Light-induced Surface Recombination Suppression, Carrier Lifetime Improvement, and Deep Defect Formation in Lead Halide Perovskites

RESEARCH ARTICLE www.lpr-journal.org Light-Induced Surface Recombination Suppression, Carrier Lifetime Improvement, and Deep Defect Formation in Lead Halide Perovskites Katarína Pekárková,* Roman Grill, Martin Ledinský, Lucie Landová, Zdenˇ ek Remeš, Eduard Belas, Petr Praus, Aleš Vlk, Jakub Pekárek, and Neda Neykova* Lead halide perovskites are promising materials in a wide range of optoelectronic devices. The internal instabilities of perovskites under illumination, however, hinder their application. Here, light-induced changes in CH3NH3PbBr3single crystals are studied, focusing on charge carrier transport properties and deep defect evolution. Using laser-induced transient current technique measurements under steady-state illumination enables the distinction of electron and hole conductivity. An improvement in charge carrier collection is observed stemming from reduced surface recombination and increased hole lifetime. Surprisingly, despite the overall optimization of carrier transport, the simultaneous formation of deep defects under illumination is recorded via photo-thermal deflection spectroscopy. Empirical models are proposed to rationalize all light-induced changes. The explanations are based on strong electron trapping, ion migration, and passivation of strong recombination centers by light-induced deep defects. 1. Introduction Lead halide perovskites AMX3, where A typically corresponds to MA+=CH3NH3+,MtoPb 2+, and X to halide anion, are nowadays prospective in various technologies, including solar cells,[1] K. Pekárková, M. Ledinský, L. Landová, Z. Remeš, A. Vlk, N. Neykova Institute of Physics Czech Academy of Sciences Cukrovarnická 10, Prague 16200, Czech Republic E-mail: [email protected];[email protected] R.Grill,E.Belas,P.Praus,J.Pekárek Faculty of MathematicsandPhysics CharlesUniversity KeKarlovu5,Prague12116,CzechRepublic L.Landová,N.Neykova CentreforAdvancedPhotovoltaics Faculty ofElectricalEngineering CzechTechnicalUniversityinPrague Technická2,Prague16627,CzechRepublic The ORCID identification number(s) for the author(s) of this article can be found under https://doi.org/10.1002/lpor.202401904 © 2025 The Author(s). Laser & Photonics Reviews published by Wiley-VCH GmbH. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. DOI: 10.1002/lpor.202401904 radiation detectors,[2]and light-emitting diodes.[3]The perovskite popularity arises from their unique properties, such as low-cost manufacturing,[1]high carrier diffusion length,[4]long carrier lifetime,[4,5]high absorption coefficient, sharp absorption edge onset,[6,7]and defect tolerance.[8]Despite constant progress, the perovskite stability under illumination remains the main obstacle to their application. Evolution of device performance under continuous illumination, known as the light-soaking effect, has been besides perovskites previously observed in various semiconductors, mostly in hydrogenated amorphous silicon (a-Si:H),[9,10]copper indium-gallium selenide (CIGS)[11]and CdTe.[12] The influence of continuous illumination on perovskite properties varies across the literature, depending mainly on exact material composition, device architecture, or growth conditions. Both improvement and deterioration of device performance were observed, as well as reversibility and irreversibility of lightinduced changes.[13–16]This points to the complex nature of light-induced changes and the lack of understanding of driving phenomena causing instability under illumination. Several explanations have been proposed, while the scientific community leans most toward mechanisms based on ion migration,[17,18] lattice expansion,[19,20]lowered trap-state density[21,22]and charge carrier depletion at contacts.[23,24]However, the origin of light-soaking phenomena in perovskites is still an open issue.[13] Moreover, most of the research was conducted on whole solar cells based on polycrystalline thin film perovskites. Polycrystalline perovskites have high defect densities and surface states due to the presence of grain boundaries.[25–28]As grain boundaries were observed to enhance ion migration[29]in perovskites, the impact of continuous illumination on the single crystals might be different. Moreover, special electron and hole transport layers in whole solar cells may also affect the material evolution under illumination. However, only a few reports were dedicated to light-soaking effects in perovskite single crystals, which are promising materials in high-energy radiation detection[2,30,31] and photo-detection.[32,33]A strong light-soaking effect observed in MAPbBr3-xIxand CH3NH3PbI3single crystals[34–36]proves that grain boundaries are not necessary for device instability under illumination. Laser Photonics Rev. 2025,19, 2401904 2401904 (1 of 10) © 2025 The Author(s). Laser & Photonics Reviews published by Wiley-VCH GmbH www.advancedsciencenews.com www.lpr-journal.org Figure 1. L-TCT experimental setup with illumination from the anode side by excitation laser pulses and transient current waveforms induced by drifting carriers through the sample. A blue line represents an ideal current waveform, and a red line corresponds to a typical current waveform in real material. Continuous illumination may be optionally used to study light-induced changes. Here, we study the light-induced changes in CH3NH3PbBr3 (MAPB) single-crystal perovskites, focusing on charge carrier transport properties. We use the non-destructive laser-induced transient current technique (L-TCT) under steady-state illumination. This approach allows us to separate electron and hole conductivity. The observed light-induced changes in surface recombination velocity and charge carrier lifetime are discussed, and possible mechanisms are proposed. Further, combining L-TCT with photo-thermal deflection spectroscopy (PDS) enables us to link the light-induced changes in transport properties with deep defect evolution. Surprisingly, an overall improvement in charge carrier collection is observed despite the simultaneous formation of deep defects. 2. Results and Discussion All measurements are performed on CH3NH3PbBr3single crystals grown by inverse temperature crystallization process in a nitrogen glovebox. Experimental details regarding the synthesis can be found in the Supporting Information. 2.1. MAPB Transport Properties To obtain general insight into the studied MAPB transport properties, we perform L-TCT measurement. This method, related to the time-of-flight measurements, records the movement of carriers generated by the laser excitation pulse. In the case of the above-bandgap excitation illumination, the sole electron or hole signal may be collected depending on the polarity of irradiated contact. The basic principle of the method is schematically shown in Figure 1. When the above-bandgap excitation pulse illuminates the anode side of a planar sample, electrons are almost immediately swept at the anode under applied bias, while photogenerated holes drift toward the opposite electrode through the sample thickness L. As carriers drift under applied bias through the material, they may be affected by surface recombination, trapping and de-trapping at shallow and deep levels, scattering at local impurities, thermal diffusion, and space charge-related effects. The average time taken by holes to travel across sample thickness under the influence of a known electric field will be further referred to as transit time tr. It manifests in the measured current waveform (CWF) by a distinct shoulder, as can be seen in Figure 1. The transient current before the transit time may be expressed by an exponential function[37] i(t)∼e−ct (1) The parameter c is given as c=(eN 𝜀0𝜀r +1 𝜇h𝜏h)𝜇h(2) here, Nis the space-charge density, 𝜖0and 𝜖rare the vacuum and relative permittivity, 𝜇hand 𝜏hare hole mobility and lifetime. The tail after transit time in typical CWF (see Figure 1)corresponds to the delayed holes captured during the drift by traps or scattered by local impurities. First, we perform standard L-TCT measurements in the dark, using only laser excitation pulses. All presented measurements are done in the configuration of anode illumination (Figure 1) when we record only signals of photo-generated drifting holes. Such a configuration was chosen because of much weaker electron L-TCT signals than the hole signals in perovskite materials, as seen in Supporting Information in Figure S3 (Supporting Information). Low electron signals are due to a significantly lower lifetime of electrons than hole lifetime.[38] We measure L-TCT in a pulsed bias regime instead of the standard continuous bias regime to minimize bias-induced ion migration. The timing diagram in Figure 2shows the pulsing bias parameters, such as bias amplitude, bias width, period, and delay time, representing the time between the rising edges of the applied bias pulse and the probing laser pulse. Unless otherwise stated, the used bias pulse has a frequency of 5 Hz, width of 1 ms, and delay time between the bias onset and laser excitation pulse of 100 μs. The anode side is excited by probing laser pulses with 3 μs width in FWHM, 5 Hz repetition frequency, and above-band-gap 514 nm wavelength. Other experimental details regarding the L-TCT measurement are in the Supporting Information. Here, we also demonstrate the sufficiency of one polarity bias pulsing in suppressing ion migration in the dark. In Figure S1 (Supporting Information), we perform L-TCT measurement Laser Photonics Rev. 2025,19, 2401904 2401904 (2 of 10) © 2025 The Author(s). Laser & Photonics Reviews published by Wiley-VCH GmbH 18638899, 2025, 18, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lpor.202401904 by Neda Neykova - Czech Technical University in Prague , Wiley Online Library on [10/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.lpr-journal.org Figure 2. The scheme of applied bias pulse (blue line) and the excitation laser pulse (yellow line) to the perovskite sample used during L-TCT measurements. at a constant 5 Hz frequency for varying widths of bias pulse of 60 V. Hysteresis effects arising from ion migration appear only for widths exceeding 3 ms. Therefore, selecting a bias width of 1 ms and frequency of 5 Hz in all following L-TCT experiments is well justified. Note that after applying a bias pulse to the perovskite sample, a transient response occurs in the circuit due to the RLC elements (the sample capacity of 2 pF and resistance of 58.5 MΩ). More details are in Supporting Information (Figure S2, Supporting Information). To estimate effective mobility and mobility-lifetime product, we measure transient currents on MAPB single crystals for different biases on the sample (see Figure 3a). As expected, the transit time is shortened with increasing applied bias. This is a consequence of the fact that charge carrier drift velocity is directly proportional to the applied electric field. Note that within the first ≈3μs, the rising edge appears in recorded current waveforms (CWFs) in Figure 3a, whose end is denoted by the dotted line. This is caused by the limitation of the used pulse generator, which enables us to set ≈3μs as the shortest possible width of the laser excitation pulse. The following initial slight increase of CWFs seen mostly at low biases corresponds to the plasma effect.[39] However, to maintain a good signal-to-noise ratio of CWFs, further excitation pulse intensity lowering to remove the plasma effect was not appropriate. From the normalized CWFs, where we multiply the horizontal time axis and divide the vertical current axis by the correFigure 3. L-TCT measurement on MAPB single crystals: a) bias dependence of transient currents; b) bias-normalized transient current waveforms; c) mobility calculated according to Equation (3); d) charge collection efficiency. Laser Photonics Rev. 2025,19, 2401904 2401904 (3 of 10) © 2025 The Author(s). Laser & Photonics Reviews published by Wiley-VCH GmbH 18638899, 2025, 18, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lpor.202401904 by Neda Neykova - Czech Technical University in Prague , Wiley Online Library on [10/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.lpr-journal.org sponding bias for each curve, one can deduce whether the internal electric field is distorted by space charge formation. Such a deformed electric field should prolong the transit time differently at different biases. Thus, the individual transit time and bias products, tr(V)⋅V, should acquire different non-constant values in a semiconductor with considerable space charge. However, in Figure 3b, the transit time and bias product, tr(V)⋅V, shows the same constant value, marked with a dashed line. This indicates an almost uniform internal electric field distribution with no significant deformation by formed space charge. Interestingly, the curves are not exactly at the same height even after dividing the vertical current axis by corresponding bias. This might be explained by surface recombination, which competes with charge carrier harvesting and is more significant at lower biases. Considering a homogenous internal electric field profile, the hole mobility 𝜇hfor individual biases Vis given as 𝜇h=L2 Vtr (3) where Lstands for the thickness of the measured sample and tr for transit time, which may be found from the inflection point of the CWF. By applying calculations according to Equation (3)to individual current waveforms in Figure 3a, with the thickness of our crystal of 1.6 mm, we obtain an average hole mobility ⟨𝜇h⟩ of 18 cm2V−1s−1(Figure 3c). Further, the hole mobility-lifetime product 𝜇h𝜏hcan be determined by fitting the bias dependency of charge collection efficiency (CCE) by Many´s equation[40,41]: CCE =1 1+shL 𝜇hV 𝜇h𝜏hV L2[1−exp (−L2 𝜇h𝜏hV)] (4) here, shcorresponds to surface recombination velocity. The bias dependency of the collected charge in Figure 3d is acquired from the integration of individual current waveforms at given biases. Using Equation (4), we obtained hole mobility-lifetime product 𝜇h𝜏h=6×10−4cm2V−1and quantity related to surface recombination velocity sh∕𝜇h=100 V∙cm−1. The corresponding hole lifetime in MAPB crystal is then 𝜏h=33 μs, and surface recombination velocity sh=1800 cm∙s−1. Even though we did not provide any specific surface treatment, the obtained surface recombination velocity is still much lower compared to typical values in IIIV and II-VI semiconductors with passivated surfaces (e.g., GaAs, InP, CdZnTe), which are in the order of 105cm.s−1.[41]Asmall deviation of the fit by Many´s Equation (4) at the lowest voltages may be given by a small space charge persisting at low voltages. Nevertheless, the evaluated ⟨𝜇h⟩=18 cm2V−1s−1,and𝜇h𝜏h= 6×10−4cm2V−1are in good agreement with 𝜇h≈15 cm2V−1 s−1and 𝜇h𝜏h≈10 ×10−4cm2V−1obtained in refs.[42,43] The typical electron current waveform (CWF) obtained by LTCT on MAPB single crystal by cathode illumination is shown in the Supporting Information in Figure S3 (Supporting Information). We can see from the comparison of electron and hole CWFs in Figure S3 (Supporting Information) and Figure 3that electrons are easily trapped in the near-surface region since most of the electron signal decreases rapidly within the initial ≈2.5μs. The perovskite surface thus probably contains deep electron traps that effectively trap electrons in the order of <1μs and have long de-trapping time in the order of tens of μs. In contrast, a distinct transit time in the hole CWFs (Figure 3) shows that holes drift from the surface to the bulk more effectively than electrons. The tendency to form deep electron traps is also consistent with the fact that lead halide perovskites usually have p-type conductivity.[44]In p-types, the Fermi level is shifted toward the valence band maximum. Since defects above the Fermi level act as electron traps, there is a higher probability of deep electron traps with long de-trapping times. On the contrary, the presence of hole-deep traps is less probable due to the Fermi-level position. Deep electron traps acting as less efficient recombination centers due to low hole capture probability may also explain the ability of perovskites to maintain high performance despite defects, referred to as defect tolerance.[45] Similar long-lived trapped electrons and slow recombination with free holes were also observed in CH3NH3PbI3 polycrystalline thin films[46]and 2D lead halide perovskite nanoplatelets.[47]Strong electron trapping was also theoretically predicted in tin halide perovskites due to the energetically favorable formation of stable electron bipolarons.[48] 2.2. Influence of Continuous Illumination on MAPB Transport Properties To investigate the influence of above-bandgap illumination on perovskite transport properties, we perform L-TCT not only in the dark (using weak excitation laser pulses) but also under additional continuous above-bandgap illumination. Comparison of the L-TCT signal without additional LED illumination and with continuous LED illumination enables us to record light-induced changes. Used LED has a wavelength of 465 nm and intensity of 450 μW.cm−2. Here, excitation pulses and continuous LED are incident on the anode side of the measured sample. We thus collect, as in the previous section, purely hole signals. It´s important to mention that the L-TCT method is sensitive only to deviations from the steady state. Continuous illumination alone leads to the steady-state distribution of free carriers. Therefore, the greater number of charge carriers generated from additional continuous illumination does not contribute to the current pulse shape recorded by L-TCT. Comparison of CWFs for selected biases 30 and 60 V are shown in Figure 4. As can be seen from a comparison of the area under individual CWFs with and without illumination, the charge carrier collection efficiency has improved under illumination. An improved charge carrier collection can be explained by light-induced suppression of surface recombination or by charge carrier lifetime or mobility enhancement. Clear evidence that surface recombination is optimized by continuous illumination may be seen from bias-normalized current waveforms when we divide the horizontal axis and multiply the vertical axis by corresponding biases. As seen in Figure 5a, individual bias-normalized CWFs recorded in the dark have significantly different amplitudes. This is a sign of substantial surface recombination. On the contrary, for CWFs obtained under continuous illumination (Figure 5b), normalized curves overlap more, which indicates that surface recombination is almost completely suppressed. Surface recombination optimization may be explained by the following mechanism shown schematically in Figure 6.ConLaser Photonics Rev. 2025,19, 2401904 2401904 (4 of 10) © 2025 The Author(s). Laser & Photonics Reviews published by Wiley-VCH GmbH 18638899, 2025, 18, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lpor.202401904 by Neda Neykova - Czech Technical University in Prague , Wiley Online Library on [10/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.lpr-journal.org Figure 4. Hole CWFs measured on MAPB single crystal by L-TCT using laser pulses only (black line) and with additional continuous LED illumination 465 nm (green line) for voltage a) 30 V and b) 60 V. Figure 5. Bias-normalized hole CWFs measured on MAPB single crystal by L-TCT a) using laser pulses only and b) with an additional continuous LED illumination at different biases. tinuous LED illumination creates electron-hole pairs. Photogenerated holes drift toward the cathode. Electrons have approximately six times shorter lifetimes than holes in MAPB perovskites, as we showed in the Supporting Information in Figure S3 (Supporting Information). Short electron lifetime manifests in the measured electron CWF by a steep current decrease with no distinct transit time. Therefore, photo-generated electrons are most probably trapped near the anode as they are subjected to much stronger trapping than holes. A negative space charge from trapped photo-electrons is formed close to the anode at a distance, denoted as x1in Figure 6a. Previously, a homogenous internal electric field (depicted by the blue line in Figure 6b) is, after illumination, locally influenced by the negative space charge. We get a higher electric field close to the illuminated anode, as depicted by the red line in Figure 6b. Locally increased electric field consequently eliminates surface recombination. Negative space charge below the illuminated anode may also originate from the accumulation of negatively charged bromine interstitials Br− i, which most probably migrate in perovskites due to their low activation energies for the migration[1]and whose forFigure 6. Scheme of surface recombination optimization phenomena induced by light. a) negative space charge formation due to trapped photogenerated electrons or accumulated light-induced Br− iand b) change of internal electric field under illumination. Laser Photonics Rev. 2025,19, 2401904 2401904 (5 of 10) © 2025 The Author(s). Laser & Photonics Reviews published by Wiley-VCH GmbH 18638899, 2025, 18, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lpor.202401904 by Neda Neykova - Czech Technical University in Prague , Wiley Online Library on [10/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.lpr-journal.org mation may be supported by additional LED illumination. As we will show later in Figure 9, in the case of the used experimental conditions of L-TCT measurements in Figures 4and 5, the mechanism from strongly trapped electrons should prevail. The substantial ion-related negative space charge forms for continuous illumination times longer than ≈2 min (in case of intensity ≈450 μW.cm−2) and for widths of bias pulse wider than ≈400 μs at 70 V. Since strong electron trapping is observed in numerous halide perovskite compositions,[46–48]the presented empirical model of the light soaking effect in Figure 6, may be thus well applicable to various perovskites, especially to those with p-type conductivity. However, as observed in works,[4,49]perovskites can also exhibit longer electron than hole lifetimes. For such perovskites without significant electron trapping, we expect a similar surface recombination suppression only in case of longer illumination times and bias pulses, sufficient for ion-related space charge formation. Reduced recombination under illumination has also been indirectly observed in perovskite thin film solar cells and has been mainly explained by ion migration[50–56]and subsequent optimization of the interface between perovskite and electron or hole transport layer. Specifically, works[50]and[51]suggest the movement of MA+and FA+cations to the hole transport layer, work[52] proposes that migrating negative ions neutralize intrinsic positive ions near the electron transport layer, and, finally, works[53–56] suggest that halide anions accumulated at electron/ hole transport layers diminish Schottky barrier, improving charge extraction. We showed that the mechanism of surface recombination exists even in single crystals regardless of the presence of special electron or hole transport layers, such as Spiro-OMeTAD or PCBM. Moreover, the possibility of one carrier type L-TCT measurement enabled us to distinguish the specifics of electron and hole transport. The severe trapping of electrons in perovskite material, observed by L-TCT, probably plays a key role in lightinduced surface recombination optimization alongside ion migration. The exact influence of light on charge carrier transport properties is not so obvious from Figure 4, as the exponential decay factor cof CWFs strongly depends not only on carrier lifetime 𝜏but, according to Equations (1)and(2), also on formed space charge. A negative space charge from trapped photo-generated electrons, formed close to the anode, contributes to the steeper exponential decline of CWFs after illumination. The effect of a negative space charge may be suppressed after switching off the continuous LED illumination when new photo-generated electrons are not created anymore and previously trapped electrons are released from traps. Therefore, we first irradiated the sample for 10 min under a bias of 45 V by the LED of 465 nm wavelength and 450 μW.cm−2intensity, then switched off illumination and recorded changes in CWFs. In Figure 7, we compare the evolution of CWFs after switching off continuous LED illumination to the CWF measured on the newly grown unexposed sample. Interestingly, we observe a milder exponential decay in the previously light-soaked sample than in the unexposed sample. Such milder exponential decay accompanied by the improved charge carrier collection efficiency may be explained by hole lifetime improvement under illumination. As seen in Figure 7, the light-induced improvement in hole lifetime is not permanent. It persists for several minutes, even hours after switching off the illumination. Figure 7. Evolution of hole CWFs after switching off continuous LED illumination with intensity of 450 μW.cm−2. CWFs were measured at a pulsed applied bias of 45 V on a MAPB single crystal. On the contrary, the effect of a negative space charge disappears much quicker, enabling us to observe the real influence of light on transport properties without parasitic space charge influence. An improved hole lifetime under illumination may originate from two mechanisms. The first one is based on hole trapping. As shown in schematic Figure 8, the perovskite sample contains, before illumination, a hole trap level ETbelow the Fermi energy level. This defect level traps holes and shortens their lifetime. After illumination, the Fermi energy level may shift downward, closer to the valence band, so that trap level ETgets above the new Fermi energy level. Consequently, holes are not trapped by this trap level ETanymore, and hole lifetime increases. The second possible mechanism that may lead to the hole lifetime improvement consists of a chemical reaction. The light may produce new charged defects, which can passivate hole trap defects and thus increase hole lifetime. Such light-induced charged defects most probably originate in Frenkel pairs of bromine interstitials Br− iand bromine vacancies V+ Br , according to the DFT calculations[57,58]and Fourier-transform photocurrent spectroscopy measurements.[59]A similar mechanism based on trap inactivation has been previously suggested to explain photoluminescence enhancement in perovskite thin films under illumination.[60–63]Here, they attributed the reduced trap density to passivation from photo-generated electrons,[22]defect-curing chemical reaction with oxygen,[62]and on the contrary, to the annihilation of Frenkel pairs.[64,65] Our explanation of surface recombination suppression and carrier lifetime increase by the light-induced formation of charged defects is also supported by the observed behavior of current responses to bias pulse. A typical dark current response of MAPB crystal to bias pulse, connected with the RLC phenomenon, is shown in Supporting Information in Figure S2 (Supporting Information) and Figure 9a. As seen in Figure 9a, the dark current response does not change during the measurement time after applying a series of 1 ms bias pulses with 5 Hz frequency. However, we observe several hysteresis in individual current responses to bias pulses under continuous illuLaser Photonics Rev. 2025,19, 2401904 2401904 (6 of 10) © 2025 The Author(s). Laser & Photonics Reviews published by Wiley-VCH GmbH 18638899, 2025, 18, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lpor.202401904 by Neda Neykova - Czech Technical University in Prague , Wiley Online Library on [10/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.lpr-journal.org Figure 8. Mechanism of carrier lifetime optimization phenomena induced by light. Energy level diagram a) before and b) after LED illumination. EC denotes the conduction band, EFFermi energy level, EThole trap level, and EVvalence band. mination, which were measured at different times after applying a series of bias pulses of 1 ms width and 5 Hz frequency (see Figure 9b). The most significant evolution of the current response under illumination is visible at longer times (≈400 μs) of one applied bias pulse. The rise in current in Figure 9b may be explained by the light-induced formation of pairs Br− i/V+ Br ,subsequent accumulation of Br− iunder the illuminated anode, and migration of V+ Br away from the illuminated anode. A negative space charge forms close to the anode after ≈400 μs and thus locally increases the internal electric field (Figure 6b). A higher electric field near the anode reduces the surface recombination and hence improves carrier collection, which manifests in increased currents. During a further 199 ms when the applied bias is zero (Figure 2), the negative space charge disappears thanks to the diffusion, returning the surface recombination rate to its previous value. The substantial negative space charge decreasing surface recombination again forms after ≈400 μs of the next bias pulse. As Br− i∕V+ Br form under illumination, the phenomenon will be most visible at longer measurement times under continuous illumination when the density of light-induced Br− i/V+ Br is highest. The observation that a substantial negative space charge from migrating ions forms after times of continuous LED illumination longer than 2 min and after ≈400 μsofonebiaspulse of 70 V (Figure 9b) has implications for the analysis of L-TCT measurements in Figures 4and 5. Here, we continuously illuminated anode for times shorter than 1 min and use 100 μs delay time between the bias onset and laser excitation pulse (Figure 2). Therefore, a negative space charge from drifting ions is negligible in L-TCT measurements in Figures 4and 5, and space charge must originate from trapped electrons. We thus prove that severe electron trapping in perovskite material plays a key role in light-induced surface recombination optimization. 2.3. Influence of Prolonged Illumination on MAPB Transport Properties Further, we study the effect of prolonged illumination, so-called light-soaking, on MAPB crystal transport properties. For this purpose, we illuminate the anode side of the sample by laser diode with an above-bandgap wavelength of 465 nm (energy of 2.7 eV) and intensity corresponding to the one-sun illumination (AM1.5G). The sequence used for light-soaking measurements is in Figure 10. Further details of the measurement are in Supporting Information. Here, we also show in section S5 (Supporting Information) that effects associated with heat generation due to light absorption should be negligible compared to the lightinduced changes. The development of hole transient currents with increasing illumination time is shown in Figure 11a. Since the area under the CWF corresponds to the collected charge, we can conclude that the charge carrier collection increases after light exposure. The increase in charge carrier collection saturates after 120 min of illumination. Another interesting feature is the rapid current drop, which is most visible on the sample without previous illumination at the start of CWF (see dashed circle). The rapid current drop gradually disappears as we prolong the steady-state illumination time. Figure 9. Current responses to 70 V bias pulse with 1 ms width and 5 Hz frequency. a) Evolution of dark currents upon measurement time after applying bias pulses, b) evolution of currents under continuous above-bandgap illumination upon measurement time after applying bias pulses. Laser Photonics Rev. 2025,19, 2401904 2401904 (7 of 10) © 2025 The Author(s). Laser & Photonics Reviews published by Wiley-VCH GmbH 18638899, 2025, 18, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lpor.202401904 by Neda Neykova - Czech Technical University in Prague , Wiley Online Library on [10/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.lpr-journal.org Figure 10. The sequence used for L-TCT measurements to study the effect of light-soaking. Figure 11. a) Evolution of hole CWFs under 1-sun illumination. b) Relaxation of CWFs after switching off illumination. Each cycle consists of the previous LS for 250 min and the subsequent 15 h of relaxation in the dark. Measured on MAPB single crystal at 5 Hz pulsed bias of 10 V and 300 μs delay time. To study the reversibility of light-induced changes, we periodically illuminate the sample for 250 min, switch off illumination, and after 15 h of storing the sample in the dark, we again record CWF. We repeat this three times. As seen in Figure 11b, the same shape of CWF is obtained repeatedly. However, it differs from the CWF measured for the first time on the pristine perovskite. Light-induced changes are thus almost reversible. The main difference compared to pristine CWF lies in the disappearance of rapid current drop and the slight improvement of charge carrier collection. To correlate light-induced changes with deep defects evolution, we measure absorptance spectra by photo-thermal deflection spectroscopy (PDS). This method uses the fact that upon the absorption of incident light by a studied sample, a part of or all excitation energy transforms into thermal energy. We use the same wavelength and illumination intensity as in the previous L-TCT measurement. Other experimental details regarding the PDS are in Supporting Information. As seen in Figure 12, the absorptance spectra for pristine perovskite are similar to those measured after sample storage for 67 h in air conditions. However, even 10 min of continuous LED illumination leads to a substantial increase in the sub-bandgap region. This corresponds to the formation of deep defect levels under illumination.[66] The observed phenomena by L-TCT and PDS, specifically the increase in sub-bandgap absorption and the disappearance of rapid transient current drop observed in L-TCT, may be explained by the following mechanism. The material may contain complex defects, which disintegrate into other defects under the influence of light. Therefore, the rapid current drop connected to these complex traps disappears in CWF under illumination. The complex defects probably acted as stronger recombination centers than new defects. Their extinction, therefore, prolongs the lifetime of holes, leading to an increased charge carrier collection under illumination. Increased sub-bandgap absorption under illumination observed in PDS is then given by the contribution of new trap levels. The saturation of charge carrier collection efficiency after 125 min of illumination is due to a finite concentration of complex defects. An equivalent explanation of a rapid current drop disappearance accompanied by an increase in deep defect density may be the light-induced formation of deep defects that passivate stronger recombination centers corresponding to a rapid current drop. As suggested above, the light-induced deep defects most probably correspond to the Frenkel pairs of bromine interstitials and vacancies due to their low formation energies.[58] Figure 12. Evolution of absorptance spectra under 1-sun illumination measured by photo-thermal deflection spectroscopy. Laser Photonics Rev. 2025,19, 2401904 2401904 (8 of 10) © 2025 The Author(s). Laser & Photonics Reviews published by Wiley-VCH GmbH 18638899, 2025, 18, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lpor.202401904 by Neda Neykova - Czech Technical University in Prague , Wiley Online Library on [10/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.lpr-journal.org 3. Conclusion We studied the transport properties of perovskite single crystals, emphasizing light-induced changes. Using L-TCT on MAPB single crystals, we determine a mobility-lifetime product of 6 ×10−4 cm2V−1, hole mobility of 18 cm2V−1s−1, and a lifetime of 33 μs. Combining L-TCT measurements with and without continuous illumination, we observe a light-induced improvement in charge carrier collection. The improvement stems from two effectsreduced surface recombination and increased hole lifetime. The suppression of surface recombination under illumination was attributed to the local increase of internal electric field due to space charge from trapped photo-electrons or accumulated Br− i. The light-induced hole lifetime improvement possibly originates from the shift of the Fermi energy level below the previous hole trap level or from the new defects, which passivate hole trap defects. The light-induced formation of mobile charged defects is also suggested by the evolution of the current response to the bias pulse under continuous illumination. The light-soaking of MAPB leads, similarly to short-term illumination, to an almost reversible improvement in charge collection efficiency. An observed gradual disappearance of a rapid current drop in CWFs under light-soaking corresponds to the passivation of a strong recombination center by new light-induced defects or the disintegration of strong recombination centers into new defects. The formation of new light-induced deep defects, most probably Frenkel pairs of bromine interstitials and vacancies, was proved via PDS by an increasing sub-bandgap absorption under lightsoaking. Supporting Information Supporting Information is available from the Wiley Online Library or from the author. Acknowledgements K.P., L.L., and N.N. acknowledge support from the Czech Science Foundation (GA ˇ CR), Grant No. 23-06543S. N.N. acknowledges support from Czech Ministry of Education, Youth and Sports (Grant No. CZ.02.01.01/00/22_008/0004617). The team members of GAˇ CR 23– 06543S provided sample fabrication, data analysis, L-TCT measurements, and manuscript writing. R.G., E.B., and P.P. acknowledge the support of the Czech Science Foundation, Project No. 23-07951S. The team members of the GAˇ CR 23-07951S provided data analysis, discussed the results, and assisted with L-TCT measurements. The authors also acknowledge the support of the projects LUASK 22202 and PVKSC 9F23003. The team members of LUASK 22202 and PVKSC 9F23003 provided PDS spectra measurements and data analysis. Conflict of Interest The authors declare no conflict of interest. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. Keywords charge carrier transport, deep defects, light-soaking, surface recombination, lifetime, perovskite, single crystal Received: November 8, 2024 Revised: April 27, 2025 Published online: June 4, 2025 [1] H. J. Snaith, J. Phys. Chem. Lett. 2013,4, 3623. [2] H.Wei,Y.Fang,P.Mulligan,W.Chuirazzi,H.H.Fang,C.Wang,B. R. Ecker, Y. Gao, M. A. Loi, L. Cao, J. Huang, Nat. Photonics 2016,10, 333. [3] D. Yang, B. Zhao, T. Yang, R. Lai, D. Lan, R. H. Friend, D. Di, Adv. Funct. Mater. 2022,32, 2109495. [4] S.D.Stranks,G.E.Eperon,G.Grancini,C.Menelaou,M.J.P.Alcocer, T. Leijtens, L. M. Herz, A. Petrozza, J. Henry, S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou, M. P. Alcocer, T. Leijtens, L. M. Herz, A. Petrozza, H. Snaith, Science 2013,342, 341. [5] G.Xing,N.Mathews,S.S.Lim,Y.M.Lam,S.Mhaisalkar,T.C.Sum, Science 2013,6960, 498. [6] S. De Wolf, J. Holovsky, S. J. Moon, P. Löper, B. Niesen, M. Ledinsky, F. J. Haug, J. H. Yum, C. Ballif, J. Phys. Chem. Lett. 2014,5, 1035. [7] M. Ledinsky, A. Vlk, T. Schonfeldova, J. Holovsky, E. Aydin, H. X. Dang, Z. Hajkova, L. Landova, J. Valenta, A. Fejfar, S. de Wolf, J. Phys. Chem. C 2020,124, 27333. [8] K. X. Steirer, P. Schulz, G. Teeter, V. Stevanovic, M. Yang, K. Zhu, J. J. Berry, ACS Energy Lett. 2016,1, 360. [9] D. L. Staebler, C. R. Wronski, Appl. Phys. Lett. 1977,31, 292. [10] R. A. Street, Appl. Phys. Lett. 1983,42, 507. [11] J.-F. Guillemoles, D. Cahen, U. Rau, A. Jasenek, H. Schock, J. Phys. Chem. B 2000,104, 4849. [12] D. Guo, D. Vasileska, 16th International Conference on Numerical Simulation of Optoelectronic Devices, NUSOD 2016, IEEE, 2016, pp. 181. [13] L. Lin, L. Yang, G. Du, X. Li, Y. N. Li, J. Deng, K. Wei, J. Zhang, ACS Appl. Energy Mater. 2023,6, 10303. [14] M. C. Brennan, S. Draguta, P. V. Kamat, M. Kuno, ACS Energy Lett. 2018,3, 204. [15] J. Holovský, A. Peter Amalathas, L. Landová, B. Dzurˇ nák, B. Conrad, M. Ledinský, Z. Hájková, O. Pop-Georgievski, J. Svoboda, T. C. J. Yang, Q. Jeangros, ACS Energy Lett. 2019,4, 3011. [16] J. Holovský, S. De Wolf, J. Werner, Z. Remeš, M. Müller, N. Neykova, M. Ledinský, L. ˇ Cerná, P. Hrzina, P. Löper, B. Niesen, C. Ballif, J. Phys. Chem. Lett. 2017,8, 838. [17] J.Liu,M.Hu,Z.Dai,W.Que,N.P.Padture,Y.Zhou,ACS Energy Lett. 2021,6, 1003. [18] J. Herterich, M. Unmüssig, G. Loukeris, M. Kohlstädt, U. Würfel, Energy Technol. 2021,9, 2001104. [19] H. Tsai, R. Asadpour, J.-C. Blancon, C. Stoumpos, O. Durand, J. W. Strzalka, B. Chen, R. Verduzco, P. Ajayan, S. Tretiak, J. Even, M. A. Alam, M. G. Kanatzidis, W. Nie, A. Mohite, Science 2018,360,67. [20] D. Ghosh, D. Acharya, L. Zhou, W. Nie, O. V. Prezhdo, S. Tretiak, A. J. Neukirch, J. Phys. Chem. Lett. 2019,10, 5000. [21] J. Peng, Y. Sun, Y. Chen, Y. Yao, Z. Liang, ACS Energy Lett. 2016,1, 1000. [22] B. Cai, X. Yang, Z. Yu, Y. Liang, Y. Shan, A. Hagfeldt, L. Sun, J. Power Sources 2020,472, 228506. [23] C. Zhao, B. Chen, X. Qiao, L. Luan, K. Lu, B. Hu, Adv. Energy Mater. 2015,5, 1500279. [24] V. Srivastava, S. H. Reddy, M. Mohan, B. Anitha, B. Adara, M. A. G. Namboothiry, J. Phys. D. Appl. Phys. 2019,52, 265302. Laser Photonics Rev. 2025,19, 2401904 2401904 (9 of 10) © 2025 The Author(s). Laser & Photonics Reviews published by Wiley-VCH GmbH 18638899, 2025, 18, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lpor.202401904 by Neda Neykova - Czech Technical University in Prague , Wiley Online Library on [10/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License