Displacement Damage

 

Two fundamental damage mechanisms take place when devices are exposed to particle fluences: ionization and lattice displacement or just displacement damage. Ionization has previously been addressed in other simulation standard articles. Neutrons, protons, alpha particles, heavy ions, and very high-energy photons cause lattice displacement, or just displacement damage. Particle bombardment can change the arrangement of the atoms in the crystal lattice creating lasting damage, and increase the number of recombination (defect) centers depleting the minority carriers and degrading the analog properties of the affected semiconductor junctions. High dose rates of particles (particles/area-s) can cause partial annealing (“healing”) of the damaged lattice, leading to a lower degree of damage than with the same doses delivered in low intensity over a longer time period.

The displacement damage effects can vary depending on parameters such as, type of particle radiation, total dose and radiation flux, combination of types of radiation, and device operating frequency, operating voltage, actual state of the device during the instant of irradiation and intrinsic and extrinsic shielding. These issues makes thorough testing difficult, time consuming, and requiring a significant number of test samples. Silvaco’s displacement damage capability with Victory Device can assist in reducing the cost of testing by pointing the way to required tests to reduce uncertainties.

The damage associated with the collision between energetic particles and atoms within the crystal lattice is defects that can trap electrons and holes. The density of defects is represented by the equation below:

NF= αD * EL*Density*Fluence (EQ.1)

where

  • NF is the defcts/cm3
  • αD is the damage factor which represents the number of defect states created per unit energy deposited in the material
  • EL is the Non-Ionizing Energy Loss in MeV-cm2/g particle
  • Density is the mass density of the material in g/cm3
  • Fluence is the particle fluence in particles/cm2 where the particle species can be alpha, electron, ion, neutron, photon, proton, or user defined.

The Displacement Damage capability within the Radiation Effects Module of Victory Device supports combinations of particle species, with separate fluence for each type.

Radiation Example, radex12, demonstrates a displacement damage of a 4T CMOS Image Sensor structure. A Victory Process file creates the 4T CMOS Image Sensor structure, which is depicted in Figure 1.

 

Figure 1. 4T CMOS image sensor.

 

This radiation example sweeps the fluence of 1.8Mev Protons from 1e8 to 1e12 and plots the electron concentration as the fluence is swept through its values.

The fluence value is initially set to 1e8 using the set construct.
set FLUENCE=1e8

The Victory Process generated structure is imported
mesh infile=radex12_0.str

The radiation statement is declared and parameterized with protons of 1.8 MeV at a fluence of 1e8 from the set statement.
radiation proton energy=1.8 fluence=$FLUENCE

Using the material statement, the damaging particles are defined to be protons with a NIEL value of 3.1MeV-cm2/g.

material damage.proton=1e3 damage.niel=3.1

The displacement damage defects model is declared by setting the fluence.model flag on the defects statement. The density of defects is calculated using EQ.1 above. This density of defects applies individually to both acceptor-like and donor-like defects states. The energy of these defect states are assumed to be uniformly distributed across the band-gap. If one wanted to describe only defects of a single type, one must explicitly set NUMA or NUMD parameter to zero. The tail state parameters SIGTAE, SIGTAH, SIGTDE, and SIGTDH are used to specify the cross sections.

defects fluence.model sigtae=1.e-17 sigtah=1.e-15 sigtde=1.e-15 sigtdh=1.e-17

Victory Device allows one to specify combinations of particle species, but each requires a separate radiation statement for each different species, and supports the bounding of defects to specific locations or material using the localization parameter within the defect statement as described in section 6.4 of the Victory Device manual.

To sweep the Fluence value, one can use a feature described in Appendix B, of the VWF Interactive Tools Manual as shown below:

solve init
solve previous

# Deplete the Image Sensor of Electrons
log outfile=radex12_0.log
solve Vcgate=3.3 ramptime=1e-6 dt=1e-8 tstop=1e-6
solve tstop=2e-6 dt=1e-8

# Dark Recovery Time
solve tstop=1 dt=1e-7

go internal

load infile=radex12_1.in

sweep parameter=FLUENCE type=list data=”1e10, 1e11, 1e12”

 

with the radex12_1.in file being:

go victorydevice

set FLUENCE=1e8

mesh infile=radex12_0.str

radiation proton energy=1.8 fluence=$FLUENCE
material dam.proton=1e3 dam.niel=3.1

defects fluence.model \
sigtae=1.e-17 sigtah=1.e-15 sigtde=1.e-15 sigtdh=1.e-17

models consrh cvt fermi

output band.param con.band val.band

probe n.conc x=4.25 y=3.25 z=1 name=cis_conc
probe potential x=1 y=4.5 z=0.002 name=fd_potential
probe potential x=4.25 y=3.25 z=1 name=cis_potential

method pam.gmres norm.scaling.local

solve init
solve previous

# Deplete the Image Sensor of Electrons
log outfile=radex12_$’FLUENCE’.log
solve Vcgate=3.3 ramptime=1e-6 dt=1e-8 tstop=1e-6
solve tstop=2e-6 dt=1e-8

# Dark Recovery Time
solve tstop=1 dt=1e-7

The simulation of increasing fluence from a base, to 1e10, 1e11 and final to 1e12 is shown by Figure 2. The Figure shows that at a fluence of 1e12 that well collection no longer exists, and the Image Sensor is no longer functioning properly.

 

Figure 2. 4T CMOS Image Sensor degrading electron concentration from base through 1e10, 1e11 and 1e12 proton fluences.