High-sensitivity X-ray detectors based on solution-grown caesium lead bromide single crystals

Article abstract of DOI:10.1039/c9tc05490a

X-ray detection is of great significance for medical diagnostics, industrial non-destructive inspection, nuclear plants and scientific research. However, high sensitivity is needed for X-ray detectors to reduce the radiation dose applied to human bodies for the applications of medical imaging and security checks. Here, we reported sensitive X-ray detectors made of solution-grown inorganic lead perovskite CsPbBr₃ single crystals and the synthesis of high-quality inorganic lead perovskite CsPbBr₃ single crystals using an improved low temperature solution method, which exhibited high transmittance, mobility and mobility-lifetime products. By designing the detector in an asymmetric electrode configuration, ion migration was effectively suppressed under a high voltage with a low dark current and an outstanding photoresponse. The optimized detector exhibited high sensitivity of 1256 μC Gy^-1 cm^-2 for 80 kVp X-ray detection under a relatively low electric field of 20 V mm^-1, which was 60 times higher than that of commercially used α-Se detectors. Due to its facile synthesis, low cost, long-term stability, and high detection sensitivity, CsPbBr₃ is expected to be an outstanding candidate for commercialized sensitive X-ray detectors.

Full text of DOI:10.1039/c9tc05490a

Journal of
Materials Chemistry C
View Article Online
View Journal
PAPER
High-sensitivity X-ray detectors based on
solution-grown caesium lead bromide single crystals†
Cite this:DOI: 10.1039/c9tc05490a
ab
a
a
a
Hongjian Zhang,
Fangbao Wang,
Yufei Lu,b Qihao Sun,
Yadong Xu,
*
Bin-Bin Zhang, *a Wanqi Jiea and Mercouri G. Kanatzidis
c
X-ray detection is of great significance for medical diagnostics, industrial non-destructive inspection,
nuclear plants and scientific research. However, high sensitivity is needed for X-ray detectors to reduce
the radiation dose applied to human bodies for the applications of medical imaging and security checks.
Here, we reported sensitive X-ray detectors made of solution-grown inorganic lead perovskite CsPbBr3
single crystals and the synthesis of high-quality inorganic lead perovskite CsPbBr3 single crystals using
an improved low temperature solution method, which exhibited high transmittance, mobility and mobility–
lifetime products. By designing the detector in an asymmetric electrode configuration, ion migration was
effectively suppressed under a high voltage with a low dark current and an outstanding photoresponse. The
optimized detector exhibited high sensitivity of 1256 mC Gyꢀ1 cmꢀ2 for 80 kVp X-ray detection under a
relatively low electric field of 20 V mmꢀ1, which was 60 times higher than that of commercially used a-Se
detectors. Due to its facile synthesis, low cost, long-term stability, and high detection sensitivity, CsPbBr3 is
expected to be an outstanding candidate for commercialized sensitive X-ray detectors.
Received 8th October 2019,
Accepted 6th November 2019
DOI: 10.1039/c9tc05490a
rsc.li/materials-c
complicated technology processes must be taken into consid-
eration for their wide applications.13
1. Introduction
X-ray detection is of utmost importance for wide applications in
Recently, metal halide perovskites with the general formula
the fields of medical diagnostics, industrial non-destructive of ABX3 (A = CH3CH3, NH2CHQNH2, and Cs; B = Pb, Bi, and Sn;
inspection, nuclear power plants and scientific research.1–3 and X = I, Br, and Cl) have been widely studied because of their
In all the X-ray detection semiconductors, high sensitivity and remarkable performances in solar cells,14–16 lasing,17,18 and
detection efficiency are paramount properties to reduce the LEDs17,19 and their facile synthesis methods. Due to their
X-ray dose that patients and workers are exposed to during high-Z elements such as Cs, Pb, Bi, I and Br and high mobi-
X-ray inspections such as computed tomography.4–6 Thus, lity–lifetime (mt) products, these perovskites are considered as
enhancing the sensitivity of X-ray detectors has important potential materials for a new generation of nuclear radiation
significance to reduce the risks of cancer and simultaneously detectors. To date, MAPbI3 thin films,20 MAPbI3 sintered
obtain high-resolution images. For semiconductor-based X-ray wafers,21 MAPbBr3 single crystals (SCs),22,23 and Cs2AgBiBr6
detectors, the current state-of-the-art materials used as X-ray SCs24 have been demonstrated to be successfully applied in
detectors include amorphous Se (a-Se),7 PbI2,8 HgI2,9,10 CdTe11 X-ray detection. However, their issues such as high density of
and CdZnTe.12 The application of commercialized a-Se is defects in their crystals, low sensitivity, ion migration under
limited because of its low absorption ability for higher energy high electric field, and poor stability are still severe enough to
X-rays over 50 keV. For PbI2 and HgI2, stability issues hinder hinder their further applications.
their further applications in X-ray detection. CdTe and CdZnTe
Accordingly, several efforts have been made to try to tackle
seem to have promising prospects; however, their high cost and these problems. For example, thermal annealing and surface
treatment have been proven to be effective measures to improve
the surface state of the crystals and carrier transportation.24,25
a State Key Laboratory of Solidification Processing, Key Laboratory of Radiation
In addition, special device designs that use asymmetric
electrodes26,27 or employ metal oxides such as aluminum zinc
oxide as electrodes25 have also been demonstrated to signifi-
cantly decrease the leakage current of the device by increasing
the Schottky barrier or forming p–n junction structures. As a
stable perovskite material without organic molecules,28–32
cesium lead bromide (CsPbBr3) has received much attention.
Detection Materials and Devices, & School of Materials Science and Engineering,
Northwestern Polytechnical University, Xi’an 710072, China.
E-mail: xyd220@nwpu.edu.cn, zbb@nwpu.edu.cn
b Institute of Flexible Electronics (IFE), Northwestern Polytechnical University (NPU),
Xi’an 710072, China
c Department of Chemistry, Northwestern University, Evanston, IL 60208, USA
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c9tc05490a
This journal is ©The Royal Society of Chemistry 2019
J. Mater. Chem. C

View Article Online
Paper
Journal of Materials Chemistry C
CsPbBr3 possesses an effective atomic number, Zeff, of 65.9, CuKa1 in the 2y range of 10–701. UV-Vis spectra of the CsPbBr3
with high attenuation coefficient for absorbing X-rays. CsPbBr3 crystals were measured on a UV-2550 spectrometer with an
microcrystals33 and quasi-monocrystalline34 thick films have integrating sphere over the spectral range of 300–1500 nm.
been reported recently as X-ray detectors, but controlling the TRIAX 550 fluorescence spectroscopy was used to collect photo-
film quality was challenging and the density of defects in the luminescence (PL) time-resolved photoluminescence (TRPL)
film was high. The optoelectronic properties related to life- spectra, with the excitation wavelength of 325 nm. Current–
times, carrier mobilities and defect tolerance of single crystals voltage (I–V) was measured using an electrical properties
show better performances than those of polycrystalline measurement system (Agilent 4155C).
materials.35,36 However, the fabrication of X-ray detectors using
2.4. Radiation detection measurements
CsPbBr3 SCs has not been reported so far.
In this work, we demonstrate the application of solution-
grown CsPbBr3 SCs to fabricate X-ray detectors with high
sensitivity. CsPbBr3 SCs were grown via an improved inverse
temperature crystallization (ITC) method with high crystal-
lization quality. By designing the device configuration into
asymmetric electrodes, the device degradation caused by ion
migration was effectively prevented under relatively high bias
voltages. Additionally, X-ray detectors with high sensitivities
were fabricated using the improved device structure.
For a-particle-induced pulse shape rise time analysis, the device
was exposed to the illumination of an uncollimated 241Am @
5.49 MeV a-particle source. The pulse shape information was
simultaneously recorded by connecting the pre-amplifier
(Model ORTEC570) and output to a high-speed waveform
digitizer card. For X-ray detection measurements, a tungsten
target X-ray tube (SPELLMAN XRB011) was used as the source.
The X-ray source was constantly operated with the tube voltage
from 40 kV to 80 kV, and the tube current was tuned from 0 to
600 mA. The X-ray dose rate was directly proportional to the tube
current, and it was carefully calibrated using a thermolumines-
cence meter. The CsPbBr3 device was placed in a dark shielding
box to prevent disturbance from ambient light. The bias voltage
was applied on the device and the induced photocurrent was
recorded by a Keithley 6517B.
2. Materials and methods
2.1. Materials
CsBr (Z99.5%), PbBr2 (Z99.5%), dimethyl sulfoxide (DMSO,
Z99.8%), dimethylformamide (DMF, 499.8%) and cyclohexanol
(CyOH, Z99.8%) were purchased from Sigma-Aldrich.
3. Results and discussion
2.2. Growth of CsPbBr3 single crystals and device fabrication
3.1. Crystal growth and characterization
PbBr2 (1 M) and CsBr (0.5 M) were dissolved in 60 mL DMSO
solution and fully stirred to obtain a 2 molar ratio of PbBr2 to CsBr.
Then 60 mL of a mixture of CyOH/DMF was added to the DMSO
solution to obtain the precursor solution. To achieve uniform
heating, a 1 ꢁ 1 ꢁ 1 m3 water bath tank was adopted as the
heating system. The water bath tank was equipped with a thermo-
meter, thermocouple, and submersible pump. During the heating
process, precise temperature compensation could be accurately
controlled based on the proportion, integral, derivative (PID)
adjustment. The precursor solution was put into the heating bath
and maintained at the temperature of 50 1C for 12 h. After heating
at a constant temperature, the solution was filtered into a new
container using 30 mm sized filter paper to remove impurities and
small CsPbBr3 crystals. After oversaturation at 50 1C, the was
heated solution at a heating rate of 1 1C per day starting from
49.5 1C. After about 10 days growth, several square-shaped CsPbBr3
crystals with a size of 4–5 mm were formed on the bottom of the
container. The as-grown crystals acted as seed crystals and con-
tinued to grow into large-sized CsPbBr3 SCs. Detector I was
prepared with a 3 ꢁ 4 ꢁ 1 mm3 CsPbBr3 SC with gold electrodes
evaporated on both sides. Detector II was prepared with a 5 ꢁ 5 ꢁ
2 mm3 CsPbBr3 SC with aluminium and gold evaporated on each
side separately. The area of each metal electrode was 3.14 mm2.
High-quality single crystals are needed for the fabrication of
high-sensitivity X-ray detectors. Among the solution growth
methods, the AVC method has been prevalent because of its
simplicity. However, the repeatability and growth speed control
of this method are not satisfactory. Thus, we employed the ITC
method to grow CsPbBr3 single crystals, and their growing
speed could be accurately controlled by adjusting the heating
rate. In our previous work,37 we systematically studied the AVC
method and found that the mole ratio of PbBr2 and CsBr (n : 1)
is the key factor to obtain a single CsPbBr3 phase and high-
quality crystals. Likewise, to make crystals grow in a Br-rich
solution and ensure no other phases precipitate, we set the
initial mole ratio of 2 : 1 in the precursor. Previous research31
showed that for the ratio of 2 : 1, the solubility of CsPbBr3
declines significantly in the range of 75–90 1C. The range of
temperature for the solubility drop can be widened to 25–110 1C
with the addition of a CyOH/DMF mixture to the precursor
solution. By using a water bath heating system (ESI,† Fig. S1),
CsPbBr3 precipitated in a specific temperature range with its
solubility decreased. The heating rate should be as low as
possible to ensure good crystallization quality. Fig. 1a shows
the single crystals grown by the ITC method in the temperature
range of 50–60 1C at the heating rate of 1 1C per day. CsPbBr3
suffers from two phase transitions, which occur at around
%
2.3. Characterization measurements
X-ray diffraction (XRD) patterns of both the powder and single 130 1C from cubic (Pm3m) to tetragonal (P4/mbm) and 88 1C
crystal samples were collected using a D/Max2500PC with from tetragonal (P4/mbm) to orthorhombic (Pbnm) phase.28
J. Mater. Chem. C
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Journal of Materials Chemistry C
Paper
Fig. 1 (a) As-grown CsPbBr3 crystals with orange color grown by the ITC method in the temperature range of 50–60 1C. (b) Schematic of the crystal
structure with the space group Pnma. (c) XRD patterns of the powder and single-crystal samples. (d) Optical transmittance spectra for as-grown CsPbBr3
crystals. The insert shows the energy gap is fitted as 2.297 eV by the Tauc plot. (e) Photoluminescence (PL) spectra of CsPbBr3 crystals at 300 K. (f) PL
lifetime decay upon two-photon excitation. The dots represent experimental data and the solid curve is the best fit with a double-exponential function.
However, in this growth process (50–60 1C), phase transi- application as a photodetector.28,31 The CsPbBr3 single crystal
tions were successfully avoided, which reduced the defects exhibited strong PL at room temperature with 325 nm excita-
originating from crystal lattice transformations. Millimeter-sized, tion. Two PL peaks located at around 530 nm and 560 nm
high-transparent CsPbBr3 single crystals were obtained within with strong intensities were observed in the PL spectrum,
10 days. The centimeter-sized single crystals were grown with as is shown in Fig. 1e, implying that near-edge defect levels
seed crystals.
originated by surface states play an important role in the
A schematic of the CsPbBr3 crystal structure at room tem- emission process.39 Fig. 1f displays the TRPL decay, which
perature is shown in Fig. 1b, which shows the perovskite crystal was fitted well to a double-exponential function. According
structure known as ABX3. Fig. 1c shows the powder and single to the spectrum, the faster decay time (t1 = 2.9 ns) and
crystal XRD patterns. The results of the XRD pattern refinement slower decay time (t2 = 25.1 ns) are speculated to be trap-
demonstrate that the resulting CsPbBr3 has the space group of assisted surface recombination and free charge carrier recom-
Pnma, with cell parameters a = 8.37 Å, b = 12.011 Å, c = 8.425 Å, a = bination in bulk, respectively. Both decay times are longer
b = g = 901. It is notable that no other phase precipitated during the than that of CsPbBr3 grown via the AVC method (t1 = 1.9 ns
growth process. The single crystal pattern shows that the as-grown and t2 = 6.8 ns).40
CsPbBr3 single crystals exhibit a set of (0 l 0) diffraction surfaces.
The optical properties of CsPbBr3 grown via the ITC method were
3.2. a-Particle-induced pulse shape rise time analysis
studied at room temperature. Fig. 1d shows the UV-Vis transmis- The time-of-flight (TOF) technique41 was adopted to evaluate
sion spectrum from 300 nm to 1500 nm of CsPbBr3. A sharp rise at the charge carrier mobility of the CsPbBr3 single crystal. The
around 550 nm is observed, which demonstrates the low density of schematic diagram of the a-particle (He2+)-induced charge carrier
the shallow energy level defects and high purity of the crystal. transport process in the CsPbBr3 crystal is shown in ESI,† Fig. S2,
From 550 nm to 1500 nm, a relatively high transmittance of B75% and the preamplifier pulse signals are analyzed. Under an external
was obtained, which is higher than that of the crystals (B57%) electric field, electrons and holes drift toward the anode and
grown via the improved AVC method37 in our previous work, which cathode immediately. For electron transport mobility calculation,
indicates higher crystalline quality. The inset in Fig. 1d shows the a-particles act on the cathode. Fig. 2a displays the paths and
Tauc plot fitting, and the band gap of the as-grown CsPbBr3 was stopping range of a-particles with an energy of 5.5 MeV in the
fitted as B2.297 eV, which is in agreement with the previous CsPbBr3 single crystals by Monte-Carlo calculations using the
crystals grown by solution methods and melting methods.37,38 SRIM software, and Fig. 2b shows the depth distributions of
The band gap value shows that CsPbBr3 is suitable for the a-particles. Due to the shallow penetration depth of the
This journal is ©The Royal Society of Chemistry 2019
J. Mater. Chem. C

View Article Online
Paper
Journal of Materials Chemistry C
Fig. 2 (a) Stopping and range of a-particles with energy of 5.5 MeV in CsPbBr3 single crystals by Monte-Carlo simulation. (b) Depth distributions of
a-particles in CsPbBr3. (c) Time-of-flight transients of electrons using a-particles. (d) Electric field-dependent electron drifting rates from 5 to 50 V cmꢀ1
.
The resulting electron mobility is 1368 ꢃ 52 cm2
V
ꢀ1 sꢀ1. (e) Time-of-flight transients of holes using a-particles. (f) Electric field-dependent hole drifting
rates from 5 to 50 V cmꢀ1. The resulting hole mobility is 2652 ꢃ 158 cm2 Vꢀ1 sꢀ1
.
a-particles of about 29 mm obtained from the simulation, the
majority of electron–hole pairs were generated and concen-
3.3. Optimization of device configuration
trated near the cathode region. Therefore, the resulting charge A sandwich electrode configuration was adopted to fabricate
signal is predominantly attributed to the drift of a single-carrier CsPbBr3-based detectors. CsPbBr3 was demonstrated to be a
polarity.41 The induced transit current signal is converted to p-type semiconductor in previous reports,38 and we considered
pulsed voltage spectra through charge sensitive preamplifier. its valence band maximum and the conduction band minimum
We collected the induced pulse voltage output from the pre- to be ꢀ5.6 and ꢀ3.3 eV, respectively. Therefore, Au (fAu = 5.1 eV)
amplifier using a high-speed waveform digitizer card. Fig. 2c is assumed to be a typical high work function electrode metal for
displays the typical voltage pulse spectrum induced by CsPbBr3 to inject holes efficiently. Fig. 3a displays a schematic
a-particles at the bias of 4, 6, 8, and 10 V for the electron diagram of the Au/CsPbBr3/Au sandwich configuration device and
transport. The 10–90% amplitude rise time of the spectrum was Au-CsPbBr3 contact energy band. j1 is the Schottky barrier for
extracted as the carrier drift time, t. The carrier drift velocity hole injection from Au to CsPbBr3 under a positive bias. VD is the
can be calculated as V = d/t, where d is the thickness of Schottky barrier for hole extraction from CsPbBr3 to Au under a
the CsPbBr3 single crystal. Combined with the equation reverse bias. As shown in Fig. 3b, this type of device has a low dark
V = mꢂE, where m is the mobility and E is the electric field, by leakage current of 0.4 nA under a 10 V bias, and the resistivity
fitting the plot of V–E, as shown in Fig. 2d, the electron mobility reached 7.85 ꢁ 109 O cm, which is quite a good performance for a
was calculated as 1368 ꢃ 52 cm2 Vꢀ1 sꢀ1. The charge transport radiation detector. However, when a high voltage was applied, the
process with holes is similar to electrons when a-particles leakage current increased dramatically with 2 mA under 100 V, and
act on the anode. As results show in Fig. 2e and f, the hole the typical I–V curve is presented in Fig. 3b. This issue is a vital
mobility was calculated as 2652 ꢃ 158 cm2 Vꢀ1 sꢀ1. The carrier challenge that exists in lead halide perovskite materials, which is
mobility obtained in our work is lower than the reported mainly attributed to vacancy defect activation and ion migra-
B4500 cm2 Vꢀ1 sꢀ1 for CsPbBr3 nanocrystals,42 but comparable tion under a high voltage, especially for halogen atoms.44–47
to that of single crystals (2290 cm2 Vꢀ1 sꢀ1 for electrons and The charge accumulation and built-in potential caused by ion
2060 cm2 Vꢀ1 sꢀ1 for holes).43 The carrier transport property in migration lead to a high leakage current and nonlinearity of the
the as-grown CsPbBr3 single crystals is excellent for high carrier I–V curve.
mobility, which guarantees a favorable performance in radia-
tion detection.
The photoresponse properties of the Au/CsPbBr3/Au device
under different bias and 420 nm LED light with a power of
J. Mater. Chem. C
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Journal of Materials Chemistry C
Paper
Fig. 3 (a) Au/CsPbBr3/Au device structure and the band diagram of Au-CsPbBr3 contact. (b) Typical I–V characteristic curves of the Au/CsPbBr3/Au
device at the range of ꢃ10 V and ꢃ100 V. (c) Photoresponse of Au/CsPbBr3/Au device under 1 V, 10 V, 50 V bias, illuminated using a LED with a
wavelength of 420 nm and a power of 1 mW. (d) Al/CsPbBr3/Au device structure and the band diagram of Al-CsPbBr3 contact. (e) Typical I–V
characteristic curves of the Al/CsPbBr3/Au device at the range of ꢃ10 V and ꢃ100 V. (f) Photoresponse of Al/CsPbBr3/Au device under 50 V bias using a
LED with a wavelength of 420 nm and a power of 1 mW.
1 mW were further investigated. It is notable that the base line suppresses ion migration significantly without increasing the
of the dark current drifted continuously with time under a contact resistance greatly, which ensures the device works
relatively high bias, and the on–off ratio also decreased due to under a high bias. Fig. 3f shows the photoresponse property
the increase in leakage current, as shown in Fig. 3c. Once the of the Al/CsPbBr3/Au device under 50 V bias and 420 nm LED
device suffered degradation caused by a high voltage, it was with 1 mW illumination. Compared with the former device,
irretrievable under a low bias. This property failure caused by the base line of the dark current was steady with time, and the
ion migration is a fatal shortcoming for its application in on–off ratio reached 102, demonstrating a good photoresponse
radiation detection since the dark current is expected to be as property. Further, the operation stability of the Al/CsPbBr3/Au
low as possible to get a high gain factor, and the applied electric device was demonstrated to be better than that of the Au/
field in the device is required to be higher to increase the CsPbBr3/Au device. The dark current under a constant 100 V
charge collection efficiency (CCE).
bias was steady within 150 s, which is greatly improved compared
Thus, to suppress ion migration and improve the perfor- with that of the Au/CsPbBr3/Au device, whose dark current
mance under a high bias, as displayed in Fig. 3d, asymmetric increased dramatically within 20 s, as shown in the ESI,†
electrodes were designed on the CsPbBr3 device by employing a Fig. S3. Therefore, long-time stability under a high voltage
lower work function metal, Al (fAl = 4.28 eV). Under the working was achieved in the optimized device.
conditions, the electrons were injected from the cathode to
3.4. X-ray detection performances of CsPbBr3 detectors
CsPbBr3, and holes injected from the anode to CsPbBr3.
Because CsPbBr3 is a p-type semiconductor, the lower work To investigate the X-ray detector performance of the typical
function of Al led to greater bending of the energy band of Au/CsPbBr3/Au device and the improved asymmetric electrodes
CsPbBr3, thus causing a higher Schottky barrier, f2, under a Al/CsPbBr3/Au device, their X-ray response and sensitivity were
positive bias for hole injection from Al to CsPbBr3. The increase measured. During the measurements, the devices were exposed
in the barrier is beneficial to suppress ion migration in the to a tungsten anode X-ray source, with the tube voltage ranging
device under a high voltage. Fig. 3e shows the I–V curves of the from 50 to 80 kV and tube current from 1 to 600 mA. To obtain
Al/CsPbBr3/Au device under a low and high bias, where both of correct sensitivity values, the radiation dose rates were carefully
them show good linearity with shape asymmetry. The dark calibrated using both an ion chamber dosimeter and thermo-
current of the new device under a low bias is the same order luminescence meters. We measured the dose rates of various
of magnitude as that of the Au/CsPbBr3/Au device, and surpris- tube currents under specific tube voltages and distances, r,
ingly, under a high voltage, the device also performed well with from the device to the X-ray source. By fitting the data, we
4 nA at 100 V. This asymmetric electrode configuration obtained a directly proportional relationship between dose rate
This journal is ©The Royal Society of Chemistry 2019
J. Mater. Chem. C

View Article Online
Paper
Journal of Materials Chemistry C
and tube current when the tube voltage and distance, r, were current. The detector sensitivity was defined as S = (ION ꢀ IOFF)/
fixed. The fitting results of the dose rate as a function of tube dose rate, which was calculated by fitting the data. As displayed
current under different tube voltages and distances, r, are in Fig. 4c, the sensitivity of detector I under different bias was
presented in the ESI,† Fig. S4. Furthermore, there is a good fitted, and the highest sensitivity was 61 mC Gyꢀ1 cmꢀ2 under
direct ratio relation between dose rate and 1/r2 according to the 8 V mmꢀ1 and 80 kVp X-rays, which is three times higher than
planar irradiation of the source (ESI,† Fig. S5). Therefore, the that of the commercially used a-Se detectors48 working at a
dose rate at any tube current and r can be calculated using much higher field of 10 V mmꢀ1 (20 mC Gyꢀ1 cmꢀ2) for 60 kVp
the calibrated results, and dose rate can be altered accurately by X-ray detection. Due to the higher CCE, the sensitivity increased
means of changing the tube current.
with an increase in the electric field. For detector I, the
A schematic illustration of the photoelectric process and sensitivity of different X-ray photoenergies was also investi-
photoconductive gain during the working state of detectors is gated under 1 V mmꢀ1, as demonstrated in the ESI,† Fig. S7. It
shown in Fig. 4a, and a picture of the CsPbBr3 single crystal showed that in the same range of dose rate, the detector has
X-ray detector is shown in Fig. 4d. As shown in the ESI,† Fig. S6, higher sensitivities for X-rays with a lower photoenergy, which
the stopping power of CsPbBr3 is similar to that of CdTe, and is caused by the longer hole drift length. By calculating the
stronger than MAPbBr3 and a-Se, due to its relatively high linear attenuation coefficient using XCOM of NIST,49 photons
average atomic number. Ignoring the very thin electrodes, the of 50 kVp and 80 kVp are adequately attenuated within the
CsPbBr3 SCs used in this work attenuated almost all the X-ray 0.2 mm and 0.7 mm CsPbBr3 samples, respectively. Due to the
radiation within the energy range of 50–80 eV. Thus, electron– thinner active layer of 50 kVp, the average hole drift length is
hole pairs were generated within the bulk SC when illuminated longer than that of 80 kVp, which contributes higher photo-
by X-rays. Under a bias, the electrons and holes separated and currents because of the higher hole mobility compared with
drifted toward the two electrodes to increase photoconductive gain. electrons (ESI,† Fig. S8). For the same photon energy of X-rays, the
Fig. 4b shows the response of the Au/CsPbBr3/Au device sensitivity and CCE were lower in the range of high dose rate,
(detector I) upon turning the X-ray source on and off at a field which is mainly due to the higher built-in electric field established
of 1 V mmꢀ1, and under continuously changed X-ray flux at by the generated charge carrier under high dose rate exposure
8 V mmꢀ1, which is the maximum bias for detector I. The (ESI,† Fig. S9). The CCE was calculated as CEE = IR/IP, where IR is
results demonstrate a sharp rise and fall edge and steady signal the induced photocurrent, and IP is theoretical photocurrent
Fig. 4 (a) Schematic illustration of the photoelectric process in the Au(Al)/CsPbBr3/Au devices under X-ray radiation. (b) X-ray photoresponse of the
Au/CsPbBr3/Au devices under different dose rates. (c) Dose rate-dependent photocurrent densities of Au/CsPbBr3/Au device under 2 V, 4 V, 6 V and 8 V.
The resulting slopes, which represent the X-ray response sensitivities, are 14 C Gyꢀ1 cmꢀ2, 29 C Gyꢀ1 cmꢀ2, 44 C Gyꢀ1 cmꢀ2 and 61 C Gyꢀ1 cmꢀ2
,
respectively. (d) CsPbBr3 single crystal X-ray detector. (e) X-ray responses of the Al/CsPbBr3/Au device under variable dose rates, which can bear a higher
electric field strength than the Au/CsPbBr3/Au device. (f) X-ray response sensitivities of the Al/CsPbBr3/Au device under 4 V, 10 V, 20 V and 40 V are
398 C Gyꢀ1 cmꢀ2, 622 C Gyꢀ1 cmꢀ2, 895 C Gyꢀ1 cmꢀ2 and 1256 C Gyꢀ1 cmꢀ2, respectively.
J. Mater. Chem. C
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Journal of Materials Chemistry C
Paper
Fig. 5 Comparison of the X-ray detection sensitivities obtain from current state-of-the-art detectors (CdZnTe, a-Se, PbI2 and HgI2), recently reported perovskite
detectors (MAPbBr3, MAPbI3 and Cs2AgBiBr6) and the CsPbBr3 detectors in this work. Data for CdZnTe, a-Se, PbI2 and HgI2 was extracted from ref. 10.
induced by X-ray ionization, which is defined as50 IP = jbe, where and the mobility–lifetime product (mt) was calculated as
e is the quantity of electric charge. The photon absorption rate j (2.5 ꢃ 0.2) ꢁ 10ꢀ3 cmꢀ2 Vꢀ1 under an X-ray dose rate of
can be defined as j = eDms/Eph, where e is the fraction of photons 7.1 m Gyꢀ1 sꢀ1. A comparison of the sensitivities obtained from
in the sample (100% for the samples in this study), D is the dose published data10,20–24 is shown in Fig. 5. It reveals that the
rate, ms is the mass of the sample, and Eph is the photon energy. b highest sensitivity obtained in detector II (1256 mC Gyꢀ1 cmꢀ2
)
is the number of carriers generated by per X-ray photon, which is is 60 times higher than that of the a-Se detector,48 15 times
given by b = Eph/D, where D can be calculated using the empirical higher than that of the Au/MAPbBr3/C60/BCP detector,22 and
model51 of D = 1.43 + 2Eg.
12 times higher than that of the Cs2AgBiBr6 detector.24 Thus,
Because of the lower dark current and higher applied bias, the CsPbBr3 is a potential material for X-ray detection with high
Al/CsPbBr3/Au device (detector II) is expected to have better sensitivity and a low working voltage.
performance. Fig. 4e shows the X-ray response of detector II
upon turning the X-ray source on and off at an electrical field
of 10 V mmꢀ1, and under continuously changed X-ray flux at
4. Conclusion
20 V mmꢀ1. Even under relatively high electric fields, the detector
exhibited good on–off ratios, and steady photocurrents. However, High quality CsPbBr3 SCs were gown via an improved low
the fall edge of the current of detector II is slower than that of temperature solution method, which exhibited high trans-
detector I with hysteresis of the photocurrent, which may mittance, mobility and carrier mobility–lifetime product.
be attributed to the slower recombination of photogenerated X-ray detectors were fabricated based on our CsPbBr3 SCs.
electron–hole pairs under a higher electric field strength and Modification of the device configuration (from Au/CsPbBr3/Au
the increase in defects in the crystal. Different from detector I, the to Al/CsPbBr3/Au) was demonstrated to be an effective method
photocurrent gain was much higher with the same increase in to suppress ion migration under high bias voltages and obtain
dose rate under a high electric field. The sensitivities of detector II long-term stability. The optimized detector could work under
increased significantly with the value of 398 mC Gyꢀ1 cmꢀ2
,
50 V with low dark currents of B3 nA and outstanding
622 mC Gyꢀ1 cmꢀ2, 895 mC Gyꢀ1 cmꢀ2 and 1256 mC Gyꢀ1 cmꢀ2 photoresponse on–off ratios of B102. Due to its high CCE at
for 4 V, 10 V, 20 V and 40 V, respectively (Fig. 4f). This improve- high bias, the sensitivity of the optimized detector reached
ment is mainly attributed to the high CCE of detector II. The ESI,† 1256 mC Gyꢀ1 cmꢀ2 under 40 V, which exceeds that of most
Fig. S9 shows the CCE values under various electric fields for state-of-the-art detectors and recently reported perovskite
detector II. Further, we fitted the data using the Hecht equation41 detectors at the same electric field. Thus, our work demon-
strates that CsPbBr3 is an excellent material for sensitive X-ray
ꢀ
ꢁ
ꢀd2
mtU
d2
detection owing to its facile synthesis and excellent detection
performance.
CCEðxÞ ¼
ꢂ
1 ꢀ emtU
(1)
This journal is ©The Royal Society of Chemistry 2019
J. Mater. Chem. C

View Article Online
Paper
Journal of Materials Chemistry C
A. A. Haghighirad and D. T. Moore, Adv. Energy Mater., 2016,
6, 1–6.
Conflicts of interest
There are no conflicts to declare.
17 Z.-K. Tan, R. S. Moghaddam, M. L. Lai, P. Docampo,
R. Higler, F. Deschler, M. Price, A. Sadhanala, L. M. Pazos
and D. Credgington, Nat. Nanotechnol., 2014, 9, 687–692.
18 H. Zhu, Y. Fu, F. Meng, X. Wu, Z. Gong, Q. Ding, M. V.
Gustafsson, M. T. Trinh, S. Jin and X. Zhu, Nat. Mater., 2015,
14, 636.
19 D. Clark, C. Stoumpos, F. Saouma, M. Kanatzidis and
J. Jang, Phys. Rev. B, 2016, 93, 195202.
20 S. Yakunin, M. Sytnyk, D. Kriegner, S. Shrestha, M. Richter,
G. J. Matt, H. Azimi, C. J. Brabec, J. Stangl and M. V.
Kovalenko, Nat. Photonics, 2015, 9, 444–449.
Acknowledgements
This work has been financially supported by National Natural
Science Foundations of China (U1631116 and 51802262),
National Key Research and Development Program of China
(2016YFE0115200), the Natural Science Foundation of Shaanxi
Province (2019JQ-459), and the Fundamental Research Funds for
the Central Universities (3102018jcc036 and 3102019TS0408). At
Northwestern University work on hard radiation detector materials
was supported by a grant from the Department of Homeland
Security ARI program (grant number 2014-DN-077-ARI086-01).
21 S. Shrestha, R. Fischer, G. J. Matt, P. Feldner, T. Michel,
A. Osvet, I. Levchuk, B. Merle, S. Golkar and H. Chen, Nat.
Photonics, 2017, 11, 436–440.
22 H. Wei, Y. Fang, P. Mulligan, W. Chuirazzi, H. Fang,
C. Wang, B. R. Ecker, Y. Gao, M. A. Loi and L. Cao, Nat.
Photonics, 2016, 10, 333–339.
23 W. Wei, Y. Zhang, Q. Xu, H. Wei, Y. Fang, Q. Wang, Y. Deng,
T. Li, A. Gruverman and L. Cao, Nat. Photonics, 2017, 11,
315–321.
Notes and references
1 A. Sakdinawat and D. Attwood, Nat. Photonics, 2010, 4,
840–848.
2 M. J. Yaffe and J. A. Rowlands, Phys. Med. Biol., 1997, 42,
1–39.
24 W. Pan, H. Wu, J. Luo, Z. Deng, C. Ge, C. Chen, X. Jiang,
W. J. Yin, G. Niu and L. Zhu, Nat. Photonics, 2017, 11, 726.
25 L. Li, X. Liu, H. Zhang, B. Zhang, W. Jie, P. J. Sellin, C. Hu,
G. Zeng and Y. Xu, ACS Appl. Mater. Interfaces, 2019, 11,
7522–7528.
3 S. O. Kasap, J. B. Frey, G. Belev, O. Tousignant, H. Mani,
J. Greenspan, L. Laperriere, O. Bubon, A. Reznik and
G. Decrescenzo, Sensors, 2011, 11, 5112–5157.
4 M. J. Yaffe and J. A. Rowlands, Phys. Med. Biol., 1997, 42,
1–39.
26 Y. He, L. Matei, H. J. Jung, K. M. McCall, M. Chen,
C. C. Stoumpos, Z. Liu, J. A. Peters, D. Y. Chung,
B. W. Wessels, M. R. Wasielewski, V. P. Dravid, A. Burger
and M. G. Kanatzidis, Nat. Commun., 2018, 9, 1609.
27 Q. Xu, W. Shao, Y. Li, X. Zhang, X. Ouyang, J. Liu, B. Liu,
Z. Wu, X. Ouyang, X. Tang and W. Jia, ACS Appl. Mater.
Interfaces, 2019, 11, 9679–9684.
28 C. C. Stoumpos, C. D. Malliakas, J. A. Peters, Z. Liu,
M. Sebastian, J. Im, T. C. Chasapis, A. C. Wibowo,
D. Y. Chung and A. J. Freeman, Cryst. Growth Des., 2013,
13, 2722–2727.
29 J. Song, Q. Cui, J. Li, J. Xu, Y. Wang, L. Xu, J. Xue, Y. Dong,
T. Tian and H. Sun, Adv. Opt. Mater., 2017, 5, 1700157.
30 M. I. Saidaminov, A. Haque, J. M. Almutlaq, S. P. Sarmah,
X. Miao, R. Begum, A. A. Zhumekenov, I. Dursun, N. Cho
and B. Murali, Adv. Opt. Mater., 2017, 5, 1600704.
31 D. N. Dirin, I. Cherniukh, S. Yakunin, Y. Shynkarenko and
M. V. Kovalenko, Chem. Mater., 2016, 28, 8470–8474.
5 S. O. Kasap and J. A. Rowlands, Proc. IEEE, 2002, 90,
591–604.
6 J. P. Moy, Nucl. Instrum. Methods Phys. Res., 2000, 442,
26–37.
7 S. O. Kasap, J. Phys. D: Appl. Phys., 2000, 33, 2853–2865.
8 R. A. Street, S. E. Ready, K. V. Schuylenbergh, J. Ho, J. B.
Boyce, P. Nylen, K. Shah, L. Melekhov and H. Hermon,
J. Appl. Phys., 2002, 91, 3345–3355.
9 G. Zentai, M. Schieber, L. Partain, R. Pavlyuchkova and
C. Proano, J. Cryst. Growth, 2005, 275, 1327–1331.
10 M. Schieber, H. Hermon, A. Zuck, A. I. Vilensky,
L. Melekhov, R. Shatunovsky, E. Meerson, Y. Saado,
M. Lukach and E. Pinkhasy, J. Cryst. Growth, 2001, 225,
118–123.
11 M. Richter and P. Siffert, Nucl. Instrum. Methods Phys. Res.,
Sect. A, 1992, 322, 529–537.
12 C. Szeles, Phys. Status Solidi B, 2004, 241, 783–790.
13 S. Shrestha, R. Fischer, G. J. Matt, P. Feldner, T. Michel,
A. Osvet, I. Levchuk, B. Merle, S. Golkar, H. Chen, S. F. 32 J. Ding, S. Du, Z. Zuo, Y. Zhao, H. Cui and X. Zhan, J. Phys.
¨
Tedde, O. Schmidt, R. Hock, M. Ru¨hrig, M. Goken,
Chem. C, 2017, 121, 4917–4923.
W. Heiss, G. Anton and C. J. Brabec, Nat. Photonics, 2017, 33 Z. Gou, S. Huanglong, W. Ke, H. Sun, H. Tian, X. Gao,
11, 436–440.
X. Zhu, D. Yang and P. Wangyang, Phys. Status Solidi RRL,
2019, 13, 1900094.
34 W. Pan, B. Yang, G. Niu, K. H. Xue, X. Du, L. Yin, M. Zhang,
H. Wu, X. S. Miao and J. Tang, Adv. Mater., 2019,
31, 1904405.
14 M. Kulbak, D. Cahen and G. Hodes, J. Phys. Chem. Lett.,
2015, 6, 2452–2456.
15 G. E. Eperon, G. M. Paterno, R. J. Sutton, A. Zampetti,
A. A. Haghighirad, F. Cacialli and H. J. Snaith, J. Mater.
Chem. A, 2015, 3, 19688–19695.
35 X. Ren, Z. Yang, D. Yang, X. Zhang, D. Cui, Y. Liu, Q. Wei,
H. Fan and S. F. Liu, Nanoscale, 2016, 8, 3816–3822.
16 R. J. Sutton, G. E. Eperon, L. Miranda, E. S. Parrott,
¨
B. A. Kamino, J. B. Patel, M. T. Horantner, M. B. Johnston, 36 G. E. Eperon and D. S. Ginger, Nat. Energy, 2016, 1, 16109.
J. Mater. Chem. C
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Journal of Materials Chemistry C
Paper
37 H. Zhang, X. Liu, J. Dong, H. Yu, C. Zhou, B. Zhang, Y. Xu 44 T. Shi, W. Yin, F. Hong, K. Zhu and Y. Yan, Appl. Phys. Lett.,
and W. Jie, Cryst. Growth Des., 2017, 17, 6426–6431. 2015, 106, 103902.
38 Y. He, L. Matei, H. J. Jung, K. M. Mccall, M. Chen, 45 C. Eames, J. M. Frost, P. R. F. Barnes, B. C. Oregan, A. Walsh
C. C. Stoumpos, Z. Liu, J. A. Peters, D. Y. Chung and
B. W. Wessels, Nat. Commun., 2018, 9, 1609.
39 D. J. As, F. Schmilgus, C. Wang, B. Schottker, D. Schikora
and K. Lischka, Appl. Phys. Lett., 1997, 70, 1311–1313.
and M. S. Islam, Nat. Commun., 2015, 6, 7497.
46 W. Ming, S. Chen and M. Du, J. Mater. Chem., 2016, 4,
16975–16981.
47 J. Kang and L. Wang, J. Phys. Chem. Lett., 2017, 8, 489–493.
40 J. Ding, S. Du, Z. Zuo, Y. Zhao, H. Cui and X. Zhan, J. Phys. 48 W. Zhao and J. A. Rowlands, Med. Phys., 1995, 22,
Chem. C, 2017, 121, 4917–4923. 1595–1604.
41 P. J. Sellin, A. W. Davies, S. Gkoumas, A. Lohstroh, M. E. 49 M. Berger, XCOM photon cross section database, NIST physics
Ozsan, J. Parkin, V. Perumal, G. Prekas and M. C. Veale, Nucl.
Instrum. Methods Phys. Res., Sect. B, 2008, 266, 1300–1306.
42 G. R. Yettapu, D. Talukdar, S. Sarkar, A. Swarnkar, A. Nag,
P. Ghosh and P. Mandal, Nano Lett., 2016, 16, 4838–4848.
laboratory, Gaithersburg, MA, 1998.
50 B. Fraboni, A. Ciavatti, F. Merlo, L. Pasquini, A. Cavallini,
A. Quaranta, A. Bonfiglio and A. Fraleonimorgera, Adv.
Mater., 2012, 24, 2289–2293.
43 J. Song, Q. Cui, J. Li, J. Xu, Y. Wang, L. Xu, J. Xue, Y. Dong, 51 R. Devanathan, L. R. Corrales, F. Gao and W. J. Weber, Nucl.
T. Tian and H. Sun, Adv. Opt. Mater., 2017, 5, 1700157.
Instrum. Methods Phys. Res., Sect. A, 2006, 565, 637–649.
This journal is ©The Royal Society of Chemistry 2019
J. Mater. Chem. C