Silicon Carbide Solid-State Photomultiplier for UV Light Detection

Transcription

1 Silicon Carbide Solid-State Photomultiplier for UV Light Detection Sergei Dolinsky, Stanislav Soloviev, Peter Sandvik, and Sabarni Palit GE Global Research 1

2 Why Solid-State? PMTs are sensitive to magnetic fields, have low quantum efficiency, are bulky and expensive. high voltage power supply and very short lifetime at elevated temperatures Why UV? flame detection, biological and chemical detection, detection of jet engines and missile plumes Bio-aerosol detection Micro Flash Ladar for navigation Deep-UV Imaging Harsh-Environment UV and Gamma Detectors Perkin Elmer Channel MP- series photomultiplier module Hamamatsu Multiple Pixel Photon Counting (MPPC) array parallel connection of individual GM-APD detectors comprising the array 2

3 Design of SiC SSPM Why SiC? Dark count rate in Si-PM increases rapidly with temperature, resulting in a maximum operating temperature below 50 o C SiC has larger bandgap (3.26 ev) - Lower leakage current - Higher operating Temperature - Higher sensitivity in UV spectra p( T1) p2( T1) p3( T1) probability of thermally produce electronhole pairs in perfect crystal. Theory Si Al x Ga 1-x As T1 273K SiC 3

4 Design of SiC SSPM dielectric Schematic cross section of individual pixel Cathode Quenching resistor interconnect Absorption n-layer P-layer Anode n-4h-sic substrate 2-D distribution of electric filed at avalanche breakdown voltage SEM images of fabricated SiC SSPM dies 4

5 Current, A Characterization of SiC SSPM Quantum efficiency, % Current, A Packaged SiC SSPM 1x10-3 1x10-4 1x10-5 1x10-6 1x10-7 1x10-8 1x10-9 1x Dark I-V curve at room temperature Area: 1x1 mm 2 Active area: 4x4 mm 2 Pixel size: 60 um 16 sub arrays Area of sub-array: 1x1 mm Bias, V o C 100 o C 200 o C Wavelength, nm Dark I-V curves vs. temperature at avalanche breakdown x10-4 1x Quantum efficiency spectra 200 o C Bias, V 20 o C 5

6 Breakdown voltage vs. Temperature Breakdown voltage, V 1E-3 1E-4 1E-5 1E-6 Single pixel, w/o R quench 25 C 100 C 200 C 250 C 300 C B Linear Fit of Data2_B 1E Current density, A/cm 2 1E-8 1E k=6.15e-2 V/C 1E E Voltage, V Temperature, o C Breakdown voltage changes with temperature 62 mv/ o C 6

7 Block diagram of setup for optical measurements Current, ua Output signal, mv Pulsed UV LED PLS 300 UV light, 300 nm PicoQuant PDL 880-D Phillips Scientific 6955 Trigger PicoQuant LED 300 nm (10 nm FWHM), <0.5 ns, 0.25 pj per pulse Oscilloscope LeCroy 1 GHz, 40 Gs/sec Bias Keithley 387 Oscilloscope SiC SSPM I-V curves of dark and photocurrents SiC SSPM with 256 pixels (1 mm 2 ) Dark Current Total Current (~40 ph) Net Photocurrent Output signal vs. bias voltage Voltage, V Bias Voltage, V 7

8 Waveforms of output signal at room temperature Fast component 5 us τ=r q C slow component 2 ns Slow component ( ~3 us) in the waveform depends on a value of quenching resistor 8

9 Impact of temperature on signal waveform Resistance, Ω 1,E+06 1,E+05 Resistance of poly-si quench resistors vs. temperature fast component in the waveform of the output signal becomes negligible with temperature increase. The time constant at 200 o C decreases to 60 ns, while peak amplitude increases up to 0.25V 1,E+04 1,E+03 quenching resistance dropped by factor of ~50 at 200 o C Temperature, o C 9

10 Single Photon Detection Oscilloscope snapshot take at room temperature 284V RT Dark counts 2 photons 2 photons 1 photon 1 photon Single Photoelectron spectrum recorded for SiC-PM with 256 pixels (1 mm 2 ) The histograms suggest discrete nature of SiC SSPM output signal when illuminated by very low level light flux 10

11 Counts PDE, % DCR, MHz Single Photon Detection Efficiency Measurements PDE = N fired N ph = N fired hν f P opt A SiC PM N fired is the average number of triggered pixels, hν is the photon energy, f is the pulse repetition rate, P opt is the optical power density, A SiC-PM is the area od SiC-PM Single Photoelectron spectrum recorded for SiC-PM with 256 pixels (1 mm 2 ) 0 ph.e. Photodetection efficiency and dark count rate as functions of voltage bias ph.e. 2 ph.e Output charge, a.u. Each peak corresponds to a certain number of photoelectrons (ph.e) Bias, V PDE increases linearly from 7 to 9% within the measured voltage range, while DCR slightly increases up to 290V and significantly grows up from ~0.4MHz/mm 2 at 290V to 2MHz/mm 2 at 296V 0 11

12 UV scintillators for SIC SSPM 12 Gamma Sensor Project

13 Quantum efficiency, % UV scintillators for SIC SSPM 0,15 0,125 0,1 0,075 0,05 0,025 0 Pr:LuAG (Furukawa) nm Quantum efficiency spectra 20 o C 100 o C 200 o C Wavelength, nm 13

14 Testing SiC SSPM with scintillator crystal (LuYAG) Gamma Source Crystal LuYAG(Pr) Keithley 387 Bias 285V SiC PM Oscilloscope Active area of SiC SSPM 2x2 mm 2 heater Gamma Source SiC SSPM Crystal LuYAG(Pr) 14

15 Output signal waveform at different temperatures Room Temp, Co 60 2 mv/div, 10 usec/div T=170 o C, Co mv/div, 500 ns/div T=70 o C, Co 60 5mV/div, 5 usec/div T=200 o C, Co mv/div, 50 ns/div SiC SSPM with LuYAG crystal demonstrated a strong response from Gamma source at 200 o C 15

16 Summary Silicon Carbide Solid-State Photomultiplier was demonstrated for the first time. Photon detection efficiency of the SiC-PM measured at 300 nm was about 8%, while a dark count rate was about 0.3MHz/mm 2 at room temperature. Time constant and peak amplitude of output signal significantly dependent on temperature, the time constant decreases from 3 us to 60 ns, while the peak amplitude increases in ~ 25 times with a temperature increase from 20 o C to 200 o C. SiC SSPM works with UV scintillators up to 200 o C 16