Ejecta Abundances Distribution and Ionization Structure in Young SNRs

The fact that the X-ray emission in young SNRs is heavily influenced by the presence of ejecta heated by a reverse shock has been known since the earliest spec-troscopic measurements with moderate resolution spectrometers [9,10]. It has also been known that their X-ray spectra are influenced by ionization conditions, specifically that the low density, shock-heated plasma is less ionized than one in collisional ionization equilibrium (CIE) at the same electron temperature. Models of so-called nonequilibrium ionization (NEI) plasmas preceded our ability to detect them in nature [44,74].

As the fluxes from prominent X-ray lines are stronger in an NEI plasma than a CIE plasma with the same electron temperature, low and modest resolution spectrometers often yield ambiguous results between ionization state and abundance. Early measurements of SNR spectra were unable to distinguish between NEI and enhanced abundances (e.g., [9,123]). The ability to accurately infer plasma conditions is further complicated by the fact that SNRs show a range of temperatures and ionization states (and abundance mixes) often along the same line of sight. This renders interpretation of spectra challenging. As a consequence, few, if any, X-ray measurements of abundances in SNRs are definitive, and all must be viewed critically. On the other hand, spectral measurements or narrow band equivalent width maps showing relative abundances can be instructive in understanding the distribution of ejecta in SNRs.

One desired use of X-ray abundances is for determining explosion type. The ejecta from core collapse SNe and deflagration/detonation (Type Ia) SNe have distinct abundance patterns. Specifically, the ejecta from core collapse SNe have high ratios compared with cosmic abundances of oxygen to iron; those from Type Ia's have the opposite. X-rays are the ideal band for seeking this signature, as broad band X-ray spectrometers cover the K lines from oxygen and iron, as well as the iron L band. The complications of abundance determination make SN typing challenging. Compounding the difficulty is our inability to know what fraction of the ejecta is visible in the X-ray band, and the fact that the oxygen abundance is difficult to determine in an unambiguous way as the oxygen line strength correlates with the column density in spectral fitting in moderate spectral resolution detectors.

Despite these difficulties, there have been some notable successes in measuring abundances, and deriving physical information from the measurement. The most successful was the use of Einstein FPCS observations of Puppis A to show an overabundance of oxygen with respect to iron, requiring the progenitor to have a mass of more than 25 [email protected] [22]. Also of note are comprehensive analysis of Tycho data requiring ejecta [46] and the analysis of an EXOSAT observation of W49B requiring the presence of a substantial amount of shocked ejecta [151]. It is with the advent of spatially resolved spectroscopy through which shock structures can be isolated that most of the advances have taken place in our knowledge of the abundances of reverse shocked ejecta and forward shocked ISM. ASCA was the pathfinder mission in this regard, but XMM-Newton and Chandra excel in these studies. The recent studies are providing real insight into the ejecta masses and their degree of mixing, as well as the explosion mechanism.

17.2.1.1 Type Ia SNRs

There are only two confirmed Galactic Type Ia remnants, Tycho (Fig. 17.2) and SN 1006 (Fig. 17.10). In Tycho, recent spatially resolved observations have removed the ambiguity of the initial spectral results, from which the strong flux of lines from Si, S, and Ar could be interpreted as either due to enhanced abundances from ejecta or NEI [9,123]. Circumstantial evidence for the presence of ejecta comes from the high mass of X-ray emitting material (5-15 [email protected]) required if solar abundance material is assumed [142]. Hamilton, et al. [46] analyzed all existing X-ray spectral data using a self-consistent model incorporating a forward shock encountering a uniform ISM and a reverse shock encountering stratified ejecta. They showed that satisfactory fits require the presence of ejecta, and that the total inferred ejecta mass is consistent with the 1.4 Mq expected from a Type Ia explosion. Using Ginga data, Tsunemi et al. showed that the centroids of the brightest lines in Tycho (and Cas A) require NEI, and included NEI effects to infer abundances substantially higher than solar [161].

Broad band model fitting of the ASCA spectrum revealed the relative ejecta abundances, and spatially resolved spectroscopy revealed their distribution [64,66]. The narrow band images in Mg, Si, S, Ar, Ca, Fe, and the continuum images all share a shell-like morphology, but each has distinct features. The Si and S ejecta have an average temperature of (0.8-1.1) x 107 K, and an average ionization age of (0.8-1.3)x 1011 cm~3 s. Azimuthal brightness variations in the two lines suggests a variation in the temperature of—1.5. The azimuthally averaged Fe K image peaks well within the other line images. The Fe K line radial surface brightness distribution and the centroid energy both indicate that the Fe ejecta have a temperature several times higher and an ionization age several times lower than the other ejecta. The non-Fe ejecta have abundances in good qualitative agreement with the predictions of standard SN Ia models. The apparently low abundance of Fe is consistent with the idea that the Fe ejecta are interior to the Si group ejecta, and largely unshocked.

Higher angular resolution views using Chandra and XMM-Newton refine this picture, but do not alter it in any fundamental way. The XMM-Newton image shows that the Fe K emission peaks at a smaller radius than the Fe L emission, verifying that the temperature with the ejecta increases toward the reverse shock [28]. The narrow band Si image corresponds well with the radio image, and probably marks the contact discontinuity, distorted by Rayleigh-Taylor instabilities. This latter conclusion is reinforced in dramatic fashion by the Chandra image of the Si emission in Tycho, which shows plume like structures throughout the interior [61,171]. Some of the plumes viewed tangentially approach the outer shock.

The line emission in SN 1006 has been a secondary consideration to the nonthermal emission arising from the bright limbs (see Sect. 17.2.6 below). Line emission is clearly observed throughout the remnant, except in the bright nonthermal limbs. Along the northwestern rim, Chandra imaging spectroscopy shows a clear separation between the forward-shocked material and the ejecta [96]. The forward shock shows material at ordinary solar abundances, shock-heated to electron temperatures of —0.6-0.7 keV. Interior to both the nonthermal northeast shock and the thermal northwest shocks are clumpy structures similar to those observed in Tycho. Their presence invite the speculation that such structure is common in Type Ia remnants. Spectral analysis of these structures reveals enhanced O, Mg, Si, and Fe abundances. No quantitative X-ray based analysis of the ejecta mass has been performed for SN 1006. Abundance measurements in SN 1006 using X-rays are compromised by the known presence of a substantial amount of high-velocity, unshocked ejecta (Fe, Si, S, and O) interior to the reverse shock [179].

The dearth of Galactic Type Ia remnants is compensated by the large Magellanic Cloud (LMC), in which at least three others have been identified. Hughes et al. used ASCA spectra from three LMC SNRs to demonstrate how remnants can be typed using their broad-band X-ray spectra [57]. Two of these remnants, 0509-67.5 and 0519-69.0, were previously thought to be Type Ia remnants, based on their Balmer-dominated optical emission. The third, N103B, was thought to be a core collapse remnant, but its spectrum more closely resembles that of a Type Ia remnant. Subsequent studies of these remnants using Chandra and XMM-Newton have facilitated

Fig. 17.2 Three-color composite Chandra image of Tychos SNR reveals a thin, hard outer rim that is the site of particle acceleration (Sect. 17.2.6) and plumes of ejecta-rich material in the projected interior, which are likely shaped by the Rayleigh-Taylor instability [171]. The red, green, and blue images correspond to photon energies in the 0.95-1.26, 1.63-2.26, and 4.1-6.1 keV bands. The angular diameter of Tycho's remnant is approximately 8 arcmin, corresponding to physical size of 5.8 pc at a distance of 2.5 kpc

Fig. 17.2 Three-color composite Chandra image of Tychos SNR reveals a thin, hard outer rim that is the site of particle acceleration (Sect. 17.2.6) and plumes of ejecta-rich material in the projected interior, which are likely shaped by the Rayleigh-Taylor instability [171]. The red, green, and blue images correspond to photon energies in the 0.95-1.26, 1.63-2.26, and 4.1-6.1 keV bands. The angular diameter of Tycho's remnant is approximately 8 arcmin, corresponding to physical size of 5.8 pc at a distance of 2.5 kpc spatially resolved spectroscopy of these distant objects with an equivalent spatial resolution to that provided by ASCA for Galactic remnants. Lewis et al. found that N103B has an ejecta distribution similar to Tycho: a bright shell dominated by Si and S, with an interior shell of hot Fe [95]. Interior to the hot Fe shell is a core of cooler Fe. The estimated masses of Si, S, Ar, Ca, and Fe do not match the predictions of any SN model, but are more consistent with a Type Ia model than a Type II. In contrast, the O, Si, and Fe mass estimates from an XMM-Newton RGS observation are markedly different and support a Type II origin [164]. Neither paper proposes a resolution to this discrepancy.

The LMC remnant 0509-67.5 also has a largely stratified structure, with ejecta confined to a shell interior to a shell of nonthermal emission [172]. The integrated spectrum shows clear evidence for enhanced metal abundances. Narrow band images reveal knots with Si and Fe abundances higher than the global average. Variations in the Fe and Si abundances from knot to knot suggest that they originated in the transition region between the Fe and Si layers of the progenitor. Inferred ejecta masses for the preferred model are 0.17 [email protected] forO; [email protected] for Si; 0.37 [email protected] forS; and 0.12 Mq for Fe. These masses are shown to be more consistent with the yields

Fig. 17.3 The LMC remnant DEM L71 has an outer forward shock structure and an interior filled with Si and Fe rich material [56]. The color scale is: red 0.3-0.7 keV, green 0.7-1.1 keV, blue 1.14.2 keV. The angular extent of the remnant is approximately 1.4 by 1.2arcmin, corresponding to 20 pc x 17 pc (Figure courtesy of Chandra X-ray Center and J. P. Hughes.)

Fig. 17.3 The LMC remnant DEM L71 has an outer forward shock structure and an interior filled with Si and Fe rich material [56]. The color scale is: red 0.3-0.7 keV, green 0.7-1.1 keV, blue 1.14.2 keV. The angular extent of the remnant is approximately 1.4 by 1.2arcmin, corresponding to 20 pc x 17 pc (Figure courtesy of Chandra X-ray Center and J. P. Hughes.)

for a delayed deflagration model for a Type Ia supernova than for a fast deflagration. The lack of Fe can be accounted for if the reverse shock has not propagated far into a cold, interior Fe layer.

DEM L71 is a much older LMC SNR; with an age >5 000 yrs, compared with less than 1000 yrs for 0509-67.5 and <1500 yrs for N103B. Nevertheless, it shows dramatic evidence of stratification [56]. In particular, its central core is markedly iron rich (Fig. 17.3). Despite its advanced age, the remnant preserves the two-shock structure observed in many young remnants. The X-ray spectra indicate a total ejecta mass of <1.5 Mq, most of which is composed of Fe.

17.2.1.2 Core Collapse Remnants

While Type Ia remnants universally seem to maintain stratification within their ejecta, the same cannot be said for core collapse remnants. They show a variety of ejecta structures, from highly stratified to nearly chaotic. Of course, Cas A, the most closely studied, shows the most anomalous structure.

Cassiopeia A: Narrow band ASCA images were the first to reveal the general ejecta distribution in Cas A [53]. The Si and S maps are virtually indistinguishable from each other or from the broad band map, but clearly different from the hard continuum. Narrow band images from the BeppoSAX MECS show clearly the difference between the line and hard continuum morphologies. Vink et al. used these data to estimate an ejecta mass of [email protected], and an X-ray emitting mass of [email protected], indicating the progenitor had a very high mass loss rate [170].

The first Chandra images revealed clear and unexpected differences in the distribution of metals (e.g., [65]). The Si, S, Ar, and Ca maps are similar to each other and to the distribution of fast optical knots. Equivalent width maps reveal the distribution of the prominent ejecta constituents. The structures in these maps contrast sharply with the 0.5-10.0 keV broadband map and the 4-6 keV continuum map. The northeastern jet, known from optical studies to contain Si group ejecta, shows up strongly in these maps. The Fe K emission has a very different morphology. In particular, in the southeast of the remnant, the Fe K emission is located at larger radii than the Si. This suggests that the inner Fe ejecta layers have been overturned and propelled beyond the Si group ejecta in this part of the remnant (Fig. 17.4). Such overturning is consistent with recent models of core collapse explosions (e.g., [17]).

Willingale et al. have used the XMM-Newton imaging data to infer the global metal abundance ratios and compare them with supernova models [178]. They show that the ratios with respect to Si of a large number of lines is most consistent with the theoretical nucleosynthesis yield for a 12 Mq progenitor.

Detailed studies of individual knots show that they have a variety of compositions. Features with distinct composition can be found on the smallest size scales. While most knots show a mix of ejecta, some are dominated by Si group elements and others by Fe. At least one knot emits Fe lines exclusively, and apparently is devoid of lower mass material. The knots also show a variety of ionization conditions, which have been used in the context of analytic hydrodynamical models to constrain the ejecta density profile, the location of the knots in mass coordinates, and the degree of explosion asymmetry [67,90]. The ejecta show a range of density profiles, from very shallow (p « r_n, where n <6) to very steep, (n — 30-50). The ejecta close to the jet show the shallowest profile, possibly due to an asymmetric explosion in which more of the energy is directed along the jet than elsewhere. For a total ejecta mass of 2Mq expected from a 20 [email protected] progenitor, the Fe-rich clumps are found to arise in a layer 0.7-0.8 [email protected] from the center. The observed composition appears to be possible only if Si burning products are mixed with O burning products.

The overall appearance of Cas A contrasts starkly with the young Type Ia remnants. Cas A consists of small knots and thin filaments, not the emission plumes observed in Tycho. The similarity between the structures in Cas A and the prediction of models involving Fe bubbles has been noted. Laming and Hwang argue that the knots are not especially overdense compared with their surroundings, and their

Fig. 17.4 Multicolor image of Cas A from a million-second Chandra observation reveals a variety of structures [62]. The red represents Si Hea (1.78-2.0keV), the blue Fe K (6.52-6.95 keV), and the green 4.2-6.4 keV continuum. In the east, the Fe ejecta are external to the Si, indicating overturning of the inner Fe ejecta layers. The hard outer rim arises from shock-accelerated electrons; projected interior hard emission has a nonthermal bremsstrahlung component. The angular diameter is approximately 5 arcmin, corresponding to 5 pc at a distance of 3.4 kpc

Fig. 17.4 Multicolor image of Cas A from a million-second Chandra observation reveals a variety of structures [62]. The red represents Si Hea (1.78-2.0keV), the blue Fe K (6.52-6.95 keV), and the green 4.2-6.4 keV continuum. In the east, the Fe ejecta are external to the Si, indicating overturning of the inner Fe ejecta layers. The hard outer rim arises from shock-accelerated electrons; projected interior hard emission has a nonthermal bremsstrahlung component. The angular diameter is approximately 5 arcmin, corresponding to 5 pc at a distance of 3.4 kpc high ionization ages and the proximity of some to the forward shock are the result of early passage through the reverse shock [90].

1E 0102.2-7219: The Small Magellanic Cloud remnant 1E 0102.2-7219 is another young core collapse remnant, with an estimated age of about 1000 yrs. ASCA observations show that its emission is dominated by four lines (He-a and Lya lines of O and Ne; [49]). Narrow band Chandra images show the remnant to be a nearly perfect ring, with a hotter, outer ring identified with the forward shock, and a cooler, inner ring of reverse-shocked ejecta [37]. Model fits to the ASCA and XMM-Newton composite spectra require multiple NEI components [49,137]. Observations using the grating instruments on Chandra and XMM-Newton have revealed the ion-ization structure and the distribution of the elements [34,127]. The brightness distribution peaks of the Lya lines from O, Ne, and Mg are exterior to those of the corresponding He-a lines, and there is a progression from smaller to larger radii of the brightness peaks of O, Ne, Mg, and Si. A simple model with constant electron temperature but variable ionization timescale accounts for the peak emission radius of all lines, suggesting that the radial structure is due to progressive ionization of a mixed plasma rather than stratification. From the Chandra grating observations,

Fig. 17.5 Multicolor Chandra image of the O-rich SNR G292.0+1.8 has a filamentary appearance similar to that of Cas A, but has a very different atomic composition. In this image, enriched ejecta are blue and normal composition material is yellow. The central pulsar wind nebula also appears in blue. The angular diameter of the remnant is 6 arcmin, corresponding to a physical size of ~11.5 pc at a distance of approximately 6 kpc (Figure courtesy of Chandra X-ray Center and J. P. Hughes.)

the mass of oxygen is estimated at <[email protected], and the neon mass is estimated to be <2Mq. The oxygen mass is consistent with a massive progenitor of <32 [email protected]

G292.0+1.8: The young (< 1600 yrs) SNR G292.0+1.8 was thought to be a Type II remnant based on its O-rich optical spectrum. Its confirmation as such resulted from the discovery using ASCA of a central synchrotron nebula and X-ray pulsar near the center [60]. The Chandra image of the remnant, shown in Fig. 17.5, reveals a thin, nearly circular outer shell of hard emission filled with an array of knots and filaments rivaling Cas A in complexity and contrasting starkly with the Type Ia remnants [111]. The composition and distribution of the shocked ejecta are different from Cas A. The ejecta consist primarily of O, Ne, and Si, with less S and Ar, and very little Fe, and are distributed primarily around the remnant's periphery. An X-ray bright equatorial band has normal composition, and is thought to be associated with presupernova mass loss.

17.2.1.3 SNRs of Uncertain Type

Kepler: Possibly the most vexing puzzle concerning young SNRs is the nature of the progenitor of Kepler's supernova (SN 1604). Strong arguments have been made for either a Type Ia or a core collapse origin. Key evidence supporting a Type Ia origin is the object's large distance (600pc) above the Galactic plane. The most compelling evidence of a Type II origin is the apparent presence of high-density material surrounding the object, thought to be the result of presupernova mass loss from the progenitor star. No compact source has been found.

The integrated ASCA spectrum shows strong emission from the K lines of Mg, Si, S, and Ca, and the L and K lines of Fe [82]. Iron appears to be in a low ionization state, with the bulk of the emission arising from Fe L lines. A model invoking a forward shock expanding freely into CSM and a reverse shock into ejecta was needed to satisfactorily fit the spectrum. High abundances of all metals are required, with the Fe abundance approximately ten times solar. Much of the Fe is thought to be still interior to the reverse shock, and so mixing of this material with exterior layers is needed to account for the Fe K line strength. A total ejecta mass of [email protected] is inferred. The relative abundances and total ejecta mass are consistent with a Type Ia origin.

Narrow band XMM-Newton images show a symmetric distribution of material, resembling Tycho more closely than Cas A [23]. The Fe L and Si images are generally consistent with each other, though the Si emission extends to a slightly larger radius. The Fe K emission peaks interior to the Fe L emission, suggesting an inward temperature gradient. There seems to be no substantial azimuthal asymmetry; az-imuthal surface brightness variations are likely associated with density variations in the circumstellar medium. Detailed analysis of the ejecta abundances from Chandra and XMM-Newton observations have not yet been published.

W49B: W49B has one of the most impressive spectra of any cosmic X-ray source, with multiple strong Si, S, and Fe emission lines [151]. Early ASCA analysis suggested that the Si and S emission arose from a shell surrounding the centrally peaked Fe [36], but subsequent, more detailed analysis indicated similar morphologies for the three line complexes [68]. Nevertheless, each element requires a unique set of plasma conditions, although all are close to collisional ionization equilibrium. In fact, Kawasaki et al. suggest that the S line ratios require the plasma in the remnant to be recombining [78]. A high density is required; a minimum of 2cm~3 is required, and the density is likely higher than this. Spectroscopy using XMM-Newton confirms the need for high density material but not a recombining plasma [102]. Both the ASCA and XMM-Newton spectra show features consistent with K lines from Cr and possibly Mn; this is the only remnant to show such lines.

The inferred abundances are high: the Si, S, Ca, Ar, Mg, and Fe all have abundance ratios to solar of 5-6. While the relative abundances support a Type Ia progenitor, a low mass Type II progenitor is consistent with the measured abundances, leaving the explosion type uncertain. On the other hand, the high-density environment in which W49B is expanding suggests a Type II progenitor. The high column to the remnant renders the O and Ne lines invisible, rendering impossible a direct O:Fe abundance comparison, and therefore a definitive determination. Kawasaki et al. argue that despite its apparent youth and anomalously high abundances, the morphology, thermodynamic state, and proximity to molecular clouds place W49B squarely in the class of mixed morphology remnants [78].

Was this article helpful?

0 0

Post a comment