Master Thesis

Studies of APD readout of BGO crystal scintillators for the ASTRO-H mission

Hiroyuki Nishioka University of Tokyo Supervisor: Professor Kazuo Makishima January 2011

Abstract In the present thesis, we studied light collection efficiency of BGO crystal scintillators read out via APDs for the ASTRO-H mission. To achieve a lower background of the Hard X-ray Imager (HXI) and the Soft Gamma-ray Detector (SGD) onboard ASTRO-H, an active shield, which consists of BGO crystal scintillators and APDs, is the most important system. Due to a high refractive index of BGO (∼2.15) and a small effective area of an APD (∼10 × 10 mm2 ), many scintillation photons are trapped in a crystal. To obtain sufficient light output, we studied the effects of changing a geometry of a crystal, reflector surrounding a crystal, and surface finish of a crystal. Especially, as to the geometrical effects, we succeeded in finding parameters which have a strong correlation with the light output, when a rectangular BGO crystal is read out by a 10 × 10 mm2 APD. In addition, we designed an APD shield assembly for ASTRO-H, and checked its performance. The results in the present thesis are fed back to the actual design of the active shield for ASTRO-H.

Contents Abstract

i

List of figures

vi

List of tables

vii

1 INTRODUCTION 1.1 The ASTRO-H Mission . . . . . . . . . . 1.1.1 Overview . . . . . . . . . . . . . 1.1.2 The Hard X-ray Imager (HXI) . . 1.1.3 Soft Gamma-ray Detector (SGD) 1.2 Active Shields of the HXI and the SGD . 1.3 BGO Crystal Scintillators . . . . . . . . 1.3.1 Crystal scintillators . . . . . . . . 1.3.2 BGO crystals . . . . . . . . . . . 1.4 Scinllator Readout via APD . . . . . . . 1.4.1 Comparison with PMT . . . . . . 1.4.2 Dependence on temperature . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

2 PLAN OF THE PRESENT RESEARCH 3 BASIC EXPERIMENTS 3.1 Experimental Setup . . . . . . . . . . . . 3.1.1 Experimental equipments . . . . 3.1.2 Measurement procedure . . . . . 3.2 Preliminary Experiments . . . . . . . . . 3.2.1 Evaluation of pulse heights . . . . 3.2.2 Reproducibility of measurements 3.2.3 Dependence on shaping time . . . 3.2.4 Dependence on temperature . . .

10

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

4 STUDIES OF LIGHT COLLECTION EFFICIENCY 4.1 Studies of BGO Shapes . . . . . . . . . . . . . . . . . . . . . 4.1.1 BGO crystal samples . . . . . . . . . . . . . . . . . . 4.1.2 Shape effects in rectangular crystals . . . . . . . . . . 4.1.3 Shape effects in non-rectangular crystals with tapers 4.2 Reflector Surrounding Crystals . . . . . . . . . . . . . . . . i

1 1 1 3 3 4 6 6 6 6 6 9

. . . . . . . .

. . . . .

. . . . . . . .

. . . . .

. . . . . . . .

. . . . .

. . . . . . . .

12 12 12 13 14 14 15 16 17

. . . . .

20 20 20 22 27 32

4.3

4.4

4.2.1 Candidates of reflector . . . . . . . . . . . 4.2.2 Measurements with different reflectors . . Surface Finish of Crystals . . . . . . . . . . . . . 4.3.1 Measurements with different surface finish 4.3.2 Interpretation of the results . . . . . . . . Optical gluing of crystals . . . . . . . . . . . . . . 4.4.1 Motivation . . . . . . . . . . . . . . . . . . 4.4.2 Gluing of rectangular crystals . . . . . . . 4.4.3 Gluing of non-rectangular crystals . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

5 DESIGN OF APD NOISE SHIELD 5.1 Selection of Material and Surface Finish . . . . . . . . . 5.1.1 Selection of material . . . . . . . . . . . . . . . . 5.1.2 Effects of aluminum surface finish on light output 5.1.3 How to cover an APD . . . . . . . . . . . . . . . 5.2 The APD Assembly for ASTRO-H . . . . . . . . . . . . 5.2.1 Actual design of the APD assembly . . . . . . . . 5.2.2 Thermal cycle test . . . . . . . . . . . . . . . . . 5.2.3 Adhesion strength . . . . . . . . . . . . . . . . . 5.2.4 Light output and shielding effect . . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

32 33 35 35 37 37 37 38 43

. . . . . . . . .

46 46 46 47 49 50 50 52 53 55

6 SUMMARY

57

7 APPENDIX 7.1 Monte Carlo Simulation Using GEANT4 . 7.1.1 Propagation of scintillation photons 7.1.2 Verification by experiments . . . . 7.2 How to Make BaSO4 Reflector . . . . . . . 7.3 Spectra of All Crystals . . . . . . . . . . .

59 59 59 60 61 62

Acknowledgments

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

64

ii

List of Figures 1.1 1.2

1.3 1.4 1.5

Configuration of the ASTRO-H satellite [1]. . . . . . . . . . . . . . Structures of the active shields of the HXI (top) and the SGD (bottom). Crystals are classified into 3 groups, top section (blue), side section (green and brown) and bottom section (red). . . . . . . . . Temperature dependence of the light output of various crystals [6]. . Energy dependence of the detector energy resolution. . . . . . . . . V-I curves (top) and gain curves (bottom) of an APD at (a)+20◦ C and (b)–20◦ C. In the top panel, green shows dark current, while red shows current when irradiated with a green Light-Emitting Diode (LED). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

5 7 8

9

2.1

A block diagram of the development of active shields for the HXI and the SGD. Numbers correspond to the sections in the present thesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

3.1 3.2

Experiment setup of basic measurements. . . . . . . . . . . . . . . . Configuration of electrical ground employed in the present experiment. Room ♯1 and ♯2 are different grounds. . . . . . . . . . . . . . A spectrum of Wide A crystal acquired at +20◦ C using a setup of Figure 3.1, by irradiating 137 Cs and feeding the preamplifier with test pulses. The shaping time is 0.5 µs. . . . . . . . . . . . . . . . . The dependence of spectra on the shaping time. These are from Wide A at (top) +20◦ C or (bottom) –20◦ C. . . . . . . . . . . . . . The peak channel and energy resolution of 137 Cs, and the peak channel and FWHM of test pulse, measured at (a)+20◦ C and (b)– 20◦ C. The photopeak of 137 Cs was not detected at shaping times of 6 µs or 10 µs at +20◦ C. . . . . . . . . . . . . . . . . . . . . . . . . The dependence on shaping time of electrical noise. . . . . . . . . . Comparison of spectra of Wide B, taken at +20◦ C (red) and –20◦ C (blue). See text for details of Wide B. The shaping times were optimized as 0.5 µs, 2.0 µs at ±20◦ C. . . . . . . . . . . . . . . . . .

3.3

3.4 3.5

3.6 3.7

4.1

13 14

15 17

18 18

19

BGO crystal samples used in the present thesis. Units are in mm. Orange are rectangular shaped objects, while blue are non-rectangular ones. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

iii

4.2

4.3

4.4

4.5 4.6

4.7 4.8 4.9

4.10

4.11 4.12 4.13 4.14 4.15 4.16

4.17

Spectra of various crystals. (Top) Spectra of 1 cm cube (♯1, black), 2.5 cm cube (♯2, blue) and a 40×66×66 mm3 one (♯5, red). (Bottom) Spectra of a 20×72×180 mm3 crystal (♯7), read out from three different sides. In both panels, the peak at ∼3,800 ch is due to the test pulse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Light output of the rectangular crystals, plotted against the surface area where the APD was attached. Red, green and blue circles are when the depth of the crystal as seen from the APD side is < 55 mm, 55 ∼100 mm, and >100 mm, respectively. . . . . . . . . . . . Same as Figure 4.3, but plotted against (top) the depth as seen from the APD, (middle) the volume, and (bottom) the total surface area of crystals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Definition of coordinates of scintillation crystals. Effects of changing each scale are illustrated at bottom. . . . . . . . . . . . . . . . . . . (Top) Light output of the rectangular crystals, plotted against Z×( X1 + 1 ) in different XY intervals. (Bottom) The same as top panel, but Y the light output is normalized by (XY )0.28 . . . . . . . . . . . . . . . Spectra of ♯7, ♯9, ♯8 and ♯10, showing effects of non-rectangular shapes. The APD was put on the smallest surface of each crystal. . Setup of the collimation experiment using ♯14. . . . . . . . . . . . Spectra of ♯14, irradiating with collimated 137 Cs . The APD is put at (top) the wider end and (bottom) the narrower end. Collimation positions of 1,2,3 and 4 in Figure 4.8are indicated by black, red, green and blue, respectively. . . . . . . . . . . . . . . . . . . . . . . The simulated results of the collimate experiment when the reflector has reflectivity of (top) 0.98 and (bottom) 1.00. Red shows the collection efficiency when the APD is put at the wider end, while blue at the narrower end. . . . . . . . . . . . . . . . . . . . . . . . Comparison of the collimated experiment (Figure 4.9) with the simulation (Figure 4.10; reflectivity=1.00). . . . . . . . . . . . . . . . . A typical BGO enclosure to be employed in the HXI and the SGD. Six configurations of the reflectors tested in the present thesis. Red box indicates the actual design for ASTRO-H. . . . . . . . . . . . . The light output of ♯7 with different reflectors. . . . . . . . . . . . Photographs of ♯8, when it is roughened only at the APD-fixing 8 × 8 mm2 window. . . . . . . . . . . . . . . . . . . . . . . . . . . . . The difference of light output by surface finish. (Top) Spectra of ♯8. The employed surface finish is polished (black), rough at opposite side of readout (red), rough at readout side (green), and rough at the APD-fixing window of 8×8 mm2 (blue). (Bottom) Spectra of ♯7’. All the 6 surfaces are polished (blue) or roughened (red). . . . . Effects of the surface finish on the light output of a BGO crystal which has similar geometry to ♯14, irradiated with 137 Cs at a position of Col. 4 in Figure 4.8. The abrasive material used in surface polish is chrome oxide (blue) and diamond slurry (red). . . . . . . .

iv

22

24

25 26

27 28 29

30

31 31 33 33 34 36

36

38

4.18 (Left) Experimental setup of separate readouts from two BGO pieces, glued together with optical grease (OKEN 6262A). (Right) An example of scintillation photons’ propagation. . . . . . . . . . . . . . 4.19 Setup of signal readout from the two PMTs in Figure 4.18. . . . . 4.20 Spectra of each crystal. . . . . . . . . . . . . . . . . . . . . . . . . 4.21 Coincidence of events. Events which occurred within 5µs are plotted. 4.22 A photograph of an edge-cut. . . . . . . . . . . . . . . . . . . . . . 4.23 Effects of the edge-cut. (Left) Results of Monte-Carlo simulations in which the edge-cut width is changed. (Right) Result of the experiment similar to Figure 4.21, after covering the edge-cut part with a black tape. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.24 Trapezoidal crystal samples. Filled red circles indicate the collimated irradiation position of 137 Cs, while yellow squares show APDs. 4.25 (Top) A spectrum of ♯9’ before division, with the irradiation position near the readout side. (Bottom) Spectra of ♯10, read out from the irradiated piece (bottom left), and from the other piece (bottom right). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.26 Spectra of ♯11 when 137 Cs is irradiated onto the read-out piece (top) and the other piece (bottom). Blue is before division and red is after. 5.1 5.2

5.3

5.4

5.5

5.6

5.7 5.8

Possible configurations of the electrical noise shield. . . . . . . . . . Spectral reflectivity of typical metals ; Ag (green), Al (red), Rh (cyan), Cu (Purple) and Au (blue). Blue region illustrates the emission spectrum of BGO. . . . . . . . . . . . . . . . . . . . . . . . . . Spectra of 137 Cs and test pulse. Red was obtained without electrical shield, while black and blue with aluminum shield (over the APD and BGO) with a thickness of 1 mm and 0.1 mm, respectively. . . . Spectra of 137 Cs and test pulse obtained by changing composition and surface finish of aluminum ; polished pure aluminum (red), rough pure aluminum (green), polished ♯5000 aluminum (blue), and roughened ♯5000 aluminum (yellow). The aluminum in all cases has an area of 20 × 72 mm2 . For comparison, black shows a result when all surface are covered with an ESR sheet. . . . . . . . . . . . . . . Photographs of the aluminum shield box for the APD. (Left) The aluminum box of 1 mm thickness and aluminum cover of 100 µm. The aluminum cover is electrically connected to the box by a φ 3 mm screw. (Right) BGO crystal, on which the APD shield box assembly (yellow) is held by fabric tape. . . . . . . . . . . . . . . . Spectra of ♯7 measured at –20◦ C. Black is when the APD is electrically unshielded, blue is when the APD is placed inside the aluminum shield box, and red is when an additional ESR sheet is inserted (see text). The power of thermostatic chamber was turned off when obtaining the spectra. The reflectors are an ESR and GORE-TEX sheets. . . . . . . . . . . . . . . . . . . . . . . . . . . . An exploded view of the APD shield box. . . . . . . . . . . . . . . . Temperature profile of a thermal test cycle. . . . . . . . . . . . . . . v

39 40 41 41 42

42 43

44 45 47

47

48

49

50

51 51 52

5.9

Photographs of the test piece after making 6 cycles. (Top left) Overview of the BGO crystal. (Top right) Glue with EPOTEK 301–2. (Bottom left) Glue a polished aluminum sheet with DC 93– 500. (Bottom right) Glue a rough aluminum sheet with DC 93–500. 5.10 Setup of the adhesion strength test. . . . . . . . . . . . . . . . . . . 5.11 Photographs of the adhesion strength test. . . . . . . . . . . . . . . 5.12 Spectra of ♯15 with or without the APD shield box, irradiated at the Col.1 and Col. 4 position of Figure 4.8. When the APD shield box is not used, the crystal and the APD were covered with aluminum chassis of 1 mm thickness. . . . . . . . . . . . . . . . . . . . . . . . 7.1 7.2

7.3

7.4

Boundary condition at a surface of a scintillator crystal. . . . . . . (Top left) Crash BaSO4 powder, using a pestle. (Top right) Paint the BaSO4 solution on a surface of a BGO crystal. (Bottom) Allow the solution to air dry on the crystal. . . . . . . . . . . . . . . . . . Spectra of rectangular crystals (♯1–♯8), read out from different side, and acquired at +20◦ C using a setup of Figure 3.1, by irradiated with 137 Cs and feeding the preamplifier with test pulses. . . . . . . Spectra of non-rectangular crystals (♯9–♯15), acquired at +20◦ C using a setup of Figure 3.1, by irradiated with a collimated 137 Cs source at positions as shown in Figure 4.8. In addition to +20◦ C, ♯14 (bottom right) was measured at –20◦ C, and the reflectors are BaSO4 paint on the one surface which has the largest area, and the others are ESR+GORE-TEX sheets. . . . . . . . . . . . . . . . . .

vi

53 54 55

56 60

61

62

63

List of Tables 1.1 1.2 1.3

Specification of instruments. . . . . . . . . . . . . . . . . . . . . . . Properties of commonly-used inorganic scintillators. . . . . . . . . . Comparison of APDs with PMTs. . . . . . . . . . . . . . . . . . . .

3.1 3.2

Typical equipments and experimental conditions at +20◦ C/–20◦ C. . 13 Results of spectra of the Wide B crystal at ±20◦ C. . . . . . . . . . 19

4.1 4.2 4.3

Sizes and shapes of BGO crystal samples. . . . . . . . . . . . . . . Light output, energy resolution and LD of rectangular crystals. . . . Light output, energy resolution and LD of non-rectangular crystals. ♯14 and ♯15 have dependence on collimation position, the value are at Col. 4 as shown in Figure 4.8. . . . . . . . . . . . . . . . . . . . Candidates of reflector. . . . . . . . . . . . . . . . . . . . . . . . . . Light output of ♯7 measured with different reflectors shown in Figure 4.13. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results of Monte-Carlo simulation. 100,000 photons were produced in the wide crystal. . . . . . . . . . . . . . . . . . . . . . . . . . . . Equipments and conditions of the experiment using 2 PMTs. The values are for ♯7/♯8. . . . . . . . . . . . . . . . . . . . . . . . . . . . Results of ♯10 and ♯11. ADC channels of ♯10 were converted from Spacewire ADC channels to Pocket MCA ones. . . . . . . . . . . .

4.4 4.5 4.6 4.7 4.8 5.1 5.2 5.3 5.4

Parameters of the spectra given in Figure 5.3, showing the effect of APD shield. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results of the light output measurements with and without aluminum shield. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The property of EPOTEK 301–2 and DC 93–500. . . . . . . . . . . Results of light output and test pulse FWHM using the ♯15 crystal with or without the APD assembly. When without an APD assembly, the crystal was put in an aluminum chassis of 1 mm thickness.

vii

2 7 7

21 23

28 32 35 39 40 45 48 50 52

56

Chapter 1 INTRODUCTION 1.1 1.1.1

The ASTRO-H Mission Overview

The ASTRO-H mission is the next astronomy X-ray mission led by the Institute of Space and Astronautical Science (ISAS) [1]. To be launched in 2014 covers a very wide energy range from 0.3 keV to 600 keV. Figure 1.1 shows the configuration of instruments onboard the satellite. ASTRO-H will carry two units of the Hard Xray Telescope (HXT) and two units of the Soft X-ray Telescope (SXT). The HXT is coupled to the Hard X-ray Imager (HXI). One of the SXT is focused on the Soft X-ray Spectrometer (SXS) and the other on the Soft X-ray Imager (SXI). In addition to these instruments, ASTRO-H will carry the Soft Gamma-ray Detector (SGD) to cover the soft γ-ray region up to 600 keV. The basic parameters of these instruments are summarized in Table 1.1.

Figure 1.1: Configuration of the ASTRO-H satellite [1].

1

Hard X-ray Imaging System (HXT+HXI) Focal Length Effective Area Energy Range Angular Resolution Effective FOV Energy Resolution Timing Resolution Detector Background Operating Temperature Soft X-ray Spectrometer System (SXT–S+XCS) Focal Length Effective Area

12 m 300 cm2 (at 30 keV) 5–80 keV