The Influence of Edge Effects on the Detection Properties of Detector Grade Cadmium Telluride

Size: px

Start display at page:

Download "The Influence of Edge Effects on the Detection Properties of Detector Grade Cadmium Telluride"

Randolf Ryan
6 years ago
Views:

1 The Influence of Edge Effects on the Detection Properties of Detector Grade Cadmium Telluride M.J. Bosma a, M.G. van Beuzekom a, S. Vähänen b, J.Visser a a. National Institute for Subatomic Physics, Nikhef, Amsterdam b. VTT Technical Research Center of Finland, Helsinki IEEE RTSD R18 3 October 28, 2011, Valencia, Spain

2 Tomorrow s digital radiography Medipix3 Back contact X ray photon Semiconductor sensor Sensor pixel Solder bump Pixel readout pad Medipix readout chip research group Medipix3 features ~1600 transistors per pixel: Charge summing circuitry Simultaneous counting and read-out no dead time Energy dispersion mode: 7 energy bands 2

3 Active detector area x cm 35 x 40 cm 2 3 inch 2.8 cm Medipix quad module with Relaxd read-out 3

4 Tile ability: Sensor edges Edge effects: Charge generalon Surface currents High field regions ConvenLonally solved by guard rings Slim edge AcLve edge Rossi, Pixel Detectors 4

Simulations: transient signals at edges ] -1 Weighting potential [cm 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.

5 Simulations: transient signals at edges ] -1 Weighting potential [cm Edge pixel Centre pixel Depth [µm] Total current [na] Edge pixel Centre pixel Time [ns] 5

6 Detector: 2 slim edge CdTe pieces Detector readout: Medipix-MXR Sensor specs (Acrorad): Quasi-Ohmic (Pt electrodes) 4.05 x 4.05 x 1 mm 3 36 x 36 pixels of (110 μm) 2 65 μm pixel-to-edge distance 6

7 Measurements Comparison between centre and edge pixels Leakage current Laser setup: Pixel response function Charge collection efficiency X-ray setup: Noise power spectrum 7

8 Dark current transients The leakage current as a funclon of Lme aver voltage stepping ( 200V 200V 200V). The current was monitored for one minute per voltage step of 40V Leakage current [A] V 0 V - 40 V - 80 V V V V V - 40 V - 80 V V V Leakage current [A] V 0 V 120 V 80 V 200 V 160 V Time [s] Time [s] Possibly caused by deep level defects 8

9 Dark current transients Flood field image ( 400V bias) Voltage stepping from 0V to 400V 9

10 Bare sensors: I V single pixels At edge and corner pixels higher leakage current. Probably caused by: More surface current More generalon current due to edge imperfeclons Leakage current [A] !10-9 Corner pixel Edge pixel Centre pixel In general, more forward than reverse current Voltage [V] 10

11 Energy calibration Count rate 4 10 Tube voltage 15 kv 20 kv 25 kv Micro focus X ray tube, 2.5 mm. Al filter. Used tail end of spectrum as a reference kv 35 kv 40 kv 45 kv 50 kv LeV: Right: E = * THL E = * THL Lower threshold DAC value LeV sensor Right sensor Lower threshold DAC value Lower threshold DAC value p ± 2.23 p ± p ± p ± Energy [kev] Energy [kev] 11

1 μm steps Sensor s physical edge Count rate [10 Hz] 1000 800 600-140V 400-170V - 200V

12 Laser data: pixel response function Threshold at half the energy equivalent intensity ensures minimal charge sharing Min. 1 μm steps Sensor s physical edge Count rate [10 Hz] V V - 200V V - 500V Count rate [10 Hz] Laser position [µm] Laser position [µm] 12

13 Laser data: charge collection efficiency Fit function described by Hecht relation: Q(U) = Q 0 µτ/l 2 (U U 0 ) [1 exp(dl/µτ(u U 0 )] Charge collection efficiency Centre pixel: (μτ) e = 1.25 E 4 Edge pixel: (μτ) e = 1.15 E 4 Centre pixel Edge pixel Bias voltage [V] 13

14 X ray data: noise power spectrum NPS tends to be flaiened due to under sampling (110 μm pixels) Low pass filtering at low threshold levels More noise at close to photopeak threshold levels Differential count rate Noise power [a.u.] Centre pixels Lower threshold level 8 kev 12.5 kev 19 kev 8 kev averaged 12.5 kev averaged 19 kev averaged Noise power [a.u.] Edge pixels Energy [kev] Lower threshold level 8 kev 12.5 kev 19 kev 8 kev averaged 12.5 kev averaged 19 kev averaged Frequency [lp/mm] Frequency [lp/mm] 14

15 Summary and outlook Summary: Dark current transients after voltage stepping Leakage current at edge and corner pixels larger than that of centre pixels Bias dependence of edge-pixel response function μτ-products of edge and centre pixels comparable Noise power at edge higher and low-frequency dominated Outlook: Laser-induced transient analysis Energy resolution measurements Determine DQE at edge (MTF from pixel response function) 15

16 Thank you High energy nuclear interaclons with Schoiky CdTe. MulLpixel spread splashes, possibly due to polarisalon or a cascade of fluorescence effects (?) Bias voltage and interaclon depth dependence under study. 16

17 BACK UP slides 17

18 Detector: I t characteristics (GaAs) The leakage current as a funclon of Lme aver voltage stepping ( 200V 200V 200V). The current was monitored for one minute per voltage step of 40V Leakage current [A] V V - 80 V V 0 V - 40 V 40 V 80 V 120 V Time [s] 18

19 X ray data: energy resolution Sekngs: 90 kv, 89 μa, 2.5 mm Al filtering, 45 cm distance, 400V sensor bias, 10 Lmes 1s acquisilon Lme per THL step Equal peak posilon, FWHM and peak to background ralo. K edges slightly less pronounced at edge pixels probably due to limited stalslcs (less pixels). Centre pixels Edge pixels Differential count rate Left sensor Filter Charge sharing Cd K edge Te K edge Right sensor Differential count rate 5000 Left sensor Right sensor Cd K edge Te K edge Energy [kev] Energy [kev] 19 0

Marten Bosma 1, Alex Fauler 2, Michael Fiederle 2 en Jan Visser Nikhef, Amsterdam, The Netherlands 2. FMF, Freiburg, Germany

Marten Bosma 1, Alex Fauler 2, Michael Fiederle 2 en Jan Visser 1 1. Nikhef, Amsterdam, The Netherlands 2. FMF, Freiburg, Germany Digital Screen film Digital radiography advantages: Larger dynamic range