⇦Chinese Physics Letters, 2019, Vol. 36, No. 11, Article code 110701 Direct ZnO X-Ray Detector with Tunable Sensitivity * Hui-Li Liang (梁会力)1**, Shu-Juan Cui (崔书娟)1,2, Wen-Xing Huo (霍文星)1,2, Tao Wang (王涛)1,2, Yong-Hui Zhang (张永晖)1,2, Bao-Gang Quan (全保刚)3, Xiao-Long Du (杜小龙)1,2,4, Zeng-Xia Mei (梅增霞)1** Affiliations 1Key Laboratory for Renewable Energy, Beijing Key Laboratory for New Energy Materials and Devices, Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100190 2School of Physical Sciences, University of Chinese Academy of Sciences, Beijing 100049 3Laboratory of Microfabrication, Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100190 4Songshan Lake Materials Laboratory, Dongguan 523808 Received 20 May 2019, online 21 October 2019 *Supported by the National Natural Science Foundation of China under Grant Nos 11675280, 11674405, 61874139 and 11875088.
**Corresponding author. Email: zxmei@iphy.ac.cn; hlliang@iphy.ac.cn Citation Text: Liang H L, Cui S J, Huo W X, Wang T and Zhang Y H et al 2019 Chin. Phys. Lett. 36 110701    Abstract Direct ZnO x-ray detectors with tunable sensitivity are realized by delicately controlling the oxygen flux during the sputtering deposition process. The photocurrents induced by x-rays from a 40 kV x-ray tube with a Cu anode increase apparently as the oxygen flux decreases, which is attributed to the introduction of $V_{\rm o}$ detects. By introducing $V_{\rm o}$ defects, the annihilation rate of the photo-generated electron-hole pairs will be greatly slowed down, leading to a remarkable photoconductive gain. This finding informs a novel way to design the x-ray detectors based on abundant oxide materials. DOI:10.1088/0256-307X/36/11/110701 PACS:07.85.Fv, 71.55.Gs, 29.40.Wk, 81.15.Cd © 2019 Chinese Physics Society Article Text X-ray detection has broad applications in medical imaging, cancer treatment, industrial non-destructive testing, public security inspection, x-ray space communication and so on. The energy of x-rays is quite higher than the common photons. For example, medical and space communication x-rays are usually in the range of a few of keV to tens of keV, meaning that they usually have low absorption rate and large penetration depth in most materials, and are quite difficult to detect. Many strategies have been explored to realize the detection of x-rays, including the Geiger–Müller tube, scintillometer, micro-channel plate, charge-coupled devices, silicon drift detectors and so on. Compared with these methods, semiconductor radiation detectors have been developed as a mainstream of x-ray detectors due to their high sensitivity, high signal/noise ratio, high energy resolution and high integration capability. So far, most of the commercialized semiconductor x-ray detectors are based on Si,[1] Se,[2] and CdZnTe.[3] However, a wide range of new materials—such as iodide and perovskite compounds,[4,5] and organic materials,[6]—are being exploited to further improve the x-ray detection performance. However, these new materials are still challenging, considering their stability and repeatability under harsh radiation environment. Wide band gap semiconductors, owning to their natural radiation hardness and low leakage current, have become another recent trend for x-ray detection. Great efforts have been devoted to investigate the performance of x-ray detectors using wide band gap materials, such as GaN,[7] SiC,[8] diamond,[9] $\beta$-Ga$_{2}$O$_{3}$ and amorphous Ga$_{2}$O$_{3}$ (a-Ga$_{2}$O$_{3}$),[10,11] and ZnO.[12–15] Among these materials, ZnO is very attractive due to its low temperature growth, easy crystallization and multi-functionality. Many groups have shown the potential of high resistant ZnO bulk crystals,[12,13] ZnO thin films[14,15] and ZnO nanowires[15] to work as an effective x-ray sensor. However, to the best of our knowledge, there are very few reports regarding the further improvement of the sensitivity of these devices by optimizing their material properties. In this work, we present x-ray detectors with tunable sensitivity using polycrystalline ZnO thin films deposited by radio frequency (rf) magnetron sputtering technique. Oxygen flux was delicately controlled during the sputtering process. Metal/semiconductor/metal (MSM) structured photodetectors with co-planar interdigital electrodes were fabricated. Irradiated by x-rays from an x-ray tube with a Cu anode biased at 40 kV, the current of the devices increased apparently as the oxygen partial pressure decreases, indicating a large photoconductive gain at the expense of a large response time. It should be noted that even though the thickness of ZnO films in our work is only 300 nm, while those in previous reports are at least 1 µm, the photocurrent of Z2 sample is still very competitive compared with recent result in Ref. [14]. This finding demonstrates that the sensitivity of ZnO x-ray detectors can be effectively tuned with a delicate control of the oxygen flux, establishing a technical foundation for further boost of radiation detectors using ZnO related materials. ZnO thin films with a thickness of 300 nm were deposited on quartz substrates under different oxygen partial pressures (Z1: $1.0\times 10^{-3}$ Pa, Z2: $7.0\times 10^{-4}$ Pa, Z3: 0) by rf magnetron sputtering system at room temperature. Our ZnO ceramic target is 4 N pure. The quartz substrates were ultra-sonically cleaned in acetone, alcohol and deionized water successively and blown dry by nitrogen gas before loading into the vacuum chamber with a base pressure of $3.0\times 10^{-4}$ Pa. Then, a high purity oxygen gas (5 N) was introduced into the sputtering chamber through a leakage valve, the amount of which was monitored by an ion gauge. After the chamber pressure was stabilized, the sputtering gas, argon (Ar), entered into the chamber through a mass flow controller (MFC) maintained at 10.0 sccm. The whole growth process lasted for 30 min with a sputtering power of 60 W and a total pressure of 0.4 Pa. The crystal structure of these ZnO thin films was investigated by x-ray diffraction (XRD) technique. MSM structured photodetectors with co-planar interdigital electrodes were fabricated on these thin films by conventional UV-lithography and lift-off technology. The prototype device has 75 pairs of fingers with 5 µm in width, 5 µm in spacing gap and 300 µm in length. The 100 nm indium tin oxide (ITO) was used as interdigital electrodes. The current-voltage ($I$–$V$) tests were carried out in dark and illuminated by UV light and x-ray, respectively, using a picoammeter (Keithley 6487). A hand-held lamp with a 365 nm output was applied as the UV light source. The x-ray source was an x-ray tube with a Cu anode in an x-ray diffraction setup (Smartlab, Rigaku), working at 40 kV and 200 mA. Figure 1 shows the XRD $\theta$–$2\theta$ scans of all of the samples. Six peaks can be clearly distinguished. They are attributed to the diffraction from wurtzite ZnO phase with different crystal planes, as labeled in the picture, indicating that all of the samples are polycrystalline no matter what the oxygen partial pressure is applied during the deposition process.
cpl-36-11-110701-fig1.png
Fig. 1. XRD $\theta$–$2\theta$ scans of all the samples. Note that the diffraction intensity is rescaled just for clarity.
cpl-36-11-110701-fig2.png
Fig. 2. (a) Microscope photograph of the MSM structured ZnO photodetector on the quartz substrate, (b) $I$–$V$ curves in darkness, (c) $I$–$V$ curves under UV 365 nm light illumination, and (d) time-dependent photoresponse of all the devices with the UV 365 nm light on and off at 5 V bias.
The MSM structured photodetectors were fabricated on the ZnO thin films. Fig. 2(a) shows the microscope photograph of the prototype device. Figure 2(b) presents the $I$–$V$ characteristics of the devices in dark. It can be found that the dark current increases gradually as the oxygen partial pressure decreases from Z1 to Z3. As reported before, $V_{\rm o}$ defects, which are responsible for the unintentional n-type conductivity, have been widely observed in oxide semiconductor materials, such as ZnO,[16] and In$_{2}$O$_{3}$.[17] Thus, as the oxygen partial pressure decreases, more $V_{\rm o}$ defects will be produced in the film, resulting in more carrier concentration and larger dark current. Figure 2(c) shows the $I$–$V$ curve under 365 nm UV light illumination. The photocurrent decreases about two orders of magnitude as oxygen partial pressure increases from 0 (Z3) to $1.0\times 10^{-3}$ Pa (Z1). It has been reported that the decreased n-type doping at the surface will promote the formation of the Schottky contact on the expense of Ohmic behavior.[18] This contact type conversion from Ohmic to Schottky has also been observed by Zhang et al. at Ti/a-Ga$_{2}$O$_{3}$ interface by tuning the conductivity of a-Ga$_{2}$O$_{3}$ thin films with a delicate control of the oxygen flux during deposition process.[19] Therefore, the decrease of the photocurrent can be ascribed to the formation of Schottky contact at ITO/ZnO interface. Figure 2(d) gives the time-dependent UV response of these devices, which demonstrates good repeatability to the periodical UV light illumination. It should be noted the recovery time (90% to 10%) increases from less than 0.11 s of Z1 to more than 15 s of Z3 accompanied by a large photocurrent increase, which is regarded as a consequence of $V_{\rm o}$ defects.[20]
cpl-36-11-110701-fig3.png
Fig. 3. (a) Time-dependent photocurrent of Z1–Z3 with x-ray source on and off at 5 V bias. (b) Photocurrent of Z2 as a function of the applied bias. The photocurrent value is taken from the 2nd cycle before x-ray is off. (c) Time-dependent photocurrent of Z2 with a 60 s-long x-ray illumination at 30 V bias. (d) X-ray-generated photocurrent at various dose rates.
We then measured their responses to x-ray illumination in air. Figure 3(a) shows the temporary response with x-rays on and off (controlled by an electro-mechanical shutter) for three cycles at 5 V bias. Similarly, as the UV response shown in Fig. 2(d), all devices demonstrate good repeatability. Importantly, the x-ray induced current also increases more than two orders of magnitude as $V_{\rm o}$ concentration increases from Z1 to Z3. Dependence of the photocurrent on the applied bias is demonstrated in Fig. 3(b). The measurement was proceeded with the maximum x-ray tube current. It can be observed that the photocurrent increases almost linearly with bias. In Fig. 3(c), we measured the photocurrent of Z2 under a monochromatic x-ray beam to evaluate its sensitivity. Since the intensity of the x-ray beam decreases a lot after monochromatization, the photocurrent also decreases from 144 nA to 44 nA when biased at 30 V with a tube current of 200 mA. Figure 3(d) exhibits the photocurrent at different dose rates. The photocurrent value is taken from the 2nd cycle before x-ray is off in Fig. 3(c). The detailed calculation of the dose rate at corresponding tube current can be found in the supporting information in Ref. [11]. From the linear fit of the experimental data, the slope of the line (i.e., the device sensitivity) is determined to be $9.19\times 10^{-3}$ µCmGy$_{\rm air}^{-1}$cm$^{-2}$. Considering the $\sim $300 nm thickness of the ZnO active layer, the normalized sensitivity with volume is 306 µCmGy$_{\rm air}^{-1}$cm$^{-3}$, which is slightly higher than the sensitivity of our previously reported a-Ga$_{2}$O$_{3}$ x-ray detector.
Table 1. A summary of the rf-sputtered ZnO x-ray detectors performance. The dark current and photocurrent values are taken from the 1st cycle before x-ray is on and off, respectively.
Sample Oxygen partial pressure (Pa) Dark current@5 V (A) X-ray photocurrent@5 V (A) Recovery time(80%–20%) (s)
Z1 $1.0\times 10^{-3}$ $1.5\times 10^{-10}$ $7.0\times 10^{-10}$ $\le 0.11$
Z2 $7.0\times 10^{-4}$ $2.0\times 10^{-9}$ $1.2\times 10^{-8}$ 43
Z3 0 $2.8\times 10^{-7}$ $7.7\times 10^{-7}$ 285
Table 1 lists all the performance parameters of the three devices, from which it can be clearly seen that both the dark current and x-ray photocurrent increase with a decreased oxygen partial pressure. More importantly, the recovery time also increases greatly from less than 0.11 s of Z1 to 285 s of Z3, demonstrating a typical persistent photoconductivity (PPC) effect. It is widely accepted that inherent PPC effect in oxide materials is related to $V_{\rm o}$ defects.[20] By introducing $V_{\rm o}$ defects, both the conductivity of the film and the Ohmic contact of the ITO/ZnO interface will be improved, which is beneficial for the enhancement of the dark current and photocurrent. However, it is unfavorable to realize a high signal/noise ratio at the expense of an increased dark current, which should be considered in future investigations. Moreover, it should be noted that the thickness of the ZnO film is only about 300 nm, meaning that only $\sim 1‰$ x-ray photons can be absorbed by such a thin active layer. Nevertheless, the Z2 device behaves as a decent x-ray detector. It may originate from the speculation that when illuminated by x-rays, energetic electrons will be produced and can impact atoms around them to transfer their excessive energy by creating new e-h pairs, which is like the cascade effect. In addition, the ionized $V_{\rm o}$ defects will suppress the recombination of the electrons and holes once they are excited by incident x-ray photons,[21] greatly slowing down the annihilation rate and leading to a remarkable photoconductive gain. All of these factors can compensate for the drawback of low absorption of x-rays in ZnO thin films. This observation is consistent with our previous work using a-Ga$_{2}$O$_{3}$ as an active layer for x-ray detection.[11]
Table 2. Comparison with the sensitivity of the various detectors presented in the references.
Material Thickness (µm) Voltage (V) Dark current Photocurrent X-ray dose rate Sensitivity in the reference
PbI$_{2}$[4] 500 50 30 pA $\sim$10 nA 140 R/min 45 nC$\cdot$mm$^{-2}$R$^{-1}$
CH$_{3}$NH$_{3}$PbI$_{3}$[5] 830 50 80 nA/cm$^{2}$ $\sim $5$\times10^{-3}$ mA/cm$^{2}$ 1 mGy/s 11 µC$\cdot$mGy$_{\rm air}^{-1}$cm$^{-2}$
Polymer(PFO)[6] 20 50 0.9 nA 4.5 nA 18.5 mGy/s 480 nC$\cdot$mGy$^{-1}$cm$^{-3}$
$\beta$-Ga$_{2}$O$_{3}$[10] 1000 $-$15 $\sim$1 µA/cm$^{2}$ $\sim $1$\times$10$^{-3}$ A/cm$^{2}$ 4.596 Gy/s
a-Ga$_{2}$O$_{3}$[11] 0.25 50 $\sim$0.2 nA $\sim$15 nA 650 mGy/s 271 µC$\cdot$mGy$_{\rm air}^{-1}$cm$^{-3}$
ZnO bulk[12] 500 50 $\sim$25 pA $\sim$30 nA 1.53 Gy/s 19.4 nC/Gy
ZnO bulk[13] 500 20 $\sim$50 pA $\sim$275 pA 307 µGy/s 1.5 µC/Gy
ZnO film[14] 3 40 $\sim$0.1 nA $\sim$30 nA 1.532 Gy/s 25 nC/Gy
ZnO film[15] 1.4 100 $\sim$0.1 mA $\sim$1.0 mA 10 mC/(kg$\cdot$s) 0.12 µC/Gy
ZnO NWs[15] 1 50 $\sim$1.5 mA $\sim$3.0 mA 10 mC/(kg$\cdot$s) 0.17 µC/Gy
ZnO (Z2)$^{\rm This work}$ 0.3 30 $\sim$10 nA 44 nA 650 mGy/s 306 µC$\cdot$mGy$_{\rm air}^{-1}$cm$^{-3}$
Finally, we further compared the sensitivity value of our device with various detectors presented in the references as listed in Table 2. From this comparison, it can be clearly found that both the dark current and sensitivity of the Z2 device are competitive to all of the other devices. In summary, we have presented rf-sputtered ZnO based x-ray detectors with tunable sensitivity. The x-ray induced photocurrent can be increased by more than two orders of magnitude as oxygen partial pressure decreases from $1.0\times 10^{-3}$ Pa to 0 Pa. The enhancement of the photocurrent is ascribed to the fact that the annihilation rate of the photon-generated carriers will be slowed down due to the introduction of $V_{\rm o}$ defects, indicating a large photoconductive gain at the expense of a long response time. This work reveals an easy method to tune the sensitivity of the x-ray detector based on oxide materials, opening a new avenue to design novel x-ray detection system. We are grateful for the technical support of the x-ray irradiation measurement from Dr. Lihong Yang, in the National Lab for Superconductivity, Institute of Physics, Chinese Academy of Sciences. References The first 25 years of silicon drift detectors: A personal viewAmorphous selenium and its alloys from early xeroradiography to high resolution X-ray image detectors and ultrasensitive imaging tubesRoom-temperature compound semiconductor radiation detectorsElectrical Behavior of X-Ray Detector Based on PbI 2 Crystal With Coplanar Electrode StructurePrintable organometallic perovskite enables large-area, low-dose X-ray imagingDirect x-ray detection with conjugated polymer devicesX-ray detectors based on GaN Schottky diodesSilicon carbide and its use as a radiation detector materialThin polycrystalline diamond for low-energy x-ray detectionSchottky x-ray detectors based on a bulk β-Ga 2 O 3 substrateFlexible X-ray Detectors Based on Amorphous Ga 2 O 3 Thin FilmsNanosecond X-ray detector based on high resistivity ZnO single crystal semiconductorFabrication and characterization of a ZnO X-ray sensor using a high-resistivity ZnO single crystal grown by the hydrothermal methodA High-Resistivity ZnO Film-Based Photoconductive X-Ray DetectorStudy of ZnO photoconductive X-ray detectorOxygen vacancies: The origin of n -type conductivity in ZnOSelf-diffusion measurements in In2O3 isotopic heterostructures: Oxygen vacancy energeticsZinc oxide: bulk growth, role of hydrogen and Schottky diodesTransition of photoconductive and photovoltaic operation modes in amorphous Ga 2 O 3 -based solar-blind detectors tuned by oxygen vacanciesAnion vacancies as a source of persistent photoconductivity in II-VI and chalcopyrite semiconductorsColossal X-Ray-Induced Persistent Photoconductivity in Current-Perpendicular-to-Plane Ferroelectric/Semiconductor Junctions
[1] Guazzoni C 2010 Nucl. Instrum. Methods Phys. Res. Sect. A 624 247
[2] Kasap S, Frey J B, Belev G, Tousignant O, Mani H, Laperriere L, Reznik A and Rowlands J A 2009 Phys. Status Solidi B 246 1794
[3] Sordo S D, Abbene L, Caroli E, Mancini A M, Zappettini A and Ubertini P 2009 Sensors 9 3491
[4] Sun H, Zhu X H, Yang D Y, Wangyang P H, Tian H B and Gao X Y 2016 IEEE Trans. Nucl. Sci. 63 1790
[5] Kim Y C, Kim K H, Son D Y, Jeong D N, Seo J Y, Choi Y S, Han I T, Lee S Y and Park N G 2017 Nature 550 87
[6] Boroum, F A, Zhu M, Dalton A B, Keddie J L, Sellin P J and Gutierrez J J 2007 Appl. Phys. Lett. 91 033509
[7] Duboz J Y, Frayssinet E, Chenot S, Reverchon J L and Idir M 2010 Appl. Phys. Lett. 97 163504
[8] Nava F, Bertuccio G, Cavallini A and Vittone E 2008 E Meas. Sci. Technol. 19 102001
[9] Conte G, Rossi M C, Salvatori S, Ascarelli P and Trucchi D 2004 J. Appl. Phys. 96 6415
[10] Lu X, Zhou L D, Chen L, Ouyang X P, Liu B, Xu J and Tang H L 2018 Appl. Phys. Lett. 112 103502
[11] Liang H L, Cui S J, Su R, Guan P F, He Y H, Yang L H, Chen L M, Zhang Y H, Mei Z X and Du X L 2019 ACS Photon. 6 351
[12] Zhao X L, Chen L, He Y N, Liu J L, Peng W B, Huang Z Y, Qi X M, Pan Z J, Zhang W T, Zhang Z B and Ouyang X P 2016 Appl. Phys. Lett. 108 171103
[13] Endo H, Chiba T, Meguro K, Takahashi K, Fujisawa M, Sugimura S, Narita S, Kashiwaba Y and Sato E 2011 Nucl. Instrum. Methods Phys. Res. Sect. A 665 15
[14] Zhou L D, Huang Z Y, Zhao X L, He Y N, Chen L, Xu M X, Zhao K, Zhang S C and Ouyang X P 2019 IEEE Photon. Technol. Lett. 31 365
[15] Zhao X L, Kang X, Chen L, Zhang Z B, Liu J L, Ouyang X P, Peng W B, He Y N 2014 Acta Phys. Sin. 63 098502 (in Chinese)
[16] Liu L S, Mei Z X, Tang A H, Azarov A, Kuznetsov A, Xue Q K and Du X L 2016 Phys. Rev. B 93 235305
[17] Tang A H, Mei Z X, Huo W X and Du X L 2018 Sci. Chin.-Phys. Mech. Astron. 61 117321
[18] Monakhov E V, Kuznetsov A Y and Svensson B G 2009 J. Phys. D 42 153001
[19] Zhang Y F, Chen X H, Xu Y, Ren F F, Gu S L, Zhang R, Zheng Y D and Ye J D 2019 Chin. Phys. B 28 028501
[20] Lany S and Zunger A 2005 Phys. Rev. B 72 035215
[21] Hu W J, Paudel T R, Lopatin S, Wang Z, Ma H, Wu K, Bera A, Yuan G, Gruverman A, Tsymbal E Y and Wu T 2018 Adv. Funct. Mater. 28 1704337