Detailed Study of the X-ray Absorption in the ISM

Transcription

1 Detailed Study of the X-ray Absorption in the ISM Javier García (Harvard-Smithsonian CfA) Efrain Gatuzz (IVIC, Caracas MPA), Claudio Mendoza (IVIC, Caracas), Tim Kallman (NASA-GSFC), Tom Gorczyca (WMU), Lia Corrales (MIT) Special thanks to Mike Nowak, Jörn Wilms & Norbert Schulz The Universe in High-resolution X-ray Spectra Chandra Workshop Cambridge, MA August 2th, 215

2 Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

3 Motivation The study of the physical and chemical properties of the interstellar medium (ISM) provides important information about the formation and evolution of the galaxies. High-energy photons interact with the ISM via scattering (by dust and grains) and/or absorption (excitation and ionization of inner K-shell electrons, mainly). Thus, X-ray observations of the ISM provide the opportunity to study the abundances and ionization fractions for a large number of elements. Previous studies on ISM absorption features in X-ray binary spectra have only used functional models to fit the spectrum (Paerels+1, Schulz+2, Takei+2, Juett+4, Liao+13) Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

4 Interstellar Medium Multiphase Structure Phase Component Temp. (K) Constituents Dust 1 2 MgSiO 3,... Cold Molecules 1 2 H 2, CO,... Neutral Gas 5 1 H i, O i,... Warm Neutral Gas 1 4 H i, O i,... Ionized Gas 1 4 H ii, O ii-v,... Hot Ionized Gas 1 6 O vi-viii, Ne ix,... X-ray absorption can provide signatures of the binding of many elements in molecules or solids; the inner-shell electronic transitions are key diagnostics since the ionization state or chemical binding shifts the line energies by a predictable amount (e.g. devries+costantini+9, Pinto+1, Costantini+12). Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

5 Oxygen Atomic Data High-energy photoionization crosssections of O ions showing the structure of the K-edge. In black the Breit Pauli R-Matrix calculation by García+5, in red those by Pradhan+3, and in green by Reilman+Manson79. The observed flux can be approximated as F (E) = F exp [ N OI OI (E)] where F is a normalization factor, N OI is the oxygen column density, and OI (E) is the photoabsorption cross section for neutral oxygen. Photoabsorption Cross Section (MBarn) O I O III O IV O V O VI Photon 62 Energy 64 (ev) O II (García+5) Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

6 Neutral Oxygen Atomic Data Photoionization cross sections for neutral oxygen from McLaughlin+Kirby98 (blue curve), Gorczyca+McLaughlin (red curve), and García+5 (black curve). This spectral region covers both the absorption K-edge and the Kα absorption line (1s-2p) from O i. All the curves have been convoluted with a.182 ev FWHM Gaussian. 1 1 Cross section (Mb) 1.1 Discrepancies in the Kα position also seen with respect to laboratory data! (Stolte+13, McLaughlin+13, Bizau+15) See B. McLaughlin s Poster (García+11) Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

7 XTE J Chandra Spectrum Spectral fit of the Chandra MEG observations of XTE J in the oxygen absorption region (21 25 Å) using a powerlaw*warmabs physical model. 1. Normalized counts s 1 Å χ (Gatuzz+13) Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

8 Benchmarking Atomic Data: Oxygen O I O II O III Stolte O VII Kα O VI Kα O II Kγ O II Kβ O I Kγ O I Kβ O I Kγ O I Kβ } O III Kα O II Kα O I Kα Cross Section (Mb) O i shift: 29 må O ii shift: 75 må Lab measurement also needs to be shifted! Theoretical photoabsorption cross sections for O i (red), O ii (green), and O iii (blue) computed by García+5 which are implemented in the warmabs model. The black solid line is the laboratory measurement by Stolte+97. (Gatuzz+13) Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

9 Benchmarking Atomic Data: Oxygen Normalized counts s -1 Å -1 χ O VII Kα O VI Kα O II Kγ XTE J O II Kβ O I Kγ O I Kβ O I Kγ O I Kβ O III Kα O II Kα O I Kα Significant improvement in the fit: χ 2 = O VI and O VII are fitted with Gaussian profiles (intrinsic to the source) Spectral fit of the Chandra MEG observations of XTE J in the oxygen absorption region (21 25 Å) using the corrected powerlaw*warmabs physical model. (Gatuzz+13) Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

10 Benchmarking Atomic Data: Neon Ne IX Kα Ne III? Ne III Kβ Ne II? Ne II? Ne II Kγ Ne II Kα Ne I Kβ Ne III Kα Ne II Kα Ne IX Kα Ne III? Ne III Kβ Ne II? Ne II? Ne II Kγ Ne II Kα Ne I Kβ Ne III Kα Ne II Kα Photons cm -2 s -1 Å Cyg X-2 Photons cm -2 s -1 Å XTE J Chandra MEG flux spectra of the X-ray binaries Cygnus X-2 and XTE J simultaneously fitted with a power-law continuum and several Gaussian profiles. Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

11 Benchmarking Atomic Data: Neon Shifts to the Kα Resonances: Ne II = 3.2 må, Ne III = 15.7 må 1 2 Ne I Ne II Ne III 1 2 Ne I Ne II Ne III Cross Section (Mb) Cross Section (Mb) Shifts to the Cross Sections: Ne I = 11 må, Ne II = +7.6 må, Ne III = 14.7 må Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

12 ISMabs: An X-ray Interstellar Absorption Model I (E) Obs = I (E) source exp ( τ) τ = i i(e)n i, where i (E): Photoelectric Cross Section and N i : Column Density of i th ion..1.9 TBnew ISMabs.6.5 TBnew ISMabs.8 Photons cm -2 s -1 Å Photons cm -2 s -1 Å Ne K region.1 O K region χ χ (Gatuzz, García et al. 215) Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

13 Analysis of All Available Sources Chandra XMM-Newton Distance (kpc) GC Sun Distance (kpc) Selection of 24 bright sources, 17 from Chandra and 15 from XMM-Newton. A total of 84 single observations were analyzed. In the case of Chandra, 2 observations were taken in TE-mode, and 29 in CC-mode. Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

14 TE- and CC-Modes in Chandra ISMabs column densities for Chandra observations versus number of counts. The CC-mode data shows large discrepancies when counts are low compared to the TE-mode data (black points). 1 Column Density (1 17 cm -2 ) 1 1 OI Column Density (1 17 cm -2 ) 1 1 NeI Column Density (1 17 cm -2 ) 1 FeI Edge Counts Edge Counts Edge Counts Oxygen columns from CC-mode data with less than 1 3 counts are unreliable Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

15 TE- and CC-Modes in Chandra: 4U Photons cm -2 s -1 Å Chandra TE-mode Photons cm -2 s -1 Å Chandra CC-mode Photons cm -2 s -1 Å XMM-Newton Possible Background contamination at the O-K edge! Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

16 Fits to the Brightest Sources The Astrophysical Journal, 8:29(13pp),215February1 Gatuzz et al. Normalized counts s -1 Å XTE J Normalized counts s -1 Å Sco X Normalized counts s -1 Å Cygnus X-2 Normalized counts s -1 Å Cygnus X Figure 4. ISMabs best broadband fits (solid red lines) for the XTE J , Sco X 1, Cygnus X 2, and Cygnus X 1 sources. For each source the observations have been combined for illustrative purposes. Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

17 Derived Column Densities: Hydrogen and Oxygen 36 XTEJ SerX-1 ScoX-1 J J J GX9+9 GX349+2 GX339-4 GS GROJ EXO CygX-2 CygX-1 AqlX-1 4U U U U U U U U U Column Density (1 21 cm -2 ) Hydrogen Column Density (1 17 cm -2 ) Oxygen Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

18 4U Hydrogen Derived Column 5 Densities: Neon and Iron Column Density (1 21 cm -2 ) Column Density (1 17 cm -2 ) Oxygen XTEJ SerX-1 ScoX-1 J J J GX9+9 GX349+2 GX339-4 GS GROJ EXO CygX-2 CygX-1 AqlX-1 4U U U U U U U U U Column Density (1 17 cm -2 ) Neon Column Density (1 17 cm -2 ) Iron Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

19 Hydrogen Column Densities 21cm (1 21 cm -2 ) DIC+9 KAL+5 WIL Column Density (1 21 cm -2 ) ISMabs H columns systematically larger than the 21-cm measurements by Dickey Fig. & Lockman 29. Hydrogen (199) column and Kalberla densities et comparison. al. (25). Y axis Better correspond agreement to Dickey with Willingale & Lockman et al. (213). But continuum modeling influences this trend! (199) while X axis correspond to ISMabs fit results. Solid lines correspond to fit functions Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

20 Spatial Variations of the Columns 4 36 Oxygen 4 36 Oxygen 4 36 Oxygen Column Density (1 17 cm -2 ) Column Density (1 17 cm -2 ) Column Density (1 17 cm -2 ) Column Density (1 17 cm -2 ) Neon Column Density (1 17 cm -2 ) Neon Column Density (1 17 cm -2 ) Neon Column Density (1 17 cm -2 ) Iron Column Density (1 17 cm -2 ) Iron Column Density (1 17 cm -2 ) Iron Distance (kpc) Latitude Longitude Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

21 Oxygen and Iron Depletion 5 Ismabs fit GRE+98 JUE+6 N O (1 17 cm -2 ) 1 N O = ( AO A Ne ) N Ne N Fe = ( AFe A Ne ) N Ne N Fe (1 17 cm -2 ) Ismabs fit GRE+98 JUE+6 ISMabs columns agree with those from Juett et al. (26) Neon does not combine (noble gas) Depletion of oxygen and iron with respect to the Solar values of Grevesse & Sauval (1998) N Ne (1 17 cm -2 ) Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

22 Ion Fractions N O III / NO I N O II / NO I N O III / NO I N O II / NO I N O III / NO I N O II / NO I N Ne II / NNe I N Ne II / NNe I N Ne II / NNe I N Ne III / NNe I N Ne III / NNe I N Ne III / NNe I Distance (kpc) Latitude Longitude Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

23 What about the Compounds? We don t see dust or molecules, but we are not really looking for them yet. But see next talk by E. Costantini A&A 521, A79 (21) Flux ( Photons m 2 s 1 Å 1 ) O K I Edge O I + O III O I O VII O II Fig. 5. O I K-edge: data and Model B. Black and grey points refer to the (Pinto+1) first and second observation, respectively. Flux ( Photons m 2 s 1 Å 1 ) Photons cm s -1 Å GS Ne IX Ne K I Ne III Ne II (Å) Fig. 7. Ne I K-edge: data and Model B. (Gatuzz+García+15) Accurate atomic data is an important step toward robust detection of compounds. 2 s 1 Å 1 ) 4 Fe I L3 Fe I L2 s m 2 s 1 Å 1 ) 12 Mg K I Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

24 Fe L-shell Absorption Only the experimental cross section for metallic Fe is included (Kortright+Kim) The Astrophysical Journal, 8:29(13pp),215February1 Gatuzz et al. Photons cm -2 s -1 Å -1 Photons cm -2 s -1 Å XTE J Cygnus X Photons cm -2 s -1 Å -1 Photons cm -2 s -1 Å Sco X Cygnus X Figure 8. ISMabs best broadband fit (solid red lines) in the Fe L-edge region for the XTE J , Sco X 1, Cygnus X 2, and Cygnus X 1 sources. For each source the observations have been combined for illustrative purposes..6.3 Accurate atomic.55 data is an important step toward robust detection of compounds cm -2 s -1 Å -1 cm -2 s -1 Å Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25

25 Final Remarks The superior spectral resolution of Chandra-HETG allowed us to benchmark theoretical cross sections, which in turn are used for detailed studies of the ISM absorption. These models are particularly relevant for Astro-H and Athena science. Our analysis of 24 bright low-mass X-ray binaries revealed absorption that includes both neutral and ionized gas, detecting for first time Kα, Kβ, and Kγ resonances from some species. The X-ray H column densities are systematically larger than those from 21-cm measurements. However, these trends can depend on the continuum model. Depletion of O and Fe is also detected. Accurate atomic data is the first necessary step for robust detection of compounds. A rather large discrepancy between several astrophysical measurements and the latest laboratory experiments still remains and needs to be solved. Javier García (Harvard-Smithsonian CfA) X-ray Absorption in the ISM Chandra Workshop, Aug 2th, / 25