Atomic data: state of the art and future perspectives - PowerPoint Presentation

Slide1

Atomic data: state of the art and future perspectives

Jelle Kaastra

with

Ton

Raassen

,

Liyi

Gu

,

Junjie

Mao,

Igone

Urdampilleta

,

Missagh

Mehdipour

SRON & Leiden University

Slide2

Introduction

X-ray emitting plasmas everywhere:

Solar system to cosmic web filamentsBroad range environments & physical conditions:collisional ionised, photo-ionised, transiently ionisedHigh resolution X-ray spectroscopy key to understand these sourcesNext year launch ASTRO-H with SXS calorimeterOld models not always most up to date atomic physicsNeed tool to model these different sources with same, consistent set atomic parameters

2

Slide3

Spectroscopic codes @ SRON

Short history:

1972

Mewe1975 Mewe-Gronenschild1985 Mewe-Gronenschild- Van den Oord

1990

Meka

1994 Mekal1992 SPEX1992 Version 12001 Version 22015 Version 3 (expected release December)Evolution from plasma model to full astrophysical model including data analysis (fitting), plotting & diagnostic output

Slide4

Need for updates

Example:

Mewe code (still core present SPEX models) approximates radiative recombination contribution to lines by local power-lawOkay for CIE but:Large deviations for recombining / ionising plasma

T

max

log T 

l

og rate



4

Slide5

Requirements for updates

Code must allow options for fast calculation yet accurate enough



Minimise number of mathematical operations & data storage for the cross sections/rates follow original strategy of Mewe: simple, accurate & fast approximations, but more accurate & complete than beforeRestrict to Z ≤ 30 (astrophysically most relevant)5

Slide6

Updates to atomic data

For full model, need updates of many processes:

Collisional & photo-

ionisation cross sectionsTransition probabilitiesAuto-ionisation ratesRecombination ratesLine energiesEtc.6

Slide7

Comparison of codes

(with

Junjie

Mao)7

Wavelength (

Å

)Wavelength (Å)SPEX V 3.0βATOMDB V3.0.2

Log T (K)

Fe

O

Slide8

Collisional

ionisation

(with

Igone Urdampilleta)8

Slide9

Motivation

In the past, several compilations

Recent one:

Dere 2007Almost always give total ionisation ratesSubshells needed for inner-shell line emissionNew data published since 2007 Revisit collisional ionisation rates9

Slide10

Collisional

ionisation

for atoms and ions of H to Zn

Direct ionization cross section fitting procedure:

Relativistic correction (Quarles 1976 and

Tinschert

et al. 1989):Excitation Autoionization fit (Mewe 1972):

where,

Ee

electron

energy

I

ionisation

potential

A, B, D, E fit parameters

C Bethe constant

R relativistic correction

Slide11

Examples of fits to collisional

ionisation

cross sections

11QI2 (10-24 m2keV2)

1s

2s

Note Dimensionless scaling 

Slide12

12

QI

2

(10-24 m2keV2)Relativistic correction 



Note dimensionless scaling I ~ Z2 for 1s shell H-sequence

Slide13

-

Cl

-

Ar

-K

-

Ca

-

Sc

-Ti

-V

-Cr

-

Mn

-Fe

-Co

-Ni

-Cu

-Zn

-

Cl

-

Ar

-K

-

Ca

-

Sc

-Ti

-V

-Cr

-

Mn

-Fe

-Co

-Ni

-Cu

-Zn

-

Cl

-

Ar

-K

-

Ca

-

Sc

-Ti

-V

-Cr

-

Mn

-Fe

-Co

-Ni

-Cu

-Zn

-

Cl

-

Ar

-K

-

Ca

-

Sc

-Ti

-V

-Cr

-

Mn

-Fe

-Co

-Ni

-Cu

-Zn

Slide14

-

Cl

-

Ar

-K

-

Ca

-

Sc

-Ti

-V

-Cr

-

Mn

-Fe

-Co

-Ni

-Cu

-Zn

Slide15

Radiative

recombination

(with

Junjie Mao)15

Slide16

Radiative recombination

Need

individual rates

to different excited shells for calculation of line spectrumAlso need cooling rate associated to the recombination (kinetic energy captured electron averaged over the recombination rate)Start with hydrogen-like systems16

Slide17

RR

C

ross Section

PI

Cross Section

Storey&Hummer (1991)

EXACT

AUTO STRUCTURE

Badnell (2006)

FAC Gu (2003)

cf

Analytic

Free e

-

distribution

Milne relation

RR

rates

Parameterisation

17

R(T) = a

0

T

-b0-c0logT

(1+a

2

T

-b2

) / (1 + a

1

T

-b1

)

Slide18

Fitting accuracy

Vast majority: accurate within few %

Very limited number outliers

Usually unimportant transitionsExample: C I n=5 1D2 level, still ~15% accuracy18

Slide19

Photoionised

plasmas

(with

Missagh Mehdipour)19

Slide20

Obscuration in NGC 5548

20

Slide21

21

Slide22

Differences photoionisation

models

(NGC 5548 obscured case)

22Ξ = Fion / nkTc

Slide23

Total radiative

recombination rates

23

Seaton approximation: simple analytic form for low to intermediate T

Fails at higher T

Previously widely used (e.g. Arnaud &

Rothenflug 1985 balance)

Slide24

Effects of update for photoionised

plasmas

24

Slide25

He-like triplets and absorption

(with

Missagh

Mehdipour)25

Slide26

He-like R-ratio in Active Galactic Nuclei

(

Seyfert

2 galaxies)26Theory: Porquet & Dubau

(2000)

Landt

et al. 2015, observations

Slide27

27

1s

2

2s1s2s(1S)2p 2P

1s

2

2s1s2s(3S)2p 2Pw

z

x

,y

Slide28

28

Slide29

29

Slide30

Line broadening

30

Slide31

Line broadening

Thermal Doppler broadening

Turbulent broadening

Natural broadening?For Fe-K, FWHM 0.2-1.0 eV (e.g. Brown et al.)Corresponds to 10-50 km/s Need Voigt profiles31Thermal broadening only

Slide32

32

Slide33

Charge transfer modeling

see talk this afternoon by

Liyi

Gu33

Slide34

Conclusions

Astrophysical sources sometimes found in remarkable areas of

parameter space

New X-ray missions like ASTRO-H (launch 2016) demand more detail & accuracy (but also Chandra & XMM-Newton benefit)Work in progress: update atomic parameters in X-ray spectral models to account for thisSPEX ( www.sron.nl/spex ) Version 3 will contain these updates (release late 2015)34