Transcript Document

Институт Физики Высоких Энергий
I n s t i t u t e of
High Energy
Physics
Influence of cooling on the working parameters of
GaAs detectors.
S. Golovnia a*, S. Gorokhov a, Y. Tsiupa a, A. Vorobiev a
O. Koretskaja b, L. Okaevich b, O. Tolbanov b
a Institute of High Energy Physics, Protvino, Russia
b Tomsk State University, Tomsk, Russia
The following semiconductor detector parameters is testing to
find the influence of cooling on it.
•
•
•
•
•
•
Detector’s Dark Current vs. Bias Voltage
Detector’s Capacitance vs. Bias Voltage
Detector’s Noise vs. Bias Voltage
Detector’s Response on β-particles and γ-rays
Detector’s Charge Collection Efficiency
Detector’s γ -Ray Detection Efficiency & Thickness
Detectors geometry and types
•
Resistive, doped Cr. Sample dimensions 3,3x3,4 mm, thickness 780 mkm.
initial indot of GaAs is ~ 0,8 – 1,2 * 10 ^17 [1/cm^3], bulk resistivity 0,5 – 0,8 *10^9
•
Epitaxial grown and Cr compensated. Sample dimensions 1,1 x 6,6 mm, thickness 500
mkm initial indot is 1*10^17 [1/cm^3] Concentrations ratio is Cr/Sn = 6.8
The Dark Current temperature dependence vs. High Voltage
The typical I-V characteristic of resistive type detector sample. The characteristic is
independent from the voltage polarity and forward and reverse branches is equal. The dark
current decreases in ten times if the detector temperature is decreasing on 20 °C
The detector capacitance dependence from the detector temperature at different modulation
frequencies
The typical C-V characteristic of resistive type GaAs detector samples. The temperature
changes is not affected on the detector capacitance if the modulation frequency higher then
10 kHz
The detector noise can be predicted in case of 2 stage integration
2


 0.66  C 2 
1    
2kT  

2 2 
q n  

  1  2   eI 
  2kT 
 Tint   C V f 
4
R  

 g m  Tint 

   1  

2
Where:
Shot noise
Current & Thermal noise
Flicker noise
λ = Tdif/Tint ~1
k = 1.380*10-23 J/K
T =~300 temperature (0K)
The contribution from different noise sources can be
easy separated
g = ~20
1.
Shot noise is proportional ~ C2/Tint
C = ~ 10 -12 total capacitance
2.
I and Thermal noise ~Tint
I = ~ 10 -9 total current
3.
Flicker noise does not depend on Tint
Vf = flicker noise voltage
The detector system noise can be predicted in case of 2 stage integrating
2


 0.66  C 2 
1    
2kT  

2 2 
q n  

  1  2   eI 
  2kT 
 Tint   C V f 
4
R  

 g m  Tint 

   1  

2
Shot noise
Where:
Current & Thermal noise
Flicker noise
λ = Tdif/Tint ~1
The contribution from different noise sources can be
easy separated
k = 1.380*10-23 J/K
T =~300 temperature (0K)
g = ~20
1.
Shot noise is proportional ~ C2/Tint
C = ~ 10 -12 total capacitance
2.
I and Thermal noise ~Tint
I = ~ 10 -9 total current
3.
Flicker noise does not depend on Tint
Vf = flicker noise voltage
2
2
d q n 3e
C dV
 Tint 
dI 16
4 dI
2
f
2
 e2 
d qn
3
 1.17 10  Tint  
dI
 nA 
The LED system setup for Noise vs. Current measurements
The schematic view of the test system setup. Both the detector and the LED put into shielded box filled
with dry gas mixture to prevent condensate. The temperatures from -40 to +50 °C can be reached. By
changing the voltage in the LED chain the photocurrent in the test sample can be easily changed.
The dependence of Sigma noise (in ADC channels) vs. detector current at different shaping times
is given for resistive type detector sample.
The dependence of Sigma noise (in ADC channels) vs. detector current is given for resistive type detector
sample.
dSigma/dI for both 50 and 100 high voltages is equal. It is mean, that noise is
put together additive, without influence on each other and noise linked with
high voltage only.
The dSigma/dI [e2/nA] vs. Shaping time is given for resistive type detector samples. The dSigma/dI = 5768,9
[e2/nA] that much more that predicted value of 1171,8 [e 2/nA] calculated before.
The epitaxial detector response on γ-rays from 241Am source
This spectrum from the radioactive source 241Am on the epitaxial type detector show us that Charge
Collection Efficiency is ~ 86-90%.
The epitaxial detector response on beta particles from 90Sr radioactive source
Charge collection efficiency vs. high voltage at different temperatures measured on
Diffusion Samples
Charge collection efficiency vs. temperature at different high voltage measured on Epitaxial Samples
Detection Efficiency vs. GaAs Detector thickness for γ-Ram Energy Region from 20 to 60 keV
The epitaxial detector active area thickness vs. detector temperature for high voltages 50 & 100 V
The resistive detector active area thickness vs. high voltage at temperature +20 oC
Conclusion
•
The presented experimental results show, that cooling both GaAs resistive & epitaxial detectors can
increase their working parameters
•
The high charge collection efficiency of epitaxial detectors together with their active area thickness
up to 100 mkm make us possible to us them as a detector for X-Ray imaging application
•
The resistive type detectors have the ability to operate with both polarity of high voltage have been
presented