Potassium Bromide Surface Passivation on CsPbI3-xBrx Nanocrystals for Efficient and Stable Pure Red Perovskite Light-Emitting Diodes

Article abstract of DOI:10.1021/jacs.9b11719

All-inorganic lead halide perovskite nanocrystals (NCs) are potential candidates for fabricating high-performance light-emitting diodes (LEDs) owing to their precisely tunable bandgaps, high photoluminescence (PL) efficiency, and excellent color purities. However, the performance of pure red (630-640 nm) all-inorganic perovskite LEDs is still limited by the halide segregation-induced instability of the electroluminescence (EL) of mixed halide CsPbI_3-xBr_x NCs. Herein, we report an effective approach to improving the EL stability of pure red all-inorganic CsPbI_3-xBr_x NC-based LEDs via the passivation of potassium bromide on NCs. By adding potassium oleate to the reaction system, we obtained potassium bromide surface-passivated (KBr-passivated) CsPbI_3-xBr_x NCs with pure red PL emission and a photoluminescence quantum yield (PLQY) exceeding 90%. We determine that most potassium ions present on the surface of NCs bind with bromide ions and thus demonstrate that potassium bromide surface passivation of NCs can both improve the PL stability and inhibit the halide segregation of NCs. Using KBr-passivated CsPbI_3-xBr_x NCs as an emitting layer, we fabricated stable and pure red perovskite LEDs with emission at 637 nm, showing a maximum brightness of 2671 cd m^-2, maximum external quantum efficiency of 3.55%, and good EL stability. The proposed KBr-passivated NC strategy will open a new avenue for fabricating efficient, stable, and tunable pure color perovskite NC LEDs.

Full text of DOI:10.1021/jacs.9b11719

Subscriber access provided by UNIVERSITY OF TECHNOLOGY SYDNEY
Article
Potassium-Bromide Surface Passivation on CsPbI3-xBrx Nanocrystals
for Efficient and Stable Pure Red Perovskite Light Emitting Diodes
Jun-Nan Yang, Yang Song, Ji-Song Yao, Kun-Hua Wang, Jing-Jing Wang, Bai-Sheng
Zhu, Ming-Ming Yao, Sami Ur Rahman, Yi-Feng Lan, Feng-Jia Fan, and Hong-Bin Yao
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.9b11719 • Publication Date (Web): 06 Jan 2020
Downloaded from pubs.acs.org on January 6, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
Potassium-Bromide Surface Passivation on CsPbI Br
3-x
x
Nanocrystals for Efficient and Stable Pure Red Perovskite Light
Emitting Diodes
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
,2
1
1,2
1,2
1,2
Jun-Nan Yang, Yang Song, Ji-Song Yao, Kun-Hua Wang, Jing-Jing Wang, Bai-Sheng
2
1,2
1
1
1,3,4
and Hong-Bin Yao*1,2
Zhu, Ming-Ming Yao, Sami Ur Rahman, Yi-Feng Lan, Feng-Jia Fan,
1Hefei National Laboratory for Physical Sciences at the Microscale, University of Science and Technology of
China, Hefei, Anhui 230026, China
2Department of Applied Chemistry, Hefei Science Center of Chinese Academy of Sciences, University of
Science and Technology of China, Hefei, Anhui 230026, China
3CAS Key Laboratory of Microscale Magnetic Resonance, University of Science and Technology of China, He
fei, Anhui 230026, China
4Synergetic Innovation Center of Quantum Information and Quantum Physics, University of Science and Tech
nology of China, Hefei, Anhui 230026, China
*
Corresponding author Email: yhb@ustc.edu.cn
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
All-inorganic lead halide perovskite nanocrystals (NCs) are potential candidates for fabricating
high-performance light-emitting diodes (LEDs) owing to their precise tunable bandgaps, high
photoluminescence (PL) efficiency and excellent color purities. However, the performance of pure
red (630 ~ 640 nm) all-inorganic perovskite LEDs is still limited by the halide segregation-induced
instability of the electroluminescence (EL) of mixed halide CsPbI Br NCs. Herein, we report an
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
-x
x
effective approach to improving the EL stability of pure red all-inorganic CsPbI Br NC-based
3
-x
x
LEDs via the passivation of potassium bromide on NCs. By adding potassium oleate to the reaction
system, we obtained potassium bromide surface passivated (K-Br passivated) CsPbI Br NCs with
3
-x
x
pure red PL emission and a photoluminescence quantum yield (PLQY) exceeding 90%. We
determine that most potassium ions present on the surface of NCs bind with bromide ions and thus
demonstrate that potassium bromide surface passivation of NCs can both improve the PL stability
and inhibit the halide segregation of NCs. Using K-Br-passivated CsPbI Br NCs as an emitting
3
-x
x
layer, we fabricated stable and pure red perovskite LEDs with emission at 637 nm, showing a
-2
maximum brightness of 2671 cd m , maximum external quantum efficiency of 3.55%, and good EL
stability. The proposed K-Br-passivated NC strategy will open a new avenue for fabricating
efficient, stable and tunable pure color perovskite NC LEDs.
Keywords: metal halide perovskite, nanocrystals, light emitting diode, segregation, stability
2
ACS Paragon Plus Environment

Page 3 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
INTRODUCTION
In the past few years, incredible advances have been made in improving the performance of
all-inorganic lead halide perovskite-based light-emitting diodes (LEDs).1, 2 The excellent
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
properties of all-inorganic lead halide perovskites, such as high PLQY, narrow full-width at
4
5
half-maximum (FWHM), and ease of bandgap tuning, make them next-generation materials
for LED displays. While the external quantum efficiency (EQE) exceeds 20% for green
6
emitters based on CsPbBr /MABr quasi-core/shell structures and red emitters based on
3
anion-exchanged
CsPbI Br
x
NCs7
and
near-infrared
emitters
based
on
3
-x
8
2
,2′-[oxybis(ethylenoxy)]-diethylamine-treated FAPbI perovskite layers have been reported,
3
the performance of pure red emitters (630 ~ 640 nm) needs to be improved to meet
9
application requirements. This is because the red channel for a display centered at the 630
nm emission peak wavelength (Rec. 2020) is a key target to realize the practical application of
LEDs for displays.10
Mixed halide CsPbI Br NCs emitting in the range of 630 ~ 640 nm are promising
3-x
x
11
candidate materials for pure red LEDs. However, under photoirradiation or with an applied
bias, the emission peak of mixed halide perovskites will present a blueshift or redshift due to
the segregation of halide ions; that is, mixed halide phases are divided into two separated
halide phases.12-14 Typically, by applying a bias voltage on CsPbI Br NC-based LEDs, the
3
-x
x
mixed phases are divided into higher-bandgap bromide-rich domains and lower-bandgap
iodide-rich domains on account of the different ionic mobilities of bromide ions and iodide
1
1
ions. After that, subsequent light emission occurs from these iodide-rich domains due to
their favorable band offsets.15 Consequently, the EL emission peak exhibits a redshift,
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
resulting in the low stability of the EL spectra of pure red CsPbI Br NC-based LEDs. To
3
-x
x
date, the Junji Kido group demonstrated the highest EQEs of pure red perovskite LEDs
7
emitting at 653 nm based on anion-exchanged CsPbI Br NCs, but the retention of red color
3
-x
x
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
purity in the as-fabricated LEDs is still a critical issue.9
To minimize photoinduced halide segregation in mixed halide perovskites, several
strategies have recently been proposed and investigated in solar cells,16 including (i)
incorporating other cations into the A site of the perovskite matrix to increase the entropic
17
2+
contributions of phase segregation, (ii) increasing the crystalline size by substituting Pb
2
+
18
with Sn cations into the B site of the perovskite matrix, and (iii) passivating the surface of
perovskites with potassium cations to form a potassium-halide layer to suppress photoinduced
1
9
ion migration in perovskite polycrystalline films. However, for LEDs, the crystalline size of
perovskites has to be confined at the nanoscale for efficient light emission, and the ion
migration caused by an electric field is more serious than the photoinduced ion migration in
solar cells, making it more difficult to stabilize the EL peak of mixed halide perovskite-based
LEDs. To suppress halide segregation in perovskite LEDs, passivation of the surface halide
vacancies in mixed halide nanocrystalline perovskites is required to inhibit ion migration
2
0
because the surface halide vacancies that trigger ion migration are suppressed. In addition,
tuning the crystalline size below the threshold size (approximately 46 nm) is a possible route
to fabricate phase-segregation-free mixed halide perovskite films by counterbalancing the
confinement of carrier diffusion into the grains with the formation and accumulation of
segregated domains.21 Last but not least, modifying the grain boundary of perovskite
polycrystals with a passivated potassium-halide layer can both significantly improve the
4
ACS Paragon Plus Environment

Page 5 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
optical properties and suppress the ion migration of a perovskite polycrystalline film, which
1
9
has potential application prospects for stabilizing the EL of light emitting diodes. Thus, a
new type of well-passivated mixed halide perovskite CsPbI Br NCs are desirable as
3
-x
x
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
emitters for pure color, stable perovskite LEDs.
Based on the above strategies,19-22 herein, we report the preparation of high-quality
potassium bromide surface passivated (K-Br passivated) mixed halide perovskite CsPbI Br
x
3
-x
NCs for the fabrication of stable and pure red perovskite LEDs. The synthesis of passivated
NCs is facilely achieved via a typical hot-injection (HI) method by adding K-oleate (K-OA) to
the precursor solution and then swiftly injecting a mixture of trimethylsilyl bromine (TMSBr)
23
and trimethylsilyl iodide (TMSI) at 180 °C. The obtained K-Br-passivated CsPbI Br NCs
3
-x
x
exhibited a high PLQY (over 90%), high dispersibility, and good stability for weeks under
ambient conditions. We fabricated pure red perovskite LEDs emitting at 637 nm using the
obtained K-Br-surface-passivated CsPbI Br NC film as an emitter, and the maximum EQE
3
-x
x
of the device reached 3.55%. Importantly, K-Br-passivated CsPbI Br NC-based pure red
3-x
x
LEDs showed high color stability during the operation.
EXPERIMENTAL SECTION
Chemicals. Cesium acetate (CH COOCs, Aladdin, 99.99%), lead acetate trihydrate
3
(
Pb(CH COO) ·3H O, Aladdin, 99.998%), potassium acetate (CH COOK, Macklin, 99.0%,
3 2 2 3
AR), sodium acetate (CH COONa, Aladdin, AR), lithium acetate (CH COOLi, Aladdin,
3
3
99%), 1-octadecene (ODE, Aldrich, 90%), oleylamine (OAm, Aldrich, 70%), oleic acid (OA,
Afla Aesar, 90%), trimethylsilyl iodide (TMSI, Aladdin, 97%, with Cu as a stabilizer),
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
trimethylsilyl bromine (TMSBr, Aladdin, 98%, with Cu as a stabilizer), PEDOT:PSS solution
(Al4083 CLEVIOS), poly-TPD (Xi’an Polymer Light Technology Corp. ), TPBi (Nichem),
toluene (C H , Sinopharm Chemical Reagent Co. Ltd. (SCRC), AR), n-octane (C H , SCRC,
7
8
8
18
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
AR), ethyl acetate (CH COOCH CH , SCRC, AR), and methyl acetate (CH COOCH ,
3
2
3
3
3
Aladdin, 99.5%, anhydrous) were used without further purification.
Material Synthesis. Preparation of K-oleate: The K-oleate (K-OA) precursor was
prepared by loading a mixture of 0.1 mmol of CH COOK, 4.5 mL of ODE and 0.5 mL of OA
3
into a 7 mL vial. Then, the solution was heated and stirred for 30 min at 80 °C to form a
transparent solution. The obtained K-OA solution was stored in dry conditions. The
concentration of K-OA in the solution was 0.02 mol L-1.
Preparation of Li-oleate: The Li-oleate (Li-OA) precursor was prepared by loading a
mixture of 0.1 mmol of CH COOLi, 4.5 mL of ODE and 0.5 mL of OA into a 7 mL vial.
3
Then, the solution was heated and stirred for 30 min at 80 °C to form a transparent solution.
The obtained Li-OA solution was stored in dry conditions. The concentration of Li-OA in the
solution was 0.02 mol L-1.
Preparation of Na-oleate: The Na-oleate (Na-OA) precursor was prepared by loading a
mixture of 0.1 mmol of CH COONa, 4.5 mL of ODE and 0.5 mL of OA into a 7 mL vial.
3
Then, the solution was heated and stirred for 30 min at 80 °C to form a transparent solution.
The obtained Na-OA solution was stored in dry conditions. The concentration of Na-OA in
the solution was 0.02 mol L-1.
Synthesis of pristine CsPbI Br NCs: First, 0.1 mmol of CH COOCs, 0.2 mmol of
3
-x
x
3
Pb(CH COO) ·3H O and 5 mL of ODE were added into a 25 mL 3-neck flask, and the
3
2
2
6
ACS Paragon Plus Environment

Page 7 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
obtained solution was degassed under vacuum for 20 min at 120 °C and then placed under
nitrogen flow for 5 min. Then, 0.5 mL of OA and 1.0 mL of OAm were injected into the
above solution under nitrogen flow. Five minutes later, the flask was placed under vacuum
again for 30 min until the solution was transparent and clear. Subsequently, the temperature
was increased to 180 °C under nitrogen flow, and a mixture of 58 μL of TMSBr and 110 μL
of TMSI was swiftly injected. After at least 10 s, the reaction mixture was cooled in an
ice-water bath.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Synthesis of K-Br-passivated CsPbI Br NCs: First, 0.1 mmol of CH COOCs, 0.2 mmol
3
-x
x
3
of Pb(CH COO) ·3H O, 1.0 mL of K-OA and 4.0 mL of ODE were added into a 25 mL
3
2
2
3
-neck flask. The remaining synthetic procedure was the same as above. Notably, the ratio of
K to Pb was 10% in our typical synthesis.
Synthesis of Li-passivated CsPbI Br NCs: First, 0.1 mmol of CH COOCs, 0.2 mmol of
3
-x
x
3
Pb(CH COO) ·3H O, 1.0 mL of Li-OA and 4.0 mL of ODE were added into a 25 mL 3-neck
3
2
2
flask. The remaining synthetic procedure was the same as above. Notably, the ratio of Li to
Pb was 10% in our typical synthesis.
Synthesis of Na-passivated CsPbI Br NCs: First, 0.1 mmol of CH COOCs, 0.2 mmol of
3
-x
x
3
Pb(CH COO) ·3H O, 1.0 mL of Na-OA and 4.0 mL of ODE were added into a 25 mL 3-neck
3
2
2
flask. The remaining synthetic procedure was the same as above. Notably, the ratio of Na to
Pb was 10% in our typical synthesis.
Purification of Prepared CsPbI Br NCs: The obtained NCs in crude solution were
3
-x
x
precipitated by adding ethyl acetate (volume ratio to that of the NC solution of 3:1), and then,
the mixture was centrifuged at 10000 rpm for 10 min. The supernatant was discarded, and the
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
precipitate was redispersed in 3.0 mL of toluene. The solution was further centrifuged for 3
min at 3000 rpm, and the supernatant was retained. Additional washing could be conducted
by adding methyl acetate (volume ratio to that of the NC solution of 3:1) for 5 min at 10000
rpm to precipitate the NCs. Finally, the obtained precipitate was redispersed in 1.5 mL of
n-octane for subsequent characterizations and tests.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CsPbI Br NC-Based LED Fabrication and Performance Test. Initially, the ITO glass
3
-x
x
substrate was ultrasonically cleaned in ITO washing solution, deionized water, isopropyl
alcohol, acetone and ethanol in sequence. After that, the substrate was dried with nitrogen
flow and treated by UV-ozone for 20 min. After the surface treatment, the PEDOT:PSS
solution was spun onto the surface of the ITO-coated glass substrate by a spin coater (Chemat
Technology spin-coater kw-4A) at 3000 rpm for 50 s and annealed at 140°C on a hot plate
(Chemat Technology kw-4AH) for 15 min under ambient conditions. A solution of poly-TPD
-1
in chlorobenzene (8 mg mL ) was spin-coated on the PEDOT:PSS layer by a spin-coater at
3
000 rpm for 50 s and annealed at 130 °C on a hot plate for 30 min under a nitrogen
atmosphere. To form an NC emission layer, the NC solution was spin-coated on top of the
poly-TPD film at a speed of 2000 rpm for 45 s. Finally, 40 nm TPBI, 1 nm LiF and 100 nm
-4
Al were thermally evaporated under a high vacuum (∼2 × 10 Pa). The device active area
2
was 4 mm . For the as-fabricated LED device performance test, a Keithley 2400 SourceMeter
was used to record the current versus voltage characteristics. The luminescence of the device
was revealed by photon flux using a silicon photodiode, which was calibrated by a PR-670
spectra scan luminance meter. The electroluminescent spectrum (EL) of the device was
collected using an Ocean Optics JAZ spectrometer. Other important parameters used to
8
ACS Paragon Plus Environment

Page 9 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
characterize LEDs were all calculated from the L-J-V and EL measurements under the
assumption that the emission of the LED exhibits a Lambertian pattern. The performance
measurement for the device was carried out under a nitrogen atmosphere in a glove box at
room temperature without any encapsulation.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Characterizations. Transmission electron microscopy (TEM) analysis was performed on a
Hitachi HT-7700 operating at an accelerating voltage of 100 kV. The high-resolution TEM
(
HRTEM) and high-angle annular dark field scanning transmission electron microscopy
HAADF-STEM) images were carried out by a Talos F200X operating at an accelerating
(
voltage of 200 kV. The energy dispersive X-ray spectroscopy (EDS) spectra were collected
on a Talos F200X with an energy dispersive detector. The scanning electron microscopy
(SEM) images were taken on a JEOL-JSM-6700 scanning electron microscope at an
acceleration voltage of 5ꢀkV. The atomic force microscopy (AFM) images were collected by
using a Veeco DI Nanoscope V system with tapping mode. The powder X-ray diffraction
(PXRD) pattern analysis was acquired by a Philips X’Pert PRO SUPER X-ray diffractometer
using Cu Kα radiation (λ = 1.54178 Å). The elemental analysis was carried out on a
PerkinElmer Optima 7300 DV inductively coupled plasma-optical emission spectrometer
(ICP-OES). X-ray photoelectron spectroscopy (XPS) was conducted on NCs loaded on a
silicon wafer by a Thermo ESCALAB 250 spectrometer equipped with a monochromatic Al
Kr radiation source (1486.7 eV). The ultraviolet-visible (UV-vis) absorption spectra of the
NC solution were measured by using a PekinElmer instrument in transmission mode. The PL
spectra and PLQYs were measured on a Hamamatsu (C11347) absolute PLQY spectrometer
using 365 nm excitation on diluted toluene suspensions in a quartz cuvette. The UV-vis
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
absorption spectra were recorded using a Varian Cary 300 UV-vis absorption
spectrophotometer. The Fourier transform infrared spectra (FT-IR) of NC powders were
1
1
obtained with a potassium bromide pellet by Nicolet 8700. H nuclear magnetic resonance ( H
NMR) spectroscopy was acquired by Bruker AVANCE AV III 400. The NMR spectra of the
samples were measured using ferrocene as an internal standard. In brief, a mixture of
perovskite NCs and ferrocene (9:1 wt/wt) in deuterated chloroform was prepared at a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
–
1
1
concentration of 10 mg ml for H NMR measurements. Time-resolved PL measurements
were obtained with our home-built time-correlated single photon counting (TCSPC) system
under ambient conditions. The CsPbI Br NC solution stored in a 1 cm cuvette and the
3-x
x
CsPbI Br NC film spin-coated on the glass substrate were excited using a 405 nm laser
3
-x
x
(Picosecond Diode Laser, PiLas). The synchrotron radiation photoemission spectroscopy
SRPES) experiments were performed at the photoemission end-station at beamline BL10B in
(
the Hefei National Synchrotron Radiation Laboratory. In the current work, SRPES spectra
were taken with synchrotron photon energy levels of 450 eV, 650 eV and 1486.7 eV.
RESULTS AND DISCUSSION
Potassium bromide-passivated CsPbI Br NC synthesis and characterization
3
-x
x
For the synthesis of all CsPbI Br NCs, we used an HI method with TMSBr and TMSI as
3-x
x
the halogen source (see the experimental section for details).23 We adopted an excessive
halogen source to obtain halide-rich NCs (molar ratio of Pb to X of 1:6).24 For the
K-Br-passivated CsPbI Br NCs, we added 1.0 mL of K-OA (0.02M) into the precursor
3
-x
x
solution before adding oleic acid (OA) and oleylamine (OAm), in which the molar ratio of K
1
0
ACS Paragon Plus Environment

Page 11 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
to Pb was 10%. Notably, unless otherwise stated, all the samples were purified twice before
measurement. Figure 1a shows the transmission electron microscopy (TEM) image of the
as-synthesized K-Br-passivated CsPbI Br NCs, demonstrating their monodispersity and
3
-x
x
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
cubic shape, as expected for Cs-based lead halide perovskite NCs.4 The typical
high-resolution TEM (HRTEM) image of K-Br-passivated CsPbI Br NCs is shown as the
3
-x
x
inset in Figure 1a. As a comparison, the TEM and HRTEM images of pristine CsPbI Br
x
3
-x
NCs are provided in Figure S1. Obviously, the lattice distances of both NCs are 0.30 nm,
25
corresponding to the (200) plane lattice spacing of cubic phase CsPbI Br . The size
3
-x
x
distribution of NCs was counted based on the TEM image, and both samples showed almost
the same average size of approximately 11.4 nm (Figure S2), indicating that the addition of
potassium ions did not obviously affect the crystalline size of NCs. Moreover, the spatial
distributions of Cs, Pb, I, Br and K species in K-Br passivated CsPbI Br NCs were further
3
-x
x
revealed by high-angle annular dark field scanning transmission electron microscopy
HAADF-STEM) and corresponding elemental mapping (Figure 1b). The homogeneous
(
distribution of main elements (Cs, Pb, I and Br) among all of the NCs suggests that halogens
are mixed well in the perovskite lattice, and the distribution of potassium in the spatial region
indicated that potassium ions were present in the CsPbI Br NCs. Energy dispersive X-ray
3
-x
x
spectroscopy (EDS) elemental quantification revealed a Cs: Pb: I: Br: K atomic ratio of 1.05:
.00: 1.84: 0.90: 0.05 (Figure S3). Elemental scan profiles of potassium and lead elements
Figure 1c) further present that the signal of potassium for a typical NC displayed relatively
high counts, implying that potassium ions mainly exist on the surface of the CsPbI Br
1
(
3
-x
x
NCs.22
1
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
To further validate the location of the potassium ions, we applied X-ray diffraction (XRD)
patterns (Figure 1d) to obtain more information on the crystal structure of the as-prepared
CsPbI Br NCs. Both NCs have cubic crystalline structures (refer to -CsPbI ICSD
3
-x
x
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
No.161481) and no formation of any impurity phases. Careful examination of the XRD
patterns reveals that the peak intensity of the K-Br-passivated CsPbI Br NCs is stronger
3
-x
x
than that of the pristine CsPbI Br NCs, which verifies that the addition of potassium ions
3
-x
x
can improve the crystallinity of the NCs. Furthermore, we determined the atomic ratio of K to
Pb in K-Br-passivated CsPbI Br NCs by inductively coupled plasma-optical emission
3
-x
x
spectroscopy (ICP-OES) (Table S1). For the sample with one cycle of purification by ethyl
acetate, the ratio of K to Pb was 10.6%. After two cycles of purification by methyl acetate,26
the ratio of K to Pb decreased to 6.0%, which well matched the elemental quantification
measured by EDS. In this context, we speculate that the potassium ions were enriched on the
surface of the NCs.19, 22
1
2
ACS Paragon Plus Environment

Page 13 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. Morphology and characterizations of the as-synthesized K-Br-passivated
CsPbI Br NCs. (a) The TEM image of K-Br-passivated CsPbI Br NCs. The inset shows
3
-x
x
3-x
x
the corresponding HRTEM image. Scale bars: 100 nm and 5 nm, respectively. (b)
HAADF-STEM image and corresponding elemental mapping images of K-Br-passivated
CsPbI Br NCs. Scale bar: 20 nm. (c) Elemental scan profiles of potassium and lead from a
3
-x
x
typical single NC as indicated by the yellow arrow in the dark-field STEM image. The counts
of the potassium element signal are tripled. Scale bar: 20 nm. (d) XRD patterns of the
prepared CsPbI Br NCs (-CsPbI ICSD No.161481).
3
-x
x
3
To figure out the states of the potassium ions present on the surface of the NCs, the high-
-resolution XPS spectra are provided in Figure 2a and Figure S5. All elements, Cs 3d, Pb 4f, I
3d, Br 3d and K 2p for the pristine CsPbI Br NCs and K-Br-passivated CsPbI Br NCs
3
-x
x
3-x
x
were calibrated with C 1s. The signal of potassium was detected after the addition of K-OA,
which further proved the existence of K on the surface of the NCs. In addition, after adding
K-OA, the Br 3d3/2 and 3d5/2 peaks shifted to higher binding energies from 69.3 and 68.3 eV
to 69.6 and 68.5 eV, respectively, while the Pb 4f5/2 and 4f7/2 peaks at 143.2 and 138.3 eV
slightly shifted to lower binding energies of 143.1 and 138.2 eV. Concurrently, no obvious
1
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
peak shift was observed for I 3d and Cs 3d spectra. We inferred that these results could be
attributed to the strong bond between potassium ions and bromide ions on the surface of the
NCs, which also weakened the bond between lead ions and bromide ions.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
To further confirm that the most of potassium ions present on the surface of the NCs were
bonded with bromide ions, we used synchrotron radiation photoemission spectroscopy
(SRPES) characterization on the uniform K-Br-passivated CsPbI Br NC film spin-coated
3
-x
x
on a silicon substrate. Utilizing incident photons with different energies can detect the
elementary information about the NC film at different thicknesses because the higher energy
photons can deeper penetrate the NCs and the detecting thickness is usually less than the size
27
of a single NC (Figure 2b). First, the SRPES spectra were taken with synchrotron photon
energy levels of 450 eV and 650 eV under a nitrogen atmosphere. As shown in Figure 2c,
when the energy of the incident photons was 450 eV, the ratio of K: Br: Pb on the surface of
the CsPbI Br NC film was calculated to be 0.141: 1.96: 1. In contrast, the ratio of K: Br: Pb
3
-x
x
decreased to 0.085: 1.14: 1 when the energy of incident photons increased to 650 eV (Figure
d). These results show that, as the detection depth increased, the ratio of K and Br to Pb
2
decreased significantly, which indicates that the potassium ions and bromide ions are enriched
on the surface of the CsPbI Br NCs. We further enhanced the incident photon energy to
3
-x
x
1
486.7 eV using a monochromatic Al Kr radiation source in order to detect deeper elementary
information in the CsPbI Br NCs. As the energy of incident photons increased to 1486.7
3-x
x
eV, the ratio of K: Br: Pb was calculated to be 0.101: 1.19: 1 (Figure 2e). The slight increase
in the ratio of K and Br to Pb can be attributed to the detection thickness variation and the
inhomogeneous distribution of potassium ions and bromide ions across the section of NCs.
1
4
ACS Paragon Plus Environment

Page 15 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Furthermore, we compared the binding energy variation of Pb, K and Br at a different
detection depth to reveal the binding type of Pb, Br and K on the surface and inside of the
NC. The binding energy of Br 3d and K 2p decreased as the depth of detection increased
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(Figure 2f and Figure 2g), while no obvious peak shift was observed for the Pb 4f spectrum
(Figure 2h), indicating that bromide and potassium have different bonding types on the
surface and in the interior of NCs. The reduction in the internal binding energy of bromide
comes from the stronger bond between bromide ions and potassium ions than that between
bromide ions and lead ions (68.70 eV for Br 3d5/2 of KBr and 68.10 eV for Br 3d5/2 of
CsPbBr3).28, 29 Additionally, the decreasing internal binding energy of potassium results from
the fact that potassium ions enter the interstitial area of the perovskite lattice in the NCs.
Therefore, we confirmed that the potassium ions were not only enriched on the surface of
NCs binding with bromide ions but also entered the perovskite lattice in a small amount.
1
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. X-ray photoelectron spectroscopy characterizations of CsPbI Br NCs. (a)
3-x
x
High-resolution XPS spectra of pristine and K-Br-passivated CsPbI Br NCs for Br 3d. (b)
3
-x
x
Schematic illustration of SRPES detected with photons of different energy levels. Photons
with higher energy can detect thicker-scale elemental information in the NC film. (c-e) The
atomic ratios of X/Pb measured by SRPES for K-Br-passivated CsPbI Br NC film with
3
-x
x
photons of different energy levels. (c) 450 eV, (d) 650 eV and (e) 1486.7 eV. (f-h) The
SRPES of K-Br-passivated CsPbI Br NC film for (f) K 2p, (g) Br 3d and (h) Pb 4f. The
3
-x
x
energy levels of incident photons are 450 eV, 650 eV and 1486.7 eV, respectively. The
intensity of the sample measured by photons with an energy of 650 eV tripled.
In addition, we characterized the optical properties of prepared mixed halide perovskite NC
solutions. Figure 3a shows the NC solution under daylight and UV light, and both samples are
well dispersed with bright pure red emission. PL spectra and UV-vis absorption of these
prepared NC solutions are exhibited in Figure 3b, and the emission peaks of both CsPbI Br
3
-x
x
1
6
ACS Paragon Plus Environment

Page 17 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
NCs are approximately 640 nm. The PLQY of K-Br-passivated CsPbI Br NCs was
3
-x
x
maintained at 93% even after two cycles of purification, signifying the good surface
passivation effect for K-Br enriched on the surface of CsPbI Br NCs. Time-resolved PL
3
-x
x
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
decay curves of these prepared NC solutions are presented in Figure 3c and Figure S7. The
PL decay curves were fitted with a biexponential function to determine the PL decay
lifetimes, which are summarized in Table S2 and Table S3. The PL lifetime of
K-Br-passivated CsPbI Br NCs (30.51 ns) is similar to that of pristine CsPbI Br NCs
3
-x
x
3-x
x
(29.02 ns), due to the low defect concentration induced by the halide-rich surface of both
samples in solution (Figure S8).
The surface organic ligands of NCs prefer to block charge injection and transport inside
LEDs due to the poor electrical conductivity of OA and OAm.7, 30 The Fourier transform
infrared spectra (FT-IR) were used to examine the presence of ligands on the surface of the
as-synthesized CsPbI Br NCs (Figure S9). The FT-IR characterizations present peaks
3
-x
x
-1
located at 2926 and 2853 cm ascribed to the C-H stretching modes of alkyl groups, which
exhibit an obvious decrease with further purification. However, after two cycles of
-1
-1
purification, the FT-IR peaks located at 1698 cm and 1632 cm belonging to the C=O and
+
N-H stretching modes still exist, indicating that a small amount of OA and OAm still exist
3
30
1
on the surface of the NCs to stabilize the NCs. We also tested the H NMR spectra of the
samples using ferrocene as an internal standard to determine the contents of OA and OAm.7
1
All H NMR signals from the NCs were normalized using the ferrocene resonance (4.19 ppm)
as the reference. As shown in Figure 3d, the intensity of the alkenyl region (δ = 5.2-5.4 ppm)
1
of the H NMR spectra obviously decreased, signifying that the total concentration of surface
1
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ligands decreased after further purification, which is consistent with the FT-IR
characterization. There is no manifest difference between pristine CsPbI Br NCs and
3
-x
x
K-Br-passivated CsPbI Br NCs in terms of surface ligands, demonstrating that K-Br surface
3
-x
x
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
passivation did not affect the condition of the surface ligands.
Figure 3. Characterization of prepared NC solutions. (a) Photos of prepared CsPbI Br
x
3
-x
NC solutions under daylight (left) and UV-light (365 nm) irradiation (right). The left sample
is the pristine CsPbI Br NC solution, and the right sample is the K-Br-passivated
3
-x
x
CsPbI Br NC solution. The PLQYs of these two samples were 85% and 93%, respectively.
3
-x
x
(
b) Absorbance and PL spectra of the prepared CsPbI Br NC solution. (c) Time-resolved
3-x x
1
PL spectra of the prepared CsPbI Br NC solution. (d) H NMR spectroscopy of pristine
3
-x
x
CsPbI Br NCs and K-Br-passivated CsPbI Br NCs with different purification cycles. A
3
-x
x
3-x
x
certain amount of ferrocene was added in the solution as an internal standard.
It should be noted that excessive potassium ions in the reaction solution have an adverse
effect on the morphology and structure of NCs. The TEM images of different ratios of K to
1
8
ACS Paragon Plus Environment

Page 19 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Pb in the reaction solution are provided in Figure S10. When the ratio of K to Pb is over 20%,
the CsPbI Br NCs are easily aggregated, and some large fused crystals are formed.
3
-x
x
Correspondingly, impurity phase KBr0
I
(ICSD No.24-1117) was detected in the products
.67 0.33
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
+
of the 20% and 30% K reaction systems, while no impurity phase was discovered in the 10%
+
K CsPbI Br NCs (Figure S11). This phenomenon illustrates that potassium ions have
3
-x
x
strong bonding with excessive halide ions, especially bromide ions, on the surface of NCs.
Simultaneously, when the ratio of K to Pb reached 20%, the bond between excess potassium
ions and surface halide ions will increase the halide vacancies on the surface of the NCs,
leading to the decrease in the PLQY (Figure S12) and the reduction in the NC solution
stability (Figure S13 and Figure S14).22 Therefore, the appropriate ratio of K to Pb is
necessary to gain well-passivated CsPbI Br NCs.
3
-x
x
Moreover, we synthesized a series of other alkali metal ion-passivated CsPbI Br NCs (Li
3
-x
x
and Na) by adding 0.02 mmol of Li-OA or 0.02 mmol of Na-OA to the reaction system, and
the ratio of alkali metal ion to lead is 1: 10. However, after storage in air for ten days, the
PLQYs of the Li- and Na-passivated NC solutions showed apparent decreased (Figure S15).
In addition, after storage in air, a large number of fused particles formed in the Li-passivated
CsPbI Br NCs, and a small amount of agglomerated particles formed in the Na-passivated
3
-x
x
CsPbI Br NCs, as shown in the TEM images (Figure S16), indicating that lithium ions and
3
-x
x
sodium ions exhibited a limited passivation effect on CsPbI Br NCs.
3-x
x
Stable, pure red emissive CsPbI Br NC films
3
-x
x
We further fabricated CsPbI Br NC solid thin films via spin-coating and characterized
3
-x
x
1
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
their optical properties and stabilities. By spin-coating NC solution on the clean glass
substrates, we obtained smooth thin films with high PLQYs (Figure 4a). The PL emission of
the as-obtained films shows a slight redshift (~1-2 nm) compared with the initial emissions
from NC solution due to formation of conduction/valence bands through agglomeration of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
21
NCs (Figure 4b). Similarly, the time-resolved PL decay curves of these prepared NC films
tested in air are presented in Figure 4c. The PL decay curves were fitted with a biexponential
function to determine the PL decay lifetimes, which are summarized in Table S4. It has been
demonstrated that surface defects will give rise to nonradiative surface states and lead to
3
1
shorter exciton lifetimes. The PL lifetime of K-Br-passivated CsPbI Br NC film (21.11
3-x
x
ns) is longer than that of the pristine CsPbI Br NC film (15.84 ns), signifying that the
3
-x
x
K-Br-passivated CsPbI Br NC film possesses fewer trap states and has better pure red light
3
-x
x
3
2
emission. It should be noted that, in the pristine CsPbI Br NC film, moisture attack on the
3
-x
x
surface of pristine NCs causes the formation of defects when they are exposed to air without
the protection of a solvent, and these defects can trap excitons, resulting in more nonradiative
recombination (Figure S17a).33 However, in the K-Br-passivated NC film, the defect
concentration on the surface is lower than that of the pristine NC films due to the passivation
effect of the K-Br enriched surface, leading to less nonradiative recombination and a longer
PL lifetime (Figure S17b). In addition, we tested the PL lifetime of NC films exposed to air
for 30 min to show the effect of K-Br surface passivation (Figure S18 and Table S4). The PL
lifetimes of freshly prepared and aged K-Br-passivated CsPbI Br NC films are 21.11 and
3
-x
x
20.70 ns, respectively. In contrast, the PL lifetimes of freshly prepared and aged pristine
CsPbI Br NC films are 15.84 and 11.57 ns, respectively. These results indicate that K-Br
3
-x
x
2
0
ACS Paragon Plus Environment

Page 21 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
passivation can provide better protection for an NC film to maintain PL in air.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Optical properties and morphologies of CsPbI Br NC thin films. (a) Photo of
3-x
x
CsPbI Br NC thin films under UV-light (365 nm) irradiation. From left to right, the samples
3
-x
x
are the pristine CsPbI Br NC film and K-Br-passivated CsPbI Br NC film, respectively.
3
-x
x
3-x
x
PLQYs of these two samples are 52% and 64%, respectively. (b) PL spectra of the
CsPbI Br NC films. (c) Time-resolved PL decay curves of CsPbI Br NC films. (d) SEM
3
-x
x
3-x
x
image of K-Br-passivated CsPbI Br NC film. Scale bar: 500 nm. (e) AFM image of
3
-x
x
K-Br-passivated CsPbI Br NC film. Scale bar: 1 m. (f) PL peak position as a function of
3
-x
x
irradiation time for CsPbBr I NC films. The ensemble films were continuously excited by a
3-x x
-2
laser emitting at 365 nm with a power density of 100 mW cm .
In addition, the morphologies of different perovskite NC thin films were investigated by
scanning electron microscopy (SEM) and atomic force microscopy (AFM). As seen from the
SEM image in Figure 4d, the K-Br-passivated CsPbI Br NC-based red perovskite NC film
3
-x
x
is uniformly covered by a number of NCs. However, the pristine CsPbI Br NC-based film
3
-x
x
displays a porous structure with a large number of mesoscopic pores (Figure S19a). The AFM
images further confirm the better quality of K-Br-passivated CsPbI Br NC thin film. The
3
-x
x
root-mean-square roughness (RMS) values are 2.43 and 1.93 nm for the pristine CsPbI Br
x
3
-x
NC film (Figure S19b) and K-Br-passivated CsPbI Br NC film (Figure 4e), respectively.
3
-x
x
2
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
High film uniformity will significantly reduce the current leakage and result in high efficiency
in the fabricated LEDs.34
To show the K-Br-enriched surface passivation effect on suppressing photoinduced halide
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
segregation of CsPbI Br NCs, we monitored the PL peak variation of CsPbI Br NCs
3-x
x
3-x
x
under continuous UV-light (365 nm) irradiation. For both the pristine CsPbI Br NC
3
-x
x
solution and K-Br-passivated CsPbI Br NC solution, we observed that almost no PL peak
3
-x
x
-2
shift occurred even after 1 hour of irradiation at 30 mW cm (Figure S20) due to the similar
segregation and recovery rate for the initial mixed halide perovskite state in NC solution.15, 21,
3
5
However, phase segregation does occur in the mixed halide CsPbI Br NC film under
3-x
x
laser irradiation.36 The agglomeration of NCs during the formation of a solid thin film
enlarges the carrier diffusion length and consequently increased the rate of phase
1
5
-2
segregation. We used a laser emitting at 365 nm with a power density of 100 mW cm (the
-2
power density of 1 sun illumination is 100 mW cm ) to evaluate the PL spectral stability of
the pristine CsPbI Br NC film and K-Br-passivated CsPbI Br NC film. The PL peak
3-x
x
3-x
x
position as a function of irradiation time for the CsPbBr I NC films is plotted in Figure 4f
3
-x x
(the time-dependent normalized PL spectra are provided in Figure S21). The PL peak of the
pristine CsPbI Br NC film presents a blueshift from the initial 640 nm towards 633 nm after
3
-x
x
4
80 s continuous laser irradiation. By contrast, the PL peak of K-Br-passivated CsPbI Br
3-x x
NC film presents a blueshift from the initial 641 nm towards 638 nm slightly after 480 s
irradiation, indicating the better PL spectral stability of the K-Br-passivated CsPbI Br NC
3
-x
x
film. In addition, the occurrence of a blueshift under continuous laser irradiation is 180 s for
the K-Br-passivated CsPbI Br NC film, while the starting time for the blueshift for the
3
-x
x
2
2
ACS Paragon Plus Environment

Page 23 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
pristine CsPbI Br NC film is as short as 60 s. As reported before, surface defects located on
3
-x
x
the grain boundary can serve as trapping sites for diffused halide ions, promoting the
formation of an iodine-rich domain because of the weaker bond between lead ions and iodine
ions. Consequently, the iodide ions are mainly channeled away from the laser excitation area,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
13
leading to a higher bandgap for carrier recombination. Thus, passivating the surface defects
of CsPbI Br NCs with a K-Br-enriched surface can inhibit halide migration and
3
-x
x
consequently stabilize the PL spectra under laser irradiation.
Performances of CsPbI Br NC-based pure red LEDs
3-x
x
We fabricated pure red LEDs using as-fabricated CsPbI Br NC films as luminescent
3
-x
x
layers. As illustrated in Figure 5a, the architecture of the device is composed of a layered
structure of ITO/PEDOT:PSS/poly-TPD/CsPbI Br NCs/TPBi/LiF/Al, in which ITO serves
3
-x
x
as the anode, PEDOT:PSS as the hole-injection layer, poly-TPD as the hole-transport layer
and electron blocking layer, TPBi as the electron-transport layer, and LiF/Al as the cathode.37
As shown in the cross-sectional SEM image (Figure 5b), the thicknesses of the PEDOT:PSS,
poly-TPD, CsPbI Br NCs, TPBi, and LiF/Al layers are ∼20, ∼40, ∼30, ∼40 and ∼100 nm,
3
-x
x
respectively. The normalized EL and PL spectra of the K-Br-passivated CsPbI Br NC LED
3
-x
x
are plotted in Figure 5c. The EL spectrum is centered at 637 nm with a narrow FWHM of 31
nm (Figure 5c, solid line), corresponding to the Commission Internationale del’Eclairage
(CIE) color coordinates of (0.70, 0.30) (Figure S22). The coordinate of a red channel for
displays is (0.71, 0.29), indicating our CsPbI Br NCs are an ideal emitter for red displays.
3
-x
x
The EL peak exhibits a slight blueshift (~4 nm) relative to the PL peak of the corresponding
2
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
CsPbI Br NC film, which is consistent with the previous reports for the EL variation of
3
-x
x
perovskite LEDs.38, 39 A photo of the LED (inset in Figure 5c) shows the bright pure red-light
emission of our fabricated LED under an applied voltage of 5 V.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Device characteristics of CsPbI Br NC-based pure red LEDs. (a) The
3
-x
x
energy-level diagram of the LED device structure proposed in this study. (b) Cross-sectional
SEM image of the as-fabricated LED device. Scale bar: 200 nm. (c) EL spectrum at an
operation voltage of 5 V and the corresponding PL spectrum for the K-Br-passivated
CsPbI Br NC-based LED. Inset shows the photo of the corresponding device at an applied
3
-x
x
voltage of 5 V. (d) Current density versus driving voltage curves for LEDs based on prepared
CsPbI Br NCs. (e) Luminance versus driving voltage curves for LEDs based on prepared
3
-x
x
CsPbI Br NCs. (f) EQEs of LEDs based on prepared CsPbI Br NCs at different
3
-x
x
3-x
x
luminance.
To demonstrate the merits of K-Br-passivated CsPbI Br NCs for LEDs, we further
3
-x
x
compared the performance of CsPbI Br NC-based pure red LEDs. The current-density–
3-x
x
voltage (J–V) characteristics of the LEDs are shown in Figure 5d. At the turn-on voltage of
the device (3.8 V for pristine CsPbI Br NC-based LED and 3.6 V for K-Br-passivated
3
-x
x
CsPbI Br NC-based LED, Figure 5e), the current density of the K-Br-passivated
3
-x
x
2
4
ACS Paragon Plus Environment

Page 25 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
CsPbI Br NC-based LED is lower than that of the pristine CsPbI Br NC-based LED,
3
-x
x
3-x
x
indicating less current leakage of the device based on the K-Br-passivated NC film. The
luminance–voltage (L–V) curves of the as-fabricated LEDs are presented in Figure 5e. The
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-2
maximum value of luminance is 2671 cd m for K-Br-passivated CsPbI Br NC-based
3-x
x
-2
devices, which is higher than that of the pristine CsPbI Br NCs of 1891 cd m . This
3
-x
x
represents a high value for pure red LEDs emitting at 630 nm ~ 650 nm to date (refer to Table
S5). As seen in Figure 5f, the K-Br-passivated CsPbI Br NC-based LED reaches a
3
-x
x
maximum EQE of 3.55%, while the pure CsPbI Br NC-based LED reaches a maximum
3
-x
x
EQE of 1.89%. Additionally, the devices based on K-Br-passivated CsPbI Br NCs show an
3-x
x
-2
average peak EQE of 3.38% and an average luminance of 2640 Cd m (Figure S23) based on
the statistics of 30 devices, indicating that the fabrication of our device is reproducible. All
the above results indicate that CsPbI Br NCs passivated by the K-Br enriched surface
3
-x
x
present a higher charge injection efficiency and good photoelectric properties, originating
from the uniform NC film and well-passivated surface.40, 41
EL stability evaluation of CsPbI Br NC-based LEDs
3
-x
x
To test the EL spectral stability of CsPbI Br NC-based pure red LEDs, we measured the
3
-x
x
emission spectra with the increase in the driving voltages. As shown in Figure 6a, the redshift
of the EL peak of the pristine CsPbI Br NC-based LED was observed when the driving
3
-x
x
voltage increased, as shown in previous reports.11, 12 In contrast, the EL peak of the device
based on the K-Br-passivated CsPbI Br NCs was maintained at nearly the same position
3
-x
x
(637 nm) even when the driving voltage increased to 6.0 V (Figure 6b). We correlated the EL
peak position with the driving voltage as shown in Figure 6c. The pristine CsPbI Br
3
-x
x
2
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
NC-based LED starts to redshift at 4.6 V and the wavelength increase reaches 20 nm when
the voltage increased to 6.0 V, while the K-Br-passivated CsPbI Br NC-based LED does
3
-x
x
not exhibit any redshift under 6.0 V, which indicates that the phase segregation caused by ion
migration is well suppressed by K-Br passivation. We tested the spectral stability of the
fabricated CsPbI Br NC-based LEDs under a continuous driving voltage of 5.0 V as well.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
-x
x
As shown in Figure 6d, the EL spectra of the LED device based on pristine CsPbI Br NCs
3
-x
x
moves swiftly from 638 nm to 661 nm within 3 min, while the emission peak of the
K-Br-passivated CsPbI Br NC-based LED moves slowly from 637 nm to 645 nm within 3
3
-x
x
min (Figure 6e). In addition, we measured the lifetime of the device by applying a constant
voltage and monitoring the evolution of luminance (Figure 6f). The half-lifetime of the
pristine CsPbI Br NC-based LED is only 30 min, while that of the K-Br-passivated
3
-x
x
CsPbI Br NC-based LED improved to 50 min, indicating the higher stability of our
3
-x
x
K-Br-passivated CsPbI Br NCs.
3
-x
x
Figure 6. Characteristics of EL stability for CsPbI Br NC-based pure red LEDs. (a)
3
-x
x
Normalized EL spectra of pure red LEDs based on pristine CsPbI Br NCs under different
3
-x
x
2
6
ACS Paragon Plus Environment

Page 27 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
driving voltages. (b) Normalized EL spectra of pure red LEDs based on K-Br-passivated
CsPbI Br NCs under different driving voltages. (c) EL peak position versus the driving
3
-x
x
voltage for CsPbI Br NC-based pure red LEDs. (d) Normalized EL spectra of the pure red
3
-x
x
LED based on pristine CsPbI Br NCs at a driving voltage of 5.0 V with different times. (e)
3
-x
x
Normalized EL spectra of the pure red LED based on K-Br-passivated CsPbI Br NCs at a
3
-x
x
driving voltage of 5.0 V with different times. (f) Operation lifetime for the devices based on
prepared CsPbI Br NCs measured at a constant driving voltage of 5.0 V. The luminance of
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
-x
x
the LEDs was normalized to their initial values.
Mechanisms for the high EL stability of K-Br-passivated CsPbI Br NC-based LEDs
3
-x
x
The mechanism for inhibiting ion migration in mixed halide CsPbI Br NC-based LEDs
3
-x
x
passivated by a K-Br enriched surface is discussed here (Scheme. 1). As reported before, the
defects in the perovskite grain boundary trap charge carriers and subsequently generate an
4
2
electric field, further leading to new formation of lower-energy iodide-rich domains. As
mentioned above, potassium ion modification should mainly happen on the surface of
CsPbI Br NCs by bonding with surface bromide ions. On the one hand, potassium ions and
3
-x
x
bromide ions possess a higher binding energy than lead ions and bromide ions, resulting in a
barrier for bromide ions to migrate under an electric field to form halide vacancy defects at
the grain boundary. On the other hand, as reported by Abdi-Jalebi et al., potassium ions and
bromide ions are more likely to passivate the defects at the grain boundary of the perovskite
NC film, consequently, potassium ions and excess halide ions are able to form
halide-sequestering species at the grain boundary, immobilizing the halide ions, thereby
inhibiting their migration.19 Thus, halide segregation can be suppressed via K-Br surface
passivation by lowering the defect concentration of the grain boundary in perovskite NC
films.
2
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme. 1 Illustration of ion migration suppressed by potassium-bromide-enriched
surface passivation.
However, a redshift is still observed in our pure red LEDs because the migration of halide
ions in mixed halide perovskite NC films under an electric field has not been totally inhibited.
This is therefore a key area that will require a significant focused research effort to resolve.1
To fabricate color stable pure light emission perovskite LEDs, additional strategies are
2
+
suggested to be applied in future study: First, substituting Pb ions with suitable cations into
the B site of the perovskite matrix to increase the crystal formation energy and stabilize the
2
5
lattice of mixed halide perovskite; second, passivating surface halide vacancies with a thin
43
inorganic crystal layer at the grain boundary of perovskite NCs to block ion migration; and
finally, optimizing the LED device structure, especially balancing the electron and the hole
injection efficiency to reduce charge accumulation inside the device.6, 44
CONCLUSION
We have developed a facile synthetic approach to prepare high-quality CsPbI Br NCs
3
-x
x
emitting at 640 nm with high PLQY (~93%) by potassium-bromide-enriched surface
passivation. Moreover, we fabricated pure red LED emitting at 637 nm by using
2
8
ACS Paragon Plus Environment

Page 29 of 33
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
K-Br-passivated CsPbI Br NCs as luminescent layers, which is the desired red emission for
3
-x
x
next-generation LED displays. The as-fabricated pure red LEDs exhibited a high brightness of
-2
2
671 Cd m and a relatively high external quantum efficiency of 3.55%. More importantly,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the color stability of pure red LEDs based on K-Br-passivated CsPbI Br NCs is higher than
3-x
x
the one based on pristine CsPbI Br NCs, showing great potential for future EL stable pure
3
-x
x
red LEDs. Our reported K-Br passivation strategy on the surface of mixed halide perovskite
NCs will provide an alternative route to fabricate stable and high color pure perovskite LEDs.
ASSOCIATED CONTENT
Supporting Information.
TEM images, energy dispersive X-ray spectroscopy, XPS spectra, PL spectra, PL lifetime,
FT-IR spectra, XRD patterns, SEM image, AFM image, photostability test and performance
summary of perovskite-based pure red LEDs.
NOTES
The authors declare no competing financial interest.
ACKNOWLEDGEMENT
We acknowledge the funding support from the National Natural Science Foundation of China
(Grants 51571184, 21875236, 51872275), the Fundamental Research Funds for the Central
Universities (Grant WK2060190085), the joint Funds from Hefei National Synchrotron
Radiation Laboratory (Grant KY2060000111), the National Key R&D Program of China
(
Grant No. 2018YFA0306600) and Anhui Initiative in Quantum Information Technologies
Grant No. AHY050000). This work was partially carried out at the USTC Center for Micro
(
and Nanoscale Research and Fabrication. We thank the Photoemission End-stations (BL10B)
at the Hefei National Synchrotron Radiation Laboratory (NSRL) for help in the SRPES
characterizations.
2
9
ACS Paragon Plus Environment