A&A 365, L242-L247 (2001)

High-resolution X-ray spectroscopy and imaging of supernova remnant N132D

E. Behar¹ - A. P. Rasmussen¹ - R. G. Griffiths² - K. Dennerl³ - M. Audard⁴ -
B. Aschenbach³ - A. C. Brinkman⁵

Send offprint request: E. Behar

1 - Columbia Astrophysics Laboratory, 550 West 120th Street, New York, NY 10027, USA
2 - X-Ray Astronomy Group, Physics and Astronomy, University of Leicester, Leicester LE1 7RH, UK
3 - Max-Planck-Institut für Extraterrestrische Physik, 85740 Garching, Germany
4 - Laboratory for Astrophysics, Paul Scherrer Institute, Würenlingen and Villigen, 5232 Villigen PSI, Switzerland
5 - Space Research Organization of the Netherlands, Sorbonnelaan 2, 3548 CA, Utrecht, The Netherlands

Received 2 October 2000 / Accepted 30 October 2000

Abstract
The observation of the supernova remnant N132D by the scientific instruments on board the XMM-Newton satellite is presented. The X-rays from N132D are dispersed into a detailed line-rich spectrum using the Reflection Grating Spectrometers. Spectral lines of C, N, O, Ne, Mg, Si, S, and Fe are identified. Images of the remnant, in narrow wavelength bands, produced by the European Photon Imaging Cameras reveal a complex spatial structure of the ionic distribution. While K-shell Fe emission seems to originate near the centre, all of the other ions are observed along the shell. A high O⁶⁺/O⁷⁺emission ratio is detected on the northeastern edge of the remnant. This can be a sign of hot ionising conditions, or it can reflect relatively cool gas. Spectral fitting of the CCD spectrum suggests high temperatures in this region, but a detailed analysis of the atomic processes involved in producing the O⁶⁺ spectral lines leads to the conclusion that the intensities of these lines alone cannot provide a conclusive distinction between the two scenarios.

Key words: atomic processes - line: formation - line: identification - techniques: spectroscopic - supernovae: individual: N132D - ISM: supernova remnants

Author for correspondance: behar@astro.columbia.edu

1 Introduction

N132D, one of the brightest soft X-ray sources in the Large Magellanic Cloud (LMC), was identified as a type II supernova remnant (SNR) by Westerlund & Mathewson (1966). The high velocities of the oxygen-rich (O-rich) ejecta near the centre of the remnant were first detected by Danziger & Dennefeld (1976) and later measured precisely by Morse et al. (1995). The fast moving ejecta is believed to originate from a core collapse supernova (SN) of a massive star. Blair et al. (2000) have recently performed a thorough optical and UV study of N132D using the Hubble Space Telescope (HST). With the exceptionally high spatial resolution of the HST, they were able to distinguish between SN ejecta and swept-up interstellar matter (ISM). Observing a knot on the northwestern rim of the remnant, they found elemental abundances that agree fairly well with those previously obtained from X-ray observations of the entire remnant (Hughes et al. 1998; Favata et al. 1998), and which are also consistent with the mean LMC abundances. In the ejecta, on the other hand, only C, O, Ne, and Mg could be detected in the optical/UV spectra. The absence of O-burning elements, such as Si, S, and Fe, in the ejecta has led Blair et al. (2000) to suggest that N132D may be a product of a type Ib SN explosion.

N132D was first observed in the X-rays with the Einstein Observatory (Mathewson et al. 1983). Using the spectrometers on board Einstein, Hwang et al. (1993) detected individual emission lines of O⁷⁺, O⁶⁺, Ne⁹⁺, Ne⁸⁺, and Fe¹⁶⁺. Comparing with the ROSAT images, Morse et al. (1995) found little overlap between the optical and soft X-ray emitting regions. The X-ray emission seemed to come from a large shell of shocked ISM gas (see also Williams et al. 1999), which is also the source of strong synchrotron radio emission (Dickel & Milne 1995). Hughes et al. (1998) fitted the ASCA CCD spectrum of N132D, obtaining mean LMC abundances, which further suggested that the X-rays in the remnant are predominantly emitted by shocked ISM.

In this work, we present the X-ray spectrum of N132D measured with the Reflection Grating Spectrometers (RGS), complemented by the images taken with the European Photon Imaging Cameras (EPIC), both on board XMM-Newton. Owing to the uniquely high dispersion of the RGS, almost all of the spectral features emitted by this extended ($\sim$120 $^{\prime \prime }$) source are resolved. Along with the RGS observation of 1E0102.2-7219 in the Small Magellanic Cloud (Rasmussen et al. 2001), these are the first highly resolved X-ray spectra of extended SNRs.

2 Observation and data reduction

The XMM-Newton X-ray observatory (Jansen et al. 2001) incorporates a payload with two identical high-resolution RGS spectrometers (den Herder et al. 2001) and three EPIC cameras (Turner et al. 2001; Strüder et al. 2001). The spectroscopic foci are separate from the telescope foci, where the cameras are located, allowing the cameras and the spectrometers to operate simultaneously. N132D was observed by XMM-Newton as part of the Performance Verification program during May, 2000. The RGS data were acquired with an effective exposure time of 53 ksec for each spectrometer. The EPIC-MOS cameras were operated with the medium filter in place, giving effective exposure times of 18 and 21 ksec. The EPIC-PN camera was operated in full-frame mode using the medium filter for an effective exposure time of 23.2 ksec.

The RGS data were processed using custom software developed at Columbia University. The background subtracted CCD events were plotted in dispersion versus pulse-height space to separate the spectral orders. An especially large (cross-dispersion) spatial extraction of 135 $^{\prime \prime }$ was required in order to cover the extended remnant. In order to determine the flux and provide the channel-to-wavelength transformation, the count-rate spectra were folded through the response matrices specifically generated for this observation, utilizing the ground calibration (Rasmussen et al. 1998). This entire procedure was performed separately for each spectrometer and for first and second orders. Note that N132D's large angular size and velocity structure tend to broaden the spectral features irregularly, thus, preventing accurate wavelength measurements of individual lines. Fortunately, this does not impede the line identifications.

The reduction of raw EPIC-MOS event lists, which was performed using the standard Science Analysis Software, involves the subtraction of hot, dead, or flickering pixels, as well as removal of events due to electronic noise. EPIC-MOS images of the whole remnant were extracted using events with pattern 0-12 only. EPIC-PN data was reduced according to standard procedures. N132D covered the CCD chips 4 and 7, which required dedicated individual corrections. For the EPIC-PN images, areas of 44 $\times$ 52 pixels were extracted. The extraction area was adaptively increased around each pixel to contain at least 200 single-pixel events in the energy range of 4.4-10 keV. In order to get reliable statistics for the spectral fits, the data were rebinned to 0.16 keV bins.

3 Results

3.1 RGS spectrum

The total RGS spectrum of N132D is plotted in Fig. 1. The data from the first and second orders in both RGS have been combined in order to improve the statistical reliability. The line identifications are presented in Table 1 for the strongest lines in accordance with the peak labels in Fig. 1. For blends, only the dominant contributing lines are given in the table. Lines of C, N, O, Ne, Mg, Si, S, and Fe are unambiguously detected in the spectrum; the strongest lines originating from K-shell O and Ne, and L-shell Fe. The ratio of the strong Fe¹⁶⁺ lines at $\sim$15 and 17 Å seem to indicate that the iron has equilibrated. At least six consecutive charge states of iron, namely Fe¹⁶⁺-Fe²¹⁺, are observed. Fe²²⁺ might also be present in the 11.75 Å blend (label 8). This wide range of Fe charge states indicates an electron temperature ( $kT_{\rm e}$) range of 0.2-1 keV in collisional equilibrium (Mazzotta et al. 1998). The absence of Fe²³⁺, which would have produced strong lines at $\sim$10.6 and 11.2 Å, provides the upper limit for the temperature, whereas the lower limit is not well constrained as charge states lower than Fe¹⁶⁺ do not regularly emit X-rays. However, line emission from Fe²⁴⁺ is detected with EPIC-PN as discussed below.

	Figure 1: The RGS fluxed spectrum of N132D. See Table 1 for line identifications
Open with DEXTER

 

 

Table 1: Spectral lines in the RGS spectrum of N132D (Fig. 1). The references for the expected wavelengths ( $\lambda _{\rm expected}$) are (1) Johnson & Soff (1985), (2) Drake (1988), (3) Brown et al. (1998) (4) Brown et al. (2000) (5) Present HULLAC (Bar-Shalom et al. 1998) calculations

Label	$\lambda _{\rm expected}$ (Å)	Reference	Ion
1	6.648	2	Si12+
	6.685	2	Si12+
	6.739	2	Si12+
2	8.421	1	Mg11+
3	9.169	2	Mg10+
	9.228	2	Mg10+
	9.314	2	Mg10+
4	9.708	5	Ne9+
5	10.238	5	Ne9+
6	11.009	5	Ne8+
7	11.554	5	Ne8+
8	11.742	4	Fe22+
	11.770	4	Fe21+
9	12.135	1	Ne9+
10	12.846	4	Fe19+
	12.864	4	Fe19+
11	13.447	2	Ne8+
	13.518	4	Fe18+
	13.550	2	Ne8+
12	13.698	2	Ne8+
13	14.208	4	Fe17+
14	15.015	3	Fe16+
15	15.261	3	Fe16+
16	16.005	1	O7+
	16.004	4	Fe17+
17	16.779	3	Fe16+
18	17.051	3	Fe16+
	17.096	3	Fe16+
19	17.623	4	Fe17+
20	18.627	5	O6+
21	18.967	1	O7+
22	21.602	2	O6+
	21.804	2	O6+
23	22.097	2	O6+
24	24.779	1	N6+
25	26.988	5	C5+
26	28.464	5	C5+
27	28.780	2	N5+
	29.082	2	N5+
28	29.535	2	N5+
29	30.436	5	S13+
30	32.436	5	S13+
	32.579	5	S13+
31	33.734	1	C5+

Modeling the profiles and fluxes of the individual spectral lines in the RGS spectrum is a rather difficult task. The line profiles are correlated with the shape of the remnant, the velocity broadening, and the point spread function of the instrument (telescope + spectrometer). The fact that the former two effects may strongly depend on the ion species makes this problem especially challenging. Therefore, with the sole exception of the O⁶⁺ lines discussed below, a detailed line-flux analysis is deferred to a later study.

3.2 EPIC-MOS soft X-ray images

Figure 2 shows the EPIC-MOS images of N132D in several narrow wavelength bands that overlap with those of the RGS. The data (3 $^{\prime \prime }$ $\times$ 3 $^{\prime \prime }$ pixels) in each band are combined from the two cameras, smoothed using a Gaussian function (sigma = 1.5 $^{\prime \prime }$), and then normalised to the brightest pixel in each image. Each narrow-band image is 2$^\prime$ $\times$ 2.5$^\prime$ wide. All of the images feature strong X-rays from the shell, in particular from the southeastern, and northwestern edges. Differences between more and less ionised regions can be noticed. The emission by the most highly ionised species Ne⁹⁺, Mg¹⁰⁺, Si¹²⁺, and Fe¹⁹⁺ is brightest on the southeastern tip, while O⁶⁺, O⁷⁺, Ne⁸⁺, Fe¹⁶⁺, and Fe¹⁷⁺ exhibit strong emission on the northwestern rim as well. This is the ISM shocked region observed by Blair et al. (2000). O⁷⁺ emission is also intense at the centre of the remnant, indicating that part of the X-ray flux may be associated with the SN ejecta. The O⁶⁺ image is unique by virtue of the bright knot on the northeastern edge, just between the two broad inlets that extend into the ambient ISM. In fact, no other element seems to be emitting X-rays in that region. The EPIC-MOS ratio image of O⁶⁺/O⁷⁺ shown in Fig. 3 illustrates the excess of O⁶⁺ emission on the northeastern rim, where the ratio attains a value of $0.79 \pm 0.01$ in contrast with $0.60 \pm 0.01$ in the centre. The RGS dispersed images for these lines (not presented) are consistent with this picture. The O⁶⁺ brightness might simply be due to relatively cold gas. However, the special morphology in that region may, alternatively, suggest that the shock wave is interacting with relatively dense and cold O-rich material, possibly from pre-SN stellar winds. In that case, this newly shocked material can be very hot and ionising. It should be noted that Banas et al. (1997) have observed a giant molecular cloud near the southern edge of N132D, whereas no clouds were detected to the north of the remnant.

	Figure 2: EPIC-MOS images of N132D in the narrow wavelength bands indicated. Each image is labeled with the principal line-emitting ion in its particular band
Open with DEXTER

	Figure 3: EPIC-MOS O6+/O7+ ratio image of N132D. The average ratios in the circular regions (15 $^{\prime \prime }$ radius) are $0.79 \pm 0.01$ near the northeastern rim and $0.60 \pm 0.01$ in the centre
Open with DEXTER

3.3 EPIC-PN hard X-ray images

The EPIC-PN camera is more sensitive to hard X-rays than the RGS and EPIC-MOS. The EPIC-PN CCD spectrum of the entire N132D remnant (not presented) shows K-shell features of Ar and Ca, as well as a prominent Fe²⁴⁺ feature at 1.86 Å ( $6.67 \pm 0.01$ keV). This Fe-K feature was also observed with BeppoSAX (Favata et al. 1998). In Fig. 4a, flux images of N132D in this line (blend) are presented. In order to separate continuum from line flux, each spectrum was fitted with a Gaussian line superimposed on a bremsstrahlung continuum.

The N132D structure as seen in the EPIC-PN images (Fig. 4) is completely different from the structure in the EPIC-MOS images (Fig. 2). Unlike the softer emitting material that exhibits a shell with some filaments, the harder Fe-K line originates in a well-defined region slightly offset from the centre of the remnant to the southeast, where there is no sign for Fe-L emission. This is consistent with an out moving blast wave; the material closer to the centre that was shocked earlier is more highly ionised. If this material is associated with the SN ejecta, it might suggest that the high-Z elements are not absent from the ejecta, but rather that they could not be detected at shorter wavelengths because they are too ionised. The northwest and northeast regions show almost no Fe²⁴⁺ emission. The continuum flux is highest near the centre (Fig. 4c), which makes N132D look like a centre-filled remnant, rather than a shell-type remnant, as suggested by the soft X-ray images. The highest equivalent width (EW) values of more than a few keV are found in a confined northern region, where the Fe abundance is probably high and the overall density low. The hottest regions in N132D are found close to the northern inlets and are devoid of any Fe-K line emission, but do exhibit some continuum flux. This can be either due to depletion of iron in those regions, but can also be explained by the temperature exceeding 10 keV or by a non-thermal continuum. Note that the bright O⁶⁺ spot (Fig. 2) seems to lie close to these hot regions, in which case O⁶⁺ would be definitely ionising.

Figure 4: N132D 150 $^{\prime \prime }$$\times$150 $^{\prime \prime }$ maps of a) the Fe-K line flux at 6.67 keV, (range 0-1.66 10$\sp {-5}$ photons cm$\sp {-2}$ s$\sp {-1}$ arcmin$\sp {-2}$), b) the EW of the Fe-K line, (0-4.7 keV), c) the differential continuum flux at 6.67 keV derived from the 4.4-10 keV spectrum, (0-1.15 10$\sp {-5}$ photons cm$\sp {-2}$ s$\sp {-1}$ keV$\sp {-1}$ arcmin$\sp {-2}$), d) the apparent temperature (0.4-3.1 keV) derived from the continuum only. All colour bars are linear

Open with DEXTER

In order to understand the absence of Fe²³⁺ in the RGS spectrum and the weakness of Fe²²⁺ and Fe²¹⁺, while emission from both lower and higher charge states is intense, the entire EPIC-PN spectrum above 2.5 keV has been fitted with a two-temperature model. An acceptable fit is obtained for $kT_{\rm e}$ = 0.89 and 6.2 keV. These results can explain the absence of Fe²³⁺, which attains its maximal fractional abundance at the intermediate temperature of $\sim$1.5 keV (Mazzotta et al. 1998). The absence of these intermediate temperatures most likely means that the hot and cold regions are spatially distinct.

Spectroscopic analysis of the $\mathsf{O^{6+}}$ lines

Seeking further evidence that the O-rich gas in the northeastern part of N132D is in the process of being ionised, we examine the RGS spectrum in the vicinity of the three strong emission lines of O⁶⁺ (He-like triplet). These lines are at 21.602, 21.804, and 22.097 Å and are referred to, respectively, as the resonant (r), intercombination (i), and forbidden (f) lines. The i line is very weak in the N132D spectrum, whereas the r and f lines are strong, as can be seen in Fig. 5. A hint for similar line structure is found for the N⁵⁺ triplet as well (Fig. 5). Fitting the O⁶⁺ spectrum with the most appropriate line-profile model available yields line intensity ratios of f:i:r = 0.62:0.12:1.0. However, since the model can not very well reproduce the observed line shapes, uncertainties of about $\pm$ 20% should be associated with these values. The quality of the data for N⁵⁺ is insufficient to allow for a similar measurement with the N⁵⁺ lines.

Decaux et al. (1997) pointed out that for He-like Fe, intense r and f lines together with a very weak i line is an indication for hot ionising conditions, because: First, at high temperature the r line is enhanced compared to the i and f lines. Second, inner-shell ionisation processes from the ground state 1s²2s of the Li-like ion to the 1s2s (J=1) level enhance the f line. In the present case of O and N, inner-shell ionisation of Li-like ions also produces He-like excited ions in the 1s2s configuration. This configuration has two levels; the J=1 level is the upper level of the f line, whereas the J=0 level cannot decay radiatively to the ground level and in the case of Fe cannot decay radiatively at all. In contrast with the Fe case, for O and N, the 1s2p_1/2 (J=1) level, which is the upper level of the i line, lies below the 1s2s (J=0) level. This allows for radiative decay that enhances the i line in parallel to the f-line enhancement.

	Figure 5: The RGS spectrum of N132D in the limited wavelength range corresponding to the He-like triplets of O6+ and N5+ ( $\lambda _{\rm expected}$ indicated by horizontal bars). Relative line intensities are given for O6+
Open with DEXTER

We use atomic-state models employing data from the HULLAC code (Bar-Shalom et al. 1998) to calculate the spectrum of O⁶⁺ in a wide range of temperatures, both in ionising plasma and in equilibrium. Indeed, it is found that both situations can produce the observed f:i:r ratios. Both hot (e.g., 0.6 keV) ionising gas with a rather low O⁶⁺/O⁵⁺ ratio (e.g., 2), as well as a cooler (0.15 keV) gas in equilibrium can emit these lines with the observed ratios. In other words, the intensities of the O⁶⁺ lines alone are insufficient for distinguishing between the equilibrium and ionising scenarios. For completeness, it should be noted that the current spectrum at least rules out a cooling plasma for N132D, because in that case, the upper levels of the i and f lines are populated by recombination resulting in much higher f/r and i/r ratios than are observed here.

5 Conclusions

The RGS spectrum of N132D features lines of C, N, O, Ne, Mg, Si, S, and Fe. K-shell Ar, Ca, and Fe are detected in the EPIC-PN spectrum. The narrow-band EPIC images of N132D give a coarse mapping of the elemental and temperature structure in the remnant. With the exception of O, the dominant part of the soft X-ray emission originates from shocked ISM along the southeastern, and northwestern edges of the expanding shell. In contrast, strong Fe-K emission is detected near the centre of the remnant, perhaps indicating that high-Z ejecta is too highly ionised to be observed in shorter wavelengths. O⁷⁺ is present throughout the remnant, while the brightest spot of O⁶⁺ emission is near the northeastern rim. The O⁶⁺ emission can be attributed, either to low temperatures, or to hot, recently shocked material that is in the process of being ionised. Fitting of the EPIC-PN spectrum indicates high temperatures in that region. Unfortunately, the emission lines of O⁶⁺ alone, although clearly seen in the RGS spectrum, cannot provide a conclusive distinction between these two scenarios, due to considerations associated with the atomic processes involved in producing these lines. A more detailed analysis of the RGS spectrum, and especially of the many strong Fe lines, will potentially provide a better insight into the temperature and non-equilibrium structure of the hot gas in N132D.

References

 
• Banas, K. R., Hughes, J. P., Bronfman, L., & Nyman, L.-Å. 1995, ApJ, 480, 607 In the text  
• Bar-Shalom, A., Klapisch, M., Goldstein, W. H., & Oreg, J. 1998, The HULLAC code for atomic physics, unpublished In the text  
• Blair, W. P., Morse, J. A., Raymond, J. C., et al. 2000, ApJ, 537, 667 In the text NASA ADS  
• Brown, G. V., Beiersdorfer, P., Liedahl, D. A., Widmann, K., & Kahn, S. M. 1998, ApJ, 502, 1015 In the text NASA ADS  
• Brown, G. V., Beiersdorfer, P., Liedahl, D. A., Widmann, K., & Kahn, S. M. 2000, LLNL preprint [UCRL-JC-136647] In the text  
• Danziger, I. J., & Dennefeld, M. 1976, ApJ, 207, 394 In the text NASA ADS  
• Decaux, V., Beiersdorfer, P., Kahn, S. M., & Jacobs, V. L. 1997, ApJ, 482, 1076 NASA ADS  
• den Herder, J. W., Brinkman, A. C, Kahn, S. M., et al. 2001, A&A, 365, L7 In the text  
• Drake, G. W. 1988, Can. J. Phys., 66, 586 In the text NASA ADS  
• Dickel, J. R., & Milne, D. K. 1995, AJ, 109, 200 In the text NASA ADS  
• Favata, F., Vink, J., Parmar, A. N., Kaastra, J., & Mineo, T. 1998, A&A, 340, 626 In the text NASA ADS  
• Hughes, J. P., Hayashi, I., & Koyama, K. 1998, ApJ, 505, 732 In the text NASA ADS  
• Hwang, U., Hughes, J. P., Canizares, C. R., & Market, T. H. 1993, ApJ, 414, 219 In the text NASA ADS  
• Jansen, F., Lumb, D., Altieri, B., et al. 2001, A&A, 365, L1 In the text  
• Johnson, W. R., & Soff, G. 1985, Atom. Data Nucl. Data Tables, 33, 405 In the text NASA ADS  
• Mathewson, D. S., Ford, V. L., Dopita, M. A., et al. 1983, ApJS, 51, 345 In the text NASA ADS  
• Mazzotta, P., Mazzitelli, G., Colafrancesco, S., & Vittorio, N. 1998, A&AS, 133, 403 In the text NASA ADS  
• Morse, J. A., Winkler, P., & Kirshner, R. P. 1995, AJ, 109, 2104 In the text NASA ADS  
• Rasmussen, A., Cottam, J., Decker, T., et al. 1998, SPIE 3444, 327 In the text  
• Rasmussen, A. P., Behar, E., Kahn, S. M., van der Heyden, K., & den Herder, J. W. 2001, A&A, 365, L231 In the text  
• Strüder, L., Briel, U. G., Dennerl, K., et al. 2001, A&A, 365, L18 In the text  
• Turner, M. J. L., Abbey, A., Arnaud, M., et al. 2001, A&A, 365, L27 In the text  
• Westerlund, B. E., & Mathewson, D. S. 1966, MNRAS, 131, 371 In the text NASA ADS  
• Williams, R. M., Chu, Y. H., Dickel, J. R., et al. 1999, ApJS, 123, 467 In the text NASA ADS