Controlling PbI2Stoichiometry during Synthesis to Improve the Performance of Perovskite Photovoltaics

Performance of Perovskite Photovoltaics

Article

Controlling PbI Stoichiometry during Synthesis to Improve the

Konstantinos Tsevas, Joel A. Smith, Vikas Kumar, Cornelia Rodenburg, Mihalis Fakis,

Abd. Rashid bin Mohd Yuso ﬀ , Maria Vasilopoulou, David G. Lidzey, Mohammad K. Nazeeruddin,

and Alan D. F. Dunbar*

Cite This: Chem. Mater. 2021, 33, 554 −566

Read Online

ACCESS

Metrics & More

Article Recommendations

sı Supporting Information

ABSTRACT: Over the past decade, remarkable progress has advanced the ﬁeld of perovskite

solar cells to the forefront of thin ﬁlm solar technologies. The stoichiometry of the perovskite

material is of paramount importance as it determines the optoelectronic properties of the

absorber and hence the device performance. However, little published work has focused on the

synthesis of fully stoichiometric precursor materials of high purity and at high yield. Here, we

report a low-cost, energy-eﬃcient, and solvent-free synthesis of the lead iodide precursor by

planetary ball milling. With our synthetic approach, we produce low-oxygen, single or multiple

polytypic phase PbI with tunable stoichiometry. We determine the stoichiometry and the

polytypes present in our synthesized materials and further compare them to commercially available materials, using X-ray diﬀraction,

X-ray photoelectron spectroscopy, and Rutherford backscattering spectroscopy. Both the stoichiometric PbI we synthesized and a

substoichiometric commercially available PbI (where the iodide content is below the optimum Pb:I atomic ratio of 1:2) were used

to grow methylammonium lead iodide microcrystals (which corrects the iodide content). Perovskite solar cells were then produced

using stoichiometric and substoichiometric PbI mixed with an equimolar amount of methylammonium iodide and compared to

devices produced from re-dissolved microcrystals. The photoactive perovskite layer deposition was processed in air, enabled by the

use of a single low-toxicity solvent (dimethyl sulfoxide) combined with vacuum-assisted solvent evaporation. We ﬁnd that the device

performance is strongly dependent upon the stoichiometry of the lead iodide precursor, reaching champion eﬃciencies over 17%,

with no obvious correlation with its polytypic phases. This work highlights the critical role of PbI stoichiometry in hybrid

perovskites as well as demonstrating synthesis methods and perovskite layer fabrication protocols suitable for low-cost solar energy

harvesting.

INTRODUCTION

PSC devices. In parallel to the challenge of stability is that of

scalability, where the possibility for air processing oﬀers the

potentially lowest-cost approach to perovskite fabrication. This

necessitates the development of fabrication methods for

materials and devices under ambient conditions using less

toxic solvents.

■

Immense academic and commercial research eﬀorts are being

undertaken into perovskite solar cells (PSCs), tapping into

their potential as the next low-cost solar power technology.

The improvement in photovoltaic performance for the most

ubiquitous hybrid organic−inorganic halide perovskite absorb-

er (methylammonium lead iodide, CH NH PbI ) has been

prodigious, with power conversion eﬃciencies (PCEs)

increasing from 3.8% to 21.2% in a decade. Mixtures of

organic and inorganic cations, predominantly methylammo-

Whilst various researchers have demonstrated the impor-

9−11

tance of perovskite phase purity and stoichiometry,

few

and

have covered the role of the precursor materials PbI

5,12−14

methylammonium iodide (CH NH I),

in particular, the

stoichiometry of PbI and its polytypic behavior. Previous

nium, formamidinium, and cesium (MA /FA /Cs ) with

various halogen anions, such as chloride, bromide, and iodide

reports have described more stoichiometrically correct

CH NH PbI thin ﬁlms, synthesized using CH NH PbI single

−

(

Cl / Br /I ) have been used to improve the performance and

crystals, in comparison with the nominally equimolar

stability of PSC devices. However, the long-term stability and

durability of PSCs have remained barriers to their commerci-

alization. These eﬀects might be related to either deﬁciency or

15,16

intermixed powders of PbI with CH NH I.

Received: September 1, 2020

Revised: December 22, 2020

Published: January 14, 2021

excess of lead iodide (PbI ) in the synthesized perovskite and a

tolerance factor closer to unity that corresponds to the cubic

−7

crystal structure.

This degradation occurs regardless of the

device architecture utilized (p-i-n or n-i-p). Controlling the

stoichiometry of the precursor materials used may help to

reduce the presence of impurities and improve the stability of

2021 American Chemical Society

Chem. Mater. 2021, 33, 554−566

Chemistry of Materials

crystal formations, as shown in Figure 1 a. In these needles,

Article

PbI is an intrinsic semiconducting material with a wide

the solid-state reaction of the same reactants. The HT

energy band gap (E ) of 2.3−2.4 eV. Its crystalline structure

synthesis of PbI results in the formation of continuous layers

is hexagonally close-packed with the space group P3

consisting of three alternating layers of lead and iodide atoms

m1,

of hexagonal PbI₂platelets decorated with needle-shaped

(

[

I−Pb−I) within the basic unit cell, which is a near-octahedron

4−

PbI ] . For the most stable at room temperature (RT) 2H

polytypic phase, the unit cell’s dimensions are a = b = 0.4557

nm and c = 0.6979 nm. However, the stacking sequence of

the (I−Pb−I) layers can vary signiﬁcantly due to the weak van

der Waals interactions between those sequential layers. This

results in the crystallochemical phenomenon of polytypism, in

which diﬀerent polytypes with variable c-axis dimensions are

formed. These variations in the c axis also result in minor

19−21

changes to a and b axes.

Polytype formation can also be

dependent on the synthetic process used, for example, co-

solvent, modiﬁed gel,₂₆sublimation, epitaxial, or vertical

Bridgman−Stockbarger synthesis. During the synthesis, the

formation of multiple polytypic phases may occur. Moreover,

reversible and irreversible transitions between diﬀerent

polytypic phases might be responsible for the coexistence of

1,27,28

several energy states in PbI₂,

in a way that is analogous

Figure 1. Scanning electron microscopy (SEM) images of lead iodide

powders as synthesized by (a) hydrothermal (HT) and (b)

mechanochemical (MC) routes or commercially available from (c)

Sigma-Aldrich (SA) and (d) Tokyo Chemical Industries (TCI); scale

bars represent 10 μm. Background and hexagonal shaped formations

to the polytypism in silicon carbide.

Despite the

identiﬁcation of more than 40 polytypic phases of PbI over

the last 60 years, primarily using X-ray diﬀraction techniques,

the exact mechanism of their formation and diversity remains

poorly understood. Further details regarding the theory of

polytypism are provided in Supporting Information Section 1.

Here, we report the use of a planetary ball milling (PBM)

correspond to PbI , while needle-shaped crystal formations

correspond to oxyiodide and lead iodide hydroxide compounds in

image (a).

process to synthesize PbI , which permits control of the

stoichiometry and the degree of polytypism. PBM utilizes high

kinetic energy, inelastic impacts between stainless steel spheres

surrounded by the reactant powders (lead(II) nitrate and

potassium iodide), and the walls of the ball milling jar. This

approach provides a controllable and energy eﬃcient way

incorporation of oxygen was detected by energy-dispersive

spectroscopy (EDS) point analysis, as can be seen in Table S1.

We believe that autoionization of water molecules occurs

during the HT synthesis of PbI (Figure 1a and Figure S1a),

caused by the strong Coulomb interactions between the ions

(

∼31.3 watts per hour per gram) to produce moderate

quantities at the lab scale (≥40 grams per run and yield of ∼96

0.5%) of PbI , without using solvents. By avoiding the use of

−

(

Pb , K , I , and NO ) and water molecules. This may result

in the formation of a mixture of lead oxyiodide and lead iodide

hydroxide compounds, with general chemical formulas Pb-

solvent during the PbI synthesis, less waste is produced and a

33−35

(I₁₋_xO ) and Pb(I (OH) ) , respectively.

x 2

1−y

y 2

potential route for impurities is removed. We then tested the

PSC performance of devices fabricated in air using a single-step

In contrast, the PBM synthesis of PbI causes extensive

deformation and fragmentation of the platelets, with reduced

particle size (average size of <4 μm), as illustrated in Figure 1b

and Figure S1b . No oxygen was detected in the MC powder, as

determined by EDS and shown in Table S1 , presumably due to

the absence of water during the synthesis. Also, for the

30,31

deposition method with vacuum-assisted evaporation

the solvent (dimethyl sulfoxide) and without the use of any

additional anti-solvent rinsing step. This fabrication method

permits the formation of high-quality perovskite thin ﬁlms

without using the highly toxic and widely used solvent

commercially available PbI powders from Sigma-Aldrich with

dimethyl formamide (DMF). PSC devices fabricated using

equimolar amounts of substoichiometric PbI and CH NH I

a purity of 99.999% (SA) and Tokyo Chemical Industries with

a purity of 99.99% (TCI), the particles (as shown in Figures 1

precursors exhibited inferior performance due to the presence

^o^f^uⁿ^r^e^a^c^t^e^d^P^b^I2 in the resultant perovskite thin ﬁlm.

dc and, respectively) are considerably larger than the MC

sample shown in Figure 1b.

Meanwhile, the polytypic phase of PbI had no profound

PbI powders synthesized by the HT method and the PBM

eﬀect on the device performance. Our results indicate that the

precise control over the stoichiometry of the precursor

materials may be a route toward more eﬃcient perovskite

solar cells.

method under three diﬀerent conditions with respect to the

rotation and duration (MC1: (200 rpm for 2 h), MC2: (400

rpm for 1 h), and MC3: (400 rpm for 4 h)) were characterized

by X-ray diﬀraction (XRD), alongside the two commercially

available PbI powders from Sigma-Aldrich (SA) and from

RESULTS AND DISCUSSION

Hydrothermal and Mechanochemical Synthesis of

■

Tokyo Chemical Industries (TCI), as shown in Figure 2.

Diﬀerences in the XRD patterns were observed between the

mechanochemical (MC1 to MC3), HT, and commercial SA

and TCI materials. The observed XRD peaks were qualitatively

labeled based on the corresponding Miller index and the

associated polytypic phases. Also, the crystallographic planes

and their corresponding Miller indices for each polytypic phase

with respect to the diﬀraction 2θ angle are summarized in

PbI . In order to vary both stoichiometry and polytypism, two

methods of synthesis of PbI have been considered in this

work: (1) the conventional hydrothermal (HT) reaction that

requires the dissolution of reactants (lead(II) nitrate and

potassium iodide) in water and (2) the water-free

mechanochemical (MC) process using PBM, which allows

Chem. Mater. 2021, 33, 554−566

Article

ratio and the amount of oxygen contained in PbI . The

stoichiometric variations between the lead and iodide

measured and the amount of oxygen are summarized in

Table 1. It should be noted that these results are averaged

Table 1. Stoichiometric Ratios of Diﬀerent Lead Iodide

Samples and Their Oxygen Content Percentage from XPS

Measurements

lead iodide

sample

synthesis conditions/

sample information

amount of

oxygen (%)

ratio of Pb:I

MC3

MC2

MC1

mechanochemical (400

rpm, 4 h)

mechanochemical (400

rpm, 1 h)

mechanochemical (200

rpm, 2 h)

1:2.1 (± 0.1)

2.91 (± 0.01)

3.97 (± 0.16)

5.21 (± 0.76)

1:2 (± 0.1)

1:2.1 (± 0.1)

TCI

hydrothermal (in water)

Tokyo Chemical

Industries, 99.99%

1:2.1 (± 0.1)

1:2.3 (± 0.1)

7.15 (± 0.21)

4.46 (± 0.14)

Figure 2. X-ray diﬀraction (XRD) patterns of commercial (SA and

TCI), hydrothermal (HT), and mechanochemical (made under

diﬀerent conditions: MC1, MC2, and MC3) PbI₂powders.

Normalization was applied with respect to the intensity of the peak

Sigma-Aldrich, 99.999%

1:1.70 (± 0.1)

4.90 (± 0.02)

The samples include commercial (SA and TCI), hydrothermal

(HT), and mechanochemically synthesized under diﬀerent conditions

(MC1 to MC3) PbI powders. The data was determined by two high-

resolution XPS scans (a survey scan followed by high-resolution scans

that corresponds to the (001) crystallographic plane for 2H PbI . All

peaks are labeled based on the polytypic phases 2H, 4H, 14H, and

0H, which can contribute to scattering at that angle as outlined in

were used for all samples).

Supporting Information Table S2.

values and do not reveal potential localized deviations as a

result of polytypism, which may be present in inhomogeneous

Table S2 . Based on the International Centre for Diﬀraction

Database, XRD peaks of crystallographic planes with Miller

indices (001), (011), and (003) correspond to the 2H

materials. X-ray-induced photolysis of PbI may also occur

during the XPS measurement process; therefore, only the

general trends in the Pb:I ratio are considered here. Further

discussion of inhomogeneity in XPS data and X-ray-induced

photolysis of PbI₂is available in Supporting Information

Section 2.

polytypic phase of PbI with code 01-071-6147, which is the

most stable PbI phase at RT, while all the other XRD peaks

correspond mostly to a mixture of 4H (04-017-4470), 14H

(

04-007-2277), and 20H (04-007-2278) polytypic phases, with

characteristic c-axis unit cell dimensions 1.3962, 4.8853, and

.979 nm, respectively. Further quantitative analysis of the

In the SA material, the lead to iodide ratio was less than the

expected 1:2. Meanwhile, for the HT, MC, and TCI samples,

the ratio varied signiﬁcantly, with values for HT, MC1, MC3,

and TCI above the stoichiometric 1:2 ratio and for MC2 being

almost stoichiometric, as shown in Table 1. This variation in

MC samples is attributed to the diﬀerent conditions used

during the PBM process. Rotation speed was found to be the

polytypic phases present was not possible due to complications

associated with overlapping of the XRD peaks that correspond

to the 2H with those of 4H, 14H, and 20H, respectively. There

may also be undetected minor fractions of PbI , which have

crystallized in a rhombohedral structure.

The MC1 sample (200 rpm for 2 h) did not show strong

diﬀraction at the secondary XRD peak positions, indicating

that it has mostly crystallized in a single polytypic phase,

closely resembling that of the SA sample. Mechanochemically

prepared PbI₂powders MC2 and MC3 show enhanced

intensity at the secondary XRD peaks as a result of the

increased rotation speed (400 rpm) and duration (4 h) of the

PBM process. The formation of more than one polytypic phase

in the mechanochemical powders (MC2 and MC3) makes

them more comparable with the TCI PbI₂sample. The

presence of additional polytypic phases in MC2 and MC3 is

attributed to the increased energy given via the inelastic

collisions occurring at the higher rotation speed between the

stainless steel spheres surrounded by the reactant powders

during the mechanochemical synthesis.

dominant factor in the process followed by duration.

Understanding the inﬂuence of using overstoichiometric

(TCI, MC1, and MC3) materials as compared to almost

stoichiometric MC2 PbI is of great future research interest.

The presence of excess iodide within the perovskite grains or at

grain boundaries caused by the use of overstoichiometric

PbI₂could have detrimental eﬀects on PSC performance.

These may vary with environmental conditions (such as

38,39

oxygen/inert gas/moisture),

or diﬀusion of excess iodide

could lead to secondary chemical reactions and degradation of

the other incorporated layers in a functioning solar cell device

40−42

(hole/electron transporting layers or metallic contacts).

Rutherford backscattering spectroscopy (RBS) character-

ization was also conducted to determine the lead to iodide

ratio of the commercial and synthesized PbI samples. RBS

Stoichiometry of the PbI Samples. Three diﬀerent

methods ((1) X-ray photoelectron spectroscopy (XPS), (2)

Rutherford backscattering (RBS), and (3) X-ray diﬀraction

XRD) studies of perovskite microcrystals made from PbI₂)

have been utilized to determine the stoichiometry of the PbI₂

samples.

XPS characterization of the commercial and synthesized

spectra with normalized yield of lead to iodide in each PbI₂

sample are shown in Figure S3. Moreover, the extracted

stoichiometric ratio between the lead and iodide measured are

summarized in Table S3. Uncertainties are given for statistical

error, while a 4% systematic error due to uncertainty in the

stopping power of the beam of ions was considered. In the SA

material, the lead to iodide ratio was 1:1.9 (± 0.008), while in

the mechanochemical MC2, it was 1:2.002 (± 0.008).

(

PbI samples was conducted to determine the lead to iodide

Chem. Mater. 2021, 33, 554−566

Article

Figure 3. XRD patterns of CH NH PbI microcrystals synthesized at diﬀerent temperatures (asterisks represent the XRD peaks of the lead iodide):

(

a) mechanochemical MC2 and CH NH I, with a 1:1 molar ratio, (b) mechanochemical MC2 and CH NH I, with a 1:1 molar ratio plus 30% by

weight of CH NH I, (c) commercial SA and CH NH I, with a 1:1 molar ratio, and (d) commercial SA and CH NH I, with a 1:1 molar ratio plus

0% by weight of CH NH I. Normalization of XRD patterns was with respect to the peak that corresponds to plane with Miller indices (110) for

3 3

tetragonal CH NH PbI in all patterns, except for (c) commercial SA and CH NH I, with a 1:1 molar ratio at 20 °C, which is normalized with

respect to the peak of the 2H lead iodide with Miller indices (001).

However, for the TCI, MC1, and MC3 samples, the ratio

shows an excess of iodide to be present.

respectively. Each reaction was permitted to proceed for 24 h

at a constant temperature of 20, 50, 70, or 90 °C.

The diﬀerent PbI materials prepared, shown to have a range

Considering the experiments where equimolar amounts of

reactants were used to prepare the CH NH PbI microcrystals,

of Pb:I ratios and a number of polytypic phases as previously

discussed, were used to synthesize CH NH PbI microcrystals.

less unreacted PbI was detected when MC2 was used, as

The system of alcohol/PbI /CH NH I, similar to Acik et al.,

compared to SA. The XRD peaks that correspond to the

was used to prepare the perovskite microcrystals which were

subsequently characterized by X-ray diﬀraction (XRD). Two

groups of perovskite microcrystals were prepared as a function

of PbI used and the corresponding amount of CH NH I at a

unreacted PbI are identiﬁed with asterisks in Figure 3 and are

more prevalent in the SA sample (Figure 3c). The percentage

of the unreacted PbI after each reaction at each temperature

was found by integrating the area under the primary XRD

peaks and comparing the ratio of (001) PbI to CH NH PbI

range of temperatures (20, 50, 70, and 90 °C). In the ﬁrst

group, the almost stoichiometric MC2 was reacted either with

with Miller indices (002)/(110) that corresponds to the

perovskite with a tetragonal crystal structure. The results are

summarized in Table S4. In contrast, when an excess of

CH NH I was introduced (30% more CH NH I by weight) at

(

a) an equimolar amount of CH NH I or (b) an excess of

3 3

CH NH I (30% more CH NH I by weight), as shown in

Figures 3a and3b, respectively. While, in the second group,

substoichiometric SA was reacted either with (c) an equimolar

amount of CH NH I or (d) an excess of CH NH I (30% more

the most, only traces of or zero unreacted PbI was detected in

all the CH NH PbI microcrystals synthesized, regardless of

CH NH I by weight), as shown in Figures 3c and 3d,

the PbI source (either MC2 or SA), as shown in Figures 33db

Chem. Mater. 2021, 33, 554−566

Chemistry of Materials

sample, as shown in Figure 4 . The large experimental error in

Article

and , respectively. We believe that methylammonium cations

iodide was detected in the SA sample as compared to the MC2

−

(

CH NH ) and iodide anions (I ) from CH NH I, primarily

3 3 3 3

react with the substoichiometric PbI , fulﬁlling the formation

of CH NH PbI and resulting in some unreacted PbI (yellow

in color), as shown in Figure S4 during ﬁltering of the

CH NH PbI microcrystals. Consequently, using substoichio-

metric PbI , partially depleted of iodide, during the fabrication

of perovskite thin ﬁlms, results in unreacted PbI . This could

provide a possible explanation for the improved device

performance noted in the literature when hydroiodic acid

(

HI) is used as an additive during the CH NH PbI formation,

3 3 3

a correction of the iodide content when the substoichiometric

PbI is used, thus reducing undercoordinated Pb.

The same system of alcohol/PbI /CH NH I was used to

synthesize CH NH PbI microcrystals with a controlled excess

of CH NH I by weight to determine quantitatively the amount

of iodide missing from PbI . These microcrystal samples were

obtained by reacting either substoichiometric SA with 10%,

5%, 20%, and 25% excess of CH NH I by weight or the

3 3

almost stoichiometrically correct MC2 with 2%, 4%, and 6%

excess CH NH I by weight for 24 h at a constant temperature

Figure 4. Number of iodide atoms in lead iodide (commercial (SA)

and mechanochemical (MC2)) as function of the characterization

technique (XPS correspond to white columns, CH NH PbI

microcrystals correspond to red columns, and RBS correspond to

blue columns).

of 90 °C. The synthesized CH NH PbI microcrystals were

characterized by XRD, as shown in Figure S5 . By gradually

increasing the amount of excess CH NH I, the amount of

unreacted PbI was reduced, as indicated in the XRD patterns

shown in Figure S5a,c. We found that when 20% excess

CH NH I was reacted with the substoichiometric SA, the weak

the XPS data could be related to X-ray-induced photolysis of

XRD peak that corresponds to unreacted PbI , as shown in

PbI (further discussion is available in Supporting Information

Figure S5b, had almost entirely disappeared. This indicates

that the required excess of CH NH I should be slightly greater

Section 2) and the penetration depth of photons that

correspond to X-rays (typically 2−5 nm). Meanwhile, in

than 20% to achieve the complete elimination of unreacted

PbI during the synthesis of CH NH PbI microcrystals when

RBS with a typical He ion penetration depth of ∼100 nm,

the acquired signal might be aﬀected by the number of

polytypic phases present and their structural density (2H ρstr.

the substoichiometric SA material is used. This value of 20%

excess of CH NH I by weight is consistent with the value of

−

−3

= 6.214 g·cm , 4H ρstr. = 5.452 g·cm , 14H ρstr. = 6.098 g·

−

−3

0% excess PbI in the PbI /CH NH I mixture used for the

cm , and 20H ρstr. = 6.1 g·cm ), as provided by the

International Centre for Diﬀraction Database.

formation of perovskite as reported by Nazeeruddin et al. In

both cases, the iodine deﬁciency is corrected, by iodine

originating from either the 20% excess of CH NH I or the 10%

Impact of Stoichiometry of PbI on Device Perform-

ance. Fabrication of PSC devices with the n-i-p (regular)

excess of PbI used, to achieve the required stoichiometry in

architecture of glass/ITO/SnO /CH NH PbI /spiro-OMe-

the ﬁnal product of CH NH PbI . Meanwhile, when 4% excess

TAD/Au was performed, where nanoparticle tin oxide

CH NH I was reacted with the almost stoichiometric MC2,

(SnO ) served as the electron transporting layer (ETL) and

this resulted in an even weaker PbI XRD peak, as shown in

spiro-OMeTAD as the hole transporting layer (HTL), to

determine how the variable stoichiometry aﬀects the device

performance. Comparisons were made between devices

Figure S5d . The stoichiometric ratios for SA and MC2 were

calculated based on the need for either the 20% more

CH NH I reacting with substoichiometric SA or the 4% more

fabricated using two varieties of PbI : SA, which is depleted

CH NH I reacting with the almost stoichiometric MC2 to

in iodide, and mechanochemical MC2, which is almost

stoichiometrically correct. PSC devices were made by

dissolving either the SA or MC2 powders with an equimolar

amount of CH NH I in dimethyl sulfoxide (DMSO) to

eliminate any signal that corresponds to PbI in the XRD data.

We found after analysis that the initial (Pb:I) ratio for SA is

approximately 1:1.80 and that for MC2 is approximately

:1.96, as presented in Supporting Information Section 3. Also,

produce photoactive thin ﬁlms with thicknesses ∼550 nm (±

50 nm) and ∼380 nm (± 70 nm) as shown in cross-sectional

SEM images, Figures S6a and S6b, respectively, after correcting

for the 45° sample angle. A second set of PSC devices was also

made by re-dissolving CH NH PbI microcrystals in DMSO.

based on all the above XRD analysis of CH NH PbI

microcrystals, a theory about the formation of unreacted

PbI was developed, as presented in Supporting Information

Section 4, accompanied by several examples, as shown in Table

S5.

The same two PbI materials (either SA or MC2) were used in

The results from the XPS and RBS characterization

techniques used to determine the lead to iodide ratio in the

the reaction with a 30% excess of CH NH I by weight. Minor

morphological diﬀerences of CH NH PbI microcrystals

commercial (SA) and mechanochemical (MC2) PbI samples

synthesized at a temperature of 90 °C, using the two varieties

are considered alongside the ratios derived from the synthesis

of CH NH PbI microcrystals with controlled excess

of PbI (either SA or MC2), are shown in Figure S7a,b. The

formation of the microcrystals made with an excess of

CH NH I and their subsequent XRD characterization (further

analysis is provided in Supporting Information Section 3). The

results from all three techniques show consistent trends; less

CH NH I helped to correct for any deﬁciency of iodide

within PbI . The PSC devices, with the photoactive layer made

by re-dissolving the CH NH PbI microcrystals (made using

Chem. Mater. 2021, 33, 554−566

Article

Figure 5. Variation of FF, J , PCE, and V_O_Cof the solar cell devices as a function of the synthesis route/materials used. Their structure was glass/

ITO/SnO /CH NH PbI /spiro-OMeTAD/Au. The devices using the CH NH PbI microcrystal synthesis involved a stoichiometry correction

step for iodide by the addition of CH NH I. The devices using equimolar amounts of PbI and CH NH I powders did not include this step.

Table 2. Performance Parameters of Solar Cell Devices with the Structure Glass ITO/SnO /CH NH PbI /spiro-OMeTAD/

Au, as a Function of the Starting Materials Used and the Synthesized CH NH PbI Microcrystals

FF (%)

J_S_C(mA/cm²)

20.98 (21.61)

V_O_C(V)

PCE (%)

dissolution of CH NH PbI microcrystals in CH NH PbI microcrystals (synthesized

68.19 (73.66)

1.03 (1.04)

14.69 (16.55)

DMSO

using MC2)

CH NH PbI microcrystals (synthesized

72.08 (75.82)

20.87 (21.65)

1.06 (1.06)

16.02 (17.40)

using SA)

dissolution of equimolar PbI /CH NH I in MC2/CH NH I (equimolar)

71.75 (75.29)

34.18 (41.01)

21.09 (21.60)

13.06 (14.96)

1.04 (1.05)

1.02 (1.02)

15.77 (17.07)

4.58 (6.26)

DMSO

SA/CH NH I (equimolar)

The photovoltaic parameters of the average values from 20 devices in bold and champion devices in parentheses.

either SA or MC2) in DMSO, had photoactive thin ﬁlm

thicknesses of ∼460 nm (± 50 nm) and ∼440 nm (± 40 nm),

as shown in cross-sectional SEM images, Figures S6c and S6d,

respectively, after correcting for the 45° sample angle. The

deposition of the photoactive thin ﬁlms was performed in

ambient air, and the black/brown perovskite phase formation

was accomplished using a vacuum-assisted method that

promotes the extraction of solvent in a single step (without

photovoltaic parameters for all the devices are summarized in

Table 2.

Notably, the photovoltaic parameters are signiﬁcantly

enhanced in PSC devices made from either CH NH PbI

microcrystals, where the iodide content was corrected by the

processing regardless of the number of polytypic phases in PbI₂

used (SA or MC2), or when equimolar amounts of almost

stoichiometric MC2 with CH NH I were used. In comparison,

the device performance parameters (FF, J , and PCE) of PSC

46,47

an anti-solvent rinsing process),

as shown in Figure S8a

devices fabricated using equimolar amounts of the SA with

CH NH I were notably inferior. This reduction in device

(

before annealing) and Figure S8b (after annealing),

respectively.

performance is attributed to the presence of a signiﬁcant

The average and champion photovoltaic parameters: ﬁll

amount of unreacted PbI between, covering or possibly within

8−50

factor (FF), short-circuit current density (J ), open-circuit

the grains of CH NH PbI .

Also, the characteristic

^v^o^l^t^a^g^e⁽^VOC), and power conversion eﬃciency (PCE), as

current density−voltage (J−V) curves in forward (0 → V_O_C)

measured during a reverse (V → 0) scan with a scan step of

and reverse (V → 0) scans of PSC devices were compared as

−

0 mV s , are illustrated in Figure 5. Additionally, the

a function of the photoactive layer fabricated using either

Chem. Mater. 2021, 33, 554−566

equimolar amounts of SA/CH NH I (blue curves).

Article

Figure 6. (a) IPCE and (b) electron current of perovskite solar cells and electron-only devices, respectively, using CH NH PbI microcrystals

(

using MC2) (black curves), CH NH PbI microcrystals (using SA) (orange curves), equimolar amounts of MC2/CH NH I (red curves), and

Figure 7. (a) Visible-infrared (Vis-IR) absorbance, (b) steady-state photoluminescence (PL) spectra, and (c, d) TRPL decay curves of perovskite

thin ﬁlms synthesized using CH NH PbI microcrystals (using MC2) (black curves), CH NH PbI microcrystals (using SA) (orange curves),

equimolar amounts of MC2/CH NH I (red curves), and equimolar amounts of SA/CH NH I (blue curves) (from ﬁve diﬀerent measured points

only in (a)). Absorbance due to the presence of unreacted lead iodide is indicated with black arrow in (a).

CH NH PbI microcrystals or equimolar amounts of diﬀerent

for the forward (0 → V_O_C) and reverse (V_O_C→ 0) scans for all

PbI with CH NH I, as shown in Figure S9. In all the plots,

devices, and that of the higher performing conditions was

worst in the case for the equimolar of reactants used. All the J−

V curves (either in reverse or forward scans) of PSC devices

fabricated using equimolar amounts of substoichiometric PbI₂

black curves correspond to the average from all J−V curves

recorded and the error bars represent the deviation across all

the devices. Hysteresis was observed between the J−V curves

Chem. Mater. 2021, 33, 554−566

Chemistry of Materials

SA) with CH NH I were notably inferior in performance, as

Article

(

(Figure 7b). The PL intensities varied signiﬁcantly and were

much lower for the perovskite ﬁlms fabricated from the

stoichiometric precursors or for those where the stoichiometry

has been adjusted through processing (microcrystals). While it

is diﬃcult to disentangle the eﬀects of interface recombination

and carrier extraction, this diﬀerence could be due to either

fewer defects in the samples fabricated using microcrystals or

eﬃcient blocking in samples based on the presence of

unreacted PbI . However, in light of the PSC device data,

we attribute this diﬀerence to suppressed electron injection

into the ETL in the SA/CH NH I sample leading to increased

bulk radiative recombination. Upon perovskite photoexcitation

shown in Figure S9d.

Less hysteresis was observed between forward and reverse

scans of PSC devices fabricated using equimolar amounts of

the substoichiometric PbI₂(SA) with CH NH I. The

hysteresis could be related to ion migration originating from

unreacted PbI likely present between the grain boundaries and

also possibly within the grains themselves, aﬀecting the

diﬀusion lengths of charge carriers. It may be masked in the

SA sample because of the considerably lower current observed.

We believe that the amount of iodide present in the

substoichiometric SA is insuﬃcient to form the exact

composition of CH NH PbI , but this can be corrected by

in the SnO /perovskite samples, electrons are transferred from

using the above method for the preparation of microcrystals to

adjust the stoichiometry.

its conduction band (lying at about −3.9 eV) to the lower

conduction band of SnO (lying about −4.1 eV), hence

In order to shed light on the diﬀerences in the device

performance, the internal photon-to-electron-conversion eﬃ-

ciencies (IPCE) of the fabricated PSC devices were measured.

The results are shown in Figure 6a and clearly demonstrate

that the photon-to-electron conversion process is more

eﬃcient for the devices based on perovskite absorbers

synthesized from the close to perfectly stoichiometric

precursors. Furthermore, electron-only devices with the

structure FTO/SnO /CH NH PbI /PCBM/Ag were fabri-

suppressing the PL signal. In the PbI -rich samples, this

electron transfer pathway is blocked due to the insulating

properties of PbI₂.

This was further corroborated by the transient PL (TRPL)

measurements taken on the same samples (detection at 770

nm) (Figure 7c,d and Table S6), which indicated faster

electron transfer from the excited perovskite toward SnO in

the equimolar MC2/CH NH I sample compared to the SA/

CH NH I counterpart. We note that the small diﬀerence in the

cated and their J−V characteristics were measured under

dark conditions, as shown in Figure 6b. These results indicate

that the electron injection is inferior in the device when the

substoichiometric PbI₂with an equimolar amount of

CH NH I were used. This can be explained by the presence

decay kinetics between equimolar and microcrystal samples

may be due to the diﬀerences in morphology.

The above results indicate that the perovskite samples

derived from substoichiometric precursors are enriched in

insulating PbI , which is formed between the grains or

of a large excess of unreacted PbI within the perovskite

combined within consecutive layers of (CH NH PbI /PbI )

synthesized from the substoichiometric precursor material.

likely formed within the grains of CH NH PbI . The

3 3 3

The presence of unreacted PbI in the perovskite ﬁlms

formation of CH NH PbI ferroelastic domains, regardless of

3 3 3

obtained when using equimolar amounts of SA/CH NH I was

the stoichiometry in PbI , has also been noted in the

5−57

also conﬁrmed by the short wavelength shoulder present in the

literature

and is conﬁrmed here by AFM (Figure S11).

visible infrared (Vis-IR) absorption spectra shown in Figure 7a.

This phenomenon may also explain the spatial variability of the

The experimentally calculated optical band gap values (E_o_b_g

for PbI (E = 2.53 eV (± 0.01 eV)) and CH NH PbI (E

)

obg

J_S_Cwithin individual CH NH PbI grains, as reported by Kutes

et al. The measured roughness of all fabricated CH NH PbI

obg

1.48 eV), as determined by Tauc plots, support the

thin ﬁlms using the vacuum-assisted method was less than 16

nm, as measured by AFM. Also, an increased concentration of

larger pinholes formed in the CH NH PbI thin ﬁlms

hypothesis that unreacted PbI might also form within the

perovskite grains because they show a deviation from the

optical band gap values that correspond to neat PbI (E

fabricated from the CH NH PbI microcrystals regardless of

obg

3 3 3

.36 eV) and CH NH PbI (E = 1.51 eV), as presented in

PbI used (either substoichiometric SA or almost stoichio-

obg

Figures S10a and S10b, respectively.

Also, the presence of unreacted PbI in the perovskite ﬁlms

obtained when equimolar amounts of SA and CH NH I were

used results in a binary mixture of CH NH PbI and PbI

metric MC2) as observed in LV-SEM images, Figure S12a,b.

This might suggest that, with further process development,

such as by further modifying the 1.13 M concentration of the

perovskite ink using the CH NH PbI microcrystals, even

(

with CH NH PbI as the major phase and PbI as the minor

better photovoltaic performance is achievable by this route.

For all the LV-SEM images in Figure S12, grain size

distribution calculations were conducted and are summarized

in Figure S13 . Speciﬁcally, the majority of grains on

CH NH PbI thin ﬁlms when microcrystals were employed

phase) with E_o_b_g= 1.48 eV and a similar binary mixture (with

PbI as the major phase and CH NH PbI as the minor phase)

with E_o_b_g= 2.53 eV (± 0.01 eV). Consequently, when

equimolar amounts of SA and CH NH I are used, the

formation of those binary mixtures, which are neither pure

CH NH PbI nor pure PbI , may lead to materials with

absorption coeﬃcients diﬀerent than their absolute values.

and fabricated using the almost stoichiometric MC2 PbI had

an average size of 110 ± 10 nm. Meanwhile, the majority of

grains had an average size of 90 ± 10 nm when microcrystals

Wang et al., and Cui et al. reported that diﬀerent precursor

ratios of PbI :CH NH I aﬀect the optoelectronic properties of

were prepared using the substoichiometric SA PbI . Also,

equimolar amounts of the almost stoichiometric MC 2 PbI₂/

the resulting CH NH PbI formed, such as the Fermi level,

CH NH I resulted in thin ﬁlms with the maximum average

charge carrier mobilities, carrier concentrations, and recombi-

grain size of 175 ± 25 nm. Meanwhile, for thin ﬁlms of

perovskite made using equimolar amounts of substoichio-

metric SA PbI /CH NH I, the average grain size was 130 ± 10

3,54

nation rates.

Further details of how this can result in

unreacted PbI in the resulting perovskite ﬁlm are given in

Supporting Information Section 4. Steady-state photolumines-

cence (PL) spectra of the diﬀerent perovskite samples

nm. The smaller grain sizes observed when using the

microcrystal processing route may be due to slightly diﬀerent

growth mechanisms occurring in the perovskite thin ﬁlms

deposited on SnO /ITO substrates were also measured

Chem. Mater. 2021, 33, 554−566

Chemistry of Materials

Article

made from CH NH PbI microcrystals and equimolar

amounts of reactants, which requires further investigation.

Oxygen reduction inside the ball milling jar was conducted by

displacing the air with argon gas (purity: 99%). This oxygen reduction

was only applied for the mechanochemical sample MC3 (400 rpm, 4

h), leading to oxygen reduction by ∼36 (± 10)%, in comparison with

the mechanochemical samples MC1/MC2 synthesized without the

introduction of argon gas. We expect that further oxygen reduction in

CONCLUSIONS

■

In conclusion, planetary ball milling is a low-cost, energy-

eﬃcient pathway to produce PbI₂with reduced oxygen

content, whereby it is possible to control both the

the mechanochemical PbI powders might be achievable by careful

drying the reactants used and by displacing the air with ultra-pure

argon gas (>99.99%) inside the ball milling jar. The yellow colored

^P^b^I2 powder was separated from the stainless steel spheres and

washed ﬁve times with 70−80 mL (each time) of deionized water

during vacuum ﬁltration (Whatman ﬁlter paper grade 5 with a pore

size of 2.5 μm) to remove any potassium nitrate byproduct. Finally,

stoichiometry and polytype of the resulting PbI by varying

the mechanochemical parameters of synthesis (revolutions per

minute and duration). The absence of water during the PbI₂

production via planetary ball milling allows for a signiﬁcant

reduction in the oxygen content present as determined by XPS

PbI was dried in a vacuum oven at 50 °C for 24 hours, before storage

analysis. Although the polytypism of the PbI precursor had no

in dark conditions and further characterization. The yield of the

produced PbI₂(MC2) was deﬁned by the relative number of

product/reactant moles between PbI and Pb(NO ) (considering the

profound eﬀect on the solar cell performance, its stoichiometry

was proven to exhibit serious implications for the device

performances. PSC devices with a photoactive layer manufac-

3 2

ideal molecular weight for PbI to be 461.01 g/mol), resulting ((40.32

−

−1

g/461.01 g·mol )/(30 g/(331.21 g·mol )) × 100% ≈ 96.5%.

CH NH PbI Microcrystal Synthesis. CH NH PbI micro-

tured using substoichiometric PbI showed reduced perform-

ance due to the large excess of PbI in the perovskite ﬁlm,

crystals were synthesized by reacting 0.7 g of the almost

unless an excess of CH NH I was used via the fabrication of

stoichiometric MC2 or substoichiometric SA PbI (used as received

CH NH PbI microcrystals. If PbI₂was stoichiometrically

and stored in dark conditions) with an equimolar amount of

CH NH I (0.2414 g) or an equimolar amount plus 30% excess by

correct, an excess of CH NH I was not necessary. Using

stoichiometrically corrected PbI , in combination with the

weight of CH NH I (0.2414 g + 0.0724 g = 0.3138 g) in 15 mL of 1-

3 3

single-step vacuum-assisted method, which is free of any anti-

solvent rinse step and does not use other toxic solvents,

promotes the formation of CH NH PbI with close to no

pentanol at variable temperatures (20, 50, 70, and 90 °C) for 24 h (in

an oil bath). In this process, 1-pentanol, which is a polar protic non-

toxic alcohol, was used to selectively dissolve only the methyl-

detectable unreacted PbI . This processing route can help

ammonium iodide but not the PbI

the CH NH PbI microcrystals. The black powder of CH NH PbI

powder and the ﬁnal product of

enable a low-cost and air-processable fabrication route for PSC

devices with repeatably high performance. This work highlights

the importance of proper material control and synthesis

procedures to achieve enhanced quality of PSC devices.

microcrystals was collected by vacuum ﬁltration (Whatman ﬁlter

paper grade 5 with a pore size of 2.5 μm) and washed with 5 mL of

chlorobenzene (Alfa Aesar) during the ﬁltration step followed by

drying in a vacuum oven for 30 min at room temperature before

further characterization.

EXPERIMENTAL DETAILS

Methylammonium Iodide Synthesis. Methylammonium iodide

■

(

^A^f^t^e^r^c^o^m^p^l^e^tⁱ^oⁿ^o^f^t^h^e^c^h^e^mⁱ^c^a^l^r^e^a^c^tⁱ^oⁿ^,^t^h^e¹^-^p^eⁿ^t^aⁿ^o^l^/^P^b^I2^/

(<2 s). Recrystallization of the excess CH

pentanol during the vacuum ﬁltration process (to separate out the

CH NH PbI microcrystals) was prevented by the use of a large

volume of 1-pentanol, which reduces the probability of the dissolved

I solution must be kept warm and ﬁltrated as fast as possible

CH NH I) was synthesized by adding dropwise 38 mL of hydroiodic

NH I dissolved in the 1-

acid (57% w/w in water, Alfa Aesar) in 30 mL of methylamine (40%

w/w in water, Sigma-Aldrich) in a 250 mL round-bottom ﬂask at 0 °C

(ice bath) and constant stirring (400 rpm) for 2 h. The round-bottom

−

ﬂask was transferred into an oil bath at a temperature of 90 °C to

evaporate water for 6−7 h. The resulting yellow/white precipitate was

further puriﬁed by recrystallizing two times from pure anhydrous

ethanol. The white/transparent crystallized ﬂakes were collected by

vacuum ﬁltration (Whatman ﬁlter paper grade 5 with a pore size of

ions (CH NH and I ) interacting. Unreacted CH NH and I ions

remain dissolved in warm 1-pentanol and are therefore removed from

the microcrystals upon ﬁltration.

Dry Dimethyl Sulfoxide Preparation. Dry dimethyl sulfoxide

(DMSO, Sigma-Aldrich) was prepared by intermixing 150 mL of as-

received dimethyl sulfoxide (DMSO) with a purity of ≥99.5% and 20

2.5 μm) and dried in a vacuum oven at 50 °C for 24 h before storage

in dark conditions. The same methylammonium iodide was used for

both the synthesis of the CH NH PbI microcrystals and perovskite

g of magnesium sulfate (MgSO , Fisher Chemical) in a 250 mL

round-bottom ﬂask. The round-bottom ﬂask was capped and left in a

fume hood at room temperature for at least 2 weeks before extracting

crystal clear dry DMSO.

Perovskite Solution Preparations. Perovskite solutions were

prepared either by dissolving 0.4 g of the almost stoichiometric MC 2

solutions.

Hydrothermal PbI Synthesis. Hydrothermal PbI was synthe-

sized by dissolving 1 g of lead(II) nitrate (Pb(NO ) , Sigma-Aldrich)

with a purity of ≥99% and 1.01 g of potassium iodide (KI, Honeywell

Fluka) with a purity of ≥99% (both precursors as received) in a 1 L

volume round-bottom ﬂask ﬁlled with 1 L of boiled (100 °C) ultra-

pure deionized water (Sigma-Aldrich) with a resistivity of 18.2 MΩ.

The round-bottom ﬂask was covered with towel made of ﬁber glass

and left to cool down slowly to room temperature for 24 h. An excess

of potassium iodide was used, to complete the reaction of the

Pb(NO ) based on the chemical reaction: Pb(NO ) + 2KI → PbI

or substoichiometric SA PbI

CH NH I (0.1379 g) in 0.620 mL of dry DMSO with a ﬁnal

concentration of 2.8 M or by dissolving 0.4 g of the CH NH PbI

microcrystals synthesized using the almost stoichiometric MC 2 or

substoichiometric SA PbI in 0.571 mL of dry DMSO with a ﬁnal

directly with an equimolar amount of

concentration of 1.13 M. Additionally, all perovskite solutions were

kept on a hot plate at a temperature of 100 °C for 10 min to ensure

complete dissolution of the reactants in DMSO before being ﬁltered

through a PTFE ﬁlter (pore size of 0.45 μm) and kept under constant

stirring (∼300 rpm) and temperature (65−70 °C) in a custom-made

aluminum block (see Figure S14 ) before deposition. The block

helped to reduce condensation of DMSO on the inside of the vial

walls, thereby preventing premature perovskite crystallization. In all

our experiments with CH NH PbI microcrystals, the solution

2KNO . The golden colored crystalline ﬂakes of PbI were collected

after vacuum ﬁltration (Whatman ﬁlter paper grade 5 with a pore size

of 2.5 μm) and dried in a vacuum oven at 50 °C for 24 h before

further characterization.

Mechanochemical PbI₂Synthesis. Mechanochemical PbI₂

powder was synthesized using a planetary ball milling machine (PM

00 Retsch) at variable revolutions per minute and time duration.

Thirty grams of Pb(NO ) with a purity of ≥99%, 30.07 g of KI with a

concentrations were calculated based on the assumption of perfect

purity of ≥99%, and 160 stainless steel spheres with a diameter of 6

CH NH PbI microcrystals with a molecular weight of 619.98 g·

−1

mm were mixed in a ball milling jar with a capacity of 500 mL.

mol , as deﬁned by the atomic mass units of each element in

Chem. Mater. 2021, 33, 554−566

Chemistry of Materials

Article

CH NH PbI and prepared for 24 h after their synthesis. Also, in all

our experiments with either commercial SA or mechanochemical

MC2, solution concentrations were calculated based on the false

measurements of powders were taken using a zero-background holder

made of monocrystalline silicon.

SEM Images and EDS Analysis. Scanning electron microscopy

(SEM) images and energy-dispersive spectroscopy (point or full

image) elemental analysis were obtained using a JSM-6010LA (JEOL)

at accelerating voltages 12 and 15 kV, respectively. Low-voltage

scanning electron microscopy images were obtained using a FEI

Helios NanoLab G3 UC SEM at an accelerating voltage of 2 kV using

a through-lens detector (TLD) at a working distance of 4.2 mm and

with a beam current of 13 pA.

assumption of stoichiometrically perfect PbI (Pb:I = 1:2, atom ratio),

−

with a molecular weight of 461.01 g·mol .

Device Fabrication. Glass substrates pre-patterned with six ITO

indium tin oxide) pixels (20 Ω/square, Ossila) were ﬁrst sonicated

(

for 10 min in a solution made of 1 mL of detergent (Hellmanex III)

and 250 mL of boiled deionized water (100 °C). Any residual

detergent was subsequently washed oﬀ by rinsing the substrates three

times with 100 mL (each time) of deionized water. The glass/ITO

substrates were further sonicated two more times for 10 min (each

time) in 250 mL of acetone and then 250 mL of isopropanol. The

cleaned substrates were then treated with oxygen plasma for 10 min to

remove organic contaminants. Clean glass/ITO substrates were

transferred into a cleanroom (class 1000) with a constant temperature

of 20 °C and humidity of 30% for deposition of the electron

transporting layer and perovskite thin ﬁlm. A colloidal dispersion of

Visible−Infrared (Vis-IR) Spectroscopy Measurements.

Absorption spectra of all CH NH PbI thin ﬁlms were obtained

using a spectrometer (USB2000 + UV−VIS-ES) equipped with a

deuterium halogen light source (UV−VIS-NIR_DT-MINI-2-GS),

both from Ocean Optics.

Time-Resolved Photoluminescence (TRPL) Measurements.

The time-resolved photoluminescence measurements were performed

using the time-correlated single photon counting (TCSPC) method

tin(IV) oxide (SnO ) in deionized water (15%, Alfa Aesar, diluted to

%) was spin coated at 3000 rpm for 30 s followed by removal of

(FluoTime 200, Picoquant) under magic-angle conditions. Excitation

SnO from part of the substrate using a cotton bud, re-exposing the

was performed by a pulsed diode laser at 470 nm, and the

instrument’s response function (IRF) was ∼80 ps. Multi-exponential

functions convoluted with the IRF were used for the ﬁtting taking into

∼

4−5 mm ITO electrode. The coated substrates were then annealed

at 150 °C for 30 min and cooled down to 20 °C, without any further

treatment. Fifty microliters of the perovskite solutions made of

either CH NH PbI microcrystals in DMSO with a concentration of

account that the χ factor should be smaller than 1.1.

X-ray Photoelectron Spectroscopy (XPS) Measurements.

Each powder was mounted for analysis by pushing the powder into

indium foil. The indium foil was then mounted directly onto the

sample holder using double-sided conducting carbon tape. The

analyses were carried out using a Kratos Supra instrument with a

monochromatic aluminum source and an energy of 1486.69 eV.

Survey scans were collected between binding energies from 1200 to 0

eV at 1 eV intervals with an acquisition time of 10 min (per point)

from two analysis points per sample. High-resolution scans were also

collected for the I 3d, O 1s, C 1s, and Pb 4f core levels at 0.1 eV

intervals with an acquisition time of 5 min for each point. The data

collected were calibrated in intensity using a transmission function

characteristic of the instrument to make the values instrument-

independent. The data was then quantiﬁed using the CasaXPS

software, determining the transmission functions as provided by the

National Physical Laboratory (NPL). Binding energy calibrations for

all survey or high-resolution scans were conducted with respect to the

Pb 4f 7/2 core level at 138.0 eV. No selected option as an escape

depth correction was applied at each survey or high-resolution scan

.13 M or solutions of (almost stoichiometric MC2 or substoichio-

metric SA) PbI reacted with equimolar amounts of CH NH I in

DMSO with a concentration of 2.8 M was statically dispersed onto

the glass/ITO/SnO samples followed by spin coating at 2500 or

100 rpm for 10 s, respectively. Each substrate was placed in a

custom-made vacuum chamber for 120 s and a pressure of ∼0.02

mbar, as shown in Figure S15. Then, each substrate was annealed on a

hot plate at 100 °C for 30 s. All glass/ITO/SnO /CH NH PbI

samples were then transferred inside a nitrogen-ﬁlled glove box for the

synthesis and deposition of the doped hole transporting layer of

,2′,7,7′-tetrakis[N,N-di(4-methoxyphenyl)amino]-9,9′-spirobiﬂuor-

ene (spiro-OMeTAD, ≥99.5%, Ossila). An 86.6 mg solution of spiro-

OMeTAD powder was dissolved in 1 mL of chlorobenzene. The

solution was then doped with lithium bis(triﬂuoromethanesulfonyl)-

imide (Li-TFSI, ≥99%, Sigma-Aldrich), 4-tert-butylpyridine (TBP

6.6%, Sigma-Aldrich), and FK209 (FK 209 Co(II) PF6, Greatcell).

The quantity of dopants used in 1 mL of the spiro-OMeTAD/

chlorobenzene solution was as follows: 20 μL of Li-TFSI (500 mg·

mL in acetonitrile), 34 μL of TBP, and 11 μL of FK209 (300 mg·

−1

mL in acetonitrile). The solution was ﬁltered with a PTFE ﬁlter

pore size of 0.2 μm), and then 50 μL was spin coated on each sample

(

before starting any further quantiﬁcation analysis. For the

dynamically at 4000 rpm for 30 s. Devices were left overnight in dry

air to allow for spiro-OMeTAD to be oxidized. Finally, 80 nm of gold

was deposited in an Edwards thermal evaporator, with a deposition

rate of 0.1 Å s for the ﬁrst 2 nm and then 1 Å s for the remaining

thickness. All un-encapsulated PSCs devices were immediately

characterized electrically.

compositional analysis of the core level spectra of I 3d 5/2, O 1s,

C 1s, and Pb 4f 7/2, the Scoﬁeld relative sensitivity factors of 16, 2.52,

1, and 13.7 were used, respectively. The binding energy regions used

for analysis were kept constant with respect to each element at each

survey or high-resolution scan. Also, for each binding energy region, a

Shirley type background function was used.

−

−1

Photocurrent Density−Voltage (J−V) Measurements. Photo-

current density−voltage (J−V) characteristics were measured under

an AM 1.5G light (1000 W m ) produced by a xenon lamp

Rutherford Backscattering Spectroscopy (RBS) Measure-

−

ments. Two commercial and four synthesized PbI samples in the

(

Newport solar simulator). The light intensity was calibrated using a

form of pellets with a diameter of ∼13 mm and thickness of ∼1.5 mm

were prepared using a press under an applied mass of 9 tons. For the

RBS measurements, a 1.7 MV Tandetron RBS linear type tandem ion

beam accelerator was used, which is located in the Laboratory of Ion

silicon reference cell (Newport). Each solar cell device was mounted

and covered with a six-pixelated cell shadow mask, with an aperture

area of 0.0256 cm per cell. Cells were scanned from −0.2 to 1.2 V

−

and then back to −0.2 V at a scan step of 0.01 V·s using a Keithley

Beam Physics at the ETH Zurich facilities in Switzerland. PbI pellets

37 source measure unit.

Powder X-ray Diﬀraction Measurements. X-ray diﬀraction

were bombarded with accelerated ions of He at 2 MeV, while a PIN

diode detector was placed at an angle of 168°, with respect to the

beam of incident ions. The resulted data were plotted using the

RUMP simulation software.

patterns of all (commercial (used as received and stored in dark

conditions), hydrothermal, and mechanochemical) lead iodide

powders and CH NH PbI microcrystals were obtained using a D2

Phaser (Bruker) diﬀractometer under monochromatic Cu Kα

radiation (λ = 1.54184 Å) and step size 2θ = 0.020273° over the

Atomic Force Microscopy (AFM) Measurements. Atomic

force microscopy of the CH NH PbI thin ﬁlms was performed using

a Dimension icon with ScanAsyst AFM (Bruker) with a cantilever

θ range from 9° to 52°. X-ray diﬀraction patterns of CH NH PbI

3 3 3

microcrystals with diﬀerent amounts of excess CH NH I were

obtained at double acquisition time for higher accuracy. Also, all

consisting of a silicon tip on a silicon nitride lever (Bruker) (f = 70

−

kHz, k = 0.4 N·m ).

Chem. Mater. 2021, 33, 554−566

Chemistry of Materials

The Supporting Information is available free of charge at

Article

ASSOCIATED CONTENT

sı Supporting Information

https://pubs.acs.org/doi/10.1021/acs.chemmater.0c03517.

Notes

■

The authors declare the following competing ﬁnancial

interest(s): D. G. Lidzey is a co-director of the company

Ossila Ltd., that retail materials and equipment used in

perovskite photovoltaic device research and development.

(

Section 1) Discussion of polytypism in PbI ; SEM−

EDS data; photograph of HT and MC PbI samples,

ACKNOWLEDGMENTS

completed ﬁlms, sample preparation, and the vacuum-

assisted drying setup; XRD peak positons for diﬀerent

■

This work was funded by the UK Engineering and Physical

Sciences Research Council (EPSRC) via a Ph.D. studentship

for K.T. at the University of Sheﬃeld from the Centre for

Doctoral Training in New and Sustainable Photovoltaics, grant

code EP/L01551X/1, in which J.A.S. also acknowledges for

funding. A.D.F.D. wishes to acknowledge support from the

EPSRC through Supergen Solar Challenge, grant EP/

M025020/1. C.R. and V.K. thank the EPSRC for funding

under grant EP/N008065/1. We appreciate the assistance

from N. Reeves McLaren from the University of Sheﬃeld for

providing the D2 Phaser (Bruker) X-ray diﬀractometer. We

also appreciate the assistance from D. Hammond, who carried

out the X-ray photoelectron spectroscopy measurements in the

Sheﬃeld Surface Analysis Centre. K.T. wishes to acknowledge

the access provided through the RADIATE GATE scheme for

RBS measurements, proposal number 19001741-ST. We also

appreciate the assistance from Hans-Arno Synal, Christof

Vockenhuber, and colleagues from the Laboratory of Ion Beam

Physics at the ETH Zurich facilities in Switzerland, who carried

PbI polytypes; (Section 2) XPS discussion and results;

RBS data and calculated stoichiometric ratios; percent-

age unreacted PbI as determined from XRD; photo-

graph of residual PbI in perovskite microcrystal powder;

microcrystal XRD data; (Section 3) description of

method to determine stoichiometry from crystallization

XRD; (Section 4) iodine deﬁciency discussion including

example calculations; SEM images: cross-sectional SEM

of SA and MC ﬁlms, SA and MC microcrystals, ﬁlm

coverage for SA, MC, SA(microcrystal), and MC-

(

microcrystal) ﬁlms plus grain size distributions; device

I(V) plots; Tauc plots for SA and MC samples; TRPL

ﬁtting parameters; and AFM images of SA and MC

samples (PDF)

AUTHOR INFORMATION

Corresponding Author

Alan D. F. Dunbar − Department of Chemical and Biological

Engineering, University of She ﬃ eld, Sheﬃeld S1 3JD, U.K.;

orcid.org/0000-0002-2313-4234; Email: a.dunbar@

she ﬃ eld.ac.uk

■

out the RBS measurements and analysis of PbI samples. M.

Vasilopoulou acknowledges support of this work by the project

“

Development of Materials and Devices for Industrial, Health,

Environmental and Cultural Applications” (MIS 5002567),

which is implemented under the “Action for the Strategic

Development on the Research and Technological Sector”,

funded by the Operational Program “Competitiveness,

Entrepreneurship and Innovation” (NSRF 2014-2020) and

co-ﬁnanced by Greece and the European Union (European

Regional Development Fund).

Authors

Konstantinos Tsevas − Department of Chemical and

Biological Engineering, University of Sheﬃeld, Sheﬃeld S1

JD, U.K.

Joel A. Smith − Department of Physics & Astronomy,

University of She ﬃ eld, Sheﬃeld S3 7RH, U.K.; orcid.org/

000-0001-6889-4408

Vikas Kumar − Department of Materials Science and

Engineering, University of She ﬃ eld, Sheﬃeld S1 3JD, U.K.;

orcid.org/0000-0002-1034-2915

REFERENCES

■

1) Kojima, A.; Teshima, K.; Shirai, Y.; Miyasaka, T. Organometal

Halide Perovskites as Visible-Light Sensitizers for Photovoltaic Cells.

J. Am. Chem. Soc. 2009, 131, 6050−6051.

2) Shin, S. S.; Yeom, E. J.; Yang, W. S.; Hur, S.; Kim, M. G.; Im, J.;

Seo, J.; Noh, J. H.; Seok, S. I. Colloidally prepared La-doped BaSnO ₃

electrodes for efficient, photostable perovskite solar cells. Science

2017, 356, 167−171.

(3) Wang, R.; Mujahid, M.; Duan, Y.; Wang, Z.-K.; Xue, J.; Yang, Y.

A Review of Perovskites Solar Cell Stability. Adv. Funct. Mater. 2019,

Cornelia Rodenburg − Department of Materials Science and

Engineering, University of Sheﬃeld, Sheﬃeld S1 3JD, U.K.

Mihalis Fakis − Department of Physics, University of Patras,

(

6504 Rio Patras, Greece; orcid.org/0000-0003-0801-

029

Abd. Rashid bin Mohd Yusoﬀ − Department of Chemical

Engineering, Pohang University of Science and Technology

4) Huang, F.; Li, M.; Siffalovic, P.; Cao, G.; Tian, J. From scalable

808843.

(

POSTECH), Pohang, Gyeongbuk 37673, Republic of Korea

solution fabrication of perovskite films towards commercialization of

solar cells. Energy Environ. Sci. 2019, 12, 518−549.

5) Jacobsson, T. J.; Correa-Baena, J.-P.; Anaraki, E. H.; Philippe, B.;

Stranks, S. D.; Bouduban, M. E. F.; Tress, W.; Schenk, K.; Teuscher,

J.; Moser, J.-E.; Rensmo, H.; Hagfeldt, A. Unreacted PbI as a Double-

Edged Sword for Enhancing the Performance of Perovskite Solar

Cells. J. Am. Chem. Soc. 2016, 138, 10331−10343.

(6) Saliba, M.; Matsui, T.; Seo, J.-Y.; Domanski, K.; Correa-Baena,

J.-P.; Nazeeruddin, M. K.; Zakeeruddin, S. M.; Tress, W.; Abate, A.;

Hagfeldt, A.; Gratzel, M. Cesium-containing triple cation perovskite

solar cells: improved stability, reproducibility and high efficiency.

Energy Environ. Sci. 2016, 9, 1989.

7) Roose, B.; Dey, K.; Chiang, Y.-H.; Friend, R. H.; Stranks, S. D.

Maria Vasilopoulou − Institute of Nanoscience and

Nanotechnology, National Centre for Scientiﬁc Research

Demokritos, 15341 Agia Paraskevi, Attica, Greece;

orcid.org/0000-0001-8893-1691

David G. Lidzey − Department of Physics & Astronomy,

University of She ﬃ eld, Sheﬃeld S3 7RH, U.K.; orcid.org/

(

000-0002-8558-1160

Mohammad K. Nazeeruddin − Group for Molecular

Engineering of Functional Materials, Institute of Chemical

Sciences and Engineering, École Polytechnique Fédérale de

Lausanne (EPFL), CH-1951 Sion, Switzerland;

orcid.org/0000-0001-5955-4786

Complete contact information is available at:

Critical assessment of the use of excess lead iodide in lead halide

https://pubs.acs.org/10.1021/acs.chemmater.0c03517

perovskite solar cells. J. Phys. Chem. Lett. 2020, 11, 6505−6512.

Chem. Mater. 2021, 33, 554−566

Chemistry of Materials

Article

8) Song, Z.; Watthage, S. C.; Phillips, A. B.; Heben, M. J. Pathways

(27) Hassan, M. A.; Jafar, M. M.; Matuchova, M.; Bulos, B. N. An

experimental evidence of some lead iodide polytypes compatible with

the dielectric functions model. J. Appl. Sci. 2010 , 10 , 3367 −3373.

(28) Salje, E.; Palosz, B.; Wruck, B. In situ observation of the

toward high-performance perovskite solar cells: review of recent

advances in organo-metal halide perovskites for photovoltaic

applications. J. Photonics Energy 2016, 6, No. 022001.

(

9) Chang, J.; Zhu, H.; Li, B.; Isikgor, F. H.; Hao, Y.; Xu, Q.;

polytypic phase transition 2H-12R in PbI : investigations of the

Ouyang, J. Boosting the performance of planar heterojunction

perovskite solar cell by controlling the precursor purity of perovskite

materials. J. Mater. Chem. A 2016, 4, 887−893.

thermodynamic structural and dielectric properties. J. Phys. C: Solid

State Phys. 1987, 20, 4077−4096.

(29) Falk, A. L.; Buckley, B. B.; Calusine, G.; Koehl, W. F.;

Dobrovitski, V. V.; Politi, A.; Zorman, C. A.; Feng, P. X. L.;

Awschalom, D. D. Polytype control of spin qubits in silicon carbide.

Nat. Commun. 2013, 4, 1819.

(10) Yao, J.; Yang, L.; Cai, F.; Yan, Y.; Gurney, R. S.; Liu, D.; Wang,

T. The impacts of PbI ₂ purity on the morphology and device

performance of one-step spray-coated planar heterojunction perov-

skite solar cells. Sustainable Energy Fuels 2018, 2, 436−443.

(

30) Li, X.; Bi, D.; Yi, C.; Dec

M.; Hagfeldt, A.; Gratzel, M. A vacuum flash-assisted solution process

for high-efficiency large-area perovskite solar cells. Science 2016, 353,

8−62.

31) Wu, C.; Zhang, Q.; Liu, Y.; Luo, W.; Guo, X.; Huang, Z.; Ting,

oppet, J.-D.; Luo, J.; Zakeeruddin, S.

(11) Kumar, V.; Barbe, J.; Schmidt, W. L.; Tsevas, K.; Ozkan, B.;

Handley, C. M.; Freeman, C. L.; Sinclair, D. C.; Reaney, I. M.; Tsoi,

W. C.; Dunbar, A.; Rodenburg, C. Stoichiometry-dependent local

(

instability in MAPbI perovskite materials and devices. J. Mater. Chem.

H.; Sun, W.; Zhong, X.; Wei, S.; Wang, S.; Chen, Z.; Xiao, L. The

Dawn of Lead-Free Perovskite Solar Cell: Highly Stable Double

Perovskite Cs AgBiBr Film. Adv. Sci. 2017, 1700759.

A 2018, 6, 23578−23586.

(12) Zhang, Y.; Lv, H.; Cui, C.; Xu, L.; Wang, P.; Wang, H.; Yu, X.;

Xie, J.; Huang, J.; Tang, Z.; Yang, D. Enhanced optoelectronic quality

of perovskite films with excess CH NH I for high-efficiency solar cells

(32) Konstantakou, M.; Perganti, D.; Falaras, P.; Stergiopoulos, T.

Anti-Solvent Crystallization Strategies for Highly Efficient Perovskite

Solar Cells. Crystals 2017, 7, 291.

in ambient air. Nanotechnology 2017, 28, 205401.

13) Mastroianni, S.; Heinz, F. D.; Im, J.-H.; Veurman, W.; Padilla,

M.; Schubert, M. C.; Wurfel, U.; Gratzel, M.; Park, N.-G.; Hinsch, A.

(

33) Bredig, M. A.; Bamberger, C. E.; Richardson, D. M. A mixed-

valency lead oxyiodide. J. Inorg. Nucl. Chem. 1978, 40, 1497 −1500.

34) Berry, F. J.; Jones, C. H. W.; Dombsky, M. An Iodine

Mossbauer Study of Lead(II) Iodide and Lead Oxyiodide. J. Solid

State Chem. 1983, 46, 41−45.

35) Tavakoli, F.; Salavati-Niasari, M.; Mohandes, F. Sonochemical

Analysing the effect of crystal size and structure in highly efficient

CH NH PbI perovskite solar cells by spatially resolved photo- and

electroluminescence imaging. Nanoscale 2015, 7, 19653−19662.

(14) Song, Z.; Watthage, S. C.; Phillips, A. B.; Tompkins, B. L.;

Ellingson, R. J.; Heben, M. J. Impact of Processing Temperature and

Composition on the Formation of Methylammonium Lead Iodide

Perovskites. Chem. Mater. 2015, 27, 4612−4619.

Synthesis and Characterization of Lead Iodide Hydroxide Micro /

nanostructures. Ultrason. Sonochem. 2014, 21 , 234 −241.

(36) Butyagin, P. Y.; Streletskii, A. N. The Kinetics and Energy

(15) Chen, Z.; Dong, Q.; Liu, Y.; Bao, C.; Fang, Y.; Lin, Y.; Tang, S.;

Balance of Mechanochemical Transformations. Phys. Solid State 2005,

7, 856−862.

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

Wang, Q.; Xiao, X.; Bai, Y.; Deng, Y.; Huang, J. Thin single crystal

perovskite solar cells to harvest below-bandgap light absorption. Nat.

Commun. 2017, 8, 1890.

Accelerated degradation of methylammonium lead iodide perovskites

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

(16) Yen, H. J.; Liang, P.-W.; Chueh, C.-C.; Yang, Z.; Jen, A. K.-Y.;

Wang, H.-L. Large Grained Perovskite Solar Cells Derived from

Single-Crystal Perovskite Powders with Enhanced Ambient Stability.

ACS Appl. Mater. Interfaces 2016, 8, 14513−14520.

38) Brenes, R.; Eames, C.; Bulovic, V.; Islam, M. S.; Stranks, S. D.

The Impact of Atmosphere on the Local Luminescence Properties of

Metal Halide Perovskite Grains. Adv. Mater. 2018, 30, 1706208.

(17) Ferreira Da Silva, A.; Veissid, N.; An, C. Y.; Pepe, I.; Barros De

(39) Wang, Q.; Chen, B.; Liu, Y.; Deng, Y.; Bai, Y.; Dong, Q.;

Oliveira, N.; Batista Da Silva, A. V. Optical determination of the direct

bandgap energy of lead iodide crystals. Appl. Phys. Lett. 1996, 69,

Huang, J. Scaling behavior of moisture-induced grain degradation in

polycrystalline hybrid perovskite thin films. Energy Environ. Sci. 2017,

(

930−1932.

(

0, 516−522.

18) Shah, K. S.; Olschner, F.; Moy, L. P.; Bennett, P.; Misra, M.;

40) Kim, S.; Bae, S.; Lee, S.-W.; Cho, K.; Lee, K. D.; Kim, H.; Park,

Zhang, J.; Squillante, M. R.; Lund, J. C. Lead iodide X-ray detection

systems. Nucl. Instrum. Methods Phys. Res., Sect. A 1996, 380, 266−

S.; Kwon, G.; Ahn, S.-W.; Lee, H.-M.; Kang, Y.; Lee, H.-S.; Kim, D.

Relationship between ion migration and interfacial degradation of

CH NH PbI perovskite solar cells under thermal conditions. Sci. Rep.

70.

(

19) Beckmann, P. A. A review of polytypism in lead iodide. Cryst.

Res. Technol. 2010, 45, 455 −460.

20) Trigunayat, G. C. A survey of the phenomenon of polytypism

in crystals. Solid State Ionics 1991, 48, 3−70.

21) Palosz, B.; Steurer, W.; Schulz, H. The structure of PbI

(

017, 7, 1200.

41) Besleaga, C.; Abramiuc, L. E.; Stancu, V.; Tomulescu, A. G.;

M.; Svavarsson, H. G.; Manolescu, A.; Pintilie, I. Iodine Migration and

Sima, M.; Trinca, L.; Plugaru, N.; Pintilie, L.; Nemmes, G. A.; Iliescu,

Degradation of Perovskite Solar Cells Enhanced by Metallic

Electrodes. J. Phys. Chem. Lett. 2016, 7, 5168−5175.

(22) Wei, Q.; Shen, B.; Chen, Y.; Xu, B.; Xia, Y.; Yin, J.; Liu, Z.

polytypes 2H and 4H: a study of the 2H-4H transition. J. Phys.:

Condens. Matter 1990, 2, 5285−5295.

(42) Wong-Stringer, M.; Game, O. S.; Smith, J. A.; Routledge, T. J.;

Alqurashy, B. A.; Freestone, B. G.; Parnell, A. J.; Vaenas, N.; Kumar,

V.; Alawad, M. O. A.; Iraqi, A.; Rodenburg, C.; Lidzey, D. G. High-

Performance Multilayer Encapsulation for Perovskite Photovoltaics.

Adv. Energy Mater. 2018, 8, 1801234.

Large-sized PbI single crystal grown by co-solvent method for visible-

light photo-detector application. Mater. Lett. 2017 , 193 , 101 −104.

(23) Patel, A. R.; Rao, A. V. An improved design to grow larger and

more perfect single crystals in gels. J. Cryst. Growth 1980, 49, 589−

90.

24) Street, R. A.; Ready, S. E.; Lemmi, F.; Shah, K. S.; Bennett, P.;

(43) Acik, M.; Alam, T. M.; Guo, F.; Ren, Y.; Lee, B.; Rosenberg, R.

A.; Mitchell, J. F.; Park, I. K.; Lee, G.; Darling, S. B. Substitutional

Growth of Methylammonium Lead Iodide Perovskites in Alcohols.

Adv. Energy Mater. 2018, 8, 1701726.

(

Dmitriyev, Y. Electronic transport in polycrystalline Pbl films. J. Appl.

Phys. 1999, 86, 2660−2667.

(44) Heo, J. H.; Song, D. H.; Han, H. J.; Kim, S. Y.; Kim, J. H.; Kim,

D.; Shin, H. W.; Ahn, T. K.; Wolf, C.; Lee, T. W.; Im, S. H. Planar

CH NH PbI Perovskite Solar Cells with Constant 17.2% Average

(25) Zheng, Z.; Liu, A.; Wang, S.; Wang, Y.; Li, Z.; Lau, W. M.;

Zhang, L. In situ growth of epitaxial lead iodide films composed of

hexagonal single crystals. J. Mater. Chem. 2005, 15, 4555 −4559.

Power Conversion Efficiency Irrespective of the Scan Rate. Adv.

26) Zhu, X. H.; Wei, Z. R.; Jin, Y. R.; Xiang, A. P. Growth and

Mater. 2015, 27, 3424−3430.

characterization of a PbI single crystal used for gamma ray detectors.

(45) Roldan-Carmona, C.; Gratia, P.; Zimmermann, I.; Grancini, G.;

Cryst. Res. Technol. 2007, 42, 456−459.

Gao, P.; Graetzel, M.; Nazeeruddin, M. K. High efficiency

Chem. Mater. 2021, 33, 554−566

Chemistry of Materials

methylammonium lead triiodide perovskite solar cells: the relevance

Article

of non-stoichiometric precursors. Energy Environ. Sci. 2015, 8, 3550−

46) Jeon, N. J.; Noh, J. H.; Kim, Y. C.; Yang, W. S.; Ryu, S.; Seok, S.

556.

I. Solvent engineering for high-performance inorganic-organic hybrid

perovskite solar cells. Nat. Mater. 2014, 13, 897−903.

(47) Xiao, M.; Huang, F.; Huang, W.; Dkhissi, Y.; Zhu, Y.;

Etheridge, J.; Gray-Weale, A.; Bach, U.; Cheng, Y.-B.; Spiccia, L. A

Fast Deposition-Crystallization Procedure for Highly Efficient Lead

Iodide Perovskite Thin-Film Solar Cells. Angew. Chem., Int. Ed. 2014,

48) Ummadisingu, A.; Grat

3, 9898−9903.

sequential deposition for metal halide perovskite formation. Sci. Adv.

018, 4, No. e1701402.

49) Philippe, B.; Saliba, M.; Correa-Baena, J. P.; Cappel, U. B.;

Turren-Cruz, S. H.; Gratzel, M.; Hagfeldt, A.; Rensmo, H. Chemical

zel, M. Revealing the detailed path of

(

Distribution of Multiple Cation (Rb , Cs , MA , and FA ) Perovskite

Materials by Photoelectron Spectroscopy. Chem. Mater. 2017, 29,

(

589−3596.

50) Du, T.; Burgess, C. H.; Kim, J.; Zhang, J.; Durrant, J. R.;

McLachlan, M. A. Formation, location and beneficial role of PbI in

lead halide perovskite solar cells. Sustainable Energy Fuels 2017, 1,

51) Da Silva Filho, J. M. C.; Ermakov, V. A.; Marques, F. C.

19−126.

Perovskite Thin Film Synthesised from Sputtered Lead Sulphide. Sci.

Rep. 2018, 8, 1563.

53) Wang, Q.; Shao, Y.; Xie, H.; Lyu, L.; Liu, X.; Gao, Y.; Huang, J.

52) Mayerho

Mixing Rules. Appl. Spectrosc. 2020, 74, 1287−1294.

fer, T. G.; Popp, J. Beyond Beer ’s Law: Spectral

Qualifying composition dependent p and n self-doping in

CH NH PbI . Appl. Phys. Lett. 2014, 105, 163508.

(54) Cui, P.; Wei, D.; Ji, J.; Huang, H.; Jia, E.; Dou, S.; Wang, T.;

Wang, W.; Li, M. Planar p −n homojunction perovskite solar cells with

efficiency exceeding 21.3%. Nat. Energy 2019, 4, 150−159.

(55) Rothmann, M. U.; Li, W.; Zhu, Y.; Bach, U.; Spiccia, L.;

Etheridge, J.; Cheng, Y. B. Direct observation of intrinsic twin

domains in tetragonal CH NH PbI . Nat. Commun. 2017, 8, 14547.

(56) Liu, Y.; Collins, L.; Proksch, R.; Kim, S.; Watson, B. R.;

Doughty, B.; Calhoun, T. R.; Ahmadi, M.; Ievlev, A. V.; Jesse, S.;

Retterer, S. T.; Belianinov, A.; Xiao, K.; Huang, J.; Sumpter, B. G.;

Kalinin, S. V.; Hu, B.; Ovchinnikova, O. S. Chemical nature of

ferroelastic twin domains in CH NH PbI perovskite. Nat. Mater.

57) Lan, C.; Zhou, Z.; Wei, R.; Ho, J. C. Two-dimensional

018, 17, 1013−1019.

perovskite materials: From synthesis to energy-related applications.

Mater. Today Energy 2019, 11, 61−82.

(58) Kutes, Y.; Zhou, Y.; Bosse, J. L.; Steffes, J.; Padture, N. P.;

Huey, B. D. Mapping the Photoresponse of CH NH PbI Hybrid

Perovskite Thin Films at the Nanoscale. Nano Lett. 2016, 16, 3434−

441.

59) Jiang, Q.; Zhang, L.; Wang, H.; Yang, X.; Meng, J.; Liu, H.; Yin,

(

Z.; Wu, J.; Zhang, X.; You, J. Enhanced electron extraction using

SnO ₂ for high-efficiency planar-structure HC(NH ) PbI -based

perovskite solar cells. Nat. Energy 2017, 2, 16177.

(60) Saliba, M.; Correa-Baena, J.-P.; Wolff, C. M.; Stolterfoht, M.;

Phung, N.; Albrecht, S.; Neher, D.; Abate, A. How to Make over 20%

Efficient Perovskite Solar Cells in Regular (n −i −p) and Inverted (p −

i −n) Architectures. Chem. Mater. 2018 , 30 , 4193 −4201.

(61) Powell, C. J.; Jablonski, A. Electron effective attenuation lengths

for applications in Auger electron spectroscopy and x-ray photo-

electron spectroscopy. Surf. Interface Anal. 2002, 33, 211−229.