Strategies for Enhancing the Stability of High-Electric-Field Perovskite Single-Crystal Radiation Detectors

Abstract

While a high electric field enhances the sensitivity and response speed of MAPbBr₃ single-crystal radiation detectors, it also triggers interfacial electrochemical corrosion and ion migration, compromising device stability. This study reveals that multivalent titanium ion (Ti⁴⁺) migration in MAPbBr₃ single-crystal radiation detectors at a high electric field requires grain boundary (GBs) assistance and is confined to the polycrystalline regions, located hundreds of nanometers from the crystal surface. Consequently, device performance can be fully restored by removing the corroded surface and redepositing the electrode. Furthermore, our results also illustrate that humidity and oxygen in the atmosphere can accelerate the electrochemical reaction at the Ti/perovskite single-crystal interface, highlighting the importance of encapsulation for perovskite single-crystal-based devices.

High-energy radiation detectors play a vital role in many fields, such as medical diagnostics and industrial nondestructive inspection. As a new generation of semiconductors, perovskites have shown great advantages in the field of radiation detection. Due to the excellent photoelectric properties such as large absorption coefficient, high carrier mobility and large resistivity, perovskite-based detectors exhibit high sensitivity and low detection limit, which contribute to improve resolution and reduce damage induced by high-energy radiation. ,

High electric field enhances the sensitivity and reduces the response time, but it brings in many issues to perovskite radiation detectors. The electric field accelerates the movement of halogen ions (such as Br^–, I^–) toward the high-potential terminal, leading to chemical reaction between halogen ions and metal electrodes, such as Au, Al, Ag. − On the one hand, ion migration assisted by electric field leads to the hysteresis of current and the decomposition of perovskite, which destabilized the stability of the device. On the other hand, the oxidized metal ions could drift into the perovskite at the electric field, introducing deep level defects in perovskite. , The corrosion of metal electrodes is also considered to be the main cause of device degradation, especially at high electric fields. ,, It has been reported that migration pathways of oxidized metal ions were correlated with their valent state. Monovalent noble metal cations (Au⁺, Ag⁺) have been reported to be easily formed in the interface of MAPbI₃/Au or Ag and migrate through the perovskite bulk via the same sites as iodine interstitials. The doping of Au⁺ modified the work function of MAPbI₃, enhancing its conductivity and leading to a significant increase in current density. While metal impurity with a higher charge state have been calculated to possess a higher interstitial diffusion barrier, meaning that their movement through the lattice is more restricted compared with monovalent species. For instance, the memory turn-on characteristics of perovskite-film devices with Al electrodes suggest that Al³⁺ ions migrate along grain boundaries (GBs). The hysteresis behavior of photocurrent and dark current is significantly more pronounced at GBs than that in grain interiors. This phenomenon can be attributed to the presence of fast ion migration pathways in polycrystalline perovskite. Although the interfacial electrochemical reactions in polycrystalline perovskite films with commonly used metal electrodes (Au, Ag, Al) have been extensively studied, the degradation mechanisms and failure pathways of single-crystal perovskite detectors at high electric fields remain insufficiently explored.

Our recent work has demonstrated that a MAPbBr₃ single-crystal X-ray detector with a Ti electrode as high-potential terminal exhibited significantly enhanced stability in dark current compared to one with an Au electrode when subjected to a reverse electric field of 2.5 V mm^–1 for 10 h. The enhanced stability can be attributed to the formation of a protective layer in the form of Ti–N bonds at the Ti/MAPbBr₃ single crystals interface. In contrast, devices with Au electrodes tend to suffer from the formation and migration of Au⁺ ions into the perovskite layer, leading to increased noise level in the devices. The Ti-based detector demonstrated significantly enhanced sensitivity when operated at a high electric field of 50 V mm^–1 while maintaining a low detection limit. Such a high electric field is beneficial for achieving high-resolution imaging; however, it has been observed that the devices exhibit a significant increase in dark current after operating for tens of minutes. The failure mechanisms of these devices at high electric fields remain unclear and require further investigation to enable the development of highly stable, high-resolution detectors for potential commercialization.

In this work, we investigated the electrochemical reactions at the interface of MAPbBr₃ single crystals/Ti electrode at a high electric field of 50 V mm^–1. Our results indicate that prolonged operation at a high electric field can break Ti–N bonds at the perovskite/Ti interface, leading to the oxidation of Ti to Ti⁴⁺ ions. These oxidized Ti⁴⁺ ions diffuse only into a shallow superficial region where the perovskite exhibits polycrystalline properties due to the crystal growth process and do not penetrate deeply into the inner single crystals without the assistance of GBs, leaving the structure and properties in the bulk undamaged. This finding can be verified by removing the damaged Ti electrode and the corrosive surface of the perovskite single crystals, followed by redepositing a fresh Ti electrode. The treated device can then be restored to its initial performance, which also presents a viable strategy for restoring the performance of devices based on perovskite single crystals. In addition, the results of GIXRD further confirm that the polycrystalline region exists in as-grown MAPbBr₃ single crystals, extending to hundreds of nanometers from the surface. The results of DFT calculations indicate a high energy barrier for interstitial diffusion of multivalent Ti⁴⁺ ions (∼2.0 eV), suggesting that their rapid migration requires the assistance of GBs. Furthermore, we find that moisture and oxygen in atmosphere can accelerate the electrochemical reaction at the interface of perovskite/Ti electrode. In contrast, the device based on MAPbBr₃ single crystal with Ti electrode in vacuum shows excellent stability, which can work stably at the electric field of 50 V mm^–1 for 30 h without electrochemical reactions. Our work contributes to a better understanding of the degradation mechanisms of perovskite detectors at high electric fields and provides the guidance for improving the stability of radiation detectors based on perovskite single crystals.

We test the long-term stability of the MAPbBr₃ single crystal devices based on a Ti electrode as a high-potential terminal at a high electric field of 50 V mm^–1, and the results are shown in Figure (a) and Figure S1. We measure four single crystals, and the curves show the consistent trend. The overall trend can be divided into four stages: (1) The current is stable around 70 nA; (2) The current increases rapidly with time; (3) The current entered a stage of slow increase; (4) The curve of current shows the great noise fluctuations. We believe that this trend of current is related to the electrochemical reaction at the interface and the migration of metal ions within the perovskite single crystals. Therefore, we used secondary ion mass spectrometry (TOF-SIMS) to test the distribution of migrating ions during the four stages, and the results are shown in Figure (b). In the first stage, the distribution of Ti⁺ in perovskite single crystal is consistent with that of fresh samples, which show that Ti⁺ ions did not move. In the second stage, Ti⁺ ions migrate rapidly in MAPbBr₃ single crystals. In the third and fourth stages, Ti⁺ ions barely continue to drift into the interior of MAPbBr₃ single crystals and almost exclusively accumulate within shallow depth regions of the MAPbBr₃ single crystal. The distributions of Br^– ions on the high-potential electrode side (Ti electrode) were also characterized for both the fresh device and the device at an electric field of 50 V mm^–1 for 24 h, as shown in the Figure S2. The results showed that with the assistance of an electric field, Br^– ions diffused into the high-potential electrode, and eventually the Ti electrode is corroded to form TiBr₄.

1.

Operational stability and degradation mechanism of Ti/MAPbBr₃ single-crystal devices under a high reverse electric field (50 V mm^–1). (a) Dark current evolution over 24 h in a pristine device. (b) TOF-SIMS depth profiles of Ti⁺ fragments after various operational durations. (c, d) XPS spectra of Pb 4f and Ti 2p at the interface after 0, 10 min, and 2 h of operation, revealing interfacial chemical evolution. (e) Dark current stability over 12 h for device based on polished MAPbBr₃ single crystal, inset shows electrode morphology transition. (f) Restored device performance after restoring, demonstrating reversibility.

XPS measurement is carried out to study the interface of Ti/perovskite single crystals during the transitional period from stage 1 to stage 2. The Pb 4f and Ti 2p spectra at the interface of Ti/MAPbBr₃ single crystals are measured, and the results are shown in Figure (c) and (d). In the first stage, the ratio of Pb⁰/Pb²⁺ is consistent with that of fresh samples, indicating that the electrochemical reaction does not occur at the interface. The results of Ti 2p spectra show that the electrochemical reaction is inhibited by a Ti–N bond protective layer at the surface of Ti/MAPbBr₃ single crystals at high electric field. However, with the increase of operational time, the Ti–N protective layer gradually disappears. Finally, the Ti electrode is oxidized to Ti⁴⁺, and Pb²⁺ in MAPbBr₃ single crystals is gradually reduced to Pb⁰. The oxidized Ti⁴⁺ ions migrate into the perovskite single crystals at high electric field, resulting in the continuous increase of dark current.

According to the results of TOF-SIMS, we believe that the transition from stage 2 to stage 3 can be attributed to the fact that the rapid migration of multivalent Ti⁴⁺ ions requires the assistance of GBs. Actually, perovskite single crystals can be divided into two parts. During the later stages of solution-based MAPbBr₃ single crystal growth, the depletion of solute leads to a reduced precursor concentration, which increases the randomness of crystallization orientations. As a result, the regions near the crystal surface exhibit polycrystalline characteristics. In contrast, the crystal interior contains fewer defects and GBs, and it can be considered as a perfect single crystal region. The fast movement of multivalent Ti⁴⁺ ions assisted by GBs in the polycrystalline region causes the current to rise rapidly in stage 2. However, this migration process of Ti⁴⁺ ions almost ceases upon reaching the perfect single crystal region due to the lack of GBs. As a result, the drift of current becomes slower in stage 3.

To confirm our speculation, we first polish the MAPbBr₃ single crystal to remove the polycrystalline region near the surface. We test the current–time curves of the device based on a polished MAPbBr₃ single crystal, and the results are shown in Figure (e). The current–time curve has only two phases: (1) The current remains stable at around 70 nA (corresponding to stage 1 of devices based on an as-grown MAPbBr₃ single crystal). (2) The current appears with a violent fluctuation (corresponding to stage 4 of devices based on as-grown MAPbBr₃ single crystal). Stage 2 and stage 3 in devices based on as-grown single crystals, which correspond to the stages of Ti⁴⁺ ion migration in the polycrystalline region near the surface of MAPbBr₃ single crystals, do not appear in the device based on polished single crystals. It is further explained that the stage 2 and stage 3 of the devices based on as-grown MAPbBr₃ single crystals are caused by the region of polycrystalline near the surface of MAPbBr₃ single crystals. After the polycrystalline region was removed by polishing, it is difficult for Ti⁴⁺ ions to enter a perfect single crystal region in the bulk due to the lack of GBs. Therefore, stage 2 (current increases rapidly) and stage 3 (current increases slowly) are not observed in the polished perovskite single crystals.

Furthermore, we investigated the changes that occur in the devices based on polished MAPbBr₃ single crystals in stage 4. We find that the electrodes of devices always show significant damage at the electrical field of 50 V mm^–1 after 24 h, so we guess that the breakdown of device performance in stage 4 caused by the damage of Ti electrode. Devices based on the polished single crystal are operated at the electric field of 50 V mm^–1 for 10 h, and the changes of electrodes are observed every 3 h. The results are shown in Figure (e) and Figure S3. When the device works for about nine h, the morphology and color of the electrode change obviously, and the hole appears in the center of the electrode. At the same time, sharp noise fluctuations emerged in the current–time curves, coinciding with the onset of stage 4, which indicates that the deterioration of the device performance in stage 4 is caused by damage to the electrode.

In addition, we measure the on–off X-ray responses of a fresh device based on polished MAPbBr₃ single crystal for a range of dose rates from 180 to 9 nGy s^–1 at different electric field, and the results are shown in Figure S4­(a). After operating at an electric field of 50 V mm^–1 for 12h, the Ti electrode is damaged. We then remove the damaged electrode and corrosive surface of MAPbBr₃ single crystal by polishing, and redeposit Ti on the surface of polished MAPbBr₃ single crystal as the fresh electrode. The trend of the dark current with time is shown in Figure (f). The on–off X-ray responses of the device in the same dose rate range are tested again at different electric fields, as shown in Figure S4­(b). The sensitivities of the fresh device and device after redepositing Ti electrodes at different electric fields are obtained by fitting the linear relationship between the X-ray response and the X-ray dose, as shown in Figure (a) and Figure S5. The signal-to-noise ratio (SNR) is also measured at different electric fields for a range of dose rates from 180 to 9 nGy s^–1, as shown in Figure S4. According to IUPAC standards, the detection limit is defined as the equivalent dose rate with an SNR value of 3. Based on the linear extrapolation signal-to-noise ratio (SNR) at different doses, the detection limits of the fresh device and device after redepositing Ti electrodes are obtained, as shown in Figure (b) and Figure S6. Sensitivity and detection limits measured at different voltages before and after Ti redeposition was summarized in a tabular format, and we calculated the recovery efficiency, as shown in Table S1. The results show that the recovery efficiency exceeds 75% for the detection limit and reaches over 95% for the sensitivity. Sensitivities and the detection limits of the device after redepositing Ti electrodes generally recover to the original state, which verifies that the appearance of stage 4 in Figure (a) and (e) is caused by the damaged electrode. This exciting results also show that the electrochemical reaction between Ti electrode and MAPbBr₃ single crystals and subsequent ion migration occur almost exclusively near the interface and have little effect on the structure and properties in the bulk of single crystal. Moreover, it also provides a method to restore the performance of the radiation detectors based on perovskite single crystals, and the failed devices based on Ti electrode can be recycled by removing the damaged electrode and the corroded surface, and redepositing the fresh Ti electrode.

2.

Radiation performance of original and recovered devices based on MAPbBr₃ single crystal with Ti electrodes as high-potential terminal. (a) Sensitivities and (b) detection limits of the device based on MAPbBr₃ single crystal with Ti electrode at different electric fields.

Further, we research the products after the Ti electrode is damaged and discuss the cause of the severe noise fluctuation in stage 4. The XPS of Br 3d, Ti 2p, and O 1s at the distance of ∼20 nm from the surface of Ti electrode (the total thickness of 70 nm) are measured by etching method for fresh device and device operating at a reverse electric field of 50 V mm^–1 for 12 h. The test region is not at the surface of the electrode, so the effect of surface contamination can be excluded. The results show that the Br^– ions in the damaged Ti electrode (corresponding to the peaks at 68.8 and 70.4 eV) increased significantly, as shown in Figure (a). The content of Ti⁰ (corresponding to the peaks at 454.0 and 460.0 eV) decreases in the damaged electrode, and the content of Ti⁴⁺ ions increases (corresponding to the peaks at 458.5 and 464.2 eV), as shown in Figure (b). Therefore, the Ti electrode is oxidized to TiBr₄ by Br^– ions in MAPbBr₃ single crystals, resulting in the breakage of the Ti electrode and the failure of the normal conductive circuit inside the device. Finally, the leakage current in the environment around the device increases and the violent noise fluctuations appear in stage 4. Impedance spectroscopy was also employed to probe the failure mechanism of the device after it was operated at 50 V mm^–1 for 10 h (Figure S7). The pristine device exhibited a characteristic semicircle at high frequencies and a 45°-sloped line at low frequencies. In contrast, the failed device displayed a near-straight line at high frequencies, signifying a shunting path with very low resistance (<10⁶ Ω), attributable to moisture in air. This, coupled with the frequency-dependent capacitive reactance, confirms severe current leakage through the external surface, rendering the internal perovskite pathway nonfunctional. The low-frequency region featured an ∼45° sloped line, resembling Warburg impedance, which is attributed to charge diffusion and accumulation at the electrode/air interface, likely involving restricted migration of ions from moisture.

3.

Electrochemical reaction of devices in stage 4. (a) Br 3d, (b) Ti 2p, and (c) O 1s spectra of Ti electrode for fresh device and device operating at a reverse electric field of 50 V mm^–1 for 12 h in atmosphere. (d) The dark current–time curves of device based on MAPbBr₃ single crystal with Ti electrode at a reverse electric field of 50 V mm^–1 in vacuum.

The critical role of the ambient environment in device failure was further confirmed by switching the measurement atmosphere to a degraded device (Figure S8­(a)). Under a high vacuum (10^–3 Pa), the current dropped to ∼25 nA, and no significant noise fluctuations were observed, as shown in Figure S8­(b). The reintroduction of air caused an immediate resurgence of the current level to ∼400 nA, accompanied by intense noise fluctuations, as shown in Figure S8­(c). These results conclusively demonstrated that the observed noise in stage 4 stemmed from a current leakage path established through the ambient environment due to electrode corrosion.

Moreover, the area surrounded by the peak (531.0 and 532.0 eV) of the O 1s spectra increases obviously, which shows that the content of adsorbed oxygen increases after the Ti electrode is damaged, as shown in Figure (c). The results show that moisture and oxygen in atmosphere also participate in this electrochemical reaction and accelerate the corrosion of the electrode, which is also consistent with previous reports. ,

Considering the influence of the environment on the electrical performance of the devices, we measure the dark current–time curves of the device based on fresh MAPbBr₃ single crystal with Ti electrode in vacuum at the electric field of 50 V mm^–1, as shown in Figure (d). Obviously, at the same electric field, the dark current in vacuum is significantly reduced. Compared to devices in atmosphere, the device in vacuum is more stable and noise fluctuations does not appear after continuous operation for 30 h. Furthermore, we measure the XPS spectra of the Ti electrode after the device operated for 30 h at an electric field of 50 V mm^–1 in a vacuum, as shown in Figure S9. Compared with the fresh device, the XPS results of the device based on MAPbBr₃ single crystals with Ti electrode operating continuously for 30h in vacuum do not show significant change, which indicates that the electrochemical reaction of device in vacuum is effectively inhibited. Our results show that the strategy of encapsulation is important for radiation detectors based on perovskite single crystals.

Microstructure of the as-grown MAPbBr₃ single crystal surface can be nondestructively analyzed by grazing incidence X-ray diffraction (GIXRD), and the results are shown in the Figure (a). The depth profile information on MAPbBr₃ single crystal structure is obtained by performing XRD with fixed incidence angle of 0.5°, 1°, and 4.5°. When the incident angle is very small (0.5° and 1°), the XRD pattern shows not only the (100) peak and (200) peak, but also the (210) peak, indicating the presence of polycrystals regions near the surface. As the incident angle increases to 4.5°, the (300) peak appears for the first time in the XRD pattern, which aligns with the XRD spectrum obtained at a high angle of 30° (reflecting the bulk phase of the as-grown MAPbBr₃ single crystal and exhibiting single-crystal characteristics). The results of GIXRD indicate that the as-grown MAPbBr₃ single crystals consist of two parts: the polycrystalline region near the surface and the perfect single crystal region in the bulk, which is consistent with the prior hypotheses. Moreover, X-ray penetration into the perfect crystal region in the bulk occurs at grazing angles above ∼4.5°.

4.

Diffusion path of Ti⁴⁺ ions. (a) GIXRD patterns of as-grown MAPbBr₃ single crystals measured with 0.5°, 1°, and 4.5° grazing angle. Powder XRD and high-angle incidence XRD (30° angle) of the as-grown MAPbBr₃ single crystal are employed for comparative analysis. (b) Schematic for the diffusion path of an interstitial metal Ti⁴⁺ ions in MAPbBr₃. (c) Potential energy evolution of Ti⁴⁺ ions along the interstitial diffusion pathway.

In the GIXRD setup, X-ray penetration depth is determined by the incident angle and material parameters. The standard definition of penetration depth refers to the distance where the intensity of X-rays decays to 1/e (approximately 37%) of its initial value upon entering the surface. According to the attenuation law of X-rays in samples, the penetration depth of X-rays, z, in perovskite materials can be calculated according to the equation:

$$\mathit{z}=\frac{\sin {\alpha }_{\mathrm{i}}}{2\mu }$$ 1

where α_i is the incidence angle and μ is the attenuation coefficient. The attenuation coefficients (μ) of MAPbBr₃ compounds are determined by applying the additive rule to constituent elemental mass attenuation coefficients, (μ/ρ)_i:

$$\mu =\rho ({\mathrm{MAPbBr}}_3)\sum _{\mathrm{i}}{\mathit{w}}_{\mathrm{i}}{(\mu /\rho )}_{\mathrm{i}}$$ 2

where w _i is the mass fraction of the certain element in the MAPbBr₃ compound. ρ­(MAPbBr₃) is the density of the MAPbBr₃ samples (ρ = 4.0 g cm^–3). The values for the elemental mass attenuation coefficients, (μ/ρ)_i, of C, H, N, Pb, and Br for the Cu–Kα radiation energy are taken from ref . Ultimately, the result of the calculated absorption coefficient, μ, is 563 cm^–1, and the results of the penetration depths corresponding to the incident angles of 0.5°, 1°, and 4.5° calculated based on eq are 78, 155, and 697 nm, respectively. Therefore, we estimate that the polycrystalline region extends to a depth of approximately 697 nm from the surface. In addition, GIXRD patterns from multiple crystal batches consistently showed the (300) diffraction peak first appeared at a grazing angle of ∼4.5°, delineating a polycrystalline surface layer, as shown in the Figure S10. The depth of this layer is reproducible across batches, being primarily governed by the controlled precursor concentration (1.6 mol L^–1) and crystal dimensions (∼6 × 6 × 2 mm³). Moreover, the presence of a polycrystalline layer was further corroborated by SEM, which directly revealed granular features on the as-grown crystal surface (Figure S10), in strong support of our hypothesis.

Furthermore, through density functional theory (DFT) calculations, we determined the interstitial diffusion energy barrier of Ti⁴⁺ ions in MAPbBr₃ single crystals. DFT are implemented in the plane-wave basis QE code. Experimental lattice constants of the room temperature cubic phase MAPbBr₃ are used: a = b = c = 5.9173 Å. The schematic diagram of the diffusion pathway is shown in Figure (b). The result, as illustrated in the Figure (c), demonstrate a Ti⁴⁺ diffusion energy barrier of ∼2.0 eV, which is higher than the interstitial diffusion energy barrier of the monovalent noble metals reported in the literature. Therefore, we can reasonably conclude that Ti⁴⁺ ion diffusion requires the assistance of GBs. Diffusion through the lattice interstitials is energetically unfavorable.

Based on the obtained data and analyses, we give a schematic diagram of the migration process of Ti⁴⁺ ions at high electric field. Perovskite single crystals can be seen as consisting of two parts: the perfect single crystal region in the bulk and the polycrystalline region near the surface, as shown in Figure (a). There are a large number of defects and GBs in the polycrystalline region, which act as channels for Ti⁴⁺ ion migration. However, in the perfect single crystal region without GBs, the Ti⁴⁺ ions migrate slowly. First, a Ti–N bond protective layer is preferentially formed at the interface of Ti/MAPbBr₃, which effectively inhibits the electrochemical reaction at the interface, as shown in Figure (b). With the increase of operating time, the protective layer of the Ti–N bond is destroyed, and the oxidized Ti⁴⁺ ions rapidly migrate into the polycrystalline region assisted by GBs at the electric field, as shown in Figure (c), causing the current to drift significantly. However, the migration process of Ti⁴⁺ ions almost ceases upon reaching the perfect single crystal region, where GBs are absent. At this time, the current drifts slowly. Finally, the Ti electrode is damaged due to the electrochemical reaction at long-term and high electric fields, and the device fails completely, as shown in Figure (d).

5.

Schematic diagram of the degradation process for device operating at high electric fields. (a) The structure of device based on MAPbBr₃ single crystal contacted with Ti electrode as a high-potential terminal. (b) Stage 1: Ti–N protective layer formed at the interface of Ti/MAPbBr₃ single crystal. (c) Stages 2 and 3: protective layer invalid and the oxidized Ti⁴⁺ ion migrating along the GBs. (d) Stage 4: Ti electrode was damaged due to electrochemical reaction.

In our study, we show the trend of current with time in MAPbBr₃ single crystal device based on Ti electrode at the high electric field of 50 V mm^–1. The trend can be divided into four stages: (1) The current curve remains stable, (2) the current increases rapidly, (3) the current enters the stage of slow increase, and (4) the current curve fluctuates violently. We believe that this regular variation is because that the rapid movement of Ti⁴⁺ ions assisted by GBs almost ceases upon reaching the perfect single crystal region in the bulk, where GBs are absent. Finally, Ti electrode is oxidized to TiBr₄ due to the electrochemical reaction, and the electrical performance of the device completely breakdown. The migration of multivalent metal ions Ti⁴⁺ does not damage the properties and structures in the bulk of perovskite single crystals. Therefore, when the damaged Ti electrode and the corrosive region at surface are removed, and fresh Ti electrodes are redeposited, the performance of MAPbBr₃ single crystal devices can be basically restored to the initial state. This provides a feasible scheme for the performance recovery of device based on perovskite single crystals. Moreover, we find that moisture and oxygen in the atmosphere can accelerate the electrochemical reaction at the interface, and the device based on MAPbBr₃ single crystals with a Ti electrode can exhibit superior stability in a vacuum, which shows that the strategy of encapsulation is important for devices based on perovskite single crystals. Our research is very significant for improving the long-term stability of radiation detectors based on perovskite single crystals.

Experimental Methods

Experimental details are provided in the Supporting Information.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsenergylett.5c02178.

• General experimental methods; Photographs of the degradation process for Ti electrode; X-ray on–off response measures at different electric field; Fitting curves of the sensitivity and detection limits of the device at different electric field; XPS spectra of Ti electrode (PDF)

The authors declare no competing financial interest.