Scott R. Messenger, Ph.D. - oma · • Provide space solar cell community an alternate method for cell level degradation analyses • Heritage method (JPL EQFLUX) requires several

Scott R. Messenger, Ph.D.US Naval Research Laboratory, Washington, DC

[email protected]‐767‐7360

Space Solar Cell Radiation Damage Modelling

mailto:[email protected]

“Prediction is very difficult, especially if it’s about the future.”

-Niels BohrDanish physicist (1885 - 1962)

*Taken from M. Xapsos, NSREC 2008

Space Radiation and Effects

Outline1. Motivation2. The Space Radiation Environment3. Solar Cell Space Radiation

Degradation Modeling-JPL Equivalent Fluence (EQFLUX)-NRL Displacement Damage Dose (SCREAM)

4. Radiation Damage in Multijunction Solar Cells

Motivation• Provide space solar cell community an alternate

method for cell level degradation analyses• Heritage method (JPL EQFLUX) requires several

ground test energies• Electrons (0.6, 1, 12 MeV)• Protons (0.02, 0.05, 0.1, 0.3, 1, 3, 10 MeV)

• Alternate Method (NRL Displacement Damage Dose, DDD) needs significantly less ground data

• Electrons (1, >2 MeV)• Protons (>1 MeV)

• AIAA S-111-2005 Space Solar Cell Qualification (Sects. 8.1 & 8.2) have been modified to include NRL DDD model as primary which can save about $150K in future space cell qualifications

Problem• To generate ground irradiation data necessary to predict

the effect of a particle spectrum (as that found in space) on a solar cell in orbit

• This is accomplished by reducing all of the ground data to a characteristic data set

Electron and Proton Fluence Data (GaAs/Ge, 1991)

108 109 1010 1011 1012 1013 1014 1015 1016 10170.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Nor

mal

ized

Max

imum

Pow

er 1 Sun, AMOGaAs/Ge

Particle Fluence (cm-2)

Protons

Electrons0.6 MeV1 MeV2.4 MeV12 MeV

9.5 MeV3 MeV1 MeV0.5 MeV0.3 MeV0.2 MeV0.1 MeV0.05 MeV

25oC

Solution• We want all data to collapse to a common basis• The JPL method uses an equivalent 1 MeV electron fluence• The NRL Method uses the displacement damage dose

JPL Data Collapse

108 109 1010 1011 1012

Displacement Damage Dose (MeV/g)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Nor

mal

ized

Max

imum

Pow

er 1 Sun, AM0T=25oC

GaAs/Ge

0.6 MeV1 MeV2.4 MeV12 MeV

Protons Electrons9.5 MeV3 MeV1 MeV0.5 MeV0.3 MeV0.2 MeV

1 MeV equiv.Neutrons

NRL Data Collapse

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.E+11 1.E+12 1.E+13 1.E+14 1.E+15 1.E+16 1.E+17

Equivalent 1 MeV Electron Fluence (cm-2)

Nor

mal

ized

Pm

ax

200keVp

300keVp

500keVp

1MeVp

3MeVp

9.5MeVp

12 MeV

2.4MeVe

1MeVe

600keVe

Proton Fit

Electron Fit

GaAs/Ge




Space Radiation Environment

• Omnidirectional• Continuous Energy Spectrum• Multi-particle (light and heavy ions, neutrons?)

p+

COVERGLASS

SUBSTRATE

ACTIVE CELL

PANEL

p+

e-

e-

e-

e-

e-

p+

p+

p+

p+

p+

p+

COVERGLASS

SUBSTRATE

ACTIVE CELL

PANEL

p+

e-

e-

e-

e-

e-

p+

p+

p+

p+

p+

COVERGLASS

SUBSTRATE

ACTIVE CELL

PANEL

p+

e-

e-

e-

e-

e-

p+

p+

p+

p+

p+

protonselectrons

Space Radiation Environment

• Omnidirectional• Continuous Energy Spectrum• Multi-particle (light and heavy ions, neutrons?)

p+

COVERGLASS

SUBSTRATE

ACTIVE CELL

PANEL

p+

e-

e-

e-

e-

e-

p+

p+

p+

p+

p+

p+

COVERGLASS

SUBSTRATE

ACTIVE CELL

PANEL

p+

e-

e-

e-

e-

e-

p+

p+

p+

p+

p+

COVERGLASS

SUBSTRATE

ACTIVE CELL

PANEL

p+

e-

e-

e-

e-

e-

p+

p+

p+

p+

p+

protonselectrons

1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

1.E+05

1.E+06

1.E+07

1.E+08

1.E+09

0 1 2 3 4 5 6 7

L Shell Value

AE8

MIN

Inte

gral

Flu

x >

E (e

lec/

cm2 /s

)0.040.10.20.30.40.50.60.70.811.251.51.7522.252.52.7533.253.53.7544.254.54.7555.566.57

AE8MIN Electron Spectra (Static)MeV

1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

1.E+05

1.E+06

1.E+07

1.E+08

1.E+09

0 1 2 3 4 5 6 7

L Shell Value

AP8

MIN

Inte

gral

Flu

x >

E (p

rot/c

m2 /s

)0.10.150.20.30.40.50.60.711.52345671015203040506070100150200300400

AP8MIN Proton Spectra (Static)MeV

Space Radiation Spectra

1.E+06

1.E+07

1.E+08

1.E+09

1.E+10

1.E+11

1.E+12

1.E+13

0.1 1 10 100 1000Proton Energy (MeV)

Inte

gral

Flu

ence

(cm

-2)

AP8MINAP8MAXCRRESPRO QuietCRRESPRO ActivePSB97Solar Protons

1.E+04

1.E+05

1.E+06

1.E+07

1.E+08

1.E+09

1.E+10

1.E+11

1.E+12

1.E+13

0.1 1 10 100 1000Proton Energy (MeV)

Diff

eren

tial F

luen

ce (c

m-2

MeV

-1)

AP8MINAP8MAXCRRESPRO QuietCRRESPRO ActivePSB97Solar Protons

1100 km, circular, 63o, 1 year duration

Integral Spectra Differential Spectra

*The space radiation spectra is generally considered accurate to ~2X. This is the dominant error source for any degradation analysis leading to …… MARGIN!!!!

What is a Coronal Mass Ejection?

Space Weather - Solar Proton Event Data October 19, 1989 Event

10/1

9/19

89

10/2

1/19

89

10/2

3/19

89

10/2

5/19

89

10/2

7/19

89

10/2

9/19

89

10/3

1/19

89

11/2

/198

9

11/4

/198

9

11/6

/198

9

11/8

/198

9

11/1

0/19

89

11/1

2/19

89

Date & Time

10-6

10-5

10-4

10-3

10-2

10-1

100

101D

iffer

entia

l Flu

x (c

m-2

s-1sr

-1M

eV-1

)

GOES-7 SEMIMP-8 GME

121-154 MeV Protons

Electron and Proton Radiation Environment

*The present radiation environment models are known to have an accuracy of at least a factor of 2. AP(E)9 will greatly update the environment data. A beta version has been created in May, 2010. The NASA Radiation Belt Storm Probe (RBSP) experiment (launch in 2012) is expected to greatly enhance the understanding.

Any model is only good to the data that is given!

(CEASE data onboard TSX5, DSP21, HEO and ICO Satellites)

1.E‐06

1.E‐05

1.E‐04

1.E‐03

1.E‐02

1.E‐01

1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

1.E+05

1.E+06

1.E+07

1.E+08

0.01 0.1 1 10

Averaged

Integral Flue

nce by

Year

(e‐ /cm

2 /sec)

Electron Energy (MeV)

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

All data0.0

0.5

1.0

1.5

2.0

2.5

1.E‐02 1.E‐01 1.E+00 1.E+01


AE8MIN/AE8MAX

SV41/AE8MAX

GPS Environment Data (LANL)

1.0E+09

1.0E+10

1.0E+11

1.0E+12

12/10/2000

12/10/2001

12/10/2002

12/10/2003

12/10/2004

12/10/2005

12/10/2006

12/10/2007

12/10/2008

12/10/2009

Daily integrated 1 MeV electron fluence

6 mil SiO2 RDC LANL

6 mil SiO2 RDC AE8MIN

6 mil SiO2 RDC AE8MAX

GPS Environment Data (LANL)

GPS Array Data with Environment Data (LANL)

0.65

0.70

0.75

0.80

0.85

0.90

0.95

1.00

12/10/2000

12/10/2001

12/10/2002

12/10/2003

12/10/2004

12/10/2005

12/10/2006

12/10/2007

12/10/2008

12/10/2009

P max/P

max0

Maximum Power (Pmax) Remaining Factor


0.65

0.70

0.75

0.80

0.85

0.90

0.95

1.00

12/10/2000

12/10/2001

12/10/2002

12/10/2003

12/10/2004

12/10/2005

12/10/2006

12/10/2007

12/10/2008

12/10/2009

P max/P

max0



0.65

0.70

0.75

0.80

0.85

0.90

0.95

1.00

12/10/2000

12/10/2001

12/10/2002

12/10/2003

12/10/2004

12/10/2005

12/10/2006

12/10/2007

12/10/2008

12/10/2009

P max/P

max0


Bottom LineThe environment is a significant source of error, depending on the mission and lifetime.




Displacement Damage

Incident particle

Scattered particle

Primary knock-on

atom (PKA)

RutherfordNuclear Elastic

Nuclear Inelastic

PROTONS

RutherfordELECTRONS

VacanciesInterstitials

SIMPLE DEFECTS

Vacancy/impurityMulti-vacancy/interstitial

Clusters

COMPLEXES

INTERACTIONS

DAMAGE

Effect of Particle-Induced Damage

p+

100 150 200 250 300Temperature (K)

-100

0

100

DLT

S S

igna

l (pF

)

H3 H4H5

H7

EDEAEC

n+p InP

Trap

Ene

rgy

0.00

0.67

1.34

H3H4

H5H7

ED EAEC

Recombination CentersDiffusion Length Degradation

E

E

Compensation CentersCarrier removal

JPL and NRL Methods•NASA Jet Propulsion Laboratory (Pasadena, CA)

–Calculate equivalent 1 MeV electron fluence for mission–Uses empirically determined relative damage coefficients (RDCs)

–Read EOL power from measured 1 MeV electron curve

•Naval Research Laboratory (Washington, DC)–Calculate displacement damage dose (Dd ) for mission

–Uses calculated values of nonionizing energy loss (NIEL)

–Read EOL power from measured “characteristic” curve

*Both methods have the same general approach.

Orbital Degradation Calculations

ppp

pppeee

e

eeelectron MeV 1 t)dE,(ERDC

dE)(Ed

Ct)dE,(ERDCdE

)(Ed

JPL Method

NRL Method

Cpe is determined empirically (75-90% degradation level)

e

1ne

ee

eeppp

p

pd dE

)MeV1(NIEL)E(NIEL)(ENIEL

dE)(EdR)dE(ENIEL

dE)(Ed

D

Rep is determined empirically (ratio Dxe /Dxp )

*Both methods currently rely on single-valued electron-proton degradation correlation. However, these methods can be adapted to account for degradation-dependent correlation (WCPEC 2006).

Heritage Model

• Data needed– Protons

• E = 0.02, 0.05, 0.1, 0.3, 1, 3,

10 MeV•

= 3e9 to 2e12 p+/cm2

– Electrons• E = 0.6, 1, 12 MeV•

= 2e13 to 1e16 p+/cm2

JPL Equivalent Fluence Method

Electron and ProtonFluence Data (GaAs/Ge, 1991)

108 109 1010 1011 1012 1013 1014 1015 1016 10170.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Nor

mal

ized

Max

imum

Pow



Protons



25oC

Electron Damage Coefficients

Proton Damage Coefficients


RDC Calculation

10-1 100 101 102


10-3

10-2

10-1

100

101

102

Rel

ativ

e P m

ax D

amag

e C

oeffi

cien

t

0 mil1 mil3 mil6 mil12 mil20 mil30 mil60 mil

Coverglass Thickness

*Relative to 1 MeV normalincidence data, w/o coverglass

Normal incidenceno coverglass

10-2 10-1 100 101 102

Proton Energy (MeV)

10-2

10-1

100

101

102

Rel

ativ

e P m

ax D

amag

e C

oeffi

cien

t0 mil1 mil3 mil6 mil12 mil20 mil30 mil60 mil


*Relative to 10 MeV proton normalincidence data, w/o coverglass

Normal incidenceno coverglass

•Data exist for Pmax, Isc, and Voc for both electrons and protons

•Solar cell technologies available include Si, GaAs/Ge, MJ (2001)

•Empirical values for Cep

(based on 75‐90% level)

•Measurement intensive ($$)

GaAs/Ge (1991)

GaAs/Ge (1991)

(10 MeV Protons & 1 MeV electrons)

108 109 1010 1011 1012 1013 1014 1015 1016 10170.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Nor

mal

ized

Max

imum

Pow



Protons



25oC80% BOL

Equivalent Fluence Analyses of GaAs/Ge Solar Cells

x01logC1

P)(P

JPL Pmax vs Phi

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.E+10 1.E+11 1.E+12 1.E+13 1.E+14 1.E+15 1.E+16 1.E+17 1.E+18Phi(10 MeV p+) and Phi(1 MeV e-) (cm-2)

Nor

m P

max

Protons1 MeV electrons5th order fit10 MeV proton fluence data1 MeV electron fluence data

JPL Voc vs Phi

0.60

0.65

0.70

0.75

0.80

0.85

0.90

0.95

1.00

1.E+10 1.E+11 1.E+12 1.E+13 1.E+14 1.E+15 1.E+16 1.E+17Phi(10 MeV p+) and Phi(1 MeV e-) (cm-2)

Nor

m V

oc

Protons1 MeV electrons5th order fit10 MeV proton fluence data1 MeV electron fluence data

value error value error value errorC 0.297 0.003 0.124 0.008 0.236 0.018

Phix 1.92E+11 5.44E+09 2.04E+11 3.88E+10 5.34E+11 1.00E+11Cep

value error value error value errorC 0.351 0.025 0.061 0.003 0.361 0.0341

Phix 2.38E+14 4.85E+13 2.79E+13 6.04E+12 4.46E+14 1.07E+14

10 MeV Proton and 1 MeV electron fit results

10 MeV Protons

1 MeV Electrons

1000 1400

Pmax Voc Isc

Pmax Voc Isc

400

0 10 20 30 40 50 60 70 80SiO2 Coverglass Thickness (mil)

1014

1015

1016

1 M

eV E

lect

ron

Flue

nce

(e- /c

m2 )

GaAs

5000 km, circular, 60o orbit (1 year duration)

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.E+11 1.E+12 1.E+13 1.E+14 1.E+15 1.E+16 1.E+17

Equivalent 1 MeV Electron Fluence (cm-2)

Nor

mal

ized

Pm

ax

200keVp300keVp500keVp1MeVp3MeVp9.5MeVp12 MeV2.4MeVe1MeVe600keVeProton FitElectron Fit

GaAs/Ge

Initial Omnidirectional Spectrum

Equivalent 1 MeV Electron Fluence

Proton Damage Coefficients

1 MeV Electron Pmax

Degradation10-2 10-1 100 101 102

Proton Energy (MeV)

10-2

10-1

100

101

Rel

ativ

e P m

ax D

amag

e C

oeff

icie

nt

0 mil1 mil3 mil6 mil12 mil20 mil30 mil60 mil

Coverglass

*Relative to 10 MeV normalincidence data, w/o

Thickness

coverglass, based on Pmax

JPL Equivalent Fluence Method

w/Cep

GaAs/Ge (1991)

10-1 100 101 102 103

Proton Energy (MeV)

5000 km, circular, 600 orbit (1 year duration)

107

108

109

1010

1011

1012

1013

1014

Flue

nce

(cm

2 MeV

)-1

JPL Model Pros/Cons• Pros:

– Heritage (developed in the 1980s)– Widely available and already incorporated into many space

radiation suites (SPENVIS, Space RadiationTM, OMERE, etc.)• Cons:

– Much ground test data needed ($$)– Requires 1 MeV electron (to 12 MeV) AND 10 MeV proton data– Currently available for Si (1982), GaAs/Ge (1996), MJ (1999)– Program not particularly user friendly in FORTRAN version– Entire calculation is technology specific (every design change

needs requalification, $$)– Calculation of omnidirectional RDCs for covered cells not

trivial and coverglass and technology specific– Assumes that the total damage can be characterized by 1 MeV

electrons which may not be appropriate for proton-dominated orbits

Proposed Model

• Data needed– Protons

• E = 3 MeV•

= 1e10 to 1e13 p+/cm2

– Electrons• E = 1 and 5 MeV•

= 1e13 to 1e16 p+/cm2

NRL Displacement Damage Dose Method


108 109 1010 1011 1012 1013 1014 1015 1016 10170.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Nor

mal

ized

Max

imum

Pow



Protons

Electrons1 MeV5 MeV

3 MeV

25oC

NonIonizing Energy Loss

NIEL= Rate at which energy is lost to nonionizing events; analogous to LET or stopping power for ionizing events (UNIT=MeV/cm or MeVcm2/g)

)T( dmin

E)]d,E)L[T(,T(d

E),(dNIEL(E)

Differential scattering cross section for displacements

Recoil energy

Lindhard partition factor

NOTE: Energy dependence of NIEL similar to experimental RDCs

10-4 10-3 10-2 10-1 100 101 102 103

Particle Energy (MeV)

10-6

10-5

10-4

10-3

10-2

10-1

100

101

Si N

IEL

(MeV

cm2 /g

)

ProtonElectronNeutron

*Td = 21 eV

Si

10-4 10-3 10-2 10-1 100 101 102 103


10-6

10-5

10-4

10-3

10-2

10-1

100

101

GaA

s N

IEL

(MeV

cm2 /g

)


*Td = 10 eV, Ga & As GaAs

NIEL in Si and GaAs

*NIEL calculation available for any charged particle in any material

*Neutron NIEL determined from Displacement Kerma calculation

NIEL (MeVcm2/g) = KERMA (MeVmb) x (10-27NA /A)

•Calculated NIEL gives energy dependence of damage

coefficients (well‐matched to RDCs)

•Characteristic curves can be fit to simple expressions

(similar to JPL handbook method)

•Characteristic curve can be generated using minimal

ground test data (only 1 proton and two electron

energies)

Characteristic Curves

w/ NIEL

Measured Data

108 109 1010 1011 1012 1013 1014 1015 1016 10170.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Nor

mal

ized

Max

imum

Pow

er

1 Sun, AMOGaAs/Ge


Protons



25oC

1 MeV equiv. Neutrons

10-4 10-3 10-2 10-1 100 101 102 103


10-6

10-5

10-4

10-3

10-2

10-1

100

101

GaA

s N

IEL

(MeV

cm2 /g

)



108 109 1010 1011 1012


0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Nor

mal

ized

Max

imum

Pow

er 1 Sun, AM0T=25oC

GaAs/Ge

0.6 MeV1 MeV2.4 MeV12 MeV

Protons Electrons9.5 MeV3 MeV1 MeV0.5 MeV0.3 MeV0.2 MeV

1 MeV equiv.Neutrons

Dd(1 MeV electron)

Dd

Displacement Damage Dose Analysis of GaAs/Ge Solar Cells

0.70

0.75

0.80

0.85

0.90

0.95

1.00

1.E+07 1.E+08 1.E+09 1.E+10 1.E+11 1.E+12Dd(prot) and Dd(1 MeV elec) (MeV/g)

Nor

m V

oc

Proton Voc1 MeV Electron VocProton Voc data1 MeV Electron Voc data

GaAs/Ge

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0


Norm

Pm

ax

Proton Pmax1 MeV Electron PmaxProton Pmax data1 MeV Electron Pmax data

GaAs/Ge

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0


Nor

m Is

c

Proton Isc1 MeV Electron IscProton Isc data1 MeV Electron Isc data

GaAs/Ge

Dd Analyses of GaAs/Ge Solar Cells

x

ddDDlogC

P)D(P 11

0

value error value error value errorC 0.2904 0.0071 0.115 0.00317 0.229 0.0158Dx 1.10E+09 8.66E+07 1.15E+09 9.31E+07 2.52E+09 4.45E+08

value error value error value errorC 0.363 0.025 0.0745 0.0078 0.343 0.03Dx 6.90E+09 1.52E+09 1.28E+09 5.32E+08 1.10E+10 2.59E+09n 1.647 0.086 2.128 0.201 1.326 0.096

Protons

Electrons

Pmax Voc Isc

Pmax Voc Isc

1

n

refd )E(NIEL

)E(NIEL)E(NIEL)E()E(D

MJ Solar Cell Radiation Response in terms of Dd

*Experimentally determined variables (C, Dxp

, Dxe

, n)

*GaAs NIEL used in the correlation

108108 109 1010 1011


0.5

0.7

0.9

1.1

Nor

m P

mp

0

0.312.5101212

Emcore 3J Cells

Energy (MeV)

electron

proton C = 0.199Dx = 1.2x109 MeV/gn = 1.8Rep = 0.17

108108 109 1010 1011


0.5

0.7

0.9

1.1

Nor

m P

mp

0

0.312.5100.6112

Spectrolab EOL 3J Cells n/p cells

Energy (MeV)

electron

Data from Marvin 2000

protonC = 0.3Dx = 3x109 MeV/gn = 1.6Rep = 0.3

1

n

refd )E(NIEL

)E(NIEL)E(NIEL)E()E(D

x

ddDDlogC

P)D(P 11

0 xp

xeep D

DR

12 MeV2 MeV1 MeV0.6 MeVEffFitElec0.05 MeV0.1 MeV0.3 MeV1 MeV2.5 MeV10 MeVEffFitProt

Emcore ATJ Solar Cell Radiation Response

*This shows that the Emcore ATJ multijunction (3J) solar cell is

well‐

behaved in that indeed a single characteristic curve is generated from the

ground data. This data set is included in SCREAM development.

108 109 1010 1011 1012


0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Nor

mal

ized

Max

imum

Pow

er 1 Sun, AM0T=25oC

GaAs/GeDd(1 MeV electron)

0 10 20 30 40 50SiO2 Thickness (mil)

109

1010

1011

1012

Dd

(MeV

/g) (1 Year Mission)

GaAs

5000 km, circular, 60o

Incident and SDS (Isotropic) NonIonizing Energy Loss (2003)

Total Mission Dose

Pmax

Degradation

NRL Displacement Damage Dose Method

10-4 10-3 10-2 10-1 100 101 102 103

Proton Energy (MeV)107

108

109

1010

1011

1012

1013

1014

1015

1016D

iffer

entia

l Flu

ence

(cm

-2M

eV-1

)

3 mil

30 mil

12 mil

Uncovered

SiO2 Coverglass Thickness

5000 km, Circular Orbit60 Inclination5 year mission

10-4 10-3 10-2 10-1 100 101 102 103

Proton Energy (MeV)

10-3

10-2

10-1

100

101

GaA

s N

IEL

(MeV

cm2 /g

)

Proton


NRL Model Pros/Cons• Pros:

– Few ground test measurements needed (3)– Ground test particle energies can be conveniently chosen– Shielding algorithm is independent– Allows for rapid analysis of emerging cell technologies– Allows for easy trade studies– Can combine data from different experiments– Allows for alternate radiation particles (neutrons, alphas, etc.)

• Cons:– Lack of heritage (developed in the mid-1990s)– More suited for sufficiently thin devices (~few mm)

– Uniform damage deposition required over active region– Program currently not available to general public– No interlaboratory cross calibration of method

Analytical Model Comparison• Proton Dd to 1 MeV electron equivalent fluence

)1(NIELD*1

DD

1 xeCC

xp

dpelectron MeV 1

e

p

• 1 MeV electron fluence to equivalent proton Dd

1

D)1(NIEL*1DD p

e

CC

xe

electron MeV 1xpdp

*General rule of thumb: 1015 1 MeV e-/cm2 ~ 1010 MeV/g

Orbit Examples (SPENVIS)

1 year, 700 x 12050 km, 63.4o, w/ solar event protons

Fluence Spectra GaAs/Ge Cell degradation vs.


700 x 12050 km, 63.4o, ESP Total Fluence, 1 year

1.E+07

1.E+08

1.E+09

1.E+10

1.E+11

1.E+12

1.E+13

1.E+14

1.E+15

1.E+16

0.01 0.1 1 10 100 1000Energy (MeV)

Inte

gral

Flu

ence

(cm

-2) Trapped Protons

Solar ProtonsTrapped Electrons

700 x 12050 km, 63.4o, 1 year (Proton-dominated)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 5 10 15 20 25 30Coverglass Thickness (mils SiO2)

Nor

mal

ized

Rem

aini

ng P

aram

eter

Pmax JPLVoc JPLNRL Proton Dd PmaxNRL Proton Dd Voc

GaAs/Ge

Orbit Examples (SPENVIS)

15 year, GEO, ESP Total Fluence

1.E+07

1.E+08

1.E+09

1.E+10

1.E+11

1.E+12

1.E+13

1.E+14

1.E+15

1.E+16

1.E+17

0.01 0.1 1 10 100 1000Energy (MeV)

Inte

gral

Flu

ence

(cm

-2)

Trapped ProtonsSolar ProtonsTrapped Electrons

15 year GEO w/ solar event protons

Fluence Spectra GaAs/Ge Cell degradation vs.


GEO, 15 year (Electrons & Protons)

0.60

0.65

0.70

0.75

0.80

0.85

0.90

0.95

1.00

0 5 10 15 20 25 30Coverglass Thickness (mils SiO2)

Nor

mal

ized

Rem

aini

ng P

aram

eter

Pmax JPLVoc JPLNRL Proton Dd PmaxNRL Proton Dd Voc

GaAs/Ge

Why Do The Models Agree?

10-2 10-1 100 101 102

Energy (MeV)

10-1

100

101

102

103

Rel

ativ

e P m

ax D

amag

e C

oeffi

cien

t SJ GaAs/Ge2J InGaP/GaAs/Ge3J InGaP/GaAs/GeCIGSNIEL GaAsJPL MJ RDCs SRIM MJ RDCs

Protons

*Parameters normalized to value at 10 MeV

10-1 100 101 102

Energy (MeV)

10-3

10-2

10-1

100

101

102

Rel

ativ

e P m

ax D

amag

e C

oeffi

cien

t Electrons

*Parameters normalized to value at 1 MeV

SJ GaAs/Ge2J InGaP/GaAs/Ge3J InGaP/GaAs/GeCIGS1 MeV Equiv. NIEL GaAs (n=1.7)1 MeV Equiv. NIEL CIGS (n=2)

The energy dependence of the experimentally determined RDCs closely match that of the calculated NIEL.

DDD Implementation History• Quickbasic code (user proton spectrum input)

– Developed by NRL with Ed Burke (mid 1990’s)– Transformed into Mathcad (mid 1990’s to present)

• Enabled both proton and electron conditions in 2005

• NASA GRC support to Maxwell Labs (AP8/AE8)– 1st

version of SAVANT (late 1990’s)– Developed for inclusion into NASA Environmental Work Bench

• 1st

NASA Living With a Star (LWS‐SET) Data Mining NRA– NRL and NASA GRC created a FORTRAN‐based, stand‐alone, WindowsTM‐

based, version of SAVANT (2003) distributed by NASA MSFC (SEE Program,

now defunct)

– Only beta version produced (large laundry list created)• SCREAM (Solar Cell Radiation Environment Analysis Models)

– Transformed MATHCAD version into MATLAB‐based executable– Available by request

• Spenvis/Mulassis (MC‐SCREAM)– All of the components are there except solar cell damage info– The enabling component is MULASSIS to calculate the SDS– User interface created by Daniel Heynderickx (DHC, Consultancy)

SAVANT DDD Analysis Code

SAVANT: Solar Array Verification and ANalysis Tool (NASA, NRL, OAI)

SCREAM• Integral Radiation Spectrum Input (user input)

– Converts the log‐log spectral data into a 10th

order polynomial fit– Differentiates the polynomial fit to get the differential spectrum

• Shielding Range data– Protons and ions: SRIM 2006– Electrons: ESTAR

• Slowed‐Down Spectrum (SDS) Calculation– Slab model adjoint calculation based on Haffner model (1967)– Benchmarked with Space RadiationTM, Mulassis

• Calculation of DDD (integral of NIEL*SDS over energy)– NIEL data from WINNIEL2 (ions) and MATHCAD (electrons)

• Calculation of Parametric Degradation– Uses 5 empirically‐determined parameters for each metric

Start-Up Page

Materials Menu Options

Cell Technology Options

Degradation Parameter Options

After selecting HEO_TrP.txt and entering, the data are plotted

After choosing CMX and 6 mils, and pressing Slowed-Down Spectrum (“Primary Spectrum” button selected)

At this time, you can save the SDS using the FILE - Save SDS menu option

After pressing Displacement Damage Dose

MC-SCREAM• Incident differential radiation spectra

(SPENVIS)

– Can choose several proton, electron, and solar proton models• Calculation of the “slowed‐down”

spectra after having passed

through shielding

(analytical, MULASSIS)– Monte Carlo model– Several material layers can be used to describe cell

• Calculation of total DDD for the mission

(SPENVIS)– NIEL data from WINNIEL2 (ions) and MCAD (electrons)– Already implemented in Mulassis

• Determination of the expected cell degradation– Uses 5 empirically‐determined parameters for each metric– Will have several cell technologies included– User entry option for proprietary/developmental cells

Output:1.Slowed down spectra for each incident behind shielding2. DDD for each particle




Multijunction Solar Cell Radiation Response

InGaP2 GaAs Ge0.8 m 3 m 300 m

Monolithic 3J InGaP2 /GaAs/Ge (AM0, 1 sun ~28%)I1 I2 I3

Monolithic: Current-limiting, Icell =minimum(I1 ,I2 ,I3 )

•Whichever cell is the softest will control the overall cell performance.•MJ cell design can alleviate this to some extent by forcing the cell to degrade by the most radiation hard subcell.•The (beginning-of-life) BOL properties are slightly sacrificed in such a design for better (end-of-life) EOL behavior.

Ground Test MJ Solar Cell Data

108 109 1010 1011 10120.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

30 keV50 keV70 keV100 keV150 keV250 keV380 keV1 MeV2 MeV3 MeV5 MeV

Proton Energy


Rem

aini

ng F

acto

r of P

max

*T. Sumita, M. Imaizumi, S. Matsuda, T. Ohshima, A. Ohi, and T. Kamiya, Proc. 19th EPVSEC, Paris, 2004.

Dd = NIEL x

Another good example of NIEL correlation!

Pmax Proton Degradation vs. Displacement Damage Dose (Dd )

108 109 1010 1011 10120.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0


Proton Energy


Rem

aini

ng F

acto

r of P

max


Two separate damage curves are apparent!

Which damage curve will apply in space?

InGaP

GaAs

Ground Test MJ Solar Cell DataPmax Proton Degradation vs. Displacement Damage Dose (Dd )

300 500 700 900 1100 1300 1500 1700 19000.0

0.2

0.4

0.6

0.8

1.0

Wavelength (nm)

Qua

ntum

Effi

cien

cy

Solid lines: UnirradiatedDashed lines: 1x1012 p+/cm2

50 keV Protons

InGaP/GaAs/Ge

300 500 700 900 1100 1300 1500 1700 19000.0

0.2

0.4

0.6

0.8

1.0

Wavelength (nm)

Qua

ntum

Effi

cien

cy


100 keV Protons

InGaP/GaAs/Ge

300 500 700 900 1100 1300 1500 1700 19000.0

0.2

0.4

0.6

0.8

1.0

Wavelength (nm)

Qua

ntum

Effi

cien

cy


400 keV Protons

InGaP/GaAs/Ge

300 500 700 900 1100 1300 1500 1700 19000.0

0.2

0.4

0.6

0.8

1.0

Wavelength (nm)

Qua

ntum

Effi

cien

cy


1 MeV Protons

InGaP/GaAs/Ge

Proton-Induced QE Degradation in MJ Cells

50 keV protons

100 keV protons

400 keV protons

1 MeV protons

SRIM 2003.26 Simulation

InGaP2 GaAs Ge0.8 m 3 m 300 m

• 3J InGaP2 /GaAs/Ge structure• 3 Proton Energies

– 63.1 keV (0.6 m range)– 251 keV (2 m range)– 1 MeV (11 m range)

• Monoenergetic, normal incidence• Monoenergetic, omnidirectional• Spectrum, omnidirectional• Spectrum, non-omnidirectional

“TRIM.DAT” SRIM input file

(“normal” SRIM run)

1.E-02

1.E-01

1.E+00

1.E+01

1.E-02 1.E-01 1.E+00 1.E+01 1.E+02Depth (um)

Ener

gy A

bsor

bed

by R

ecoi

ls (k

eV/u

m) 63 keV (norm.)

251 keV (norm.)1 MeV (norm.)

GaAs GeInGaP

108 109 1010 1011 10120.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0


Proton Energy


Rem

aini

ng F

acto

r of P

max

Monoenergetic, Unidirectional Irradiations

• Typical ground test conditions

(not space conditions)• Nonuniform vacancy distribution – Bragg Peak at end of track• Different energies can preferentially degrade one sub‐junction• This effect is not seen in 1 MeV electron irradiation (longer ranges)

InGaP degradation

GaAs degradation


*Results from SRIM 2003 v.26 (www.srim.org)

Monoenergetic, Omnidirectional Irradiation

• More uniform vacancy distribution than a unidirectional beam• Bragg peak not seen

*Results from SRIM 2003 v.26 using special input file (TRIM.DAT) which specifies random incident angles (via direction cosines) over 2

geometry1.E-02

1.E-01

1.E+00

1.E+01

1.E-02 1.E-01 1.E+00 1.E+01 1.E+02Depth (um)

Ener

gy A

bsor

bed

by R

ecoi

ls (k

eV/u

m)

63 keV (omni.)251 keV (omni.)1 MeV (omni.)

GaAs GeInGaP

Energy Spectrum, Omnidirectional Irradiation

• Representative of exposure in the space radiation environment• The vacancy distribution profile is nearly uniform over active region

*Results from SRIM 2003 v.26 using special input file (TRIM.DAT) which specifies random incident angle and energy to simulate HEO spectrum (3 mil SiO2 )

No special effects due to low energy protons apparent!

1.E-03

1.E-02

1.E-01

1.E+00

1.E-02 1.E-01 1.E+00 1.E+01 1.E+02Depth (um)

Ener

gy A

bsor

bed

by R

ecoi

ls (k

eV/u

m)

HEO orbit,3 mils SiO2

GaAs GeInGaP

4 mil CMG Coverglass

2 mil DC 93-500 Adhesive

11 m IMM (assumed Ge)5 m Ag Back Metallization

2 mil GE566 Adhesive

2 mil Kapton

Assumed Structure for Monte Carlo N-Particle eXtended (MCNPX) Simulation

Blanket Structure for Monte Carlo Simulation

Monte Carlo (MCNPX) simulation of Tacsat4 Radiation Environment on IMM blanket structure

TACSAT4 Omnidirectional Proton Irradiation (Entire Structure)

Top-Side RadiationBottom-Side Radiation

4 mil CMG

2 mil DC93-500

11 m IMM5 m Ag

2 mil Kapton

2 mil GE566

Ionizing Energy Deposition In Blanket

Energy Deposition vs. Depth in Blanket

cell

Irradiation from bottom

Irradiation from top

Bottom-SidedTop-SidedSum

Effect of Thin Shielding on Spectrum IrradiationOmnidirectional Irradiation

0.0

0.1

0.2

0.3

0.4

0.5

0.6

1.E-02 1.E-01 1.E+00 1.E+01Depth (um)

Ener

gy G

oing

to R

ecoi

ls (k

eV/u

m/io

n)

No shielding1 um ZnO10 um SiO23 mils SiO2

Circular, 5093 km, 57o ,1 year Orbit

(3 mils ~ 75 m)

Documents

Scott R. Messenger, Ph.D. - oma · • Provide space solar cell community an alternate method for cell level degradation analyses • Heritage method (JPL EQFLUX) requires several