Professional Documents
Culture Documents
Cite This: ACS Appl. Mater. Interfaces 2018, 10, 21985−21990 www.acsami.org
■ INTRODUCTION
Perovskite solar cells (PSCs) are a potential disruptive
through encapsulation,8 light-induced degradation remains as a
key challenge that needs to be tackled in order to realize more
technology in the field of photovoltaics (PV) because of stable devices.
their excellent power conversion efficiencies (PCEs), and Reports discussing the light-induced instability of PSCs
promise of low costs due to the ease of solution processing.1 emphasize different causes of light induced degradation so far,
However, any PV manufacturer today that wishes to be however the majority conclude that ultraviolet (UV) radiation
competitive in large-scale solar power generation needs to is the most damaging region of the solar spectrum.9,10 In
guarantee that their PV module will produce more than 80% of particular, the most commonly used electron transport layer
its original rated power after a period of 25 years.2 This is the (ETL), titanium dioxide (TiO2), is reported to reduce device
key aspect of PSCs in which they fall far behind their inorganic stability under UV illumination in two ways. First, the
market-dominating counterpartssuch as silicon (Si), copper photocatalytic effect at the surface of the titanium dioxide
indium gallium diselenide (CIGS), or cadmium telluride surface is reported to decompose the methylammonium lead
(CdTe)because of the intrinsic material instability at iodide (CH3NH3PbI3) perovskite to PbI2 in an irreversible
elevated temperatures3 and their vulnerability to degrade process.11−13 Second, upon UV exposure, deep trapping at
during prolonged light exposure,4 humidity,5 and oxygen.6 The oxygen vacancies on the TiO2 surface is reported to strongly
underlying degradation processes occur not only within the enhance recombination and, thus, decreases performance of
perovskite absorber itself, but rather they are cumulative the PSC. The oxygen vacancies originate from holes in the
phenomena originating from other layers and interfaces as well. valence band which recombine with adsorbed oxygen causing
For instance, the most commonly used hole transport layer desorption and leaving positively charged trap states of long
(HTL) 2,2′,7,7′-Tetrakis[N,N-di(4-methoxyphenyl)amino]- lifetime.14,15 To overcome these problems various approaches
9,9′-spirobifluorene (C81H68N4O8, or commonly known as have been suggested:9,16−18 (i) replacing TiO2 with another
spiro-OMeTAD) is neither moisture stable nor can it ETL material; (ii) insertion of interfacial layers between the
withstand elevated temperatures.7 Since exposure to moisture
and oxygen can be effectively prevented via proper Received: February 20, 2018
encapsulation5 as other successful technologies (e.g., CIGS Accepted: June 11, 2018
solar cells) have overcome susceptibility to water vapors Published: June 11, 2018
TiO2 ETL and perovskite absorber; and (iii) the use of UV-
absorbing filters and introduction of luminescent downshifting
(LDS) layers to improve UV stability.
Despite the general understanding of the degradation
mechanism at the TiO2 surface upon UV illumination, the
literature remains inconclusive in defining the harmfulness of
UV degradation for PSCs.19−21 This is due to the multitude of
coexisting degradation mechanisms occurring across different
layers and interfaces of the PSC, including metal infiltration
from top contact to perovskite,4 deterioration of chemical
bonding between HTL and gold (Au) electrode,10 photo-
induced oxidation of spiro-OMeTAD22 and the dependence of
perovskite film thickness variation on photoinduced perform-
ance degradation.23 In addition, there is a lack of standardized
long-term illumination test conditions for PSC. Even the “class
A” spectral match of the standardized test conditions (STC)
for solar simulators does not provide an intensity specification
for UV light of wavelengths shorter than 400 nm.24 As a
consequence, the literature reports use of different UV spectra,
ranging from studies with light stress of one sun to studies
using only UV light stress.9,14,16 Moreover, various illumination
levels are used ranging from 1000 W m−2 of a 325 nm UV laser
light (>250 suns) to simulated one-sun and 100-suns spectra
without specifying in detail the UV content.18,25,26 The existing
standardized test with regard to UV stress is the International
Electrotechnical Commission (IEC) qualification test 61215,
which uses the air-mass 1.5 global (AM 1.5G) spectrum, but
this test does not ensure that a PV module will withstand >20
years of UV radiation.27 Thus, a comparative study of the
spectral dependence of the UV photostability of PSCs with
different device architectures is missing, as it is not clear which
UV wavelengths and at what intensities are detrimental to
device stability.
■
deposited via spin coating; (iii) nanoparticles (np-TiO2) grown via
wet-chemical synthesis; (iv) TiO2 deposited via electron-beam RESULTS AND DISCUSSION
evaporation (e-TiO2); and (v) post annealed at 450 °C meso-TiO2
on top of compact TiO2. We have provided further details on device The schematic illustration of a perovskite solar cell under UV
fabrication in the Supporting Information (SI). The two different UV illumination, highlighting the two different UV illumination
spectra used are 310−317 nm (emitted from a narrowband UV−B spectra tested in this study, along with the AM 1.5G spectrum
fluorescent Philips lamp, hereafter referred to as 311 nm UV) and are shown in Figure 1a. All the devices are exposed to UV light
360−380 nm (emitted from an actinic Philips UV lamp in the UV-A from the glass/ETL side, while the light exposure times are
region, hereafter referred to as 370 nm), shown in Figure 1a. The recorded in terms of sun-hours. The transmittance of the ETL
choice of these UV lamps is justified as the scarcity of the UV sources coated substrates varies in the two wavelength ranges of
with broad spectral emission limits such degradation studies to span interest (310−317 nm and 360−380 nm) as shown in Figure
over many UV spectral bands. Initial experiments are “‘accelerated”’
such that the UV light intensities are significantly greater than the
S2. It should be noted that average transmission for all ETLs in
respective UV content in AM 1.5G spectrum (112 times for 311 nm the former wavelength range is low (∼20% on average for all
and 57 times for 370 nm UV light). However, in a later control ETLs) because of absorption by the soda lime glass substrate,
experiment, we also verify the findings for lower intensities (11 suns while in the longer wavelength range the transmission is >70%
for 311 nm). In order to isolate only the effects from light-induced on average. The stress tests are performed on a number of
degradation, all experiments are carried out in nitrogen (N2) filled devices always including two kinds of reference configurations:
21986 DOI: 10.1021/acsami.8b03024
ACS Appl. Mater. Interfaces 2018, 10, 21985−21990
ACS Applied Materials & Interfaces Research Article
(i) the dark reference devices (6 in total), which are wrapped dark and under UV filters. It should be noted that the devices
with aluminum foil so that they experience the same heat stress with np-TiO2 ETL appear intrinsically less stable with regard
as other samples, but are not exposed to any light; and (ii) to PCE-5 min, although the PCE from voltage−current scan is
devices covered with optical long-pass filters (4 in total), which not degraded (see Figure S3). Furthermore, all TiO2 ETL
cutoff wavelength of 400 nm. The third set of devices (6 in based devices exhibit similar performance degradation under
total) are the ones exposed to UV light for degradation testing. 311 nm UV. Thus, it can be concluded that devices fabricated
Both types of references (dark and UV filters covered) are used using any of the ETLs investigated here exhibit similar
here to isolate the temperature effects (if any) as some studies degradation trends when exposed to 311 nm UV irradiance,
report the decreased device performance due to adverse effect while the reference devices retain their initial PCEs-5 min.
of temperature cycles (20−70 °C) on spiro-OMeTAD/ Having demonstrated a strong spectral dependence of the
perovskite interface.30 UV degradation for PSCs with TiO2 and SnO2 ETLs for high
The results of the experiments demonstrate that the devices UV exposure intensities at 311 nm, the impact of UV intensity
exposed to 311 nm UV experience severe decrease in short- is investigated. Since the intensities of both UV light sources
circuit current density (JSC). In contrast, devices exposed to used in this study are high, ranging from 50 to 100 times more
370 nm UV experience no degradation in general and a light than respective UV content in the AM 1.5G spectrum, it is
soaking effect in the case of the c-TiO2 devices is prominent. important to understand whether the observed degradation is
Light soaking generally describes the improved performance as also present at lower intensities. To demonstrate that this
a result of photoinduced structural transformation due to trap degradation is solely caused by the photon energy instead of
filling by photogenerated carriers upon prolonged light the high UV light intensity, polymer based UV long-pass filters
exposure.31 Figure 1(b) shows JSC measured for devices with were used allowing only the 10% of incoming 311 nm UV light
SnO2. The reference devices do not exhibit any decrease in JSC, (11 times of 311 nm UV content in AM 1.5G spectrum) to be
while devices exposed to 311 nm UV experience a significant transmitted toward the devices under test. As demonstrated in
decrease in JSC of up to 64% over the 1750 sun-hour testing Figure 2, the degradation trend remains the same under 90%
period. We attribute this reduction to the decomposition of the
perovskite absorber layer.32 Charge generation and extraction
remains unaltered for the devices exposed to 370 nm UV
wavelength for equal number of sun-hours. The observed loss
in JSC at 311 nm UV resulted in reduced PCEs, measured by
tracking the power output of the PSC for 5 min at constant
voltage close to the maximum power point (hereafter referred
to as PCE-5 min). The open circuit voltage (VOC) is less
affected as compared to JSC, while the fill factor (FF) decreased
due to the decrease in JSC as shown in Figures S3−S5.
Additional data on the devices, like RS, RSh, and cross-sectional
scanning electron microscope (SEM) images are presented in
Figures S6−S8. The initial and final values of PCE-5 min for all
the devices exposed to both UV spectra are given in Table S1. Figure 2. UV degradation of perovskite solar cells based on a SnO2 as
A similar trend of decreased JSC under 311 nm UV light is ETL under ∼11 suns UV irradiation. Reference devices covered with
observed for all other ETLs (c-TiO2, e-TiO2, np-TiO2) optical long-pass filters (left) are compared with the devices exposed
decreasing 70.3%, 72.2% and 81.8% respectively, as shown in to UV light (right).
Figure S3. This underlines the fact that degradation under
UV−B light is present for all types of ETLs. Interestingly less intense 311 nm UV radiation, indicating that the UV light
among all the studied ETLs, SnO2 appeared to be 43% more intensity itself is not playing a major role here. Assuming a
stable in terms of PCE-5 min as compared to np-TiO2 (the linear degradation rate with respect to intensity, a PSC will
least stable). The findings reported here are also applicable in degrade at a rate of 1.27% /100 sun-hours of exposure time to
devices based on meso-TiO2 as ETL which do not have state 311 nm UV light. Due to their impact, these UV−B photons
of the art PCE as shown in Figure S9. cannot be ignored, since sunlight contains ∼1.6 W m−2 in the
The trend in JSC degradation for the devices based on SnO2 UV−B band of 300−320 nm. This observation suggests that
and tested under 311 nm UV radiation translates into this degradation trend may be the cause of instability of PSCs
reduction of PCE as illustrated in Figure 1c. The PCE-5 min under simulated one-sun solar irradiance.
decrease is proportional to the JSC decrease for the same The remedies suggested to prevent this 311 nm UV-
devices. All devices with other ETLs (c-TiO2, e-TiO2, np-TiO2, stimulated degradation include the employment of a UV filter
meso-TiO2) also show a decrease in stabilized PCE-5 min with or a LDS layer. The former approach is already commonly
increasing exposure times when illuminated with 311 nm UV employed in PV applications. For example, the front glass sheet
light. In contrast, all the devices exposed to 370 nm UV show of PV modules in extra-terrestrial applications is doped with
no degradation in PCEs-5 min when compared to the trivalent cerium (Ce3+) ions to minimize the impact of high
reference devices kept in dark or stored under UV filters. energy radiation,33 while for terrestrial applications UV
For the PSCs prepared with SnO2, e-TiO2, and c-TiO2, the absorbers are used in commercial PV encapsulants like
initial PCE-5 min is maintained throughout the testing period. ethylene vinyl acetate (EVA).34 However, while UV filtering
A degradation in PCE-5 min is only observed for PSCs with will protect devices from photodegradation, but the energy
np-TiO2 ETL and stressed under 370 nm UV, as shown in content of the UV photons is not harnessed.
Figure 1c. However, this decrease is not caused by the UV The addition of a LDS layer will also protect the devices
stress, as it is also apparent from the reference devices kept in assuming total absorption of the harmful photons. However, it
21987 DOI: 10.1021/acsami.8b03024
ACS Appl. Mater. Interfaces 2018, 10, 21985−21990
ACS Applied Materials & Interfaces Research Article
■ AUTHOR INFORMATION
Corresponding Authors
*E-mail: amjad.farooq@kit.edu.
*E-mail: ulrich.paetzold@kit.edu.
Figure 3. Loss and gain in integrated JSC by filtering out specific UV ORCID
wavelengths, maximum recoverable and the realistically recoverable Amjad Farooq: 0000-0002-8604-8536
JSC via LDS materials.
Ihteaz M. Hossain: 0000-0001-6533-1757
Jonas A. Schwenzer: 0000-0001-8795-4875
These results show that if we cut off the photons of
Tobias Abzieher: 0000-0002-2733-0136
wavelengths shorter than 360 nm, then the total lost JSC, Bryce S. Richards: 0000-0001-5469-048X
maximum recoverable JSC and realistically recoverable JSC are Notes
0.53, 0.45, and 0.22 mA cm−2, respectively. Hence, it is The authors declare no competing financial interest.
concluded that the LDS technology can afford the protection
from UV light and at the same time afford a small gain in terms
of photocurrent. This, however, is less than 0.3 mA cm−2 as
■ ACKNOWLEDGMENTS
The authors would like to acknowledge financial support of the
denoted by the realistic scenario in Figure 3, which still German Federal Ministry for Education and Research (BMBF
includes favorable assumptions. Here, it should also be noted project PeroSol FKZ 03SF0483B); the Helmholtz Association
that up-to-date there are no luminescent materials that will for the Young Investigator Group of Dr. U. W. Paetzold;
exhibit 100% PLQY for the long period of time that a PV Recruitment Initiative of Prof. B. S. Richards; the Helmholtz
module is expected to withstand sunlight exposure in the field Energy Materials Foundry (HEMF); and the Science and
(>20 years). Hence, it is concluded that the approach of Technology of Nanostructures research programme); as well
filtering the UV light appears to be the simplest, leading to as the Karlsruhe School of Optics and Photonics (KSOP).The
protection of the devices at the cost of a very small fraction of authors express their gratitude to the great spirit of the “KIT
the available photocurrent. perovskite PV taskforce” and the KIT Young Investigator
21988 DOI: 10.1021/acsami.8b03024
ACS Appl. Mater. Interfaces 2018, 10, 21985−21990
ACS Applied Materials & Interfaces Research Article
Network. Mr. Amjad Farooq thankfully acknowledges the Correa-Baena, J.-P. Highly Efficient and Stable Planar Perovskite Solar
financial support from DAAD (Deutscher Akademischer Cells by Solution-Processed Tin Oxide. Energy Environ. Sci. 2016, 9
Austauschdienst/German academic exchange service) for his (10), 3128−3134.
doctoral research work under personal reference number (17) Qiu, L.; Ono, L. K.; Jiang, Y.; Leyden, M. R.; Raga, S. R.; Wang,
91604868. S.; Qi, Y. Engineering Interface Structure to Improve Efficiency and
■
Stability of Organometal Halide Perovskite Solar Cells. J. Phys. Chem.
B 2018, 122, 511.
REFERENCES (18) Fernandes, R. V.; Bristow, N.; Stoichkov, V.; Anizelli, H. S.;
(1) Correa-Baena, J.-P.; Saliba, M.; Buonassisi, T.; Grätzel, M.; Duarte, J. L.; Laureto, E.; Kettle, J. Development of Multidye UV
Abate, A.; Tress, W.; Hagfeldt, A. Promises and Challenges of Flters for OPVs Using Luminescent Materials. J. Phys. D: Appl. Phys.
Perovskite Solar Cells. Science 2017, 358 (6364), 739−744. 2017, 50 (2), 025103.
(2) Dunlop, E. D.; Halton, D.; Ossenbrink, H. A. 20 Years of Life (19) Huang, F.; Jiang, L.; Pascoe, A. R.; Yan, Y.; Bach, U.; Spiccia,
and More: Where is the End of Life of a PV Module? Conf. Rec. L.; Cheng, Y. B. Fatigue Behavior of Planar CH3NH3PbI3 Perovskite
Thirty-first IEEE Photovolt. Spec. Conf. 2005 2005, 1593. Solar Cells Revealed by Light On/Off Diurnal Cycling. Nano Energy
(3) Aitola, K.; Domanski, K.; Correa-Baena, J. P.; Sveinbjörnsson, 2016, 27, 509−514.
K.; Saliba, M.; Abate, A.; Grätzel, M.; Kauppinen, E.; Johansson, E. M. (20) Bracher, C.; Freestone, B. G.; Mohamad, D. K.; Smith, J. A.;
J.; Tress, W.; Hagfeldt, A.; Boschloo, G. High Temperature-Stable Lidzey, D. G. Degradation of Inverted Architecture CH3NH3PbI3‑xClx
Perovskite Solar Cell Based on Low-Cost Carbon Nanotube Hole Perovskite Solar Cells Due to Trapped Moisture. Energy Energy Sci.
Contact. Adv. Mater. 2017, 29 (17), 1606398. Eng. 2018, 6 (1), 35−46.
(4) Guarnera, S.; Abate, A.; Zhang, W.; Foster, J. M.; Richardson, (21) Khenkin, M. V.; Anoop, K. M.; Visoly-Fisher, I.; Galagan, Y.; Di
G.; Petrozza, A.; Snaith, H. J. Improving the Long-Term Stability of Giacomo, F.; Patil, B. R. R.; Sherafatipour, G.; Turkovic, V.; Rubahn,
Perovskite Solar Cells with a Porous Al2O3 Buffer Layer. J. Phys. H.-G.; Madsen, M.; Merckx, T.; Uytterhoeven, G.; Bastos, J. P.;
Chem. Lett. 2015, 6 (3), 432−437. Aernouts, T.; Brunetti, F.; Lira-Cantu, M.; Katz, E. Reconsidering
(5) Dong, Q.; Liu, F.; Wong, M. K.; Tam, H. W.; Djurišić, A. B.; Ng, Figures of Merit for the Performance and Stability of Perovskite
A.; Surya, C.; Chan, W. K.; Ng, A. M. C. Encapsulation of Perovskite Photovoltaics. Energy Environ. Sci. 2018, 11, 739−743.
Solar Cells for High Humidity Conditions. ChemSusChem 2016, 9 (22) Sanchez, R. S.; Mas-Marza, E. Light-Induced Effects on Spiro-
(18), 2518. OMeTAD Films and Hybrid Lead Halide Perovskite Solar Cells. Sol.
(6) Rong, Y.; Liu, L.; Mei, A.; Li, X.; Han, H. Beyond Efficiency: the Energy Mater. Sol. Cells 2016, 158, 189−194.
Challenge of Stability in Mesoscopic Perovskite Solar Cells. Adv. (23) Jena, A. K.; Kulkarni, A.; Ikegami, M.; Miyasaka, T. Steady
Energy Mater. 2015, 5 (20), 1501066. State Performance, Photo-Induced Performance Degradation and
(7) Lei, H.; Yang, G.; Zheng, X.; Zhang, Z.-G.; Chen, C.; Ma, J.; Their Relation to Transient Hysteresis in Perovskite Solar Cells. J.
Guo, Y.; Chen, Z.; Qin, P.; Li, Y.; Fang, G. Incorporation of High- Power Sources 2016, 309 (March), 1−10.
Mobility and Room-Temperature-Deposited CuxS as a Hole Trans- (24) Guidelines for PV Power Measurement in Industry; Taylor, N.,
port Layer for Efficient and Stable Organo-Lead Halide Perovskite Ed.; European Commission, 2010.
Solar Cells. Sol. RRL 2017, 1 (6), 1700038. (25) Chen, T. P.; Lin, C. W.; Li, S. S.; Tsai, Y. H.; Wen, C. Y.; Lin,
(8) Carcia, P. F.; McLean, R. S.; Hegedus, S. Encapsulation of W. J.; Hsiao, F. M.; Chiu, Y. P.; Tsukagoshi, K.; Osada, M.; Sasaki, T.;
Cu(InGa)Se2 Solar Cell with Al2O3 Thin-Film Moisture Barrier Chen, C. W. Self-Assembly Atomic Stacking Transport Layer of 2D
Grown by Atomic Layer Deposition. Sol. Energy Mater. Sol. Cells 2010, Layered Titania for Perovskite Solar Cells with Extended UV Stability.
94 (12), 2375−2378. Adv. Energy Mater. 2018, 8, 1701722.
(9) Bella, F.; Griffini, G.; Correa-Baena, J.-P.; Saracco, G.; Gratzel, (26) Misra, R. K.; Aharon, S.; Li, B.; Mogilyansky, D.; Visoly-Fisher,
M.; Hagfeldt, A.; Turri, S.; Gerbaldi, C. Improving Efficiency and I.; Etgar, L.; Katz, E. A. Temperature- and Component-Dependent
Stability of Perovskite Solar Cells with Photocurable Fluoropolymers. Degradation of Perovskite Photovoltaic Materials under Concentrated
Science (Washington, DC, U. S.) 2016, 354 (6309), 203−206.
Sunlight. J. Phys. Chem. Lett. 2015, 6 (3), 326−330.
(10) Wei, D.; Wang, T.; Ji, J.; Li, M.; Cui, P.; Li, Y.; Li, G.;
(27) Kempe, M. D. Accelerated UV Test Methods and Selection
Mbengue, J. M.; Song, D. Photo-Induced Degradation of Lead Halide
Criteria for Encapsulants of Photovoltaic Modules. 2008 33rd IEEE
Perovskite Solar Cells Caused by the Hole Transport Layer/Metal
Photovolatic Specialists Conference (PVSC) 2008, 1.
Electrode Interface. J. Mater. Chem. A 2016, 4 (5), 1991−1998.
(28) Mahmood, K.; Sarwar, S.; Mehran, M. T. Current Status of
(11) Li, W.; Zhang, W.; Van Reenen, S.; Sutton, R. J.; Fan, J.;
Electron Transport Layers in Perovskite Solar Cells: Materials and
Haghighirad, A. A.; Johnston, M. B.; Wang, L.; Snaith, H. J. Enhanced
UV-light Stability of Planar Heterojunction Perovskite Solar Cells Properties. RSC Adv. 2017, 7 (28), 17044−17062.
With Cesium Bromide Interface Modification. Energy Environ. Sci. (29) Haque, M. A.; Sheikh, A. D.; Guan, X.; Wu, T. Metal Oxides as
2016, 9 (2), 490−498. Effcient Charge Transporters in Perovskite Solar Cells. Adv. Energy
(12) Lee, S. W.; Kim, S.; Bae, S.; Cho, K.; Chung, T.; Mundt, L. E.; Mater. 2017, 7 (20), 1602803.
Lee, S.; Park, S.; Park, H.; Schubert, M. C.; Glunz, S. W.; Ko, Y.; Jun, (30) Sheikh, A. D.; Munir, R.; Haque, M. A.; Bera, A.; Hu, W.;
Y.; Kang, Y.; Lee, H. S.; Kim, D. UV Degradation and Recovery of Shaikh, P.; Amassian, A.; Wu, T. Effects of High Temperature and
Perovskite Solar Cells. Sci. Rep. 2016, 6, 38150. Thermal Cycling on the Performance of Perovskite Solar Cells:
(13) Zhang, P.; Wu, J.; Wang, Y.; Sarvari, H.; Liu, D.; Chen, Z. D.; Acceleration of Charge Recombination and Deterioration of Charge
Li, S. Enhanced Efficiency and Environmental Stability of Planar Extraction. ACS Appl. Mater. Interfaces 2017, 9 (40), 35018−35029.
Perovskite Solar Cells by Suppressing Photocatalytic Decomposition. (31) Zhang, T.; Cheung, S. H.; Meng, X.; Zhu, L.; Bai, Y.; Ho, C. H.
J. Mater. Chem. A 2017, 5, 17368−17378. Y.; Xiao, S.; Xue, Q.; So, S. K.; Yang, S. Pinning Down the Anomalous
(14) Leijtens, T.; Eperon, G. E.; Pathak, S.; Abate, A.; Lee, M. M.; Light Soaking Effect toward HighPerformance and Fast-Response
Snaith, H. J. Overcoming Ultraviolet Light Instability of Sensitized Perovskite Solar Cells: The IonMigration-Induced Charge Accumu-
TiO2 with Meso-Superstructured Organometal Tri-Halide Perovskite lation. J. Phys. Chem. Lett. 2017, 8 (20), 5069−5076.
Solar Cells. Nat. Commun. 2013, 4, 2885. (32) Tress, W.; Yavari, M.; Domanski, K.; Yadav, P. K.; Niesen, B.;
(15) Ahn, N.; Kwak, K.; Jang, M. S.; Yoon, H.; Lee, B. Y.; Lee, J. K.; Correa-Baena, J.-P.; Hagfeldt, A.; Grätzel, M. Interpretation and
Pikhitsa, P. V.; Byun, J.; Choi, M. Trapped Charge-Driven Evolution of Open-Circuit Voltage, Recombination, Ideality Factor
Degradation of Perovskite Solar Cells. Nat. Commun. 2016, 7, 13422. and Subgap Defect States During Reversible Light-Soaking and
(16) Anaraki, E. H.; Kermanpur, A.; Steier, L.; Domanski, K.; Irreversible Degradation of Perovskite Solar Cells. Energy Environ. Sci.
Matsui, T.; Tress, W.; Saliba, M.; Abate, A.; Grätzel, M.; Hagfeldt, A.; 2018, 11, 151−165.