BACKGROUND
I. Field of the Invention
The present invention relates to the manufacture of devices for the detection of high-energy electromagnetic radiation (X- and γ-rays). More particularly, the present invention relates to the manufacture and the use of the high-spectral resolution virtual Frisch-grid radiation detectors based on CdZnTe, CdTe, CdMnTe, HgI2, TlBr, or other semiconductors capable of operating as single-carrier-transport devices.
II. Background of the Related Art
Semiconductor nuclear radiation detectors have experienced a rapid development in the last few years. They are now used in a large variety of fields, including nuclear physics, X-ray and gamma ray astronomy, and nuclear medicine. Their imaging capabilities, good energy resolution, and the ability to fabricate compact systems are very attractive features, in comparison with other types of detectors, such as gas detectors and scintillators. In recent years, a substantial effort has been invested in developing a range of compound semiconductors with wide band gap and high atomic number for X-ray and gamma ray detectors. These compound semiconductors are generally derived from elements of groups III and V (e.g. GaAs) and groups II and VI (e.g. CdTe) of the periodic table. However, besides binary compounds, ternary materials have been also produced, e.g. CdZnTe and CdMnTe. Among these compound semiconductors, cadmium telluride (CdTe) and cadmium zinc telluride (CdZnTe) are two of the most promising materials for radiation detectors with good energy resolution, high detection efficiency, and room temperature operation.
The drawback of the compound semiconductor based detectors of the prior art is that the amplitude of output signal is affected by the immobile holes left at the interaction point produced by the incident gamma ray. Since gamma rays interact randomly inside the detector's volume, the output signals depend on the locations of interaction points. Such behavior of the output signals, caused by immobile holes, is called the induction effect. This effect degrades spectral resolution of semiconductor detectors unless special measures are implemented to neutralize the holes.
There are two common ways to minimize the induction effect: (1) subtracting the fraction of the charge signal contributed by the stationary holes and (2) electrostatic shielding of the stationary holes. The validity of both techniques is consistent with the Ramo-Shockley theorem (S. Ramo, Proc. IRE 27, p. 584, 1939; W. Shockley, J. Appl. Phys. 9, p. 635, 1938; each of which is hereby incorporated herein by reference in its entirety).
In the first technique, the induced charge contributed by the holes is measured and electronically subtracted from the total output signal. (USSR Patent No. SU-1264723A; incorporated herein by reference in its entirety). The holes-induced signal can be measured with one or several electrodes adjacent to the collecting one. The electrodes can be comprised of pixels or strips and, depending on the locations of interaction points, the same electrode can be used for measuring collected or induced-only charges. The coplanar-grid devices proposed for CdZnTe, (P. N. Luke, Appl. Phys. Lett. 65 (22), pp. 2884-2886, 1994; U.S. Pat. No. 5,530,249; each of which is hereby incorporated herein by reference in its entirety), and fluid Xe detectors (A. Bolotnikov, et al., IEEE Trans. Nucl. Sci., Vol. 51, n.3, pp. 1006-1010, 2004; incorporated herein by reference in its entirety), are special cases of using this technique.
The second technique is based on developing an electrostatic shielding of the stationary holes. There are two types of devices which employ the electrostatic shielding: Frisch-grid and virtual Frisch-grid detectors. In the first type detectors, e.g., in a classic Frisch-grid ionization chamber, a metal grid (or mesh) is used for electrostatic shielding of the collecting electrode (or several collecting electrodes) from the positive ions (holes in the case of semiconductors). In the second type detectors, i.e., the virtual Frisch-grid detectors, the special grounded (or virtually grounded) electrode (or several electrodes) is made to produce essentially the same shielding effect as if a real Frisch-grid were placed inside the detector. In the past, several designs of the virtual Frisch-grid detectors were proposed for CdZnTe semiconductor: pixel detectors (H. H. Barrett, et al., Phys. Rev. Lett. 75 (1), p. 156, 1995; incorporated herein by reference in its entirety), CAPture™ (K. Parnham, et al., in Hard X-Ray, Gamma-Ray and Neutron Detector Physics, Proceedings of SPIE, 1999; incorporated herein by reference in its entirety), hemispherical, (C. Szeles, et al., in Hard X-Ray and Gamma-Ray Detector Physics VIII, edited by Larry A. Franks, et al., Proceedings of SPIE Vol. 6319 (SPIE, Bellingham, Wash., 2006); incorporated herein by reference in its entirety), and Frisch-ring, (U.S. Pat. No. 6,175,120; G. Montemont, et al., IEEE Trans. NucL ScL, Vol. 48, pp. 278-281, 2001; each of which is incorporated herein by reference in its entirety).
Pixel Detectors
In the pixel detectors, the charge signal induced on a given pixel by the stationary charges is greatly reduced, because it is shared between other pixels. Therefore, for each individual pixel the other pixels (virtually grounded) act as an electrostatic shield (it is also called the “small-pixel effect”).
CAPture™ and Hemispherical Detectors
The CAPture™ and hemispherical detectors as shown in FIG. 1A are produced by extending the cathode electrode up the sides of the detector body. In both devices, the extended electrodes are in a physical contact with the semiconductor surfaces. As a result, a wide area of the bare surface surrounding the anode is required to keep the surface leakage current below an acceptable level.
Frisch-Ring Detectors
As shown in FIG. 1B, in the existing Frisch-ring detectors (U.S. Pat. No. 6,175,120, Montemont, 2001; Bolotnikov, 2006), the cathode is also extended up the sides of the detector as in the CAPture™ and hemispherical detectors (see FIG. 1A), but the extended portion of the cathode is physically separated from the semiconductor surfaces by a thin layer of insulating material. The non-contacting electrode (also known as the non-contacting Frisch-ring; U.S. Pat. No. 6,175,120; incorporated herein by reference) is the main feature that makes Frisch-ring detectors different from CAPture™ and hemispherical-type devices.
In the Frisch-ring detectors the non-contacting electrode is placed on the side surfaces of the crystal and physically, connected to the cathode, as illustrated in FIG. 1B. (U.S. Pat. No. 6,175,120, Montemont, 2001; Bolotnikov, 2006). A common high-voltage bias (with respect to the anode) is applied to the cathode and the ring, while the output signal is read out from the anode. This configuration requires a gap of about 2-3-mm of the unshielded surface to be left near the anode's contact to prevent high leakage current or even possible discharge in the area between these two electrodes. As a result, an electrostatically unshielded area of the surface exists near the anode.
Overall, previously known designs of virtual Frisch-grid detectors (e.g., FIG. 1A and FIG. 1B) have two common problems that affect their spectroscopic performances. The first problem is the presence of the unshielded area near the anode (see FIG. 1A and FIG. 1B) and the second problem is related to the fact that the original virtual Frisch-grid detectors are essentially two-terminal devices in which the cathode signal cannot provide the particle's interaction depth information to correct the electron losses due to electron trapping (Bolotnikov, 2006). For thin detectors the electron trapping has little effect on the detection performance and can be neglected. However, this problem must be solved for thick detectors longer than about 10 mm, usually employed by virtual Frisch-grid detectors, where electron trapping can be significant.
Recently, a new design of virtual Frisch-grid detectors has been proposed to address the above problems, which lead to improved performance of this type of devices (Bolotnikov, et al., in Proceedings of SPIE Hard XRay and Gamma-Ray Detector Physics VIII, Vol. 6702, edited by L. A. Franks, et al., (SPIE, Bellingham, Wash., 2007); incorporated herein by reference). A schematic of the device is shown in FIG. 1C. A rectangular shaped crystal (bar) 102 has the geometrical aspect ratio (a ratio of its length to its width) of 2 or more as in the existing Frisch-ring devices, e.g., FIG. 1B (Bolotnikov, 2006). The cathode metallization 101 is extended 2-3 mm up to the side surfaces. The non-contacting ring 104 covers the device's side surfaces up to the cathode's 101 edge and is kept at the same potential as the anode 105. The insulating layer 103, e.g., the ultra-thin polyester shrink tube, covers the entire area of the side surfaces. This provides decoupling of the non-contacting ring 104 and the cathode 101 and, at the same time, allows one to apply high differential bias (up to 3000 V and higher) between the cathode 101 and the shielding electrode 104 (non-contacting ring). In turn, the decoupling of the cathode allows for implementing the cathode readout scheme, and enables those skilled in the art, first, to correct for electron trapping and, second, to reject the events interacting near the anode 105 which contribute to the background.
However, this design is not optimal for achieving the full capabilities of the cathode readout scheme to correct the charge loss due to trapping. There are contradicting requirements in this approach. To ensure strong shielding effect of the virtual Frisch grid, the shielding electrode on the crystal's side should cover the entire area of the device's surface as shown in FIG. 1C. But when the entire detector is shielded, very few signals can be detected by the cathode. This makes the cathode insensitive to the events interacting deep inside the detector. On the other hand, it is critical for the good detector performance to be able to detect all events within the crystal and even those that interact close to the anode.
Therefore, it will be desirable to have a virtual Frisch-grid detector(s) and an array of virtual Frisch-grid detectors that effectively reduce the electron trapping problem in the moderately to substantially thick detector crystals, while avoiding the shortcomings of the prior art.
SUMMARY
Embodiments of the invention provide for virtual Frisch-grid detector arrays. In one embodiment, a Frisch-grid detector array includes: a circuit board; a plurality of virtual Frisch-grid detectors electrically connected to the circuit board wherein each virtual Frisch-grid detector includes a semiconductor capable of operating as a single-carrier-transport device having a first surface and a second surface; an anode connected to the first surface of the semiconductor and the circuit board; a common cathode for the plurality of virtual Frisch-grid detectors, wherein the common cathode is connected to the second surface of the semiconductors; a shielding electrode producing the virtual Frisch-grid effect spanning the perimeter of the semiconductor between the first surface and the second surface; and a capacitor electrically connected to the common cathode and the circuit board.
Embodiments of the invention also include radiation detector systems. In one embodiment, a radiation detector system includes the virtual Frisch-grid detector array as described above; an application-specific-integrated-circuit (ASIC) connected to the virtual Frisch-grid detector array operable to receive a signal from said array; and at least one microprocessor connected with the ASIC operable to control the ASIC.
In certain embodiments, the capacitor and circuit board are connected via a wire passing between the plurality of virtual Frisch-grid detectors.
Embodiments also include the wire passing between the plurality of virtual Frisch-grid detectors in an insulated conduit.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1A illustrates the cross-sectional views of the CAPture™ detector according to Parnham (1999).
FIG. 1B illustrates the cross-sectional views of the Frisch-ring detector according to U.S. Pat. No. 6,175,120, Montemont (2001) and Bolotnikov (2006).
FIG. 1C illustrates the cross-sectional views of the virtual Frisch-grid detector according to Bolotnikov (2007).
FIG. 2 illustrates a block diagram of a radiation detector system, which comprises an interconnected virtual Frisch-grid detector array, an application-specific-integrated-circuit (ASIC), and a microprocessor.
FIG. 3 illustrates a diagram of a plurality of 4×4 arrays of virtual Frisch-grid detectors according to FIG. 2.
FIG. 4A illustrates one row of a 6×6 virtual Frisch-grid detector array that reduces the electron trapping problem.
FIG. 4B illustrates one row of a 6×6 virtual Frisch-grid detector array that reduces the electron trapping problem similar to FIG. 4A, however, wherein the length of the shielding electrode differs between the internal and external modules.
FIG. 4C illustrates one row of a 6×6 virtual Frisch-grid detector array that reduces the electron trapping problem similar to FIG. 4B, wherein the edge modules use the full length of the shielding electrode.
FIG. 5A illustrates the cross-sectional view of a virtual Frisch-grid detector with a shielding electrode extending a partial length of a semiconductor near the anode.
FIG. 5B illustrates the cross-sectional view of the virtual Frisch-grid detector of FIG. 5A with the cathode metallization extended up to the side surfaces of the semiconductor.
FIG. 5C illustrates the cross-sectional view of the virtual Frisch-ring detector of FIG. 5A with the shielding electrode extending the full length of a semiconductor used in the external module of the radiation detector array.
FIG. 5D illustrates the cross-sectional view of the virtual Frisch-grid detector of FIG. 5B with the shielding electrode extending the full length between the edge of the anode and the edge of the extended cathode used in the external module of the radiation detector array.
FIG. 6A graphically illustrates an example of a correlation between the signal amplitude from the cathode and the anode of the CdZnTe type detector array.
FIG. 6B graphically illustrates an example of a correlation between the signal amplitude from the cathode and the anode of the CdZnTe type detector array.
FIG. 7A graphically illustrates a pulse-height spectrum produced by an uncollimated 137Cs source with a 14-mm long CdZnTe type detector before correction.
FIG. 7B graphically illustrates a pulse-height spectrum produced by an uncollimated 137Cs source with a 14-mm long CdZnTe type detector after correction relative to FIG. 7A.
FIG. 7C graphically illustrates a pulse-height spectrum produced by an uncollimated 137Cs source with a 14-mm long CdZnTe type detector before correction.
FIG. 7D graphically illustrates a pulse-height spectrum produced by an uncollimated 137Cs source with a 14-mm long CdZnTe type detector after correction relative to FIG. 7C.
FIG. 8 illustrates one row of a virtual Frisch-grid detector array with common cathodes and capacitors.
FIG. 9 illustrates 2×2 virtual Frisch-grid detector array with a common cathode and capacitor.
FIG. 10A illustrates a 6×6×15 mm3 CdZnTe type detector encapsulated in a ultra-thin polyester shell.
FIG. 10B illustrates the encapsulated detector of FIG. 10A and furnished with two spring contacts and a 4 mm wide shielding electrode (aluminum tape with a thin conducting adhesive layer).
FIG. 10C illustrates an array of the detectors of FIG. 10B.
FIG. 11A graphically illustrates pulse-height spectrums produced by an uncollimated 137Cs source on detectors with a common cathode shown in FIG. 10C before correction (results for 3 detectors shown).
FIG. 11B graphically illustrates pulse-height spectrums produced by an uncollimated 137Cs source on detectors with a common cathode shown in FIG. 10C after correction using drift times (results for 3 detectors shown).
FIG. 11C graphically illustrates pulse-height spectrums produced by an uncollimated 137Cs source on detectors with a common cathode shown in FIG. 10C after correction using depths of interactions (results for 3 detectors shown).
FIG. 11D graphically illustrates examples of the dependencies of the anode's amplitudes versus the drift times of the CdZnTe type detector array shown in FIG. 10C (results for 3 detectors shown).
FIG. 11E graphically illustrates examples of the dependencies of the anode's amplitudes versus the cathode amplitudes of the CdZnTe type detector array shown in FIG. 10C (results for 3 detectors shown).
FIG. 12 graphically illustrates correlations between the anode signals readout from the array of FIG. 10C in which the amplitudes exceed the noise thresholds.
FIG. 13 graphically illustrates improvements of the pulse-height spectra measured from nine detectors of the array of FIG. 10C after applying the charge-loss correction and rejecting ICC events from single-interaction point events. All the spectra are plotted using the same scale to emphasize the detectors' different photopeak efficiencies. The numbers in the top left corners indicate the detectors' ratings that were assigned based on their relative photoefficiency.
FIG. 14 graphically illustrates the pulse-height spectra from nine detectors of the array of FIG. 10C plotted before and after rejecting ICC events.
FIG. 15 graphically illustrates the ICC events-rejection algorithm for the array of FIG. 10C. The first column shows the pulse-height spectra from a 137Cs source after the charge-loss correction for three representative detectors. The second and third columns, respectively, show the spectra containing only accepted and rejected events. The last (fourth) column contains scatter plots of all events in coordinates: Ratio of cathode to anode, r=C/A, versus drift times, t.
FIG. 16 graphically illustrates the distributions as in FIG. 15 but plotted for the low-energy gamma rays from a 133Ba source.
DETAILED DESCRIPTION
The present radiation detector system disclosed herein is directed to effectively solving the electron trapping problem of the prior art detectors by optimizing the configuration of each individual virtual Frisch-grid detector in an array configuration.
As shown in FIG. 2, in one embodiment, the radiation detector system comprises a virtual Frisch-grid detector array 700 made of a plurality of virtual Frisch-grid detectors, an application-specific-integrated-circuit (ASIC) 800 connected to the virtual Frisch-grid detector array 700 for receiving a signal from said array, and a microprocessor 900 connected with the ASIC 800 for operating the ASIC, where signals received by the array 700, passed through the ASIC 800, and processed by the microprocessor 900. The configuration and components of the present radiation detector system are discussed in detail herein below.
Virtual Frisch-Grid Detector Array
It will be appreciated and understood by those skilled in the art that in order to overcome the problem caused by electron trapping, the best practical approach is to use the cathode signal to correct the anode's charge loss due to trapping. However, since the shield in the virtual Frisch-grid detectors may extend from the cathode down to the anode, i.e., full length, or the cathode may be extended further down the length of the semiconductor, the ability to measure cathode signals is substantially reduced. To overcome this limitation, the inventors have discovered that each virtual Frisch-grid detector, or more specifically each shielding electrode of each detector may be optimized individually by varying the length of the shield and the location of the detector within the detector array. In a preferred embodiment of the present invention, the location of the shield is close to the anode. A schematic illustration of a top plan view of a virtual Frisch-grid detector array of such embodiment is shown in FIG. 3. The virtual Frisch-grid detector array 700 comprises one or more sub-array(s) 600 with a varying number of the virtual Frisch-grid detectors within each sub-array. In one non-limiting embodiment, every detector within an array or sub-array is connected to form a common cathode. However, it is also envisioned that each array or sub-array may have two or more subgroups of virtual Frisch-grid detectors that are connected to a common cathode within that particular subgroup. In one embodiment, the virtual Frisch-grid detector array comprises one (1) sub-array. In another embodiment, the virtual Frisch-grid detector array may comprise more than one (1) sub-array, for example, 2, 4, 6, 8, or 9 sub-arrays. It will be appreciated and understood by those skilled in the art that the number of sub-arrays within the virtual Frisch-grid detector array of the present invention will depend on the desired parameters and extrinsic requirements, e.g., the overall size of the radiation detection system, and should not be limited to the embodiments exemplified herein.
The virtual Frisch-grid detector array (or sub-array) comprises a plurality of virtual Frisch-grid detectors that are configured in a matrix A(m,n), where m and n independently represent the number of rows and columns that may range from 1 to 1000 or more. While there is no specific requirement of how the plurality of virtual Frisch-grid detectors must be presented in the array, in one exemplary embodiment, A(4,4) describes a 4×4 matrix 600 with 4 rows and 4 columns as shown in FIG. 3. In another embodiment, A(6,6) describes a 6×6 matrix, which is shown as a cross section view in FIGS. 4A, 4B, and 4C.
In one embodiment, a virtual Frisch-grid detector array 300 comprises a plurality of virtual Frisch-grid detectors in a 6×6 matrix, i.e., 36 detectors. FIG. 4A shows one of the rows 1-6. Each virtual Frisch-grid detector within the array 300 has a shielding electrode 314 near the anode extending a partial length of a semiconductor 312, thus, leaving the semiconductor 312 within the virtual Frisch-grid detector partially unshielded. The actual length of the shielding electrode 314 depends on the aspect ratio of the semiconductor 312. The longer the semiconductor 312, a relatively shorter shielding electrode 314 can be used leaving large fraction of device's volume for the cathode 301. Preferably, the shielding electrode should be at least equal to the width of the detector and independent of the device length. In one particular embodiment, the length of the shielding electrode 314 is about ⅓ of the total length of the semiconductor 312, thus, leaving about ⅔ of the semiconductor 312 unshielded. In another embodiment, the length of the shielding electrode 314 is about ¼ of the total length of the semiconductor 312, thus, leaving about ¾ of the semiconductor 312 unshielded. In yet another embodiment, the length of the shielding electrode 314 ranges between about ¼ and about ⅓ of the total length of the semiconductor 312, thus, leaving about ¾ to about ⅔ of the semiconductor 312 unshielded. The shield 314 may be longer or shorter depending whether the shield 314 is sufficient to provide the shielding of the anode, which can be optimized for a particular configuration of the semiconductor detector. A cathode 311 of each detector in the array 300 is attached to a conductive board (or other conductive materials) 301 via a conductive connector (e.g., bond) 302. The conductive connector 302 may be, for example, an indium bump, a conductive epoxy, a spring connector, or a gold stump. The cathode 301 of each individual detector in the array 300 can either be read together, separately, or in sub-combination of detectors. In one embodiment, the cathode 301 of each individual detector in the array 300 is read together. In such configuration, the interaction depth information derived from the cathode signal of the virtual Frisch-grid detector array is used to correct the anode's charge loss due to electron trapping. Both the ratio between the cathode and anode signals and drift time given by the difference between the starting time of cathode signal and arrival time of the anode signal can be used to implement corrections of the cathode signals.
In another embodiment, a virtual Frisch-grid detector array 400 comprises an internal module 410 and an external (or edge) module 420 configured in a 6×6 matrix. FIG. 4B shows one of the rows 2-5. The internal module 410 comprises a plurality of virtual Frisch-detectors (e.g., 16 detectors for 4×4 internal matrix) with a shielding electrode 414 extending a partial length of a semiconductor 412 near the anode 415, e.g., ¼ to ⅔ of the length of the semiconductor, thus allowing the semiconductor 412 to remain partially unshielded, e.g., ¾ to ⅔ unshielded. The edge module 420, on the other hand, comprises a plurality of virtual Frisch-detectors (e.g., 20 detectors) with a shielding electrode 424 extending further than the shielding electrode 414 of the detector(s) in the internal module 410, but still does not cover the semiconductor 422 completely. In one particular embodiment, if the length of the shielding electrode 414 is about ¼ to about ⅓ of the total length of the semiconductor 412, the length of the shielding electrode 424 of the edge module 420 is anywhere between the length of the shield 414 and the length of the semiconductor 422, e.g., 1>L(424)>¼. One skilled in the art will appreciate that the length of the shield 424 can be optimized for a particular configuration of the semiconductor detector array as long as the shield 424 is sufficient to provide the shielding of the overall array. A cathode 411 and 421 of each detector in the array 400 is attached to a conductive board 401 via a conductive bond 402. Similar to an array described in FIG. 4A, the cathode 411 and 421 of each individual detector in the array 400 can either be read together, separately, or in sub-combination of detectors. In one embodiment, the cathode 411 of each individual detector in the internal module 410 is read together with the cathode 421 of each individual detector in the external module 420 in the array 400.
In a preferred embodiment, a virtual Frisch-grid detector array 500 comprises an internal module 510 and an external (or edge) module 520 configured in a 6×6 matrix. FIG. 4C shows one of the rows 2-5. The internal module 510 comprises a plurality of virtual Frisch-detectors (e.g., 16 detectors for 4×4 internal matrix) with a shielding electrode 514 extending a partial length of a semiconductor 512 near the anode 515, thus allowing the semiconductor 512 to remain partially unshielded. The edge module 520, on the other hand, comprises a plurality of virtual Frisch-detectors (e.g., 20 detectors) with a shielding electrode 524 extending the full length of the semiconductor 512 between the edges of the cathode 521 and the anode 525. A cathode 511 and 521 of each detector in the array 500 is attached to a conductive board 501 via a conductive connector (or bond) 502. A top view of a similar configuration, albeit shown as a 4×4 matrices are presented in FIG. 3, where the internal modules comprise the detectors with partial shield 614 and the external modules comprise the detectors with a full shield 624, and all the detectors are connected to the conductive board 601, e.g., via a conductive connector or bond 602. Similar to an array described in FIG. 4A and FIG. 4B, the cathode 511 and 521 of each individual detector in the array 500 in FIG. 4C can either be read together, separately, or in sub-combination of detectors. In such configuration, the interaction depth information derived from the cathode signal of the internal module in the virtual Frisch-grid detector array is used to correct the anode's signal due to electron trapping. In one embodiment, the cathode 511 of each individual detector in the internal module 510 is read together with the cathode 521 of each individual detector in the external module 520 in the array 500. Specifically, the interaction depth information derived from the cathode signal of the internal module 510 in the virtual Frisch-grid detector array 500 can be used to correct the anode's charge loss due to electron trapping. The correlation between the amplitude of the signals, Aanode and Acathode (see FIGS. 6A-6B), generated by monoenergetic gamma rays and readout from the anode and the cathode in the virtual Frisch-grid detector array produces a negative slope of the narrow-dots continuum corresponding to the total energy deposition events, which is an indication of the electron charge loss that can be corrected as
A
cor
=A
anode
+KA
cathode (1)
Where K is the slope of the correlation curve. Moreover, the ratio between the cathode and anode signals, Acathode/Aanode, can be used to reject the events interacting close to the anode. Such events, represented by the dots located close to the vertical axis, degrade the spectral response of virtual Frisch-grid detector array. This provides a practical approach to use the cathode signal of the internal module to correct the anode's signal loss due to trapping in the whole system. The anode's signal loss due to electron trapping may also be corrected by using the drift time measured as a difference between starting time of the cathode signal and arrival time of the anode signal.
Virtual Frisch-Grid Detector
A schematic of various virtual Frisch-grid detector(s) of the present invention are shown, for example, in FIGS. 5A-5B, to be used in the array 300 shown in FIG. 4A and FIG. 4B or the internal module 510 of the array 500 as shown in FIG. 4C. The virtual Frisch-grid detector 200 (referenced to FIGS. 5A-5B) comprises a compound semiconductor 202 having a first surface (also can be referred to as top) and a second surface (also can be referred to as bottom). An anode 205 is connected to the first side of the semiconductor 202, and a cathode 201 is connected to the second side of the semiconductor 202. In one, non-limiting, embodiment, the cathode 201 and the anode 205 are made from platinum or other conducting material known in the art. The virtual Frisch-grid detector 200 further comprises a shielding electrode 204 surrounding the perimeter of the semiconductor 202 between the anode 205 and the cathode 201, which produces the virtual Frisch-grid effect. The shielding electrode 204 is positioned near the edge of the anode 205 and extends a partial length of the semiconductor 202, never reaching the edge of the cathode 201. Since a signal is induced at the anode 205 by the charge motion between the shielding electrode 204 and the anode 205, the shielding electrode 204 screens out the induced signal from slow moving positive ions drifting towards the cathode 201. Hence, placing the shielding electrode 204 near the anode 205 ensures that the origin of induced signal is from those electrons that drifted from the detector volume into the measurement region, thereby causing the signal to form mainly from electron motion. In one embodiment, the shielding electrode 204 in the detector 200 spans about 5% to about 95% of the semiconductor 202 length positioned at the edge of the anode 205. In more preferred embodiment, the shielding electrode 204 in the detector 200 spans about 15% to about 75% of the semiconductor 202 length positioned at the edge of the anode 205. In even more preferred embodiment, the shielding electrode 204 in the detector 200 spans about 25% to about 35% of the semiconductor 202 length positioned at the edge of the anode 205. We recognized that in the present detectors, the minimum size of the shield may be selected depending only on the device aspect ratio and placed next to the anode. The same shield size can be used for all detectors from the array (or internal module of the array). By contrast, the conventional detector arrays of prior art require different shields for each detector in the array because the conventional detectors need to be carefully optimized to achieve the best energy resolution depending on the properties of the individual crystals, applied bias, geometry, etc.
The virtual Frisch-grid detector 200 further comprises a side insulating layer 203, e.g., the ultra-thin polyester shrink tube, between the surface of the semiconductor 202 and the shielding electrode 204. The non-contacting virtual Frisch-grid detector eliminates grid-to-anode leakage current while still achieving single-carrier performance (see U.S. Pat. No. 6,175,120; incorporated herein by reference in its entirety). The side insulating layer 203 has a thickness adequate to provide excellent electrical isolation and to avoid dielectric breakdown between the anode 205 and any surfaces with a differential bias. In one particular embodiment, the side insulating layer has a thickness of about 10-2000 microns. In one embodiment, the insulating layer is from about 100 microns to about 2000 microns. In another embodiment, the insulating layer is from about 200 microns to about 1000 microns. In yet another embodiment, the insulating layer is about 300 microns. The side insulating layer 203 is made from an insulating material possessing very high electrical resistivity, high threshold for dielectric breakdown, and preferably high corrosion resistance. In one embodiment of the instant invention, the insulating material is machined or casted. The insulating material can be made from a metal oxide (e.g., an alumina, zirconium oxide or quartz), glass-ceramic (e.g., Macor™), polyimide (Kapton™), photoresist, polymers (e.g., polytetrafluoroethylene or Teflon™), plastics, carbide ceramics (e.g., SiC or B4C), nitride ceramics (e.g., Si3N4, AlN, or BN) or a combination thereof.
In an alternative design, shown in FIG. 5B, the semiconductor 202, as in the virtual Frisch-grid detectors defined in FIG. 5A, has a first surface and a second surface. The cathode metallization 201 is extended down to the sides by about 1-30% of the total length of the semiconductor 202, for example, 2-3 mm. The non-contacting shielding electrode 204 covers the device's side surfaces up to the cathode's 201 edge and is kept at the same potential as the anode 205. The insulating layer 203 covers the entire area of the side surfaces. This provides decoupling of the non-contacting ring 204 and the cathode 201 and, at the same time, allows one to apply high differential bias (up to 3000 V and higher) between the cathode 201 and the shielding electrode 204.
The virtual Frisch-grid detectors of the external or edge module, as shown in FIG. 5C and FIG. 5D, may resemble the configuration of the virtual Frisch-grid detectors of the prior art as long as the entire perimeter of the semiconductor 202 is covered by the shielding electrode 204, i.e., 100% from the first side to the second side of the semiconductor 202.
In either configuration of the detector, the semiconductor 202 is selected based on its ability to operate as a single-carrier-transport device. In particular, the compound semiconductors that may be employed in the present invention are generally derived from elements of groups III and V (e.g. GaAs) and groups II and VI (e.g. CdTe) of the periodic table. Besides binary compounds, ternary materials also may be used as the compound semiconductors capable of operating as a single-carrier-transport, e.g., Cd1xZnxTe and Cd1xMnxTe, where 0≦x≦1. It is common practice to omit the fractional subscripts when referring to the alloy families; such practice is followed in describing the present invention. Among these compound semiconductors and their alloys, in one embodiment, cadmium telluride (CdTe), cadmium zinc telluride (CdZnTe), cadmium manganese telluride (CdMnTe), thallium bromide (TlBr), or mercuric iodide (HgI2) is used. However, it will be appreciated and understood by those skilled in the art that any compound or element may be used in the present invention as long as it is capable of operating as single-carrier-transport device, such as in addition to the semiconductors listed above, for example, HgCdTe, HgZnSe, GaAs, PbI2, AlSb, InP, ZnSe, ZnTe, PbO, BiI3, SiC, HgxBr1-x,I2, HgxCd1-xI2, wherein x is greater than 0 and less than 1, InI2, Ga2Se3, Ga2Te3, TlPbI3, Tl4HgI6, Tl3As2Se3, TlGaSe2, and AgGaTe2. In one particular embodiment, a semiconductor 202 is made from a cadmium zinc telluride (CdZnTe) crystal available from eV Microelectronics (373 Saxonburg Blvd., Saxonburg, Pa. 16056). One skilled in the art will appreciate that the semiconductor may be larger or smaller and vary in shape depending upon the design specifications. In one embodiment, the size of the semiconductor 202 is limited by the availability of a single crystal of highly uniform compound semiconductor. In another, non-limiting, embodiment, the aspect ratio of the semiconductor 202 is at least 2. In yet another embodiment, the semiconductor 202 has an aspect ratio of about 2.5 and the size parameters of about 6 mm by about 6 mm by 15 mm thick. In still another embodiment, the semiconductor 202 has an aspect ratio of about 2.8 and the size parameters of about 7 mm by 7 mm by 20 mm thick. In a preferred embodiment, the semiconductor 202 has an aspect ratio of about 2-2.4 and the size parameters of about 5-7 mm by 5-7 mm and 10 to 15 mm thick. The surface of the semiconductor 202 is prepared by surface passivation to provide a lower surface leakage current and stability over time, which can be achieved, for example, by polishing the semiconductor 202 followed by immersion into a suitable chemical. The suitable chemical for surface passivation can be selected from a bromine and methanol solution, hydrogen peroxide solution, ammonium fluoride, ammonium sulfide, HumiSeal™, or a combination thereof. Once, the semiconductor 202 is polished, the surface can also be processed with dry methods, such as coating with AlN or SiN, as a replacement to the immersion in a suitable chemical, or following the chemical immersion treatment.
An embodiment of an integrated detector array is shown in FIG. 8. The integrated detector array includes a circuit board 1000 on which a plurality of virtual Frisch-grid detectors are attached. The virtual Frisch-grid detectors may be arranged and include all the components as described for FIGS. 2-5. Embodiments include at least one ASIC 1010 attached to the circuit board 1000. The ASIC 1010 may be any ASIC as described herein. At least one microprocessor is connected with the ASIC to control the ASIC. The at least one ASIC 1010 and/or circuit board 1000 may be thermally connected to a cooling board 1020. Cooling board 1020 may be any cooling board commonly known in the art. The anode of the virtual Frisch-grid detectors may be attached to the circuit board 1000 via a conductive connector, for example, an indium bump, a conductive epoxy, a spring connector, or a gold stump. A conductive connector 1030 is connected to a common cathode 1040. The common cathode 1040 may be any cathode as described herein. Electrically attached to common cathode is a capacitor 1060. Capacitor 1060 may be directly attached to the common cathode 1040 or attached through a conductive wire 1050. From capacitor 1060 a conductive wire 1070 connects the capacitor 1060 to the circuit board 1000. The conductive wire may pass between virtual Frisch-grid detectors in the array e.g., it may pass in between the corners of four detectors. The conductive wire 1070 may be insulated. Furthermore, whether insulated or not, the conductive wire 1070 may pass through an insulating conduit 1080 disposed between the virtual Frisch-grid detectors in the array, such as between the corners of four detectors. The insulating conduit may be made from a metal oxide (e.g., an alumina, zirconium oxide or quartz), glass-ceramic (e.g., Macor™), polyimide (Kapton™), photoresist, polymers (e.g., polytetrafluoroethylene or Teflon™), plastics, carbide ceramics (e.g., SiC or B4C), nitride ceramics (e.g., Si3N4, AlN, or BN) or a combination thereof. The location of the wire between virtual Frisch-grid detectors in the array, such as between the corners of four detectors, may reduce cross-talk between the common cathode and the anodes, and may reduce disturbance between each detector of the array, thus assuring normal operation of the detectors.
Another embodiment of an integrated detector array is shown in FIG. 9, and shows a 3D view of a 2×2 array (with the fourth Frisch-grid detector not shown for clarity). In addition to the circuit board 1000, the common cathode 1040, the capacitor 1060, the conductive wire 1070, the insulating conduit 1080 shown in FIG. 8, an optional bias resistor 1090 is shown attached to the common cathode 1040. In an embodiment, the detectors, each furnished with two flat spring contacts, are slid inside a holder where they fit tightly, and then gently pressed to the circuit board 1000 using a thin conductive plate (or board) placed on the top of them. This plate also connects together the cathodes of the detectors within the sub-array, and carries the small capacitor 1060 and bias resistor 1090.
While the virtual Frisch-grid detector and a detector array of the present invention have been described in connection with what is presently considered to be the most practical and preferred embodiment, it is to be understood that the invention is not to be limited to the disclosed embodiments, but on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.
EXAMPLES
The examples set forth below also serve to provide further appreciation of the invention but are not meant in any way to restrict the scope of the invention.
Example 1
To overcome the problem caused by electron trapping, the best practical approach is to use the cathode signal to correct the anode's signal loss due to electron trapping. In the design of virtual Frisch-grid detectors of prior art, the shield is extended up to the cathode, which, although enhances the shielding of the semiconductor, reduces the ability to measure cathode signals. Therefore, in the present invention we optimized the virtual Frisch-grid design for the individual CdZnTe bars and the CdZnTe array configuration.
The CdZnTe crystals were purchased from eV-Microelectronics, Inc. (373 Saxonburg Blvd., Saxonburg, Pa. 16056). The crystals were re-shaped into bar detectors with a geometric aspect ratio between approximately 2 and 3. Fabrication of the bar-shaped CZT detectors was conducted as described in Bolotnikov, et al. (Proc. SPIE, vol. 5540, 33-45, 2004, incorporated herein by reference in its entirety). The original crystals were hand polished and etched briefly with a 2% bromine/methanol solution. Electroless gold (Au) contacts were applied only to the ends, thereby forming the anode and cathode contacts. Afterwards, the side surfaces were polished further to reduce side-surface leakage current. To improve the device performance, the side surfaces of some of the samples were treated with a NH4F/H2O2 solution (Wright, et al. Proc. SPIE, vol. 5198, pp. 306-313, 2004, incorporated herein by reference in its entirety). After the bar-shaped crystals were fabricated, their side surfaces were wrapped in Teflon tape followed by Cu tape. During the measurements, the detector under test was placed inside a standard eV-Products device holder (Bolotnikov, 2004) or a custom-made test box.
For those virtual Frisch-grid detectors acting as internal modules of the array, we reduced the shielding electrode length and left ⅔ of the CdZnTe bulk unshielded (see FIG. 4C and FIG. 5A). At the same time, we maintained the current virtual Frisch-grid design for those virtual Frisch-grid CdZnTe detectors at the edges of the array (see FIG. 4C and FIG. 5C), whose long shielding electrode ensures a good shielding efficiency for the whole array.
In this way, the interaction depth information can be achieved by reading the cathode signal of the internal modules, which was used to correct the anode's signal loss due to electron trapping. This is particularly important for long detectors, >10 mm, where electron trapping is significant. FIG. 6A is a graph representation that shows the correlation between the amplitude of the signals, Aanode and Acathode, generated by monoenergetic gamma rays and read out from the anode and the cathode in the 14-mm CdZnTe detector. The negative slope of the narrow-dots continuum corresponding to the total energy deposition events is an indication of the electron charge loss. Such events, represented by the dots located close to the vertical axes, degrade the spectral response of virtual Frisch-grid detectors.
A standard 137Cs radioactive source was used to observe the spectroscopic properties of the detectors. The signals were measured with an eV Products 5092 preamplifier. The data acquisition system included a spectroscopy shaping amplifier, MCA card, digital oscilloscope to store waveforms readout from a charge-sensitive preamplifier, and standard NIM electronics. FIGS. 7A and 7B show two pulse-height spectra measured from a 137Cs source with a 14-mm long Frisch-grid detector before and after applying the interaction depth correction and rejecting the events interacting close to the anode. The data was evaluated by measuring the waveforms from preamplifiers and applying the pulse-shape analysis. The low energy tail seen in the original spectrum (see FIG. 7A) is substantially reduced without losing photopeak counts by selecting the events with an Acathode/Aanode ratio greater than 0.005. As a result, the energy resolution (full width at half-maximum, FWHM) improves from 2.4% to 1.3% at 662 keV for the detector designed based on the present invention as seen in FIG. 7B.
Example 2
The Experiment presented in Example 1 was repeated to demonstrate reproducibility of the present invention. FIG. 6B is a graph representation that shows the reproducibility in the correlation between the amplitude of the signals, Aanode and Acathode generated by monoenergetic gamma rays and read out from the anode and the cathode. A standard 137Cs radioactive source was used to observe the spectroscopic properties of the detectors. FIGS. 7C and 7D show two pulse-height spectra measured from a 137Cs source with a 15-mm long Frisch-grid detector before and after applying the interaction depth correction and rejecting the events interacting close to the anode. The data was evaluated by measuring the waveforms from preamplifiers and applying the pulse-shape analysis. The low energy tail seen in the original spectrum (see FIG. 7C) is substantially reduced without losing of photopeak counts by selecting the events with a ratio greater than 0.005. As a result, the energy resolution (full width at half-maximum, FWHM) improves from 3.4% to 1.4% at 662 keV for the detector designed based on the present invention as seen in FIG. 7D.
Thus, in this novel design, the deleterious effects caused by the electron trapping may be corrected without losing the shielding efficiency. The energy resolution and the detection efficiency improve accordingly.
Example 3
An integrated 3×3 detector array was assembled using CZT detectors as described in Example 1. The 6×6×15 mm3 CZT detectors were encapsulated in an ultra-thin polyester shell (FIG. 10A). The encapsulated detectors were furnished with two spring contacts and a 4-mm-wide shielding electrode (aluminum tape with a thin conducting adhesive layer) (FIG. 10B). The spring contacts cover the entire surfaces of the detector's cathodes and anodes to prevent damaging of the crystals' surfaces when the springs are compressed. This configuration assures the facile replacement of individual detectors, while conferring resistance to any shocks or external impacts. The spring contacts preclude the need to use conductive epoxy to connect the detectors. The detectors were then placed inside the cells of the plastic housing mounted on a fanout substrate, and the common cathode secured on top. As seen in FIG. 10C, a Teflon-insulated wire is used to connect the common cathodes to the sockets on the fanout substrate. The wire passes through the small gap between the corners of four adjacent detectors, and is soldered directly to the capacitor attached to the common cathode.
The application of charge-loss correction techniques for the 3×3 array is shown in FIG. 11 (only 3 detectors are shown): Column (a) shows the raw spectra, column (b) shows the corrected spectra based on the drift times, column (c) shows corrected spectra based on the cathode signals (interaction depths), and, columns (d), and (e) show dependencies of the anode's amplitudes versus the drift times and the cathode amplitudes, respectively. The solid lines represent the fitted curves used to correct the anodes' amplitudes. As seen in FIG. 11, corrections to both the drift time and interaction depth equally can be applied in the cases where the electric field inside the detector is uniform and, hence, the drift time is a linear function of the interaction depth. However, if the electric field is not uniform, it is more appropriate to use the drift time since the carriers' loss is proportional to the electrons' drift time. Alternatively, the drift-time dependence versus the interaction depth must be calibrated before making corrections.
FIG. 12 shows the correlations between the anode signal readouts from any two detectors whose amplitudes exceed noise thresholds. In these plots, the photopeak events follow straight lines as would be expected for multiple interaction events with no charge sharing. For pixel detectors, such lines would have a curvature indicating the signal losses. The Compton edges clearly are evident in these plots.
FIG. 13 demonstrates the improvements of the pulse-height spectra measured from the nine detectors of the 3×3 array after applying the charge-loss correction and rejecting incomplete charge collection (ICC) events from single interaction point events. All the spectra are plotted using the same scale to emphasize the detectors' different photopeak efficiencies. The corrected spectra show good or acceptable responses with energy resolution in the range of 1.0-1.5% FWHM at 662 keV. One detector showed an energy resolution of 1.6% FWHM at 662 keV. The numbers in the top left-hand corners indicate the detectors' ratings that were assigned based on their spectral resolution and relative photo-peak efficiency. All spectra look similar when normalized to the same height of the photopeaks as illustrated in FIG. 14, but the detectors have different photo-peak efficiencies; the numbers in the top left corners indicate the relative number of events under the photopeaks in percentages of the number found for the best detector. The reduction in photopeak events is caused by defects in the crystal that “move” ICC events from a photopeak area into the Compton continuum. Such events can be identified and rejected from the pulse-height spectra, so peak-to-Compton ratios stay the same for all detectors. Thus, the detectors' sensitive volumes have effectively been reduced by vetoing the defective regions inside which otherwise contribute to the continuous background. On the other hand, as seen in FIG. 14, the pulse-height spectra plotted before and after rejection have the same heights meaning the rejection affects ICC events only.
The rejection of ICC events allows substantially every detector to be employed (without compromising an overall array's performance) as long as the detector provides an acceptable energy resolution even though only a fraction of the crystal is actually used. The array's effective area can be maintained by selecting detectors with smaller fractions of defective regions or using a larger number of less expensive and easily available detectors; it is a tradeoff between the cost and effective area.
Referring now to FIG. 15, the first column shows the pulse-height spectra from a 137Cs source after the charge-loss correction for three representative detectors. The second and third columns, respectively, show spectra containing only accepted and rejected events. The last column contains scattered plots of all events in coordinates: Ratio of cathode to anode, r=C/A, versus drift times, t. As can be seen, the majority of the events (dots) concentrate along the curves whose shapes slightly vary from one detector to another. These curves, which may be called the correlation functions, are evaluated by selecting the photoabsorption events that are not affected by the crystal's defects, and fitting their distributions with spline functions (the center solid lines). Two parallel lines represent the ˜100 keV wide acceptance intervals; all events outside these intervals were rejected. As seen, the majority of the rejected events are from the Compton continuum region while the photopeaks remain uncut.
The same correlation functions, r, also can be used to reject the ICC events over the entire energy range starting from very low energies. FIG. 16 shows similar plots as in FIG. 15 but for the events generated by the low-energy photons from the 133Ba source. Again, as previously, results from three detectors are shown. As can be seen, events unaffected by the crystal defects follow the same correlation curves used in the previous case (FIG. 15). Also apparent are the small event losses from the low-energy photopeaks, <100 keV.
It will be appreciated by persons skilled in the art that the present invention is not limited to what has been particularly shown and described. Rather, the scope of the present invention is defined by the claims which follow. It should further be understood that the above description is only representative of illustrative examples of embodiments. For the reader's convenience, the above description has focused on a representative sample of possible embodiments, a sample that teaches the principles of the present invention. Other embodiments may result from a different combination of portions of different embodiments.
The description has not attempted to exhaustively enumerate all possible variations. The alternate embodiments may not have been presented for a specific portion of the invention, and may result from a different combination of described portions, or that other undescribed alternate embodiments may be available for a portion, is not to be considered a disclaimer of those alternate embodiments. It will be appreciated that many of those undescribed embodiments are within the literal scope of the following claims, and others are equivalent. Furthermore, all references, publications, U.S. patents, and U.S. patent application Publications cited throughout this specification are hereby incorporated by reference as if fully set forth in this specification.
Patent Application
20140217297
