The Total Ionizing Dose Effects on Perovskite CsPbBr3 Semiconductor Detector

Perovskite CsPbBr3 semiconductors exhibit unusually high defect tolerance leading to outstanding and unique optoelectronic properties, demonstrating strong potential for γ-radiation and X-ray detection at room temperature. However, the total dose effects of the perovskite CsPbBr3 must be considered when working in a long-term radiation environment. In this work, the Schottky type of perovskite CsPbBr3 detector was fabricated. Their electrical characteristics and γ-ray response were investigated before and after 60Co γ ray irradiation with 100 and 200 krad (Si) doses. The γ-ray response of the Schottky-type planar CsPbBr3 detector degrades significantly with the increase in total dose. At the total dose of 200 krad(Si), the spectral resolving ability to γ-ray response of the CsPbBr3 detector has disappeared. However, with annealing at room temperature for one week, the device’s performance was partially recovered. Therefore, these results indicate that the total dose effects strongly influence the detector performance of the perovskite CsPbBr3 semiconductor. Notably, it is concluded that the radiation-induced defects are not permanent, which could be mitigated even at room temperature. We believe this work could guide the development of perovskite detectors, especially under harsh radiation conditions.


Introduction
The accurate detection of γ-rays from radioactive sources in high resolution at room temperature is of significant interest for industrial, medical imaging, and space applications [1][2][3]. However, the γ-ray photons are in very low flux. Notably, most semiconductors are susceptible to even a small number of defects and impurities that invariably act as signal-killing carrier traps. The traditional γ-ray detector materials are very few and face many challenges. The emergence of inorganic perovskite CsPbBr 3 with surprisingly good defect tolerance, which is very promising for high-resolution X-ray and γ-ray detection at room temperature, has attracted extensive attention [4][5][6][7][8][9][10][11][12][13][14][15]. Additionally, the CsPbBr 3 detector has shown remarkable energy resolving capability under both X and γ rays, particularly achieving 3.2% (3.9 keV, FWHM) energy resolution for 122 keV 57 Co γ-ray [10,16]. Furthermore, the lifetime of the hole in CsPbBr 3 detector-grade single crystal was observed to be over 25 µs [16].
For an excellent detector, device stability in a radiation environment is also critical. In addition, electronic devices for space, scientific and medical applications must be tolerant of a wide range of total ionizing doses (TID). However, the spectral degradation of In addition, electronic devices for space, scientific and medical applications must be erant of a wide range of total ionizing doses (TID). However, the spectral degradation the perovskite CsPbBr3 detector working in high irradiation environments has not be investigated [17][18][19][20].
In this work, we report a first investigation of the spectral response variation of C bBr3 detectors under total dose irradiation. The CsPbBr3 detector was fabricated with asymmetric electrode design, in which the γ-ray response of the Schottky-type pla CsPbBr3 detector was revealed under different total doses. Additionally, the effect of irradiation damage was evaluated through annealing experiments. Therefore, it is c cluded that this device can work stably under the total ionizing dose of 100 krad(Si). believe that our work will significantly promote and provide guidance for further inv tigations on the application of halide perovskites.

CsPbBr3 Detector Fabrication
High-quality CsPbBr3 single crystals were grown from melt using the Bridgm method, as shown in Figure 1a. After mechanical polishing, the surfaces were deposi with the metal contact (Au and Ga/In), and the typical thickness of gold contact was nm. After electrode preparation, carbon paste was used to connect the metal electro with the collection circuit for the front-end electronics. Finally, the test sample is sea with wax. And the top-view photograph of the CsPbBr3 detectors was illustrated in Fig  1.b. More information on the samples is detailed in our previous work [16].

Irradiation
The CsPbBr3 detectors were irradiated by γ-rays derived from the 60 Co source (at Northwest Institute of Nuclear Technology, Xi'an, China) and had an energy distribut centered typically around 1.25 MeV. The detector was unbiased, with all pins float during 60 Co γ radiation, and tested at a dose rate of 50.0 rad(Si)/s. The sample's curre voltage (IV) characteristics were measured at the total ionizing doses of 100 and krad(Si), respectively. The spectrum response of the sample was measured at the to ionizing doses of 100 and 200 krad(Si). Finally, the sample's tested parameters were me ured within two hours after each radiation step. As indicated in the insert, the dimension of the crystal is 3 × 3 mm 2 , where the thickness is 1 mm.

Irradiation
The CsPbBr 3 detectors were irradiated by γ-rays derived from the 60 Co source (at the Northwest Institute of Nuclear Technology, Xi'an, China) and had an energy distribution centered typically around 1.25 MeV. The detector was unbiased, with all pins floating during 60 Co γ radiation, and tested at a dose rate of 50.0 rad(Si)/s. The sample's currentvoltage (IV) characteristics were measured at the total ionizing doses of 100 and 200 krad(Si), respectively. The spectrum response of the sample was measured at the total ionizing doses of 100 and 200 krad(Si). Finally, the sample's tested parameters were measured within two hours after each radiation step.

Measurements
The current-voltage (IV) characteristics of the CsPbBr 3 detector were measured using a Keysight B1500A semiconductor device analyzer. The response was tested with a 59.5 keV 241 Am γ-ray source. As shown in Figure 2, the CsPbBr 3 detectors were positioned inside a vacuum chamber to detect γ-rays. The response signal induced in the CsPbBr 3 detector was, in turn, transmitted to the following equipment through coaxial cables: an ORTEC-142A preamplifier, an ORTEC-672 amplifier (shaping time of 3 µs, gain of 1500 times), an ORTEC-ASPEC-927 multi-channel analyzer (MCA), and the MAESTRO software was used to acquire the γ-ray spectrally. An SRS PS350, a high-voltage supply, was used to provide the reverse bias to the CsPbBr 3 detector through the preamplifier.

Measurements
The current-voltage (IV) characteristics of the CsPbBr3 detector were measured using a Keysight B1500A semiconductor device analyzer. The response was tested with a 59.5 keV 241 Am γ-ray source. As shown in Figure 2, the CsPbBr3 detectors were positioned inside a vacuum chamber to detect γ-rays. The response signal induced in the CsPbBr3 detector was, in turn, transmitted to the following equipment through coaxial cables: an OR-TEC-142A preamplifier, an ORTEC-672 amplifier (shaping time of 3 μs, gain of 1500 times), an ORTEC-ASPEC-927 multi-channel analyzer (MCA), and the MAESTRO software was used to acquire the γ-ray spectrally. An SRS PS350, a high-voltage supply, was used to provide the reverse bias to the CsPbBr3 detector through the preamplifier.

I-V and I-T Characteristics
As shown in Figure 3, The pristine I-V curve of the CsPbBr3 detector on the reverse side was compared with the reverse side I-V curves taken after the irradiation with the total dose of 100 krad and 200 krad, respectively. During the test, the sweep voltage was set from −100 V to 0 V, with 4 s intervals between each voltage step. As indicated in Figure  3, with the increase in total dose, there was a decrease in the dark current of the CsPbBr3 detector. Notably, the higher the bias voltage applied, the more the radiation effect will be.

I-V and I-T Characteristics
As shown in Figure 3, The pristine I-V curve of the CsPbBr 3 detector on the reverse side was compared with the reverse side I-V curves taken after the irradiation with the total dose of 100 krad and 200 krad, respectively. During the test, the sweep voltage was set from −100 V to 0 V, with 4 s intervals between each voltage step. As indicated in Figure 3, with the increase in total dose, there was a decrease in the dark current of the CsPbBr 3 detector. Notably, the higher the bias voltage applied, the more the radiation effect will be.

Measurements
The current-voltage (IV) characteristics of the CsPbBr3 detector were measured using a Keysight B1500A semiconductor device analyzer. The response was tested with a 59.5 keV 241 Am γ-ray source. As shown in Figure 2, the CsPbBr3 detectors were positioned inside a vacuum chamber to detect γ-rays. The response signal induced in the CsPbBr3 detector was, in turn, transmitted to the following equipment through coaxial cables: an OR-TEC-142A preamplifier, an ORTEC-672 amplifier (shaping time of 3 μs, gain of 1500 times), an ORTEC-ASPEC-927 multi-channel analyzer (MCA), and the MAESTRO software was used to acquire the γ-ray spectrally. An SRS PS350, a high-voltage supply, was used to provide the reverse bias to the CsPbBr3 detector through the preamplifier.

I-V and I-T Characteristics
As shown in Figure 3, The pristine I-V curve of the CsPbBr3 detector on the reverse side was compared with the reverse side I-V curves taken after the irradiation with the total dose of 100 krad and 200 krad, respectively. During the test, the sweep voltage was set from −100 V to 0 V, with 4 s intervals between each voltage step. As indicated in Figure  3, with the increase in total dose, there was a decrease in the dark current of the CsPbBr3 detector. Notably, the higher the bias voltage applied, the more the radiation effect will be.   The diffusion model has been utilized for analyzing the I-V characteristics of the CsPbBr 3 detector [21][22][23][24][25]. In the diffusion model, the current density J under the reverse bias V in a metal-semiconductor Schottky device can be described as: where q is the electron charge, µ is the carrier mobility, N c is the effective density of states in the conduction band, N i is the concentration of the ionized donor/acceptor centers, ε is the electrical permittivity in the crystal, d is the thickness of the device, V in is the built-in internal electric field, k is Boltzmann's constant, T is temperature, and φ is Schottky barrier [24]. Accordingly, the J-V curve under the reverse bias of a CsPbBr 3 device was fitted according to the above J-V equation, as plotted in Figure 4.
Sensors 2023, 23,2017 The diffusion model has been utilized for analyzing the I-V characteristics CsPbBr3 detector [21][22][23][24][25]. In the diffusion model, the current density J under the bias V in a metal-semiconductor Schottky device can be described as: where q is the electron charge, μ is the carrier mobility, Nc is the effective density o in the conduction band, Ni is the concentration of the ionized donor/acceptor cente the electrical permittivity in the crystal, d is the thickness of the device, Vin is the b internal electric field, k is Boltzmann's constant, T is temperature, and ϕ is Schottky [24]. Accordingly, the J-V curve under the reverse bias of a CsPbBr3 device wa according to the above J-V equation, as plotted in Figure 4. Accordingly, the J-V curve under the reverse bias of an MSM device was fit cording to the above J-V equation, as plotted in Figure 2. The J-V curve in the me range followed the tendency predicted by the diffusion model. Before irradiati Schottky barrier extracted was 0.72 eV, while u is 53.6 cm 2 /Vs. Following irradiati Schottky barrier was not changed, while the μ was dropped with the total dose, a cm2/Vs for 100 krad(Si) 38.6 cm2/Vs for 200 krad(Si). Thus, we considered that the tion induced defects in the CsPbBr3 detector and is the main reason for the decr mobility.
The total ionizing dose impacted the photoelectric response of the CsPbBr3 d Figure 5 shows the time-dependent currents under the light pulses and 50 V bi dark current was about 0.7-1.9 nA. Before radiation, the photocurrent was around and the ON-OFF ratio was about 80. Then, with the increase in total dose, the photo and the ON-OFF ratio dropped. For the total dose of 200 krad(Si), the photocurre around 32 nA, and the ON-OFF ratio was reduced to 35. Therefore, this finding in that electrical-active defects were likely formed after total irradiation, resulting in th ping of photo-generated carriers.  Accordingly, the J-V curve under the reverse bias of an MSM device was fitted according to the above J-V equation, as plotted in Figure 2. The J-V curve in the measured range followed the tendency predicted by the diffusion model. Before irradiation, the Schottky barrier extracted was 0.72 eV, while u is 53.6 cm 2 /Vs. Following irradiation, the Schottky barrier was not changed, while the µ was dropped with the total dose, and 44.9 cm 2 /Vs for 100 krad(Si) 38.6 cm 2 /Vs for 200 krad(Si). Thus, we considered that the radiation induced defects in the CsPbBr3 detector and is the main reason for the decrease in mobility.
The total ionizing dose impacted the photoelectric response of the CsPbBr 3 detector. Figure 5 shows the time-dependent currents under the light pulses and 50 V bias. The dark current was about 0.7-1.9 nA. Before radiation, the photocurrent was around 85 nA, and the ON-OFF ratio was about 80. Then, with the increase in total dose, the photocurrent and the ON-OFF ratio dropped. For the total dose of 200 krad(Si), the photocurrent was around 32 nA, and the ON-OFF ratio was reduced to 35. Therefore, this finding indicates that electrical-active defects were likely formed after total irradiation, resulting in the trapping of photo-generated carriers.

γ-ray Spectral Performance
The γ-ray spectral performance of the CsPbBr3 detector was also influenced ionizing dose radiation. The energy resolution is a critical characteristic for evalua ability of energy spectral measurement of the detectors. The detector's energy re is usually defined as the ratio between the FWHM and the position of the peak c which is a dimensionless fraction expressed as a percentage. Figure 6 shows the resolution of CsPbBr3 detectors to 241 Am γ-ray with various bias voltages (-50 V and -150 V) before and after 60 Co γ ray total dose irradiation. During the energy test, the radiation-induced charge Q collection was on the Au cathode, while Ga/I was grounded, and the γ-ray irradiated from the anode. Figure 6a illustrates the spectra under applied voltage (−50 V) with a shaping time of 3 μs. Before radia CsPbBr3 detector showed a well-resolved spectroscopic response with an energy tion of 21.76% for 59.5 keV 241 Am γ-ray. Then the energy resolution degraded to 34.2% after γ-ray irradiation with a total dose of 100 krad(Si). At the same time, t tral performance disappeared when the total dose increased to 200 krad(Si). Afte diation device had been annealed at room temperature for one week, the γ-ray response was restored with an energy resolution of 25.35%. During the annealing pins of the device were in floating bias and were placed at room temperature. F shows that when the applied bias was -100 V, the TID effects on the spectral perfo of the CsPbBr3 detector. Before irradiation, the CsPbBr3 detector showed a more solved spectroscopic response with an energy resolution of 13.89%. After a cum total dose of 100 krad(Si), the energy resolution was 20.49%. When the total dose 200 krad, the spectral performance disappeared. After annealing, the device's mance under this voltage bias has not been recovered.
The test result with -150 V bias voltage is demonstrated in Figure 6c. The the bias voltage applied the better spectral performance was achieved. Howe der higher bias voltage, the influence of irradiation becomes more signific shown in Figure 6c, the energy resolution was 12.05% before irradiation and in

γ-ray Spectral Performance
The γ-ray spectral performance of the CsPbBr 3 detector was also influenced by total ionizing dose radiation. The energy resolution is a critical characteristic for evaluating the ability of energy spectral measurement of the detectors. The detector's energy resolution is usually defined as the ratio between the FWHM and the position of the peak centroid, which is a dimensionless fraction expressed as a percentage. Figure 6 shows the energy resolution of CsPbBr 3 detectors to 241 Am γ-ray with various bias voltages (−50 V, −100 V, and −150 V) before and after 60 Co γ ray total dose irradiation. During the energy spectral test, the radiation-induced charge Q collection was on the Au cathode, while Ga/In anode was grounded, and the γ-ray irradiated from the anode. Figure 6a illustrates the energy spectra under applied voltage (−50 V) with a shaping time of 3 µs. Before radiation, the CsPbBr 3 detector showed a well-resolved spectroscopic response with an energy resolution of 21.76% for 59.5 keV 241 Am γ-ray. Then the energy resolution degraded to around 34.2% after γ-ray irradiation with a total dose of 100 krad(Si). At the same time, the spectral performance disappeared when the total dose increased to 200 krad(Si). After the radiation device had been annealed at room temperature for one week, the γ-ray spectral response was restored with an energy resolution of 25.35%. During the annealing test, the pins of the device were in floating bias and were placed at room temperature. Figure 6b shows that when the applied bias was −100 V, the TID effects on the spectral performance of the CsPbBr 3 detector. Before irradiation, the CsPbBr 3 detector showed a more well-resolved spectroscopic response with an energy resolution of 13.89%. After a cumulative total dose of 100 krad(Si), the energy resolution was 20.49%. When the total dose reached 200 krad, the spectral performance disappeared. After annealing, the device's performance under this voltage bias has not been recovered.
particle energy of Pu (5156 keV) and γ-ray energy of Am (59.59 keV), respectively. The EdSi and EdCsPbBr3 represent the mean ionization energy for CsPbBr3 (5.3 eV) and Si (3.6 eV) [26], respectively. The CCESi is the charge collection efficiency of the reference silicon detector (~100%). The CCE at 150 V bias of the pre-irradiated CsPbBr3 detector was 91.75% and decreased to 84.56% with a total dose of 100 krad (Si). The degradation of energy resolution and CCE can be ascribed to increased trapping centers under gamma radiation, which decrease the mean free path of holes and electrons.

Photoluminescence
PL spectroscopy was used to characterize the optical properties of the perovskite CsPbBr3 crystal. Figure 7 shows the PL spectrum of as-grown and 60 Co γ-ray irradiated samples (the total dose = 165 rad(Si)), and the PL intensities at various temperatures were analyzed in detail. It is evident that the peak value of the PL spectrum has a red shift in The test result with −150 V bias voltage is demonstrated in Figure 6c. The higher the bias voltage applied the better spectral performance was achieved. However, under higher bias voltage, the influence of irradiation becomes more significant. As shown in Figure 6c, the energy resolution was 12.05% before irradiation and increased to 17.62% after irradiation with a total dose of 100 krad(Si). Similar to the test results under −100 V bias, there was no energy spectrum characteristic when the cumulative total dose was 200 krad(Si) and the test after annealing.
The charge collection efficiency (CCE) is also influenced by the TID effect, as shown in Figure 6d. The CCE of CsPbBr 3 detectors can be calculated by using a Si detector as a reference sample, and CCE can be expressed as in Formula (2) [26][27][28].
The P Si and P CsPbBr3 represent the channel numbers of the peak centroid measured by the CsPbBr 3 detector and Si (697.52) detector, respectively. The G Si and G CsPbBr3 represent the gain for CsPbBr 3 (50) and Si (1500), respectively. The E α and E γ and present the αparticle energy of 239 Pu (5156 keV) and γ-ray energy of 241 Am (59.59 keV), respectively. The Ed Si and Ed CsPbBr3 represent the mean ionization energy for CsPbBr 3 (5.3 eV) and Si (3.6 eV) [26], respectively. The CCE Si is the charge collection efficiency of the reference silicon detector (~100%). The CCE at 150 V bias of the pre-irradiated CsPbBr 3 detector was 91.75% and decreased to 84.56% with a total dose of 100 krad (Si). The degradation of energy resolution and CCE can be ascribed to increased trapping centers under gamma radiation, which decrease the mean free path of holes and electrons.

Photoluminescence
PL spectroscopy was used to characterize the optical properties of the perovskite CsPbBr 3 crystal. Figure 7 shows the PL spectrum of as-grown and 60 Co γ-ray irradiated samples (the total dose = 165 rad(Si)), and the PL intensities at various temperatures were analyzed in detail. It is evident that the peak value of the PL spectrum has a red shift in the 80 K temperature test, and with the increase of test temperature, the peak value of the PL spectrum after irradiation appeared to have a blue shift. The possible reason for this phenomenon is the total ionizing dose irradiation changes the potential energy at the top of the valence band and at the bottom of the guide band of the halide material [26]. The origin of this difference needs to be further investigated.
Sensors 2023, 23,2017 the 80 K temperature test, and with the increase of test temperature, the peak valu PL spectrum after irradiation appeared to have a blue shift. The possible reason phenomenon is the total ionizing dose irradiation changes the potential energy a of the valence band and at the bottom of the guide band of the halide material [ origin of this difference needs to be further investigated.

Discussion
This work investigated the 60 Co γ ionization radiation effects on the Schot CsPbBr3 detector at the dose rate of 50.0 rad(Si)/s. The ionization radiation has lit on the I-V curve of the CsPbBr3 detector. With the increase in total dose, the dark decreased slightly. The I-T character was impacted significantly after ionization ra and the pre-irradiated CsPbBr3 detector demonstrated a high ON-OFF ratio of 8 200 krad(Si) radiation, the ON-OFF ratio of the CsPbBr3 detector was reduced t 100 krad(Si), the energy spectral characteristics of the device were still maintaine ever, when the total dose accumulated to 200 krad(Si), the spectral performanc tested CsPbBr3 detector disappeared.
As shown in Table 1, with various bias voltages (-50, -100, and -150 V) durin ray spectral test, there were different responses of energy resolution to irradiati R1 − R2, where R1 represents the resolution of post-irradiation and R2 represents olution of pre-irradiation. When tested under extensive voltage conditions, the resolution caused by radiation after accumulating 100 krad(Si) was deduced at ab Additionally, the performance did not recover within one week of room tempera nealing. Then, under low-voltage test conditions (50 V), the spectral resolution increased by 1.1% after accumulating a total dose of 100 krad(Si). In addition, t

Discussion
This work investigated the 60 Co γ ionization radiation effects on the Schottky type CsPbBr3 detector at the dose rate of 50.0 rad(Si)/s. The ionization radiation has little effect on the I-V curve of the CsPbBr 3 detector. With the increase in total dose, the dark current decreased slightly. The I-T character was impacted significantly after ionization radiation, and the pre-irradiated CsPbBr 3 detector demonstrated a high ON-OFF ratio of 80. After 200 krad(Si) radiation, the ON-OFF ratio of the CsPbBr 3 detector was reduced to 35. At 100 krad(Si), the energy spectral characteristics of the device were still maintained. However, when the total dose accumulated to 200 krad(Si), the spectral performance of the tested CsPbBr 3 detector disappeared.
As shown in Table 1, with various bias voltages (−50, −100, and −150 V) during the γ-ray spectral test, there were different responses of energy resolution to irradiation ∆R = R1 − R2, where R1 represents the resolution of post-irradiation and R2 represents the resolution of pre-irradiation. When tested under extensive voltage conditions, the energy resolution caused by radiation after accumulating 100 krad(Si) was deduced at about 6%. Additionally, the performance did not recover within one week of room temperature annealing. Then, under low-voltage test conditions (50 V), the spectral resolution slightly increased by 1.1% after accumulating a total dose of 100 krad(Si). In addition, the main difference was that when tested under low voltage conditions, the energy spectral performance of the CsPbBr 3 detector recovered within one week of annealing after irradiation. Additionally, the energy resolution after annealing degraded by about 4.77% compared to that obtained before irradiation. However, this recovery process was not observed when testing at high voltage. Finally, when the total dose was 200 krad(Si), the detector had no spectral characteristics tested under all bias conditions.

Conclusions
The 60 Co γ ionizing radiation may induce electrically active defects in the bulk of the metal-semiconductor interface of the CsPbBr 3 detector. Those defects will likely alter the detector's performance. The difference in energy spectrum under different test bias voltage may also be due to the defect induced by ionizing radiation in the Schottky interface region. With the increase of the bias applied during the testing process, the Schottky barrier becomes wider, and the space charge region increases, leading to the decrease of carrier lifetime during the energy spectrum testing. Therefore, our future work will continue to study the radiation effects and damage mechanism of the CsPbBr 3 detectors to meet the increasing demands for radiation detection in intense radiation fields.