1: %\documentclass[preprint]{aastex}
2: \documentclass[apj,twocolumn]{emulateapj}
3: \usepackage{natbib}
4:
5: \citestyle{aa}
6:
7: \newcommand{\msun}{M$_\odot$}
8: \newcommand{\kms}{km s$^{-1}$}
9: \newcommand{\masyr}{mas yr$^{-1}$}
10:
11: \slugcomment{Submitted to ApJ.}
12:
13: \shorttitle{IRS Spectral Mapping of SNR E\,0102}
14: \shortauthors{Sandstrom et al.}
15:
16: \begin{document}
17:
18: %%%%%%%%%%%%%%
19: % TITLE PAGE %
20: %%%%%%%%%%%%%%
21:
22: \title{Measuring Dust Production in the Small Magellanic Cloud
23: Core-Collapse Supernova Remnant 1E\,0102.2$-$7219}
24:
25: \author{Karin M. Sandstrom\altaffilmark{1},
26: Alberto D. Bolatto\altaffilmark{2},
27: Sne\v{z}ana Stanimirovi\'c\altaffilmark{3},
28: Jacco van Loon\altaffilmark{4}
29: and J. D. T. Smith\altaffilmark{5}}
30:
31: \affil{$^1$Astronomy Department, 601 Campbell Hall, University of
32: California, Berkeley, CA 94720, USA}
33: \affil{$^2$Department of Astronomy and Laboratory for Millimeter-wave
34: Astronomy, University of Maryland, College Park, MD 20742, USA}
35: \affil{$^3$Department of Astronomy, University of Wisconsin,
36: Madison, WI 53706, USA}
37: \affil{$^4$Astrophysics Group, Lennard-Jones Laboratories,
38: Keele University, Staffordshire ST5 5BG, UK}
39: \affil{$^5$Ritter Astrophysical Research Center, University of
40: Toledo, OH 43603, USA}
41:
42: \email{karin@astro.berkeley.edu}
43:
44: %%%%%%%%%%%%
45: % ABSTRACT %
46: %%%%%%%%%%%%
47:
48: \begin{abstract}
49:
50: We present mid-infrared spectral mapping observations of the
51: core-collapse supernova remnant 1E\,0102.2$-$7219 in the Small
52: Magellanic Cloud using the InfraRed Spectrograph (IRS) on the Spitzer
53: Space Telescope. The remnant shows emission from fine structure
54: transitions of neon and oxygen as well as continuum emission from
55: dust. Comparison of the mid-IR dust emission with observations at
56: x-ray, radio and optical wavelengths shows that the dust is associated
57: with the supernova ejecta and is thus newly formed in the remnant.
58: The spectrum of the newly formed dust is well reproduced by a model
59: that includes $3\times 10^{-3}$ \msun\ of amorphous carbon dust at 70
60: K and $2\times 10^{-5}$ \msun\ of Mg$_2$SiO$_4$ (forsterite) at 145 K.
61: Our observations place a lower limit on the amount of dust in the
62: remnant since we are not sensitive to the cold dust in the unshocked
63: ejecta. We compare our results to observations of other core-collapse
64: supernovae and remnants, particularly Cas A where very similar
65: spectral mapping observations have been carried out. We observe a
66: factor of $\sim 10$ less dust in E\,0102 than seen in Cas A, although
67: the amounts of amorphous carbon and forsterite are comparable.
68:
69: \end{abstract}
70:
71: \keywords{dust, extinction --- infrared: ISM --- supernova remnants}
72:
73: %%%%%%%%
74: % BODY %
75: %%%%%%%%
76:
77: \section{Introduction}
78: \label{intro}
79:
80: Dust is a crucial component of the interstellar medium (ISM). It
81: provides a site for interstellar chemistry, in particular the
82: formation of molecular hydrogen; it regulates thermal balance in
83: various phases of the ISM; and it provides shielding for dense clouds,
84: critically influencing the process of star-formation. The
85: characteristics of interstellar dust depend on the balance between its
86: formation and destruction and on the processing that dust grains
87: endure in the ISM. The mechanisms for dust production, their
88: timescales and efficiencies and the effect of metallicity on these
89: quantities determine the cosmic history of dust production and are
90: mostly not well characterized.
91:
92: In particular, the relative amount of dust that core-collapse
93: supernovae (CCSN) contribute to the ISM is poorly constrained and a
94: matter of much debate. Because of the short lifetimes of their
95: massive progenitor stars (a few tens of Myrs), CCSN are often cited as
96: an important source of dust early in the history of the Universe
97: \citep{morgan03,dwek07,todini01}. Observational evidence for
98: substantial dust production in supernovae, however, is scant and
99: controversial.
100:
101: Our understanding of dust formation in the early Universe and the role
102: of CCSN has recently been challenged by observations of large dust
103: masses around high redshift quasars \citep{bertoldi03,beelen06}.
104: Because the redshifts of these quasars are so high and the dust masses
105: are so large, the dust had to be produced efficiently and on a short
106: timescale. For the $z=6.42$ quasar SDSS J1148$+$5251, the age of the
107: Universe is only $\sim 900$ Myr and more than $10^8$ \msun\ of dust
108: are observed in its vicinity (although there are substantial
109: uncertainties associated with this dust mass, see the discussion in
110: \citet{omont01}). There is some indication that dust in some of these
111: high redshift galaxies may follow an extinction curve unlike that in
112: either the Milky Way or Small Magellanic Cloud \citep{maiolino04}.
113: \citet{hirashita05} claim that the extinction curve of dust in the
114: $z=6.19$ quasar SDSS J1048$+$4637, for example, can be reproduced by
115: the species and size distribution of dust formed in models of the
116: unmixed ejecta of $\sim 10-30$ \msun\ supernovae.
117:
118: The models of dust production in supernovae typically predict on the
119: order of $0.1-1$ \msun\ of dust formed in an average CCSN \citep[see
120: for instance,][]{todini01,nozawa03}. \citet{dwek07} estimate that to
121: produce the mass of dust observed around SDSS J1148$+$5251 requires an
122: average yield of $\sim 1$ \msun\ of dust per CCSN. As of yet, however,
123: observations of newly-formed dust in supernovae and their remnants
124: have shown orders of magnitude less dust than would be expected based
125: on these models. If we cannot substantiate the predicted dust
126: production efficiency of CCSN, our explanation of the high-z quasar
127: observations may need to rely more heavily on the contributions of red
128: supergiants and massive AGB stars to dust production
129: \citep{vanloon05,vanloon08}.
130:
131: Part of the discrepancy between expectations and observations of dust
132: production in CCSN and their remnants stems from the difficulty of
133: observing newly-formed dust. The condensation of dust is thought to
134: happen between $\sim 400-800$ days after the explosion
135: \citep{kozasa89,todini01,nozawa03}. While the newly-formed dust is
136: still warm, its near- and mid-IR emission may be detectable.
137: Observations of the thermal emission newly-formed from dust in SN
138: 1987A show $\sim 10^{-4}$ \msun\ of dust produced by 775 days after
139: the explosion \citep{roche93,wooden93,moseley89}. \citet{sugerman06}
140: claimed to detect mid-IR emission from $\sim 0.02$ \msun\ of
141: newly-formed dust in SN 2003gd, although this claim has been disputed
142: by \citet{meikle07} after their reanalysis of the Spitzer observations
143: that showed only $4\times 10^{-5}$ \msun\ of dust. At these early
144: times, near- and mid-IR emission can also arise from IR light echoes,
145: as the supernova's radiation heats pre-existing circumstellar dust
146: \citep{bode80,dwek83,pozzo04}. If there is a contribution by an IR
147: light echo, the mass of newly-formed dust inferred from the early-time
148: infrared emission may be greatly overestimated. On the other hand,
149: clumpiness of the ejecta could lead to an underestimate of the dust
150: mass based on the infrared emission \citep{ercolano07}.
151:
152: As the supernova ejecta expand and cool, the newly-formed dust
153: becomes detectable only at far-IR and/or submillimeter wavelengths at
154: which time confusion with cold dust in nearby star-forming regions
155: becomes a major issue. \citet{dunne03} claimed a detection of $\sim
156: 2$ \msun\ of newly-formed dust in Cas A with 850 \micron\ SCUBA
157: observations. \citet{krause04}, however, used far-IR and molecular
158: line observations to argue that there is at least an order of
159: magnitude less newly-formed dust in Cas A and that most of the
160: submillimeter emission arises in an intervening molecular cloud.
161: Similarly, submillimeter observations of Kepler's remnant were
162: used by \citet{morgan03} to argue for $0.1-3$ \msun\ of newly-formed
163: dust although these results have been recently disputed by
164: \citet{blair07} who find evidence for only $5\times 10^{-4}$ \msun\
165: using far-IR observations from Spitzer.
166:
167: Eventually, the interaction of the ejecta with the surrounding
168: circumstellar and/or interstellar medium (CSM/ISM) produces a reverse
169: shock which propagates back into the ejecta. The reverse shock
170: reheats the newly-formed dust so that once again it is detectable in
171: the mid-IR and may destroy some of the dust via sputtering, as well.
172: Observations at 24 \micron\ of 1E\,0102.2$-$7219, the same SNR we
173: discuss here, by \citet{stanimirovic05} showed evidence for $\sim
174: 8\times 10^{-4}$ \msun\ of dust. Recently, \citet{rho08} used IRS on
175: Spitzer to make a spectral map of Cas A. They claim evidence for
176: $\sim 0.02$ \msun\ of newly-formed dust in the remnant. Since only
177: the reverse-shocked dust should be visible in the mid-IR, only a
178: fraction of the newly-formed dust can be detected in this way.
179: Additionally, circumstellar or interstellar dust that is interacting
180: with the forward shock should also be detected in the mid-IR.
181: \citet{williams06} observed four supernova remnants in the Magellanic
182: Clouds with the IRAC and MIPS instruments on Spitzer and concluded
183: that the mid-infrared emission was from shocked circumstellar dust
184: rather than newly-formed dust. Despite these limitations mid-IR
185: observations are very useful for studying the dust production in
186: supernovae because they typically have higher angular resolution,
187: which allows the remnant to be more easily separated from foreground
188: and background emission. Mid-IR spectroscopy also has the potential
189: to provide some information on the composition of the newly-formed
190: dust in a remnant from the distinctive mid-IR spectral features of
191: some dust species. In this paper we present mid-IR spectral mapping
192: observations of the Small Magellanic Cloud remnant 1E\,0102.2$-$7219,
193: which we use to learn about the mass and composition of dust produced
194: in CCSN.
195:
196: SNR 1E\,0102.2$-$7219 (E\,0102) is a young, oxygen-rich remnant
197: located on the outskirts of the emission-line nebula N\,76. It is
198: thought to be the product of a Type Ib/Ic or IIL/b supernova
199: \citep{blair00,chevalier05} due to the lack of hydrogen and helium in
200: the remnant's spectrum and the high abundance of oxygen and neon in
201: its ejecta. E 0102 has been observed extensively at x-ray, UV,
202: optical, infrared and radio wavelengths. Chandra x-ray imaging of the
203: remnant by \citet{gaetz00} showed a textbook remnant structure with a
204: blast wave radius of $\sim$ 22'' and a reverse shock radius of $\sim$
205: 15''. \citet{amy93} observed the remnant at 6 cm using the Australia
206: Telescope Compact Array (ATCA) and found a shell of radio emission
207: with a radius of $\sim$ 20''. This ring lies between the measured
208: radii of the faint rim of the blast wave and the bright ring of the
209: reverse shock, and is attributed to material in the forward shock.
210: The full-width velocity dispersion of the optical [O III] emission
211: from the remnant was measured to be $\sim 6000$ km/s by
212: \citet{tuohy83}, implying an age of $\sim 1000$ years given the
213: remnant's size. \citet{hughes00} use three epochs of x-ray
214: observations from the \textit{Einstein}, \textit{ROSAT} and
215: \textit{Chandra} observatories to measure an expansion rate for the
216: whole remnant in agreement with the results of \citet{tuohy83}.
217: Recent work by \citet{finkelstein06} used two epochs of optical [O
218: III] observations to determine proper motions for ejecta filaments and
219: find an age of $2050\pm 600$ years for the remnant. The results of
220: the following work do not depend sensitively on the age, so we adopt a
221: value of 1000 years for simplicity.
222:
223: \citet{stanimirovic05} presented the first infrared detection of SNR E
224: 0102 from the Spitzer Survey of the Small Magellanic Cloud (S$^3$MC)
225: observations. S$^3$MC covered the star-forming regions of the SMC Bar
226: and Wing with imaging in all MIPS and IRAC bands \citep{bolatto07}.
227: The remnant was only clearly detected at 24 \micron.
228: \citet{stanimirovic05} used the 24 \micron\ detection along with the
229: upper limits in the other bands to argue that the dust in the remnant
230: has a temperature of $\sim 120$ K and they determined an upper limit
231: of $8\times 10^{-4}$ \msun\ of warm dust. The main uncertainties in
232: their analysis were the fraction of the MIPS 24 \micron\ emission
233: coming from the [O IV] line at 24.8 \micron\ and the temperature of
234: the dust.
235:
236: In the following, we present Spitzer IRS spectral mapping observations
237: of SNR 1E\,0102.2$-$7219 obtained as part of the Spitzer Spectroscopic
238: Survey of the Small Magellanic Cloud (S$^4$MC). These observations
239: allow us to map the emission lines and dust continuum from the remnant
240: and place a lower limit on the amount of newly formed dust in E 0102.
241: In Section~\ref{sec:data} we describe the observations and their
242: reduction. In Section~\ref{sec:spec} we present the spectrum of the
243: remnant and discuss the spatial distribution of its various
244: components. In Section~\ref{sec:dustmodel} we use a model for the
245: reverse-shocked dust to measure a mass of $3\times 10^{-3}$ \msun\
246: and we discuss the composition of the dust as revealed by its
247: spectrum. Section~\ref{sec:discussion} discusses the implications of
248: our measurement for the understanding of CCSN dust production and
249: compares our results with previous work.
250:
251: \section{Observations and Data Reduction}
252:
253: \subsection{Spitzer Observations}\label{sec:data}
254:
255: SNR E 0102 was mapped with all LL and SL orders of IRS on the Spitzer
256: Space Telescope as part of the S$^4$MC project (Spitzer Spectroscopic
257: Survey of the Small Magellanic Cloud, GO 30491). The
258: wavelength coverage of the SL and LL orders extends from 5.2 to 38.0
259: \micron\ with a spectral resolution between 60 and 120. Ramp times of
260: 14 and 30 seconds for SL and LL were chosen as a compromise between
261: sensitivity, spatial coverage and observation length.
262:
263: SNR E 0102 was covered in the spectral maps of the N\,76 region,
264: observed on 9 and 12 December 2006 for SL and LL, respectively. The LL
265: map consists of $6\times 75$ pointings covering an area of $474\times
266: 381$ arcseconds$^2$ in both LL1 and LL2. The SL observations
267: consisted of $5\times 120$ pointings covering $260\times 222$
268: arcseconds$^2$ for the SL1 and SL2 orders. The roll angle was left
269: unconstrained for our observations. The SL and LL maps have different
270: orientations because of the slit positions in the focal plane of the
271: telescope. Figure~\ref{fig:n76} shows a three-color image of the
272: region from the S$^3$MC dataset overlayed with the coverage of the
273: SL2, SL1, LL2 and LL1 maps. The spectral maps are completely sampled
274: in both LL and SL by stepping perpendicular to the slit by one-half
275: slit width (5.08 and 1.85'' for LL and SL, respectively). In the SL
276: orders we have stepped by a full slit width parallel to the slit
277: (52'') in order to increase the spatial coverage of the maps at the
278: expense of some redundancy. In the LL orders we have used half-slit
279: width steps parallel to the slit (79''). More redundancy in the LL
280: maps is necessary due to the increasingly large numbers of bad pixels
281: at long wavelengths.
282:
283: \begin{figure*}[t]
284: \centering
285: \epsscale{1.0}
286: \includegraphics[width=5in]{fig1.eps}
287: \caption{Three-color image of the N 76 region
288: overlayed with the coverage of the IRS spectral mapping observations.
289: The S$^3$MC MIPS and IRAC maps at 24, 8.0 and 3.6 \micron\ are shown
290: in red, green and blue. The white solid and dashed lines show the
291: coverage of the LL1 and LL2 spectral maps. The cyan solid and dashed
292: lines show the coverage of the SL1 and SL2 maps. The 1E\,0102.2$-$7219
293: supernova remnant is located at R.A. $01^{\rm h}04^{\rm m}02\fs1$,
294: Dec. $-72^\circ01\arcmin52\farcs5$ (J2000).}
295: \label{fig:n76}
296: \end{figure*}
297:
298:
299: \subsection{Data Reduction}\label{sec:reduction}
300:
301: The spectral maps were calibrated and assembled in IDL using the
302: Cubism package \citep{smith07}. Cubism takes as input
303: pipeline-processed two-dimensional spectral images from IRS and
304: produces three dimensional spectral data cubes using a reprojection
305: algorithm which employs two steps of polygon clipping (in which the
306: exact geometrical overlap of input pixels, or pseudo-rectangles in the
307: case of the IRS spectral images, is used to calculate its weight in an
308: output pixel grid). Cubism also applies an aperture loss correction
309: function (ALCF) and a slit loss correction function (SLCF) to adjust
310: the calibration which is based on point-source observations rather
311: than extended sources. The output pixel scales of the LL and SL cubes
312: are 5.08'' and 1.85'', respectively. Images of the IRS data in this
313: work are shown on the IRS pixel scale rather than aligned to R.A. and
314: Dec to avoid unnecessary interpolation.
315:
316: The Galactic and zodiacal light foregrounds were removed from the IRS
317: spectra using a dedicated ``off'' position. The same position was
318: used for both the LL and SL modules. The background observations were
319: done immediately following the mapping observations. The ``off''
320: position was chosen as a place in the S$^3$MC 24 \micron\ MIPS mosaic
321: with no extended emission or point sources, located at R.A. $1^{\rm
322: h}09^{\rm m}40\fs00$, Dec. $-73^\circ31\arcmin30\farcs00$.
323:
324: The subtraction of the ``off'' position, in addition to removing the
325: Galactic and zodiacal foregrounds, mitigates the effects of rogue
326: pixels (i.e. time variable bad pixels) in the IRS detectors, although
327: it does not eliminate them. Further bad pixel removal was done in
328: Cubism by examining each wavelength for striping (caused by global bad
329: pixels) and other artifacts. The effect of bad pixels increases at
330: the longer wavelengths, degrading the data quality beyond about 35
331: \micron.
332:
333: \subsection{Resolution Matching and Alignment of
334: Cubes}\label{sec:conv}
335:
336: In order to directly compare the spatial distribution of supernova
337: emission at each wavelength as well as properly extract a spectrum
338: when the extraction region is on the order of a resolution element, it
339: was necessary to convolve all the individual slices of the spectral
340: cube to the same resolution. This was achieved by creating
341: convolution kernels based on theoretical point-spread functions (PSFs)
342: for the IRS spectrograph generated by the program sTinyTim
343: \footnote{\url{http://ssc.spitzer.caltech.edu/archanaly/contributed/stinytim/index.html}}.
344:
345: The convolution kernels $k$ were defined in the following way:
346: \begin{equation}
347: k = \mathrm{FFT}^{-1}(R/I) ,
348: \end{equation}
349: where $R$ is the 2-dimensional Fast Fourier Transform (FFT) of the PSF
350: of the longest wavelength slice of the LL1 cube and $I$ is the FFT of
351: the PSF of the slice in question. The convolution kernels were masked
352: in Fourier space such that there was no power at spatial frequencies
353: higher than the Nyquist frequency of the lowest resolution kernel in
354: order to avoid the amplification of high-frequency noise. Each slice
355: of the spectral data cube was then convolved with the appropriate
356: kernel. Although the PSF of IRS is not well explored (i.e. the
357: predictions from sTinyTim are mostly untested) and the PSF that Cubism
358: recovers can be somewhat elongated in a way that varies with
359: wavelength, we found that the results of this technique were adequate
360: for our purposes. The convolution process was tested on the cube of
361: the N 22 region of the SMC which contains a very bright point source
362: for which the first two diffraction rings are visible for essentially
363: every slice in the cube from SL2 through LL1. The correspondence
364: between wavelength slices after our convolution is excellent as shown
365: in Figure~\ref{fig:n22}. For each cube, the associated uncertainty
366: cube was also convolved and the errors appropriately propagated.
367:
368: \begin{figure*}
369: \centering
370: \epsscale{1.0}
371: \includegraphics[width=5in]{fig2.eps}
372: \caption{Results of the convolution described in
373: Section~\ref{sec:conv} for a bright point source in the N\,22 region.
374: The panels show the point source at (1) 23.0 \micron, (2) 30.0 \micron,
375: and (3) 36.7 \micron. The bottom panels (1b), (2b) and (3b), show the
376: point source before the convolution. The top panels, labeled (1a),
377: (2a) and (3a) show the source after convolution. The image stretch is
378: logarithmic to highlight the faint diffraction rings around the point
379: source. After the convolution, the diffraction rings are aligned at
380: all wavelengths, showing the effectiveness of our sTinyTim derived
381: convolution kernels.}
382: \label{fig:n22}
383: \end{figure*}
384:
385: The LL1, LL2, SL1 and SL2 cubes were aligned using a technique based
386: on the polygon-clipping algorithm implemented in Cubism. In
387: polygon-clipping the weight assigned to each pixel in the original
388: image is calculated based on the geometric overlap of that pixel with
389: the grid of output pixels. Aligning the cubes using polygon clipping
390: minimizes interpolation error and allows for direct propagation of
391: the associated uncertainty cubes. The LL2, SL1 and SL2 cubes were
392: aligned to match the astrometry and pixel scale of the LL1 map after
393: resolution matching.
394:
395: \subsection{Local Background Removal}
396:
397: More difficult to remove than the Galactic foreground is the variable
398: and complex local foreground and background emission (for simplicity
399: we will refer to this as a background, though there are contributions
400: from the entire line of sight through the SMC). E 0102 is located
401: near the large emission-line complex N\,76. Emission from the
402: outskirts of this region must be removed from the supernova remnant
403: spectrum to accurately measure the dust content and emission line
404: flux.
405:
406: An accurate background subtraction is not possible over a large area
407: due to the proximity of N\,76 and the complexity of the emission in the
408: region. For this reason we focus on the area near the supernova
409: remnant and attempt to subtract a model for the local emission. We
410: first attempted to fit planar or higher order surfaces to the emission
411: in the region of the remnant. The steep profile on the side of N\,76
412: nearest the remnant was poorly represented in these fits, resulting in
413: large residuals. Polynomial fits to each row and column, as done by
414: \citet{stanimirovic05}, were a better representation of the local
415: emission but also suffered from issues with reproducing the steepness
416: of the N\,76 shoulder, interfering with the goal of determining the
417: background in the immediate vicinity of the remnant. Rather than
418: trying to reproduce the local emission we argue that a better
419: representation of the background comes from interpolating across the
420: supernova remnant, which eliminates the need to reproduce the complex
421: emission structures with polynomials or surfaces.
422:
423: For each slice of the convolved and aligned cubes, we mask out the
424: immediate region of the remnant, using a mask 60'' in diameter to
425: exclude any emission from the remnant in our background determination.
426: We then linearly interpolate in the $x$ and $y$ directions across the
427: masked region, using two pixels on either side of the remnant in the
428: given row or column. Finally, we average together the results
429: obtained by interpolating in the $x$ and $y$ direction. This
430: technique has advantages and disadvantages. Since we only use the
431: four pixels nearest the remnant, there is some random noise in the
432: background value. We believe this disadvantage is far
433: outweighed by the difficulty in doing any kind of more complex fit
434: given the spatial intensity variations of the background. In
435: addition, the column based interpolation also tends to remove any
436: residual striping in the LL cubes that results from the effects of
437: rogue detector pixels.
438:
439: A plot of the background spectrum is shown in the bottom panel of
440: Figure~\ref{fig:snr}. Subtracting the background eliminates the
441: contributions of [S III], [Si II] and PAHs entirely from the SNR
442: spectrum showing that emission from these species is not related to
443: the remnant. All of these emission features are typical for the ISM
444: irradiated by the average interstellar radiation field (ISRF). The
445: local background spectrum is slightly negative at short wavelengths.
446: This is most likely due to the gradient in the zodiacal and Galactic
447: foregrounds between the location of the dedicated ``off'' position and
448: the remnant. Subtracting the local background, however, corrects for
449: this over-subtraction in the spectrum of the remnant. In the
450: continuum, background emission accounts for 40 to 60\% of the
451: emission. For the emission lines, the background subtraction removes
452: a few percent of the integrated strength of [Ne III] and [O IV].
453:
454: \begin{figure*}
455: \centering
456: \epsscale{1.0}
457: \includegraphics[width=5in]{fig3.eps}
458: \caption{This figure shows (a) the spectrum of SNR E 0102
459: after the subtraction of the local background and (b) the local
460: background spectrum. The spectrum of the supernova remnant shows dust
461: continuum and emission lines from neon and oxygen. The background
462: spectrum shows lines from sulfur, silicon, neon and oxygen as well as
463: emission from polycyclic aromatic hydrocarbons around 8 \micron\ and
464: dust continuum. The contributions from silicon, sulfur and PAHs are
465: entirely removed from the remnant's spectrum after background
466: subtraction. The background spectrum is slightly negative at short
467: wavelengths due to oversubtraction of the Galactic and zodiacal light
468: foregrounds. The background subtraction corrects this oversubtraction
469: in the spectrum of the remnant.}
470: \label{fig:snr}
471: \end{figure*}
472:
473:
474: In general, the local emission in each wavelength slice has a very
475: similar spatial structure, except at the wavelengths of strong
476: emission lines. In the strong emission line from [O IV] and/or [Fe
477: II] at 25.8 \micron, a diffuse halo is evident around the supernova
478: remnant. This ionized gas cannot be from material that is
479: collisionally interacting with the remnant since it is well outside
480: the extent of the forward shock. A similar structure is seen at the
481: wavelengths of the other spectral lines. The ionized material
482: surrounding the remnant has been observed previously in optical
483: studies \citep{tuohy83,blair00,finkelstein06}. The source of the
484: ionization for this surrounding material has been discussed by
485: \citet{tuohy83}. It is possible that the material was ionized by the
486: supernova shock breakout or is currently being ionized by emission
487: from the radiative shocks in dense knots of ejecta in E 0102. The
488: surroundings of the remnant will be discussed further in
489: Section~\ref{sec:csm}.
490:
491: \section{Analysis}
492:
493: \subsection{The Mid-IR Spectrum}\label{sec:spec}
494:
495: \subsubsection{Sources of Dust Continuum Emission}\label{sec:temps}
496:
497: We now consider the possible sources for dust continuum emission from
498: SNR E 0102: circumstellar/interstellar dust in the forward shocked
499: region, newly formed dust in the reverse shocked ejecta or newly
500: formed dust in the unshocked ejecta. The contribution of each of
501: these regions to the spectrum depends on the mass, temperature and
502: composition of the dust present.
503:
504: The dominant heating process for dust in the shocked ejecta and
505: shocked CSM/ISM will be collisions with electrons
506: \citep{dwek81,arendt99}. The equilibrium temperature of a dust grain
507: that is collisionally heated in a plasma can be found by equating the
508: collisional heating rate of the grain with its radiative cooling rate.
509: The collisional heating rate for a dust grain with size $a$ by
510: particles with mass $m$, density $n$ and temperature $T$ is given by
511: \citet{dwek87}:
512: \begin{equation}
513: H(a,T,n) = (32/\pi m)^{1/2} \pi a^2 n(kT)^{3/2} h(a,T) .
514: \end{equation}
515: Here $h(a,T)$ is a dimensionless parameter which describes the grain
516: ``heating efficiency'' and depends on the fractional amount of the
517: incident energy of the particle that is deposited in the grain by the
518: collision. \citet{dwek87} develop an approximation for $h(a,T)$ based
519: on a comparison of the effective thickness of the grain $4a/3$ and the
520: experimentally measured stopping distance of an electron or ion in a
521: solid. We use their formulae in this analysis. The radiative cooling
522: rate of the grain is given by:
523: \begin{equation}
524: L_{gr}(a, T_d) = 4\pi a^2 \sigma T_{d}^{4} <Q(a,T_d)> ,
525: \end{equation}
526: where $<Q(a,T_d)>$ is the Planck-averaged value of the absorption
527: coefficient for a dust grain of the given size and temperature.
528: Equating these heating and cooling rates yields the equilibrium
529: temperature of the grain.
530:
531: In the remnant, shocked CSM/ISM dust would be located in a shell with
532: an outer radius of 22'' and an inner radius at the contact
533: discontinuity, the exact position of which has not been determined.
534: X-ray spectroscopy of the blast wave region shows post-shock material
535: with abundances consistent with the SMC, electron temperatures around
536: $1$ keV, and ionization timescales $\tau = n_e t \sim 3\times 10^{10}$
537: s cm$^{-3}$ \citep{hayashi94,hughes00,sasaki06}. Assuming a timescale
538: of $\sim 1000$ years (the age of the remnant) gives electron densities
539: in the forward shock region of $n_e \sim 1$ cm$^{-3}$ . Collisionally
540: heated dust grains in this region would have equilibrium temperatures
541: of between $30$ and $70$ K for silicate and carbonaceous grains with
542: sizes between $0.001$ and $1.0$ \micron\ \citep[assuming the optical
543: constants for interstellar graphite and silicate grains
544: from][]{laor93}. Figure~\ref{fig:teq} shows the equilibrium
545: temperature of dust grains in the forward shocked plasma as a function
546: of grain size. The collision rates in this plasma, as well as in the
547: reverse shock, are low enough that the grains are not heated above
548: their sublimation temperature and destruction of the grains via
549: evaporation.
550:
551: \begin{figure}
552: \centering
553: \epsscale{1.0}
554: \includegraphics[width=3in]{fig4.eps}
555: \caption{The equilibrium temperature of collisionally heated grains
556: of different compositions in the forward and reverse shocked x-ray
557: emitting plasma and the hot, post-shock region of a dense knot. We
558: plot only amorphous carbon and forsterite in the reverse shocked ejecta
559: based on the results of our dust model fits in Section~\ref{sec:fit}.
560: We assume that dust grains in the shocked CSM/ISM will have the
561: optical properties characteristic of interstellar graphite and
562: silicate from \citet{laor93}.}
563: \label{fig:teq}
564: \end{figure}
565:
566:
567: Shocked ejecta dust would be located in a shell with an outer radius
568: at the contact discontinuity and an inner radius at 15''. The heating
569: mechanism of the dust in the shocked ejecta depends on its density
570: structure. When the reverse shock encounters more diffuse parts of
571: the ejecta its passage creates hot x-ray emitting plasma. X-ray
572: spectroscopy of the shocked ejecta shows a plasma composed of oxygen,
573: neon, magnesium and silicon with electron temperatures around 0.4 keV
574: ($5\times 10^6$ K) and ionization timescales of $n_e t \sim 10^{12}$ s
575: cm$^{-3}$. Assuming $t\sim 1000$ years, as before, gives densities in
576: the x-ray emitting reverse shocked ejecta of $\sim 20$ cm$^{-3}$.
577: Figure~\ref{fig:teq} also shows the equilibrium temperature of
578: collisionally heated amorphous carbon and forsterite (Mg$_2$SiO$_4$)
579: grains as a function of size given these plasma parameters, using
580: optical constants from \citet{rouleau91} for amorphous carbon and
581: \citet{jaeger03} for forsterite and assuming the heating efficiencies
582: for carbonaceous and silicate dust from \citet{dwek87}.
583:
584: The interaction of the reverse shock with material in dense clumps of
585: ejecta drives radiative shocks that produce optical and infrared line
586: emission, discussed further in Section~\ref{sec:lines}, and heat the
587: dust grains by a combination of radiative and collisional processes.
588: \citet{arendt99} compared the collisional and radiative heating for
589: dust in the hot, post-shock region of a knot in Cas A assuming the
590: density, temperature and cooling function predicted by the models of
591: \citet{sutherland95} for shock velocities between $150-200$ km
592: s$^{-1}$---the same models which are found to best reproduce the
593: optical and UV emission from E 0102. They found that the collisional
594: heating exceeds radiative heating under these conditions by factors of
595: a few hundred. Thus, for dust in the immediate post-shock part of
596: dense knots, we can ignore radiative heating. Figure~\ref{fig:teq}
597: shows the equilibrium temperatures for collisionally heated dust
598: grains in this region, using $T_e = 10^{6.64}$ K, $n_e = 400$
599: cm$^{-3}$ and the gas cooling function $\Gamma = 10^{-17.5}$ ergs
600: cm$^{3}$ s$^{-1}$ from \citet{sutherland95}. On the other hand, in
601: the cooled post-shock gas of the dense knots and the in cold pre-shock
602: gas, the dust is likely heated by the radiation field generated by the
603: shock front. \citet{bouchet06} calculate the radiative heating for
604: dust in similar conditions in SN 1987A and find that it can reach
605: temperatures of $\sim 125$ K. Given the angular resolution of our
606: observations we will not be able to distinguish regions where dust in
607: knots is collisionally or radiatively heated, but we note that the
608: equilibrium dust temperatures in these circumstances are all between
609: $\sim$ $100-200$ K.
610:
611: If the dust grains in these various collisionally heated regions are
612: small enough, a collision with a single electron may be sufficient to
613: raise the grain temperature well above the equilibrium value for a
614: short time. After this temperature spike the grain can cool to below
615: the equilibrium temperature, resulting in a large range of
616: temperatures for a given grain size. An important consideration when
617: dealing with this stochastically heated dust is that only some
618: fraction of the grains will be warm at a given time, so our mid-IR
619: observations may miss a substantial fraction of the dust. Grains will
620: be stochastically heated in this manner when two conditions are met:
621: 1) the collisional rate is slow compared with the cooling rate of the
622: grain and 2) the amount of energy deposited by a single electron is
623: large compared with the internal energy of the grain.
624:
625: For the plasma conditions in the forward and reverse shocks,
626: Figure~\ref{fig:stoch} shows the ranges of grain size for which we
627: expect the dust to be in equilibrium with the plasma based on the
628: ratios between the energy deposited by one electron and the grain's
629: total internal energy ($\Delta E/U$) and the ratio between the cooling
630: time and collision time ($\tau_{cool}/\tau_{coll}$). For simplicity
631: we show only one grain species for the forward and reverse shocks:
632: graphite in the forward shock and amorphous carbon in the reverse
633: shock. The size at which equilibrium is reached is a factor of $\sim
634: 2$ smaller for silicate grains in the forward shock and a factor of
635: $\sim 1.5$ smaller for forsterite grains in the reverse shock. For
636: this calculation we have used the heat capacity of graphite grains
637: from \citet{dwek86} and silicate from \citet{draine85}, both of which
638: are based on fits to experimental data, and the same optical constants
639: used in Figure~\ref{fig:teq}. The graphite heat capacity is used for
640: amorphous carbon and the silicate heat capacity is used for
641: forsterite. For the estimates presented here, this level of accuracy
642: should be sufficient, though the heat capacity of a dust grain does
643: depend on its composition, shape and ``fluffiness'' among other
644: factors.
645:
646: \begin{figure}
647: \centering
648: \epsscale{1.0}
649: \includegraphics[width=3in]{fig5.eps}
650: \caption{This plot illustrates the range of dust grain sizes that will
651: be in equilibrium in the forward and reverse shocks. The dotted lines
652: are for amorphous carbon in the reverse shock and the dashed lines are
653: for graphite in the forward shock, assuming the densities and
654: temperatures discussed in the text. Two ratios are shown on the plot:
655: the ratio between the cooling time and the collision time
656: ($\tau_{cool}/\tau_{coll}$) and the ratio between the energy deposited
657: by a collision with one electron and the internal energy of the grain
658: ($\Delta E/U$). When the ratio of the collision time to the cooling
659: time is large (i.e. collisions happen faster than the dust grain can
660: cool) and the ratio of the energy deposited by one electron to the
661: total internal energy of the grain is small (i.e. the grain
662: temperature does not change appreciably from one collision) then the
663: grains can be said to be in equilibrium. The grain size for which
664: these conditions are met are $> 0.01$ \micron\ for amorphous carbon in
665: the reverse shock and $> 0.02$ \micron\ for graphite in the forward
666: shock. This size is a factor of $\sim 1.5$ smaller for forsterite in
667: the reverse shock and $\sim 2$ smaller for silicate dust in the
668: forward shock.}
669: \label{fig:stoch}
670: \end{figure}
671:
672:
673: The unshocked ejecta are located inside the reverse shock radius of
674: 15''. The density, temperature and structure of the unshocked ejecta
675: are unobserved. We can place limits on the temperature of the gas and
676: dust by considering the processes responsible for heating and cooling
677: the ejecta. In the initial phases of the remnant's expansion the
678: ejecta are heated by the decay of radioactive isotopes produced in the
679: explosion. As the remnant expands, the ejecta cool adiabatically,
680: reaching a temperature cool enough to allow dust condensation after
681: about 400 to 800 days post-explosion \citep{todini01,nozawa03}.
682: Assuming adiabatic cooling with an index $\gamma = 1.25$ as used by
683: \citet{kozasa89} and a temperature of $\sim 1000$ K 400 days after
684: the supernova, the current gas temperature would be less than the
685: temperature of the cosmic microwave background. Before reaching that
686: temperature, of course, the dust and gas would come into equilibrium
687: with the local radiation field which has contributions from both the
688: general SMC radiation field and the emission from E 0102's shock
689: fronts. The equilibrium temperature of big grains in the interstellar
690: radiation field of the SMC is $\sim 20$ K \citep{bot04,leroy07}, so we
691: consider this a lower limit on the temperature of the unshocked ejecta
692: dust.
693:
694: \subsubsection{Oxygen and Neon Fine-Structure Emission
695: Lines}\label{sec:lines}
696:
697: We observe emission from fine-structure transitions of [Ne II], [Ne III],
698: [Ne V] and [O IV] in the mid-IR spectrum of E 0102. At the spectral
699: resolution of the LL1 order, the [O IV] line overlaps with [Fe
700: II] at 25.988 \micron. However, given the predominance of oxygen in
701: the supernova remnant as observed in the optical, x-ray and UV and the
702: dearth of iron emission observed at any wavelength, we consider it
703: very likely that the 25.8 \micron\ emission line is due to [O IV].
704: Table~\ref{tab:lines} lists the detected emission lines and their
705: strengths. Examination of the spatially resolved velocity information
706: shows a range of velocities in agreement with previous spectroscopic
707: measurements of optical line emission.
708:
709: \begin{deluxetable*}{lccc}
710: \tablewidth{0pt}
711: \tabletypesize{\scriptsize}
712: \tablecolumns{4}
713: \tablecaption{Emission Lines in the Mid-IR Spectrum of E 0102\label{tab:lines}}
714: \tablehead{ \multicolumn{1}{l}{Line} &
715: \multicolumn{1}{l}{Measured Wavelength} &
716: \multicolumn{1}{l}{Central Intensity} &
717: \colhead{Integrated Strength} \\
718: \colhead{} &
719: \multicolumn{1}{c}{(\micron)} &
720: \multicolumn{1}{c}{(MJy sr$^{-1}$)} &
721: \multicolumn{1}{c}{($10^{-7}$ W m$^{-2}$ sr$^{-1}$)} }
722: \startdata
723: \hbox{[Ne II]} & 12.8053 $\pm$ 0.0005 & 6.287 $\pm$ 0.035 & 2.585 $\pm$ 0.015 \\
724: \hbox{[Ne III]} & 15.5514 $\pm$ 0.0010 & 2.295 $\pm$ 0.019 & 0.917 $\pm$ 0.022 \\
725: \hbox{[Ne V]} & 24.3460 $\pm$ 0.0053 & 0.566 $\pm$ 0.015 & 0.234 $\pm$ 0.017 \\
726: \hbox{[O IV]} & 25.9022 $\pm$ 0.0010 & 4.662 $\pm$ 0.019 & 2.132 $\pm$ 0.014 \\
727: \enddata
728: \tablecomments{These measured values are from Gaussian fits to the IRS spectra.}
729: \end{deluxetable*}
730:
731:
732: The contribution of [O IV] to the 24 \micron\ flux was one of the major
733: uncertainties for \citet{stanimirovic05} in their analysis of the mass
734: of newly formed dust in the remnant. We find that [O IV] is
735: responsible for $\sim$ 15\% of the 24 \micron\ flux from the remnant,
736: significantly less than the 60\% they were forced to assume due to the
737: lack of spectroscopic information.
738:
739: The line emission from oxygen-rich supernova remnants like E 0102 is
740: thought to arise from slower, radiative shocks created when the
741: reverse shock encounters dense clumps of ejecta. Due to the high
742: abundance of oxygen and other heavy elements in the nuclear-processed
743: stellar ejecta, the cooling time behind these shocks is shorter than
744: the recombination time, producing a plasma that is much cooler than
745: its ionization state would indicate. The relative ionization levels
746: in the post-shock ejecta depend upon the details of the shock front
747: and the ionization state of the material prior to encountering the
748: reverse shock. \citet{sutherland95} found that the optical emission
749: spectrum of E 0102 could be fit relatively well with a model for the
750: reverse shock entering a dense ejecta knot with velocities between 150
751: and 200 km s$^{-1}$ and creating a precursor photoionization front
752: that determines the initial ionization state of the pre-shock gas.
753: Using the observed mid-IR emission line strengths to determine
754: physical conditions in the ejecta would require shock modeling outside
755: the scope of this paper.
756:
757: \subsection{Spatial Decomposition of the Mid-IR
758: Emission}\label{sec:decomp}
759:
760: At the resolution and signal-to-noise of our observations it is
761: difficult to spatially separate the emission from
762: circumstellar/interstellar dust in the forward shock from the emission
763: from newly formed dust heated by the reverse shock. The outer radius
764: of the forward shock, as measured by \citet{gaetz00} is $\sim$ 22''
765: while the inner radius of the reverse shock is $\sim$ 15''. In
766: between, the two plasmas are separated by a contact discontinuity, the
767: location of which has not been measured. The resolution of the IRS
768: instrument in the LL1 order ranges between 6 and 9'' and the pixel
769: scale is $\sim$ 5''. The angular resolution inhibits our ability to
770: separate newly formed dust from dust heated by the forward shock,
771: without more detailed modeling. Measuring the amount of forward
772: shocked dust is also an interesting quantity on its own
773: (Section~\ref{sec:csm}).
774:
775: The close correspondence between the $0.3-10$ keV Chandra map and the
776: 24 \micron\ dust map, as noted by \citet{stanimirovic05}, suggests
777: that the dust emission is mainly from the reverse shocked material and
778: is thus newly formed in the remnant. In this section, we quantify
779: the fraction of the dust emission at each wavelength in our cube that
780: can be attributed to dust in the ejecta of the supernova by
781: decomposing the remnant spatially using x-ray, radio and optical
782: observations as templates for the sources of emission in the remnant.
783: As discussed in the previous section, there are three mechanisms for
784: which we might expect to detect mid-IR dust emission: CSM/ISM dust in
785: the forward shocked region, newly formed dust in the x-ray emitting
786: regions of reverse shocked gas and newly formed dust in the dense
787: knots of ejecta. This approach is similar to that used by
788: \citet{arendt92} in studying dust emission from the Cygnus Loop. We
789: will first discuss the template images we use to represent these three
790: emission mechanisms.
791:
792: Emission from the forward shocked material can be traced by radio
793: continuum emission. The radio continuum is synchrotron emission from
794: electrons interacting with the compressed CSM/ISM magnetic field in
795: the region between the blast wave and the contact discontinuity.
796: \citet{amy93} observed the remnant at 6 cm using the Australia
797: Telescope Compact Array. They see a shell structure which fills the
798: region between the faint outer boundary of the blast wave as traced by
799: x-ray emission and the outer boundary of the reverse shock. We
800: use the radio map as a tracer of the forward shocked circumstellar and
801: interstellar material.
802:
803: We have used the $0.3-10$ keV Chandra x-ray map of \citet{gaetz00} to
804: trace the x-ray emitting reverse shocked material. X-ray spectroscopy
805: of E 0102 has demonstrated that the bright x-ray ring contains
806: material composed primarily of oxygen and neon with smaller
807: contributions from magnesium and silicon, consistent with the
808: nucleosynthetic products of a massive star
809: \citep{flanagan04,sasaki06}. Finally, to trace the spatial
810: distribution of dust in dense knots of reverse shocked material we use
811: the optical [O III] 5007 \AA\ emission from HST ACS imaging by
812: \citet{finkelstein06} that arises from the aforementioned radiative
813: shocks.
814:
815: The templates were convolved to the resolution of the longest
816: wavelength point spread function of the IRS cube and then resampled to
817: the same pixel grid. For the [O III] template a few additional
818: reduction steps were necessary prior to the convolving and alignment.
819: The processed line (F475W) and continuum (F550M) images from HST were
820: provided to us by S. Finkelstein. The wide F475W filter was used to
821: ensure detection of the high velocity emission from the remnant.
822: Relative to the point sources in the image, the remnant emission is
823: quite faint. Thus, to properly construct a template it was necessary
824: to remove the point sources from the image by subtracting the
825: continuum image from the line image. This was done by first
826: cross-convolving the two images with theoretical ACS PSFs from TinyTim
827: \citep{krist95} for the given wavebands, scaling, and subtracting.
828: For bright point sources, the PSF subtraction often left large
829: residuals, which were removed with a combination of median filtering
830: and masking by hand for the brightest, saturated stars. The three
831: templates are shown in Figure~\ref{fig:templates} after convolution
832: and resampling to the LL pixel grid.
833:
834: \begin{figure*}
835: \centering
836: \epsscale{1.0}
837: \includegraphics[width=5in]{fig6.eps}
838: \caption{Convolved and aligned x-ray, optical and [O III] templates used in
839: decomposing the mid-IR emission. Data are from \citet{gaetz00} in the
840: x-ray, \cite{amy93} in the radio and \citet{finkelstein06} in the
841: optical.}
842: \label{fig:templates}
843: \end{figure*}
844:
845:
846: The decomposition at each wavelength was done with a least squares fit
847: that constrained the coefficient of each template image to be greater
848: than or equal to zero. The fit returns the best coefficients of the
849: linear combination of the three templates that match the observed
850: surface brightness distribution. Examples of the decomposition are
851: shown in Figures~\ref{fig:stackcont} and~\ref{fig:stackline} for two
852: images which show $20.5-22.3$ \micron\ dust continuum and the [O IV]
853: emission line at 25.9 \micron, respectively. To obtain the images
854: used for these plots we have binned the cube in wavelength to increase
855: the signal-to-noise of the image. Figure~\ref{fig:decomp} shows the
856: results of the decomposition for all wavelengths. The error bars
857: on the plot include covariance between the three templates.
858: One can interpret each of these results as the spectrum of the
859: emission from the region described by each template: the x-ray
860: emitting reverse-shocked ejecta, the [O III] emitting reverse shocked
861: ejecta in dense knots and the forward shocked CSM/ISM material.
862:
863: \begin{figure*}
864: \centering
865: \epsscale{1.0}
866: \includegraphics[width=5in]{fig7.eps}
867: \caption{Results of the spatial decomposition for a stack of continuum
868: images between 20.5 and 22.3 \micron. We show the stacked image in
869: order to increase the signal-to-noise ratio for display purposes.
870: Panel a) shows the stack, Panel b) shows the best fit combination of
871: the three templates and Panel c) shows the residuals of the fit. Note
872: that the residual image has a different grey scale than the images in the
873: Panels a) and b). While the linear combination of the templates
874: reproduces most of the features of the mid-IR images, the residuals do
875: have some persistent features.}
876: \label{fig:stackcont}
877: \end{figure*}
878:
879: \begin{figure*}
880: \centering
881: \epsscale{1.0}
882: \includegraphics[width=5in]{fig8.eps}
883: \caption{Same as in Figure~\ref{fig:stackcont} but for the [O IV] emission
884: line at 25.9 \micron. Note that the grey scale for the residual image
885: has a different scale than the other images. The ring of emission
886: visible in the residual image is most likely due to a slight
887: undersubtraction of the background in the vicinity of a bright
888: emission line.}
889: \label{fig:stackline}
890: \end{figure*}
891:
892: \begin{figure*}
893: \centering
894: \epsscale{1.0}
895: \includegraphics[width=7in]{fig9.eps}
896: \caption{Results of the spatial decomposition of the
897: mid-IR spectrum. Each panel shows the spectrum of the emission
898: associated with each tracer. Error bars include the covariance
899: between the tracers. Spikes in the x-ray and radio results in the
900: vicinity of spectral lines are artifacts of the decomposition
901: technique. The dotted line overplotted in each panel shows the total
902: remnant spectrum. The [O III] template traces the emission from
903: dense knots of reverse shocked ejecta, the x-ray template traces
904: shocked ejecta and the radio template traces shocked CSM/ISM.}
905: \label{fig:decomp}
906: \end{figure*}
907:
908:
909: It is clear from the residuals shown in Figures~\ref{fig:stackcont}
910: and~\ref{fig:stackline} that the correspondence between the templates
911: and the observations, while good, misses some features of the infrared
912: emission. There is a persistently bright region in the lower left of
913: the images which has no counterpart in the combined template. At the
914: wavelength of strong emission lines the combined template subtraction
915: leaves a ring of emission which most likely represents an
916: undersubtraction of the background at these wavelengths. Because of
917: these issues, when determining the best fit dust mass we consider the
918: most reliable answer to be from the fit to the total remnant spectrum,
919: rather than the combination of the spectra associated with x-ray and
920: [O III] emitting ejecta.
921:
922: Despite these uncertainties there are some strong conclusions that can
923: be drawn from the decomposition results. One results that is a good
924: test for the technique is that all of the line emission is found to
925: come from the regions traced by optical [O III] emission, as expected.
926: Other results of the decomposition are: 1) that the spectrum
927: associated with the x-ray emitting ejecta shows dust continuum but no
928: emission lines, 2) there is some dust continuum in the region traced
929: by [O III] emission, which seems to peak at a wavelength around 22
930: \micron\ and 3) the forward shock spectrum also has a small dust
931: emission component that peaks around 20 \micron. We discuss the
932: implications of the forward shocked dust further in
933: Section~\ref{sec:csm}.
934:
935: \subsection{Modeling the Dust Emission}\label{sec:dustmodel}
936:
937: \subsubsection{Constructing the Dust Model}
938:
939: Having established that the mid-IR dust continuum emission from E 0102
940: mainly originates in the reverse shocked ejecta of the remnant, we now
941: proceed to model the dust emission to attempt to place limits on the
942: total amount of dust formed in the CCSN. Because we only see emission
943: from material that has been reheated by the reverse shock, our model
944: for the dust must include only the types of dust that can form out of
945: the nucleosynthetic materials present in the layers of ejecta that are
946: shocked \citep{kozasa89,nozawa03}. A crucial observation is that the
947: ejecta in SNR E 0102 do not appear to be mixed. Observations at x-ray
948: \citep{flanagan04,sasaki01,gaetz00,rasmussen01,hayashi94}, ultraviolet
949: \citep{sasaki06}, optical
950: \citep{finkelstein06,tuohy83,blair00,blair89} and now infrared only
951: detect emission from oxygen, neon, carbon, magnesium and silicon.
952: This is in contrast to the similar O-rich remnant Cas A, which shows
953: emission from sulfur, argon, calcium and iron as well
954: \citep{hwang04,fesen06,rho08}. Either the nucleosynthetic layers of
955: the progenitor star appear to have undergone some macroscopic mixing
956: in Cas A or the reverse shock has encountered the oxygen-burning
957: layers and the carbon-burning layers, for instance, in different parts
958: of the supernova, simultaneously \citep{ennis06}. The same is not
959: true of E 0102 where we see only nucleosynthetic products from the
960: carbon-burning layers. This is most likely a consequence of the
961: ejecta being stratified and unmixed.
962:
963: \citet{nozawa03} model the condensation of dust in the unmixed
964: ejecta of core-collapse supernovae. The types of dust they predict
965: are amorphous carbon in the He-rich layers; Al$_2$O$_3$, Mg$_2$SiO$_4$
966: (forsterite) and MgO in the O-Mg-Si layer; MgSiO$_3$ and SiO$_2$ in
967: the O-Si-Mg layer; and silicon and iron rich species in the deeper
968: nucleosynthetic layers. To construct our model for the dust in E 0102
969: we first consider how deeply into the ejecta the reverse shock has
970: propagated, and thus what species of dust we should include in our
971: model. Measurements of the abundances of oxygen, neon, magnesium and
972: silicon in the x-ray spectra of the shocked ejecta demonstrate that
973: magnesium is $\sim 2$ times more abundant than silicon
974: \citep{flanagan04}. The relative amounts of these two elements
975: locate the reverse shock in the O-Mg-Si layer. For this reason, in
976: our dust model we assume the primary species are amorphous carbon,
977: Al$_2$O$_3$, forsterite and MgO as predicted for the O-Mg-Si layer and
978: the He-rich layer by \citet{nozawa03}.
979:
980: The importance of this approach to the determination of the dust mass
981: in E 0102 arises from trying to find a physically motivated and yet
982: unique model for the dust composition. Given enough dust species out
983: of the complicated mineralogy of dust condensation, one could imagine
984: any number of degenerate fits to the mid-IR spectrum. By restricting
985: the dust model to amorphous carbon, Al$_2$O$_3$, MgO and forsterite we
986: hope to obtain physically meaningful values for the dust mass of each
987: species. As we will discuss further below, even with these
988: constraints the fit is not entirely unique, demonstrating the
989: challenges of dealing with dust continuum emission.
990:
991: \subsubsection{Upper Limit at 70\micron}
992:
993: An important constraint on our determination of the dust mass from the
994: mid-IR spectrum is the upper limit for the surface brightness of the
995: supernova remnant at 70 \micron\ from the S$^3$MC MIPS observations.
996: Contamination by the local foreground is higher at longer wavelengths
997: because of the increasing contribution from cooler dust and the lower
998: resolution at 70 \micron\ means it is more difficult to separate the
999: remnant from the shoulder of N\,76. For these reasons determining the
1000: upper limit is not straightforward. There is no obvious emission
1001: visible from the supernova remnant in the MIPS 70 \micron\ image. To
1002: place a limit on the remnant's surface brightness we subtract a local
1003: background using the same technique outlined in
1004: Section~\ref{sec:reduction} with a mask region 40'' in radius to avoid
1005: contamination from the remnant, since the angular resolution at 70
1006: \micron\ is 18''. The background subtraction leaves emission
1007: consistent with zero, with an error of $\sim 1.0$ MJy sr$^{-1}$ in the
1008: mean. We conservatively adopt a 2-$\sigma$ upper limit of 2.0 MJy
1009: sr$^{-1}$ at 70 \micron\ in the dust fitting to be conservative.
1010:
1011: \subsection{Dust Model Fit Results}\label{sec:fit}
1012:
1013: We performed a non-linear least squares fit to determine the best fit
1014: dust model including the upper limit at 70 \micron. We first fit to
1015: the total spectrum of the remnant extracted over a 22'' radius region
1016: from the background subtracted cube. Our dust model involves the four
1017: grain species discussed above with a fixed size of 0.1 \micron. For
1018: dust grains in the Rayleigh limit the dust mass is independent of the
1019: grain size, so fixing the size at 0.1 \micron\ does not affect the
1020: fit. The parameters of the model are the mass of dust in each species
1021: and its temperature. We also performed the fit on the total remnant
1022: spectrum minus the spectrum associated with the decomposed radio
1023: emission shown in Figure~\ref{fig:decomp}. The results from these two
1024: fits are very similar, as shown in Table~\ref{tab:fitresults}.
1025: Removing the spectrum associated with the shocked CSM/ISM slightly
1026: increases the steepness of the long wavelength continuum leading to
1027: more cool dust in the fit.
1028:
1029: \begin{deluxetable*}{llrc}
1030: \tablewidth{0pt}
1031: \tabletypesize{\scriptsize}
1032: \tablecolumns{4}
1033: \tablecaption{Fit Results\label{tab:fitresults}}
1034: \tablehead{ \multicolumn{1}{l}{Spectrum} &
1035: \multicolumn{1}{l}{Model} &
1036: \multicolumn{1}{c}{Mass} &
1037: \multicolumn{1}{c}{Temperature} \\
1038: \multicolumn{1}{c}{} &
1039: \multicolumn{1}{c}{} &
1040: \multicolumn{1}{c}{(\msun)} &
1041: \multicolumn{1}{c}{(K)} }
1042: \startdata
1043: Total Spectrum & Al$_2$O$_3$ & $<1 \times 10^{-6}$ & $>$ 70 \\
1044: & Am. Carbon & $\phantom{<}3 \times 10^{-3}$ & \phm{$>$}\phn 70 \\
1045: & MgO & $<1 \times 10^{-5}$ & $>$ 70 \\
1046: & Mg$_2$SiO$_4$ & $\phantom{<}2 \times 10^{-5}$ & \phm{$>$}145 \\[0.15in]
1047: Total Spectrum & Mg$_2$SiO$_4$ & $\phantom{<}2 \times 10^{-5}$ & \phm{$>$}145 \\
1048: & Mg$_2$SiO$_4$ & $\phantom{<}1 \times 10^{-2}$ & \phm{$>$}\phn 54 \\[0.15in]
1049: Total Spectrum $-$ CSM/ISM & Am. Carbon & $\phantom{<}6 \times 10^{-3}$ & \phm{$>$}\phn 60 \\
1050: & Mg$_2$SiO$_4$ & $\phantom{<}2 \times 10^{-5}$ & \phm{$>$}145 \\[0.15in]
1051: X-ray Emitting & Am. Carbon & $\phantom{<}4 \times 10^{-3}$ & \phm{$>$}\phn 60 \\
1052: Reverse Shocked Ejecta & Mg$_2$SiO$_4$ & $\phantom{<}8 \times 10^{-6}$ & \phm{$>$}150 \\[0.15in]
1053: \hbox{[O III]} Emitting & Mg$_2$SiO$_4$ & $\phantom{<}2 \times 10^{-6}$ & \phm{$>$}180 \\
1054: Reverse Shocked Ejecta & & & \\[0.15in]
1055: CSM/ISM & ISM Graphite & $\phantom{<}8 \times 10^{-6}$ & \phm{$>$}170 \\
1056: \enddata
1057: \tablecomments{Optical constants for amorphous carbon from
1058: \citet{rouleau91} and ISM graphite from \citet{laor93}. The sources
1059: for the other optical constants can be found in
1060: Table~\ref{tab:colddust}.}
1061: \end{deluxetable*}
1062:
1063:
1064: The spectrum of the remnant is well fit by a combination of forsterite
1065: (Mg$_2$SiO$_4$) and amorphous carbon dust as shown in
1066: Figure~\ref{fig:allfit}. Neither Al$_2$O$_3$ nor MgO play an
1067: important role in the best fit model. Upper limits on the masses of
1068: Al$_2$O$_3$ and MgO are listed in Table~\ref{tab:fitresults} at a
1069: temperature of 70 K. At higher temperatures the limits on the masses
1070: are stricter. The best fit forsterite component has a temperature of
1071: 145 K while the amorphous carbon component, which contains most of the
1072: dust mass we detect, has a temperature of 70 K. However, the fit
1073: results for the cool dust component are not unique. A fit of similar
1074: quality using two temperature components of forsterite is listed in
1075: Table~\ref{tab:fitresults} as well. The continuum shape around 20
1076: \micron\ is not well reproduced without forsterite in the dust model,
1077: however the longer wavelength continuum shape does not distinguish
1078: strongly between forsterite and amorphous carbon. In the following we
1079: proceed by assuming that the dust is amorphous carbon because 1) it
1080: provides a more conservative estimate of the dust mass (i.e. a lower
1081: limit) and 2) theoretical predictions suggest that there should be a
1082: substantial amount of amorphous carbon which, if present, should all
1083: be in the reverse shocked outer layers and visible to us in the
1084: mid-IR.
1085:
1086: \begin{figure*}
1087: \centering
1088: \epsscale{1.0}
1089: \includegraphics[width=5in]{fig10.eps}
1090: \caption{Dust model fit to the spectrum showing $3\times 10^{-3}$
1091: \msun\ of 70 K amorphous carbon and $2\times 10^{-5}$ \msun\ of 145 K
1092: forsterite. The 145 K forsterite component is necessary to reproduce
1093: the continuum shape below $\sim 30$ \micron. The 60 K dust component
1094: is less constrained and could be replaced with forsterite at a
1095: comparable temperature with no impact on the quality of the fit. The
1096: arrow shows the upper limit at 70 \micron.}
1097: \label{fig:allfit}
1098: \end{figure*}
1099:
1100:
1101: The 70 \micron\ upper limit does not affect the amount of forsterite
1102: we measure. On the other hand, the limit does affect both the mass
1103: and temperature of the amorphous carbon dust. Decreasing the limit
1104: tends to make the amorphous carbon dust warmer and less massive,
1105: increasing it makes the amorphous carbon cooler and more massive. We
1106: note that for limits below about 1 MJy sr$^{-1}$ the fit no longer
1107: converges properly, because the limit is inconsistent with the slope
1108: of the data past 30 \micron. We predict that with a small increase in
1109: signal-to-noise and/or angular resolution the remnant should be
1110: detectable at 70 \micron.
1111:
1112: We also perform individual fits to the spectra associated with the
1113: x-ray, [O III] and radio templates in order to learn about the dust in
1114: the x-ray emitting reverse shocked ejecta, the dense knots of reverse
1115: shocked ejecta and the CSM/ISM dust in the forward shock.
1116: Figure~\ref{fig:zoomed} shows the dust continuum associated with the
1117: radio and [O III] templates and the fit listed in
1118: Table~\ref{tab:fitresults}. The continuum shape and signal-to-noise
1119: in the spectrum associated with the radio template do not strongly
1120: constrain the species of dust present in the CSM/ISM material. We show
1121: a representative fit with interstellar graphite from \citet{laor93}.
1122: The fits imply $\sim 10^{-6} - 10^{-5}$ \msun\ of dust with a
1123: temperature around $150-180$ K. We discuss this component further in
1124: Section~\ref{sec:csm}. The dust continuum associated with the [O III]
1125: emitting knots has a better defined shape that is well reproduced by
1126: forsterite but not amorphous carbon. Fits to that component yield a
1127: dust mass of $\sim 2\times 10^{-6}$ \msun\ of forsterite at 180 K.
1128:
1129: \begin{figure*}
1130: \centering
1131: \epsscale{1.0}
1132: \includegraphics[width=5in]{fig11.eps}
1133: \caption{This figure shows a zoomed in version of the spectra
1134: associated with the [O III] and radio templates from the spatial
1135: decomposition along with a fit to the dust continuum. The fit to the
1136: spectrum associated with the radio template is representative. Since
1137: we expect it to be interstellar dust we have used interstellar
1138: graphite from \citet{laor93}. Silicate also reproduces the continuum
1139: shape at the same level of quality. For the [O III] knots, however,
1140: the shape of the continuum is not well reproduced by amorphous carbon.
1141: The fit using forsterite is quite good, given the uncertainties
1142: involved.}
1143: \label{fig:zoomed}
1144: \end{figure*}
1145:
1146:
1147: Using a simplified version of the model which involves only forsterite
1148: and amorphous carbon, we performed the fit at each pixel in the map in
1149: order to learn about the spatial distribution of the dust species.
1150: These results are shown in Figures~\ref{fig:pbpmasses}
1151: and~\ref{fig:pbptemps}. The forsterite to amorphous carbon ratio
1152: tracks the intensity of the x-ray ring of the reverse shock. This
1153: resemblance may be an indication that forsterite is more abundant in
1154: the deeper layers of the ejecta which are now encountering the reverse
1155: shock.
1156:
1157: \begin{figure*}
1158: \centering
1159: \epsscale{1.0}
1160: \includegraphics[width=5in]{fig12.eps}
1161: \caption{Dust masses determined at each pixel. The first panel shows
1162: the mass of amorphous carbon, the second shows the mass of forsterite
1163: and the third shows the ratio of forsterite to amorphous carbon. The
1164: ratio is highest at the boundary of the reverse shock which agrees
1165: with the idea that the reverse shock is just beginning to encounter
1166: the O-Mg-Si layer where forsterite has been produced.}
1167: \label{fig:pbpmasses}
1168: \end{figure*}
1169:
1170: \begin{figure*}
1171: \centering
1172: \epsscale{1.0}
1173: \includegraphics[width=5in]{fig13.eps}
1174: \caption{Dust temperatures determined at each pixel. The temperature
1175: of the amorphous carbon dust is consistent with grains of $\sim 0.1$
1176: \micron\ in equilibrium with the x-ray emitting reverse-shocked
1177: plasma. The forsterite component has temperatures which are higher
1178: than the predicted equilibrium in the x-ray emitting plasma. This may
1179: be the result of the forsterite abundance increasing deeper in the
1180: ejecta, as suggested by the forsterite to amorphous carbon abundance
1181: shown in Figure~\ref{fig:pbpmasses}, where the recent passage of the
1182: reverse shock has not yet destroyed smaller grains which can be
1183: stochastically heated.}
1184: \label{fig:pbptemps}
1185: \end{figure*}
1186:
1187:
1188: \section{Discussion}\label{sec:discussion}
1189:
1190: \subsection{Newly Formed Dust}
1191:
1192: We detect $3\times 10^{-3}$ \msun\ of 70 K amorphous carbon and
1193: $2\times 10^{-5}$ \msun\ of 145 K forsterite in the ejecta of E 0102.
1194: The equilibrium temperatures of grains in the x-ray emitting plasma
1195: are between 50 and 100 K for sizes between 1.0 and 0.001
1196: \micron. The temperature we measure for the amorphous carbon
1197: component corresponds to the equilibrium temperature for grains with
1198: sizes of $\sim 0.1$ \micron. Figure~\ref{fig:stoch} verifies that
1199: grains of this size should be in equilibrium.
1200:
1201: The temperature of the forsterite dust component is hotter than the
1202: equilibrium temperature that could be achieved by grains in the x-ray
1203: emitting plasma of the reverse shock, but is consistent with
1204: temperatures in the dense [O III] emitting knots. In fitting the
1205: decomposed spectrum associated with the [O III] emission we do see
1206: forsterite with temperatures around 180 K, typical for grains with
1207: sizes around 0.01 \micron\ under the expected conditions. If,
1208: however, most of the forsterite is located in the x-ray emitting
1209: plasma, as the results of our spatial decomposition suggest, the
1210: temperature would imply small dust grains that are stochastically
1211: heated. In the following section we show that the sputtering
1212: timescale in the reverse shocked ejecta for small grains is less than
1213: the age of the remnant. The presence of small, stochastically heated
1214: dust grains in the x-ray emitting plasma can be explained if the dust
1215: has only recently been through the reverse shock, such that it has
1216: been in the plasma for less than the sputtering time. Given the
1217: results shown in Figure~\ref{fig:pbpmasses}, i.e. that the abundance
1218: of forsterite seems to increase near the bright, inner boundary of the
1219: reverse shock, this may very well be the explanation for why we see
1220: forsterite grains out of equilibrium with the plasma.
1221:
1222: \subsection{Dust Sputtering and Grain Size Constraints}\label{sec:sput}
1223:
1224: The dust we see in E 0102 at present may not be the amount that was
1225: initially produced in the supernova or what was initially swept up by
1226: the blast wave. Dust grains in shock waves can be destroyed by
1227: sputtering, where impacts by hot gas particles erode the grain.
1228:
1229: For the dust in the forward shocked ISM/CSM we can estimate the
1230: sputtering time with the relation
1231: \begin{equation}
1232: \tau_{sp} \sim \frac{a}{da/dt} \sim 10^6
1233: \frac{a}{\micron}\left(\frac{n_e}{\mathrm{cm}^{-3}}\right)^{-1}
1234: \mathrm{yr} ,
1235: \end{equation}
1236: from \citet{draine79} for gas with temperatures greater than $\sim
1237: 10^6$ K and solar abundances. In this regime the majority of the
1238: sputtering is done by hydrogen ions and $n_e \approx n_H$, so the
1239: abundance of heavier elements does not significantly affect the
1240: calculations and we can use these results for the SMC metallicity.
1241: Following the discussion in Section~\ref{sec:temps}, we estimate
1242: densities in the x-ray emitting forward shocked material of $\sim 1$
1243: cm$^{-3}$. Thus, the sputtering timescale for dust grains in the
1244: forward shock is $\sim 10^5$ years for an 0.1 \micron\ grain and $\sim
1245: 10^3$ years for an 0.001 \micron\ grain. Given the age of the
1246: remnant, dust grains with sizes below 0.001 \micron\ should have been
1247: destroyed behind the shock.
1248:
1249: In the reverse shocked gas the sputtering of dust is due to collisions
1250: with oxygen ions, since oxygen is the dominant component of the
1251: ejecta. We can use a simple scaling argument based on the sputtering
1252: time relation above to estimate the lifetime of grains in the reverse
1253: shocked material. The sputtering rate $R$ in (g s$^{-1}$) is given by
1254: \begin{equation} R \sim \sigma n v Y , \end{equation} where $n$ and
1255: $v$ are the density and velocity of the sputtering particles, $\sigma$
1256: is the grain cross section and $Y$ is the sputtering yield per
1257: collision in units of mass per collision. Therefore, the change of
1258: grain radius with time, assuming constant density for the grain and a
1259: thermal velocity for the colliding particles $v\sim (kT/m)^{1/2}$,
1260: goes as: \begin{equation} \frac{da}{dt} \propto n T^{1/2} m^{-1/2}Y .
1261: \end{equation} Scaling the sputtering between the forward shock and
1262: reverse shock gas involves a number of assumptions. First, the
1263: density of oxygen ions in the reverse shock is not simply related to
1264: the electron density, because the ionization state of the gas is not
1265: in equilibrium (in which case the number of free electrons donated per
1266: oxygen atom would be a predictable function of the temperature).
1267: Second, the temperature of the oxygen ions is not an easily determined
1268: quantity either, given the uncertain equilibration timescale between
1269: the electrons and the ions after the shock. As an estimate we can
1270: assume $n_{\mathrm{O}} \sim n_e/3$ as a reasonable value for the
1271: number of free electrons donated by each oxygen atom, and
1272: $T_{\mathrm{O}} \sim T_e$, which should be the case if the
1273: equilibration timescale between the ions and electrons is less than
1274: $\sim 1000$ years. In the case where the ions are hotter than the
1275: electrons, the sputtering timescale will decrease. \citet{nozawa07}
1276: investigated the sputtering of dust grains in the reverse shocks of
1277: supernova remnants and via fits to experimental data for a variety of
1278: grain compositions and collider properties found that sputtering by
1279: oxygen ions has $\sim 50$ times higher yield than sputtering by
1280: hydrogen ions. Using these values ($n_{\mathrm{O}} \sim n_e/3$, $T_e
1281: \sim T_{\mathrm{O}}$, and $Y_{\mathrm{O}} \sim 50 Y_{\mathrm{H}}$) we
1282: estimate that the sputtering timescale in the reverse shocked ejecta,
1283: which has an electron density of $\sim 20$ cm$^{-3}$, for a grain of
1284: size $0.1$ \micron\ is $\sim 1000$ years. Thus, in the x-ray emitting,
1285: oxygen-rich ejecta, the destruction time for grains smaller than 0.1
1286: \micron\ is shorter than the age of the remnant and we expect that the
1287: grain size distribution in the plasma has been altered significantly.
1288:
1289: The amorphous carbon dust, which makes up the majority of the mass we
1290: detect from the reverse shocked ejecta, has a temperature which
1291: suggests emission from $\sim 0.1$ \micron\ grains. Based on our
1292: estimate of the sputtering timescale in the reverse shocked plasma,
1293: these grains would be able to survive the lifetime of the remnant.
1294: The warmer forsterite dust component is found to be associated
1295: with the dense [O III] knots, which have a higher equilibrium
1296: temperature due to their high densities, and with the most recently
1297: reverse-shocked ejecta, in which smaller dust grains may still exist
1298: and be stochastically heated to higher temperatures.
1299:
1300: \subsection{The Possibility of Cold Dust}
1301:
1302: Because our observations are only sensitive to warm, recently shocked
1303: dust, we cannot quantify the total amount of newly formed dust
1304: produced in the remnant. To understand the limits that the 70
1305: \micron\ observation places on the mass of cold dust, we construct a
1306: model for the unshocked dust in the supernova remnant based on the
1307: condensation models of \citet{nozawa03} for a 25 \msun\ progenitor.
1308: We list the dust species included in the model, the source for their
1309: optical constants, and the masses predicted by \citet{nozawa03} in
1310: Table~\ref{tab:colddust}. The total mass of dust predicted by their
1311: models is 0.6 \msun, minus the amorphous carbon component which we
1312: argue must primarily be in the outer, shocked layers of the ejecta.
1313: In Figure~\ref{fig:colddust} we show the remnant's spectrum, the best
1314: fit model with amorphous carbon and forsterite from
1315: Section~\ref{sec:fit}, and the spectrum of the cold dust at
1316: temperatures of 20 (blue) and 35 K (black). It is clear that our limit
1317: at 70 \micron\ does not preclude the possibility of a substantial mass
1318: of cold dust in the remnant, since at a temperature of 20 K, the
1319: entire predicted mass of dust can be hidden at 70 \micron.
1320: Determining the total amount of newly formed dust in E 0102 will
1321: require longer wavelength observations.
1322:
1323: \begin{deluxetable*}{lccc}
1324: \tablewidth{0pt}
1325: \tabletypesize{\scriptsize}
1326: \tablecolumns{4}
1327: \tablecaption{Cold Dust Model Components\label{tab:colddust}}
1328: \tablehead{ \multicolumn{1}{l}{Species} &
1329: \multicolumn{1}{l}{Mass} &
1330: \multicolumn{1}{l}{Optical Constants Reference} \\
1331: \multicolumn{1}{c}{} &
1332: \multicolumn{1}{c}{(\msun)} &
1333: \multicolumn{1}{c}{} }
1334: \startdata
1335: Mg$_2$SiO$_4$ & 0.2 & \citet{jaeger03} \\
1336: Amorphous Silicon & 0.2 & \citet{piller85} \\
1337: SiO$_2$ & 0.07 & \citet{philipp85} \\
1338: FeS & 0.07 & \citet{begemann94} \\
1339: MgO & 0.06 & \citet{roessler91} \\
1340: Fe & 0.02 & \citet{lynch91} \\
1341: MgSiO$_3$ & 0.003 & \citet{dorschner95} \\
1342: Al$_2$O$_3$ & 0.002 & \citet{begemann97} \\
1343: \enddata
1344: \end{deluxetable*}
1345:
1346: \begin{figure*}
1347: \centering
1348: \epsscale{1.0}
1349: \includegraphics[width=5in]{fig14.eps}
1350: \caption{This figure shows predictions for the far-IR spectrum of dust
1351: in the remnant using the cold dust model components listed in
1352: Table~\ref{tab:colddust}. The cold dust spectrum contains the total
1353: dust mass predicted by the \citet{nozawa03} models for a 25 \msun\
1354: progenitor excluding amorphous carbon. The dashed line shows this
1355: model assuming the cold dust has a temperature of 20 K and the
1356: dotted line shows the same model with a temperature of 35 K. The
1357: black dash-dot and long-dashed lines show the total spectrum of cold
1358: dust plus the warm dust we detect in the mid-IR. This plot
1359: demonstrates that all of the predicted $\sim 0.6$ \msun\ of dust can
1360: be present in the remnant if it has a temperature of 20 K, showing
1361: that our current observations alone can not strongly constrain the
1362: dust condensation efficiency.}
1363: \label{fig:colddust}
1364: \end{figure*}
1365:
1366:
1367: %If we assume that the condensation efficiency in the ejecta is uniform
1368: %and that the mass of dust in the reverse-shocked layer is close to
1369: %what it initially was before the passage of the reverse shock (two
1370: %very debatable propositions) we can estimate the dust mass that might
1371: %be cold and undetectable. \citet{flanagan04} argue for $\sim 6$
1372: %\msun\ of oxygen in the reverse shocked ejecta and $\sim 2$ \msun\ of
1373: %neon. The abundances of silicon and magnesium are found to be much
1374: %lower.
1375:
1376: \subsection{Implications for Dust Production in CCSN and Comparison
1377: with Previous Results}
1378:
1379: \citet{stanimirovic05} found a mass of $8\times 10^{-4}$ \msun\ of
1380: dust in E 0102 based on the flux at 24 \micron. Using upper limits at
1381: 8 and 70 \micron, they argued that the dust had to have a temperature
1382: around 120 K. They assumed that 60\% of the 24 \micron\ emission was
1383: from the contribution of [O IV], a value we measure to be 15\%. These
1384: differences alone do not explain the $\sim 4$ times more dust that we
1385: find compared to \citet{stanimirovic05}. Instead, the greater
1386: contribution of cooler dust with a temperature of around 70 K
1387: accounts for most of the difference. Our measurement of the
1388: temperature and mass of forsterite dust are comparable to their
1389: predictions. The spectrum past 30 \micron, however, is not compatible
1390: with only a $\sim 120$ K component.
1391:
1392: Similar mid-infrared spectral mapping observations have been carried
1393: out on Cas A, another young, oxygen-rich CCSN remnant, by
1394: \citet{rho08}. There are some important differences between the
1395: observations. Cas A's relative proximity and brightness allows more
1396: detailed fits to the spectrum of the remnant. Cas A also has emission
1397: from S, Si, and Ar in its spectrum in addition to O, Ne, and Mg.
1398: This could be the result of macroscopic mixing, as suggested by
1399: \citet{douvion99}, or of velocity inhomogeneity (Smith et al., in
1400: prep.), causing different layers of the ejecta to encounter the
1401: reverse shock simultaneously. In both of these situations, the
1402: overall composition of the nucleosynthetic layers is preserved and the
1403: same species of dust should have formed in the ejecta as in E 0102. It
1404: may not be possible, however, to separate different nucleosynthetic
1405: layers spatially, making the modeling more complex.
1406:
1407: \citet{rho08} find on the order of $0.02-0.05$ \msun\ of dust in the
1408: Cas A remnant, with contributions from a variety of dust species.
1409: They find three distinct spectra in Cas A, a 21 \micron\ peak spectrum,
1410: a weak 21 \micron\ spectrum and a featureless spectrum. The weak 21
1411: \micron\ spectrum has associated strong neon lines, similar to what we
1412: see in E\,0102. The dominant dust components by mass that
1413: \citet{rho08} fit to the weak 21 \micron\ spectrum are two temperature
1414: components of amorphous carbon at $\sim 80$ and $\sim 200$ K totaling
1415: $\sim 1-2\times 10^{-3}$ \msun\ and $\sim 0.6-1.4\times 10^{-2}$
1416: \msun\ of FeO at $\sim 60$ K. The mass of amorphous carbon we find is
1417: comparable to their results, however we have no evidence of FeO in our
1418: spectrum and it seems unlikely that a substantial amount of FeO could
1419: be produced, since it requires microscopic mixing of the ejecta.
1420: Comparing with the total amount of dust that \citet{rho08} find in Cas
1421: A, we see similar amounts of both amorphous carbon and Mg$_2$SiO$_4$,
1422: on the order of $10^{-3}$ and $10^{-5}$ \msun\ respectively. However,
1423: they argue for significant amounts of Fe, FeO, FeS and SiO$_2$,
1424: producing a total dust mass of between $0.02-0.05$ \msun\ of
1425: newly-formed dust---an order-of-magnitude larger than what we observe
1426: in E\,0102. The progress of the reverse shock into the deeper
1427: nucleosynthetic layers of Cas A may partially account for the
1428: difference in the observed dust mass between the two remnants. It is
1429: possible that similar dust masses could be present but the iron- and
1430: silicon-rich dust species are in unshocked layers of the ejecta of
1431: E\,0102 and are too cold for us to detect in these observations.
1432:
1433: Given the progress of the reverse shock into the ejecta, we can claim
1434: with good confidence that we should see essentially all of the
1435: amorphous carbon dust that has been produced in the mid-IR. This
1436: being the case, we find substantially less amorphous carbon dust than
1437: predicted by dust condensation models. The models of \citet{nozawa03}
1438: predict on the order of $10^{-1}-10^{-2}$ \msun\ of amorphous carbon,
1439: whereas we only see $\sim 3\times 10^{-3}$ \msun. We estimate that
1440: the sputtering time for dust in the reverse-shocked ejecta is less
1441: than the age of the remnant, so we may be observing less dust then was
1442: initially formed. \citet{nozawa07} calculate the dust destruction via
1443: sputtering in the reverse shock for the same models quoted above using
1444: a simple prescription for the interaction of the ejecta with the
1445: surrounding ISM. They find that for an ambient density of $0.1$
1446: cm$^{-3}$, only 8\% of the amorphous carbon dust is destroyed for
1447: their 25 \msun\ progenitor. Most of the mass of carbon dust in that
1448: model is in grains with sizes greater than 0.05 \micron, which have
1449: longer sputtering lifetimes. We have no constraints on the initial
1450: grain size, so it is difficult to estimate how much of the dust in the
1451: remnant has been destroyed up to this point.
1452:
1453: Previous observations of newly formed dust in CCSN have yielded dust
1454: masses in the range of $10^{-3}$ to $10^{-5}$ \msun, with some
1455: controversial measurements claiming higher masses. The majority of
1456: the measurements come from CCSN of Type II or their remnants, which
1457: have hydrogen envelopes and red supergiant progenitors. Dust
1458: formation has been observed in two Type Ib events: SN 1990I
1459: \citep{elmhamdi04} and the peculiar event SN 2006jc
1460: \citep{smith08,nozawa08} which shows evidence for dust formation at
1461: very early times, around 50 days after the explosion. The
1462: masses determined from observations of 2006jc range from a few
1463: $10^{-6}$ to a few $10^{-3}$ depending on the observation date,
1464: wavelength and dust model. Recently, \citet{nozawa08} have modeled
1465: dust condensation in a Type Ib event in order to assess the effects of
1466: a small or non-existent hydrogen envelope on the dust condensation
1467: efficiency. The high expansion velocities mean that the density drops
1468: quickly, leading to dust condensation at earlier times in these
1469: events. The condensation efficiency is high, however, yielding $0.7$
1470: \msun\ of amorphous carbon in their model of SN 2006jc, which is in
1471: contrast to the relatively small amount of dust detected in the
1472: near-IR and mid-IR observations. Our results for E 0102 lie near the
1473: upper bounds of the observed dust mass in SN 2006jc. It may not be the
1474: case, however, that dust is forming in the ejecta of SN 2006jc.
1475: \citet{smith08} argue that the dust condensation occurs in the
1476: cooling region of the shocks created by the interaction of the
1477: blast wave with an expanding shell of circumstellar material. If this
1478: is the case, we may not expect to see comparable species or amounts of
1479: dust in SN 2006jc and E 0102.
1480:
1481:
1482: \subsection{The Circumstellar Environment of SNR E
1483: 0102}\label{sec:csm}
1484:
1485: The progenitor of SNR E 0102 was most likely a massive star that
1486: lost most of its hydrogen/helium envelope before exploding as a Type
1487: Ib/Ic or IIb/L supernova \citep{blair00,chevalier05}. As such, it
1488: would have significantly impacted the circumstellar medium in the
1489: form of winds and mass loss. Here we briefly discuss what we can
1490: learn about the CSM/ISM from our observations.
1491:
1492: The spectrum associated with the radio continuum template, as
1493: discussed in Section~\ref{sec:decomp}, shows the dust continuum
1494: emission from forward shocked CSM/ISM material. Fits to this spectrum
1495: with interstellar graphite and silicate grains give masses in the
1496: range $10^{-5}-10^{-6}$ \msun\ and temperatures of $\sim 180$ K. If
1497: the remnant were expanding into the average SMC interstellar medium
1498: with a density of $n \sim 0.1$ cm$^{-3}$, we would expect
1499: approximately 2 \msun\ of swept up gas and $\sim 3\times 10^{-3}$
1500: \msun\ of swept up dust at this point in time \citep[using a dust to
1501: gas ratio of 1/700 as determined by][]{leroy07}. Although the fit
1502: dust mass is far less than this expected value, the temperature
1503: indicates that the dust we see is not in collisional equilibrium and
1504: therefore we are only detecting the fraction of the dust that is warm
1505: at a given time (see Figure~\ref{fig:teq} for the predicted
1506: equilibrium temperature in the forward shock).
1507:
1508: For stochastic heating, the fraction of grains which are warm at any
1509: time is given approximately by the ratio of the collision time to the
1510: cooling time $f \sim \tau_{cool}/\tau_{coll}$. This ratio is shown as
1511: a function of grain size in Figure~\ref{fig:stoch}. Grains with $a >
1512: 0.02$ \micron\ will not be stochastically heated and will have
1513: equilibrium temperatures between 40 and 70 K. In
1514: Section~\ref{sec:sput}, we estimated that grains with $a < 0.001$
1515: \micron\ should be destroyed by sputtering on timescales shorter than
1516: the age of the remnant. Within these limits $0.001 < a < 0.02$, we
1517: can estimate that $f \sim 0.01-1$, i.e. that we detect between 1 and
1518: 100\% of the dust mass in the forward shocked CSM/ISM. Thus, the
1519: amount of dust in the forward shocked CSM/ISM is essentially
1520: unconstrained by our observations, ranging from $10^{-6} - 10^{-3}$
1521: \msun.
1522:
1523: Depending on the progenitor star and its winds, we may not expect to
1524: see any dust in the forward-shocked plasma. This would be the case if
1525: the blast wave has not yet reached the pristine ISM and the
1526: circumstellar medium has been cleared of dust by the winds of the
1527: progenitor star. \citet{chevalier05} argue that E\,0102 is still
1528: interacting with the wind of its progenitor star based on the
1529: positions of the forward and reverse shocks and the lack of limb
1530: brightening in the x-ray emission from the forward shock. The dust
1531: content of the circumstellar medium will depend on the duration of the
1532: wind and the previous evolution of the progenitor (for example,
1533: whether there were outbursts similar to those observed from the
1534: progenitor of SN 2006jc \citep{pastorello07}) among other factors.
1535: Further observational constraints on the progenitor mass and evolution
1536: will be necessary to understand whether the dust we see associated
1537: with the forward shock fits with the overall picture of the evolution
1538: of E\,0102.
1539:
1540: \section{Summary \& Conclusions}\label{sec:conclusions}
1541:
1542: To summarize, we present mid-IR spectral mapping observations of the
1543: supernova remnant 1E\,0102.2$-$7219 in the vicinity of N\,76 in the
1544: Small Magellanic Cloud. The $\sim 1000$ year old, oxygen-rich remnant
1545: is thought to be the product of a Type Ib/Ic or IIb/L supernova.
1546: Currently the reverse shock, produced by the dynamical interaction of
1547: the ejecta with the surrounding ISM/CSM, is propagating back through
1548: the ejecta. Based on spectroscopic observations at x-ray,
1549: ultraviolet, optical and now infrared wavelengths, the ejecta down to
1550: the depth of the reverse shock appear to be composed of oxygen, neon,
1551: magnesium, silicon and carbon. This composition indicates that the
1552: reverse shock has only penetrated into the deeper layers of
1553: nucleosynthetic products and that no large-scale mixing of deeper
1554: layers has occurred.
1555:
1556: The mid-IR spectrum of the remnant shows fine-structure emission lines of
1557: oxygen and neon on top of emission from warm dust. We use observations
1558: of the remnant at x-ray, radio and optical wavelengths as templates to
1559: decompose the infrared emission into components associated with the
1560: newly-formed dust in the reverse shocked ejecta and ISM/CSM dust in
1561: the forward shocked region. Our decomposition of the spectrum shows
1562: (1) dust continuum associated with the x-ray emitting reverse shocked
1563: ejecta, (2) oxygen and neon fine-structure emission lines and warm dust
1564: continuum associated with the optical [O III] emitting gas in dense
1565: knots of ejecta where the reverse shock is slow and radiative and (3)
1566: a small amount of dust emission associated with the forward shock.
1567: The majority of the dust is associated with the x-ray emitting reverse
1568: shocked material.
1569:
1570: We construct a model for the newly-formed dust in the unmixed ejecta.
1571: Using x-ray measurements of the abundance of magnesium and silicon we
1572: estimate the location of the reverse shock to be in the O-Mg-Si layer.
1573: We model the dust emission with amorphous carbon, forsterite
1574: (Mg$_2$SiO$_4$), Al$_2$O$_3$, and MgO as appropriate for the
1575: composition of the ejecta down to the depth of the O-Mg-Si layer,
1576: motivated by the results of dust condensation models. The best fit
1577: model contains $3\times 10^{-3}$ \msun\ of amorphous carbon at a
1578: temperature of 70 K and $2\times 10^{-5}$ \msun\ of forsterite with
1579: temperature of 145 K. The masses of Al$_2$O$_3$ and MgO are limited
1580: to be less than $1\times 10^{-6}$ and $1\times 10^{-5}$ \msun\ at a
1581: temperature of 70 K, respectively, since they do not contribute
1582: substantially to the continuum shape in the mid-IR. The temperature
1583: of amorphous carbon agrees with expectations for 0.1 \micron\ grains
1584: in the x-ray emitting plasma behind the reverse shock, which agrees
1585: with our estimate that smaller grains have been destroyed by
1586: sputtering. The temperature of the forsterite component suggests that
1587: the emission is from smaller, stochastically heated grains which have
1588: recently been heated by the passage of the reverse shock or are
1589: residing in dense knots of ejecta material.
1590:
1591: If the reverse shock is presently encountering material from the
1592: O-Mg-Si layer of the progenitor we should be detecting the majority of
1593: the newly-formed amorphous carbon dust, since it forms primarily in
1594: the outermost layers of the ejecta. Comparison with the
1595: \citet{nozawa03} results show substantially less amorphous carbon dust
1596: in E 0102 than would be predicted based on the progenitor mass. Our
1597: results for the masses of amorphous carbon and forsterite are
1598: comparable to work on Cas A by \citet{rho08}.
1599:
1600: Our mid-IR observations are not sensitive to cold dust present in the
1601: remnant. We show that if the dust has a temperature of 20 K, $\sim
1602: 0.6$ \msun\ of dust could be present in the remnant and undetectable
1603: in our observations. Observations at longer wavelengths by Herschel,
1604: LABOCA on APEX or ALMA will be necessary to determine the amount of
1605: cold dust present in the remnant. Without a substantial mass of
1606: hidden cold dust, the dust production in E\,0102 falls orders of
1607: magnitude below what would be necessary to explain the observations of
1608: dust around high redshift quasars.
1609:
1610: \acknowledgements
1611:
1612: The authors would like to thank S. Finkelstein for providing us with
1613: the [O III] map. K.S. would like to thank Eli Dwek, Maryam Modjaz,
1614: and Nathan Smith for useful conversations and Adam Leroy for providing
1615: a quiet place to work at a crucial time. K.S. acknowledges the
1616: support of an NSF Graduate Research Fellowship. This work is based on
1617: observations made with the Spitzer Space Telescope, which is operated
1618: by the Jet Propulsion Laboratory, California Institute of Technology
1619: under a contract with NASA. This research was supported in part by by
1620: NASA through awards issued by JPL/Caltech (NASA-JPL Spitzer grant
1621: 1264151 awarded to Cycle 1 project 3316, and grants 1287693 and
1622: 1289519 awarded to Cycle 3 project 30491).
1623:
1624: {\it Facilities:} \facility{Spitzer ()}
1625:
1626: %%%%%%%%%%%%%%
1627: % REFERENCES %
1628: %%%%%%%%%%%%%%
1629:
1630: %\bibliographystyle{apj}
1631: %\bibliography{e0102_refs}
1632: \begin{thebibliography}{73}
1633: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
1634:
1635: \bibitem[{{Amy} \& {Ball}(1993)}]{amy93}
1636: {Amy}, S.~W. \& {Ball}, L. 1993, \apj, 411, 761
1637:
1638: \bibitem[{{Arendt} {et~al.}(1992){Arendt}, {Dwek}, \& {Leisawitz}}]{arendt92}
1639: {Arendt}, R.~G., {Dwek}, E., \& {Leisawitz}, D. 1992, \apj, 400, 562
1640:
1641: \bibitem[{{Arendt} {et~al.}(1999){Arendt}, {Dwek}, \& {Moseley}}]{arendt99}
1642: {Arendt}, R.~G., {Dwek}, E., \& {Moseley}, S.~H. 1999, \apj, 521, 234
1643:
1644: \bibitem[{{Beelen} {et~al.}(2006){Beelen}, {Cox}, {Benford}, {Dowell},
1645: {Kov{\'a}cs}, {Bertoldi}, {Omont}, \& {Carilli}}]{beelen06}
1646: {Beelen}, A., {Cox}, P., {Benford}, D.~J., {Dowell}, C.~D., {Kov{\'a}cs}, A.,
1647: {Bertoldi}, F., {Omont}, A., \& {Carilli}, C.~L. 2006, \apj, 642, 694
1648:
1649: \bibitem[{{Begemann} {et~al.}(1997){Begemann}, {Dorschner}, {Henning},
1650: {Mutschke}, {G\"uertler}, {K\"oempe}, \& {Nass}}]{begemann97}
1651: {Begemann}, B., {Dorschner}, J., {Henning}, T., {Mutschke}, H., {G\"uertler},
1652: J., {K\"oempe}, C., \& {Nass}, R. 1997, \apj, 476, 199
1653:
1654: \bibitem[{{Begemann} {et~al.}(1994){Begemann}, {Dorschner}, {Henning},
1655: {Mutschke}, \& {Thamm}}]{begemann94}
1656: {Begemann}, B., {Dorschner}, J., {Henning}, T., {Mutschke}, H., \& {Thamm}, E.
1657: 1994, \apjl, 423, L71
1658:
1659: \bibitem[{{Bertoldi} {et~al.}(2003){Bertoldi}, {Carilli}, {Cox}, {Fan},
1660: {Strauss}, {Beelen}, {Omont}, \& {Zylka}}]{bertoldi03}
1661: {Bertoldi}, F., {Carilli}, C.~L., {Cox}, P., {Fan}, X., {Strauss}, M.~A.,
1662: {Beelen}, A., {Omont}, A., \& {Zylka}, R. 2003, \aap, 406, L55
1663:
1664: \bibitem[{{Blair} {et~al.}(2007){Blair}, {Ghavamian}, {Long}, {Williams},
1665: {Borkowski}, {Reynolds}, \& {Sankrit}}]{blair07}
1666: {Blair}, W.~P., {Ghavamian}, P., {Long}, K.~S., {Williams}, B.~J., {Borkowski},
1667: K.~J., {Reynolds}, S.~P., \& {Sankrit}, R. 2007, \apj, 662, 998
1668:
1669: \bibitem[{{Blair} {et~al.}(2000){Blair}, {Morse}, {Raymond}, {Kirshner},
1670: {Hughes}, {Dopita}, {Sutherland}, {Long}, \& {Winkler}}]{blair00}
1671: {Blair}, W.~P., {Morse}, J.~A., {Raymond}, J.~C., {Kirshner}, R.~P., {Hughes},
1672: J.~P., {Dopita}, M.~A., {Sutherland}, R.~S., {Long}, K.~S., \& {Winkler},
1673: P.~F. 2000, \apj, 537, 667
1674:
1675: \bibitem[{{Blair} {et~al.}(1989){Blair}, {Raymond}, {Danziger}, \&
1676: {Matteucci}}]{blair89}
1677: {Blair}, W.~P., {Raymond}, J.~C., {Danziger}, J., \& {Matteucci}, F. 1989,
1678: \apj, 338, 812
1679:
1680: \bibitem[{{Bode} \& {Evans}(1980)}]{bode80}
1681: {Bode}, M.~F. \& {Evans}, A. 1980, \mnras, 193, 21P
1682:
1683: \bibitem[{{Bolatto} {et~al.}(2007){Bolatto}, {Simon}, {Stanimirovi{\'c}}, {van
1684: Loon}, {Shah}, {Venn}, {Leroy}, {Sandstrom}, {Jackson}, {Israel}, {Li},
1685: {Staveley-Smith}, {Bot}, {Boulanger}, \& {Rubio}}]{bolatto07}
1686: {Bolatto}, A.~D., {Simon}, J.~D., {Stanimirovi{\'c}}, S., {van Loon}, J.~T.,
1687: {Shah}, R.~Y., {Venn}, K., {Leroy}, A.~K., {Sandstrom}, K., {Jackson}, J.~M.,
1688: {Israel}, F.~P., {Li}, A., {Staveley-Smith}, L., {Bot}, C., {Boulanger}, F.,
1689: \& {Rubio}, M. 2007, \apj, 655, 212
1690:
1691: \bibitem[{{Bot} {et~al.}(2004){Bot}, {Boulanger}, {Lagache}, {Cambr{\'e}sy}, \&
1692: {Egret}}]{bot04}
1693: {Bot}, C., {Boulanger}, F., {Lagache}, G., {Cambr{\'e}sy}, L., \& {Egret}, D.
1694: 2004, \aap, 423, 567
1695:
1696: \bibitem[{{Bouchet} {et~al.}(2006){Bouchet}, {Dwek}, {Danziger}, {Arendt}, {De
1697: Buizer}, {Park}, {Suntzeff}, {Kirshner}, \& {Challis}}]{bouchet06}
1698: {Bouchet}, P., {Dwek}, E., {Danziger}, J., {Arendt}, R.~G., {De Buizer},
1699: I.~J.~M., {Park}, S., {Suntzeff}, N.~B., {Kirshner}, R.~P., \& {Challis}, P.
1700: 2006, \apj, 650, 212
1701:
1702: \bibitem[{{Chevalier}(2005)}]{chevalier05}
1703: {Chevalier}, R.~A. 2005, \apj, 619, 839
1704:
1705: \bibitem[{{Dorschner} {et~al.}(1995){Dorschner}, {Begemann}, {Henning},
1706: {J\"ager}, \& {Mutschke}}]{dorschner95}
1707: {Dorschner}, J., {Begemann}, B., {Henning}, T., {J\"ager}, C., \& {Mutschke},
1708: H. 1995, \aap, 300, 503
1709:
1710: \bibitem[{{Douvion} {et~al.}(1999){Douvion}, {Lagage}, \&
1711: {Cesarsky}}]{douvion99}
1712: {Douvion}, T., {Lagage}, P.~O., \& {Cesarsky}, C.~J. 1999, \aap, 352, L111
1713:
1714: \bibitem[{{Draine} \& {Anderson}(1985)}]{draine85}
1715: {Draine}, B.~T. \& {Anderson}, N. 1985, \apj, 292, 494
1716:
1717: \bibitem[{{Draine} \& {Salpeter}(1979)}]{draine79}
1718: {Draine}, B.~T. \& {Salpeter}, E.~E. 1979, \apj, 231, 77
1719:
1720: \bibitem[{{Dunne} {et~al.}(2003){Dunne}, {Eales}, {Ivison}, {Morgan}, \&
1721: {Edmunds}}]{dunne03}
1722: {Dunne}, L., {Eales}, S., {Ivison}, R., {Morgan}, H., \& {Edmunds}, M. 2003,
1723: \nat, 424, 285
1724:
1725: \bibitem[{{Dwek}(1983)}]{dwek83}
1726: {Dwek}, E. 1983, \apj, 274, 175
1727:
1728: \bibitem[{{Dwek}(1986)}]{dwek86}
1729: ---. 1986, \apj, 302, 363
1730:
1731: \bibitem[{{Dwek}(1987)}]{dwek87}
1732: ---. 1987, \apj, 322, 812
1733:
1734: \bibitem[{{Dwek} {et~al.}(2007){Dwek}, {Galliano}, \& {Jones}}]{dwek07}
1735: {Dwek}, E., {Galliano}, F., \& {Jones}, A.~P. 2007, \apj, 662, 927
1736:
1737: \bibitem[{{Dwek} \& {Werner}(1981)}]{dwek81}
1738: {Dwek}, E. \& {Werner}, M.~W. 1981, \apj, 248, 138
1739:
1740: \bibitem[{{Elmhamdi} {et~al.}(2004){Elmhamdi}, {Danziger}, {Cappellaro}, {Della
1741: Valle}, {Gouiffes}, {Phillips}, \& {Turatto}}]{elmhamdi04}
1742: {Elmhamdi}, A., {Danziger}, I.~J., {Cappellaro}, E., {Della Valle}, M.,
1743: {Gouiffes}, C., {Phillips}, M.~M., \& {Turatto}, M. 2004, \aap, 426, 963
1744:
1745: \bibitem[{{Ennis} {et~al.}(2006){Ennis}, {Rudnick}, {Reach}, {Smith}, {Rho},
1746: {DeLaney}, {Gomez}, \& {Kozasa}}]{ennis06}
1747: {Ennis}, J.~A., {Rudnick}, L., {Reach}, W.~T., {Smith}, J.~D., {Rho}, J.,
1748: {DeLaney}, T., {Gomez}, H., \& {Kozasa}, T. 2006, \apj, 652, 376
1749:
1750: \bibitem[{{Ercolano} {et~al.}(2007){Ercolano}, {Barlow}, \&
1751: {Sugerman}}]{ercolano07}
1752: {Ercolano}, B., {Barlow}, M.~J., \& {Sugerman}, B.~E.~K. 2007, \mnras, 375, 753
1753:
1754: \bibitem[{{Fesen} {et~al.}(2006){Fesen}, {Hammell}, {Morse}, {Chevalier},
1755: {Borkowski}, {Dopita}, {Gerardy}, {Lawrence}, {Raymond}, \& {van den
1756: Bergh}}]{fesen06}
1757: {Fesen}, R.~A., {Hammell}, M.~C., {Morse}, J., {Chevalier}, R.~A., {Borkowski},
1758: K.~J., {Dopita}, M.~A., {Gerardy}, C.~L., {Lawrence}, S.~S., {Raymond},
1759: J.~C., \& {van den Bergh}, S. 2006, \apj, 636, 859
1760:
1761: \bibitem[{{Finkelstein} {et~al.}(2006){Finkelstein}, {Morse}, {Green},
1762: {Linsky}, {Shull}, {Snow}, {Stocke}, {Brownsberger}, {Ebbets}, {Wilkinson},
1763: {Heap}, {Leitherer}, {Savage}, {Siegmund}, \& {Stern}}]{finkelstein06}
1764: {Finkelstein}, S.~L., {Morse}, J.~A., {Green}, J.~C., {Linsky}, J.~L., {Shull},
1765: J.~M., {Snow}, T.~P., {Stocke}, J.~T., {Brownsberger}, K.~R., {Ebbets},
1766: D.~C., {Wilkinson}, E., {Heap}, S.~R., {Leitherer}, C., {Savage}, B.~D.,
1767: {Siegmund}, O.~H., \& {Stern}, A. 2006, \apj, 641, 919
1768:
1769: \bibitem[{{Flanagan} {et~al.}(2004){Flanagan}, {Canizares}, {Dewey}, {Houck},
1770: {Fredericks}, {Schattenburg}, {Markert}, \& {Davis}}]{flanagan04}
1771: {Flanagan}, K.~A., {Canizares}, C.~R., {Dewey}, D., {Houck}, J.~C.,
1772: {Fredericks}, A.~C., {Schattenburg}, M.~L., {Markert}, T.~H., \& {Davis},
1773: D.~S. 2004, \apj, 605, 230
1774:
1775: \bibitem[{{Gaetz} {et~al.}(2000){Gaetz}, {Butt}, {Edgar}, {Eriksen},
1776: {Plucinsky}, {Schlegel}, \& {Smith}}]{gaetz00}
1777: {Gaetz}, T.~J., {Butt}, Y.~M., {Edgar}, R.~J., {Eriksen}, K.~A., {Plucinsky},
1778: P.~P., {Schlegel}, E.~M., \& {Smith}, R.~K. 2000, \apjl, 534, L47
1779:
1780: \bibitem[{{Hayashi} {et~al.}(1994){Hayashi}, {Koyama}, {Ozaki}, {Miyata},
1781: {Tsumeni}, {Hughes}, \& {Petre}}]{hayashi94}
1782: {Hayashi}, I., {Koyama}, K., {Ozaki}, M., {Miyata}, E., {Tsumeni}, H.,
1783: {Hughes}, J.~P., \& {Petre}, R. 1994, \pasj, 46, L121
1784:
1785: \bibitem[{{Hirashita} {et~al.}(2005){Hirashita}, {Nozawa}, {Kozasa}, {Ishii},
1786: \& {Takeuchi}}]{hirashita05}
1787: {Hirashita}, H., {Nozawa}, T., {Kozasa}, T., {Ishii}, T.~T., \& {Takeuchi},
1788: T.~T. 2005, \mnras, 357, 1077
1789:
1790: \bibitem[{{Hughes} {et~al.}(2000){Hughes}, {Rakowski}, \&
1791: {Decourchelle}}]{hughes00}
1792: {Hughes}, J.~P., {Rakowski}, C.~E., \& {Decourchelle}, A. 2000, \apjl, 543, L61
1793:
1794: \bibitem[{{Hwang} {et~al.}(2004){Hwang}, {Laming}, {Badenes}, {Berendse},
1795: {Blondin}, {Cioffi}, {DeLaney}, {Dewey}, {Fesen}, {Flanagan}, {Fryer},
1796: {Ghavamian}, {Hughes}, {Morse}, {Plucinsky}, {Petre}, {Pohl}, {Rudnick},
1797: {Sankrit}, {Slane}, {Smith}, {Vink}, \& {Warren}}]{hwang04}
1798: {Hwang}, U., {Laming}, J.~M., {Badenes}, C., {Berendse}, F., {Blondin}, J.,
1799: {Cioffi}, D., {DeLaney}, T., {Dewey}, D., {Fesen}, R., {Flanagan}, K.~A.,
1800: {Fryer}, C.~L., {Ghavamian}, P., {Hughes}, J.~P., {Morse}, J.~A.,
1801: {Plucinsky}, P.~P., {Petre}, R., {Pohl}, M., {Rudnick}, L., {Sankrit}, R.,
1802: {Slane}, P.~O., {Smith}, R.~K., {Vink}, J., \& {Warren}, J.~S. 2004, \apjl,
1803: 615, L117
1804:
1805: \bibitem[{{J{\"a}ger} {et~al.}(2003){J{\"a}ger}, {Dorschner}, {Mutschke},
1806: {Posch}, \& {Henning}}]{jaeger03}
1807: {J{\"a}ger}, C., {Dorschner}, J., {Mutschke}, H., {Posch}, T., \& {Henning}, T.
1808: 2003, \aap, 408, 193
1809:
1810: \bibitem[{{Kozasa} {et~al.}(1989){Kozasa}, {Hasegawa}, \& {Nomoto}}]{kozasa89}
1811: {Kozasa}, T., {Hasegawa}, H., \& {Nomoto}, K. 1989, \apj, 344, 325
1812:
1813: \bibitem[{{Krause} {et~al.}(2004){Krause}, {Birkmann}, {Rieke}, {Lemke},
1814: {Klaas}, {Hines}, \& {Gordon}}]{krause04}
1815: {Krause}, O., {Birkmann}, S.~M., {Rieke}, G.~H., {Lemke}, D., {Klaas}, U.,
1816: {Hines}, D.~C., \& {Gordon}, K.~D. 2004, \nat, 432, 596
1817:
1818: \bibitem[{{Krist}(1995)}]{krist95}
1819: {Krist}, J. 1995, in Astronomical Society of the Pacific Conference Series,
1820: Vol.~77, Astronomical Data Analysis Software and Systems IV, ed. R.~A.
1821: {Shaw}, H.~E. {Payne}, \& J.~J.~E. {Hayes}, 349--+
1822:
1823: \bibitem[{{Laor} \& {Draine}(1993)}]{laor93}
1824: {Laor}, A. \& {Draine}, B.~T. 1993, \apj, 402, 441
1825:
1826: \bibitem[{{Leroy} {et~al.}(2007){Leroy}, {Bolatto}, {Stanimirovi\'c}, {Mizuno},
1827: {Israel}, \& {Bot}}]{leroy07}
1828: {Leroy}, A., {Bolatto}, A., {Stanimirovi\'c}, S., {Mizuno}, N., {Israel}, F.,
1829: \& {Bot}, C. 2007, \apj, 658, 1027
1830:
1831: \bibitem[{{Lynch} \& {Hunter}(1991)}]{lynch91}
1832: {Lynch}, D.~W. \& {Hunter}, W.~R. 1991, {Handbook of Optical Constants of
1833: Solids II}, ed. E.~D. {Palik} (San Diego: Academic Press), 385--396
1834:
1835: \bibitem[{{Maiolino} {et~al.}(2004){Maiolino}, {Schneider}, {Oliva}, {Bianchi},
1836: {Ferrara}, {Mannucci}, {Pedani}, \& {Roca Sogorb}}]{maiolino04}
1837: {Maiolino}, R., {Schneider}, R., {Oliva}, E., {Bianchi}, S., {Ferrara}, A.,
1838: {Mannucci}, F., {Pedani}, M., \& {Roca Sogorb}, M. 2004, \nat, 431, 533
1839:
1840: \bibitem[{{Meikle} {et~al.}(2007){Meikle}, {Mattila}, {Pastorello}, {Gerardy},
1841: {Kotak}, {Sollerman}, {Van Dyk}, {Farrah}, {Filippenko}, {H{\"o}flich},
1842: {Lundqvist}, {Pozzo}, \& {Wheeler}}]{meikle07}
1843: {Meikle}, W.~P.~S., {Mattila}, S., {Pastorello}, A., {Gerardy}, C.~L., {Kotak},
1844: R., {Sollerman}, J., {Van Dyk}, S.~D., {Farrah}, D., {Filippenko}, A.~V.,
1845: {H{\"o}flich}, P., {Lundqvist}, P., {Pozzo}, M., \& {Wheeler}, J.~C. 2007,
1846: \apj, 665, 608
1847:
1848: \bibitem[{{Morgan} {et~al.}(2003){Morgan}, {Dunne}, {Eales}, {Ivison}, \&
1849: {Edmunds}}]{morgan03}
1850: {Morgan}, H.~L., {Dunne}, L., {Eales}, S.~A., {Ivison}, R.~J., \& {Edmunds},
1851: M.~G. 2003, \apjl, 597, L33
1852:
1853: \bibitem[{{Moseley} {et~al.}(1989){Moseley}, {Dwek}, {Glaccum}, {Graham}, \&
1854: {Loewenstein}}]{moseley89}
1855: {Moseley}, S.~H., {Dwek}, E., {Glaccum}, W., {Graham}, J.~R., \& {Loewenstein},
1856: R.~F. 1989, \nat, 340, 697
1857:
1858: \bibitem[{{Nozawa} {et~al.}(2007){Nozawa}, {Kozasa}, {Habe}, {Dwek}, {Umeda},
1859: {Tominaga}, {Maeda}, \& {Nomoto}}]{nozawa07}
1860: {Nozawa}, T., {Kozasa}, T., {Habe}, A., {Dwek}, E., {Umeda}, H., {Tominaga},
1861: N., {Maeda}, K., \& {Nomoto}, K. 2007, \apj, 666, 955
1862:
1863: \bibitem[{{Nozawa} {et~al.}(2008){Nozawa}, {Kozasa}, {Tominaga}, {Sakon},
1864: {Tanaka}, {Suzuki}, {Nomoto}, {Maeda}, {Umeda}, {Limongi}, \&
1865: {Onaka}}]{nozawa08}
1866: {Nozawa}, T., {Kozasa}, T., {Tominaga}, N., {Sakon}, I., {Tanaka}, M.,
1867: {Suzuki}, T., {Nomoto}, K., {Maeda}, K., {Umeda}, H., {Limongi}, M., \&
1868: {Onaka}, T. 2008, \apj, 684, 1343
1869:
1870: \bibitem[{{Nozawa} {et~al.}(2003){Nozawa}, {Kozasa}, {Umeda}, {Maeda}, \&
1871: {Nomoto}}]{nozawa03}
1872: {Nozawa}, T., {Kozasa}, T., {Umeda}, H., {Maeda}, K., \& {Nomoto}, K. 2003,
1873: \apj, 598, 785
1874:
1875: \bibitem[{{Omont} {et~al.}(2001){Omont}, {Cox}, {Bertoldi}, {McMahon},
1876: {Carilli}, \& {Isaak}}]{omont01}
1877: {Omont}, A., {Cox}, P., {Bertoldi}, F., {McMahon}, R.~G., {Carilli}, C., \&
1878: {Isaak}, K.~G. 2001, \aap, 374, 371
1879:
1880: \bibitem[{{Pastorello} {et~al.}(2007){Pastorello}, {Smartt}, {Mattila},
1881: {Eldridge}, {Young}, {Itagaki}, {Yamaoka}, {Navasardyan}, {Valenti}, {Patat},
1882: {Agnoletto}, {Augusteijn}, {Benetti}, {Cappellaro}, {Boles}, {Bonnet-Bidaud},
1883: {Botticella}, {Bufano}, {Cao}, {Deng}, {Dennefeld}, {Elias-Rosa},
1884: {Harutyunyan}, {Keenan}, {Iijima}, {Lorenzi}, {Mazzali}, {Meng}, {Nakano},
1885: {Nielsen}, {Smoker}, {Stanishev}, {Turatto}, {Xu}, \&
1886: {Zampieri}}]{pastorello07}
1887: {Pastorello}, A., {Smartt}, S.~J., {Mattila}, S., {Eldridge}, J.~J., {Young},
1888: D., {Itagaki}, K., {Yamaoka}, H., {Navasardyan}, H., {Valenti}, S., {Patat},
1889: F., {Agnoletto}, I., {Augusteijn}, T., {Benetti}, S., {Cappellaro}, E.,
1890: {Boles}, T., {Bonnet-Bidaud}, J.-M., {Botticella}, M.~T., {Bufano}, F.,
1891: {Cao}, C., {Deng}, J., {Dennefeld}, M., {Elias-Rosa}, N., {Harutyunyan}, A.,
1892: {Keenan}, F.~P., {Iijima}, T., {Lorenzi}, V., {Mazzali}, P.~A., {Meng}, X.,
1893: {Nakano}, S., {Nielsen}, T.~B., {Smoker}, J.~V., {Stanishev}, V., {Turatto},
1894: M., {Xu}, D., \& {Zampieri}, L. 2007, \nat, 447, 829
1895:
1896: \bibitem[{{Philipp}(1985)}]{philipp85}
1897: {Philipp}, H.~R. 1985, {Handbook of Optical Constants of Solids}, ed. E.~D.
1898: {Palik} (San Diego: Academic Press), 749--763
1899:
1900: \bibitem[{{Piller}(1985)}]{piller85}
1901: {Piller}, H. 1985, {Handbook of Optical Constants of Solids}, ed. E.~D. {Palik}
1902: (San Diego: Academic Press), 571--586
1903:
1904: \bibitem[{{Pozzo} {et~al.}(2004){Pozzo}, {Meikle}, {Fassia}, {Geballe},
1905: {Lundqvist}, {Chugai}, \& {Sollerman}}]{pozzo04}
1906: {Pozzo}, M., {Meikle}, W.~P.~S., {Fassia}, A., {Geballe}, T., {Lundqvist}, P.,
1907: {Chugai}, N.~N., \& {Sollerman}, J. 2004, \mnras, 352, 457
1908:
1909: \bibitem[{{Rasmussen} {et~al.}(2001){Rasmussen}, {Behar}, {Kahn}, {den Herder},
1910: \& {van der Heyden}}]{rasmussen01}
1911: {Rasmussen}, A.~P., {Behar}, E., {Kahn}, S.~M., {den Herder}, J.~W., \& {van
1912: der Heyden}, K. 2001, \aap, 365, L231
1913:
1914: \bibitem[{{Rho} {et~al.}(2008){Rho}, {Kozasa}, {Reach}, {Smith}, {Rudnick},
1915: {DeLaney}, {Ennis}, {Gomez}, \& {Tappe}}]{rho08}
1916: {Rho}, J., {Kozasa}, T., {Reach}, W.~T., {Smith}, J.~D., {Rudnick}, L.,
1917: {DeLaney}, T., {Ennis}, J.~A., {Gomez}, H., \& {Tappe}, A. 2008, \apj, 673,
1918: 271
1919:
1920: \bibitem[{{Roche} {et~al.}(1993){Roche}, {Aitken}, \& {Smith}}]{roche93}
1921: {Roche}, P.~F., {Aitken}, D.~K., \& {Smith}, C.~H. 1993, \mnras, 261, 522
1922:
1923: \bibitem[{{Roessler} \& {Huffman}(1991)}]{roessler91}
1924: {Roessler}, D.~M. \& {Huffman}, D.~R. 1991, {Handbook of Optical Constants of
1925: Solids II}, ed. E.~D. {Palik} (San Diego: Academic Press), 919
1926:
1927: \bibitem[{{Rouleau} \& {Martin}(1991)}]{rouleau91}
1928: {Rouleau}, F. \& {Martin}, P.~G. 1991, \apj, 377, 526
1929:
1930: \bibitem[{{Sasaki} {et~al.}(2006){Sasaki}, {Gaetz}, {Blair}, {Edgar}, {Morse},
1931: {Plucinsky}, \& {Smith}}]{sasaki06}
1932: {Sasaki}, M., {Gaetz}, T.~J., {Blair}, W.~P., {Edgar}, R.~J., {Morse}, J.~A.,
1933: {Plucinsky}, P.~P., \& {Smith}, R.~K. 2006, \apj, 642, 260
1934:
1935: \bibitem[{{Sasaki} {et~al.}(2001){Sasaki}, {Stadlbauer}, {Haberl},
1936: {Filipovi{\'c}}, \& {Bennie}}]{sasaki01}
1937: {Sasaki}, M., {Stadlbauer}, T.~F.~X., {Haberl}, F., {Filipovi{\'c}}, M.~D., \&
1938: {Bennie}, P.~J. 2001, \aap, 365, L237
1939:
1940: \bibitem[{{Smith} {et~al.}(2007){Smith}, {Armus}, {Dale}, {Roussel}, {Sheth},
1941: {Buckalew}, {Jarrett}, {Helou}, \& {Kennicutt}}]{smith07}
1942: {Smith}, J.~D.~T., {Armus}, L., {Dale}, D.~A., {Roussel}, H., {Sheth}, K.,
1943: {Buckalew}, B.~A., {Jarrett}, T.~H., {Helou}, G., \& {Kennicutt}, Jr., R.~C.
1944: 2007, \pasp, 119, 1133
1945:
1946: \bibitem[{{Smith} {et~al.}(2008){Smith}, {Foley}, \& {Filippenko}}]{smith08}
1947: {Smith}, N., {Foley}, R.~J., \& {Filippenko}, A.~V. 2008, \apj, 680, 568
1948:
1949: \bibitem[{{Stanimirovi{\'c}} {et~al.}(2005){Stanimirovi{\'c}}, {Bolatto},
1950: {Sandstrom}, {Leroy}, {Simon}, {Gaensler}, {Shah}, \&
1951: {Jackson}}]{stanimirovic05}
1952: {Stanimirovi{\'c}}, S., {Bolatto}, A.~D., {Sandstrom}, K., {Leroy}, A.~K.,
1953: {Simon}, J.~D., {Gaensler}, B.~M., {Shah}, R.~Y., \& {Jackson}, J.~M. 2005,
1954: \apjl, 632, L103
1955:
1956: \bibitem[{{Sugerman} {et~al.}(2006){Sugerman}, {Ercolano}, {Barlow}, {Tielens},
1957: {Clayton}, {Zijlstra}, {Meixner}, {Speck}, {Gledhill}, {Panagia}, {Cohen},
1958: {Gordon}, {Meyer}, {Fabbri}, {Bowey}, {Welch}, {Regan}, \&
1959: {Kennicutt}}]{sugerman06}
1960: {Sugerman}, B.~E.~K., {Ercolano}, B., {Barlow}, M.~J., {Tielens}, A.~G.~G.~M.,
1961: {Clayton}, G.~C., {Zijlstra}, A.~A., {Meixner}, M., {Speck}, A., {Gledhill},
1962: T.~M., {Panagia}, N., {Cohen}, M., {Gordon}, K.~D., {Meyer}, M., {Fabbri},
1963: J., {Bowey}, J.~E., {Welch}, D.~L., {Regan}, M.~W., \& {Kennicutt}, R.~C.
1964: 2006, Science, 313, 196
1965:
1966: \bibitem[{{Sutherland} \& {Dopita}(1995)}]{sutherland95}
1967: {Sutherland}, R.~S. \& {Dopita}, M.~A. 1995, \apj, 439, 381
1968:
1969: \bibitem[{{Todini} \& {Ferrara}(2001)}]{todini01}
1970: {Todini}, P. \& {Ferrara}, A. 2001, \mnras, 325, 726
1971:
1972: \bibitem[{{Tuohy} \& {Dopita}(1983)}]{tuohy83}
1973: {Tuohy}, I.~R. \& {Dopita}, M.~A. 1983, \apjl, 268, L11
1974:
1975: \bibitem[{{van Loon} {et~al.}(2005){van Loon}, {Cioni}, {Zijlstra}, \&
1976: {Loup}}]{vanloon05}
1977: {van Loon}, J.~T., {Cioni}, M.-R.~L., {Zijlstra}, A.~A., \& {Loup}, C. 2005,
1978: \aap, 438, 273
1979:
1980: \bibitem[{{van Loon} {et~al.}(2008){van Loon}, {Cohen}, {Oliveira}, {Matsuura},
1981: {McDonald}, {Sloan}, {Wood}, \& {Zijlstra}}]{vanloon08}
1982: {van Loon}, J.~T., {Cohen}, M., {Oliveira}, J.~M., {Matsuura}, M., {McDonald},
1983: I., {Sloan}, G.~C., {Wood}, P.~R., \& {Zijlstra}, A.~A. 2008, \aap, 487, 1055
1984:
1985: \bibitem[{{Williams} {et~al.}(2006){Williams}, {Borkowski}, {Reynolds},
1986: {Blair}, {Ghavamian}, {Hendrick}, {Long}, {Points}, {Raymond}, {Sankrit},
1987: {Smith}, \& {Winkler}}]{williams06}
1988: {Williams}, B.~J., {Borkowski}, K.~J., {Reynolds}, S.~P., {Blair}, W.~P.,
1989: {Ghavamian}, P., {Hendrick}, S.~P., {Long}, K.~S., {Points}, S., {Raymond},
1990: J.~C., {Sankrit}, R., {Smith}, R.~C., \& {Winkler}, P.~F. 2006, \apjl, 652,
1991: L33
1992:
1993: \bibitem[{{Wooden} {et~al.}(1993){Wooden}, {Rank}, {Bregman}, {Witteborn},
1994: {Tielens}, {Cohen}, {Pinto}, \& {Axelrod}}]{wooden93}
1995: {Wooden}, D.~H., {Rank}, D.~M., {Bregman}, J.~D., {Witteborn}, F.~C.,
1996: {Tielens}, A.~G.~G.~M., {Cohen}, M., {Pinto}, P.~A., \& {Axelrod}, T.~S.
1997: 1993, \apjs, 88, 477
1998:
1999: \end{thebibliography}
2000:
2001: \end{document}
2002: