New approach to simulate radiation damage to single

Report
New approach to simulate radiation damage to singlecrystal diamonds with SILVACO TCAD
Florian Kassel, Moritz Guthoff, Anne Dabrowski, Wim de Boer
CERN
Institut für Experimentelle Kernphysik (IEKP), KIT
KIT – Universität des Landes Baden-Württemberg und
nationales Forschungszentrum in der Helmholtz-Gemeinschaft
www.kit.edu
Outline
2
•
Introduction to SILVACO TCAD
•
Simulation of a diamond sensor
•
Benchmarking of the simulation
•
Approach to simulate the polarization effect of the diamond
•
Conclusion
19.11.2014
ADAMAS – 3rd Collaboration Meeting, Trento
Florian Kassel
Introduction to SILVACO TCAD
SILVACO TCAD:
Tool to simulate the electrical, optical and thermal
behavior of semiconductor devices
3
•
Is used in CMS silicon strip community to simulate
electrical properties of the sensors
•
Individual sensor designs of arbitrary materials possible
•
Simulation can be done in 2D, quasi-3D and 3D
•
Possibility to include a large selection of physical laws
19.11.2014
ADAMAS – 3rd Collaboration Meeting, Trento
Florian Kassel
Introduction to SILVACO TCAD
Advantages of SILVACO TCAD in understanding radiation
damage
•
Electrical field distribution is directly linked to the detector
efficiency
•
Understanding of electrical field distribution is crucial!
•
Electrical field properties depend on:
 Drift behavior of charge carriers
 Properties of traps in diamond bulk
 Creation of space charges
can be simulated
with SILVACO TCAD
 etc, …
Real measurement of electrical field possible with Transient
Current Technique (TCT) measurement!
Exact simulation of TCT pulse
necessary !
4
19.11.2014
ADAMAS – 3rd Collaboration Meeting, Trento
Florian Kassel
Simulation of an alpha particle hit (TCT)

Quasi
-3D
2D
3D
y
y
r
z
y
φ
x
 Real charge amount
 Overestimated
charge density
 z extension set to
fixed value
5
19.11.2014
z
x
 Charge amount and
density meet real
value
 Has to be rotationally
symmetric
ADAMAS – 3rd Collaboration Meeting, Trento
 Charge amount and
density meet real
value
 All design possible
Florian Kassel
Simulation of an alpha particle hit (TCT)
Initial charge deposition simulated in SILVACO:
 Based on FLUKA simulation (5.0 MeV)
 Cut of charge distribution in radial length to increase
simulation speed
Alpha 5.0 MeV in diamond
Cut
y – depth (cm)
6
19.11.2014
ADAMAS – 3rd Collaboration Meeting, Trento
Florian Kassel
Benchmarking:
2D - SIMULATION
Comparing of simulation results with a TCT measurement of
a real sCVD diamond.
 Charge carriers density
dependent on thickness
of 2D slice
 charge carrier influence
on electric field no
longer negligible
 High particle density
leads to electrical field
free regions
 Diffusion becomes
dominant
7
19.11.2014
ADAMAS – 3rd Collaboration Meeting, Trento
Florian Kassel
Benchmarking:
Quasi-3D SIMULATION (200V)
Measurement e
Measurement h
Simulation e
Simulation h
0.10
Voltage (V)
• Quasi-3D like model
designed
• Parameters:
 Thickness: 530 µm
 Radial length:
100µm
• No traps simulated
• Mobility parameters
for e and h drift
used from thesis of
M.Pomorski
0.15
0.05
0.00
0
2
4
6
8
10
12
14
16
Time (ns)
8
19.11.2014
ADAMAS – 3rd Collaboration Meeting, Trento
Florian Kassel
18
20
Hole drift through diamond bulk
0.05
0.0 ns
0.3
3.0
5.0
ns
200V
Diamond
19.11.2014
9 Trento
ADAMAS – 3rd Collaboration Meeting,
530µm

0V
100µm
9
19.11.2014
ADAMAS – 3rd Collaboration Meeting, Trento
Florian Kassel
Benchmarking:
Quasi-3D SIMULATION (400V)
Deviation between
simulation and
measurements:
• Reduced alpha
particle energy
simulated
• Local cuts of
energy distribution
of alpha particle:
 >  >  µ
 >  > .  µ
 Further reduction
of alpha particle
energy.
10
19.11.2014
ADAMAS – 3rd Collaboration Meeting, Trento
Optimization of alpha particle
simulation necessary !
Florian Kassel
Benchmarking:
3D - SIMULATION
…still calculating
11
19.11.2014
ADAMAS – 3rd Collaboration Meeting, Trento
Florian Kassel
Quasi-3D:
Simulation of diamond polarization
Configuration:
•
•
Traps added to the diamond bulk, values assumed for first
tests
1
,
³
1
1.01014
,
³
 Acceptor traps1:
 = 1.23 ,  = 1.01013
ℎ = 1.01014 ²
 Donor traps:
 = 1.23 ,  =
ℎ = 1.01014 ²
Simulation of a MIP particle background
 Lab measurements: 3.7x106 Bq source used. Time order polarization
~1000s
 SILVACO TCAD: 4x1010 Bq source simulated. Speeding up of simulation
by a factor of 10.000 Polarization within 0.1s
•
Simulation of electron or hole drift by injection of charge
either on top or bottom of diamond.
1) M.Bruzzi et al., Deep levels and trapping mechanisms in chemical vapor
deposited diamond, Journal of applied Physics 91, 9 (2002), 5765 - 5774
12
19.11.2014
ADAMAS – 3rd Collaboration Meeting, Trento
Florian Kassel
Quasi-3D:
Simulation of diamond polarization
8000
Electrical field (V/cm)
• Different e/h-trap parameters
lead to an asymmetrical
ionization of the traps
 Asymmetrical electrical field
distribution
• Increased recombination in low
field region.
Electrical field
9000
7000
6000
5000
4000
3000
2000
1000
0
0
100
200
300
400
500
Thickness (µm)

7x1011
6x1011
Donor
Acceptor
5x1011
4x1011
3x1011
2x10
11
1x1011
Recombination (#/cm³s)
Ionized traps density (/cm³)
7x1017
6x10
Recombination
17
5x1017
4x1017
3x1017
2x1017
1x1017
0
-1x1017
0
50
100
150
200
250
300
350
400
450
500
Thickness (µm)
13
19.11.2014
ADAMAS – 3rd Collaboration Meeting, Trento
0
100
200
300
400
Thickness (µm)
Florian Kassel
500
Quasi-3D:
Resulting TCT pulse for electron and hole drift
0.12
Example sCVD_2012:
0.10
 2.5E14 24GeV peq at CMS
 Thickness: 410µm
 Mean after 750s (3.7MBq)
hole
electron
0.16
0.14
0.06
0.12
Signal (V)
Signal (V)
0.08
0.04
0.02
hole
electron
0.10
0.08
0.06
0.04
0.02
0.00
0.00
0
5
10
15
Time (ns)
0
5
10
15
20
25
30
Time (ns)
14
19.11.2014
ADAMAS – 3rd Collaboration Meeting, Trento
Florian Kassel
20
Cross checking of simulation results
Irradiation study planned:
•
Investigation of the radiation impact on the electrical
properties of a un irradiated sCVD diamond
•
Stepwise irradiation (~5e12 neq/cm²) till 5e13 neq/cm²
•
Proton irradiation performed at KIT (~23 MeV)
•
Systematic measurement of electrical field effects:
 Transient current technique (TCT)
 Charge collection efficiency (CCE)
•
15
Comparison of results with SILVACO TCAD simulation
19.11.2014
ADAMAS – 3rd Collaboration Meeting, Trento
Florian Kassel
Conclusion
•
Quasi-3D simulation are feasible to simulate electrical properties
of a diamond detector.
•
Simulation of an alpha particle in good agreement to real
measurement
•
First approach of simulating the polarization effect in diamond
 Increased density of donor traps assumed
 MIP particle background simulated
 Change of electrical field distribution and hence a change of
the shape of the TCT signal observed.
Outlook:
Simulation of MIP particle in order to calculate the CCE
• Modifying of traps in order to get closer to real results
• Simulation of improved diamond designs:
 Split pads (BCM1F), etc…
16
19.11.2014
ADAMAS – 3rd Collaboration Meeting, Trento
Florian Kassel

similar documents