Diffuse Hot Gas under X-rayAbsorption line SpectroscopyDaniel Wang and Yangsen Yao (UMass)in collaboration withTaotao Fang (Berkeley),Wei Cui (Purdue),Fabrizio Nicastro (CfA), &Rupert Croft (Carnegie Mellon)See also the poster by Yangsen Yao and Daniel WangWhy interesting?SdScSbSaStarburstWhy absorption lines?• Add the velocityand/or depth• Measure the columndensity, thus the mass• Direct line diagnostics• Independent of coolgas absorptionTarget Sight LinesNo detectionX-ray binaryAGNChandra Grating InstrumentPropertiesFWHM ~ 5x102 km/sAbsorption line modelb =100 km/s is actually used!Implementation of anaccurate model inXSPEC:I(ν)=Ic(ν) exp[-τ(ν)]τ(ν)∝NHfafi(T)fluφ(ν,ν0,b)b=(2kT/mi+ξ2)1/2We assume solarabundance and CIE.Ionization fractionAssuming CIE and solar abundancesOVIOVIIIOVIINe IXNe VIIINe IXMkn 421PKS 2155 3C 273LETG/ACISLETG/HRCLMC X-3• 100 ks exposure.• The EW is about thesame as those seen inAGN spectra!Results from extragalactic sourcesLog[NH(cm-2)]Log[T(K)]Source20.1(19.2-21.0)6.2(5.7-6.4)LMC X-320.0(19.7-20.2)6.3(6.0-6.4)3C 27319.5(15.3-19.7)6.37(6.33-6.42)PKS 2125-30419.2(19.1-19.3)6.27(6.22-6.30)MRK 421 No evidence for significant X-ray absorptionbeyond the LMC!!!Galactic source selection andChandra observations• LMXBs with |b| > 2o• S/N > 7 per bin at ~0.6 keV• Excluding sources with identifiedintrinsic emission/absorption features• Ten LMXBs with 17 observations (6with the LETG)Also detection of OVIIKα, Kβ,and OVIII KαlinesMean T ~ 106.3+/-0.2 KnH~3.4(-1.4,+1.7)x10-3 cm-3(assuming filling factor =1)Ne IX Kα absorptionlines in LMXB spectraExponential distribution modelsDisk model•nH = 5.0(-1.8,+2.6)x10-3 cm-3 exp[-|z|/1.1(-0.5,+0.7) kpc]•Total NH~1.6 x1019 cm-2Sphere model•nH = 6.1(-3.0,+3.6)x10-2 cm-3 exp[-R/2.7(-0.4,+0.8) kpc]~3 x 10-3 cm-3 at the Sun•Total NH~6.1 x1019 cm-2•MH~7.5(2.5-16)x108 MsunX-ray absorption is primarily around the Galactic disk within a few kpc!Comparisons with other measurements• OVI absorption (e.g., Savage et al. 2003)– Comparable to the scale height and velocity dispersionof OVI-absorbing gas– The predicted NOVI is ~1/4 of the observed (assumingCIE and cosmic abundances)• Pulsar dispersion measures (e.g., Gomez et al. 2001)– Ne=6x1019 cm-2 (Galactic pulsars only)– Ne=2x1020 cm-2 (including MC pulsars)– lower limit to the metal abundance of hot gasNH/Ne ~ 0.06 – 0.3 solar– Excluding the HII contribution, the abundance is likelyto be solar.– Or a substantial fraction of the pulsar DM is due to hotgas.• X-ray emission– dI/dΩ=4.8(+-0.8) ph/(s cm2 sr) inOVII triple, averaged over a largefield (McCammon et al. 2002)– Marginally consistent with theexpected from our inferredexponential distribution, assumingT = 106.3+-0.2 K– Correlation with diffuse _-keVband intensityOVI- and X-ray-absorbingand emitting hot gases mayall be associated with eachother.Disk dist.Uniform dist.ObservedSummaryA systematic analysis of Chandra grating spectra of XBs andAGNs to study the OVII, OVIII, and Ne IX absorption,accounting for line saturation and multiple line detections• No evidence for significant X-ray absorption beyond the LMC(probably ~< 1019 cm-2, assuming the solar abundance)• X-ray absorption primarily around the Galactic disk within afew kpc• Vertical scale height ~1-2 kpc, similar to the values for OVIabsorbers and free electrons from pulsar dispersion measures• Metal abundance probably ~ solar• Total absorbing gas mass ~ (0.2-1.6)x109 Msun• Mean temperature ~ 106.3+-0.2 K, density ~ 3 x 10-3 cm-3• Predicts ~0.4 x10-8 cm-2 of OVI, ~_ of the observed.Chandra grating observationsHigh Energy Transmission GratingSpectrometerLines of sightNo detectionX-ray