1: \documentclass[referee]{chjaa}
2: %\documentclass[preprint]{chjaa}
3:
4: \usepackage{graphicx,times} %for PS/EPS graphics inclusion, new
5: \input{epsf.sty} %for PS/EPS graphics inclusion, old
6: \input{psfig.sty} %for PS/EPS graphics inclusion, old
7: \usepackage{latexsym}
8:
9: %\shorttitle{{\sl Chandra} observation of Cas A}
10: %\shortauthors{X.J.Yang et al.}
11:
12:
13: \begin{document}
14:
15: \title{High spatial resolution X-ray spectroscopy of SNR Cassiopeia A with {\sl Chandra}}
16:
17: \volnopage{Vol.0 (200x) No.0, 000--000} %%preserved for Editor. DOn't remove!
18: \setcounter{page}{1} %%starting page, preserved for Editor. DOn't remove!
19:
20: \author{Xue-Juan Yang
21: \inst{1,2}\mailto{yangxj@mail.ihep.ac.cn}
22: %% Please move "\mailto{}" to the corresponding author of the paper.
23: %% For single author or all the authors from an institute, use "\inst{}" only
24: %% Here is an example of three authors come from different institutes.
25: \and Fang-Jun Lu
26: \inst{1}
27: \and Li Chen
28: \inst{2}
29: }
30:
31: \institute{Particle Astrophysics Center, Institute of High Energy Physics,
32: Chinese Academy of Sciences, Beijing 100049, China\\
33: \email{yangxj@mail.ihep.ac.cn}
34: %% Please give the E-mail address of the author, to whom future correspondence and
35: %% offprint requests will be sent. Note to pair \mailto{} with \email{}
36: \and
37: Department of Astronomy, Beijing Normal University, Beijing
38: 100875, China\\
39: }
40:
41: \date{Received~~2007 Oct. 10; accepted~~year~~month day}
42:
43: \abstract{
44: We present high spatial resolution X-ray spectroscopy of supernova
45: remnant Cassiopeia A with the {\sl Chandra} observations. The X-ray
46: emitting region of this remnant was divided into 38 $\times$ 34
47: pixels with a scale of 10$\arcsec$ $\times$ 10$\arcsec$ each.
48: Spectra of 960 pixels were created and fitted with an absorbed two
49: component non-equilibrium ionization model. With the spectral
50: analysis results we obtained maps of absorbing column density,
51: temperatures, ionization ages, and the abundances for Ne, Mg, Si, S,
52: Ca and Fe. The Si, S and possibly Ca abundance maps show obviously
53: jet structures, while Fe doesn't follow the jet but seems to be
54: distributed perpendicular to it. In the range of about two orders of
55: magnitude, the abundances of Si, S and Ca show tight correlations
56: between each other, suggesting them to be ejecta from explosive
57: O-burning and incomplete Si-burning. Meanwhile, Ne abundance is well
58: correlated with that of Mg, indicating them to be the ashes of
59: explosive C/Ne burning. The Fe abundance is positively correlated
60: with that of Si when Si abundance is lower than 3 solar abundances,
61: but a negative correlation appears when the Si abundance is higher.
62: We suggest that such a two phase correlation is the results of
63: different ways in which Fe is synthesized.
64: \keywords{ISM: supernova remnants -- ISM: individual: Cassiopeia A}
65: }
66:
67: \authorrunning{X.-J. Yang et al.} %author_head in even pages
68: \titlerunning{{\sl Chandra} observation of Cas A} % title_head in odd pages
69: \maketitle
70:
71: %
72: %________________________________________________________________
73:
74: \section{Introduction}
75:
76: The young core-collapse supernova remnant (SNR) Cassiopeia A (Cas A)
77: is a perfect laboratory for studying the ejecta and shock in
78: remnants, for its age and distance are well-determined and it is
79: very bright across the whole electromagnetic spectrum. The age of
80: Cas A is around 300 years (Thorstensen et al. 2001) and the distance
81: about 3.4 kpc (Reed et al. 1995). It appears today as a nearly
82: circular 3 arcmin diameter bright ring, with a low surface
83: brightness 5 arcmin diameter radio and X-ray plateau. As a young
84: SNR, it is believed that thermal X-ray emission mainly comes from
85: the forward shocked interstellar medium (ISM) (Shklovsky 1973) and
86: the reverse shocked supernova (SN) ejecta (McKee 1974). The X-ray
87: image of Cas A is consistent with this scenario as suggested by
88: Gotthelf et al. (2001). They discovered a thin, bright X-ray wisp
89: that is interpreted as the forward shock, and a sharp rise in radio
90: and X-ray line emissivity at the inner edge of the bright ring,
91: which is associated with the reverse shock.
92:
93: Numerical models (e.g., Woosley \& Weaver 1995; Thielemann et al.
94: 1996) predict that nucleosynthesis in core-collapse supernovae (SNe)
95: occurs in an ``onionskin'' manner. Explosive Si burning occurs near
96: the core, where the shock temperatures is the highest. This process
97: would burn up Si and form ejecta dominated by $^{56}$Ni, which
98: decays to $^{56}$Co and finally $^{56}$Fe. Further out, as the shock
99: temperature decreases, Si burning is incomplete and the main
100: products include not only Fe, but also much Si, S, Ar and Ca. Even
101: further out, explosive O burning occurs, leading to a composition
102: dominated by O and Si with very little or no Fe. At the outmost the
103: explosive Ne/C burning occurs and forms mostly O. Willingale et al.
104: (2002) did a spectral mapping of Cas A using the data collected by
105: the XMM-Newton X-ray observatory ({\sl XMM-Newton}). They found that
106: the distributions of Si, S, Ar and Ca are very similar to each other
107: but distinct from Ne and Mg. This supports the above explosive
108: nucleosynthesis network. However, Hughes et al. (2000) and
109: Willingale et al. (2002) showed that the Fe-rich ejecta lie outside
110: the Si-rich material, using data from the Chandra X-ray observatory
111: ({\sl Chandra}) and {\sl XMM-Newton} respectively. It was concluded
112: that the ejecta in Cas A have undergone a spatial inversion of the
113: explosive O- and Si-burning products. That is to say, the materials
114: from the predicted ``onion'' layers have mixed together. Hughes et
115: al. (2000) proposed this inversion to be the result of
116: neutrino-driven convection during the initiation of the SN
117: explosion.
118:
119: {\sl Chandra} performed a 1 Ms observation on Cas A as a Very Large
120: Project (VLP), which was largely motivated by the two papers: Laming
121: \& Hwang (2003) and Hwang \& Laming (2003), focusing on the
122: suggested jet structure of the ejecta and the iron synthesis. Hwang
123: et al. (2004) showed the first sight of the observation. A jet
124: structure of the Si-rich ejecta was clearly found in the enhanced
125: Si-K emission line image. So it is proposed that Cas A was formed by
126: an asymmetric explosion, which is also indicated by the observed
127: kinematics and the high $^{44}$Ti yield (Vink 2004). Fe-rich ejecta,
128: however, was not found in the jet area, but perpendicular to it
129: instead. This contradicts to jet-induced explosion (Khokhlov et al.
130: 1999), but might be the case for collapsars (Nagataki et al. 2003;
131: Zhang et al. 2004). The explosion energy for Cas A was about $2 \sim
132: 4 \times 10^{51}$ ergs, suggesting it to be a normal SN (Laming \&
133: Hwang 2003). Cas A thus provides strong evidence that jets may also
134: be produced by normal SNe (Hwang et al. 2004).
135:
136: In this paper, we present high resolution X-ray spectroscopy of Cas
137: A using the {\sl Chandra} observations. Willingale et al. (2002) did
138: a similar work with the {\sl XMM-Newton} data. Here we would like to
139: cross-check their results. In our work, we use a smaller pixel size
140: (10$\arcsec$ $\times$10$\arcsec$, 1/4 of that used by Willingale et
141: al. (2002)), which is closer to the typical size of the ejecta
142: knots of Cas A. Another point is that the so-called ``jet'' region
143: is included in the analysis, while not in Willingale et al. (2002).
144: The high spatial and spectral resolutions of {\sl Chandra} together with
145: the abundant archived data permit us to study this famous SNR in
146: detail. In $\S$ 2, we briefly describe the observation and data
147: reduction. In $\S$ 3, we show the results, while some discussion in
148: $\S$ 4, and summary in $\S$ 5.
149:
150:
151:
152: \section{Observation and Data Reduction}
153:
154: Cas A was observed by {\sl Chandra} for about 1 Ms in 2004 as a VLP.
155: The details of the observations were described by Hwang et al.
156: (2004). In this paper, we used data from two deep ACIS exposure
157: segments of the VLP. One was performed from April 14 to 16
158: (Observation ID: 4638) and the other from May 25 to 26 (Observation
159: ID: 4639), with the exposure time of about 167 ks and 80 ks
160: respectively. The reasons that we used two segments are to ensure
161: good statistic for the spatially resolved analysis and to reduce the
162: mis-calibration effects, by using data collected in different
163: epochs.
164:
165:
166: The X-ray data were analyzed using CIAO software package (version
167: 3.3, with the CALDB version 3.2.1 and the ATOMDB version 1.3.1). In
168: order to do spatially resolved analysis, the X-ray emission region
169: was divided into pixels with size about 10$\arcsec$ $\times$
170: 10$\arcsec$, corresponding to a grid of 38 $\times$ 34. Figure 1
171: shows the image for observation ID 4639 with the grid overlaid. We
172: retained events within energies 0.5-10 keV using the archived level
173: 2 event file and created spectra for 960 pixels that contain at
174: least 3000 counts. The frequency distribution of the counts in each
175: pixel is shown in Figure 2 (left panel). We can see that most pixels
176: contain more than 10000 counts, which allows statistical reliable
177: spectral analysis. The background spectrum was created from the
178: off-source region.
179:
180: The spectral analysis was performed using the XSPEC (version 11.2)
181: package (Arnaud 1996). The spectra for each pixel of the two
182: observations were jointly fitted with two non-equilibrium ionization
183: components (VNEI, Borkowski et al. 2001). The free parameters are
184: the temperatures, emission measures, ionization ages ($\tau$={\sl
185: n$_e$t}) and abundances of O, Ne, Mg, Si, S, Ca and Fe for each of
186: the two components. The abundances are in units of solar abundances
187: given by Anders \& Grevesse (1989). We also introduced a uniform
188: redshift for each component to study the dynamics. The WABS model
189: (Morrison \& MacCammon 1983) was included to take care of
190: interstellar photo-electric absorption. A Gaussian line was added at
191: the energy of 3.4 keV to account for the Ar line emission, which is
192: not included in the VNEI model in XSPEC version 11.2 we used.
193:
194: Here we note two points. One is that the nonthermal emission
195: contributes to the 4 $\sim$ 6 keV continuum without any doubt, which
196: could be as high as 25$\%$ in Cas A (Willingale et al. 2002).
197: However, since its X-ray emission is dominated by the thermal
198: component (Laming 2001; Bleeker et al. 2001) and we are mainly
199: interested in the study of line emission in this ppaer, we believe
200: that our model is appropriate. The same spectral model was also used
201: by Willingale et al. (2002). The other is that Cas A is an O-rich
202: SNR, so that O contributes a significant fraction of the electrons
203: (Vink et al. 1996; Willingale et al. 2002). In this case, the O
204: abundance is coupled with the emission measure. Here we set the O
205: abundances of both components free in the fitting process, but we
206: will not discuss about the O abundance in this paper as it might be
207: contaminated.
208:
209:
210: \section{Results}
211:
212: The frequency distribution of the 960 reduced ${\chi}^2$ values of
213: the spectral fits is given in Figure 2 (right panel). It has a peak
214: around 0.7, which suggests our fitting results to be acceptable
215: statistically. In Figure 3, we give the spectra along with the
216: fitting residuals of several typical regions (marked in Figure 1).
217: The emission lines are also marked. Region A is located at the
218: outmost of the jet in the northeast. It has very strong Si, S, Ar
219: and Ca lines, but the Fe-K line is absent. Therefore it should be
220: dominated by O-burning products, but we cannot rule out the
221: incomplete Si-burning, as there seem to be the Fe-L lines around 1
222: keV (c.f. $\S$ 1). Region B is from the counter-jet, which is very
223: strong in Si and S. The Fe-K line is also clearly shown. This is
224: very consistent with the incomplete Si-burning yield out. Region C
225: is the Fe-rich region from the southeast. From the spectra we can
226: see very strong Fe-K and Fe-L lines, which are believed to be from
227: complete explosive Si-burning. Region D is located at the out rim,
228: which is weak in all the emission lines of the main elements in Cas
229: A and dominated by the continuum (c.f. Hughes et al. 2000, Figure
230: 3). These spectra show that the X-ray properties change
231: significantly across the remnant. In the following we present the
232: statistical studies of the spectral fitting results.
233:
234: (I) We got the spatial distribution of the absorbing column density
235: (c.f. Fig. 4), as well as temperatures and ionization ages for the
236: cool and hot components (c.f. Fig. 5). In order to show the large
237: scale structures more clearly, we smoothed all these maps with a
238: two-dimension Gaussian with FWHM 40$\arcsec$ $\times$ 40$\arcsec$,
239: and overplotted the smoothed contours on these maps. The column
240: density of the west part is higher than that of the east. This is
241: consistent with the former studies (Keohane et al. 1996; Willingale
242: et al. 2002), in which it was suggested that the absorption to the
243: west is higher due to the interaction of the remnant with the
244: molecular cloud. The column density lies in the range
245: $0.8-1.5\times10^{22}$ cm$^{-2}$ with the mean value of
246: 1.19$\times10^{22}$ cm$^{-2}$. The typical statistical error is
247: $\sim 5\%$ at 90\% confidence level. The column density is a little
248: smaller than 1.50$\times10^{22}$ cm$^{-2}$ derived by Willingale et
249: al. (2002), but consistent with that from Keohane et al. (1996)
250: given by the equivalent width of HI and OH ($1.05-1.26
251: \times10^{22}$ cm$^{-2}$).
252:
253: The temperature distributions for the cool and hot components are
254: not very similar. It seems that the higher temperature of the hot
255: components appears at the out rim and Fe-rich regions (c.f. Fig 6).
256: This is not surprising, since the hot component is responsible for
257: most of the Fe K emission and also dominates continuum above 4 keV
258: (Willingale et al. 2002).
259:
260: The ionization age maps for both components are much similar to the
261: corresponding ones given by Willingale et al. (2002). We can see
262: that the map of the hot component is relatively uniform over the
263: whole remnant, while the cool one shows some structures. The
264: brightest shell, which has the largest density, doesn't have the
265: highest ionization age. This leads to the conclusion that the
266: brightest shell is later shocked. Considering that the cool
267: component is associated with the reverse shock (Vink et al. 1996,
268: Willingale et al. 2002), this suggests that the reverse shocked
269: ejecta may be stratified and thus shocked at different time.
270:
271:
272: (II) We obtained the spatial distribution of all the elements
273: concerned. The elemental abundance of each pixel is the mean of the
274: two components weighted by their emission measures.
275:
276: Figure 6 displays the abundance maps of Si, Ca and Fe. The contours
277: have the similar meaning with those in Fig. 4. We can find that Si
278: and possibly Ca abundances show a jet-like structure in the
279: northeast and a counterpart in the southwest. This ``jet'' structure
280: was first suggested by using the Si emission line equivalent width
281: image (Hwang et al. 2000) and was confirmed with the ratio image of
282: the Si He $\alpha$ (plus continuum, $1.78\sim2.0$ keV) and
283: $1.3\sim1.6$ keV (mostly weak Mg He $\alpha$, Fe L plus continuum)
284: (Hwang et al. 2004). Our abundance maps further confirm their
285: results. Meanwhile, from the Fe abundance map, we can find that Fe
286: is relatively poor in the jet area but seems to be rich in the
287: direction perpendicular to it. It is also clearly shown in the map
288: that the Fe-rich knots lie outside the Si- and S-rich ejecta, which
289: is consistent with the results of Hughes et al. (2000) and
290: Willingale et al. (2002).
291:
292: (III) Figure 7 $-$ 9 are the correlation plots of the abundances of
293: Ne, Si, and Fe with the other elements. S abundance is strongly
294: correlated with that of Si. Its correlation with other elements are
295: very similar to those of Si, and are therefore not displayed. We
296: note that the minimum points are excluded, for they represent
297: regions which are dominated by the the continuum emission and thus
298: have no (or very weak) corresponding lines. In order to show the
299: overall trends of the correlations more clearly, we binned the data
300: points into 10 channels and overplotted them in the figures with big
301: pentacles. The corresponding coherent coefficients are given in
302: Table 1. Apparently Si is well correlated with S, as well as with
303: Ca. There is also a good correlation between Ne and Mg.
304:
305: We noticed that in the abundance correlation maps of Si versus Fe,
306: there seems to be a two phase correlation with the break at Si
307: abundance equaling to 3. When the Si abundance is lower than 3,
308: there is a positive correlation with Fe, while negative when higher.
309: All these features would give us information about the (explosive)
310: nucleosynthesis of the SN, as will be discussed in detail in $\S$
311: 4.3.
312:
313: (IV) The Doppler map (c.f. Fig. 4) we got is different from those of
314: Willingale et al. (2002). They showed that the ejecta in the
315: Southeast is generally blue shifted, while in the North red shifted.
316: However, in our map, most regions are red shifted and only a small
317: fraction in the Southeast are blue shifted. The Doppler map shows
318: prominent ``bar''-like features from the southeast to the northeast.
319:
320: We found that the Doppler map is mostly the result of
321: mis-calibration. The direction of the ``bar''-like feature is
322: exactly along the CCD columns. Since the CCD pixels in one column
323: share one read-out circuit, the red shift ``bar'' is probably due to
324: the inefficiency of the read-out circuit. The generally accepted
325: velocity for the knots is in the order of 2000 km/s (Willingale et
326: al. 2002), corresponding to an energy shift of $\sim$15 eV @ 2 keV
327: (near the Si and S lines). However, the resolution of {\sl
328: Chandra}-ACIS is about an order of magnitude higher than this value.
329: The calibration and/or the performance of {\sl Chandra}-ACIS is
330: probably not good enough to determine such a small energy shift.
331: Therefore, we conclude that it is very difficult to derive a
332: reliable Doppler map with the {\sl Chandra}-ACIS data.
333:
334:
335:
336: \section{Discussion}
337:
338: \subsection{Jet structure}
339:
340: The ejecta structure and the explosion asymmetry of Cas A have been
341: widely studied. It has been long suggested that the core-collapse SN
342: explosion process is intrinsically asymmetric, from both the
343: observational and theoretical point of views (Khokhlov et al. 1999
344: and reference therein). For Cas A, it shows rapidly moving
345: oxygen-rich material outside the nominal boundary (Fesen \&
346: Gunderson 1996) and evidence for two oppositely directed jets (Reed
347: et al. 1999). The aspherical distribution of its ejecta suggests
348: that the SN explosion, which produced Cas A, may be asymmetric.
349:
350: Hwang et al. (2000, 2004) have further supported the jet emission of
351: Si-rich ejecta using the {\sl Chandra} data. These works are based
352: on the equivalent width (EW) images of the corresponding emission
353: lines. However the correlation between EW and the abundance can be
354: distorted by temperature, column density and ionization age. In
355: other words, the same abundance can be detected as significantly
356: different EW because of the different line emissivity that is a
357: function of the local plasma conditions. Willingale et al. (2002)
358: gave the abundance maps from the {\sl XMM-Newton} data, but the jet
359: structure is not seen in their results, due to the spatial
360: resolution limits and their bigger pixel size.
361:
362: In this paper, we gave directly the element abundances map, among
363: which Si, S and probably Ca show the jet structure as expected. The
364: ejecta in the jet area are very enriched in Si, S and Ca, and
365: relatively poor in Fe (still exists, c.f. Figure 3(B)). This
366: confirms not only the presence of jet itself but also that these
367: elements come from incomplete explosive Si-burning rather than
368: O-burning (KhoKhlov et al. 1999; Hughes et al. 2000). Meanwhile, the
369: Fe in the counter-jet region (Figure 3(B)) is less enriched than the
370: off-jet area (Figure 3 (C)), which also implies that the jet
371: material did not emerge from as deep in the progenitor as other
372: observed ejecta(Hughes et al. 2000 and reference therein).
373:
374:
375: \subsection{Iron distribution}
376:
377: As far as we know, $^{56}$Fe mainly comes from the decay of
378: $^{56}$Ni, which is synthesized in the (incomplete) explosive
379: Si-burning. Hwang \& Laming (2003) identified a region of nearly
380: pure Fe ejecta, which might be a candidate for the site of
381: $\alpha$-rich freezeout (see also Vink 2004). Early {\sl Chandra}
382: observation showed that the Fe emission in Cas A is associated with
383: ejecta and mainly distributed in the faint region outside the bright
384: Si-dominated shell in the east. The lack of Si beyond the Fe
385: emission, as well as the lack of Fe inside the Si emission, argues
386: against projection effects (Hughes et al. 2000).
387:
388: From our abundance maps, we further confirmed that the ejecta which
389: is highly enriched of Fe lies outside the Si and S dominated ejecta
390: in the east. According to the onion model, the lighter an element
391: is, the larger radius it should be located (Aschenbach 2002). As
392: long as the projection effect may not play, at least an important
393: role, this means that the ejecta in Cas A may have suffered some
394: kind of overturn, either during or after the explosion. This
395: overturn might be the result of the neutrino-driven convection
396: during the initiation of the SN explosion (Hughes et al. 2000).
397:
398: In the previous {\sl Chandra} results (Hwang \& Laming 2003; Laming
399: \& Hwang 2003), it has been suggested that the Fe-rich ejecta have
400: higher characteristic ionization age than the O/Si knots by factors
401: of a few to 10, implying either a correspondingly higher density or
402: an earlier shock time. As large density enhancement (or deficits) is
403: subject to a number of hydrodynamic instabilities that might destroy
404: the knots within a few shock crossing time scales (Klein et al.
405: 1994; Klein et al. 2003; Wang \& Chevalier 2001; Poludnenko et al.
406: 2002; Poludnenko et al. 2004), such knots are thus not expected to
407: be seen now. So Hwang \& Laming (2003) claimed that early shock time
408: and very modest (or no) density enhancement are the most consistent
409: explanation for the Fe knots that have survived to be seen today.
410: Comparing Figure 4 (middle panel) and the Fe abundance map in Figure
411: 6, we can find that the Fe-rich knots in the southeast do have
412: higher ionization age. This further confirms that they might be
413: early shocked, and might be additinal evidence for the overturn
414: described previously.
415:
416:
417: \subsection{Nucleosynthesis}
418:
419: In the range of about two orders of magnitude, the abundances of Si,
420: S and Ca show tight correlations with each other (c.f. Fig. 8).
421: Willingale et al. (2002) also presented the similar results using
422: the {\sl XMM-Newton} data and suggested this to be strong evidence
423: for the nucleosynthesis of these elements by explosive O-burning and
424: incomplete explosive Si-burning due to the shock heating of these
425: layers in the core-collapse supernova. This kind of good correlation
426: can also be found between abundances of Ne and Mg, which means that
427: they might be the ashes of explosive C/Ne burning. All these results
428: are generally consistent with the (explosive) nucleosynthesis theory
429: of a massive star (Woosley et al. 2002; Woosley \& Janka 2006).
430:
431: Fe abundance is not correlated with any other element, with one
432: exception that there seems to be a possible two phase correlation
433: with Si (c.f. Fig. 8 or 9). It is contaminated by extremum data
434: points, for it is difficult to reliably derive the Fe abundances
435: from Fe-L lines (e.g. Suh et al. 2005). So we picked regions whose
436: spectra show clearly Fe K lines, and re-draw the Si versus Fe
437: abundance plot in Figure 10. Now the two phase correlation is more
438: clearly presented. From Figure 6 we can easily find that the regions
439: with Si abundance higher than 3 generally concentrate at the
440: Si-dominated bright shell, which is natural, and also at the jet and
441: its counterpart. These regions are believed to be the ejecta of
442: incomplete explosive Si-burning or explosive O-burning. We know that
443: explosive O-burning would lead to products with little or no Fe
444: while in the incomplete explosive Si-burning Fe is one of the main
445: products. From the Si versus Fe correlation plot, we can see that Fe
446: is not as enriched as Si, but still abundant. This would support
447: that these ejecta mainly come from incomplete explosive Si-burning,
448: but also mix with O-burning products. The regions with lower Si
449: abundance are mainly the relatively faint parts of the remnant and
450: might be dominated by the shocked circumstellar medium (CSM). We
451: note here that the southeast ejecta, which are highly enriched in Fe
452: and are believed to be synthesized in the complete explosive
453: Si-burning, are only a small fraction of the total remnant. They
454: don't affect much to the correlation we discussed above.
455:
456:
457: \section{Summary}
458:
459: We did a spatially resolved X-ray spectroscopy of SNR Cas A. An
460: obvious jet structure can be found in the Si, S and Ca abundance
461: maps, further confirming the former suggestions. However, the Fe map
462: shows that it doesn't follow the jet, but distributes somehow
463: perpendicular to it. Meanwhile, it lies outside the lighter elements
464: (such as Si and S), which is consistent with the previous results
465: and might be due to the neutrino-driven convection during the
466: initiation of the SN explosion. The tight positive correlations of
467: Si, S and Ca abundances (c.f. Figure 8) suggest that these elements
468: come from explosive O-burning and incomplete Si-burning. Ne and Mg
469: abundances also show a good positive correlation, which means that
470: they should be the ashes of explosive C/Ne burning. The Fe abundance
471: is positively correlated with that of Si when Si abundance is lower
472: than 3 solar value, but when the Si abundance is higher, it appears
473: a negative correlation. We propose this two phase correlation the
474: result of the ways in which Fe is synthesized. The highly
475: Si-enriched ejecta concentrated in the jet and the bright shell are
476: probably a mixture of the explosive O-burning and the incomplete Si-
477: burning products. The rest part, in contrast, might be dominated by
478: the shocked CSM.
479:
480:
481: \begin{acknowledgements}
482: We acknowledge the use of data obtained by Chandra. The Chandra
483: Observatory Center is operated by the Smithsonian Astrophysical
484: Observatory for and on the behalf of NASA. This work is supported by
485: the National Science Foundation of China through grants 10533020 and 10573017.
486: \end{acknowledgements}
487:
488:
489: \begin{thebibliography}{}
490:
491: \bibitem[2002]{Aschenbach} Aschenbach B., ~2002, in Neutron Stars, Pulsars, and Supernova Remnants, ed. Becker W., Lesch H., and Tr\"umper J. (MPE Rep. 278; Garching: MPE), 13
492:
493: \bibitem[1996]{Arnaud} Arnaud K. A., ~1996, ASPC, 101, 17
494:
495: \bibitem[1989]{Anders} Anders E., Grevesse N., ~1989, Geochimica et Cosmochimica Acta, 53, 197
496:
497: \bibitem[2001]{Bleeker} Bleeker J. A. M., Willingale R., van der Heyden K. J. et al., ~2001, A\&A, 365, L225
498:
499: \bibitem[2001]{Borkowski} Borkowski K. J., Lyerly W. J., Reynolds S. P., ~2001, ApJ, 548, 820
500:
501: \bibitem[1996]{Fesen} Fesen, R. A., Hurford A. P., ~1996, ApJS, 106, 563
502:
503: \bibitem[2001]{Gotthelf} Gotthelf E. V., Koralesky B., Rudnick L. et al., ~2001, ApJ, 552, L39
504:
505: \bibitem[2000]{Hughes} Hughes J. P., Rakowski C. E., Burrows D. N. et al., ~2000, ApJ, 528, L109
506:
507: \bibitem[2000]{Hwang} Hwang U., Holt S. S., Petre R., ~2000, ApJ, 537, L119
508:
509: \bibitem[2003]{Hwang} Hwang U., Laming J. M., ~2003, ApJ, 597, 362
510:
511: \bibitem[2004]{Hwang} Hwang U., Laming J. M., Badenes C. et al., ~2004, ApJ, 615, L117
512:
513: \bibitem[1996]{Keohane} Keohane J. W., Rudnick L., Anderson M. C., ~1996, ApJ, 466, 309
514:
515: \bibitem[2003]{Klein} Klein R. I., Budil K. S., Perry T. S. et al., ~2003, ApJ, 583, 245
516:
517: \bibitem[1994]{Klein} Klein R. I., McKee C. F., Colella P., ~1994, ApJ, 420, 213
518:
519: \bibitem[1999]{Khokhlov} Khokhlov A. M., H\"oflich P. A., Oran E. S. et al., ~1999, ApJ, 524, L107
520:
521: \bibitem[2001]{Laming} Laming J. M., ~2001, ApJ, 546, L1149
522:
523: \bibitem[2003]{Laming} Laming J. M., Hwang U., ~2003, ApJ, 597, 347
524:
525: \bibitem[1974]{Mckee} Mckee C. F., ~1974, ApJ, 188, 335
526:
527: \bibitem[1983]{Morri} Morrison R., McCammon D., ~1983, ApJ, 270, 119
528:
529: \bibitem[2003]{Nagataki} Nagataki S., Mizuta A., Yamada S. et al., ~2003, ApJ, 596, 401
530:
531: \bibitem[2002]{Poludnenko} Poludnenko A. Y., Frank A., Blackman E. G., ~2002, ApJ, 576, 832
532:
533: \bibitem[2002]{Poludnenko} Poludnenko A. Y., Dannenberg K. K., Drake R. P. et al., ~2004, ApJ, 604, 213
534:
535: \bibitem[1995]{Reed} Reed J. E., Hester J. J., Fabian A. C. et al., ~1995, ApJ, 440, 706
536:
537: \bibitem[1973]{Shklovsky} Shklovsky I. S., ~1973, Soviet Astron. 16, 749
538:
539: \bibitem[2005]{Suh} Suh J. A., Audard M., G\"udel M. et al., ~2005, ApJ, 630, 1074
540:
541: \bibitem[1996]{Thielemann} Thielemann F.-K, Nomoto K., Hashimoto M., ~1996, ApJ, 460, 408
542:
543: \bibitem[2001]{Thorstensen} Thorstensen J.R., Fesen R.A., van den Bergh S., ~2001, AJ, 122, 297
544:
545: \bibitem[1996]{Vink} Vink J., Kaastra J.S., Bleeker J.A.M., ~1996, A\&A, 307, L41
546:
547: \bibitem[2004]{Vink} Vink J., ~2004, New Astro. Rev., 48, 61
548:
549: \bibitem[2001]{Wang} Wang C. Y., Chevalier R. A., ~2001, ApJ, 549, 1119
550:
551: \bibitem[2002]{Willingale} Willingale R., Bleeker J. A. M., van der Heyden K. J. et al., ~2002, A\&A, 381, 1039
552:
553: \bibitem[2002]{Woosley} Woosley S. E., Heger A., Weaver T. A., ~2002, Rev. Mod. Phys, 74, 1015
554:
555: \bibitem[2005]{Woosley} Woosley S. E., Janka T., ~2005, Nat. Phys., 1, 147
556:
557: \bibitem[1995]{Woosley} Woosley S. E., Weaver T. A., ~1995, ApJS, 101, 181
558:
559: \bibitem[2004]{Zhang} Zhang W., Woosley S. E., Heger A., ~2004, ApJ, 608, 365
560:
561: \end{thebibliography}
562:
563:
564: \begin{table*}
565: \caption[]{coherent coefficient for data points excluding extremum ones.}
566: \label{obslog}
567: \begin{tabular}{cccccccc}
568: \noalign{\smallskip} \hline \hline \noalign{\smallskip}
569: Element&O&Ne&Mg&Si&S&Ca&Fe\\
570: \noalign{\smallskip} \hline \noalign{\smallskip}
571:
572: O & - & 0.33 & 0.41 & 0.33 & 0.26 & 0.14 & 0.13 \\
573: Ne & 0.33 & - & 0.49 & 0.19 & 0.16 & 0.05 & 0.25 \\
574: Mg & 0.41 & 0.49 & - & 0.48 & 0.39 & 0.16 & 0.24 \\
575: Si & 0.33 & 0.19 & 0.48 & - & 0.86 & 0.35 & 0.23 \\
576: S & 0.26 & 0.16 & 0.39 & 0.86 & - & 0.33 & 0.11 \\
577: Ca & 0.14 & 0.05 & 0.16 & 0.35 & 0.33 & - & 0.05 \\
578: Fe & 0.13 & 0.25 & 0.24 & 0.23 & 0.11 & 0.05 & - \\
579:
580: \noalign{\smallskip} \hline \noalign{\smallskip}
581: \end{tabular}
582: \end{table*}
583:
584:
585: \begin{figure*}
586: \includegraphics[width=8cm,clip]{figure1.eps}
587: \caption{The pixel grid used in our analyses superimposed on the
588: {\sl Chandra} image of Cas A (Observation ID: 4639). Region A, B, C
589: and D are described in $\S$ 3 and their spectra are presented in
590: Fig. 3. }
591: \end{figure*}
592:
593:
594: \begin{figure*}
595: \includegraphics[width=6cm,clip]{figure2a.ps}
596: \includegraphics[width=6cm,clip]{figure2b.ps}
597: \caption{Frequency distributions of the 960 pixel counts (left
598: panel) and ${\chi}^2$ values obtained in the spectra fits (right
599: panel).}
600: \end{figure*}
601:
602: \begin{figure*}
603: \includegraphics[width=16cm,clip]{figure3.eps}
604: \caption{Spectra created from regions marked in Fig. 1. Spectrum A represents the spectrum of
605: the explosive O-burning products, while B and C of incomplete and complete explosive Si-burning
606: products respectively. Spectrum D is that of the forward shocked ISM. }
607: \end{figure*}
608:
609:
610: \begin{figure*}
611: \includegraphics[width=15cm,clip]{figure4.eps}
612: \caption{The absorbing column density ($N_H$, $10^{22}$ cm$^{-2}$,
613: left panel), ionization age (Log10 $n_{e}t$ cm$^{-3}$ s, middle
614: panel), and red shift (right panel) maps of Cas A.
615: The negative value of redshift means blueshift. The coding
616: used is shown on the top of each panel. The overplotted contours
617: represent the same dataset but smoothed by a two-dimension Gaussian
618: component with FWHM of $40\arcsec \times 40\arcsec$. The contour
619: levels are as follows: 1.1, 1.2, 1.3 for $N_H$; 11.0, 11.2, 11.3 for
620: Log10 $n_{e}t$; and 0.01, 0.012, 0.015 for redshift. }
621: \end{figure*}
622:
623:
624: \begin{figure*}
625: \includegraphics[width=12cm,clip]{figure5.eps}
626: \caption{The maps of temperatures and ionization ages for the cool
627: (index 1) and hot (index 2) components of Cas A. a: $kT_1$ (keV), b:
628: $kT_2$ (keV), c: Log10 $(n_{e}t)_1$ (cm$^{-3}$ s), d: Log10
629: $(n_{e}t)_2$ (cm$^{-3}$ s). The contours represents the
630: corresponding images smoothed by the same Gaussian component as used
631: in Fig. 4. Contour levels are 1.0, 1.1, 1.2 for $kT_1$, and 2.6,
632: 2.8, 3.0 for $kT_2$, while 11.0, 11.2, 11.3 for the two Log10
633: $(n_{e}t)$. }
634: \end{figure*}
635:
636:
637: \begin{figure*}
638: \includegraphics[width=15cm,clip]{figure6.eps}
639: \caption{Si, Ca and Fe abundance (in units of solar abundance) maps
640: of Cas A. Contours overplotted represent the same dataset but
641: smoothed the same Gaussian component as used in Fig. 4. The contour
642: levels are 2.0, 3.0, 4.0; 1.0, 2.0, 3.0, 4.0; and 0.3, 0.6, 0.9, 1.2
643: respectively. }
644: \end{figure*}
645:
646:
647: \begin{figure*}
648: \includegraphics[width=6cm,clip]{figure7.ps}
649: \caption{The abundance correlation plots between Ne and those of
650: other elements. All the points were binned into 10 channels and
651: overplotted in the figure with the big pentacles connected with the
652: solid line. }
653: \end{figure*}
654:
655: \begin{figure*}
656: \includegraphics[width=6cm,clip]{figure8.ps}
657: \caption{The Si abundance correlation plots versus those of other elements.
658: The big pentacles and solid line have the same meaning as in Figure 7.}
659: \end{figure*}
660:
661: \begin{figure*}
662: \includegraphics[width=6cm,clip]{figure9.ps}
663: \caption{The Fe abundance correlation plots versus those of other elements.
664: The big pentacles and solid line have the same meaning as in Figure 7. }
665: \end{figure*}
666:
667: \begin{figure*}
668: \includegraphics[width=8cm,clip]{figure10.ps}
669: \caption{The Si versus Fe abundance correlation from regions with
670: significant Fe-K lines. }
671: \end{figure*}
672:
673:
674: \end{document}
675: