Applied Radiation and Isotopes 94 (2014) 314–318
Contents lists available at ScienceDirect
Applied Radiation and Isotopes journal homepage: www.elsevier.com/locate/apradiso
Design and study of a coplanar grid array CdZnTe detector for improved spatial resolution Yuedong Ma n, Shali Xiao, Guoqiang Yang, Liuqiang Zhang Key Laboratory of Optoelectronic Technology and Systems, Ministry of Education, Chongqing University, Chongqing 400044, China
H I G H L I G H T S
A novel structure of coplanar grid CdZnTe detector was designed to evaluate the possibility of applying the detector to gamma-ray imaging applications. The best spatial resolution of coplanar grid CdZnTe detectors ever reported has been achieved, along with good spectroscopic performance. Depth correction of the energy spectra using a new algorithm is presented.
art ic l e i nf o
a b s t r a c t
Article history: Received 23 April 2014 Received in revised form 31 August 2014 Accepted 8 September 2014
Coplanar grid (CPG) CdZnTe detectors have been used as gamma-ray spectrometers for years. Comparing with pixelated CdZnTe detectors, CPG CdZnTe detectors have either no or poor spatial resolution, which directly limits its use in imaging applications. To address the issue, a 2 2 CPG array CdZnTe detector with dimensions of 7 7 5 mm3 was fabricated. Each of the CPG pairs in the detector was moderately shrunk in size and precisely designed to improve the spatial resolution while maintaining good energy resolution, considering the charge loss at the surface between the strips of each CPG pairs. Preliminary measurements were demonstrated at an energy resolution of 2.7–3.9% for the four CPG pairs using 662 keV gamma rays and with a spatial resolution of 3.3 mm, which is the best spatial resolution ever achieved for CPG CdZnTe detectors. The results reveal that the CPG CdZnTe detector can also be applied to imaging applications at a substantially higher spatial resolution. & 2014 Elsevier Ltd. All rights reserved.
Keywords: Coplanar grid CdZnTe Spatial resolution Gamma-ray detector
1. Introduction Cadmium zinc telluride (CdZnTe) has several attractive properties, such as its high atomic number and relatively large band gap, making it suitable for room-temperature high-energy-radiation detection (Zhang et al., 2013). Nevertheless, one shortcoming of this material is the poor mobility-lifetime product for holes, which leads to substantial amounts of charge trapping in planar CdZnTe detectors. To circumvent the tailing in the low energy section of the energy spectrum caused by hole trapping, a series of single polarity CdZnTe detectors in which the induced signals on the anode mainly depend on the movement of the electrons have been designed. As a single polarity device, the coplanar grid (CPG) CdZnTe detector was first proposed in 1994 (Luke, 1994, 1995). Under
n
Corresponding author. Tel.: þ 86 18716427286; fax: þ 86 2365105287. E-mail address: [email protected] (Y. Ma).
http://dx.doi.org/10.1016/j.apradiso.2014.09.003 0969-8043/& 2014 Elsevier Ltd. All rights reserved.
operating conditions, the signals from the non-collecting grid (NCG) are subtracted from that collected by the collecting grid (CG). In addition, the measured signals of the CPG CdZnTe detectors mainly depend on the movement of the electrons in the near grid region. One advantage of the CPG CdZnTe detectors is the highly uniform charge induction efficiency (CIE) distribution, which demonstrates the good intrinsic spectral performance (Luke, 1996). Furthermore, gamma-ray detection can be performed by this device using relatively simple electronic readout systems (He et al., 2005). In addition to a good spectroscopic performance, the radiation detectors must have a good spatial sensitivity for several applications. However, CPG CdZnTe detectors exhibit a poor spatial resolution compared with pixelated CdZnTe detectors. Few attempts have been made to improve the spatial resolution of CPG detectors. In the pursuit of good energy resolution and high detection efficiency, Luke assembled four independent detector modules, each consisting of a 1 cm3 CdZnTe detector and associated electronics, to form a 2 2 CPG detector array (Luke et al., 2001). For the purpose of improving the detection efficiency while
Y. Ma et al. / Applied Radiation and Isotopes 94 (2014) 314–318
maintaining good energy resolution, He et al. 2005 employed four independent CPG pairs on the anode surface of a single CdZnTe crystal with dimensions of 30 30.5 12 mm3 . Both researchers independently improved the spatial resolution of the CPG detectors to approximately 14 mm. In this paper, a novel 2 2 CPG array CdZnTe detector has been precisely designed and fabricated to improve the spatial resolution to the best resolution yet achieved while maintaining good energy resolution. The spectroscopic performance of the detector was measured, and the energy resolution at different interaction depths was obtained using a depth-sensing method.
2. Detector design To improve the spatial resolution of the CPG CdZnTe detector, we need to reduce the size of each CPG pair. However, to what extent we can decrease the size to obtain a better performance by the detector remains unknown. When we shrink the size of each CPG pair, the detection efficiency of the detector will be degraded because only four CPG pairs are employed in the current detector. The detection efficiency can be improved by employing more CPG pairs to form a large-scale CPG array CdZnTe detector without degrading the spatial resolution. In addition, in this paper, the size of the detector was precisely designed to avoid severe degradation of the detection efficiency. Another restriction is caused by charge loss occurring on the surface between the strips of each CPG pair, which should be similar to that in pixelated CdZnTe detectors (Bolotnikov et al., 1999). Ideally, the surface resistivity of the CdZnTe material is supposed to be extremely high; therefore, all the electric field lines will terminate on the electrodes. The charge generated between two collecting strips of the CG will be split and shared by the two collecting strips by virtue of the inter-grid bias. In reality, however, the electric field distribution will be significantly altered even with a fairly small surface conductivity. Therefore, a portion of the electric field lines that originally terminated on the electrodes will intersect the surface between the strips, leading to charge loss at the surface. To reduce this type of charge loss, we decreased the size of the gap between the strips and set the strip pitch of the CPG pairs to be larger than the diameter of the electron cloud. Because the larger size of the electron cloud enhances the possibility of charge sharing between multiple collecting strips, the probability of charge loss is increased.
315
When a gamma ray is deposited in the CdZnTe detector, an electron cloud appears in the vicinity of the interaction point. The size of the electron cloud expands as it drifts through the crystal, mainly due to the diffusion caused by the thermal motion. For a 137 Cs gamma ray, the mean diameter of the initial electron cloud after energy deposition is 200 mm, and the expanded diameter (ΔD) of the electron cloud is given by Kim et al. (2011). rffiffiffiffi ld ð1Þ ΔD ¼ 0:529 V where l is the drift distance, d (5 mm) is the detector thickness, and V ( 1000 V) is the cathode bias. Because the expanded diameter is calculated to be 85 mm, the final diameter of the pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi electron cloud is 2002 þ852 ¼ 220 mm. Fig. 1(a) shows the photo mask picture of the CPG array CdZnTe detector with dimensions of 7 7 5 mm3. The detector consists of four independent CPG electrode pairs (P11, P12, P21, P22) forming a 2 2 CPG array on the anode surface, which is separated by a boundary electrode. The gap between the strips of each CPG pair is 100 mm. In addition, the strip pitch is 300 mm, which is larger than the limitation of 220 mm. To reduce the impact of the edge effect, the distance between the peripheral electrode of the CPG pairs and the boundary electrode is designed to be 400 mm. Fig. 1(b) shows the details of the detector. It is worth noting that our design is different from the multipair CdZnTe detector (He et al., 2005) in two respects. We decreased the size of the two outermost strips of the CPG pairs and adopted the conventional geometry of the CPG electrode, rather than the coplanar anodes design of generation 3 (He et al., 1998), for the convenience of conducting each CPG pairs in such a limited space. In addition, a planar electrode, instead of another four multi-pair CPG pairs, was adopted to improve the weighting potential uniformity toward the cathode surface. The structure of the detector was designed using COMSOL Multi-physics 4.2, a large multi-physics finite element analysis software package, which enables us to calculate the weighting potential distribution in the detector. The final design was obtained considering the optimization of the weighting potential uniformity and the limitation of the spatial resolution. Fig. 2 shows the weighting potential distribution of the collecting grid across the section formed by the X axis shown in Fig. 1(a) and the depth axis (Z axis). As shown in Fig. 2, the weighting potential of the CG increases linearly from 0 in the cathode surface to 1/2 in the
Fig. 1. (a) The photo mask picture of the anode electrode of the 2 2 CPG array CdZnTe detector. The X axis and the depth axis (Z axis) form the section over which the calculated weighting potential is shown in Fig. 2. (b) Schematic representation of the real CPG array CdZnTe detector.
316
Y. Ma et al. / Applied Radiation and Isotopes 94 (2014) 314–318
4. Results and discussion 4.1. I–V Characteristics and surface resistance We first measured the leakage current of the detector under different biases. The I–V curves for the four CPG pairs were similar. Thus, in Fig. 3, we only show the I–V curve of P11. For comparison, the I–V properties of the planar CdZnTe detector, which was previously fabricated on the same crystal, were also measured and are presented. As shown in Fig. 3, the leakage current of the CPG array CdZnTe detector is more than twice that of the planar grid detector, which reveals a high degree of noise in the CPG detector. However, it is believed that the noise will decrease with decreasing CPG size, which can be observed as another advantage of using relatively small CPG CdZnTe detectors, apart from obtaining a better spatial resolution.
4.2. Spectroscopic performances Fig. 2. Weighting potential of the collecting grid along the section formed by the X axis shown in Fig. 1 (a) and the depth axis (Z axis), where both the X axis and Z axis have been normalized.
vicinity of the anode surface. In the region near the anode surface, the weighting potential of the CG increases sharply to unit magnitude, while the weighting potentials of the NCG and the boundary electrode decrease dramatically to zero.
3. Experimental methods The CdZnTe crystals employed in the present study were grown using the modified vertical Bridgman method. To perform a preliminary qualification of the materials, the spectroscopic performances of the planar detectors were measured under irradiation by alpha particles. Only the detector with the best energy resolution was adopted, and the coplanar grid electrode was eventually reprocessed on the detector through a photo mask. The CPG array CdZnTe detector was placed in a box shielded against interference from the external environment. Appropriate biases were applied to the CG, NCG, boundary electrode and cathode. The signals from the CG and the NCG were concurrently read out by two preamplifiers and then directly digitized by a GWInstek GDS-2204A digital storage oscilloscope. The sampled pulses were acquired and processed under the control of Labview. A depth sensing method was employed to correct the energy spectrum, where a new algorithm was applied to obtain more accurate interaction depth information. Depth sensing in coplanar grid CdZnTe detectors was previously performed by reading out the signals from the CG and NCG. The depth ratio could then be obtained by (SC þSNC)/(SC SNC), where SC and SNC are the signals from the CG and NCG, respectively (He et al., 2005). However, the interaction depth obtained by this depth sensing algorithm is not sufficiently precise because electron trapping has been ignored in the calculation. To obtain a more accurate interaction depth, a recently proposed algorithm, which has not yet been verified experimentally, was used. In the new algorithm (Fritts et al., 2013), the interaction depth z is 1 SC þ SNC z ¼ λ ln 1 þ λ SC SNC
The biases of the detector were set as follows: the cathode was biased to 1000 V, the NCG was biased to 50 V, and the CG and the boundary electrode were set to ground. The energy spectra at 662 keV for the four CPG pairs were obtained separately. It is noteworthy that regardless of the pair that the signal was read out from, the other three pairs were biased at the same level, which was performed to avoid a weighting potential non-symmetry under operation. The signal pulses were first read out from P11, and an energy resolution of 2.7% for 137Cs gamma rays was obtained with the depth correction method. Partly because the thickness of the detector is only 5 mm, the improvement in the energy resolution caused by the depth correction with the new depth sensing algorithm is not obviously greater than that obtained with the conventional algorithm. The depth-corrected spectrum utilizing the new depth sensing algorithm is shown in Fig. 4(a) along with the uncorrected spectrum. In contrast to the energy resolution of 3.7% obtained by the multi-pair CPG detectors, the CPG array CdZnTe detector achieves better spectroscopic performance. Ignoring the difference caused by the material variance, two effects may contribute to this improvement. First, the weighting potential uniformity near the cathode surface is better when employing a planar electrode rather than when using four additional CPG pairs. In addition, with a smaller CPG size, the degradation of the spectroscopic performance caused by the noise will be reduced. As we mentioned above, the depth sensing method cannot only be applied to correct the energy spectrum, but it can also provide
ð2Þ
where λ is the mean trapping length, which only depends on the electron mobility-lifetime product (meτe) and on the electric field intensity E.
Fig. 3. The I–V characteristics of the planar CdZnTe detector and the CPG array CdZnTe detector; two detectors were fabricated using the same CdZnTe crystal.
Y. Ma et al. / Applied Radiation and Isotopes 94 (2014) 314–318
317
Fig. 4. (a) The uncorrected spectrum along with the depth-corrected spectrum of P11 for 662 keV gamma rays. (b) Energy resolution as a function of interaction depth.
Fig. 5. Depth-corrected energy spectra of
137
us with information about the interaction depth; therefore, we can obtain energy spectra at different interaction depths. Fig. 4 (b) illustrates the energy resolution as a function of the interaction depth of the detector. The detector was divided into 12 parts from the anode surface to the cathode surface. As shown in Fig. 4(b), the best energy resolution was observed near the cathode surface. Then, the energy resolution degrades almost linearly as the depth decreases, with a depth index ranging from 12 to 4. We found the worst energy resolution near the anode surface. This is mainly because the weighting potential non-uniformity is the most severe in the vicinity of the anode surface. The spectroscopic performances of the other three CPG pairs (P12, P21 and P22) were also measured with the same bias level.
Cs for the four CPG pairs P11, P12, P21 and P22.
The peak-to-Compton (p/c) and peak-to-valley (p/v) ratios for the three pairs as well as P11 were calculated separately, as shown in Fig. 5. The energy resolutions for 662 keV gamma rays of the four CPG pairs are 2.7%, 3.65, 3.1% and 3.9%, respectively. The difference in energy resolution of the four CPG pairs may result from the varying quality of the CdZnTe crystals in the four regions. In addition, the p/c and p/v ratios are P11 (p/c¼4.6, p/v¼39.9), P12 (p/c¼ 3.5, p/v¼ 33.1), P21 (p/c¼4.5, p/v¼36.0) and P22 (p/c¼2.8, p/v¼28.5). Interestingly, the CPG pair with a higher energy resolution also shows a higher p/c ratio and a lower p/v ratio. The desired spectroscopic performances of the CPG array CdZnTe detector have been achieved, with an intrinsic spatial resolution of 3.3 mm, indicating that CPG CdZnTe detectors can obtain a much better
318
Y. Ma et al. / Applied Radiation and Isotopes 94 (2014) 314–318
spatial resolution with a similar energy resolution. The results imply the possibility of CPG CdZnTe detectors being applied to imaging applications. 5. Conclusions A 2 2 CPG array CdZnTe detector with improved spatial resolution and good spectroscopic performance was designed and fabricated. The preliminary results demonstrated 2.7–3.9% energy resolutions for the four CPG pairs using 662 keV gamma rays at a spatial resolution of 3.3 mm, indicating that CPG CdZnTe detectors can achieve a substantially higher spatial sensitivity with similar spectroscopic performance. These results demonstrate the potential for CPG CdZnTe detectors being applied to imaging applications in practice. Further investigation on optimizing the size and structure of the detector will be pursued in the future to improve the spatial resolution of the CPG array CdZnTe detector. In addition, additional CPG pairs can be assembled to form a large-scale CPG array CdZnTe detector so that imaging applications utilizing CPG CdZnTe detectors with more picture elements and better detection efficiency can be realized experimentally. As the number of CPG pairs increases, a simpler readout scheme is obtained. This is because, through the proper designing of the geometry of the CPGs, one can obtain a good energy resolution by collecting signals from only one electrode (Amman and Luke, 1997). Acknowledgments This work is supported by the National Natural Science Foundation of China (No. 61274048).
References Zhang, Q., Zhang, C., Lu, Y., Yang, K., Ren, Q., 2013. Progress in the development of CdZnTe unipolar detectors for different anode geometries and data corrections. Sensors 13, 2447–2474. Luke, P.N., 1994. Single‐polarity charge sensing in ionization detectors using coplanar electrodes. Appl. Phys. Lett. 65, 2884–2886. Luke, P.N., 1995. Unipolar charge sensing with coplanar electrodes-application to semiconductor detectors. IEEE Trans. Nucl. Sci. 42, 207–213. Luke, P.N., 1996. Electrode configuration and energy resolution in gamma-ray detectors. Nucl. Instrum. Methods Phys. Res. Sect. A: Accel. Spectrom., Detect. Assoc. Equip. 380, 232–237. He, Z., Sturm, B.W., Rhodes, E., 2005. Investigation of a large volume multi-pair coplanar grid CdZnTe detector for improved detection efficiency. In: Proceedings of the IEEE Nuclear Science Symposium Conference Record, 2005, pp. 1159–1162. Luke, P.N., Amman, M., Lee, J.S., Ludewigt, B.A., Yaver, H., 2001. A CdZnTe coplanargrid detector array for environmental remediation. Nucl. Instrum. Methods Phys. Res. Sect. A: Accel. Spectrom. Detect. Assoc. Equip. 458, 319–324. Bolotnikov, A.E., Cook, W.R., Harrison, F.A., Wong, A.S., Schindler, S.M., Eichelberger, A.C., 1999. Charge loss between contacts of CdZnTe pixel detectors. Nucl. Instrum. Methods Phys. Res. Sect. A: Accel. Spectrom. Detect. Assoc. Equip. 432, 326–331. Kim, J.C., Anderson, S.E., Kaye, W., Zhang, F., Zhu, Y., Kaye, S.J., He, Z., 2011. Charge sharing in common-grid pixelated CdZnTe detectors. Nucl. Instrum. Methods Phys. Res. Sect. A: Accel. Spectrom. Detect. Assoc. Equip. 654, 233–243. He, Z., Knoll, G., Wehe, D., Du, Y., 1998. Coplanar grid patterns and their effect on energy resolution of CdZnTe detectors. Nucl. Instrum. Methods Phys. Res. Sect. A: Accel. Spectrom. Detect. Assoc. Equip. 411, 107–113. Fritts, M., Durst, J., Göpfert, T., Wester, T., Zuber, K., 2013. Analytical model for event reconstruction in coplanar grid CdZnTe detectors. Nucl. Instrum. Methods Phys. Res. Sect. A: Accel. Spectrom. Detect. Assoc. Equip. 708, 1–6. Amman, M.S., Luke, P.N., 1997. Coplanar-grid detector with single-electrode readout. Proc. SPIE3115, 205–213.
Contents lists available at ScienceDirect
Applied Radiation and Isotopes journal homepage: www.elsevier.com/locate/apradiso
Design and study of a coplanar grid array CdZnTe detector for improved spatial resolution Yuedong Ma n, Shali Xiao, Guoqiang Yang, Liuqiang Zhang Key Laboratory of Optoelectronic Technology and Systems, Ministry of Education, Chongqing University, Chongqing 400044, China
H I G H L I G H T S
A novel structure of coplanar grid CdZnTe detector was designed to evaluate the possibility of applying the detector to gamma-ray imaging applications. The best spatial resolution of coplanar grid CdZnTe detectors ever reported has been achieved, along with good spectroscopic performance. Depth correction of the energy spectra using a new algorithm is presented.
art ic l e i nf o
a b s t r a c t
Article history: Received 23 April 2014 Received in revised form 31 August 2014 Accepted 8 September 2014
Coplanar grid (CPG) CdZnTe detectors have been used as gamma-ray spectrometers for years. Comparing with pixelated CdZnTe detectors, CPG CdZnTe detectors have either no or poor spatial resolution, which directly limits its use in imaging applications. To address the issue, a 2 2 CPG array CdZnTe detector with dimensions of 7 7 5 mm3 was fabricated. Each of the CPG pairs in the detector was moderately shrunk in size and precisely designed to improve the spatial resolution while maintaining good energy resolution, considering the charge loss at the surface between the strips of each CPG pairs. Preliminary measurements were demonstrated at an energy resolution of 2.7–3.9% for the four CPG pairs using 662 keV gamma rays and with a spatial resolution of 3.3 mm, which is the best spatial resolution ever achieved for CPG CdZnTe detectors. The results reveal that the CPG CdZnTe detector can also be applied to imaging applications at a substantially higher spatial resolution. & 2014 Elsevier Ltd. All rights reserved.
Keywords: Coplanar grid CdZnTe Spatial resolution Gamma-ray detector
1. Introduction Cadmium zinc telluride (CdZnTe) has several attractive properties, such as its high atomic number and relatively large band gap, making it suitable for room-temperature high-energy-radiation detection (Zhang et al., 2013). Nevertheless, one shortcoming of this material is the poor mobility-lifetime product for holes, which leads to substantial amounts of charge trapping in planar CdZnTe detectors. To circumvent the tailing in the low energy section of the energy spectrum caused by hole trapping, a series of single polarity CdZnTe detectors in which the induced signals on the anode mainly depend on the movement of the electrons have been designed. As a single polarity device, the coplanar grid (CPG) CdZnTe detector was first proposed in 1994 (Luke, 1994, 1995). Under
n
Corresponding author. Tel.: þ 86 18716427286; fax: þ 86 2365105287. E-mail address: [email protected] (Y. Ma).
http://dx.doi.org/10.1016/j.apradiso.2014.09.003 0969-8043/& 2014 Elsevier Ltd. All rights reserved.
operating conditions, the signals from the non-collecting grid (NCG) are subtracted from that collected by the collecting grid (CG). In addition, the measured signals of the CPG CdZnTe detectors mainly depend on the movement of the electrons in the near grid region. One advantage of the CPG CdZnTe detectors is the highly uniform charge induction efficiency (CIE) distribution, which demonstrates the good intrinsic spectral performance (Luke, 1996). Furthermore, gamma-ray detection can be performed by this device using relatively simple electronic readout systems (He et al., 2005). In addition to a good spectroscopic performance, the radiation detectors must have a good spatial sensitivity for several applications. However, CPG CdZnTe detectors exhibit a poor spatial resolution compared with pixelated CdZnTe detectors. Few attempts have been made to improve the spatial resolution of CPG detectors. In the pursuit of good energy resolution and high detection efficiency, Luke assembled four independent detector modules, each consisting of a 1 cm3 CdZnTe detector and associated electronics, to form a 2 2 CPG detector array (Luke et al., 2001). For the purpose of improving the detection efficiency while
Y. Ma et al. / Applied Radiation and Isotopes 94 (2014) 314–318
maintaining good energy resolution, He et al. 2005 employed four independent CPG pairs on the anode surface of a single CdZnTe crystal with dimensions of 30 30.5 12 mm3 . Both researchers independently improved the spatial resolution of the CPG detectors to approximately 14 mm. In this paper, a novel 2 2 CPG array CdZnTe detector has been precisely designed and fabricated to improve the spatial resolution to the best resolution yet achieved while maintaining good energy resolution. The spectroscopic performance of the detector was measured, and the energy resolution at different interaction depths was obtained using a depth-sensing method.
2. Detector design To improve the spatial resolution of the CPG CdZnTe detector, we need to reduce the size of each CPG pair. However, to what extent we can decrease the size to obtain a better performance by the detector remains unknown. When we shrink the size of each CPG pair, the detection efficiency of the detector will be degraded because only four CPG pairs are employed in the current detector. The detection efficiency can be improved by employing more CPG pairs to form a large-scale CPG array CdZnTe detector without degrading the spatial resolution. In addition, in this paper, the size of the detector was precisely designed to avoid severe degradation of the detection efficiency. Another restriction is caused by charge loss occurring on the surface between the strips of each CPG pair, which should be similar to that in pixelated CdZnTe detectors (Bolotnikov et al., 1999). Ideally, the surface resistivity of the CdZnTe material is supposed to be extremely high; therefore, all the electric field lines will terminate on the electrodes. The charge generated between two collecting strips of the CG will be split and shared by the two collecting strips by virtue of the inter-grid bias. In reality, however, the electric field distribution will be significantly altered even with a fairly small surface conductivity. Therefore, a portion of the electric field lines that originally terminated on the electrodes will intersect the surface between the strips, leading to charge loss at the surface. To reduce this type of charge loss, we decreased the size of the gap between the strips and set the strip pitch of the CPG pairs to be larger than the diameter of the electron cloud. Because the larger size of the electron cloud enhances the possibility of charge sharing between multiple collecting strips, the probability of charge loss is increased.
315
When a gamma ray is deposited in the CdZnTe detector, an electron cloud appears in the vicinity of the interaction point. The size of the electron cloud expands as it drifts through the crystal, mainly due to the diffusion caused by the thermal motion. For a 137 Cs gamma ray, the mean diameter of the initial electron cloud after energy deposition is 200 mm, and the expanded diameter (ΔD) of the electron cloud is given by Kim et al. (2011). rffiffiffiffi ld ð1Þ ΔD ¼ 0:529 V where l is the drift distance, d (5 mm) is the detector thickness, and V ( 1000 V) is the cathode bias. Because the expanded diameter is calculated to be 85 mm, the final diameter of the pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi electron cloud is 2002 þ852 ¼ 220 mm. Fig. 1(a) shows the photo mask picture of the CPG array CdZnTe detector with dimensions of 7 7 5 mm3. The detector consists of four independent CPG electrode pairs (P11, P12, P21, P22) forming a 2 2 CPG array on the anode surface, which is separated by a boundary electrode. The gap between the strips of each CPG pair is 100 mm. In addition, the strip pitch is 300 mm, which is larger than the limitation of 220 mm. To reduce the impact of the edge effect, the distance between the peripheral electrode of the CPG pairs and the boundary electrode is designed to be 400 mm. Fig. 1(b) shows the details of the detector. It is worth noting that our design is different from the multipair CdZnTe detector (He et al., 2005) in two respects. We decreased the size of the two outermost strips of the CPG pairs and adopted the conventional geometry of the CPG electrode, rather than the coplanar anodes design of generation 3 (He et al., 1998), for the convenience of conducting each CPG pairs in such a limited space. In addition, a planar electrode, instead of another four multi-pair CPG pairs, was adopted to improve the weighting potential uniformity toward the cathode surface. The structure of the detector was designed using COMSOL Multi-physics 4.2, a large multi-physics finite element analysis software package, which enables us to calculate the weighting potential distribution in the detector. The final design was obtained considering the optimization of the weighting potential uniformity and the limitation of the spatial resolution. Fig. 2 shows the weighting potential distribution of the collecting grid across the section formed by the X axis shown in Fig. 1(a) and the depth axis (Z axis). As shown in Fig. 2, the weighting potential of the CG increases linearly from 0 in the cathode surface to 1/2 in the
Fig. 1. (a) The photo mask picture of the anode electrode of the 2 2 CPG array CdZnTe detector. The X axis and the depth axis (Z axis) form the section over which the calculated weighting potential is shown in Fig. 2. (b) Schematic representation of the real CPG array CdZnTe detector.
316
Y. Ma et al. / Applied Radiation and Isotopes 94 (2014) 314–318
4. Results and discussion 4.1. I–V Characteristics and surface resistance We first measured the leakage current of the detector under different biases. The I–V curves for the four CPG pairs were similar. Thus, in Fig. 3, we only show the I–V curve of P11. For comparison, the I–V properties of the planar CdZnTe detector, which was previously fabricated on the same crystal, were also measured and are presented. As shown in Fig. 3, the leakage current of the CPG array CdZnTe detector is more than twice that of the planar grid detector, which reveals a high degree of noise in the CPG detector. However, it is believed that the noise will decrease with decreasing CPG size, which can be observed as another advantage of using relatively small CPG CdZnTe detectors, apart from obtaining a better spatial resolution.
4.2. Spectroscopic performances Fig. 2. Weighting potential of the collecting grid along the section formed by the X axis shown in Fig. 1 (a) and the depth axis (Z axis), where both the X axis and Z axis have been normalized.
vicinity of the anode surface. In the region near the anode surface, the weighting potential of the CG increases sharply to unit magnitude, while the weighting potentials of the NCG and the boundary electrode decrease dramatically to zero.
3. Experimental methods The CdZnTe crystals employed in the present study were grown using the modified vertical Bridgman method. To perform a preliminary qualification of the materials, the spectroscopic performances of the planar detectors were measured under irradiation by alpha particles. Only the detector with the best energy resolution was adopted, and the coplanar grid electrode was eventually reprocessed on the detector through a photo mask. The CPG array CdZnTe detector was placed in a box shielded against interference from the external environment. Appropriate biases were applied to the CG, NCG, boundary electrode and cathode. The signals from the CG and the NCG were concurrently read out by two preamplifiers and then directly digitized by a GWInstek GDS-2204A digital storage oscilloscope. The sampled pulses were acquired and processed under the control of Labview. A depth sensing method was employed to correct the energy spectrum, where a new algorithm was applied to obtain more accurate interaction depth information. Depth sensing in coplanar grid CdZnTe detectors was previously performed by reading out the signals from the CG and NCG. The depth ratio could then be obtained by (SC þSNC)/(SC SNC), where SC and SNC are the signals from the CG and NCG, respectively (He et al., 2005). However, the interaction depth obtained by this depth sensing algorithm is not sufficiently precise because electron trapping has been ignored in the calculation. To obtain a more accurate interaction depth, a recently proposed algorithm, which has not yet been verified experimentally, was used. In the new algorithm (Fritts et al., 2013), the interaction depth z is 1 SC þ SNC z ¼ λ ln 1 þ λ SC SNC
The biases of the detector were set as follows: the cathode was biased to 1000 V, the NCG was biased to 50 V, and the CG and the boundary electrode were set to ground. The energy spectra at 662 keV for the four CPG pairs were obtained separately. It is noteworthy that regardless of the pair that the signal was read out from, the other three pairs were biased at the same level, which was performed to avoid a weighting potential non-symmetry under operation. The signal pulses were first read out from P11, and an energy resolution of 2.7% for 137Cs gamma rays was obtained with the depth correction method. Partly because the thickness of the detector is only 5 mm, the improvement in the energy resolution caused by the depth correction with the new depth sensing algorithm is not obviously greater than that obtained with the conventional algorithm. The depth-corrected spectrum utilizing the new depth sensing algorithm is shown in Fig. 4(a) along with the uncorrected spectrum. In contrast to the energy resolution of 3.7% obtained by the multi-pair CPG detectors, the CPG array CdZnTe detector achieves better spectroscopic performance. Ignoring the difference caused by the material variance, two effects may contribute to this improvement. First, the weighting potential uniformity near the cathode surface is better when employing a planar electrode rather than when using four additional CPG pairs. In addition, with a smaller CPG size, the degradation of the spectroscopic performance caused by the noise will be reduced. As we mentioned above, the depth sensing method cannot only be applied to correct the energy spectrum, but it can also provide
ð2Þ
where λ is the mean trapping length, which only depends on the electron mobility-lifetime product (meτe) and on the electric field intensity E.
Fig. 3. The I–V characteristics of the planar CdZnTe detector and the CPG array CdZnTe detector; two detectors were fabricated using the same CdZnTe crystal.
Y. Ma et al. / Applied Radiation and Isotopes 94 (2014) 314–318
317
Fig. 4. (a) The uncorrected spectrum along with the depth-corrected spectrum of P11 for 662 keV gamma rays. (b) Energy resolution as a function of interaction depth.
Fig. 5. Depth-corrected energy spectra of
137
us with information about the interaction depth; therefore, we can obtain energy spectra at different interaction depths. Fig. 4 (b) illustrates the energy resolution as a function of the interaction depth of the detector. The detector was divided into 12 parts from the anode surface to the cathode surface. As shown in Fig. 4(b), the best energy resolution was observed near the cathode surface. Then, the energy resolution degrades almost linearly as the depth decreases, with a depth index ranging from 12 to 4. We found the worst energy resolution near the anode surface. This is mainly because the weighting potential non-uniformity is the most severe in the vicinity of the anode surface. The spectroscopic performances of the other three CPG pairs (P12, P21 and P22) were also measured with the same bias level.
Cs for the four CPG pairs P11, P12, P21 and P22.
The peak-to-Compton (p/c) and peak-to-valley (p/v) ratios for the three pairs as well as P11 were calculated separately, as shown in Fig. 5. The energy resolutions for 662 keV gamma rays of the four CPG pairs are 2.7%, 3.65, 3.1% and 3.9%, respectively. The difference in energy resolution of the four CPG pairs may result from the varying quality of the CdZnTe crystals in the four regions. In addition, the p/c and p/v ratios are P11 (p/c¼4.6, p/v¼39.9), P12 (p/c¼ 3.5, p/v¼ 33.1), P21 (p/c¼4.5, p/v¼36.0) and P22 (p/c¼2.8, p/v¼28.5). Interestingly, the CPG pair with a higher energy resolution also shows a higher p/c ratio and a lower p/v ratio. The desired spectroscopic performances of the CPG array CdZnTe detector have been achieved, with an intrinsic spatial resolution of 3.3 mm, indicating that CPG CdZnTe detectors can obtain a much better
318
Y. Ma et al. / Applied Radiation and Isotopes 94 (2014) 314–318
spatial resolution with a similar energy resolution. The results imply the possibility of CPG CdZnTe detectors being applied to imaging applications. 5. Conclusions A 2 2 CPG array CdZnTe detector with improved spatial resolution and good spectroscopic performance was designed and fabricated. The preliminary results demonstrated 2.7–3.9% energy resolutions for the four CPG pairs using 662 keV gamma rays at a spatial resolution of 3.3 mm, indicating that CPG CdZnTe detectors can achieve a substantially higher spatial sensitivity with similar spectroscopic performance. These results demonstrate the potential for CPG CdZnTe detectors being applied to imaging applications in practice. Further investigation on optimizing the size and structure of the detector will be pursued in the future to improve the spatial resolution of the CPG array CdZnTe detector. In addition, additional CPG pairs can be assembled to form a large-scale CPG array CdZnTe detector so that imaging applications utilizing CPG CdZnTe detectors with more picture elements and better detection efficiency can be realized experimentally. As the number of CPG pairs increases, a simpler readout scheme is obtained. This is because, through the proper designing of the geometry of the CPGs, one can obtain a good energy resolution by collecting signals from only one electrode (Amman and Luke, 1997). Acknowledgments This work is supported by the National Natural Science Foundation of China (No. 61274048).
References Zhang, Q., Zhang, C., Lu, Y., Yang, K., Ren, Q., 2013. Progress in the development of CdZnTe unipolar detectors for different anode geometries and data corrections. Sensors 13, 2447–2474. Luke, P.N., 1994. Single‐polarity charge sensing in ionization detectors using coplanar electrodes. Appl. Phys. Lett. 65, 2884–2886. Luke, P.N., 1995. Unipolar charge sensing with coplanar electrodes-application to semiconductor detectors. IEEE Trans. Nucl. Sci. 42, 207–213. Luke, P.N., 1996. Electrode configuration and energy resolution in gamma-ray detectors. Nucl. Instrum. Methods Phys. Res. Sect. A: Accel. Spectrom., Detect. Assoc. Equip. 380, 232–237. He, Z., Sturm, B.W., Rhodes, E., 2005. Investigation of a large volume multi-pair coplanar grid CdZnTe detector for improved detection efficiency. In: Proceedings of the IEEE Nuclear Science Symposium Conference Record, 2005, pp. 1159–1162. Luke, P.N., Amman, M., Lee, J.S., Ludewigt, B.A., Yaver, H., 2001. A CdZnTe coplanargrid detector array for environmental remediation. Nucl. Instrum. Methods Phys. Res. Sect. A: Accel. Spectrom. Detect. Assoc. Equip. 458, 319–324. Bolotnikov, A.E., Cook, W.R., Harrison, F.A., Wong, A.S., Schindler, S.M., Eichelberger, A.C., 1999. Charge loss between contacts of CdZnTe pixel detectors. Nucl. Instrum. Methods Phys. Res. Sect. A: Accel. Spectrom. Detect. Assoc. Equip. 432, 326–331. Kim, J.C., Anderson, S.E., Kaye, W., Zhang, F., Zhu, Y., Kaye, S.J., He, Z., 2011. Charge sharing in common-grid pixelated CdZnTe detectors. Nucl. Instrum. Methods Phys. Res. Sect. A: Accel. Spectrom. Detect. Assoc. Equip. 654, 233–243. He, Z., Knoll, G., Wehe, D., Du, Y., 1998. Coplanar grid patterns and their effect on energy resolution of CdZnTe detectors. Nucl. Instrum. Methods Phys. Res. Sect. A: Accel. Spectrom. Detect. Assoc. Equip. 411, 107–113. Fritts, M., Durst, J., Göpfert, T., Wester, T., Zuber, K., 2013. Analytical model for event reconstruction in coplanar grid CdZnTe detectors. Nucl. Instrum. Methods Phys. Res. Sect. A: Accel. Spectrom. Detect. Assoc. Equip. 708, 1–6. Amman, M.S., Luke, P.N., 1997. Coplanar-grid detector with single-electrode readout. Proc. SPIE3115, 205–213.
