Characterization of CdZnTeSe Nuclear Detector Chemically Etched in Bromine Methanol

Semiconductor nuclear radiation detectors made from tertiary and quater-nary compounds of cadmium telluride (CdTe) can operate at room temperature without cryogenic cooling. One of such materials that have become of great interest is cadmium zinc telluride selenide (CdZnTeSe). Compared to other CdTe-based materials, such as cadmium zinc telluride (CdZnTe), CdZnTeSe can be grown with much less Te inclusions and sub-grain boundary networks. Chemical etching is often used to smoothen wafer surfaces during detector fabrication. This paper presents the characterization of CdZnTeSe that is chemically etched using bromine methanol solution. Infrared imaging shows that the wafer has no sub-grain boundary networks that often limit detector performance. The current-voltage (I-V) characterization experiment gave a resistivity of 4.6 × 10 10 Ω-cm for the sample. The I-V curve was linear in the ±10 to ±50 volts range. An energy resolution of 7.2% was recorded at 100 V for the 59.6-keV gamma line of 241 Am.


Introduction
The growing scientific, technological, medical, and industrial applications of rad-Materials Sciences and Applications iation detectors have attracted researchers to work to achieve design and fabrication of nuclear detectors of high resolution and precision that is operable at room temperature. There have been consistent and pungent research efforts worldwide in the past forty years that have been channeled to identify semiconductors for room-temperature x-ray and γ-ray detection [1]. Early efforts in the development of room-temperature semiconductor nuclear detectors include mercuric iodide (HgI 2 ), cadmium telluride (CdTe), cadmium zinc telluride (CdZnTe or CZT), and thallium bromide (TlBr) [2] [3] [4] [5] [6]. Among these materials, CdZnTe has been widely accepted due to its ability for high-resolution nuclear detection at room temperature without cryogenic cooling. CdTe and CdZnTe have found success commercially. Major companies that produce CdZnTe detector devices include Kromek Group PLC [7], Redlen Technologies [8], FLIR Systems [9], Eurorad [10], and H3D [11].
Notwithstanding many years of intense research, innovative improvement, and dominance in the market for its physical and optoelectronics properties, CdZnTe is still hurt by three major damaging imperfections which include high concentration of sub-grain boundaries and their networks, Te inclusion/precipitates and compositional inconsistency that severely degrade the energy resolution of detectors [1] [12] [13] [14]. The indiscriminate distribution of sub-grain boundary networks in the CdZnTe matrix and associated defects are responsible for uneven spatial distribution of trapping centers, which causes spatial variability in the charge transport properties throughout the active volume of CdZnTe devices [1] [15]. Such spatial non-uniformity of the charge collection efficiency is responsible for variability in the pulse-height spectra, depending on the drift paths of the ionized electrons, resulting in broadened photopeak [15]. Thus [27]. The performance of CdZnTeSe and other CdTe-based detector could be limited by surface defects which are often produced during device fabrication. Most surface defects come from the cutting of detector wafers from ingots and during mechanical polishing. The processes for reducing and eliminating the fabrication-induced defects and residues include surface polishing, chemical etching, and chemical passivation [28]- [33]. In this paper, we present the characterization of CdZnTeSe that is chemically etched using bromine methanol solution.

CdZnTeSe Material
The CdZnTeSe is a large-grain polycrystalline material with a grain area of approximately 2.5 × 1.75 cm 2 being the largest [1]. The material was grown by the

Cutting of Detector Wafer
The fabrication of CdTe-based nuclear radiation (x-rays and γ-rays) detection devices starts with the cutting of wafers of desired sizes from the ingot. We used a cutting machine that is equipped with a diamond impregnated wire saw. The wafer used in this study was cut from as-grown CdZnTeSe ingot. The cutting machine is programmable and can be set to cut wafers of desired dimensions. It consists of a movable platform on which the ingot or large wafer is mounted.
The platform is automatically moved between cutting cycles based on the dimensions set in the controller. There is a water pump attached to the system that runs water on the sample through a nuzzle. The water serves as a coolant and a lubricant. The cutting process often introduces surface damages and defects.
These could increase surface leakage currents and result in the trapping of charge carriers generated by ionizing radiation, and thus increase electronic noise, which in turn degrade the energy resolution of the detector [34] [35] [36].
The three major steps used in removing fabrication-induced damages and defects are mechanical polishing, chemical etching, and surface passivation. Chemo-mechanical polishing is sometimes used as an alternative to chemical etching. The sample used in this study has the dimensions of 6.7 × 5.7 × 1.8 mm 3 after mechanical polishing.

Surface Polishing
Mechanical polishing is used as the first stage of removing surface damages caused by cutting and dicing the wafers. The rough surfaces from the cutting process could sometimes create new defects that act as trapping centers for charge carriers transported between the detector's electrodes [31]. The surface

Chemical Etching
Chemical etching is deployed at the second stage to remove fine residual me-  [40]. Chattapadhyay et al. [39], in an earlier study, reported that the effect of increasing the surface current by etching in Br:MeOH solution was lowered by passivating the sample in a potassium hydroxide (KOH) aqueous solution, which leaves a more stoichiometric and smoother surface.
The CdZnTeSe sample used in the present study has dimensions of 6.7 × 5.7 × 1.8 mm 3 . It was chemically etched using Br:MeOH solution. The etching process involve dipping the wafer into a 2% solution of bromine methanol for 2 minutes, and then dried with compressed air.

Surface Passivation
Surface passivation can be regarded as the third stage of removing impactions and surface damages on the CdZnTeSe wafer to enhance the performance of an x-ray and γ-ray nuclear detector at room-temperature. In surface passivation, the chemical treatment is used to minimize surface oxidation, and increase the detector's shelf and operational life [3]. Kim et al. [38] used ammonium sulfide to passivate CdZnTe after etching in a 2% bromine-methanol solution and found that the apparent mobility-lifetime product of the 2-mm-thick CdZnTe detector improved by 21%, with a noticeable decline in the leakage current, surface-state densities, and improved performance of the detector. It was also reported that, by removing the space-charge accumulated on the detectors' side surfaces, the ammonium sulfide passivation of the CdZnTe 9-mm thick detectors, supports establishing a "focusing" internal electric field within the detector's cavity that coxed electrons away from the surface and towards the anode [38]. Further re-

Deposition of Electrical Contacts
Gold electrical contacts were deposited on the two opposite 6.7 × 5.7 mm 2 surfaces of the CdZnTeSe wafer. This involved pipetting gold chloride solution on each surface. After the gold chloride solution reacts with the surface, the excess solution is removed using a felt paper.
In the boundary layer between metals (contacts) and semiconductors there exist thermal stress at the contact-wafer interface caused by the mismatch of the thermal expansion coefficient between the electrode's material and the CdZnTeSe crystal. Thus, it plays an important role in the determination of the electrical and transport characteristics of the contacts [42]. The method of electroless contact deposition is generally used because of its simplicity and it creates a stronger chemical bond between the metal and the semiconductor compared to sputtering and thermal evaporation methods of which both can create defects in the interface region, leading to trapping effects [3] [43]. It has been reported in literatures that gold (Au) is found to be the contact of choice, because it is sufficiently stable chemically and mechanically to allow bonding and prolonged operation, thus reducing current leakage [44]. Zhang et al. [45], reported that the electroless Au deposition method creates more inter-diffusion between the contact and CdZnTe material, so resulting in a better gamma response compared to other metal contacts and techniques. Thus, we adopted and used the electroless Au contact deposition technique in this study.

Infrared Transmission Imaging
An infrared transmission imaging was performed after chemical etching, prior to the deposition of electrical contacts. A Nikon Eclipse LV100 microscope was used. The microscope is equipped with an infrared light source, a motorized xyz-translation stage, and an infrared camera that is connected to a computer. It comes with a software for image capturing and analysis. The matrix of the CdZnTeSe wafer is transparent to the infrared light while Te inclusions are opaque. Hence, any Te inclusion present will appear as a dark spot as the infrared light transmits through the bulk of the wafer. The infrared image is shown in Figure 1. The image shows that the CdZnTeSe wafer is almost free with Te inclusions and there are no sub-grain boundary networks. Since Te inclusions trap charge carriers, this image indicates a very good wafer for charge collection uniformity, and hence good radiation detection efficiency [4].

Current-Voltage Measurements
Current-voltage (I-V) measurements were carried out to determine the I-V characteristics and resistivity of the wafer. High resistivity of the order of 10 8 Ω-cm and above is required for CdTe-based detectors to have low leakage current in the nanoamp range [23].

Detector Response Characterization
The detector repose measurement system consists of a standard eV Product sample holder connected to a high voltage power supply, a pre-amplifier, a shaping amplifier, and a multichannel analyzer (MCA) that is connected to a computer with a software for spectral analysis. See schematic diagram in Figure 3. The sample holder in Figure 3 is made of brass and has a beryllium window against    which a sealed 241 Am radiation source is placed. The applied voltage was 100 V.
The recorded spectrum is shown in Figure 4. The detector resolution measured as full-width-at-half-maximum (FWHM) for the 49.6-keV gamma line of 241 Am is 7.2%.

Conclusion
We have characterized a CZTS nuclear detector that is chemically etched using bromine methanol solution. The infrared transmission imaging result showed that the CZTS crystal has no grain boundary network, and it is mostly free of Te inclusions. This implies that the CZTS matrix does not have defects that could adversely affect the transport of charge carriers. The current-voltage curve is linear in the ±10 to ±50 volts range and gave a resistivity of 4.6 × 10 10 Ω-cm. The recorded energy resolution is 7.2% at 100 V bias for the 59.6-keV gamma line of 241 Am. This CZTS has shown to be an effective material for the detection of γ-rays at room temperature. In future studies, we plan to focus on surface passivation, and on the variations in surface current with time after chemical etching.