modeling of charge collection efficiency degradation in ... · three main types of effects: -...

120 June 2016, Loughborough; E. Vittone

Modeling of charge collection efficiency degradation in semiconductor devices induced by

MeV ion beam irradiation

Ettore VittonePhysics Department

University of Torino - Italy


COOPERATION AND MUTUAL

UNDERSTANDING LEAD TO GROWTH

AND GLOBAL ENRICHMENT

IAEA Coordinate Research Programme (CRP) F11016 (2011-2015)

“Utilization of ion accelerators for studying and modeling of radiation

induced defects in semiconductors and insulators”

NUS

Ruđer Bošković Inst.

Delhi Univ.

ANSTOHelsinki Univ.

CAN

JAEA-Kyoto Univ.

Malesian Nuclear Agency

Torino Univ.

Surrey Univ.

SANDIA

Leipzig Univ.


Object of the research

Study of the radiation hardness of semiconductors

ToolFocused MeV Ion beams

to induce the damage

and

to probe the damage


Radiation damage is the general alteration of the

operational properties of semiconductor devices induced

by ionizing radiation

Three main types of effects:

- Transient ionization. This effect produces electron-hole pairs; particle

detection with semiconductors is based on this effect.

-Long term ionization. In insulators (oxides), the material does not return to

its initial state, if the electrons and holes produced are fixed, and charged

regions are induced.

- Displacements. Dislocations of atoms from their normal

sites in the lattice, producing less ordered structures, with

long term effects on semiconductor properties.


Frank Hartmann, Silicon tracking detectors in high-energy physics

Nuclear Instruments and Methods in Physics Research A 666 (2012) 25–46

M. Bruzzi, M. Moll, RD50, 2010


Modeling radiation degradation in solar cells extends satellite lifetimeRobert J. Walters, Scott Messenger, Cory Cress, Maria Gonzalez and Serguei Maximenko

A physics-based model of the effect of radiation on the performance of solar cells in

space may enhance the on-orbit lifetime of Earth-orbiting spacecraft. SPIE 2011

http://spie.org/x43655.xml

Space environment →

→ wide spectrum of ions

(protons) and electrons.

To understand the

performance of a solar cell in

the space radiation

environment, it is necessary to

know how cell degradation

depends on the energy of the

irradiating particle.

http://spie.org/x43655.xml


Characterization of radiation induced damage:

ded

0

DK1K1Y

Y

Device characteristics

before irradiation

Equivalent

damage factor

Particle

Fluence

Displacement

dose

Device characteristic after irradiation

First order: proportionality, independent of the particle, between the damage factor

and the particle NIEL

NIEL approach:

measurement of Ked only for one

particle (at one specific energy)

Ked can be estimated for all

the particles and energies


IBIC: Ion Beam Induced Charge

Transport)Carrier Free Energy, Deposited(FVout

Incoming radiation

V

Q

Incoming radiation

V

Vout

MeasuredWell known Material Characterization

Deposited Energy

Free charge generation

and transport

Output Electrical

Signal Vout


ded

0

DK1K1Q

QCCE

Induced Charge

before irradiation

Induced Charge after irradiation


Equivalent

damage factor

Particle

Fluence

Displacement

dose

MeV Ion beams

to induce the damage → DIB=DAMAGING IONS

And

to probe the damage→ PIB=PROBING IONS


CCE degradationinduced by ion irradiation

1 10 100 1000

0,85

0,90

0,95

1,00

Cl 11 MeV

CC

E

Fluence (m-2)

Hamamatsu photodiode

Vbias = 100 V

Is a function of the damaging ion fluence

ded

0

DK1K1Y

Y

DIB= Cl 11 MeV

PIB= He 1.4 MeV


1 10 100 1000

0,85

0,90

0,95

1,00

Cl 11 MeV

O 4 MeV

Li 2.15 MeV

CC

E

Fluence (m-2)

He 1.4 MeV


Vbias = 100 V


Is a function of the ion energy and mass

ded

0

DK1K1Y

Y

PIB= He 1.4 MeV


0,80

0,85

0,90

0,95

1,00

1 10 100 1000 10000

4H-SiC

Schottky diode

Hamamatsu

p-i-n diode

P-type Fz-Si

Fluence (m-2)

N-type Fz-Si

CC

ECCE degradation

induced by ion irradiationIs a function of the material and/or device

ded

0

DK1K1Y

Y


10 100 1000

0,85

0,90

0,95

1,00

100 V

20 V

50 V

10 V

CC

E

Fluence (m-2)

He 1.4 MeV


Vbias = 100 V

Vbias

dedbias

0

DK1)V(K1Y

Y


Is a function of the polarization state of the device


100 1000 10000

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1,0

Damage induced by 8 MeV He

1 MeV H

2 MeV H

4.5 MeV H

8 MeV He

12 MeV HeCC

E

Fluence (m-2)

n-type Fz silicon diode

Vbias = 50 V

Probing ions

dedbias

0

DK1) PIB,V(K1Y

Y

CCE degradationinduced by ion irradiationIs a function of the probing ions (PIB)


Summary

CCE degradation

DIBs(12 MeV and 8 MeV He)

PIBs (1, 2, 4.5 H; 8, 12 MeV He)

Different bias voltages (10,20,50 V)








NUS


Delhi Univ.

ANSTOHelsinki Univ.

CAN

JAEA-Kyoto Univ.


Torino Univ.

Surrey Univ.

SANDIA

Leipzig Univ.


Model for charge pulse

formation

(IBIC theory)

Model for CCE

degradation

(SRH model)

Experimental protocol

• To correlate the effect of different kinds of radiation on the

properties of materials and devices

• To extract parameters directly correlated with the radiation

hardness of the material

Goals


Model for charge pulse formation

(IBIC theory)

• Formalism based on the Shockley-Ramo-Gunn

theorem

• Adjoint equation method: the CCE is the

solution of the Adjoint Equation

T.H.Prettyman, Nucl. Instr. and Meth. in Phys. Res. A 422 (1999) 232-237.


Gunn’s theorem

Pulse shapes calculation

V-qI

Ev

Shockley-Ramo theorem

d

1-qI v

Weighting field

Gunn theorem


Vi

p0n0

t

0

iV

)r(E)r(v)r;'t,r(p)r(v)r;'t,r(nrd'dtq)t(Q

charge induced the Evaluate

constant held are

conductorsother theall of potentials The

E

equation sPoisson’ thesolvingby

V

E

field g weightinsGunn' theEvaluate

i

Initial conditions

For mapping charge pulses

0tat point Generation r

)t()rr(G

0

0pn,

Formalism based on the Gunn’s theoremB

oun

da

ry

co

nd

itio

ns

p

pp0p

n

nn0n

0

pGpDp

t

p

nGnDn

t

n

charge space ,conditions boundary

by defined potential

the using equations continuity the Solve



(IBIC theory)

• Formalism based on the Shockley-Ramo-Gunn

theorem

• Adjoint equation method: the CCE is the

solution of the Adjoint Equation

T.H.Prettyman, Nucl. Instr. and Meth. in Phys. Res. A 422 (1999) 232-237.


equation continuityelectron for thefunction sGreen' theis

V

)()()r;'t,(nd'dtq)t(Q

electrons from Induced Charge

Vi

n0

t

0

in

rErvrr

The continuity equation

involves linear operators

The charge induced from

electrons can be evaluated by

solving a single, time

dependent adjoint equation.

n

n*

n0n

nGnDn

t

n

i

nn

in

V

EG

Qn

Short-cutAdjoint equation Method


d

0x

0

x

y nnS

d

x

y

x ppS

S

v

1dzexp

V

yFdy

v

1dzexp

V

yFdy

xdxqQ

Electrons

Holes

Gunn’s

weighting field

Drift lengths


(IBIC theory)

Ionization profile Fully depleted device

No diffusion

Ramo Theorem

1D


Capture

coefficient

)x(Vac11

0

Basic assumption:

1) In the linear regime, the ion induced damage affects mainly the carrier lifetime

2) The ion induced trap density is proportional to the VACANCY DENSITY

Fluence

Vacancy Density Profile

Model for CCE degradation

Shockley-Read-Hall model


The experimental protocol

Z. Pastuovic et al., IEEE Trans on Nucl. Sc. 56 (2009) 2457; APL (98) 092101 (2011)


Samples under studyn- and p- type Fz p-i-n Si diodes

Fabricated by the Institute of Physics, University of Helsinki

16 floating guard rings

The frontal electrode and the guard rings

are coated with Al (0.5 µm]).

The Al electrode has a hole in the center, 1 mm diameter.

Different dimensions: 5 or 2.5 mm

MeV ions


Electrical

characterization

C-V characteristics

Depletion width-voltageExperimental

protocol



Electrical

characterization

Electrostatic

modeling

Experimental

protocol


Electron drift

velocity profiles

hole drift velocity

profiles

Gunn’s weighting

potential

Gunn’s weighting

field


MeV scanning

ion microbeam

Spot size < 3 m

PROBING THE PRISTINE SAMPLE


IBIC map on a pristine diode

probed with a scanning

1.4 MeV He microbeam;

Uniform CCE map

Electrical

characterization

Electrostatic

modeling

IBIC map on

pristine sample

Experimental

protocol



Ion microbeams

Different ion mass/energy

Spot size < 3 m

DAMAGING SELECTED AREAS

100X100 m2


500 m

100 m

1 2

4 5

7 8

3

6

9

Electrical

characterization

IBIC map on

pristine sample

Irradiation of 9

regions at

different fluences

Experimental

protocol





ZOOM in view of the selected area for focused

ion beam irradiation at different fluences


He ion microbeam

Energy 1.4 MeV

Spot size < 3 m

0 1 2 3 4 5 6 70.0

0.1

0.2

0.3

Depth (m)

No

rmali

zed

Io

niz

ing

En

erg

y l

oss (

1/

m)

Data from SRIM

1.4 MeV He in Si

PROBING DAMAGED AREAS


a measured 2D distribution

of the IBIC signal amplitude

after irradiation

Electrical

characterization

IBIC map on

pristine sample

Irradiation of 9

regions at

different fluences

IBIC map of

irradiated regions

Experimental

protocol


500 m

100 m

1 2

4 5

7 8

3

6

9







after irradiation

Electrical

characterization

IBIC map on

pristine sample

Irradiation of 9

regions at

different fluences

IBIC map of

irradiated regions

Experimental

protocol


500 m

100 m

1 2

4 5

7 8

3

6

9




0

50

100

420 440 460 4800

50

100

Co

un

ts

Fluence

10 V

100 V

Co

un

ts

Pulse Height (Channels)

f

IBIC spectra

(bias voltage =

10 V and 100 V)

from the central

regions of four

of the areas

shown in Fig. c




after irradiation

Electrical

characterization

IBIC map on

pristine sample

Irradiation of 9

regions at

different fluences

IBIC map of

irradiated regions

Experimental

protocol


500 m

100 m

1 2

4 5

7 8

3

6

9




10 100 1000

0,85

0,90

0,95

1,00

100 V

20 V

50 V

10 V

CC

E

Fluence (m-2)

He 1.4 MeV


Vbias


PIB = Probing ion beam

DIB = Damaging ion beam

DIB: Vacancy profiles

PIB: Ionization profiles

Different bias voltages


0.0 2.0x1012

4.0x1012

6.0x1012

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

DIB=8 MeV He

Vbias

= 50 V

8 MeV He

12 MeV He

CC

E

Fluence (cm-2

)

2 MeV H

PIB

Fixed DIB

Fixed Vbias

Variable PIBs

0.0 2.0x1012

4.0x1012

6.0x1012

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

10 V

20 V

50 V

Vbias

(V)

CC

E

Fluence (cm-2

)

DIB=8 MeV He

PIB=2 MeV H

0.0 2.0x1012

4.0x1012

6.0x1012

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

12 MeV He

8 MeV He

DIB

CC

EFluence (cm

-2)

PIB=2 MeV H

Vbias

= 50 V

Variable DIB

Fixed PIB

FIXED Vbias

Fixed DIB

Fixed PIB

Variable Vbias


0 10 20 30 40 50

0

20

40

60

80

100

Vacan

cy/i

on

/m

Depth (m)

Vacancy

profile

P+N N+

DIB


0 10 20 30 40 50

0

20

40

60

80

100

Vacan

cy/i

on

/m

Depth (m)

Vacancy

profile

0 10 20 30 40 500

50

100

Ion

izati

on

en

erg

y l

oss (

keV

m)

Depth (m)

P+N N+

PIB

Generation

profile

Short range PIB


0 10 20 30 40 500

50

100

Ion

izati

on

en

erg

y l

oss (

keV

m)

Depth (m)

P+N N+

PIB

Electric field

Electron motionHole motion

Generation

profile

0 2 4 6 8 100,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1,0C

CE

Fluence (x1012

cm-2

)

PIB=1 MeV H

DIB=8 MeV He


0 2 4 6 8 100,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1,0C

CE

Fluence (x1012

cm-2

)

PIB=1 MeV H

DIB=8 MeV He

Residual map

d

0

d

x

y

x

n

0nS

S )x(Vac1

v

1dzexp

V

yFdyxdxqQ

n Free parameter


0 10 20 30 40 50

0

20

40

60

80

100

Vacan

cy/i

on

/m

Depth (m)

Vacancy

profile

P+N N+

DIB


0 100 200

0

30

60

90

Ioni

zatio

n en

ergy

loss

(keV

m

)

Depth (m)

0 10 20 30 40 50

0

20

40

60

80

100

Vacan

cy/i

on

/m

Depth (m)

Vacancy

profile

P+

N

N+

PIB

Generation

profile

Long range PIB


0 100 200

0

30

60

90

Ioni

zatio

n en

ergy

loss

(keV

m

)

Depth (m)

0 10 20 30 40 50

0

20

40

60

80

100

Vacan

cy/i

on

/m

Depth (m)

Vacancy

profile

P+

N

N+

PIB

Generation

profile

Electron motionHole motion

0 2 4 6 8 100,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1,0C

CE

Fluence (x1012

cm-2

)

PIB=4.5 MeV H

DIB=8 MeV He


0 10 20 30 40 50

0

20

40

60

80

100

Vacan

cy/i

on

/m

Depth (m)

Vacancy

profile

0 2 4 6 8 100,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1,0C

CE

Fluence (x1012

cm-2

)

PIB=4.5 MeV H

DIB=8 MeV He

Residual map

d

0d

x

y

x

n

0nS

x

0

x

y

p

0pS

S

)x(Vac1

v

1dzexp

V

yFdy

)x(Vac1

v

1dzexp

V

yFdy

xdxqQ


Short range PIB Long range PIB

n=1700 m3/s

p=130 m3/s

Bias Voltage = 50 V


Damaging ions: 8 MeV He

Probing ions: 1,2,4.5 MeV H, 12 MeV He

Bias Voltages: 10,20 50 V

n-type Fz silicon diode

N-typeP-type


Fz silicon diode

Capture coefficient

n-type

n = (2500±300) m3/s

p = ( 210±160) m3/s

p-type

n = (1310± 90) m3/s

p = (2200±300) m3/s


N-type silicon

DLTS measurements

singly V2(−/0) negatively charged divacancy

σn≈5·10-15 cm2

0 5 10 15 20 25 30 35 40 45 50 55 6010

-3

10-2

10-1

100

101

102

8 MeV He

Va

ca

nc

y/Io

n/

m

Depth (m)

Vacancy Marlowe

di-Vacancy Marlowe

Vacancy SRIM

From MARLOWE

simulation

n=vthσn

σn≈(3.6±0.4)·10-15 cm2

C. R. Crowell,

Appl. Phys. 9, 79-81, 1976

Vth= 1.8 ·107 m/s


d

0d

x

y

x

n

0nS

x

0

x

y

p

0pS

S

)x(Vac1

v

1dzexp

V

yFdy

)x(Vac1

v

1dzexp

V

yFdy

xdxqQ

Solution of the adjoint equations

For very low level of radiation

Linearization vs.

Effective fluence *


Very low level of damage

E. Vittone et al. / Nuclear Instruments and Methods in Physics Research B 372 (2016) 128–142


Derivation of the Non Ionizing Energy Loss

(NIEL) displacement damage formula

Constant vacancy profile

Low displacement damage

ded DK1CCE

z

0

d

z

z

0 S

eff

pp

d

z S

eff

nn

R

0

ed xdxV

yFdy)z(kxdx

V

yFdy)z(kdz

MK

Ked = equivalent damage factor depends on

Electrostatics of the device

Carrier transport and recombination

Ion probe ionization profile


Limits of applicability

)x(Vacv1

)x(Vac11

thh,e

h,e,0

p,n

h,e,0h,e

Basic Hypotheses

DIB : low level of damage

“linear model”

Independent traps, no clusters

Unperturbed electrostatics (i.e. doping profile) of the device

PIB : ion probe

CCE is the sum of the individual e/h contributions

No plasma effects induced by probing ions


An experimental protocol has been proposed to study the radiation hardness

of semiconductor devices

Under the assumption of low damage level,

the CCE degradation of a semiconductor device induced by ions of different

mass and energy can be interpreted by means of a model based on

•The Shockley-Ramo-Gunn theorem for the charge pulse formation

•The Shockley-Read-Hall model for the trapping phenomena

If the generation occurs in the depletion region, an analytical solution of the

adjoint equation can be calculated.

Adjusted NIEL scaling can be derived from the general theory in the case of

constant vacancy profile.

The model leads to the evaluation of the capture coefficient.

CONCLUSIONS

The capture coefficient is directly related to the

radiation hardness of the material








NUS


Delhi Univ.

ANSTOHelsinki Univ.

CAN

JAEA-Kyoto Univ.


Torino Univ.

Surrey Univ.

SANDIA

Leipzig Univ.



ded

0

DK1K1Y

Y

Device characteristics

before irradiation

Equivalent

damage factor

Particle

Fluence

Displacement

dose

Device characteristic after irradiation

First order: proportionality, independent of the particle, between the damage factor

and the particle NIEL

NIEL approach:

measurement of Ked only for one

particle (at one specific energy)

Ked can be estimated for all

the particles and energies


ded

0

DK1K1Q

QCCE

Induced Charge

before irradiation

Induced Charge after irradiation


Equivalent

damage factor

Particle

Fluence

Displacement

dose


n

n*

n0n

nGnDn

t

n

Excess carrier lifetime

thtrap vN

1

Trap density Thermal velocity

Capture cross section

K

11

0

kNN traptrap

0 Trap density induced by

radiation

Trap density in pristine

material


Damaging ions: 8 MeV He

Probing ions: 1,2,4.5 MeV H, 12 MeV He

Bias Voltages: 10,20 50 V

n-type Fz silicon diodeBias Voltage = 50 V

Bias Voltage = 20 V

Bias Voltage = 10 V

n = (2500±300) m3/s

p = (210±160) m3/s

CAPTURE COEFFICIENTS

modeling of charge collection efficiency degradation in ... · three main types of effects: -...

Documents