Title: Capacitance at Different Temperature: Simulation and Measurements
1Capacitance at Different Temperature Simulation
and Measurements
- 6nd ROSE Workshop CERN 23.10.2000
- A. Santocchia, B. MacEvoy, G. Hall
- Imperial College London
- F. Moscatelli, D. Passeri
- Perugia University
2Outline
- Defect Kinetics Model
- Point Defects - Cluster Defects
- Setup Measurements
- Data VS Kinetics Model Comparison
- Simulation Model
- Status Future Work
3Standard SRH
- Defect Reactions
- Interstitial - Vacancy
- Impurity Concentrations
- Phosphorus
- Oxygen
- Carbon
-
4Gamma and Neutron Results
- Good agreement between model and data for gamma
irradiation - V2O responsible for changes in Neff
- BUT
- Problems with Hadron Interaction
- a and Neff are too small
5Now what? ? Clusters
- Gamma irradiation
- Point Defects
- Hadron irradiation
- Point Defects Cluster Defects
-
- CLUSTER When a Primary Knock-on Atom loses the
last 5 KeV of recoil energy ? many defects in a
small spatial region - Some CLUSTER Numbers
- Typical Dimension 5-10 nm
- Silicon Atoms in a Cluster 104
- Defects in a Cluster 10
- Typical Density 1019 cm-3
- Average Defect Separation 1 nm
- Introduction Rate 0.4 cm-1
6Charge Exchange (Not SRH!)
Is direct charge exchange between defects
forbidden? NO What if we include this new
phenomenon in the Model?
7Charge Exchange 2
Increase in carrier generation rate and
acceptor-like contribution to Neff predicted for
expected density of defects in Clusters
8Proton and Neutron Results
- Neutron Irradiation
- All materials equally rad-hard
- Proton Irradiation
- Oxygen-enriched More rad-hard
- Carbon-enriched Less rad-hard
Why ? V2O more important for p
9Room VS Cryogenic Temperature
- The model performs well _at_ RT
- Is model still good _at_ T 100K - 300K ?
- Is there any chance to find other evidence of
Charge Direct Exchange ? - Already experimentally seen but with complex
spectroscopy techniques (ODMR, ESE-EPR)
- Cryogenic System Built to check the Model
10Experimental Setup
- HP LCR Meter 4284A with variable frequency (20 Hz
- 1MHz) - Small Heating Rate Cryogenic System (0.5K/min)
- CV scan of Irradiated Diode
- T (100 - 300)K Vbias (0-150) V
- f 200Hz, 1KHz, 10KHz
11Diode Characteristics
- Standard Hammamatsu Process
- 310 mm thick - 5?5 mm2 guarded
- Resistivity 5.8 KWcm Vdep54V
- Fluence 8.7?1013 n/cm2
- Beneficial annealing completed and then kept _at_
T-20C
12Model and Data _at_ RT
SRH CC, CO, P, VP, VO, V2O, V3O non SRH V2,
E70, E170
- First check that Model and Data are in agreement
for the chosen diode - Complete Diode characterization performed _at_ RT
- Lets start the game!
13Neff VS T (Model Data)
- Watts Da Via observe Type Inversion _at_ T190K
for fluence 1014 n/cm2 - The model shows inversion _at_ T225K but the
fluence is lower No Big discrepancy!
14Comments
- Can Neff be extracted from data?
- Look at CV plots at different Temperature and try
to guess where is Vdep - And what about the Frequency al low T?
15Temperature 100K 200K
16Temperature 250K 300K
17Simulation
- Finite Element Simulator Silvaco
- Simple Diode Structure (50mm width 300mm Thick)
- 3 Defects with standard SRH Statistic and the
following parameters
Energy Defect sn (cm-2) sp (cm-2) H (cm-1)
Ec-0.42 V2 -/0 10-16 2x10-15 1
Ec-0.55 V20 10-16 10-15 0.1
Ev0.36 CO 10-15 10-16 1
18Current Consideration
- The current depend on the level occupancy AND on
the defects density - In the simulation the V2 level (Ec-0.42) is the
major responsable for the current - The Increasing of the V2 Introduction Rate hide a
very clear and simple explanation
The simulation doesnt take into account any
direct inter-level charge exchange
19Occupancy (Level Ec-0.42)
20Occupancy (Level Ec-0.55)
21Occupancy (Level Ev0.36)
22Occupancy (All Defects)
23Capacitance (DataSimulation)
24Conclusion
- Measurements and Simulation show that SRH is not
enough (no news here!) - Work in progress
- Experimental Setup ready at IC for Signal
measurements _at_ T (100 - 300)K indipendent value
for Vdep soon available as a cross-check - CV curve Interpretation difficult but could give
interesting information compared with Signal
measurements and Device Simulation