IEEE NSS 2013

IEEE NSS 2013 27 October – 2 November 2013

Seoul, Korea

T. Basaglia1, M. Batic2, M. C. Han3, G. Hoff4, C. H. Kim3, H. S. Kim3, M. G. Pia5, P. Saracco5

1CERN

2Sinergise, Ljubljana, Slovenia 3Hanyang University, Seoul, Korea

4Pontificia Universidade Catolica do Rio Grande do Sul, Porto Alegre, Brazil 5INFN Genova, Italy

Physics methods for the simulation of photoionisation

IEEE NSS 2013

Rationale

2

State-of-the-art simulation of photon interactions

Elastic scattering : Published Photoelectric effect Compton scattering

pair production : first results

New theoretical calculations and parameterisations were

released recently.

à Are they accurate than old one?

New physics Very low energy(~eV scale) & micro/nano dosimetry

ex. FLUKA, Geant4-DNA, MOCA, OREC/NOREC, PARTRAC, Penelope, PTB-code,

Trion etc.

à How accurate are models?

New trends

The simulation of photon physics is

well-established.

à Is there any quantitative validation?

Simulation physics

Project to validate a wide set of simulation modeling options against a large collection of experimental data

IEEE NSS 2013 3

Cross sections Angular distribution EGS5 PHOTX Sauter

EGSnrc Storm-Israel Fit to XCOM EPDL97 (subshell)

Sauter

FLUKA EPDL97 Sauter

Geant4 Revised Biggs-Lighthill (Henke) EPDL97

Sauter-Gavrila Same direction

ITS Scofield 1973 Fischer+Sauter

MCNP(X) EPDL89, EDPL97 ENDFB/IV+Storm-Israel

Penelope EPDL97 Sauter (K shell)

Photoionisation in Monte Carlo codes

IEEE NSS 2013

lowenergy::G4LivermorePhotoElectricModel

utils::G4VEmModel

lowenergy::G4LivermorePolarizedPhotoElectricModel

lowenergy::G4PenelopePhotoElectricModel

standard::G4PEEffectFluoModel

polarisation::G4PolarizedPEEffectModel

polarisation::G4VPolarizedCrossSection

polarisation::G4PolarizedPEEffectCrossSection

lowenergy::G4PhotoElectricAngularGeneratorSauterGavrila

utils::G4VEmAngularDistribution

lowenergy::G4PhotoElectricAngularGeneratorSimple

lowenergy::G4PhotoElectricAngularGeneratorPolarized

Photoelectric effect

Geant4 9.6MGP 9/9/2013 reverse engineered

standard::G4PhotoElectricEffect

- isInitialised :G4bool

G4VDiscreteProcessutils::G4VEmProcess

utils::G4VAtomDeexcitation

-anglModel-currentModel

-fAtomDeexcitation-fAtomDeexcitation

-fAtomDeexcitation

-fAtomDeexcitation

4

Sauter-Gavrila

Sauter-Gavrila

same as incident γ

Base class for atomic deexcitation

Penelope 2008 EPDL97

polarized EPDL97

polarized Livermore EPDL97

Biggs-Lighthill Em models

Em process

Packages •  lowenergy •  polarisation •  standard •  utils

Photoionisation in Geant4 9.6

IEEE NSS 2013 5

Year Compilation Energy Z (sub)Shell Method 1967-1988 Biggs-Lighthill 10 eV – 100 GeV 1-100 - parameterised

1992 Brennan-Cowan 30 eV – 700 keV 3-92 - tabulated

2000 Chantler 10 eV – 433 keV 1-92 K tabulated

2003 Ebel 1 keV – 300 keV 1-92 all parameterised

2002 Elam 100 eV – 1 MeV 1-98 - tabulated

1997 EPDL97 (Scofield) 10 eV – 100 GeV 1-100 all tabulated

1982-1993 Henke 10 eV – 30 keV 1-92 - tabulated

1970-2006 McMaster/Shaltout 1 keV – 700 keV 1-94 - tabulated

1989 PHOTX (Scofield) 1 keV – 100 MeV 1-100 tabulated

2001 RTAB 10 eV – 30 keV 1-99 all tabulated

1973 Scofield 1 keV – 1.5 MeV 1-100 all tabulated

1970 Storm-Israel 1 keV – 100 GeV 1-100 - tabulated

1973 Veigele 100 eV – 100 MeV 1-94 - tabulated

1987-2010 XCOM (Scofield) 1 keV – 100 GeV 1-100 - tabulated

e.g. Chantler’s exchange potential in his DHF calculation is different from Scofield’s

Different methods and calculations

Cross section sources

IEEE NSS 2013

" Evaluate a large number of available modeling options

" Suitable for use in Monte Carlo simulation codes ‒  Tabulated theoretical calculations

‒  Simple analytical formulations, with documented parameters

" All options evaluated in the same computational environment ‒  Minimize dependencies on other software parts (not always components)

" Quantitative, objective evaluation based on statistical methods

" Establish state-of-the-art for the simulation of photoionisation on objective ground

" Computational performance measured along with physical accuracy

6

Strategy

※ We only focus on the simulation with non-polarised photons case.

IEEE NSS 2013

TCrossSectionTFinalState

G4TPhotoionisation

G4CsTabula G4FsPhotoionisation G4IPhotoelectronGenerator

G4PhotoelectronSauter G4PhotelectronSauterGavrila G4PhotoelectronSimple

G4VProcessprocesses-management::G4VDiscreteProcess

G4AtomDeexcitation

Photoionisation

Strategy pattern

First design iterationMGP January 2013

or G4CsPhotoIoniBiggs,or G4CsPhotoIoniEbel

«bind»«bind»

7

Streamlined software design consistent with Geant4 kernel

Policy-based class design (à la Alexandrescu, Modern C++ design, 2001) §  minimize dependencies §  lightweight unit tests for validation

Sharp domain decomposition Clearly identified responsibilities

No duplication of code nor of functionality

Computational environment

IEEE NSS 2013

" Collected from the literature ‒  Total cross sections ‒  Partial cross sections ‒  Angular distributions

" Data types ‒  Pure experimental cross sections: direct measurements ‒  Semi-empirical cross sections: involve theoretical manipulations ▻  e.g. subtraction of calculated scattering contribution (Compton and elastic)

" Format ‒  Tables, text ‒  Figures: digitized, digitization error estimated

" Evaluation of experimental data ‒  Systematic effects: identified whenever possible ‒  Outliers

8

> 150 references > 5000 data points ~ 3700 σ total ~ 1400 σ shell

Experimental data

IEEE NSS 2013

Systematic effect?

9

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

EPDL E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviationsFr

actio

n

−10 −7 −4 −1 1 3 5 7 9

Chantler E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

Henke E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

RTAB E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

Scofield E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

Penelope E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

PHOTX E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

Storm E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

XCOM E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

VeigeleL E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

VeigeleH E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

Elam E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

Brennan E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

Biggs E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

McMaster E > 1 keVexp.semiemp.

0.0

0.1

0.2

0.3

Number of standard deviations

Frac

tion

−10 −7 −4 −1 1 3 5 7 9

Ebel E > 1 keVexp.semiemp.

Difference between calculated and

“experimental” total cross sections,

expressed in terms of number of standard

deviations: pure experimental

and semi-empirical

data

Only pure experimental data used in the

validation process

IEEE NSS 2013

" Two-stage statistical analysis 1.  Compatibility of each cross section calculation method with experiment 2.  Comparison of compatibility with experiment across modeling categories

" Quantitative appraisal of capabilities and differences

10

Compatibility with experiment

Goodness-of-fit test χ2 test α = 0.01

α  ≥ 0.01 pass α  < 0.01 fail

Difference across categories

Contingency tables Fisher exact test

Barnard test Pearson χ2 test

α = 0.05

as appropriate

Data analysis method

𝐞𝐟𝐟𝐢𝐜𝐢𝐞𝐧𝐜𝐲=   𝐍↓𝑝𝑎𝑠𝑠 /𝐍↓𝑡𝑒𝑠𝑡  𝑐𝑎𝑠𝑒𝑠  

IEEE NSS 2013

" Most calculation methods exhibit similar compatibility with experiment for E>250 eV ‒  Chantler, Brennan-Cowan look worse

" Degraded accuracy below 250 eV

11

●

●

●

●● ● ●

●

●

● ●

●

●

● ●●

0.0

0.2

0.4

0.6

0.8

1.0

Cross section model

Effic

ienc

y

EPD

LC

hant

ler

Hen

keRT

ABSc

ofie

ldPe

nelo

pePH

OTX

Stor

mXC

OM

Veig

eleL

Veig

eleH

Elam

Bren

nan

Sand

iaM

cMas

ter

Ebel

● E>250 eVE<250eVpreliminary

Analysis of contingency tables EPDL

Chantler EPDL

Brennan-Cowan Fisher 0.044 0.011 Pearson χ2 0.033 0.007 Barnard 0.035 0.007

●●●●

●●●●●●●●●●

●●●●●

●●

●●

● ● ● ● ●

0.014 0.016 0.018 0.020

0

2

4

6

8

E (keV)C

ross

sec

tion

(M

b)

●

●

●

●

H, Z=1Beynon1965Beynon1966Kohl1978Palenius1976EPDLRTABChantlerBiggs

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●● ●

0.02 0.04 0.06 0.08 0.10

0

5

10

15

20

E (keV)

Cro

ss s

ectio

n (

Mb)

●●●●●●●●●●●●●

●●

●●

●●

●

●●●●

●

●●●●

●

●

●

●●

●

●

●●●●

●●

●

●

●

O, Z=8Cole1978aAngel1988Cairns1965Samson1985EPDLRTABChantlerHenkeBrennanBiggsH O

●

●●●●●●

●

●

●●

●

●●●

●●

●●●

5 10 15 20 25

0

10

20

30

40

E (keV)

Cro

ss s

ectio

n (

Kb)

●

●

●

●

●

●

●

Fe, Z=26Dachun1992Murty1998DelGrande1986EPDLRTABChantlerStormHenkeVeigele HElamBrennanBiggsEbelFe

E<250 eV Results - Total cross sections

E>250 eV

E (keV) E (keV) E (keV)

Cro

ss s

ectio

n (k

b)

Cro

ss s

ectio

n (M

b)

Cro

ss s

ectio

n (M

b)

IEEE NSS 2013 12

● ●●●●●

●

●●●●●●●●●●

●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●

30 40 50 60 70

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Z

Cro

ss s

ectio

n (

Kb)

●●●●●●●●●●●●●

●

●●

●

●

●

●●

●●

●

●

●

59.54 keVArora1981Karabulut2005Ertugrul2003EPDLRTABscaled RTABChantlerEbel

●●●●●●●●●●●●●

●●●●●●●●●●●

●●

●●●●

●●

●●

●●

40 50 60 70 80 90

0

100

200

300

400

500

600

700

Z

Cro

ss s

ectio

n (

b)

●

59.54 keVKarabulut2005Karabulut2002EPDLRTABscaled RTABEbel

●●

●

●●

●●

●●

●●

●●

●

0.7 0.8 0.9 1.0

0.0

0.5

1.0

1.5

E (keV)

Cro

ss s

ectio

n (

Mb)

●

Xe, Z=54Becker1987EPDLRTABscaled RTAB

● ●● ●●●●●●●●●●●●●●●● ●●●●●●●●●

0.07 0.08 0.09 0.10 0.11 0.12

0.0

0.2

0.4

0.6

0.8

1.0

1.2

E (keV)

Cro

ss s

ectio

n (

Mb)

●

Ba, Z=56Bizau1989EPDLRTAB

shell EPDL Chantler RTAB scRTAB Ebel K 0.209 0.350 <0.001 0.315 <0.001 L1 0.075 <0.001 0.069 0.964 L2 0.339 <0.001 0.299 0.154 L3 1 <0.001 1 1 M1 <0.001 <0.001 <0.001 M4 0.031 <0.001 <0.001 M5 <0.001 <0.001 <0.001 N1 <0.001 <0.001 <0.001 N6 <0.001 <0.001 <0.001 <0.001 N7 <0.001 <0.001 <0.001 <0.001 O1 <0.001 <0.001 <0.001 <0.001 O2 <0.001 <0.001 <0.001 <0.001 O3 <0.001 <0.001 <0.001 <0.001 P1 <0.001 <0.001 <0.001 <0.001

p-value χ2 test

Systematic effect observed with RTAB shell cross sections

(presumably a missing factor in the calculation)

Calculated inner shell cross sections compatible with experiment

Outer shell cross sections inconsistent with experimental data Beware: small data sample, limited data

sources

K

L3

M4

O1

Results - Shell cross sections

IEEE NSS 2013 13

0 30 60 90 120 150 180

0.0

0.2

0.4

0.6

0.8

1.0

Θ ang le  (deg rees )

Normarilz

ed  cross

 sec

tion

 E xp.  G 4P ola r  G 4S auter  G E ANT  3

A luminiumK -­‐s hell1170  keV

Qualitative appraisal Limited experimental sample

Experimental systematic effects (corrected/uncorrected data)

Option à la GEANT 3 (Sauter) evaluated along with other Geant4 options

Results - Angular distribution

0 30 60 90 120 150 180

0.0

0.2

0.4

0.6

0.8

1.0

Θ ang le  (deg rees )

Normarilz

ed  cross

 sec

tion

 E xp.  G 4P ola r  G 4S auter  G E ANT  3

G oldL 2-­‐s hell412  keV

0 30 60 90 120 150 180

0.0

0.2

0.4

0.6

0.8

1.0

Θ ang le  (deg rees )

Normarilz

ed  cross

 sec

tion

G oldL 3-­‐s hell412  keV

 E xp.  G 4P ola r  G 4S auter  G E ANT  3

0 30 60 90 120 150 180

0.0

0.2

0.4

0.6

0.8

1.0  E xp.  G 4P ola r  G 4S auter  G E ANT  3

K ryptonM1-­‐s hell1.2536  keV

Θ ang le  (deg rees )

Normarilz

ed  cross

 sec

tion

IEEE NSS 2013

" Large scale effort to evaluate quantitatively physics methods for photoionisation simulation ‒  Part of a wider project for quantitative assessment of state-of-the-art

simulation of photon interactions

" Total cross section ‒  Most calculation methods exhibit similar behaviour ‒  More recent calculations (Chantler, Brennan-Cowan) do not appear more

accurate than old Scofield’s 1973 (unrenormalized) " Inner shells

‒  EPDL, (corrected) RTAB appear equivalent, Ebel’s parameterisation inconsistent with experimental K shell data

" Outer shells ‒  No calculation method appears adequate to reproduce experimental data

" Photoelectron angular distribution ‒  Scarce data and experimental systematics prevent a quantitative discrimination

14 All results will be documented in detail in a forthcoming publication

Conclusion

IEEE NSS 2013

…a big THANK YOU to the CERN Library!