Spatial distribution of X-ray emitting ejecta in Tycho's SNR: indications of shocked Titanium
aa r X i v : . [ a s t r o - ph . H E ] M a r D raft version J uly
17, 2017
Preprint typeset using L A TEX style emulateapj v. 05 / / SPATIAL DISTRIBUTION OF X-RAY EMITTING EJECTA IN TYCHO’S SNR: INDICATIONS OF SHOCKED TITANIUM
M. M iceli , S. S ciortino , E. T roja , S. O rlando Draft version July 17, 2017
AbstractYoung supernova remnants show a characteristic ejecta-dominated X-ray emission that allows us to probe theproducts of the explosive nucleosynthesis processes and to ascertain important information about the physicsof the supernova explosions. Hard X-ray observations have recently revealed the radioactive decay lines of Ti at ∼ . ∼ . XMM-Newton archiveobservations of the Tycho’s SNR. We produce equivalent width maps of the Fe K and Ca XIX emission linesand find indications for a stratification of the abundances of these elements and significant anisotropies. We thenperform a spatially resolved spectral analysis by identifying five di ff erent regions characterized by high / lowvalues of the Fe K equivalent width. We find that the spatial distribution of the Fe K emission is correlated withthat of the Cr XXII. We also detect the Ti K-line complex in the spectra extracted from the two regions with thehighest values of the Fe and Cr equivalent widths. The Ti line emissions remains undetected in regions wherethe Fe and Cr equivalent widths are low. Our results indicate that the post-shock Ti is spatially co-locatedwith other iron-peak nuclei in Tycho’s SNR, in agreement with the predictions of multi-D models of Type Iasupernovae. Keywords:
X-rays: ISM — ISM: supernova remnants — ISM: individual object: Tycho’s SNR INTRODUCTION
Supernova remnants (SNRs) govern the physical and chem-ical evolution of our Galaxy. An exploding star releases ∼ erg of kinetic energy through some solar masses ofejecta that expand supersonically and drive powerful shocksback and forth in the ambient medium and ejecta themselves.The X-ray emission from young SNRs is a powerful diagnos-tic tool to study the imprint of the supernova explosion in theevolution of the remnant. The X-ray emission of young SNRsis, in fact, ejecta-dominated, being mainly associated with themetal-rich material expelled in the supernova explosion andheated up to X-ray emitting temperatures by the interactionwith the reverse shock. The ejecta carry information aboutthe explosive nucleosynthesis processes and can ”keep mem-ory” of the physics of the explosion itself (e. g., Badenes et al.2008b, Miceli et al. 2006).In particular, the iron-group elements (e. g., Cr, Mn, Ti,together with Fe and Ni) are synthesized in the inner lay-ers of the exploding star and can provide important informa-tion on the progenitor (Badenes et al. 2008a; Yamaguchi et al.2014). It has been shown that there is a correlation in the cen-troids of X-ray line complexes of Cr, Mn, and Fe in a largenumber of SNRs, including Kepler, W49B, N103B, Tycho,G344.7-0.1, and Cas A (Yang et al. 2013). This result seemsto suggest that these elements are spatially co-located in theexplosions in this large sample of SNRs, that includes bothcore-collapse and Type Ia SNRs. On the other hand, it hasbeen recently shown that the spatial distribution of Ti is sig-nificantly di ff erent from that of Fe in the Cassiopeia A SNR(Grefenstette et al. 2014). However, the radioactive emissiontrace the whole amount of Ti, concentrated in the unshocked Dipartimento di Fisica & Chimica, Universit`a di Palermo, Piazza delParlamento 1, 90134 Palermo, Italy [email protected] INAF-Osservatorio Astronomico di Palermo, Piazza del Parlamento 1,90134 Palermo, Italy NASA, Goddard Space Flight Center, Greenbelt, MD 20771, USA Department of Physics and Department of Astronomy, University ofMaryland, College Park, MD 20742, USA interior of the remnant, while the X-ray emission from Feoriginates only from the ejecta shocked by the reverse shock.Therefore, the di ff erent morphologies may be due to the factthat we only observe a small fraction of Fe in X-rays.The radioactive hard X-ray signature of Ti (whose ra-dioactive decay lines are at 67.86 keV and 78.36 keV)has been recently observed in Tycho’s SNR through theanalysis of
S wi f t / BAT (Troja et al. 2014) and
INT EGRAL (Wang & Li 2014) observations. Tycho’s SNR is the remnantof a Type Ia SN explosion (Badenes et al. 2006; Krause et al.2008) occurred in 1572 AD and presents clear signatures ofe ffi cient particle acceleration (see, e.g., Eriksen et al. 2011;Bykov et al. 2011; Morlino & Caprioli 2012; Slane et al.2014). Besides regions characterized by strong synchrotronX-ray emission, its thermal X-ray radiation is dominated bythe ejecta (Cassam-Chena¨ı et al. 2007).In Type Ia SNRs, delayed-detonation models (e. g.,Gamezo et al. 2005) predict that Fe-group elements are lo-cated in the inner parts of the ejecta profile, surrounded byintermediate-mass elements (e. g., Si, S, and Ca). More re-cently, three-dimensional delayed-detonation models devel-oped by Seitenzahl et al. (2013) revealed the details of the el-ement stratification, by showing that Fe-group elements canhave velocities higher than those of Ni (which, after the de-cay, produces the bulk of Fe-rich ejecta), but still lower thanthose of the intermediate-mass elements, which are then ex-pected to expand in an outer shell. Three-dimensional defla-gration models (e. g., R¨opke & Hillebrandt 2005) generallysuggest a more e ffi cient mixing in the abundances distributionthan classical 1-D deflagration models (Nomoto et al. 1984).We here take advantage of the deep set of XMM-Newton archive observations of the Tycho’s SNR to study the spatialdistribution of the heavy elements in the shocked ejecta. Wealso look for the Ti K-emission line complex at ∼ . iceli et al .He α emission line in the AS CA spectrum of W49B has beenreported (Hwang et al. 2000), though it was not confirmedby the subsequent
XMM-Newton observations (Miceli et al.2006). RESULTS
We analyzed the archive
XMM-Newton
EPIC observa-tions 0096210101, 0310590101, 0310590201, 0412380101,0412380201, 0412380301, 0412380401, 0511180101, allhaving pointing coordinates α J = h m . s , δ J =+ ◦ ′ . ≤
12 for the MOS cameras, PATTERN ≤ = . ff ects. Images were produced byadopting the procedure described in Miceli et al. (2006) (seetheir Sect. 2). Spectral analysis was performed in the energyband 3 . − . ∼ . ff er-ent observations were fitted simultaneously. For each spec-trum, we subtracted a background spectrum extracted froma nearby region immediately outside of the SNR shell andwe verified that the best-fit values do not depend significantlyon the choice of the background region. In all the fittings,the column density of the interstellar absorption was fixed to N H = × cm − in agreement with Cassam-Chena¨ı et al.(2007). Small local variations of the N H can be present acrossthe remnant (Slane et al. 2014), but we do not expect any sig-nificant e ff ects on our results, given the relatively hard energyband considered here. Image Analysis
To trace the spatial distribution of the chemical abundancesin the shocked ejecta, we produced equivalent width (EW)maps. X-ray line and continuum emission both scale with theplasma emission measure, and EW maps allow us to disen-tangle higher element abundances from higher emission mea-sure. Though the EW of an emission line increases with theelement abundance, it also varies with the plasma temperatureand ionization age (e.g, van Paradijs & Bleeker 1999). There-fore, EW maps provide indications of the distribution of theabundances, but do not give quantitative information and needto be tested with spatially resolved spectral analysis (as we doin Sect. 2.2).To identify the Fe-group elements, we produced the equiv-alent width (EW) map for the Fe K line complex, which,thanks to the high statistics of the
XMM-Newton data, hasa much higher signal-to-noise ratio than that presented inHwang et al. (2002). We also produce, for the first time, theEW map of the Ca XIX emission lines and, for comparison,of the Si XIII lines. To produce these maps we divided thecontinuum-subtracted line images (in the 1 . − .
05 keV, 3 . − .
05 keV, and 6 . − . pn spectrum of the remnant in a continuum bandadjacent to the line emission with a phenomenological power-law model. In particular, we considered the 4 . − . Γ = . . − .
65 keV band for the Si line (in this case, witha best-fit
Γ = . ff ect these maps, ourspatially resolved spectral analysis shows that the EW mapsprovide reliable results (see Sect. 2.2).Fig. 1 shows the continuum (4 . − . Chandra by Hwang et al. 2002), our Fe and Ca EW maps re-veal strong anisotropies, with high values at North and rela-tively low values in the center and elsewhere in the rim. Itis not easy to ascertain the origin of these inhomogeneities,which may be intrinsic or resulting from the interaction ofthe remnant with the ambient medium. In fact, Tycho’s SNRevolves in a structured environment (see, e.g., Williams et al.2013; Chiotellis et al. 2013) that may produce anisotropiesand corrugations in the reverse shock front. On the other hand,these anisotropies are not present in the Si-rich ejecta, there-fore the observed Ca and Fe inhomogeneities may be intrinsicin the ejecta structure.The Ca XIX EW presents a bright arc immediately behindthe northern border of the shell (region 4 in Fig. 1). A largeregion characterized by strong Fe EW is clearly present be-hind this bright Ca arc, closer to the center of the remnant(region 1 in Fig. 1). Fig.2 shows the comparison between theCa and Fe EWs (and the continuum emission) to highlight thedi ff erences in the distribution of these elements. The spatialdistribution of the Si, Ca, and Fe EWs is suggestive of a pos-sible stratification of the SN ejecta, with Fe confined withinthe inner region and lower-Z elements forming an outer enve-lope. An indirect spectroscopic indication of a stratificationbetween Fe-rich ejecta and lighter elements was proposed onthe basis of the analysis of the global spectrum of the remnant(e.g., Hwang et al. 1998) and of the spectrum extracted froma large region in the eastern part of the shell (Badenes et al.2006). But with our EW map (together with the spatially re-solved spectral analysis presented in Sect. 2.2), we can di-rectly separate out the di ff erent elements spatially, thus re-vealing that, besides the element stratification, there are alsostrong anisotropies in the Ca and Fe distributions.A stratification in the ejecta abundances is predicted by3-D simulations of delayed-detonation Type Ia SNe, whichshow a turbulent inner region characterized by iron-group el-ements, surrounded by a smooth distribution of intermediate-mass elements (including Ca) in the outer layers of the ejecta(Kasen et al. 2009). This stratification between Fe and Ca isalso observed in Type Ia SNe (e.g., Tanaka et al. 2011). As forthe inhomogeneities in the Ca map, the light-echo spectrumof Tycho’s SN has revealed a high velocity component in theCa-rich ejecta (Krause et al. 2008). Also, the presence of fastCa-rich knots deduced by Krause et al. (2008), may explainwhy the Ca EW presents a much higher level of anisotropythan the Si EW. Our results then suggest that the remnant has To exclude the contribution of the synchrotron emitting limbs, we ex-tracted the spectrum from a circular region slightly smaller than the shell. patial distribution of the ejecta in T ycho ’ s SNR 3
Si EW5 43 2 1
Ca EW5 43 2 1
Fe EW5 43 2 1
Figure 1.
Upper left panel:
EPIC count-rate images (MOS and pn mosaic) of Tycho’s SNR in the continuum band 4 . − , Upper right panel:
Si EW mapobtained in the 1 . − .
05 keV keV band.
Lower left panel:
Ca EW map obtained in the 3 . − .
05 keV band.
Lower right panel: the Fe K EW map obtained inthe 6 . − . ′′ and the images are all adaptively smoothed to a signal-to-noise ratio R =
10, but the Si EW map , where R =
25. Wesuperimposed the regions selected for the spectral analysis (in red) together with the contour levels of the continuum image at 10% , , , kept memory of the pristine ejecta distribution.We also confirm the presence of the bright Fe-rich east-ern knot at south-east (Vancura et al. 1995). Our map showsthat this knot (region 3 in Fig. 1) has the highest Fe K EWobserved in the whole remnant (see also Sect. 2.2 and Fig.4). We find that the projected distance between this knot andthe approximate center of the remnant (indicated by a yel-low cross in Fig. 1) is 30% higher than that of region 1 (i.e. of the “smooth” distribution of shocked Fe-rich ejecta).This di ff erence can be explained by considering the Fe-richknot as a moderately overdense shrapnel (a density inhomo-geneity in the ejecta profile) that moves beyond the Fe-richejecta shell, as modelled by Miceli et al. (2013). Ejecta shrap- nels are common in core-collapse SNRs, but localized clumpsof ejecta have been observed also in Type Ia SNe and theirevolution has been modelled with hydrodynamic simulations(Wang et al. 2003, 2006, Orlando et al. 2012), so we can ar-gue that the eastern knot traces an ejecta clump originated inthe inner layers of the exploding progenitor star. Spectral Analysis
We performed a spatially resolved spectral analysis by se-lecting 5 regions defined on the basis of the Ca and Fe EWmaps and shown in Fig. 1 and in Fig. 2. Region 1 and 3are very bright in the Fe K EW map and are expected to beFe-rich; region 2 traces a part of the shell characterized by a M iceli et al . Figure 2.
Color-composite image showing the EPIC count-rate images inthe 4 . − , green ), together with the Ca EW map (in red ), andthe Fe K EW map (in blue ). The regions selected for the spectral analysis aresuperimposed. no r m a li z ed c oun t s s − k e V − χ Energy (keV)
Ti Cr MnCa Fe
Figure 3.
Summed pn spectrum extracted from region 1 of Fig. 1, togetherwith its best fit model and residuals. The elements contributing to the emis-sion lines described in the text are labelled in blue. bright continuum and low EW for both Ca and Fe; region 4is where the Ca EW is the highest, while region 5 shows lowvalues of the continuum and of the EW of Fe and Ca.Figure 3 shows the pn spectrum of region 1 obtained bysumming all the observations (the analysis has been per-formed simultaneously on all the di ff erent spectra, we showthe summed spectrum only for visibility reasons). We foundthat, in this spectrum and in all the other spectra, both the Caand the Fe line complexes are quite broadened, with the Caclearly showing a double peaked feature. We interpret thisbroadening as an e ff ect of the superposition of di ff erent con-ditions in the ionization states of the plasma and we model thespectra with two thermal components of optically thin plasmain non-equilibrium of ionization (VNEI model in XSPEC)with di ff erent temperature, abundances, and ionization time-scales, τ (this model describes the spectra significantly bet-ter than a single-temperature model). We associate the low τ
50 100 150Cr EW (eV)010002000300040005000 F e E W ( e V )
12 345
Figure 4.
Fe EW in the 6 . − . . − . component with the ejecta closer to the reverse shock and thehigh τ component with the material shocked at earlier stages,i. e. the ejecta at higher distances from the center of the shell.For each component, we let only the Ca and Fe abundancesfree to vary and we impose solar abundances for all other ele-ments (but Ni, which is assumed to have the same abundanceas Fe). We also add three gaussian components to model theline emission from shocked Ti (if any), Cr, and Mn.Table 1 shows the best-fit parameters for the five spectralregions. The temperatures of the low τ component (labelled“1”) are significantly lower than those of the high τ compo-nent. This result suggests that the temperature of electrons in-creases as they move away from the reverse shock, in agree-ment with what expected in the case of electron heating byCoulomb collisions with hotter protons in the post-shock flow.Though the Ca and Fe abundances in the low τ componentare poorly constrained, the abundances in the high τ compo-nent show much smaller errors. This is because the high τ component dominates the flux in the line bands in all the re-gions, being in the range ∼ −
74% of the total in the3 . − .
05 keV band, and 65% −
85% of the total in the 6 . − . and Fe (see Table 1). Also, the Fe equiv-alent width, measured from the spectral analysis in regions1 − S uzaku global spectrum ofthe remnant. Our spatially resolved spectral analysis allowsus to find spatial variations in the intensity of these lines. Wemeasured the EW of the Cr and Mn emission lines in all theregions. While the Mn EW presents large error bars, we revealsignificant variations in the EW of the Cr XXII line. Figure4 shows that the Cr EW derived by our fittings increases inregions with high Fe EW. This correlation strongly indicatesthat Fe and Cr are spatially co-located, thus confirming whatproposed by Yang et al. (2013) on the basis of the global spec- patial distribution of the ejecta in T ycho ’ s SNR 5
Table 1
Results of the spectral analysis for the regions shown in Fig. 1.Parameter Region 1 Region 2 Region 3 Region 4 Region 5 kT (keV) 1 . + . − . . + . − . . + . − . . ± . . + . − . τ (10 s cm − ) 3 . ± . . ± . + − . . ± . . ± . EM (10 cm − ) 9 + − + − + − + − + . − . Ca ±
300 200 ±
140 600 ±
500 600 ±
400 140 + − Fe ±
200 70 ±
60 200 ±
150 200 ±
160 300 ± kT (keV) 4 . ± . . ± . . ± . . + . − . . ± . τ (10 s cm − ) 26 ± ± + − + − . + . − . EM (10 cm − ) 8 ± + − . + − . ± . + . − . Ca + − . + . − . + − + − . ± . + − . + . − . + − ±
10 5 . ± . E Ti (keV) 4 . . − . . ** . ± .
06 4 . ** . ** N Ti (10 − cm − / s) ± < . ± . + . − . < . E Cr (keV) 5 . + . − . . ± .
15 5 . + . − . . + . − . . ± . N Cr (10 − cm − / s) 6 . ± . . ± . . + . − . . ± . . + . − . E Mn (keV) 6 . ± .
03 6 . + . − . . ** . ± .
06 6 . ± . N Mn (10 − cm − / s) 3 . + . − . . ± . . ± . . ± . . + . − . Reduced χ (dof) 1 .
05 (1855) 0 .
97 (2022) 1 .
01 (203) 0 .
98 (1399) 1 .
07 (2521)
Note . — All errors are at the 90% confidence level. * Emission measure per unit area. ** Unconstrained. trum of the remnant.We also found that the quality of the fits in regions 1 and3 improves significantly by adding to the model a Gaussiancomponent with energy E ∼ .
90 keV, corresponding to sometransitions to the ground level of Ti (Kramida et al. 2014).The normalization of this component is higher than zero at al-most 3 sigmas ( ∆ χ ∼
8) in region 1 and at more than two sig-mas in region 3 ( ∆ χ ∼ ff ects a few bins of the spectra, it still produces a significantreduction in the χ (which is calculated in the broad 3 . − . ∼ . Chandra
ACIS spectra of region 1 extracted fromobservations 10093-97, 10902-4, and 10906 (all performed in2009, for a total of 734 ks) . The combined fits of the pnand ACIS spectra is reasonably good (reduced χ = .
26 with3644 dof), though it presents clear residuals corresponding tothe Ca and Fe line complexes, possibly associated with thedi ff erent PSFs of the two telescopes (a larger contaminationfrom bright nearby regions is expected in the XMM-Newton spectra). These residuals make the Ti line unconstrained, sowe focus on the 4 . − . . − . χ = .
06 with 1626 dof, and the normaliza-tion of the Ti line is N Ti = . + . − . × − cm − / s, in agreementwith the value in Table 1. Also, N Ti > Chandra data have been reprocessed with CIAO 4.7 and spectra havebeen extracted with the SPECEXTRACT script. sigmas. DISCUSSION AND CONCLUSIONS
We analyzed several archive
XMM-Newton observations ofTycho’s SNR to study the spatial distribution of the shockedejecta. We found strong indications for anisotropies in the dis-tributions of Fe-rich and Ca-rich shocked ejecta, which appearto be mainly localized in the northern part of the remnant. Wealso found a radial stratification of the ejecta chemical compo-sition, with Ca and Si localized in an outer shell with respectto Fe.Our spatially resolved analysis shows that the EW of the Crand Fe lines are correlated in the di ff erent regions of the rem-nant, with regions having the highest Fe abundances show-ing also the highest Cr equivalent width. Theoretical mod-els of delayed-detonation show that, besides the explosiveSi-burning regime, di ff erent yields of Fe-group elements canbe synthesized in incomplete Si-burning layers, depending onvariations in the details of the transition from deflagration todetonation (e. g., Iwamoto et al. 1999). The indications of aspatial co-location of Fe and Cr obtained by us suggest thatthe bulk of shocked Fe-group elements in Tycho’s SNR hasbeen synthesized in the explosive Si-burning regime (as inKepler’s SNR, see Park et al. 2013).We also found indications for the presence of Ti line emis-sion, confirmed by the joint Chandra and
XMM-Newton dataanalysis. We verified that this emission is concentrated in re-gions characterized by bright emission from Cr and Fe K. Wethen conclude that the spatial distribution of the Ti-rich ejectashould follow that of the Fe-rich ejecta in Tycho’s SNR. Thisclearly suggests that Fe-peak nuclei are spatially co-locatedin the remnant, in agreement with the predictions of multi-Dmodels of Type Ia SN explosions.The
S wi f t / BAT observations of radioactive emission from Ti points toward a total mass M Ti > − M ⊙ (Troja et al.2014). This value is consistent with that expected from adelayed-detonation explosion, which appear to be particularlysuited for Tycho’s SNR (e.g., Badenes et al. 2006). Delayed- M iceli et al .detonation models (e.g., Iwamoto et al. 1999) generally pre-dict much larger yields of Ti (5 − × − M ⊙ ) and Ti( ∼ × − M ⊙ ) than Ti. We then expect these heavy iso-topes to contribute predominantly to the X-ray line emission.We can evaluate whether the observed Ti line flux reportedin Table 1 is sound by comparing it to the flux of the Cr emis-sion line. In particular, in region 1 we obtain a line flux ratioof the Ti to Cr emission lines N Ti / N Cr = . ± . N Ti / N Cr ∼ M Ti / M Cr × E Ti / E Cr , where M Ti , Cr indicate themass of the shocked (i. e., X-ray emitting) Ti and Cr, respec-tively, and E Ti , Cr are the corresponding emissivities per ion. Ifwe consider that Ti and Cr are spatially co-located, we can as-sume that the ratio M Ti / M Cr is the same as the ratio of the total(shocked and unshocked) masses of Ti to Cr synthesized at theexplosion. This mass ratio is predicted to be M Ti / M Cr ∼ . α emissivitiesof Si, S, Ar, Ca, Fe, and Ni for a plasma in non-equilibriumof ionization having the best-fit temperature and ionizationtimescale that we obtained in region 1 for the high τ compo-nent (see Table 1), which is the one that mainly contribute tothe line emission. We thus obtain E Ti / E Cr ∼
2. Therefore,the expected Ti to Cr line flux ratio is N Ti / N Cr ∼ .
12, whichis consistent with that observed in region 1. Though large un-certainties are involved in this estimate, we conclude that theobserved Ti flux appears to be reasonable and consistent withexpectations.Further observations are necessary to study in details thespatial distribution of the Ti-rich ejecta in Tycho’s SNR. HardX-ray observations performed with the
NuS T AR telescopewill allow us to trace the Ti emission with high spatial res-olution and to verify if it is consistent with that of the Fe Kemission (shown in blue in Fig. 1), as suggested by our analy-sis. However, we point out that, the abundance of neutron-richelements is highly sensitive to the electron captures takingplace in the central layers of the exploding star, so, in princi-ple, the spatial distribution of Ti and Ti may not coincidewith that of the radioactive Ti (though an e ffi cient mixing isexpected).A major leap forward will be provided by the next gen-eration of X-ray telescopes. As an example, we simulatedan 80 ks observations of the Fe-rich region of Tycho’s SNR,performed with the Soft X-ray Spectrometer of the forthcom-ing Astro − H mission (Takahashi et al. 2014 and referencestherein). We verified that it will be possible to detect the Tiline in the whole field of view of the telescope with a veryhigh statistical confidence (5 sigmas). A detailed study of thespatial distribution of the shocked Ti will be possible withthe Athena telescope (Nandra et al. 2013; Decourchelle et al.2013). We simulated a 50 ks observation of Tycho’s SNRperformed with
Athena
X-IFU (Ravera et al. 2014) and foundthat, by assuming an average line flux equal to that observedin region 1, the X-IFU spectra will allow us to detect the Tiemission line in 1 arcmin regions at more than 5 σ . We thank the anonymous referee for their comments andsuggestions. This paper was partially funded by the PRININAF 2014 grant. M. M. thanks M. Dadina for discussionsabout the X-IFU instrumental background.REFERENCES Badenes, C., Borkowski, K. J., Hughes, J. P., Hwang, U., & Bravo, E. 2006,ApJ, 645, 1373Badenes, C., Bravo, E., & Hughes, J. P. 2008a, ApJ, 680, L33Badenes, C., Hughes, J. P., Cassam-Chena¨ı, G., & Bravo, E. 2008b, ApJ,680, 1149Bykov, A. M., Ellison, D. C., Osipov, S. M., Pavlov, G. G., & Uvarov, Y. A.2011, ApJ, 735, L40Cassam-Chena¨ı, G., Hughes, J. P., Ballet, J., & Decourchelle, A. 2007, ApJ,665, 315Chiotellis, A., Kosenko, D., Schure, K. M., Vink, J., & Kaastra, J. S. 2013,MNRAS, 435, 1659Decourchelle, A., Costantini, E., Badenes, C., et al. 2013, ArXiv e-prints,arXiv:1306.2335Eriksen, K. A., Hughes, J. P., Badenes, C., et al. 2011, ApJ, 728, L28Gamezo, V. N., Khokhlov, A. M., & Oran, E. S. 2005, ApJ, 623, 337Grefenstette, B. W., Harrison, F. A., Boggs, S. E., et al. 2014, Nature, 506,339Hwang, U., Decourchelle, A., Holt, S. S., & Petre, R. 2002, ApJ, 581, 1101Hwang, U., Hughes, J. P., & Petre, R. 1998, ApJ, 497, 833Hwang, U., Petre, R., & Hughes, J. P. 2000, ApJ, 532, 970Iwamoto, K., Brachwitz, F., Nomoto, K., et al. 1999, ApJS, 125, 439Kasen, D., R¨opke, F. K., & Woosley, S. E. 2009, Nature, 460, 869Kramida, A., Yu. Ralchenko, Reader, J., & and NIST ASD Team. 2014,NIST Atomic Spectra Database (ver. 5.2), [Online]. Available: http://physics.nist.gov/asd [2014, November 26]. NationalInstitute of Standards and Technology, Gaithersburg, MD.Krause, O., Tanaka, M., Usuda, T., et al. 2008, Nature, 456, 617Miceli, M., Decourchelle, A., Ballet, J., et al. 2006, A&A, 453, 567Miceli, M., Orlando, S., Reale, F., Bocchino, F., & Peres, G. 2013, MNRAS,430, 2864Morlino, G., & Caprioli, D. 2012, A&A, 538, A81Nandra, K., Barret, D., Barcons, X., et al. 2013, ArXiv e-prints,arXiv:1306.2307Nomoto, K., Thielemann, F.-K., & Yokoi, K. 1984, ApJ, 286, 644Orlando, S., Bocchino, F., Miceli, M., Petruk, O., & Pumo, M. L. 2012, ApJ,749, 156Park, S., Badenes, C., Mori, K., et al. 2013, ApJ, 767, L10Ravera, L., Barret, D., den Herder, J. W., et al. 2014, in Society ofPhoto-Optical Instrumentation Engineers (SPIE) Conference Series, Vol.9144, Society of Photo-Optical Instrumentation Engineers (SPIE)Conference Series, 2R¨opke, F. K., & Hillebrandt, W. 2005, A&A, 431, 635Seitenzahl, I. R., Ciaraldi-Schoolmann, F., R¨opke, F. K., et al. 2013,MNRAS, 429, 1156Slane, P., Lee, S.-H., Ellison, D. C., et al. 2014, ApJ, 783, 33Takahashi, T., Mitsuda, K., Kelley, R., et al. 2014, ArXiv e-prints,arXiv:1412.2351Tamagawa, T., Hayato, A., Nakamura, S., et al. 2009, PASJ, 61, 167Tanaka, M., Mazzali, P. A., Stanishev, V., et al. 2011, MNRAS, 410, 1725Troja, E., Segreto, A., La Parola, V., et al. 2014, ArXiv e-prints,arXiv:1411.0991van Paradijs, J., & Bleeker, J. A. M., eds. 1999, Lecture Notes in Physics,Berlin Springer Verlag, Vol. 520, X-Ray Spectroscopy in AstrophysicsVancura, O., Gorenstein, P., & Hughes, J. P. 1995, ApJ, 441, 680Wang, L., Baade, D., H¨oflich, P., & Wheeler, J. C. 2003, ApJ, 592, 457Wang, L., Baade, D., H¨oflich, P., et al. 2006, ApJ, 653, 490Wang, W., & Li, Z. 2014, ApJ, 789, 123Williams, B. J., Borkowski, K. J., Ghavamian, P., et al. 2013, ApJ, 770, 129Yamaguchi, H., Badenes, C., Petre, R., et al. 2014, ApJ, 785, L27Yang, X. J., Tsunemi, H., Lu, F. J., et al. 2013, ApJ, 766, 44 We have used the ATOMDB database and the APEC non-equilibrium ionization library Libapecnei, see