Monolithically-grained perovskite solar cell with Mortise-Tenon structure for charge extraction balance

Abstract

Although the power conversion efficiency values of perovskite solar cells continue to be refreshed, it is still far from the theoretical Shockley-Queisser limit. Two major issues need to be addressed, including disorder crystallization of perovskite and unbalanced interface charge extraction, which limit further improvements in device efficiency. Herein, we develop a thermally polymerized additive as the polymer template in the perovskite film, which can form monolithic perovskite grain and a unique “Mortise-Tenon” structure after spin-coating hole-transport layer. Importantly, the suppressed non-radiative recombination and balanced interface charge extraction benefit from high-quality perovskite crystals and Mortise-Tenon structure, resulting in enhanced open-circuit voltage and fill-factor of the device. The PSCs achieve certified efficiency of 24.55% and maintain >95% initial efficiency over 1100 h in accordance with the ISOS-L-2 protocol, as well as excellent endurance according to the ISOS-D-3 accelerated aging test.

Full text

Article https://doi.org/10.1038/s41467-023-38926-3
Monolithically-grained perovskite solar cell
with Mortise-Tenon structure for charge
extraction balance
Fangfang Wang
1
,MubaiLi
1
,QiushuangTian
1
, Riming Sun
1
, Hongzhuang Ma
1
,
Hongze Wang
1
, Jingxi Chang
1
,ZihaoLi
1
,HaoyuChen
1
,JiupengCao
1
,AifeiWang
1
,
Jingjin Dong
1
,YouLiu
1
, Jinzheng Zhao
1
,YingChu
1
,SuhaoYan
1
,ZichaoWu
1
,
Jiaxin Liu
1
,YaLi
1
, Xianglin Chen
1
,PingGao
1
,YueSun
1
, Tingting Liu
1
,WenboLiu
1
,
Renzhi Li
1
,JianpuWang
1
,Yi-bingCheng
2
, Xiaogang Liu
3
,
Wei Huang
1,4,5
& Tianshi Qin
1
Although the power conversion efficiency values of perovskite solar cells
continue to be refreshed, it is still far from the theoretical Shockley-Queisser
limit. Two major issues need to be addressed, including disorder crystal-
lization of perovskite and unbalanced interface charge extraction, which limit
further improvements in device efficiency. Herein, we develop a thermally
polymerized additive as the polymer template in the perovskite film, which can
form monolithic perovskite grain and a unique “Mortise-Tenon”structure after
spin-coating hole-transport layer. Importantly, the suppressed non-radiative
recombination and balanced interface charge extraction benefitfromhigh-
quality perovskite crystals and Mortise-Tenon structure, resulting in enhanced
open-circuit voltage and fill-factor of the device. The PSCs achieve certified
efficiency of 24.55% and maintain >95% initial efficiency over 1100 h in accor-
dance with the ISOS-L-2 protocol, as well as excellent endurance according to
the ISOS-D-3 accelerated aging test.
Perovskite solarcells (PSCs) have been considered the most promising
emerging photovoltaic technology1–3due to the expressive power
conversion efficiency (PCE) up to 26%4. Although the PCE values are
approaching the efficiency of monocrystalline silicon solar cells, there
are still significant gaps co mpared to the theoretical Shockley-Queisser
limit5. Numerous studies have been devoted to investigating and
analyzing the perovskite composition6–8, crystallization kinetics9–13,
and film morphology14–16 of different perovskite systems, and many
successful strategies have been developed to improve the efficiency of
PSCs. By analyzing most of the issues with perovskite, there are two in-
depth factor that need to be addressed that limit further improve-
ments in device efficiency.
One of the main factors is the often-mentioned issue of perovskite
defects, which can lead to non-radiative recombination and thus
degrade device performance17. As perovskite films deposited by solu-
tion processes are typically polycrystalline18, leading to high number of
structural defects in the bulk19,20, on the surface21, and at the grain
boundaries (GBs) of the films22,23. In fact, most of the defects in/on the
perovskite mainly stems from disordered crystallization of perovskite
films during fabrication. Many solutions have been studied to enhance
Received: 3 February 2023
Accepted: 19 May 2023
Check for updates
1
Key Laboratory of Flexible Electronics (KLOFE), Institute of Advanced Materials (IAM) & School of Flexible Electronics (Future Technologies), Nanjing Tech
University (NanjingTech), 30 South Puzhu Road, Nanjing 211816, China.
2
Advanced Technology for Materials Synthesis and Processing, Wuhan University of
Technology, Wuhan, Hubei 430070, China.
3
Department of Chemistry, National University of Singapore, Singapore 117543, Singapore.
4
KeyLaboratoryfor
Organic Electronics & Information Displays (KLOEID) & Institute of Advanced Materials (IAM), Nanjing University of Posts and Telecommunications, Nanjing,
Jiangsu210023, China.
5
Frontiers Science Centerfor Flexible Electronics& Institute of Flexible Electronics (IFE), Northwestern PolytechnicalUniversity (NPU),
Xi’an, Shanxi 710072, China. e-mail: iamwhuang@nwpu.edu.cn;iamtsqin@njtech.edu.cn
Nature Communications | (2023) 14:3216 1
1234567890():,;
1234567890():,;
Content courtesy of Springer Nature, terms of use apply. Rights reserved

the crystallinity and surface morphology of the active layer, such as
modulating perovskite formulation, optimizing deposition
techniques24–26, additive engineering27–30, and compensatory interface
passivation31.
Another factor is the unbalanced charge extraction between the
perovskite and the charge transport layer (CTL), however, not enough
attention has been paid to it yet32–34. As a typical sandwich-structured
device, not only the photoactive perovskite layer but also the charge
transport layers including electron transport layer (ETL) and hole-
transport layer (HTL) have a significant impact on the performanceof
the device, especially on reducing open-circuit voltage (V
OC
)losses
and suppressing hysteresis35,36.V
OC
and fill factor (FF) are often wea-
kened due to undesired carrier losses at the perovskite/CTL interface
during charge extraction and transport21. In particular, a large number
of defects are predominantly located on the upper surface and at the
GBs of the perovskite films, resulting in a serious trap-assisted non-
radiative recombination at the perovskite/HTL interface37,38. Conse-
quently, the hole extraction efficiency at the interface is substantially
lower than the electron extraction efficiency, causing the interfacial
space charge to form and accumulate39. Most of the research focused
on developing new HTLs with high hole mobility or adding interfacial
layers to provide gradient energy levels40,41,however,itisstillnot
possible to efficiently achieve balanced carrier extraction.
Herein, we used a thermally polymerized additive N-vinyl-2-pyr-
rolidone (NVP) as a polymer template in the perovskite film, followed
by a conventional HTL/Chlorobenzene (CB) solution spin-coating
process to remove the residual miscellaneous phases and open the
GBs to form monolithic perovskite grains, thereby suppressing the
defect-related non-radiative recombination. Furthermore, this process
results in the formation of a novel “Mortise-Tenon”(M-T) structure for
perovskite/HTL composite film (Fig. 1a, b), which provides an
obviously larger contact area between perovskite and HTL, thereby
facilitating hole extraction to achieve balanced charge management.
Based on the high-quality perovskite crystalline and the unique M-T
architecture that effectively enhances V
OC
and FF, PSCs can achieve
certified efficiencies of 24.55% for reverse scan and 24.25% for a for-
ward scan. Moreover, NVP-based PSCs maintain >95% initial efficiency
over 1100 h in accordance with the ISOS-L-2 protocol, as well as
excellent endurance according to the ISOS-D-3 accelerated aging test.
Results
Crystallographic characterization of monolithic perovskite
grain and Mortise-Tenon structure
We used NVP as additive in the perovskite precursor, which could
straightforwardly convert into polyvinylpyrrolidone (PVP) via atom
transfer radical polymerization (ATRP) during perovskite annealing
step. As in traditional n-i-p PSCs, the perovskite films were then cov-
ered with Spiro-OMeTAD/CB solution. To explore the effect of the
additive on the architecture of the perovskite film, we operated a
scanning transmission electron microscope (STEM) on the cross-
sectional perovskite film fabricated by focused-ion-beam (FIB). In the
overall view of STEM (Fig. 1a), we found that the perovskite film
3.18 Å <002>
HTL-Tenon
Mortise-Tenon Structure
Mortise-Tenon Structure
Mortise
Pb
Perovskite
Control
HTL
IC
390 nm 560 nm
340 nm 280 nm
10 nm
PVP
w/NVP w/NVP
Perovskite-Mortise
Tenon
ab
c
ef
d
500 nm 500 nm 500 nm 500 nm
qxy [Å-1]qxy [Å-1]
qz [Å-1]
qz [Å-1]
1μm
Fig. 1 | Crystallization, architecture, and morphology of perovskite/NVP film.
aCross-sectional STEM image of an ultra-thin perovskite slice (<100 nm thickness)
fabricated by FIB with an architecture (from top to bottom) of sputtered Pt/spiro-
OMeTAD (HTL)/perovskite with NVP/SnO
2
(ETL)/FTO/glass. The depth of the
inserted HTL depth is 1/2~2/3 of the total perovskite grain height. The yellow
dashed line is the boundary between spiro-OMeTAD and dielectric barrier of PVP.
bDiagram of Mortise-Tenon structure. cEnlarged images of STEM image, the
labeled triangular area may be miscellaneous phases in PVP matrix.
dCorresponding EDS elemental mapping of Pb, I, and C from the HAADF image
shown in c.eTEM image clearly shows that perovskite grains are surrounded by
PVP. The crystal lattice distance of perovskite was 3.18 Å, and the SAED pattern of
corresponding interdigital perovskite films as illustrated in the inset. fThe dif-
fraction patterns of control and perovskite/NVP films were collected by GIWAXS at
a small angle.
Article https://doi.org/10.1038/s41467-023-38926-3
Nature Communications | (2023) 14:3216 2
Content courtesy of Springer Nature, terms of use apply. Rights reserved

possessed monolithic grains with the average length and width of
grains exceed 1 μm. In contrast, the cross-sectional SEM images of the
control film (Supplementary Fig. 1) showed small and disordered
crystals. X-ray diffraction (XRD) (Supplementary Fig. 2) exhibited that
the control perovskite film had a strong PbI
2
signal at 12.8°42, whereas
the perovskite with NVP addition (denoted as perovskite/NVP in the
following discussion) featured a sharp peak of α-phase and no PbI
2
signal was observed. Furthermore, surprisingly, the perovskite/HTL
composite film showed a M-T structure. As shown in Fig. 1b, M-T
structure has been used by woodworkers for centuries, because the
concave and convex parts joint each other and possess a large con-
nection area. As shown in Fig. 1a, c, and Supplementary Fig. 3, the
detailed boundary line between the inserted HTL and the perovskite
was clearly observed by high-angle annular dark-field (HAADF), where
the depth of the inserted HTL depth ranged from 390 to 560 nm,
representing 1/2~2/3 of the total perovskite grain height. The per-
ovskite/HTL composite films formed a unique M-T structure thatcould
provide a larger contact area between perovskite and HTL, thereby
facilitating the hole extracting. Its corresponding elemental distribu-
tions were displayed via energy dispersive X-ray spectroscopy (EDS)
(Fig. 1d and Supplementary Fig. 4). Pb and I were predominantly and
uniformly located in the monolithic perovskite grains, with a small
distribution in the triangular area labeled in Fig. 1c. C derived from
Spiro-OMeTAD was mainly found in the upper layer of perovskite and
extended deep into the GBs. It was worth noting that C also dispersed
in the triangular area, which might belong to PVP. This overlap in the
triangular region of Pb, I, and C element distribution might be the
uncoordinated miscellaneous phases in PVP matrix, which could be
considered as a dielectric zone between the HTL and ETL. We further
performed high-resolution transmission electron microscopy (HR-
TEM) on perovskite/NVP sample (Fig. 1e). TEM images clearly showed
that the polymerized NVP/PVP gels surrounded perovskite crystal
grains as amorphous phases, in which the selected area diffraction
manifested an perovskite crystal latticewith <002> plane of α-phase at
3.18 Å43. The polymerizing could be verified by Fourier transform
infrared spectroscopy (FTIR) (Supplementary Fig. 5), in which C=C
stretching (1623 cm−1) disappeared after heating at 100 °C for 1 h44.The
grazing-incidence wide-angle x-ray scattering (GIWAXS) was further
measured (Fig. 1f and Supplementary Fig. 6). The neat perovskite
sample demonstrated PbI
2
peak at q
z
=0.90Å
−1,andδ-phase rings at
q
xy
=1.62and1.83Å
−1, whereas perovskite/NVP film exhibited a pure α-
phase without these miscellaneous signals13. Additionally, the diffrac-
tion halo around q
z
=1.2~2.0Å
−1was attributed to PVP in the per-
ovskite films.
The formation analysis of Mortise-Tenon structure
The above crystallographic analysis allows us to propose the main
processes involved in the formation of monolithic perovskite grains
and M-T structures of the perovskite/NVP film, as shown in the sche-
matic diagram in Fig. 2a. As NVP exhibited excellent solubility for PbI
2
and FAI, and miscibilitywith DMSO, as shown in Fig. 2b, the solution of
perovskite precursors with NVP/or DMSO was transparent and could
Control
CB extraction
Control
Control
w/NVP
w/NVP
144 142 140 138 136
Binding energy (eV)
+
++
-
--
PSK/NVP
Control
pristine state
w/NVP
CB extraction
w/NVP
pristine state
DMF/DMSO
Spiro-OMeTAD/CB
PbI
PVP
Enhanced Hole Extration
Monolithic grain
Perovskite (PSK)
Electron Extration CBD-SnO2
FTO
PSK/NVP/DMSO
PSK/PVP
PVP PVP/CB
PSK/PVP/DMSO
Nucleation
Crystalization
HTL Coverage
ab
cd
e
f
1 μm 1 μm 1 μm
1 μm
PSK
NVP
Pb⁰ 4f
5/2
Pb 4f
5/2
Pb 4f
7/2
Pb⁰ 4f
7/2
0 2000 4000 6000
10
−3
10
−2
10
−1
10
0
PL Intensity (arb. units)
Time (ns)
Fig. 2 | Formation mechanism of Mortise-Tenon (M-T) structure. a Schematic
illustration of the formation mechanism of M-T structure of perovskite/NVP and
upper HTL. bPerovskite precursor is solublein pure NVP andis miscible withDMSO
(top); perovskite precursor in NVP is polymerized to PVP after heating at 100°C for
20 min and the solidified product is not dissolved in DMSO (middle); PVP can be
dissolved in CB (bottom). SEM images of c, the control perovskite films and
d, perovskite/NVP film before and after CB extraction. ePb 4fand I 3d XPS spectra
of the control and perovskite/NVP film, respectively. fTime-resolved photo-
luminescence decay curves (excitation: 520 nm, 2.26 nJ cm−2, 0.1 MHz). Solid lines
were fitted from the generic kinetic model to obtain the trap density of
perovskite films.
Article https://doi.org/10.1038/s41467-023-38926-3
Nature Communications | (2023) 14:3216 3
Content courtesy of Springer Nature, terms of use apply. Rights reserved

be coated uniformly. After spin-coating process, NVP could still sur-
round the perovskite crystal seeds at the initial nucleation step owing
to strong interaction between NVP and perovski te precursors (PbI
2
and
FA+) proved by 1H NMR, IR spectra, and a solubility experiment as
shown in Supplementary Fig. 5 and Supplementary Discussion 1.
During annealing process, NVP gradually polymerized into PVP and
still surrounded the perovskite crystal grains, and some uncoordinated
miscellaneous phases were also solidified at the GBs. As PVP is inso-
luble in DMSO, but soluble in CB (Fig. 2b). Therefore, after spin-
coating the HTL/CB solution, CB removed the residual miscellaneous
phases and opened the GBs to form monolithic perovskite grains,
while remaining some PVP undissolved on the bottom as the dielectric
layer. At the same time, HTL layer was covered on the perovskite films
forming the unique M-T structures, which providedan obviously larger
contact area between perovskite and HTL.
To verify our speculation, the scanning electron microscopy
(SEM) images of control perovskite film and perovskite/NVP film were
measured before and after CB extraction. HTL/CB solution was
replaced with pure CB solvent, as solutions containing solute would
affect the resolution of SEM surface tests. Itcould be observed that the
miscellaneous phases (irregular bright domains) were invariably exis-
ted at perovskite grain boundaries before and after CB extraction;
besides, pinholes and cracks were formedafter CB extraction (Fig. 2c).
For the perovskite/NVP film (Fig. 2d), the PVP and the miscellaneous
phases on the perovskite and in the boundaries were effectively
washed away after CB extraction, leaving distinct boundary gaps
between perovskite grains, and eventually forming monolithic per-
ovskite grains. Small molecular N-methyl-2-pyrrolidone (NMP) and PVP
polymer were used as additives in perovskite films. As shown in Sup-
plementary Fig. 7, the unpolymerized NMP additives showed a similar
morphology as the control perovskite film. In contrast, the PVP poly-
mer additive showed very small and disordered crystallization of
perovskite due to rapid precipitation during annealing. Both NMP and
PVP could not form the same morphological structure as NVP after CB
extraction.
We further probed the chemical structure on the surface of the
CB-extracting perovskite films by X-ray photoelectron spectroscopy
(XPS) (Fig. 2e). The main peaks of Pb 4f
5/2
and Pb 4f
7/2
of perovskite/
NVP film shifted by 0.54 eV towards lower binding energy compared to
the control sample. In addition, after CB extracting, the Pb0peaks,
which was considered to be the origin of deep defect energy level, was
clearly observed in the control perovskite film at binding energy of
141.35 and 136.52 eV, whereas they completely disappeared in the
perovskite/NVP films. This indicated that there was still residual PVP
covered the perovskite film or in the GBs, and the lone pair electrons of
the carbonyl group on PVP could effectively coordinate with Pb ions45.
Similarly, as shown in Supplementary Fig. 8, a shift to lower binding
energy was also observed in the I 3d spectra. The control film exhibited
a C-C=O signal related to oxygen/moisture in perovskite at 288.25eV21,
which was eliminated by adding NVP, in meanwhile, both new C=O
signal at 287.63 eV and new C–N signal at 285 .53 eV were detected due
to carbonyl and pyrrole units on NVP molecule. In addition, perovskite
films with different molar ratios of NVP addition from 15% to 60% were
also measured by XPS as shown in Supplementary Fig. 8. Further shift
to lower binding energy indicated that the passivation effect improved
with increasing content of NVP. Furthermore, as shown in Fig. 2f,
perovskite/NVP film exhibited a longer perovskite lifetime (1138.42 ns)
and lower trap density (1.58 × 1015cm−3) compared to the control film
(375.69 ns and 8.11 × 1015cm−3) measured by time-correlated single
photon counting (TCSPC) and fitted by a generic kinetic model46,47.
Supplementary Fig. 9 showed that with increasing the excitation
density, the PL quantum efficiencies (PLQEs) gradually reached a
maximum value owing to the filling of defects. Perovskite/NVP film
showed a higher PLQE values with a maximum of 10.20% compared to
the control film (8.38%). TCSPC and PLQE measurement confirmed
that perovskite/NVP film with the monolithic grain and passivation
effect synergistically suppressed the defect-related non-radiative
recombination.
Balanced charge extraction and increased built-in electric field
Atomic force microscopy (AFM, Fig. 3a) and PeakForce tunneling
atomic force microscopy (PF-TUNA, Fig. 3b) images were measured
under a bias voltage of 5 V to provide synchronous morphology
roughness and spatially resolved electronic properties of control and
perovskite/NVP films after CB extracting. The AFM topography of
perovskite/NVP film depicted a higher altitude intercept (486 nm) than
the control counterpart (130 nm) (Fig. 3a) owing to the M-T structure.
The PF-TUNA image of perovskite/NVP film was brighter than that of
the control sample, indicating that more conducting current flow
through the perovskite layer in the former than in the latter, which was
generally attributed to the increased conductivity and thus enhanced
the charge transport48. The improved charge transport in overall per-
ovskite/NVP films was ascribed to the high-quality of monolithic per-
ovskite grains. Furthermore, the GBs of perovskite/NVP film showed
brighter contrast in Fig. 3b, indicating that they carried current more
efficiently. This was probably due to the thinner thickness of the filmat
GBs. Therefore, in this case, GBs might contribute to improving PV
performance because it facilitates charge transport rather than acting
as a recombination center as it usually does22. To better understand the
charge extraction enhancement induced by M-T structure, device
simulations were carried out using the commercial software Silvaco. A
device with a perovskite grain size of 780 nm was designed for simu-
lation (Supplementary Fig. 10, Supplementary Tables 1 and 2). GB
grooves were assumed to be 390 nm and 560 nm deep, which was
consistent with the STEM measurements (Fig. 1a). To simulate the
effect of HTL extraction ability in PSCs, a 100 nm Spiro-OMeTAD was
added onto the surface of perovskite film. High hole currents con-
ducted from the GBs in the perovskite filmwith M-T structurecould be
clearly observed, which also c onfirmed the PF-TUNA measurement. To
further verify the effect of M-T structure on the charge extraction of
the perovskite film, we investigated the time-correlated single photon
counting (TCSPC) characterizations on both perovskite/NVP and the
control samples, with glass/FTO/ETL/PSK/HTL configurations (Fig. 3c,
d and Supplementary Table 3). For the HTL side incidence test (Fig.
3c), the perovskite/NVP sample presented almost 15 times faster
photo-induced luminescence lifetime (1.51ns) than the control coun-
terpart (22.42 ns). To further verify the origin of the enhancement of
the hole extraction, ultraviolet photoelectron spectroscopy (UPS) was
carried out to detect the interfacial energy level structure of the per-
ovskite films. As shown in Supplementary Fig. 11, the control and NVP-
based perovskite film exhibited comparable valence band maximum.
Therefore, the effective enhancement on hole extraction came mainly
from the M-T structural contact between perovskite and HTL. On the
other hand, the ETL side incidence test (Fig. 3d) demonstrated similar
lifetime for both interdigital (1.23 ns) and planar (1.37ns) samples. The
charge extraction balance beneficial from M-T architecture has the
potential to achieve high-performance PSCs, particularly in terms of
enhancing V
OC
and FF.
Surface contact potential difference (CPD) was carried out by
Kelvin probe force microscopy (KPFM) with a schematic illustration in
Fig. 3e, f and Supplementary Figs. 12 and 13. The height sensor dia-
grams (Fig. 3e) confirmed that both perovskite/NVP (667 nm) and
control perovskite (630 nm) films were similar in thickness to each
other. The corresponding surface CPD (Fig. 3f, g) across the per-
ovskite/SnO
2
interface exhibited that perovskite/NVP film (288 mV)
had a bigger CPD value than that of the control film (167 mV),
respectively. In addition, Mott-Schottky measurements on entire
devices consisting of glass/FTO/SnO
2
/perovskite/spiro-OMeTAD/Au
were carried out to investigate this effect on the built-in electric field
(V
bi
)(Fig. 3h),inwhichNVP-basedPSCrepresentedahigherV
bi
value
Article https://doi.org/10.1038/s41467-023-38926-3
Nature Communications | (2023) 14:3216 4
Content courtesy of Springer Nature, terms of use apply. Rights reserved

(0.89 V) than the control counterpart (0.75 V). The space charge
depletion width (W) could be calculated from a plot of C−2versus V.The
NVP-based PSC demonstrated more than doubled Wvalue (55.1 nm)
than control PSC (24.1 nm), owing to the larger expanded spacecharge
depletion region of the NVP-perovskite. Enhanced V
bi
and Wcouldwell
facilitate charge separation and prevent carrier recombination, which
was expected to realize high-performance photovoltaics, especially on
improving V
OC
49.
Photovoltaic performance and stability of PSCs
We further explored the photovoltaic performance of PSCs with
addition of NVP. PSCs were fabricated with conventional structures of
glass/FTO/SnO
2
/perovskite(FAPbI
3
)
0.95
(MAPbBr
3
)
0.05
/spiro-OMeTAD/
Au. Different molar ratios of NVP (15%, 30%, 60%, and 100%) have been
used in the perovskite to obtain optimized PSCs, the current density-
voltage (J–V) characteristics of PSCs were recorded under AM 1.5 G
simulated solar illumination of 100 mW cm−2as shown Supplementary
Fig. 14 and Supplementary Table 4. Among different NVP addition
ratios, PSCs with 30 mol% NVP exhibited the highest device perfor-
mance with the longest perovskite lifetime measured by TCSPC
(Supplementary Fig. 15 and Supplementary Table 5). SEM images in
Supplementary Fig. 16 clearly showed that further increasing the
proportion of NVP to 60% and 100%, large amount of NVP addition
probably affected the crystallization of perovskite, and thus decreased
the device efficiency. As shown in Fig. 4a, the control device had a
maximum power conversion efficiency PCE of 22.91% with a V
OC
of
1.151V, short circuit current density (Jsc) of 25.21 mA cm−2and a FF of
78.94%. The champion PSCs with 30mol% NVP showed a maximum
PCE of 24.69% with a V
OC
of 1.195 V, Jsc of 25.77 mAcm−2and a FF of
80.19%. To verify the PCE, our best NVP-based device was certificated
by an independent certification laboratory (PWQC, China). This device
was tested using a shadow mask with a certified size of 8.925 mm2
(Supplementary Fig. 17). SupplementaryFig. 18 represented a certified
PCE of 24.25% with V
OC
of 1.188 V, J
SC
of 25.41 mA cm−2and FF of 80.35%
at foward scan (FS), and the reverse scan (RS) shows a certified PCE of
24.55%, with V
OC
of 1.187 V, J
SC
of 25.66 mA cm−2and FF of 80.64%. The
negligible hysteresis result both in-house and certificate authority
confirmed the M-T architecture of the PSCs in balancing hole and
electron extraction in PSCs. It was worth noting that the NVP-based
PSC exhibited excellet V
OC
and FF than the control device, which was
attributed to the better crystallization of perovskite and balanced
charge extraction of the M-T structural device. Electrochemical
impedance spectra (EIS) in Supplementary Fig. 19 and Supplementary
Table 6 showed that the NVP-based PSCs had a larger recombination
w/NVP
130 nm
0 nm
1 μm
Control
486 nm
0 nm
5.0 pA 5.0 pA
w/NVP
SnO
Control
ΔH = 667 nm
w/NVP SnO
ΔH = 630 nm
ΔE = 288 mV
w/NVP
Control
ΔE = 167 mV
0246810
0.0
0.1
0.2
0.3
Distance (μm)
CPD (V)
Control
w/NVP
0.4 0.6 0.8 1.0 1.2
0
10
20
30
40
50
Voltage (V)
C
-2
(10
16
cm
4
F
-2
)
Control
w/NVP
a
b
c
d
e
f
gh
-4.0 pA -4.0 pA
Control
+
+
+
hv
−
−
−
hv
1 μm
1 μm
1 μm
1μm 1μm
1μm
1μm
0 50 100 150 200
10
-3
10
-2
10
-1
10
0
Time (ns)
PL Intensity (arb. units)
Control
w/NVP
0 50 100 150 200
10
-3
10
-2
10
-1
10
0
Time (ns)
PL Intensity (arb. units)
Control
w/NVP
Fig. 3 | Optoelectronic properties of the control and perovskite/NVP films.
aHeight images of perovskite films (5 × 5 μm). bPF-TUNA images of perovskite
films ( 5 × 5 μm). TCSPC spectra of perovskite films were excited with laser (exci-
tation: 520 n m, 55 nJ cm −2,1MHz)cfrom the HTL side and d, from the ETL side.
eSurfaceheight imagesand f, KPFM images ofthe interface between the perovskite
and SnO
2
layers. gCorresponding values of surface contact potential difference
(CPD). hPlots of C−2versus applied voltage by Mott–Schottky analysis in control
and NVP-based perovskite solar cells.
Article https://doi.org/10.1038/s41467-023-38926-3
Nature Communications | (2023) 14:3216 5
Content courtesy of Springer Nature, terms of use apply. Rights reserved

resistance (R
rec
) than the control device, which stems from the
reduced non-radiative Recombination50. We also simulated the pho-
tovoltaic performance of the M-T structured PSC and the control
device. We found that the simulated performance was very similar to
our experimental results (Supplementary Table 1) with significant
improvements in the V
OC
and FF of the devices. The monochromatic
incident photon-to-electron conversion efficiency (IPCE) spectra
showed that the integrated J
SC
values (<5% deviation) matched the J–V
measured data (Fig. 4b). Moreover, the NVP-based PSC exhibited
steady power outputs at maximum power point (Supplementary Fig.
20) with an average PCE of 23.7% compared to their control counter-
parts of 22.3% (Supplementary Fig. 21). The larger PSC (1.0 cm2)
showed PCEs up to 23.06% (RS) and 22.71% (FS), indicating that this
M-T architecture of PSC is scalable (Fig. 4c).
By using poly[bis(4-phenyl)(2,4,6-trimethylphenyl)amine] (PTAA)
instead of Spiro-OMeTAD as the hole transporting materials21,51–55,the
NVP-based PSC exhibited an excellent light exposure stability with <5%
PCE loss after 1100 h at 1-sun illumination at 65 °C in N
2
atmosphere
(ISOS-L-2 protocol) (Fig. 4d)56. In contrast, the control PSC started to
decompose dramatically until the PCE loss reached 40% after 400 h.
Furthermore, we also conducted the accelerated damp-heat tests of
encapsulated PSCs subjected to 85% relative humidity (RH) and 85°C,
in damp-heat chamber (ISOS-D-3 protocol)56 (Fig. 4e), where the
encapsulated NVP-based PSC kept >90% initial efficiency over 1100 h.
This superior stability might be attribute to the monolithicperovskite/
NVP grains covered by polymerized network of PVP, which also con-
firmed by XPS measurement discussed above. To evaluate whether the
polymer network would suppress lead leakage from the device,
unencapsulated PSCs with different ratio of NVP were immersed into
deionized water for 1 hour. The leadconcentration of the solution was
analyzed via atomic absorption spectrometer (AAS) as shown in Sup-
plementary Fig. 22. The Pb concentration of the control PSC was
4.86ppm, and as the molar ratio of NVP increased from 15% to 100%,
the Pb concentration decreased to 1.85 ppm, which was one-fourth of
that in the control counterpart. Furthermore, owing to the excellent
solubility of NVP for perovskite precursor, unencapsulated PSCs used
pure NVP as the only solvent to fabricate perovskite film. The unen-
capsulated PSC demonstrated PCE of 17.24% and maintained >80% of
its initial PCE over 600 minutes, and the control device degraded
immediately (Supplementary Fig. 23). Water-vapor test provided a
better visualization of the role of perovskite with NVP addition in
retarding perovskite decomposition. We fumigated the unencapsu-
lated control and perovskite/pure NVP films in hot water vapor and
could see that the control films rapidly turned yellow while the NVP
films remained in the black phase (Fig. 4f and Supplementary Movie 1).
Discussion
Herein, we demonstrate an effective strategy of using a thermal poly-
merized NVP-perovskite/HTL composite film to form the monolithic
perovskite grains and a unique M-T structure, which facilitate
improved V
OC
and FF of PSCs due to the suppressed non-radiative
recombination and balanced charge extraction. The corresponding
PSCs exhibited excellent certified PCE up 24.55% (RS) and 24.25% (FS).
The negligible hysteresis is attributed to the balance of hole and
electron extraction in the PSC due to the M-T structure. Furthermore,
NVP-based PSCs with polymerized network exhibited excellent illu-
mination, moisture, andthermal stability in accordance with the ISOS-
L-2 and ISOS-D-3 protocol. Our work highlights the role of increasing
0.00.20.40.60.81.0
1.2
0
5
10
15
20
25
Voltage (V)
w/NVP- Reverse
w/NVP- Forward
0.0 0.2 0.4 0.6 0.8 1.0 1.2
0
3
6
9
12
15
18
21
24
27
Voltage (V)
Control-Reverse
Control-Forward
w/NVP- Reverse
w/NVP- Forward
300 400 500 600 700 800 900
0
20
40
60
80
100
Control
w/NVP
Wavelength (nm)
EQE (%)
0
5
10
15
20
25
0 100 200 300 400 500 600 700 800 900 1000 1100
50
60
70
80
90
100
Nor. PCE (%)
Time (hour)
w/NVP
Control
0 100 200 300 400 500 600 700 800 900 1000 1100
50
60
70
80
90
100
Nor. PCE (%)
Time (hour)
w/NVP
Control
Intergrated Jsc (mA cm
-2
)
Current density (mA cm
-2
)
Current density (mA cm
-2
)
VOC JSC FF PCE
[V] [mA cm-2][%] [%]
1.151
RS
FS
RS
FS
Control
w/NVP
25.21 78.94 22.91
1.139 25.16 76.92 22.04
1.195 25.77 80.19 24.69
1.194 25.76 79.94 24.59
VOC JSC FF PCE
[V] [mA cm-2][%] [%]
1.187
RS
w/NVP FS
24.51 79.29 23.06
1.182 24.46 78.52 22.71
ISOS-D-3
ISOS-L-2
1-sun illumination , 65°C, N
85% RH and 85°C, in damp-heat chamber
0s
6s 25s
4s
abc
d
e
f
2s
8s
Control
w/NVP
Fig. 4 | Device performance and stability. a J−Vcurves and photovoltaic para-
metersof champion devices of the control and NVP-based PSCsmeasured in-house.
bRepresentative EQEs and integrated J
SC
valuesof the control andNVP-based PSCs.
cJ−Vcurves of a larger-area NVP-based PSC. The active area is 1.0 cm2,defined by a
mask aperture. The inset is a photograph of the device with dotted outlines
representing the active areas.dNormalized PCE of encapsulated PSCs according to
ISOS-L-2 protocol (1-sun illumination and 65 °C, in N
2
atmosphere). eNormalized
PCE of encapsulated PSCsaccording to the ISOS-D-3 protocol (85% RHand 85 °C, in
a damp-heatchamber). All error bars represent the standard deviation of six
devices. fScreenshots of unencapsulated control andperovskite/NVPfilms in w ater
steam test (100% RH and 100°C) captured from Movie S1.
Article https://doi.org/10.1038/s41467-023-38926-3
Nature Communications | (2023) 14:3216 6
Content courtesy of Springer Nature, terms of use apply. Rights reserved

the contact area of perovskite layer and HTL and proposes a unique
strategy to realize interfacial charge-extracting balance in PSCs, which
may be an important approach to achieve efficient device in the future.
Methods
Materials
Formamidinium iodide (FAI, ≥99.5%) and methylammonium bromide
(MABr, ≥99.5%) were purchased from Hangzhou Perovs Optoelectronic
Technology Corp (China). Methylammonium chloride (MACl, ≥99.5%),
lead bromide (PbBr
2
, 99.99%), 2,2’,7,7’-tetrakis (N,N-di-pmethox-
yphenylamine)−9,9’-spirobifluorene (spiro-OMeTAD, 99.5%), lithium
bis(trifluoromethanesulfonyl)imide salt (Li-TFSI, ≥99%), FK209-Co(III)-
TFSI (≥99%) were purchased from Xi’an Polymer Light Technology Corp
(China). Lead iodide (PbI
2
99.99%), N-vinyl-2-pyrrolidinone (NVP, ≥99%),
azodiisobutyronitrile (AIBN, ≥98%), N,N-dimethylformamide (DMF,
>99.5%), dimethyl sulfoxide (DMSO, >99.0%), chlorobenzene (CB,
>98.0%), ethyl acetate (EA, >99.5%), isopropanol (IPA, >99.5%), acetoni-
trile (ACN, >99.5%), urea (>99.0%), and 4-tert-butyl-pyridine (TBP,
>96.0%) were purchased from TCI Shanghai (China). Stannous chloride
SnCl
2
·2H
2
O (99.99%), thioglycolic acid (TGA, 98%), and urea (≥99.5%)
were purchased from Sigma‐Aldrich (USA). All materials were used as
received without further modification.
Instruments
Fourier transform infrared spectroscopy (FTIR) was performed using a
Scientific Nicolet iS50 spectrometer (Thermo, America). Scanning
electron microscopy (SEM) was performed using JSM-7800F (JEOL,
Japan). Transmission electron microscopy (TEM) was performed using
JEM-1400PLUS (JEOL, Japan). GIWAXS experiments were performed on
SAXS/WAXS beamline at the Australian Synchrotron. X-ray photoelec-
tron spectroscopy (XPS) measurement was carried out on a Thermo-
Fisher ESCALAB 250Xi system with a monochromatized Al Kα(for XPS
mode) under a pressure of 5.0 × 10−7Pa. Focused-ion-beam (FIB) was
performed on Helios G4 CX (FEI, USA). Scanning transmission electron
microscope (STEM) was performed on Themis Z (FEI, USA). The time-
resolved PL measurements were performed by a combination of a
TimeHarp 260 PICO module (PicoQuant), aiHR320 spectrometer (Hor-
iba),andCOUNT-100T-FCsinglephotoncountingmodules(Laser
Components GmbH). PeakForce Tunneling AFM (PF-TUNA) and Kelvin
probe force microscopy (KPFM) measurements were operated with the
probe model SCM-PIT-V2 (material: 0.01–0.025 Ohm-cm Antimony (n)
doped Si, cantilever: T=2.8μm, L=225μm, W=35μm, f
c
=75kHz,
k= 3N/m, coating: front: conductive PtIr, back: reflective PtIr) with the
DimensionFastScanAFMsystem,Bruker Corporation. Mott-Schottky
measurement was recorded using CHI760E electrochemical workstation
(CH Instruments Ins, USA). Current density-voltage (J−V) curves in-house
were measured using a class 3 A solar simulator (XES-40S3, SAN-EI)
under AM 1.5 G standard light equipped with a Keithley 2400 source
meter. The incident photon-to-electron conversion efficiency (IPCE)
measurements were carried out by a QE-R-900AD system (Nanjing Ouyi
Optoelectronics Technology). The standard silicon solar cell (QE-B1)
calibrated by NIM was used to calibrate the light intensity to AM 1.5 G
irradiance (100 mW/cm2). The device was measured and certified at the
Quality Testing Center for Photovoltaic and Wind Power Systems of the
Chinese Academy of Sciences (test report No. PWQC-WT-P21110821-1R).
This certified device was tested with a shadow mask with a certified size
of 8.925 mm2provided by the National Institute of Metrology, China
(testreportNo.CDjc2021-15963).
Preparation of solubility measurement
To measure the solubility of NVP in the perovskite precursor, 1.474M
(FAPbI
3
)
0.95
(MAPbBr
3
)
0.05
precursor was dissolved in 1 mL of pristine
NVP solvent (0.3 wt.% AIBN),then heated to 60 °C and stirred for 2 h in
anitrogen-filled glovebox. The perovskite precursor in polymerizing
NVP showed a viscous fluidity and adhered to the vial wall. Upon
stirring at 60°C for another 2 h in the glovebox, the perovskite pre-
cursor in polymerized NVP became an immobilizing gel. For solubility
and miscibility tests, all samples were dissolved in 1 mL of DMSO each.
For measuring the polymerized NVP solubility in CB, 1 mL of neat NVP
solvent each was heated at 60°C for 4h to avoid precipitation of the
perovskite precursor in CB anti-solvent. Afterward, the polymerized
NVP was completely dissolved in 1 mL of CB.
Chemical bath deposition of SnO
2
layer [33]
The FTO glass was cleaned ultrasonically for 20 min with detergent,
pure water, and ethanol, respectively. Then they were dried with a
streamof dry nitrogen,followed by treatment with UVO for 15 min. The
compact SnO
2
film was achieved by chemical bath deposition (CBD).
The CBD solution was prepared by mixing 1.25 g of urea, 1.25 mL of
HCl, 25 μLofTGA,and275mgofSnCl
2
·2H
2
Oper100mLoficedeio-
nized (DI) water to form a 0.012 M solution. The cleaned FTO glass was
soaked in the diluted SnCl
2
·2H
2
O solution (0.002 M) for 2 h at 90 °C
and cleaned via sonication with DI water and IPA for 5 min each. It was
then annealed at 170 °C for 1 h, followed by spin-coating with 20 mM
KCl in DI water at 3000 rpm for 30 s (2000 rpm ramp) and annealing at
100 °C for 10 min.
Preparation of the perovskite layer
The perovskite precursor of (FAPbI
3
)
0.95
(MAPbBr
3
)
0.05
was prepared
by dissolving 240.76 mg FAI, 706.9 mg PbI
2
, and 33.76 mg MACl,
8.21 mg MABr, 27.05 mg PbBr
2
salts in DMF/DMSO (8:1 v/v) mixed
solvent in 1.5 M concentration. For the NVP additive system, the
molar ratios of perovskite to NVP (with 0.3 wt.% AIBN) were 15%,
30%, 60%, and 100%. For typical measurements, 30 mol% was
adopted. For the pure NVP system, pure NVP was used as the only
solvent to dissolve the perovskite precursors. The perovskite solu-
tion was deposited on CBD-SnO
2
/FTO by two consecutive spin-
coating steps of 1000 rpm for 10s and 5000 rpm for 30 s, respec-
tively. During the second spin-coating step (5000 rpm), 100 μLofEA
was quickly poured onto the substrate after 20 s. The films were then
annealed at 100 °C for 1 h.
Preparation of the hole-transport layer
The hole-transport layer was prepared by dissolving 30 μL of TBP,
18 μL of Li-TFSI solution (520 mg Li-TFSI in 1.0 mL acetonitrile), 29 μL
of FK209-Co (III)-TFSI solution (300 mg FK209-Co (III)-TFSI in 1.0mL
acetonitrile), and 73 mg of spiro-OMeTAD in 1.0 mL CB. The hole-
transport layer was deposited on perovskite films by spin-coating a
spiro-OMeTAD solution at 3000 rpm for 30 s. For the device thermal
stability test, PTAA doped with 4-Isopropyl-4’-methyldiphenyliodo-
nium Tetrakis (pentafluorophenyl) borate (TPFB) was used to replace
spiro-OMeTAD as the hole-transport layer. The concentration of PTAA
was 30 mg mL−1and the weight ratio of PTAA/TPFB was 10:1. The PTAA
was deposited on top of the perovskite layer at a spin rate of
3000 rpm. for 30 s.
Sample preparation for SEM, XPS, PF-TUNA, KPFM, and GIWAXS
measurement
On the FTO/SnO
2
substrate, 20 μL of perovskite precursor solutions
(with/without NVP) were spin-coated respectively and annealed at
100 °C for 1 h. Afterward, 100 μLofneatCBsolventwasdroppedonto
the annealed film, followed by spin-coating at 3000 rpm for 30 s.
These CB-extracted perovskite films were then used for
measurements.
Sample preparation for TEM
On the FTO/SnO
2
substrate, 20 μL the perovskite precursor was spin-
coated and annealed at 100°C for 1 h. the perovskite powder sample
was scraped from the perovskite film and dispersed in n-hexane. Then
directly used for TEM measurement.
Article https://doi.org/10.1038/s41467-023-38926-3
Nature Communications | (2023) 14:3216 7
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Sample preparation for cross-sectional STEM and TCSPC
measurement
On the FTO/SnO
2
substrate, 20 μL of perovskite precursor was spin-
coated and annealed 100 °C for 1 h. Then HTL in CB was spin-coated on
the cooled samples. For the NVP system, the HTL/CB solution was
dropped on perovskite film and followed by a spin-coating process at
3000 rpm for 30 s. TCSPC spectra were tested separately from the
Spiro-OMeTAD side or SnO
2
side of the sample. For the preparation of
cross-sectionsamples,wepre-cuttheFTOsubstratewithaglasscutter
before film deposition and broke the sample after film deposition to
obtain vertical, ordered cross-sectional samples.
Focused-ion beam (FIB) for perovskite/NVP cross-sections for
scanning transmission electron microscopy (STEM)
The NVP-based perovskite was transferred to the chamber of the FEI
Helios G4 CX dual-beam system and focused using the SEM model.
8μm×3μm platinum was sputtered onto the perovskite/NVP sample
to protect the internal morphology of the target slice. The sample was
dug with a Ga+ion beam into two 4-μm deep holes adjacent to the
target zone. A tungsten needle was welded to the target slice by
sputtering platinum in between. The target zone was then inclined and
cut off the bottom and side connections to sample substrate by Ga+ion
beam. The tungsten needle transferred target slice to a TEM sample
holder and welded together by sputtering platinum. Finally, the
thickness of target slice was further reduced to 100 nm by using low-
powered Ga+ion beam from the top platinum side, resulting in the
ultra-thin target sample for HAADF-STEM.
High-angle annular dark-field (HAADF) STEM
The structures of the NVP-based perovskite were characterized using a
FEI Talos F200X microscope in STEM mode at 200kV, equipped with
an EDS detector and a high-angle annular dark-field (HAADF) detector.
Ultra-thin samples for STEM were prepared using FIB in a FEI Helios G4
CX dual beam microscope.
Time-correlated single photon counting
TCSPC spectra of perovskite films were excited with 520 nm laser
(55 nJ cm−2fluence and 1 MHz repetition rate) impinged on either the
HTL or ETL side.
PSC fabrication for PCE measurements
TheperovskitesolutionwasdepositedonCBD-SnO
2
/FTO by two
consecutive spin-coating steps of 1000 rpm and 5000 rpm for 10 s and
30 s, respectively. During the second spin-coating step (5000rpm),
100 μL of EA was quickly poured onto the substrate after 20 s. The
films werethen annealed at 100 °C for 1 h. The hole-transport layer was
deposited by spin-coating a spiro-OMeTAD solution at 3000 rpm for
30 s. Finally, the gold electrode (80nm) was deposited by thermal
evaporation.
Device characterization
Current density-voltage (J−V) curves were obtained using a solar
simulator (class 3 A, XES-40S3, SAN-EI) under AM 1.5 G standard light
equipped with a Keithley 2400 source meter. The standard silicon
solar cell (QE-B1) calibrated by NIM was used to calibrate the light
intensity to AM 1.5 G irradiance (100 mW/cm2). A shadow mask was
used to define the effective aperture area of the device to 0.09 cm2.
PSCs with an 0.1 cm2active area were fabricated by evaporating top-
electrodes with mask sizes of 0.32 cm × 0.32 cm. PSCs with a 1.0 cm2
active area were fabricated by evaporating top-electrodes with mask
sizes of 1.1 cm × 1.4 cm and then measured using a 1.0 cm × 1.0 cm
aperture mask. All materials, solutions, and preparation processes
for large-area perovskite films were the same as for the small-
sized PSCs.
Encapsulated PSC stability measurements
PSCs were encapsulated by covering a glass slide (overlap size:
1.2 cm ×1.2 cm) on the device (substrate size: 1.4 cm ×1.4 cm, device
area: 0.32 cm × 0.32 cm) with UV-curable resin (ThreeBond 3042B) for
stability measurements. For ISOS-L-2 protocol, encapsulated PSCs
were continuously illuminated under a solar simulator and heated on
hot-plate at 65 °C in a N
2
glovebox. For the ISOS-D-3 protocol, encap-
sulated PSCs were stored in a damp-heat chamber (BPS-50CL, Shang-
hai BluePard, China) with a setting temperature of 85 °C and a relative
humidity of 85%. PSCs were periodically measured (once per day
except Sunday for 7 weeks) using a solar simulator (class 3A, XES-
40S3, SAN-EI).
Unencapsulated PSC stability measurement and Water-vapor
measurement
The damp-heat (85 °C, 85% relative humidity) tests under dark condi-
tions by unencapsulated PSCs fabricated with NVP as the only solvent
device were performed using in the environment test chamber (BPS-
50CL, Shanghai BluePard, China) for 720 min.
For water-vapor test, 500 mL of deionized water was added to a
beaker and heated to boiling on a hot plate, and the unencapsulated
control and perovskite/pure NVP films were fumigated in hot water
vapor to observe the film changes.
Lead leakage measurement
To measure the lead leakage from the perovskite devices, the unen-
capsulated control and PCSs with 15% NVP, 30% NVP, 60% NVP, and
100% NVP were immersed in glass vials containing 10 mL of deionized
water, respectively, and the devices were removed after 60min. All
samples in the glass vials were analyzed by flame atomic absorption
spectrometer (iCETM 3500, Thermo Fisher).
Reporting summary
Further information on research design is available in the Nature
Portfolio Reporting Summary linked to this article.
Data availability
All data generated in this study are provided in the article and Sup-
plementary Information and the raw data supporting this study are
available from the Source Data file. Source data are provided with
this paper.
References
1. Ashworth, C. Reproducible, high-performance perovskite solar
cells. Nat. Rev. Mater. 6,293–293 (2021).
2. Kim,J.Y.,Lee,J.W.,Jung,H.S.,Shin,H.&Park,N.G.High-efficiency
perovskite solar cells. Chem. Rev. 120,7867–7918 (2020).
3. Juarez-Perez, E. J. & Haro, M. Perovskite solar cells take a step for-
ward. Science 368, 1309 (2020).
4. Best Research-Cell Efficiency Chart (NREL, 2023); https://www.
nrel.gov/pv/interactive-cell-efficiency.html
5. Wang, K. et al. Overcoming shockley-queisser limit using halide
perovskite platform? Joule 6,756–771 (2022).
6. Jeon, N. J. et al. Compositional engineering of perovskite materials
for high-performance solar cells. Nature 517,476–480 (2015).
7. Saliba, M. et al. Incorporation of rubidium cations into perovskite
solar cells improves photovoltaic performance. Science 354,
206–209 (2016).
8. Yang, W. S. et al. Iodide management informamidinium-lead-
halide–basedperovskite layers for efficientsolar cells. Science 356,
1376–1379 (2017).
9. Li, M. et al. Orientated crystallization of FA-based perovskite via
hydrogen-bonded polymer network for efficient and stable solar
cells. Nat. Commun. 14,573(2023).
Article https://doi.org/10.1038/s41467-023-38926-3
Nature Communications | (2023) 14:3216 8
Content courtesy of Springer Nature, terms of use apply. Rights reserved

10. Hu, Q. et al. In situ dynamic observations of perovskite crystal-
lisation and microstructure evolution intermediated from [PbI
6
]4-
cage nanoparticles. Nat. Commun. 8,15688(2017).
11. Qin, M. et al. Manipulating the mixed-perovskite crystallization
pathwayunveiledbyinsituGIWAXS.Adv. Mater. 31,
e1901284 (2019).
12. Zhang, K. et al. A prenucleation strategy for ambient fabrication of
perovskite solar cells with high device performance uniformity. Nat.
Commun. 11, 1006 (2020).
13. Song, J. et al. Manipulating the crystallization kinetics by additive
engineering toward high‐efficient photovoltaic performance. Adv.
Funct. Mater. 31, 2009103 (2021).
14. Bi, C. et al. Non-wetting surface-driven high-aspect-ratio crystalline
graingrowthforefficient hybrid perovskite solar cells. Nat. Com-
mun. 6,7747(2015).
15. Lin, Y. et al. Unveiling the operation mechanism of layered per-
ovskite solar cells. Nat. Commun. 10,1008(2019).
16. Bai,Y.etal.Initializingfilm homogeneity to retard phase segrega-
tion for stable perovskite solar cells. Science 378,747–754 (2022).
17. Chen, B., Rudd, P. N., Yang, S., Yuan, Y. & Huang, J. Imperfections
and their passivation in halide perovskite solar cells. Chem. Soc.
Rev. 48,3842–3867 (2019).
18. Zhao, L. et al. Enabling full-scale grain boundary mitigation in
polycrystalline perovskite solids. Sci. Adv. 8, eabo3733 (2022).
19. Chang, Q. et al. Ferrocene-induced perpetual recovery on all ele-
mental defects in perovskite solar cells. Angew. Chem. Int. Ed. Engl.
60,25567–25574 (2021).
20. Yang, Y. et al. Suppressing vacancy defects and grain boundaries
via ostwald ripening for high-performance and stable perovskite
solar cells. Adv. Mater. 32,e1904347(2020).
21. Jiang, Q. et al. Surface passivation of perovskite film for efficient
solar cells. Nat. Photonics 13, 460–466 (2019).
22. Son, D.-Y. et al. Self-formed grain boundary healing layer for highly
efficient CH
3
NH
3
PbI
3
perovskite solar cells. Nat. Energy 1,
16081 (2016).
23. Zong, Y. et al. Continuous grain-boundary functionalization for
high-efficiency perovskite solar cells with exceptional stability.
Chem 4,1404–1415 (2018).
24. Liu, K. et al. Moisture-triggered fast crystallization enables efficient
and stable perovskite solar cells. Nat. Commun. 13, 4891 (2022).
25. Lu, H. et al. Vapor-assisted deposition of highly efficient, stable
black-phase FAPbI3 perovskite solar cells. Science 370,74
(2020).
26. Sun, R. et al. Over 24% efficient poly(vinylidene fluoride) (PVDF)-
coordinated perovskite solar cells with a photovoltage up to 1.22 V.
Adv. Funct. Mater. 33, 2210071 (2023).
27. Li, H. et al. Sequential vacuum-evaporated perovskite solar cells
with more than 24% efficiency. Sci. Adv. 8, eabo7422 (2022).
28. McMeekin, D. P. et al. Intermediate-phase engineering via dime-
thylammonium cation additive for stable perovskite solar cells. Nat.
Mater. 22,73–83 (2023).
29. Jeong, J. et al. Pseudo-halide anion engineering for alpha-FAPbI3
perovskite solar cells. Nature 592,381–385 (2021).
30. Park, B.-w et al. Stabilization of formamidinium lead triiodide α-
phase with isopropylammonium chloride for perovskite solar cells.
Nat. Energy 6,419–428 (2021).
31. Zhou, H. et al. Interface engineering of highly efficient perovskite
solar cells. Science 345,542–546 (2014).
32. Jiang, Q. et al. Enhanced electron extraction using SnO
2
for high-
efficiency planar-structure HC(NH
2
)
2
PbI
3
-based perovskite solar
cells. Nat. Energy 2, 16177 (2016).
33. Yoo, J. J. et al. Efficient perovskite solar cells via improved carrier
management. Nature 590,587–593 (2021).
34. Peng, J. et al. Centimetre-scale perovskite solar cells with fill factors
of more than 86 per cent. Nature 601,573–578 (2022).
35. Wang, F. F. et al. Materials toward the upscaling of perovskite solar
cells: progress, challenges, and strategies. Adv. Funct. Mater. 28,
1803753 (2018).
36. Rombach, F. M., Haque, S. A. & Macdonald, T. J. Lessons learned
from Spiro-OMeTAD and PTAA in perovskite solar cells. Energy
Environ. Sci. 14,5161–5190 (2021).
37. Luo, D. et al. Enhanced photovoltage for inverted planar hetero-
junction perovskite solar cells. Science 360, 1442–1446
(2018).
38. Lin, R. et al. All-perovskite tandem solar cells with improved grain
surface passivation. Nature 603,73–78 (2022).
39. Ni, Z. et al. Resolving spatial and energetic distributions of trap
states in metal halide perovskite solar cells. Science 367,
1352–1358 (2020).
40. Jung, E. H. et al. Efficient, stable and scalable perovskite solar cells
using poly(3-hexylthiophene). Nature 567,511–515 (2019).
41. Xu, D. et al. Constructing molecular bridge for high-efficiency and
stable perovskite solar cells based on P3HT. Nat. Commun. 13,
7020 (2022).
42. Yun, Y. et al. A nontoxic bifunctional (anti)solvent as digestive-
ripening agent for high-performance perovskite solar cells. Adv.
Mater. 32, e1907123 (2020).
43. Lee, J. W. et al. Solid-phase hetero epitaxial growth of alpha-phase
formamidinium perovskite. Nat. Commun. 11,5514(2020).
44. Li, X. et al. In-situ cross-linking strategy for efficient and oper-
ationally stable methylammoniun lead iodide solar cells. Nat.
Commun. 9,3806(2018).
45. Niu, B. et al. Mitigating the lead leakage of high-performance per-
ovskite solar cells via in situ polymerized networks. ACS Energy Lett.
6, 3443–3449 (2021).
46. Cao, Y. et al. Perovskite light-emitting diodes based on sponta-
neously formed submicrometre-scale structures. Nature 562,
249–253 (2018).
47. Stranks, S. D. et al. Recombination kinetics in organic-inorganic
perovskites: excitons, free charge, and subgap states. Phys. Rev.
Appl. 2,034007(2014).
48. Si, H. et al. Emerging conductive atomic force microscopy for metal
halide perovskite materials and solar cells. Adv. Energy Mater. 10,
1903922 (2020).
49. Jiang, C. S. et al. Carrier separation and transport in perovskite solar
cells studied by nanometre-scale profiling of electrical potential.
Nat. Commun. 6,8397(2015).
50. Chen, Q. et al. Rapid microwave-annealing process of hybrid per-
ovskites to eliminate miscellaneous phase for high performance
photovoltaics. Adv. Sci. 7, 2000480 (2020).
51. Kim, J. et al. An effective method of predicting perovskite solar cell
lifetime–Case study on planar CH
3
NH
3
PbI
3
and HC(NH
2
)
2
PbI
3
per-
ovskite solar cells and hole transfer materials of spiro-OMeTAD and
PTAA. Sol. Energy Mater. Sol. Cells 162,41–46 (2017).
52. Yang, T. et al. One-stone-for-two-birds strategy to attain beyond
25% perovskite solar cells. Nat. Commun. 14,839(2023).
53. Hui, W. et al. Stabilizing black-phase formamidinium perovskite
formation at room temperature and high humidity. Science 371,
1359–1364 (2021).
54. Wang, T. et al. Room temperature nondestructive encapsulation via
self-crosslinked fluorosilicone polymer enables damp heat-stable
sustainable perovskite solar cells. Nat. Commun. 14,1342
(2023).
55. Fan, W., Deng, K., Shen, Y., Bai, Y. & Li, L. Moisture-accelerated
precursor crystallisation in ambient air for high-performance per-
ovskite solar cells toward mass production. Angew. Chem. Int. Ed.
61, e202211259 (2022).
56. Khenkin, M. V. et al. Consensus statement for stability assessment
and reporting for perovskite photovoltaics based on ISOS proce-
dures. Nat. Energy 5,35–49 (2020).
Article https://doi.org/10.1038/s41467-023-38926-3
Nature Communications | (2023) 14:3216 9
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Acknowledgements
We thank the Analytical & Testing Center at Northwestern Polytechnical
University for FIB-STEM measurement, SAXS/WAXS beamline at the
Australian Synchrotron for GIWAXS measurement, Hong Kong Poly-
technic University for the academic license of Silvaco software. This
work is supported financially by the National Natural Science Foundation
of China (62075094 T.Q., 52003118 F.W., 62205143 W.H.); Natural Sci-
ence Foundation of Jiangsu Province (BK20211537 T.Q.).
Author contributions
T.Q. and F.W. conceived the idea and designed the experiment; F.W.,
T.Q., and W.H. supervised the work; M.L., Q.T., R.S., and H.M. fabricated
the perovskite devices and carried out the PV performance character-
izations; F.W. provided STEM data analysis; M.L. and R.S. carried out the
KPFM and PF-TUNA measurement; Q.T., H.W., and J.C. carried out the
TCSPC, SCLC, and EIS measurement; Z.L. carried out the SEM measure-
ment with the assistance of Z.W.; J.C., and H.C. provided architecture
simulation; A.W. carried out the HR-TEM measurement with the assis-
tance of J.Z.; J.D. provided GIWAXS data analysis with the assistance of
Y.C.; Yo.L. and H.M. carried out built-in electric field testing; J.L. and Ya.L.
performed the IR measurement; X.C., P.G., and S.Y. performed the XRD
and XPS measurement; Y.S., T.L., and W.L. performed the Trap density
measurement and PLQE measurement with the assistance of Q.T.; F.W.
performed the data analysis and wrote the manuscript. T.Q. provided
some revisions. R.L., J.W., Y-B.C., and X.L. gave some useful suggestions.
All authors discussed the results and commented on the manuscript.
Competing interests
The authors declare no competing interests.
Additional information
Supplementary information The online version contains
supplementary material available at
https://doi.org/10.1038/s41467-023-38926-3.
Correspondence and requests for materials should be addressed to Wei
Huang or Tianshi Qin.
Peer review information Nature Communications thanks Jinwei Gao,
Chang-Zhi Li and the other, anonymous, reviewer(s) for their
contribution to the peer review of this work. A peer review file is
available.
Reprints and permissions information is available at
http://www.nature.com/reprints
Publisher’s note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as
long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license, and indicate if
changes were made. The images or other third party material in this
article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not
included in the article’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted
use, you will need to obtain permission directly from the copyright
holder. To view a copy of this license, visit http://creativecommons.org/
licenses/by/4.0/.
© The Author(s) 2023
Article https://doi.org/10.1038/s41467-023-38926-3
Nature Communications | (2023) 14:3216 10
Content courtesy of Springer Nature, terms of use apply. Rights reserved

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com