy

RESULTS AND DISCUSSION

■

Fabrication and Characterization of Substituted

Hybrid Organic−Inorganic Perovskite Derivatives. The

three large ammonium cations that we chose to replace the

MA ion (ionic radius = 2.17 Å) in the A-site of the archetypical

defects which alters the band gap of the material. In this

work, we do not observe signiﬁcant alterations in the band gap

after the incorporation of larger cations; thus, we conclude the

ﬁlms retain the 3D structure through the addition of larger

cations (Table S2 ).

MAPbI were (in order of increasing ionic radius) imidazolium

(2.58 Å), dimethylammonium (2.72 Å), and guanidinium

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

In PXRD, the perovskite ﬁlms showed preferred orientation

along (110) plane, so the (110) peak, which appears at a 2θ

value of ∼14, was the most intense in the PXRD pattern

have the same black-brown color as the MAPbI control, even

in compositions with higher percentages of the larger ion

where we observed that other phases coexist with the 3D

perovskite in PXRD. As the ion substitution increased, the

samples have a fainter brown color. The 100% substituted

samples were pale yellow for dimethylammonium and

imidazolium perovskites and bright yellow for the guanidinium

perovskite.

(Figure S4 , top). This peak was used to track the presence/

absence of the 3D mixed perovskite phase, both initially and as

these samples degraded. PXRD patterns of MAPbI along with

samples prepared from precursor solutions containing 10%

imidazolium, dimethylammonium, and guanidinium ions are

shown in Figure S4 .

The amount of a particular ion that can be incorporated into

the lattice before any measurable quantity of secondary phases

are formed did not appear to decrease with increasing size. The

smallest ion, imidazolium, formed 3D perovskite up to 10%

incorporation, whereas the largest ion, guanidinium, formed

primarily 3D perovskite up to 20% incorporation. The

dimethylammonium ion, which is of intermediate size, fell in

the middle at 15%. However, all three ions showed the

homeotypic 3D perovskite phase as the majority phase up to a

similar percentage (20% for dimethylammonium, and 30% for

guanidinium and imidazolium ions). We surmise that the

formation of secondary phases may be more strongly related to

the formation energy and stability of these lower-dimensional

secondary phases than the instability of the 3D perovskite

In samples prepared from precursor solutions containing up

to 10% imidazolium ions, PXRD showed that the 3D

perovskite structure was the exclusive crystalline phase, with

only small PbI₂impurity peaks arising from incomplete

conversion during fabrication (see Figures S5 −S7 ). For

samples prepared with more than 10% imidazolium ions in

the precursor solution, the 3D perovskite phase was still

present alongside lower-dimensional perovskite phases. The

D phase remained the majority crystalline phase for samples

prepared with 30% imidazolium ion in the precursor solution

based on relative peak intensities). Peaks attributable to the

D phase were observed even in samples prepared from

solutions containing 80% imidazolium ions. Samples with

00% of imidazolium ion as the A-site ion showed only peaks

in the PXRD that corresponds to the 1D chain perovskite

(

44−48

structure.

A comparison of the PXRD patterns of samples containing

0, 50, and 100% dimethylammonium ions is shown in Figure

S11 . For samples containing 20−80% of dimethylammonium

ions, every peak can be attributed to either the 3D tetragonal

perovskite or 1D chain perovskite phase, suggesting these two

phases coexist. Perovskite samples containing imidazolium ions

also behaved similarly. Guanidinium-containing perovskites,

however, have more than one phase present at 100%

substitution and have peaks in the midrange substitution

percentages, which cannot be attributed to either the 100%

phases or the 3D tetragonal phase.

structure, as previously observed in the literature. PXRD

patterns of samples with 100% of the A-site cation as

imidazolium, dimethylammonium, and guanidinium ions are

shown in Figures S8 and S9 .

In samples prepared from precursor solutions containing a

mixture of 85% methylammonium and 15% dimethylammo-

nium ions, PXRD showed peaks attributable exclusively to the

D perovskite structure. PbI₂impurity peaks were not

observed in the dimethylammonium containing perovskites.

For precursor solution compositions containing greater than

5% dimethylammonium ions, the 3D mixed perovskite phase

We also observed shifts of the (110) and (220) peaks to

lower 2θ angles in the PXRD patterns of the substituted hybrid

organic−inorganic perovskites, indicating an increase in lattice

size (Figure 2a). This increase was roughly correlated to size,

as larger ions led to a larger shift. Samples containing

dimethylammonium ions did show a larger shift at lower

substitution percentages (up to 20%) compared to samples

with guanidinium ions, but these ions are much more similar in

size (0.06 Å diﬀerence in ionic radius) than methylammonium

or imidazolium is to either of them. Guanidinium-containing

perovskites showed a larger shift than dimethylammonium

containing perovskites at substitution percentages of over 30%.

was still present alongside lower-dimensional perovskite

phases. The 3D phase remained the majority phase up to

compositions containing 20% dimethylammonium ions. The

D phase remained through compositions containing 80%

dimethylammonium ions. PXRD peaks attributable to the 3D

phase was not observed in the sample prepared from solutions

0% dimethylammonium ions. The sample with 100%

dimethylammonium ion had the 1D chain perovskite structure,

as previously observed in literature with a preferred (010)

orientation (see Figure S8b ).

In samples prepared from precursor solutions containing a

mixture of methylammonium and up to 20% guanidinium ions,

PXRD showed peaks attributable exclusively to the 3D

perovskite structure with only small PbI₂impurity peaks

coming from incomplete conversion during fabrication. At

compositions above 20% guanidinium ions, the 3D mixed

perovskite phase was still present alongside other lower-

dimensional perovskite phases (see Figure S7 ). The 3D phase

remained the majority phase through precursor solution

compositions up to 30% guanidinium ion. Some of this 3D

phase remained through compositions containing 80%

guanidinium ions. PXRD peaks attributable to the 3D phase

was not observed in samples from precursor solution

compositions with 90% guanidinium ions. The sample with

An example of the (110) peak shift in MAPbI and the 20%

substituted samples is shown in Figure 2. The peaks shifted

further to lower 2θ with increasing substitution percentages for

all three ions; the d₁₁₀versus substitution percentage is shown

in Figure 2. The spacing between these planes increased with

increasing substitution for all three large ions, until ∼15−20%

incorporation for imidazolium and dimethylammonium and

40% incorporation for guanidinium, at which point it leveled

out to a relatively constant value.

The increase in lattice size, which trends with both ionic size

and increasing substitution percentage, indicates that the large

ions are incorporated into the A-site of the 3D lattice and not

merely present in grain boundaries or on the surface.

Additionally, the lattice size stays relatively constant even at

high substitution percentages where we know that the lower-

dimensional perovskite phases coexist with the 3D perovskite

phase. This indicates that larger ions are not excluded in favor

00% guanidinium ions had 1D and 2D perovskite structures,

46−48

as previously observed in literature (see Figure 9).

Photographs of the samples with mixed A-cation composi-

tions are included in Figure S10. A majority of the samples

Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Figure 3. (a) Area of (110) PXRD peak vs time exposed to light for hybrid organic−inorganic perovskites with varying percentages of

dimethylammonium (DMA). Hybrid organic−inorganic perovskites substituted with all three ions showed these “S”-shaped curves, indicating an

autocatalytic degradation reaction mechanism. (b) Schematic of the phases of the autocatalytic reaction which give rise to the “S”-shaped curves

and equation ﬁtting these curves. We extract a rate coeﬃcient k for the degradation reaction from the ﬁtting for each of these samples.

of a mixed 3D perovskite with lower substitution percentage

even when other crystal structure pathways are present for the

large cations. Thus, for ensuing discussions, we surmise that a

deﬁned fraction of the methylammonium ion has been

substituted with the large organic cation.

would improve the stability of these compounds under light.

To test our hypothesis, we exposed the 0−100% substituted

thin ﬁlms to a very bright (∼5 Sun) LED light for 24 h under

ambient conditions and monitored their degradation. The high

brightness was chosen in order to (1) accelerate the

degradation reaction so that we could observe the entire

degradation over the 24 h period and (2) ensure that light was

the primary stressor in causing degradation over other

environmental factors such as heat or humidity since the

experiments were carried out in ambient air. Control samples

were kept in the dark in the same ambient conditions as the

degraded samples to determine whether other conditions such

as temperature or humidity did cause any degradation over the

24 h period. These samples did not show any degradation by

PXRD analysis over the 24 h period, indicating that the

degradation measured in the primary samples is light-induced

(Figure S13). Photographs of the samples before and after 24 h

of light exposure, as well as the dark control samples, are

included in Figure S10. After complete degradation, all of the

ﬁlms turned yellow. This color comes from a mix of bright

The lattice parameters and increase in unit cell volume for

all substituted compounds up to 20% were calculated based on

PXRD data, as shown in Figure 2 and detailed in Table S1. As

expected from the (110) and (220) peak shifts, the volume of

the unit cell increased with increasing substitution percentage

to accommodate the larger ions. The trends for dimethylam-

monium and imidazolium ions seem to correlate to their size,

as both increase at similar rates, but imidazolium ion gave a

smaller unit cell volume than did the dimethylammonium ion.

Guanidinium ion, however, increased in unit cell volume much

more slowly than did dimethylammonium or imidazolium ion

at low substitution percentages, but eventually reached a

volume greater than that of imidazolium and similar to that of

dimethylammonium ion. MAPbI gave a unit cell volume of

85 Å . In samples with 20% of the large cations, we obtained a

unit cell volume of 1002 Å (1.7% increase) with for

imidazolinium, 1004 Å (1.7% increase) for guanidinium,

and 1005 Å (2.3% increase) for dimethylammonium ions.

yellow PbI , the primary degradation product of MAPbI , and

the lower dimensional perovskites (imidazolium and dimethy-

lammonium lead triodides are pale yellow compounds,

whereas guanidinium lead triodide is bright yellow). Images

of the setup for the light degradation experiments, the

spectrum of the LED light, and the variations in the intensity

and temperature over time are included in Figures S1 and S2.

The intensity and temperature were reasonably constant over

experiment time.

Lattice parameters were only reﬁned up to 20% substitution,

but on the basis of the peak shifts for 30−40% guanidinium

ion, it is likely that the lattice of guanidinium-based perovskites

continues to expand to the largest overall value after

dimethylammonium and imidazolium ions volumes become

constant.

UV−vis spectra were recorded for all of the substituted

compounds up to 20%, and the band gaps were calculated

from Tauc plots of this data, shown in Figure S12. While the

samples showed a slight increase in absorption from 350 to 550

nm and a slight decrease in absorption from 550 to 750 nm,

the overall absorption proﬁle remained very similar for all

substituted compounds. The calculated band gaps varied by

We used PXRD to track the light-induced degradation of the

3D perovskite structure by measuring the samples every 4 h

through a 24 h period and calculated the decrease in (110)

peak area (the most intense peak of the 3D perovskite phase).

Using this peak area versus time data, we were able to rule out

simple zero-, ﬁ rst-, and second-order kinetic mechanisms

(shown in Figure S14). We noticed a characteristic “S”-

sigmoidal shape of all of the peak area versus time plots, an

1% for all substituted compounds and did not show any

49−51

signiﬁcant trend. The calculated band gaps are listed in Table

example of which is shown in Figure 3.

This shape is

S2. This maintenance of the band gap of MAPbI in the

characteristic of an autocatalytic reaction, where one of the

reaction products acts as a catalyst to speed up the reaction.

Autocatalytic decomposition in solid-state systems can be ﬁt to

substituted hybrid organic−inorganic perovskites, which is

well-suited for photovoltaic applications, is promising for the

viability of these new compounds.

49−51

eq 1 below:

Light-Induced Degradation Kinetics. We hypothesized

that substituting the MA ion in MAPbI₃with larger

imidazolium, dimethylammonium, and guanidinium cations

a₀

a =

1 + e⁻^k⁽^t⁺^t^c⁾

(

Inorg. Chem. XXXX, XXX, XXX−XXX

Figure 4. (a) Normalized rate coeﬃcients (k) for the autocatalytic light-induced degradation reaction of the substituted hybrid organic−inorganic

perovskites. All three of the ions at every substitution percentage showed slower light-induced degradation than MAPbI . The material with 5%

guanidinium (GA) showed the slowest degradation, a 62.48% decrease in the rate coeﬃcient from MAPbI . (b) Normalized rate coeﬃcients (k) for

formation of PbI during the light-induced degradation reaction. Increasing percentages of guanidinium (GA) and dimethylammonium (DMA)

formed PbI more slowly than MAPbI , as expected from their slower degradation. Imidazolium (IA), however, showed faster PbI formation than

MAPbI despite degrading more slowly.

where a = area of (110) peak, a = area of (110) peak at time

Interestingly, the decrease in k did not directly correspond

to an increase in ionic size. The largest ion, guanidinium, led to

the most stable compounds at nearly all substitution

percentages. However, the smallest of the three ions,

imidazolium, led to more stable compounds than did the

larger dimethylammonium ion. This indicates that factors

beyond size must be considered in evaluating the impact of

new organic cations on stability under light.

zero, k = composite rate coeﬃcient, t = time (hours), and t =

critical point. Using this to ﬁt our peak area versus time data,

we were able to obtain a composite rate coeﬃcient k for the

light-induced degradation reaction of each of the substituted

derivatives. A schematic of how this reaction mechanism and

equation lead to the observed “S”-shaped sigmoidal curve is

shown in Figure 3. Rate coeﬃcients were only calculated up to

Guanidinium-containing perovskites exhibited behavior that

deviated from expected size trends both in initial structure and

degradation kinetics. Guanidinium ion diﬀers from the other

ions in that it has no dipole and has several more potential

0% substitution, as many of the higher substitution

percentages degraded before 24 h and thus had fewer data

points. All of the calculated rate coeﬃcients and their percent

change compared to MAPbI can be found in Table S3. This

complex rate coeﬃcient may capture more than one

degradation pathway, as discussed below, but it oﬀers a useful

tool for comparing the relative degradation rate of the

materials with one another.

hydrogen bonding sites. The decrease in k did trend with the

number of amine functional groups capable of hydrogen

bonding: dimethylammonium (smallest decrease in k) has one,

imidazolium (middle k) has two, and guanidinium (largest

decrease in k) has three. Size undoubtedly plays a role,

however. Although dimethylammonium and methylammo-

nium ions have the same number of amine functional groups,

the amine group on the methylammoninum ion is less

sterically hindered; thus, the ammonium group on the

methylammonium ion can form stronger hydrogen bonds

than dimethylammonium. Therefore, if only hydrogen bonding

were responsible for the increase in stability, then dimethy-

lammonium-substituted compounds would likely be less stable

We found that every one of the substituted derivatives up to

0% had a smaller k, corresponding to a slower light-induced

degradation reaction, compared to MAPbI . The k numbers as

a function of large cation substitution percentage normalized

to their MAPbI control are shown in Figure 4a. All three

cations showed a steep initial decrease in k from ∼0−5%

substitution, after which they either leveled out or increased

slightly. However, none ever went back up to the level of

MAPbI . Dimethylammonium-containing compounds showed

than MAPbI . The strength of hydrogen bonds in these

the smallest decrease in rate coeﬃcients compared to MAPbI₃,

with a maximum decrease in k of 37% at 20% dimethylammo-

nium substitution. The maximum decrease in k for

imidazolium-containing perovskites was 51% at 7.5% imidazo-

lium substitution. Guanidinium containing perovskites showed

the largest decrease in k of all of the large cations, with a

maximum decrease of 62.48% with 5% guanidinium sub-

stitution. Even the smallest decrease in k (14.6% for 2.5%

dimethylammonium substitution) represents signiﬁcantly

materials is diﬃcult to quantify, as factors such as geometry,

rotational freedom, polarizability, and potential for concerted

hydrogen bonding could all play a role.

Since PbI is the primary degradation product of MAPbI

and we know that the substituted samples degrade more slowly

than MAPbI , we might expect that the rate of PbI formation

would decrease with increasing large cation substitution. The

rate of PbI formation shown in Figure 4b was calculated based

on ﬁtting the area of the (001) peak versus time plot (which

showed the same unique S-shape) and is shown as a function

of organic cation substitution. This does seem to be the case

for dimethylammonium and guanidinium. Both showed slower

PbI formation than MAPbI , and the rate decreased more in

improved light stability compared to MAPbI . This is strong

evidence supporting our hypothesis that substituting any of the

three larger imidazolium, dimethylammonium, and guanidi-

nium cations for MA ions would lead to more stable hybrid

organic−inorganic perovskite compounds.

guanidinium than it did in dimethylammonium, which was

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

consistent with trends in degradation rate. Samples with

imidazolium ions, however, showed the opposite behavior,

with the PbI formation rate increasing as a percentage of

imidazolium substitution.

The rate of PbI formation does not give a complete picture

of degradation products, however, as alternative lower-

dimensional degradation products can also form when these

larger cations are introduced. Diﬀerences in the formation

energy and stability of these alternative products could

inﬂuence the rates of formation of PbI in addition to the

overall degradation rate. We were unable to quantify the rate of

formation for these 1D products since for the majority of the

lower percent substitution samples; this peak did not appear at

all or did not appear until very late in degradation (values listed

in Table S4 ). We also note that if amorphous products form

during the degradation then PXRD will not be useful to

quantify their rates of formation. General observations point to

diﬀerences in 1D product formation between the three ions.

Additionally, amorphous forms of these materials present in

the original sample or as a product of degradation could play a

role, which we could not quantify due to the limitations of

PXRD. Analysis of the role of these amorphous materials will

be the subject of future study.

We observed no evidence of the most intense (110) peak

from the 1D imidazolium perovskite phase in imidazolium

samples up to 10% substitution. Samples with 12.5−20%

imidazolium had this peak, but its area did not change with

time of light exposure and did not strongly correlate to the

percent of imidazolium substitution. Samples with 0−15%

dimethylammonium ions did not show any evidence of the

most intense (010) peak from the 1D dimethylammonium

phase, except for a very small peak that appeared at 12−16 h

and disappeared at 20 h. Samples containing 20% dimethy-

lammonium ion, however, showed a 1D peak, which degraded

with increasing light exposure time. None of the 0−20%

guanidinium containing samples showed the most intense

4,55

(

120) 1D guanidinium phase peak at 0 h.

The 0−5%

guanidinium-containing samples never showed this peak

through degradation, but it appeared for the 20% sample at

h, the 10, 12.5, and 15% samples at 8 h, and the 7.5% sample

at 12 h. Additionally, unlike the 1D peaks in imidazolium or

dimethylammonium containing samples, the area of the 1D

peaks for all of the guanidium-containing samples increased

with both increasing degradation and increased guanidinium

substitution percentage. This may point to a lower formation

energy of 1D guanidinium degradation products compared to

Figure 5. Normalized area of (110) peak in MAPbI

and hybrid

organic−inorganic perovskites substituted with 7.5, 10, and 12.5%

imidazolium (IA) (a), guanidinium (GA) (b), and dimethylammo-

nium (DMA) (c). The ﬁrst point is the initial value, and the second

point is after 4 h of light-induced degradation. The third point is after

PbI , which seems to be favorable given that guanidinium

shows the largest increase in stability.

To gain further insight into the mechanism of light-induced

degradation, the reversibility of this reaction was investigated.

4 h of resting in the dark. MAPbI degraded more than any of the

substituted samples in 4 h. However, MAPbI₃degradation was

partially reversible, while the substituted hybrid organic−inorganic

perovskites either stayed the same or degraded slightly more.

Samples of MAPbI and 7.5, 10, and 12.5% large cation

substituted hybrid organic−inorganic perovskites were meas-

ured by PXRD, exposed to light for 4 h and measured again,

then placed in the dark for 24 h and measured a ﬁnal time

degradation factors that aﬀect the substituted hybrid organic−

(

Figure 5). After the 4 h light degradation, MAPbI (black

inorganic perovskites diﬀerently than MAPbI .

curve in Figure 5a−c) degraded to 79% of its initial value,

signiﬁcantly more than any of the substituted samples, as

Light-activated ion transport could act as a reversible

4,28,29,31,56

degradation mechanism.

The more stable substi-

expected. After resting for 24 h in the dark, MAPbI recovered

tuted samples may show less ion migration out of the lattice

to 87% of its original value, indicating the partial reversibility of

compared to MAPbI because of increased hydrogen bonding

this reaction, as observed in previous studies. However, none

of the substituted samples showed any reversibility in their

degradation, either staying constant or degrading slightly more.

This points to a combination of both reversible and irreversible

with the inorganic lattice, increased steric hindrance for

methylammonium ions, increased activation energy for the

movement of larger cations, or passivation of grain boundaries

as a path for ion movement. However, by the same argument,

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

after the light is turned oﬀ, the ions which moved out of the

structure may not be able to diﬀuse back into the structure.

This would lead to a lack of recoverability compared to the

smaller, more mobile methylammonium ion in MAPbI₃

samples.

dimethylammonium, and guanidinium ions provides some new

insights. There are both reversible and irreversible components

to the degradation. The substituted hybrid organic−inorganic

perovskites appear to degrade by the irreversible mechanism,

but the eﬀects of the reversible mechanism are substantially

reduced. Reactive iodide species may be formed under light,

and ﬁnding organic cations which do not undergo these

irreversible reactions (either because of steric hindrance,

Chemical reactions could act as irreversible degradation

mechanisms, and logically, these diﬀerent molecules would

have diﬀerent reactivities. One study conducted by Pont et al.

observed autocatalytic degradation of mixed halide hybrid

increased pK , or diﬀerent molecular geometries) should

organic−inorganic perovskites under light in ambient air.

increase stability. All of the cations tested here could undergo

They ruled out oxygen diﬀusion through the ﬁlm as the cause

for the delayed onset of light-induced degradation, instead

hypothesizing that one of the degradation reaction products

was acting as the catalyst. They noted, however, that ﬁlms with

bromide were much more light-stable than those with iodide.

chemical reactions to some extent, but the low pK of

dimethylammonium and guanidinium ions may be contribu-

ting to their increased stability. Chemical reactions do not

seem to be the only source of degradation in MAPbI ,

however, as a distinct reversibility is observed in these samples

which is not present in the substituted derivatives. This

reversible degradation seems to be the primary pathway that is

avoided in the substituted hybrid organic−inorganic perov-

skites. Increased activation energy for ion transport with the

larger ions may lead to this decrease in reversible light-induced

degradation; this will be investigated in future works.

I is a common reactive species invoked in hybrid organic−

inorganic perovskites, but there is disagreement on how it is

58−62

formed and how it reacts.

Fu et al. identiﬁed an autocatalytic light-induced degradation

reaction in FA MA Pb(I Br ) hybrid organic−inorganic

1−x

1−y 3

perovskite photovoltaic devices and assigned I as the catalyst

in their systems. They observed that their encapsulated

devices ruptured from the built-up pressure of I after a certain

CONCLUSIONS

■

amount of light-induced degradation, while bare ﬁlms showed

signiﬁcantly increased stability. They measured the gases which

We successfully fabricated hybrid organic−inorganic perov-

skites with 0−100% substitution of the methylammonium

(MA) cation in archetypical MAPbI with the larger

evolved from their system and found I and CH I. However,

the reaction where I⁻displaces NH may not be able to

imidazolium, dimethylammonium, and guanidinium com-

proceed similarly with all the compounds we used in this work.

Dimethylammonium ion could potentially react similarly with

pounds. These substituted hybrid organic−inorganic perov-

skites maintained a MAPbI -like 3D tetragonal perovskite

CH NH as the leaving group rather than NH . Imidazolium

structure as a primary or secondary phase up to high

substitution percentages, despite predictions that these cations

were too large to ﬁt into the methylammonium ion lattice site.

These compounds showed increasing lattice size with

increasing substitution percentage as well as with increasing

ionic radius, suggesting their integration into the lattice itself.

We also do not see any evidence that the largest ion we tested,

guanidinium (2.78 Å), represents the upper limit of the size of

ion that can be incorporated into the 3D lattice. Therefore, it is

possible that ions even larger than guanidinium could be

substituted for the methylammonium ion to create new hybrid

organic−inorganic perovskite derivatives. The band gaps of

these compounds remained constant at the same value as

−

and guanidinium ions, however, are unlikely to react with I

ions via a substitution reaction.

In contrast, Wang et al. proposed that I could form I

−

under light, which would react with CH NH to form

CH NH , I , and H . All of the species used in this work have

protons which could be removed. We examined the pK values

for the ions used in this work. In order from most to least

acidic were imidazolium (6.95), methylammonium (10.6),

dimethylammonium (10.8), and guanidinium (12.5) ions. This

does not exactly follow the degradation trends, as imidazolium-

substituted samples degrade more slowly than MAPbI₃.

However, it does trend with the rate of PbI formation, so

deprotonation may be involved in the pathway of the

MAPbI , which is well-suited for the desired photovoltaic

degradation reaction, which forms PbI as a product.

applications. After characterizing their degradation under

illumination, we found that every one of the hybrid organic−

inorganic perovskite derivatives up to 20% substitution showed

Light is a primary driving factor for the accelerated

autocatalytic degradation observed across these studies, in

hybrid organic−inorganic perovskites with both mixed cations

increased stability compared to MAPbI , with a maximum

and mixed halides. Fu et al. conducted their work in an N

decrease in the degradation rate of 62% for 5% guanidinium

ion substitution. We observed a unique autocatalytic

mechanism of this light-induced degradation, which will

provide important insights into further understanding of how

these cations improve hybrid organic−inorganic perovskite

stability and how to design even more stable materials. We also

noted how characteristics of the larger cations other than size,

such as dipole and hydrogen bonding, may play a role in

stability. This was especially noticeable for guanidinium ion,

the only ion with no dipole and three amine sites, which

showed noticeable deviations from size trends in both structure

and degradation kinetics. Overall, we have conﬁrmed our

hypothesis that substitution of imidazolium, dimethylammo-

nium, and guanidinium ion for the methylammonium ion in

MAPbI₃hybrid organic−inorganic perovskites yield com-

pounds with improved light stability which are promising

atmosphere under high heat (80 °C) and noted that the

degradation would not proceed without light but that even a

low-intensity light would trigger it. Pont et al. observed the

same autocatalysis, but at room temperature under ambient

conditions with a very low-intensity ﬂuorescent light (∼2 mW

−

cm intensity at 550 nm). We conducted our studies in

ambient air at 35 °C and 35−45% RH with a very intense LED

light and observed this same autocatalytic degradation, and we

did not observe any degradation under these same conditions

in the dark. Therefore, we surmise that the visible light is a

necessary factor that initiates autocatalytic decomposition.

The inﬂuence of light on hybrid organic−inorganic perov-

skites that leads to degradation is complex and multifaceted,

but the increased stability of these new hybrid organic−

inorganic perovskite derivatives substituted with imidazolium,

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Docampo, P. Capturing the Sun: A Review of the Challenges and

Article

alternatives for use in hybrid organic−inorganic perovskite

Perspectives of Perovskite Solar Cells. Adv. Energy Mater. 2017, 7,

photovoltaic applications.

(

700264.

4) Grat

Mater. 2014, 13, 838−842.

5) Correa-Baena, J.-P.; Abate, A.; Saliba, M.; Tress, W.; Jacobsson,

J. T.; Gratzel, M.; Hagfeldt, A. The rapid evolution of highly efficient

perovskite solar cells. Energy Environ. Sci. 2017, 10, 710−727.

6) Sherahilo, T. Oxford PV perovskite solar cell achieves 28%

zel, M. The light and shade of perovskite solar cells. Nat.

ASSOCIATED CONTENT

sı Supporting Information

The Supporting Information is available free of charge at

https://pubs.acs.org/doi/10.1021/acs.inorgchem.0c01133.

■

(

Molecular properties of organic cations tested, PXRD of

substituted compounds, pictures of pristine and

degraded samples, lattice parameters, experimental

setup, spectrum of light used in degradation experi-

ments, demonstration of lack of ﬁt to simple kinetic

models, rate coeﬃcients, peak areas of 1D products

e ﬃ ciency. Oxford PV . https://www.oxfordpv.com/news/oxford-pv-

perovskite-solar-cell-achieves-28-e ﬃ ciency, 2018.

Nickel, N. H. Influence of Radiation on the Properties and the

7) NREL. Best Research-Cell Eﬃciency Chart. https://www.nrel.

gov/pv/cell-e ﬃ ciency.html, 2019.

8) Lang, F.; Shargaieva, O.; Brus, V. V.; Neitzert, H. C.; Rappich, J.;

(

Stability of Hybrid Perovskites. Adv. Mater. 2018, 30, 1702905.

(PDF)

(9) Deretzis, I.; Smecca, E.; Mannino, G.; La Magna, A.; Miyasaka,

AUTHOR INFORMATION

Corresponding Author

D. Venkataraman − University of Massachusetts Amherst,

Department of Chemistry, Amherst, Massachusetts 01003,

United States; orcid.org/0000-0003-2906-0579;

Email: dv@umass.edu

T.; Alberti, A. Stability and Degradation in Hybrid Perovskites: Is the

■

Glass Half-Empty or Half-Full? J. Phys. Chem. Lett. 2018, 9, 3000−

(

007.

10) Rong, Y.; Hu, Y.; Mei, A.; Tan, H.; Saidaminov, M. I.; Seok, S.;

McGehee, M. D.; Sargent, E. H.; Han, H. Challenges for

commercializing perovskite solar cells. Science 2018, 361, eaat8235.

(11) Boyd, C. C.; Cheacharoen, R.; Leijtens, T.; McGehee, M. D.

Understanding Degradation Mechanisms and Improving Stability of

Authors

Perovskite Photovoltaics. Chem. Rev. 2019, 119, 3418−3451.

Christie L. C. Ellis − University of Massachusetts Amherst,

Department of Chemistry, Amherst, Massachusetts 01003,

United States

Hamza Javaid − University of Massachusetts Amherst,

Department of Chemistry, Amherst, Massachusetts 01003,

United States

Emily C. Smith − University of Massachusetts Amherst,

Department of Chemistry, Amherst, Massachusetts 01003,

United States

(12) Smith, E. C.; Ellis, C. L. C.; Javaid, H.; Renna, L. A.; Liu, Y.;

Russell, T. P.; Bag, M.; Venkataraman, D. Interplay between Ion

Transport, Applied Bias, and Degradation under Illumination in

Hybrid Perovskite p-i-n Devices. J. Phys. Chem. C 2018, 122, 13986−

13994.

(13) Nickel, N. H.; Lang, F.; Brus, V. V.; Shargaieva, O.; Rappich, J.

Unraveling the Light-Induced Degradation Mechanisms of

700158.

14) Egger, D. A.; Rappe, A. M.; Kronik, L. Hybrid Organic −

Inorganic Perovskites on the Move. Acc. Chem. Res. 2016, 49, 573−

81.

15) Cao, H.; Li, J.; Dong, Z.; Su, J.; Chang, J.; Zhao, Q.; Li, Z.;

NH PbI Perovskite Films. Adv. Electron. Mater. 2017, 3,

3 3

Complete contact information is available at:

https://pubs.acs.org/10.1021/acs.inorgchem.0c01133

(

Author Contributions

Yang, L.; Yin, S. Reducing Defects in Perovskite Solar Cells with

White Light Illumination-Assisted Synthesis. ACS Energy Lett. 2019,

†

H.J. and E.S. contributed equally to this work

, 2821−2829.

16) Khenkin, M. V.; Katz, E. A.; Abate, A.; Bardizza, G.; Berry, J. J.;

Brabec, C.; Brunetti, F.; Bulovic, V.; Burlingame, Q.; Di Carlo, A.;

Cheacharoen, R.; Cheng, Y. B.; Colsmann, A.; Cros, S.; Domanski, K.;

Dusza, M.; Fell, C. J.; Forrest, S. R.; Galagan, Y.; Di Girolamo, D.;

̈ ̈

Gratzel, M.; Hagfeldt, A.; von Hauff, E.; Hoppe, H.; Kettle, J.; Kobler,

H.; Leite, M. S.; Liu, S.; Loo, Y. L.; Luther, J. M.; Ma, C. Q.; Madsen,

M.; Manceau, M.; Matheron, M.; McGehee, M.; Meitzner, R.;

Funding

(

We gratefully acknowledge the ﬁnancial support of the US

Army CCDC Solider Center through contract no.

W911QY1820002 for this work. The powder X-ray diﬀrac-

tometer used in this study was acquired through the National

Science Foundation Major Research Instrumentation Program

(

NSF CHE-1726578).

Notes

Nazeeruddin, M. K.; Nogueira, A. F.; Odabas i̧ , Ç.; Osherov, A.; Park,

The authors declare no competing ﬁnancial interest.

N. G.; Reese, M. O.; De Rossi, F.; Saliba, M.; Schubert, U. S.; Snaith,

H. J.; Stranks, S. D.; Tress, W.; Troshin, P. A.; Turkovic, V.; Veenstra,

S.; Visoly-Fisher, I.; Walsh, A.; Watson, T.; Xie, H. B.; Yildirim, R.;

Zakeeruddin, S. M.; Zhu, K.; Lira-Cantu, M. Consensus statement for

stability assessment and reporting for perovskite photovoltaics based

on ISOS procedures. Nat. Energy 2020, 5, 35−49.

ACKNOWLEDGMENTS

■

We thank Professor Mark Turnbull of Clark University for his

help in gathering initial PXRD data of the substituted

compounds. We thank Dr. Ravi Mosurkal, Professor Jayant

Kumar, Professor Lian Li, and Dr. Youngju Bae of University

of Massachusetts Lowell for fruitful discussions.

(17) Nie, W.; Blancon, J.-C.; Neukirch, A. J.; Appavoo, K.; Tsai, H.;

Chhowalla, M.; Alam, M. A.; Sfeir, M. Y.; Katan, C.; Even, J.; et al.

Light-activated photocurrent degradation and self-healing in perov-

skite solar cells. Nat. Commun. 2016, 7, 11574.

(18) Kwak, K.; Lim, E.; Ahn, N.; Heo, J.; Bang, K.; Kim, S.; Choi, M.

An atomistic mechanism for the degradation of perovskite solar cells

by trapped charge. Nanoscale 2019, 11, 11369−11378.

(19) Akbulatov, A. F.; Frolova, L. A.; Griffin, M. P.; Gearba, I. R.;

Dolocan, A.; Vanden Bout, D. A.; Tsarev, S.; Katz, E. A.; Shestakov, A.

F.; Stevenson, K. J.; Troshin, P. A. Effect of Electron-Transport

Material on Light-Induced Degradation of Inverted Planar Junction

Perovskite Solar Cells. Adv. Energy Mater. 2017, 7, 1700476.

REFERENCES

■

2) Jena, A.; Kulkarni, A.; Miyasaka, T. Halide Perovskite

1) Green, M. A.; Ho-Baillie, A.; Snaith, H. J. The emergence of

perovskite solar cells. Nat. Photonics 2014, 8, 506 −514.

Photovoltaics: Background, Status, and Future Prospects. Chem.

Rev. 2019, 119, 3036−3103.

(3) Petrus, M. L.; Schlipf, J.; Li, C.; Gujar, T. P.; Giesbrecht, N.;

Muller-Buschbaum, P.; Thelakkat, M.; Bein, T.; Huttner, S.;

̈ ̈

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Blocking Layer of Sb S on Nanocrystalline TiO for CH NH PbI

Article

(

20) Ito, S.; Tanaka, S.; Manabe, K.; Nishino, H. Effects of Surface

(37) Boote, B. W.; Andaraarachchi, H. P.; Rosales, B. A.; Blome-

Fernandez, R.; Zhu, F.; Reichert, M. D.; Santra, K.; Li, J. Z.; Petrich, J.

W.; Vela, J.; Smith, E. A. Unveiling the Photo- and Thermal-Stability

of Cesium Lead Halide Perovskite Nanocrystals. ChemPhysChem

2019, 20, 2647−2656.

Perovskite Solar Cells. J. Phys. Chem. C 2014, 118, 16995−17000.

(21) Song, Z.; Wang, C.; Phillips, A. B.; Grice, C. R.; Zhao, D.; Yu,

Y.; Chen, C.; Li, C.; Yin, X.; Ellingson, R. J.; Heben, M. J.; Yan, Y.

Probing the origins of photodegradation in organic −inorganic metal

halide perovskites with time-resolved mass spectrometry. Sus. Energy

Fuels. 2018, 2, 2460−2467.

(38) Charles, B.; Dillon, J.; Weber, O. J.; Islam, M. S.; Weller, M. T.

Understanding the stability of mixed A-cation lead iodide perovskites.

J. Mater. Chem. A 2017, 5, 22495−22499.

(22) Shlenskaya, N.; Belich, N. A.; Grat

zel, M.; Goodilin, E. A.;

(39) Zhang, Y.; Grancini, G.; Feng, Y.; Asiri, A. M.; Nazeeruddin, M.

Tarasov, A. B. Light-induced reactivity of gold and hybrid perovskite

as a new possible degradation mechanism in perovskite solar cells. J.

Mater. Chem. A 2018, 6, 1780−1786.

K. Optimization of Stable Quasi-Cubic FA MA ₁ ₋_x PbI ₃ Perovskite

Structure for Solar Cells with Efficiency beyond 20%. ACS Energy Lett.

23) Juarez-Perez, E. J.; Ono, L. K.; Maeda, M.; Jiang, Y.; Hawash,

017, 2, 802−806.

(

40) Binek, A.; Hanusch, F. C.; Docampo, P.; Bein, T. Stabilization

Z.; Qi, Y. Photodecomposition and thermal decomposition in

methylammonium halide lead perovskites and inferred design

principles to increase photovoltaic device stability. J. Mater. Chem.

A 2018, 6, 9604−9612.

of the Trigonal High-Temperature Phase of Formamidinium Lead

Iodide. J. Phys. Chem. Lett. 2015, 6, 1249−1253.

(41) Kieslich, G.; Sun, S.; Cheetham, A. K. Solid-state principles

applied to organic −inorganic perovskites: new tricks for an old dog.

(24) Bag, M.; Renna, L. A.; Adhikari, R. Y.; Karak, S.; Liu, F.; Lahti,

Chem. Sci. 2014, 5, 4712−4715.

P. M.; Russell, T. P.; Tuominen, M. T.; Venkataraman, D. Kinetics of

Ion Transport in Perovskite Active Layers and Its Implications for

Active Layer Stability. J. Am. Chem. Soc. 2015, 137, 13130−13137.

(42) Ke, W. J.; Stoumpos, C. C.; Zhu, M. H.; Mao, L. L.;

Spanopoulos, I.; Liu, J.; Kontsevoi, O. Y.; Chen, M.; Sarma, D.;

Zhang, Y. B.; et al. Enhanced photovoltaic performance and stability

(25) Smith, E. C.; Ellis, C. L. C.; Javaid, H.; Arden, B. G.;

with a new type of hollow 3D perovskite {en}FASnI . Sci. Adv. 2017,

Venkataraman, D. The use of ion-selective membranes to study cation

transport in hybrid organic −inorganic perovskites. Phys. Chem. Chem.

Phys. 2019, 21, 20720−20726.

(

, e1701293.

43) Spanopoulos, I.; Ke, W. J.; Stoumpos, C. C.; Schueller, E. C.;

Kontsevoi, O. Y.; Seshadri, R.; Kanatzidis, M. G. Unraveling the

Chemical Nature of the 3D “Hollow ” Hybrid Halide Perovskites. J.

Am. Chem. Soc. 2018, 140, 5728−5742.

(26) Domanski, K.; Roose, B.; Matsui, T.; Saliba, M.; Turren-Cruz,

S.-H.; Correa-Baena, J.-P.; Carmona, C.; Richardson, G.; Foster, J. M.;

De Angelis, F.; et al. Migration of cations induces reversible

performance losses over day/night cycling in perovskite solar cells.

Energy Environ. Sci. 2017, 10, 604−613.

(44) W ę c ł awik, M.; G ą gor, A.; Piecha, A.; Jakubas, R.; Medycki, W.

Synthesis, crystal structure and phase transitions of a series of

imidazolium iodides. CrystEngComm 2013, 15, 5633−5640.

(27) Hoke, E. T.; Slotcavage, D. J.; Dohner, E. R.; Bowring, A. R.;

(45) Hartono, N. T. P.; Sun, S.; Gelvez-Rueda, M. C.; Pierone, P. J.;

Karunadasa, H. I.; McGehee, M. D. Reversible photo-induced trap

formation in mixed-halide hybrid perovskites for photovoltaics. Chem.

Sci. 2015, 6, 613−617.

Erodici, M. P.; Yoo, J.; Wei, F.; Bawendi, M.; Grozema, F. C.; Sher,

M.-J.; Buonassisi, T.; Correa-Baena, J.-P. The effect of structural

dimensionality on carrier mobility in lead-halide perovskites. J. Mater.

Chem. A 2019, 7, 23949−23957.

(28) Xing, J.; Wang, Q.; Dong, Q.; Yuan, Y.; Fang, Y.; Huang, J.

Ultrafast ion migration in hybrid perovskite polycrystalline thin films

under light and suppression in single crystals. Phys. Chem. Chem. Phys.

(46) Daub, M.; Haber, C.; Hillebrecht, H. Synthesis, Crystal

Structures, Optical Properties, and Phase Transitions of the Layered

Guanidinium-Based Hybrid Perovskites [C(NH ) ] MI ; M = Sn, Pb.

29) Mosconi, E.; De Angelis, F. Mobile Ions in Organohalide

016, 18, 30484−30490.

Eur. J. Inorg. Chem. 2017, 2017, 1120 −1126.

Perovskites: Interplay of Electronic Structure and Dynamics. ACS

Energy Lett. 2016, 1, 182−188.

(47) Szafranski, M.; Katrusiak, A. Phase transitions in the layered

structure of diguanidinium tetraiodoplumbate. Phys. Rev. B: Condens.

Matter Mater. Phys. 2000, 61, 1026 −1035.

(30) Yuan, Y.; Huang, J. Ion Migration in Organometal Trihalide

Perovskite and Its Impact on Photovoltaic Efficiency and Stability.

(48) Wilke, M.; Casati, N. Insight into the Mechanochemical

Acc. Chem. Res. 2016, 49, 286−293.

Synthesis and Structural Evolution of Hybrid Organic −Inorganic

(31) Ellis, C. L. C.; Smith, E.; Javaid, H.; Berns, G.; Venkataraman,

Guanidinium Lead(II) Iodides. Chem. - Eur. J. 2018, 24, 17701−

D. Chapter 6 - Ion Migration in Hybrid Perovskites: Evolving

Understanding of a Dynamic Phenomenon. In Perovskite Photo-

voltaics; Thomas, S., Thankappan, A., Eds.; Academic Press, 2018; pp

17711.

50) Brown, M. E.; Glass, B. D. Pharmaceutical applications of the

49) Brown, M. E. The Prout-Tompkins rate equation in solid-state

kinetics. Thermochim. Acta 1997, 300 , 93 −106.

Prout −Tompkins rate equation. Int. J. Pharm. 1999 , 190 , 129 −137.

51) Prout, E. G.; Tompkins, F. C. The thermal decomposition of

potassium permanganate. Trans. Faraday Soc. 1944, 40, 488 −498.

52) Giorgi, G.; Fujisawa, J.-I.; Segawa, H.; Yamashita, K. Organic −

63−196.

32) Brennan, M. C.; Ruth, A.; Kamat, P. V.; Kuno, M.

Photoinduced Anion Segregation in Mixed Halide Perovskites. Trends

Chem. 2020, 2, 282−301.

(33) Elmelund, T.; Seger, B.; Kuno, M.; Kamat, P. V. How Interplay

between Photo and Thermal Activation Dictates Halide Ion

Segregation in Mixed Halide Perovskites. ACS Energy Lett. 2020, 5,

Inorganic Hybrid Lead Iodide Perovskite Featuring Zero Dipole

Moment Guanidinium Cations: A Theoretical Analysis. J. Phys. Chem.

C 2015, 119, 4694−4701.

(

6−63.

34) Sun, Q.; Fassl, P.; Becker-Koch, D.; Bausch, A.; Rivkin, B.; Bai,

(53) Pandey, S. K.; Manogaran, D.; Manogaran, S.; Schaefer, H. F.

S.; Hopkinson, P. E.; Snaith, H. J.; Vaynzof, Y. Role of Microstructure

in Oxygen Induced Photodegradation of Methylammonium Lead

Triiodide Perovskite Films. Adv. Energy Mater. 2017, 7, 1700977.

Quantification of Hydrogen Bond Strength Based on Interaction

Coordinates: A New Approach. J. Phys. Chem. A 2017, 121, 6090−

6103.

(35) Bryant, D.; Aristidou, N.; Pont, S.; Sanchez-Molina, I.;

Chotchunangatchaval, T.; Wheeler, S.; Durrant, J. R.; Haque, S. A.

Light and oxygen induced degradation limits the operational stability

of methylammonium lead triiodide perovskite solar cells. Energy

Environ. Sci. 2016, 9, 1655−1660.

(54) Fu, Y. P.; Hautzinger, M. P.; Luo, Z. Y.; Wang, F. F.; Pan, D. X.;

Aristov, M. M.; Guzei, I. A.; Pan, A. L.; Zhu, X. Y.; Jin, S.

Incorporating Large A Cations into Lead Iodide Perovskite Cages:

Relaxed Goldschmidt Tolerance Factor and Impact on Exciton-

Phonon Interaction. ACS Cent. Sci. 2019, 5, 1377−1386.

(55) Jodlowski, A. D.; Roldan-Carmona, C.; Grancini, G.; Salado,

M.; Ralaiarisoa, M.; Ahmad, S.; Koch, N.; Camacho, L.; de Miguel,

G.; Nazeeruddin, M. K. Large guanidinium cation mixed with

(

36) Ito, S.; Tanaka, S.; Vahlman, H.; Nishino, H.; Manabe, K.;

Lund, P. Carbon-Double-Bond-Free Printed Solar Cells from TiO ₂ /

CH NH PbI /CuSCN/Au: Structural Control and Photoaging

Effects. ChemPhysChem 2014, 15, 1194−1200.

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

methylammonium in lead iodide perovskites for 19% efficient solar

Article

cells. Nat. Energy 2017, 2, 972−979.

(56) Yun, J. S.; Seidel, J.; Kim, J.; Soufiani, A.; Huang, S.; Lau, J.;

Jeon, N.; Seok, S.; Green, M. A.; Ho-Baillie, A. Critical Role of Grain

Boundaries for Ion Migration in Formamidinium and Methylammo-

nium Lead Halide Perovskite Solar Cells. Adv. Energy Mater. 2016, 6,

(

600330.

57) Pont, S.; Bryant, D.; Lin, C.-T.; Aristidou, N.; Wheeler, S.; Ma,

X.; Godin, R.; Haque, S. A.; Durrant, J. R. Tuning CH NH Pb-

I Br ) perovskite oxygen stability in thin films and solar cells. J.

1 ‑x x 3

Mater. Chem. A 2017, 5, 9553−9560.

(

58) Kim, G. Y.; Senocrate, A.; Yang, T.-Y.; Gregori, G.; Gratzel, M.;

Maier, J. Large tunable photoeffect on ion conduction in halide

perovskites and implications for photodecomposition. Nat. Mater.

59) Juarez-Perez, E. J.; Hawash, Z.; Raga, S. R.; Ono, L. K.; Qi, Y.

018, 17, 445−449.

Thermal degradation of CH NH PbI perovskite into NH and CH I

gases observed by coupled thermogravimetry −mass spectrometry

analysis. Energy Environ. Sci. 2016, 9, 3406−3410.

(60) Motti, S. G.; Meggiolaro, D.; Barker, A. J.; Mosconi, E.; Perini,

C. A. R.; Ball, J. M.; Gandini, M.; Kim, M.; De Angelis, F.; Petrozza,

A. Controlling competing photochemical reactions stabilizes perov-

skite solar cells. Nat. Photonics 2019, 13, 532−539.

(61) Wang, S.; Jiang, Y.; Juarez-Perez, E. J.; Ono, L. K.; Qi, Y.

Accelerated degradation of methylammonium lead iodide perovskites

induced by exposure to iodine vapour. Nat. Energy 2017, 2, 16195.

(

62) Fu, F.; Pisoni, S.; Jeangros, Q.; Sastre-Pellicer, J.; Kawecki, M.;

Paracchino, A.; Moser, T.; Werner, J.; Andres, C.; Duchene, L.; Fiala,

P.; Rawlence, M.; Nicolay, S.; Ballif, C.; Tiwari, A. N.; Buecheler, S. I

vapor-induced degradation of formamidinium lead iodide based

perovskite solar cells under heat −light soaking conditions. Energy

Environ. Sci. 2019, 12, 3074−3088.