1: \documentclass{emulateapj}

2: %\documentclass[12pt,preprint]{aastex}

3: \usepackage{apjfonts}

4: \usepackage{epsfig}

5: \usepackage{natbib}

6: 

7: %\voffset=0.03in

8: 

9: % My latex definitions

10: \newcommand{\chisq}{\ensuremath{\chi^2}}

11: \newcommand{\msun}{\ensuremath{M_{\odot}}}

12: 

13: 

14: \slugcomment{To Appear in {\it The Astrophysical Journal}}

15: \shorttitle{X-rays and Dust Emission from Kes 75}

16: \shortauthors{Morton et al.}

17: 

18: \begin{document}

19: 

20: \title{Observations of X-rays and Thermal Dust Emission from the Supernova 

21: Remnant Kes~75}

22: 

23: \author{Timothy D. Morton\altaffilmark{1}, Patrick Slane\altaffilmark{2},

24: Kazimierz J. Borkowski\altaffilmark{3}, Stephen P.

25: Reynolds\altaffilmark{3}, David J. Helfand\altaffilmark{4}, 

26: B.~M.~Gaensler\altaffilmark{5,2}, and John~P.~Hughes\altaffilmark{6}}

27: \altaffiltext{1}{Harvard College, Cambridge, MA 02138.}

28: \altaffiltext{2}{Harvard-Smithsonian Center for Astrophysics, 60 Garden

29: Street, Cambridge, MA 02138.}

30: \altaffiltext{3}{Department of Physics, North Carolina State University, Box

31: 8202, Raleigh, NC 27695-8202.}

32: \altaffiltext{4}{Columbia Astrophysics Laboratory, Columbia University, 550

33: West 120th Street, New York, NY 10027.}

34: \altaffiltext{5}{School of Physics A29, The University of 

35: Sydney, NSW 2006, Australia.}

36: \altaffiltext{6}{Department of Physics and Astronomy, Rutgers University, 136

37: Frelinghuysen Road, Piscataway, NJ 08854.}

38: 

39: \begin{abstract}

40: 

41: We present {\sl Spitzer Space Telescope} and {\sl Chandra X-ray

42: Observatory} observations of the composite Galactic supernova remnant

43: Kes 75 (G29.7$-$0.3). We use the detected flux at 24 $\micron$ and hot

44: gas parameters from fitting spectra from new, deep X-ray observations

45: to constrain models of dust emission, obtaining a dust-to-gas mass

46: ratio $M_{dust}$/$M_{gas}\sim$10$^{-3}$. We find that a two-component

47: thermal model, nominally representing shocked swept-up interstellar or

48: circumstellar material and reverse-shocked ejecta, adequately fits the

49: X-ray spectrum, albeit with somewhat high implied densities for both

50: components. We surmise that this model implies a Wolf-Rayet progenitor

51: for the remnant. We also present infrared flux upper limits for the

52: central pulsar wind nebula.

53: 

54: \end{abstract}

55: 

56: \keywords{dust, extinction -- ISM: individual (SNR G29.7-0.3) -- supernova 

57: remnants -- X-rays: ISM}

58: 

59: \section{Introduction}

60: 

61: The Galactic supernova remnant (SNR) Kes~75, together with its associated

62: pulsar (PSR J1846-0258) and pulsar wind nebula (PWN), is a prototypical

63: example of a composite supernova remnant. Located at a distance of $d

64: \sim 19 d_{19}$~kpc (based on neutral hydrogen absorption measurements

65: -- Becker \& Helfand 1984) the pulsar luminosity is extremely high --

66: second only to the Crab for Galactic pulsars. PSR J1846-0258 also has an

67: exceptionally strong dipole magnetic field ($\sim 5\times 10^{13}$~G),

68: as inferred from the pulsar spin properties (Gotthelf et al.~2000),

69: and the X-ray luminosity from the associated PWN is a remarkably high

70: fraction of the pulsar's total spindown energy loss (20\%, compared to

71: a typical value of $\sim0.1$\%)\footnote{

72: $L_{x,psr}(0.5-10 {\rm\ keV}) = 4.4 \times 10^{35}{\rm\ ergs\ s}^{-1}$; 

73: $L_{x,neb}(0.5-10 {\rm\ keV}) = 1.7 \times

74: 10^{36} d_{19}^2 {\rm\ ergs\ s}^{-1}$ (H03); $\dot E = 8.3 \times

75: 10^{36} {\rm\ ergs\ s}^{-1}$ (Livingstone et al. 2006)}. The

76: characteristic age of the pulsar is extremely young -- only 723 years

77: (Gotthelf et al.~2000) -- and recent timing measurements of the braking

78: index, n, have yielded an estimated upper limit to its true age of only 884

79: years (assuming n is constant; Livingstone et al.~2006), placing it among 

80: the youngest known rotation-powered pulsars in the Galaxy. The

81: remnant is also quite large ($\sim 10$~pc in radius), implying an

82: average expansion velocity of $\sim$10$^4$ km/s -- typical (or even

83: in excess) of values for undecelerated SNR expansion.

84: 

85: Kes~75 has been studied in considerable detail in both the radio and X-ray

86: bands, but high extinction due to its large distance and location near

87: the Galactic plane prevent its detection at optical wavelengths.  In the

88: radio, a synchrotron shell is observed, with a flux density of 10 Jy at

89: 1 GHz, and spectral index $\alpha = 0.7$ (energy flux $S_{\nu} \propto

90: \nu^{-\alpha}$), as well as flatter ($\alpha \sim 0.25$) emission from the

91: PWN (Becker \& Kundu 1976).  Observations with the {\em ASCA} observatory

92: reveal distinct spectral components from the PWN and the remnant shell

93: (Blanton \& Helfand 1996), although the SNR plasma parameters are poorly

94: determined due, in part, to mixing of these two spectral components as a

95: result of the poor angular resolution.  Similar morphology to that seen

96: in the radio band has been detected in the X-rays (Helfand et al. 2003 -

97: hereafter H03), with spatial coincidence of the thermal X-ray emission

98: and radio shell (see Figure 1), and an axisymmetric PWN structure with a

99: photon index of $\Gamma = 1.92$ (photon number flux $F_{\gamma} \propto

100: \nu^{-\Gamma}$).

101: 

102: While radio and X-ray observations provide information regarding the age,

103: energetics, and ambient conditions for SNRs, infrared observations of

104: remnants are also instructive because they can reveal the presence of

105: heated dust, both shock-heated swept-up dust from the circumstellar

106: or interstellar medium (CSM/ISM), and potentially, dust created by

107: the supernova itself. The Galactic Legacy Infrared Mid-Plane Survey

108: Extraordinaire (GLIMPSE; Benjamin et al. 2003), which uses all four bands

109: of the Infrared Array Camera (IRAC) on the {\sl Spitzer Space Telescope}

110: ($SST$; 3.6, 4.5, 5.8, and 8 $\micron$), encompassed $\sim 100$ known

111: radio/X-ray SNRs, about 10-15\% of which were detected with confidence

112: (Lee et al. 2005, Reach et al. 2006). Kes~75 was not detected in the

113: GLIMPSE data, nor in earlier infrared SNR surveys (e.g. Arendt 1989,

114: Saken et al. 1992).

115: 

116: As part of a general {\sl SST} survey of LMC SNRs, Borkowski et al.~(2006:

117: B06) and Williams et al.~(2006a; W06) reported detection of four Type

118: Ia remnants and four core-collapse remnants, respectively, in the 24

119: $\mu$m and 70 $\mu$m MIPS (Multiband Imaging Photometer for Spitzer)

120: bands, with stringent upper limits on emission from the IRAC bands.

121: Using these observations in connection with dust emission models, they

122: find evidence of substantial dust destruction in the blast waves of both

123: types of remnant: 30\% -- 40\% of the total dust mass, and as much as

124: 90\% of the mass in grains smaller than about 0.04 $\mu$m. They find

125: no evidence for emission associated with ejecta. Both studies report

126: pre-shocked dust/gas mass ratios lower by a factor of several than are

127: generally assumed for the LMC.

128: 

129: In order to explore evolutionary scenarios and investigate the dust

130: content in Kes~75, we present X-ray spectral fits for the remnant from

131: new $Chandra$ observations, as well as the first infrared detection of

132: this remnant from new $SST$ observations. These infrared observations have

133: high enough angular resolution that, in addition to determining properties

134: of the remnant shell, we are also able to place constraints on infrared

135: emission from the central PWN in order to investigate the need for the

136: spectral break that most PWNe exhibit between the radio and X-ray bands.

137: 

138: %%%%%%%%%%%%%%%%%%%%%%%%%% Figure 1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

139: \begin{figure}[t]

140: \includegraphics*[width=3.3in]{f1.eps}

141: %\includegraphics*[width=3.39in]{f1.eps}

142: %\epsscale{0.4}

143: %\plotone{f1.eps}

144: \caption{

145: Top: VLA image of Kes~75 at 1.4~GHz from the MAGPIS project (Helfand et al.

146: 2006). The greyscale is linear, covering a range of 0 to 0.1 Jy~beam$^{-1}$.

147: Middle: 24 $\micron$ MIPS image of Kes~75. X-ray contours are overlaid

148: to highlight the similar morphology. The PWN is not detected in the

149: infrared, nor is any part of the shell of Kes~75 detected in any of the

150: IRAC bands. The greyscale is logarithmic, from 100 to 200 MJy~sr$^{-1}$.

151: The bright, circular sources in the image, as well as numerous fainter

152: sources, are consistent with point sources along the line of sight to

153: Kes~75.

154: Bottom: $Chandra$ X-ray image of Kes~75. Intensity is displayed

155: logarithmically from 10 to 140 cnts~arcsec$^{-2}$. Extraction regions

156: for the shell and background are shown.

157: }

158: \end{figure}

159: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

160: 

161: 

162: \section{Observations and Data Reduction}

163: 

164: X-ray observations of Kes~75 were carried out in October 2000, for

165: a total of $\sim 39$~ks, and in June 2006, over four pointings with

166: a total exposure time of $\sim 156$~ks. Observations were made with

167: the Advanced CCD Imaging Spectrometer ($ACIS$) instrument on board the

168: $Chandra$ $X$-$ray$ $Observatory$ ($CXO$), providing a spatial resolution

169: of 0.5\arcsec over the range 0.5-10 keV. All data were re-processed

170: using the most recent calibration files, and cleaned for episodes of

171: high background. The total good exposure time for the five pointings was

172: $\sim 188$~ks. Figure 1 (bottom) shows the X-ray image displayed with

173: a logarithmic intensity scale, showing the both the central PWN and the

174: remnant shell.

175: 

176: Infrared observations were made on September 15, 2005, using both the

177: IRAC and MIPS cameras on the $SST$, and were processed and mosaiced

178: using standard pipelines. Images were obtained at 3.6, 4.5, 5.8, 8, and

179: 24 $\micron$, the shorter wavelength images from the IRAC instrument

180: and the 24 $\micron$ image from MIPS. The pixel size is 1.2$\arcsec$

181: for the IRAC bands and 2.45$\arcsec$ for MIPS, roughly matched to the

182: (wavelength-dependent) diffraction-limited point spread function of the

183: telescope. For each of the IRAC bands, we obtained a 36-point dither

184: pattern of pointings at the source, with a 30 s integration time per

185: pointing, yielding a total exposure of 1080~s. For the MIPS data, we

186: obtained a total of 4.5 ks using a series of 30 s integrations.

187: 

188: The remnant is not detected in any of the IRAC bands, but the two limbs

189: of the shell apparent in the radio and X-ray images are clearly visible

190: at 24 $\micron$. Figure 1 (middle) shows the 24 $\micron$ image, overlaid

191: with X-ray contours to highlight the similarities in morphology. Because

192: Kes~75 lies in the Galactic plane, there is considerable contamination

193: from foreground and background emission. The central PWN is not detected

194: in any of the infrared bands.

195: 

196: To extract X-ray spectra, we used the `specextract' routine in the

197: CIAO software package, grouping to obtain a minimum of 25 counts/bin.

198: We extracted individual spectra for each observation segment using

199: extraction regions corresponding to the southeast (SE) and southwest

200: (SW) limbs and to the PWN, with a background region outside the shell

201: (Figure 1, bottom). We used these same regions to extract infrared

202: fluxes from the $SST$ images. X-ray spectral fitting (discussed in

203: greater detail below) was done using the XSPEC software package.

204: 

205: 

206: \section{Analysis}

207: 

208: To characterize the X-ray spectrum of the remnant shell, we fit the data

209: to a non-equilibrium ionization (NEI) collisional plasma plane-shock

210: model (an updated version of the XSPEC model `vpshock') with foreground

211: absorption. We fit the spectrum of each limb independently, jointly

212: fitting spectra from each of the observation segments. We find significant

213: excess at the high energy end of the spectrum, consistent with previous

214: fits of X-ray data for Kes~75 (H03). H03 discuss possible origins of

215: this hard tail, such as cosmic-ray acceleration at the shock front or

216: foreground dust-scattering of emission from the PWN, and add a power law

217: component to the model to account for this emission. We find that even

218: with the addition of a power law, the model fits to the spectra are poor

219: (with a reduced chi-squared $\chi_r^2 > 1.6$), the primary problem being

220: significant residuals around prominent emission lines, and in a broad

221: band near 1~keV.

222: 

223: As an alternative to the power law component, we investigated fits with a

224: second, higher temperature thermal component. As we describe below, this

225: two-component thermal model yields significantly better fits, with the

226: high-temperature component accommodating the hard emission. This component

227: requires enhanced abundances of Si, S, Ar, and Fe, suggesting a scenario

228: in which the low-temperature component corresponds to forward-shocked

229: circumstellar/interstellar material while the high-temperature component

230: originates from reverse-shocked ejecta.

231: 

232: %%%%%%%%%%%%%%%%%%%%%% Table 1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

233: 

234: \begin{deluxetable}{lccccc}

235: \tablecolumns{6}

236: \tabletypesize{\scriptsize}

237: \tablewidth{0pc}

238: \tablecaption{Infrared Flux Densities}

239: \tablehead{

240: \colhead{region} &

241: \colhead{3.6 $\micron$} &

242: \colhead{4.5 $\micron$} &

243: \colhead{5.8 $\micron$} &

244: \colhead{8 $\micron$} &

245: \colhead{24 $\micron$}

246: }

247: \startdata

248: Thermal shell & $<$0.04 & $<$0.05 & $<$1.92 & $<$0.72 & $18.9\pm 3.6$ \\

249: PWN & $<$0.068 & $<$0.061 & $<$0.044 & $<$0.062 & $<$0.66 \\

250: \enddata

251: \tablecomments{

252: All flux densities are extinction-corrected and quoted in Jy. The 3.6,

253: 4.5, 5.8, and 8 $\micron$ data are taken with the IRAC instrument,

254: and the 24 $\micron$ data with the MIPS camera.

255: }

256: \end{deluxetable}

257: 

258: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

259: 

260: Modeling collisionally heated dust requires the infrared data as input,

261: as well as the hot electron gas density $n_e$, electron temperature

262: $T_e$, ion temperature $T_i$, shock timescale $\tau = n_e t_{shock}$,

263: and volume emission measure $EM = n_e n_H V$. We are only able to

264: measure directly $T_e$ from the X-ray spectrum, and we assume $T_i=T_e$.

265: This assumption may be incorrect, particularly given the young age of

266: the remnant, but we note that at temperatures of about 1 keV, electron

267: collisions are expected to dominate grain heating (Dwek and Arendt 1992),

268: and so any additional effect from a slightly higher ion temperature

269: would make only a marginal contribution to an increase in emission

270: from the dust. All these parameters, together with standard grain size

271: and composition distribution assumptions (Weingartner \& Draine 2001),

272: give a predicted infrared emission spectrum shape, the normalization of

273: which can be adjusted to fit the observed infrared flux. This, in turn,

274: gives a value of the dust-to-gas mass ratio. With this in mind, we note

275: that if ion temperatures are higher than electron temperatures, we will

276: overestimate the amount of dust present, since higher temperatures would

277: require less dust to produce the same emission at a given wavelength.

278: The reader is referred to B06 for a detailed description of the dust

279: modeling.

280: 

281: B06 also used the measured infrared flux values and the 70/24 $\micron$

282: flux ratio to further constrain these models and to make a prediction

283: about dust destruction (since a larger-than-expected 70/24 $\micron$

284: ratio would imply destruction of the smaller grains). However, in the

285: absence of data in the $70 \micron$ band, we had to rely on our $24

286: \micron$ detection and IRAC band upper limits for our constraints.

287: 

288: In order to extract infrared fluxes, we defined source regions consistent

289: with the emission observed in the Chandra images, excluded any obvious

290: point sources from the infrared data, and then calculated a total

291: background-subtracted flux with the ``funtools'' package in ds9, using

292: three different background regions for each infrared image. For the 24

293: $\micron$ detection, we averaged the calculated background-subtracted flux

294: for the three different background regions, while for the IRAC images,

295: we used the highest value attained from the three different backgrounds

296: to get conservative flux upper limits. PWN upper limits were obtained in

297: a similar manner. Formal errors for the infrared flux extractions were

298: determined from the width of histograms of pixel values in the region,

299: and were determined to be on the order of 10\%. The errors we quote,

300: however, are dominated by uncertainties in the background caused by

301: confusion along the line of sight, which we quantify as the largest

302: deviation from the mean of the three background-subtracted results.

303: 

304: To calculate the hot gas densities required to constrain the dust models,

305: we use the volume emission measure ($n_e n_H V$), taken directly from

306: the normalization of the XSPEC 'vpshock' model:

307: \begin{equation} K = 10^{-14}\frac{n_e n_H V}{4 \pi d^2} {\rm\

308: cm}^{-5} \end{equation} 

309: where $n_e$ is the electron density, $n_H$ is the gas density, $V$ is the

310: volume of the emitting region, and $d$ is the distance to the remnant. To

311: calculate the volume of the emitting regions, we take the area of the

312: regions indicated in Figure 1 (bottom) and assume a slab-like volume

313: whose depth along the line of sight is equal to the length of the long

314: axes of the regions. We expect this method to give an upper limit to the

315: volume estimate, so we introduce a filling factor $0<f<1$ to represent

316: the total fraction of the volume that is actually emitting; the filling

317: factors could be considerably smaller than 1 if there is significant

318: clumping in the gas. We then assume that $n_e / n_H = 1.18$, which is

319: true for a fully ionized gas with solar abundances, and that $d = 19$ kpc.

320: 

321: \section{Results}

322: 

323: \subsection{SNR Infrared flux}

324: 

325: We do not detect Kes~75 in any of the four IRAC bands, but we do find

326: a clear detection in the 24-$\micron$ MIPS band. This is consistent

327: with the B06 and W06 LMC SNR results, for which detections were only

328: found in the MIPS bands, and with the GLIMPSE SNR results which detected

329: only $\sim10$\% of known Galactic SNRs in the IRAC bands. Infrared flux

330: results are presented in Table 1, where we have corrected the observed

331: flux values using extinction corrections from Indebetouw et al. (2005).

332: 

333: 

334: %%%%%%%%%%%%%%%%%%%%%%%%%% Figure 2 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

335: \begin{figure}[t]

336: %\epsscale{0.7}

337: %\plotone{f2.eps}

338: \includegraphics*[width=3.5in]{f2.eps}

339: \caption{

340: The spectrum of the Kes~75 thermal shell (SE limb, taken from a 54~ks

341: segment of {\sl Chandra} data), with a two-component thermal model. Dashed

342: (dotted) curves show the low (high) temperature thermal model components.

343: }

344: \end{figure}

345: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

346: 

347: 

348: \subsection{X-ray spectral fitting}

349: 

350: As a starting point to spectral fitting of the shell emission from

351: Kes~75, we determined the column density of the foreground gas through

352: fits to the PWN spectrum. We extracted spectra from the five observation

353: segments using a circle centered on the pulsar, with a 7~arcsec radius,

354: but excluding emission from the pulsar itself. An absorbed power

355: law provided an excellent fit to the $\sim 64\,000$-count spectrum

356: ($\chi^2_r = 1.01$), with $\Gamma = 1.96 \pm 0.04$ and $N_H = (4.03 \pm

357: 0.07) \times 10^{22}$ cm$^{-2}$. The $\sim 40\,000$-count spectrum from

358: an annulus extending from a radius of 7~arcsec to 15 arcsec, excluding a

359: small circle around a jet-like feature, gave best-fit values of $\Gamma

360: = 2.11 \pm 0.05$ and $N_H = (4.04 \pm 0.10) \times 10^{22}$ cm$^{-2}$

361: ($\chi^2_r = 0.99$). Details of the spectrum and structure of the PWN are

362: the subject of a forthcoming paper. Based on the above fits, we adopt

363: a value of $N_H = 4 \times 10^{22}$ cm$^{-2}$ for the shell of Kes~75.

364: We note that this value is in excellent agreement with that derived by

365: H03 through fits to the PWN.

366: 

367: The shape of the high energy continuum (above 3~keV) for the SNR shell

368: is consistent with a bremsstrahlung model with $kT \sim 1.8$~keV. With

369: $N_H$ fixed as described above, extrapolating back to lower energies

370: yields a large deficit relative to the observed flux in the $1-2$~keV

371: band; an additional soft thermal component is required. We thus

372: investigated a model with two vpshock model components, one with a

373: temperature near that from the bremsstrahlung model, and one set lower

374: to accommodate the soft emission. We allowed the temperature,

375: ionization timescale, and normalization to vary for each of the thermal

376: components, and find that significant residuals remain around the lines

377: of Si, S, and Ar, as well as in a broad band near $\sim 1$~keV, where

378: Fe-L emission dominates. When the abundances of these elements are

379: allowed to vary (independently) in the high temperature component, we

380: find a much-improved fit with large enhancements above solar values. In

381: Figure 2 we plot the spectrum of the SE limb from one of the

382: observation segments. Also shown is the best-fit model, as well as each

383: individual emission component.  Fitting results are summarized in Table

384: 2, where the errors on the fit parameters represent 90\% confidence

385: intervals.\footnote{We note that H03 find similar $\chi_r^2$ values for

386: an ad hoc model of bremsstrahlung emission accompanied by gaussian

387: lines at discrete energies, but reject this model as non-physical based

388: on the lack of connection between derived temperature and the centroids

389: and relative strengths of the lines.}

390: 

391: %%%%%%%%%%%%%%%%%%%%%%%%% Table 2 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

392: 

393: \begin{deluxetable}{lcccc}

394: \tablecolumns{3}

395: \tabletypesize{\scriptsize}

396: \tablewidth{0pc}

397: \tablecaption{Spectral Fits to the Kes~75 Thermal Shell}

398: \tablehead{

399: \colhead{Parameter} &

400: \colhead{SE Rim} &

401: \colhead{SW Rim} \\

402: }

403: \startdata

404: $N_H$ ($\times$10$^{22}$ cm$^{-2}$) & 4.0 (fixed) & 4.0 (fixed) \\

405: $kT_1$ (keV) & $0.25 \pm 0.01$ & $0.22\pm^{0.02}_{0.01}$ \\

406: Abundances & solar & solar \\

407: $\tau_1$ (s cm$^{-3}$) & $>1.5\times 10^{12}$ & $>1.5\times 10^{12}$ \\

408: $F_1$\tablenotemark{a} ($\times$10$^{-10}$ ergs cm$^{-2}$ s$^{-1}$) & 4.6 & 2.2 \\

409: \\

410: $kT_2$ (keV) & $1.38\pm 0.02$ & $1.50\pm^{0.98}_{0.07}$ \\

411: $[$Si$]$\tablenotemark{b} & $3.7\pm 0.2 $ & $3.6\pm 0.5$ \\

412: $[$S$]$ & $13.5 \pm 0.7$ & $10.2\pm 1.3$ \\

413: $[$Ar$]$ & $10.5\pm 2.3$ & $6.9\pm^{3.4}_{3.2}$ \\

414: $[$Fe$]$ & $14.5\pm 1.8$ & $14.2\pm^{5.2}_{4.3}$\\

415: $\tau_2$ (s cm$^{-3}$) & $(7.0\pm 0.3) \times 10^9$ & $(6.1\pm 0.4) \times 10^{9}$\\

416: $F_2$ ($\times$10$^{-10}$ ergs cm$^{-2}$ s$^{-1}$) & 2.2 & 1.0 \\

417: \\

418: $\chi^{2}_{r}$ (dof) & 1.34 (840) & 1.31 (522)\\

419: \enddata

420: \tablenotetext{a}{X-ray fluxes are unabsorbed in the 0.5 - 10 keV band.}

421: \tablenotetext{b}{Abundances are relative to solar values.}

422: \end{deluxetable}

423: 

424: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

425: 

426: 

427: While the X-ray spectrum from an SNR is, in reality, comprised of

428: emission from a distribution of temperatures, compositions, and

429: ionization states, we broadly associate the low-temperature component

430: found here with the forward-shocked CSM/ISM, and the high-temperature,

431: metal-enriched component with the reverse-shocked ejecta. The ionization

432: timescales for the CSM/ISM component are poorly constrained on the high

433: end, and we are only able to reliably obtain lower limits. The fits are

434: remarkably similar for the two limbs, particularly given the considerable

435: difference in X-ray/IR flux ratios observed in Figure 1. The best-fit

436: model for the SE limb, which has a higher X-ray flux than the SW limb,

437: yields a slightly higher swept-up CSM/ISM temperature and a slightly

438: lower ejecta temperature, but these variations are not of high statistical

439: significance. We do not find any evidence for spectral variations within

440: the limb regions, although the reduction in the number of counts in each

441: spectrum results in a corresponding increase in the fit uncertainties,

442: limiting our ability to discern small spatial variations that might be

443: expected from the two emission components.

444: 

445: 

446: \subsection{Densities and Shock Ages}

447: 

448: As described above, we can estimate the densities of the emission

449: components from the normalization provided by the X-ray model fits along

450: with an estimate of the emitting volume (Eq.~1). Using values from Table

451: 2, we find filling-factor-dependent post-shock electron densities for both

452: the swept-up CSM/ISM component and ejecta component from each limb. The

453: values we get are fairly high, especially for the CSM/ISM component

454: (58$f_1^{-1/2}d_{19}^{-1/2}$ cm$^{-3}$ and 36$f_1^{-1/2}d_{19}^{-1/2}$

455: cm$^{-3}$ for the SE and SW limbs, respectively). In addition, even

456: with these high densities, the forward shock ages implied from the

457: measured ionization timescales ($>$2400 yrs) are significantly higher

458: than current best estimates of the pulsar age from spindown measurements

459: ($<$900 yrs). For the reverse shock, we obtain more modest densities of

460: $\sim 9.6 f_2^{-1/2}$ cm$^{-3}$ ($\sim 2.2 f_2^{-1/2}$ cm$^{-3}$) for

461: the SE (SW) region, and an age of around 75 years, which, though young,

462: is not completely implausible for recently shocked ejecta. We scale

463: quantities with different filling factors $f_1$ and $f_2$ to indicate

464: the possibility that the two emission components may occupy different

465: fractions of the total volume (but we treat these as being the same in

466: the SE and SW limbs). We note that the density estimate for the ejecta

467: component (which comprises the bulk of the observed flux, but

468: a considerable minority of the unabsorbed flux) is similar to that

469: obtained for a single thermal component by H03. It is a factor of

470: $\sim 5$ higher than that obtained by Blanton \& Helfand (1996), but

471: this is primarily because they assumed that the emission completely

472: filled a shell around the PWN (since ASCA was unable to resolve the

473: emission and show that the thermal component was primarily from the

474: much smaller SE and SW limbs). 

475: 

476: %%%%%%%%%%%%%%%%%%%%%%%%%% Figure 3 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 

477: \begin{figure}[t]

478: %\includegraphics*[width=3.2in]{f3.eps}

479: \plotone{f3.eps}

480: \caption{

481: Dereddened IR emission from the shell of Kes 75, along with the dust-emission

482: model described in the text. Inverted triangles indicate upper limits.

483: Filled symbols and the solid curve correspond to the SW shell while

484: open symbols and the dashed curve correspond to the SE shell. The model

485: is derived from the electron temperature and density implied by the

486: X-ray measurements, with the $24 \mu$m measurements setting the normalization

487: for the hydrogen density.

488: }

489: \end{figure}

490: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

491: 

492: \subsection{Dust models}

493: 

494: The measured 24 $\mu$m fluxes of 2.4 Jy and 2.8 Jy in the SE and SW

495: limbs, respectively, can be used to estimate dust masses present in

496: the shocked ambient medium. Here we treat the IR emission as arising

497: exclusively from dust. This is consistent with the X-ray results that

498: indicate a swept-up CSM/ISM mass that is a factor of $\sim 6$ larger

499: than the X-ray-emitting ejecta mass.  The high ($4 \times 10^{22}$

500: cm$^{-2}$) column density $N_H$ toward Kes 75 implies a very high

501: ($A_V \sim 22^{\rm m}$) optical extinction, and substantial extinction

502: ($A_{24 \mu{\rm m}}=1.4$) even in the MIPS 24 $\mu$m band (we used

503: an infrared extinction curve from Chiar \& Tielens 2006 valid for the

504: local diffuse ISM). The dereddened 24 $\mu$m fluxes are 8.7 Jy and 10.2

505: Jy in the SE and SW limbs, respectively, Because the shape of the IR

506: spectrum is not known, we must rely on modeling in making dust mass

507: estimates. As described in Section 3 (and, in more detail, in B06),

508: the spectral shape is determined by the grain temperatures, which depend

509: primarily on the temperatures and densities of the X-ray emitting plasma. The

510: electron temperatures are 0.25 keV and 0.22 keV in the SE and SW limbs

511: (Table 2).  A lower limit to the electron densities, 50 cm$^{-3}$, is obtained

512: from limits on $\tau$ listed in Table 2 and the SNR age $t_{SNR}$ of

513: less than 884 yr (Livingstone et al. 2006). Electron densities derived

514: from emission measures depend on the unknown filling fraction $f$; by

515: setting $f=1$ and assuming a distance of 19~kpc, we obtain comparable

516: density lower limits of 58 cm$^{-3}$ and 36 cm$^{-3}$ in SE and SW

517: limbs, respectively. The spectral energy distribution predicted from

518: these inputs, and constrained by the observed $24 \mu$m flux, is plotted

519: in Figure 3. Triangles indicate upper limits from the IRAC data, and

520: squares represent the measured flux from the MIPS data.

521: 

522: The estimated dust masses are 0.03 M$_\odot$ and 0.05 M$_\odot$ in SE

523: and SW, for plane shocks with age of 884 yr, and temperatures and density

524: upper limits just quoted. These dust masses should be considered as upper

525: limits to the actual dust masses in the limbs, as less dust is required

526: to produce the observed fluxes at higher plasma densities (which would

527: be implied for filling factors smaller than the assumed value). About

528: 40\%\ of the preshock dust mass was destroyed by sputtering in these

529: shock models; dust destruction rates are higher at higher densities.

530: In addition, IR line emission from [Fe II] or [O IV], which been observed

531: in N49 (Williams et al. 2006b) and has been suggested as a contributor

532: to the emission from 1E~0102.2-7219 (Stanimirovi\'{c} et al. 2005), may

533: contribute a small fraction of the flux from the ejecta component in

534: Kes~75. If present, this will further reduce our dust mass estimates.

535: We note that we have assumed a grain size distribution and composition

536: that is typical of the Milky Way ISM. With a detection at only one

537: wavelength in the IR, the data are not sufficient to probe potential

538: variations from these values.

539: 

540: \subsection{PWN spectrum}

541: 

542: We do not detect the central PWN in any of the $SST$ bands we observed,

543: but we list upper limits to the flux in Table 1, and we plot these (in

544: Figure 4) against current radio and X-ray data from Salter et al. (1989),

545: Bock and Gaensler (2005), and H03. We find that the IRAC upper limits do

546: not introduce stringent constraints on the spectral break frequency. While

547: a change in spectral index is required between the radio and X-ray bands,

548: with extrapolation of the radio and X-ray spectra indicating a break

549: somewhere near the IR band, the {\sl SST} data do not constrain where

550: this break occurs.

551: 

552: %%%%%%%%%%%%%%%%%%%%%%%%%% Figure 4 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 

553: \begin{figure}[t]

554: \plotone{f4.eps}

555: %\includegraphics*[width=3.2in]{f4.eps}

556: \caption{

557: The PWN spectrum. Note the spectral break necessary between the radio

558: points (from Salter et al. 1989; Bock and Gaensler 2005) and the X-ray

559: regime (Helfand et al. 2003). Our infrared upper limits are

560: consistent with both X-ray and radio data (marginally so with X-ray at

561: the 8 $\micron$ point) suggesting a spectral break in or near the infrared.

562: }

563: \end{figure}

564: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

565: 

566: 

567: \section{Discussion}

568: 

569: \subsection{The remnant's shell}

570: 

571: Using our density estimates from both limbs to calculate the total gas

572: masses, we find $\sim 94f_1^{1/2}d_{19}^{5/2}$ $\msun$ for the cool

573: forward-shock component and $\sim 15f_2^{1/2}d_{19}^{5/2}$ $\msun$

574: for the ejecta component, where we have added the contributions for the

575: two shells. The value for the CSM/ISM component is particularly high,

576: especially given the large expansion velocity implied by the inferred

577: size and age, although clumping of the gas (i.e., a smaller filling

578: factor) would yield smaller values. We also note that these estimates

579: correspond only to the observed shell segments, which occupy roughly

580: 20-30\% of the circumference of the remnant.

581: 

582: While the adopted two-component thermal model provides an adequate fit

583: to the X-ray spectrum of Kes~75, it is clearly only an approximation

584: for the forward and reverse shock emission components. We are unable

585: to constrain more complex distributions of the temperature, ionization

586: state, and abundances, particularly given the very high absorption, but

587: we believe that the distinct components identified in the spectral fits

588: provide adequate evidence to identify the presence of shocked CSM/ISM

589: material in the presence of enriched ejecta. We note, however, that

590: the high Fe abundance inferred for the ejecta component is problematic

591: in that it implies much higher amounts of Fe than expected for a very

592: massive stellar progenitor. While the uncertainties listed in Table 2

593: correspond to formal $\chi^2$-fitting errors, the enhanced Fe abundance

594: is preferred by the fits solely to reduce residuals in the broad Fe-L

595: region near $\sim 1$~keV, where the high absorption makes it particularly

596: difficult to rule out a more complex temperature distribution for the

597: forward shock component that could potentially reduce these residuals.

598: 

599: \subsection{The remnant's dust content}

600: 

601: The dust/gas mass ratios we find are $5 \times 10^{-4}$ and $1.4 \times

602: 10^{-3}$ for the SE and

603: SW limbs, respectively.  These are considerably smaller than the value

604: of $\sim 0.7\%$ normally assumed for the Galaxy. The nominal Galactic

605: value is calculated by measuring the extinction from nearby bright stars;

606: using SNRs to probe dust provides a different sample of the ISM dust

607: content in the Galaxy, potentially affected by local grain destruction

608: by the SNR shock (which may represent as much as 40\% of the dust mass

609: in the preshock material, as described in Section 4.4). We note that

610: the dust mass estimates are based on the assumption that all of the IR

611: emission results from dust. Any contribution from line emission would

612: reduce the dust/gas mass ratio even further. We also note, though, that

613: the gas mass values are derived from the low-temperature component of

614: a heavily absorbed X-ray spectrum.

615: 

616: Our low dust/gas ratio is consistent with the results reported by B06

617: and W06 in the LMC, who found discrepancies of factors of $\sim 5$

618: between their deduced dust/gas ratios and the (lower than Galactic)

619: values commonly assumed for the LMC. W06 speculate that if grains are

620: porous, smaller dust masses could provide the observed extinctions,

621: while larger masses would be required to produce the observed IR emission

622: because of more efficient dust destruction, bringing the two types of

623: estimates closer together.

624: 

625: In order to place better constraints on the dust component, additional

626: infrared observations are required.  Infrared spectra would more tightly

627: constrain the dust models and allow us to state more confidently the

628: dust composition of the remnant. A spectrum would also reveal how much,

629: if any, of the 24 $\micron$ flux might be from emission lines (such as

630: [Fe II] or [O IV]) rather than emission from heated dust.

631: 

632: 

633: \subsection{The remnant's progenitor}

634: 

635: Chevalier (2005) discusses the origins and evolution of massive star

636: core-collapse supernovae, to which class Kes~75 belongs, and notes in

637: particular that Kes~75 has characteristics suggestive of an origin from

638: a Type Ib or Type Ic event. SN Ib/c are born from Wolf-Rayet (WR)

639: stars, which are characterized by high mass-loss rates, and especially

640: high-velocity stellar winds. These WR winds are 100-200 times faster

641: than winds from earlier mass-loss stages, and so they sweep up the

642: circumstellar medium and slow-velocity wind material, and clear out a

643: bubble around the star, bounded by a dense and clumpy shell extending

644: up to $\geq$10 pc over a WR lifetime of $\sim$$2\times 10^5$~yr. Thus,

645: when the star finally collapses and produces a SN Ib/c, the ejecta

646: travel initially through very low-density gas, and then eventually

647: begin to interact with this dense WR shell.

648: 

649: Assuming a WR picture for a progenitor for Kes~75 answers several

650: questions, most prominently the large size of the remnant: if ejecta

651: are able to expand uninhibited through a low-density medium (cleared

652: out by the WR wind), then they will not sweep up mass quickly, and thus

653: will not slow down significantly. This indication, and the young shock

654: age of the reverse-shocked ejecta (implying recent shocking of the WR

655: shell), both point to this progenitor model. This bubble model might

656: also help explain the high density and clumpiness that we seem to be

657: observing, and mixing of ejecta and swept-up material across the contact

658: discontinuity due to the interaction with a dense shell may explain the

659: lack of any obvious spatial separation of these components. Finally,

660: dust depletion by the WR wind could also contribute to the very low

661: inferred dust-to-gas ratio in Kes~75.

662: 

663: The X-ray luminosity for Kes~75, based on the flux values from Table~2,

664: is extremely high ($L_x = 4.3 \times 10^{37} d_{19}^2 {\rm\ ergs\

665: s}^{-1}$). Nearly 70\% of this luminosity can be attributed to the

666: low-temperature emission component, and this raises the concern that most

667: of this inferred emission is actually hidden by the high column density.

668: As noted above, a more complex temperature distribution for this component

669: could yield a significantly different unabsorbed result. However,

670: the luminosity for the higher temperature component is exceptionally

671: large as well, as is that for the pulsar and the PWN. This could

672: point to an error in the distance estimate of $19$~kpc. However,

673: a rather large error would be required. To reduce $L_{x,pwn}/\dot E$

674: to a value of 1\% -- similar to that for the Crab, but still a factor

675: of 10 larger than the (very broad) average for PWNe -- would require a

676: distance reduction by a factor of 4.5. This would reduce the remnant

677: radius and inferred free-expansion speed by the same factor, and reduce

678: the mass estimates by a factor of nearly 40. However, this distance

679: would double the already-high inferred density of the postshock gas,

680: and the observed column density would be anomalously high. If, instead,

681: the interpretation of a distant WR progenitor is correct (and we infer

682: this based on several lines of reasoning), then Kes~75 is the remnant

683: of a relatively rare event, and perhaps the fact that its properties

684: stand out from the rest of the population is not unexpected.

685: 

686: \subsection{The remnant's PWN}

687: 

688: Extrapolation of the radio and X-ray spectra for the PWN in Kes

689: 75 indicate a spectral break near the IR band. The upper limits we

690: derive for the infrared emission are consistent with such a break,

691: but do not provide additional constraints because the values are

692: above the extrapolated break at $\nu_b \approx 5 \times 10^{14}$~Hz.

693: If interpreted as a synchrotron break, this would indicate a magnetic

694: field $B \sim 100 \mu$G given the age of the pulsar. This is somewhat

695: large value for a PWN, and corresponds to a magnetic energy within the

696: nebula that comprises nearly the entire spin-down energy deposited by

697: the pulsar over its estimated age. The required change in spectral index

698: for such a break is $\Delta \alpha \approx 0.7$, which is larger than

699: the change of 0.5 expected purely from synchrotron losses.  However, a

700: wide variety of effects can lead to spectral curvature or complex breaks,

701: including structure in the input particle spectrum, nonuniform magnetic

702: fields, and time dependence in the pulsar input power. Considerably

703: deeper IRAC observations could constrain the spectral behavior between

704: the radio and X-ray bands. Similarly, observations in the submillimeter

705: and TeV bands could potentially probe the mid-frequency behavior of the

706: spectrum as well as the high energy cut-off.

707: 

708: 

709: \section{Conclusions}

710: 

711: Using updated versions of fitting codes and new $Chandra$ observations,

712: we model the X-ray spectrum of Kes~75 using a two-component thermal

713: model in which we broadly associate a cool, solar-abundance component

714: with forward-shocked CSM/ISM, and a hot, metal-rich component with

715: reverse-shocked ejecta. The implied gas masses are somewhat high,

716: particularly for the cool component, suggesting that the forward shock has

717: encountered dense ambient material. The fits give ionization timescales

718: for the cool component of $\tau > 10^{12}{\rm\ s\ cm}^{-3}$, implying

719: an age of $> 2400$ years (older than the measured spindown age), and of

720: $\tau < 10^{10} {\rm\ s\ cm}^{-2}$ for the hotter, enriched component,

721: implying very recently shocked material ($\sim$75 years). This, combined

722: with the large size and high average expansion velocity for the SNR,

723: suggests a Wolf-Rayet progenitor that cleared out a $\sim$10 pc bubble

724: before exploding in a type Ib/c supernova.

725: 

726: We also report the detection of Kes~75 at 24 $\micron$; the morphology

727: of the infrared emission is spatially coincident with the observed

728: X-ray shell emission. We use this detection and the parameters derived

729: from the X-ray fits to model the dust emission of this remnant, finding

730: a dust/gas mass ratio of about 10$^{-3}$ or less, considerably lower than the

731: value of $\sim$0.7\% for the Galaxy derived from optical and ultraviolet

732: extinction measurements. Additional infrared observations of Kes~75

733: at different wavelengths would further constrain this ratio. We do not

734: detect the remnant at shorter wavelengths, indicating the destruction

735: of small grains by the shock. We suggest that the use of high-resolution

736: infrared imaging and X-ray spectroscopy of SNRs to constrain dust models

737: might provide a new sample of Galactic ISM dust measurements, distinct

738: from extinction measurements.

739: 

740: We do not detect the pulsar wind nebula in Kes~75 in any of our infrared

741: observations, primarily due to the high Galactic background. The upper

742: limits we derive are consistent with unbroken extrapolations of the both

743: the radio and X-ray spectra of the nebula, suggesting a spectral break

744: either in the infrared or between the radio and infrared.

745: 

746: 

747: \acknowledgments

748: The work presented here was supported in part by {\sl Spitzer} Grants

749: JPL CIT 1264892 (DJH), JPL RSA 1264893 (SPR), and JPL 1265776 (POS), as well

750: as Chandra Grant GO6-7053X (POS) and NASA Contract NAS8-39073 (POS).

751: 

752: 

753: \begin{thebibliography}{}

754: 

755: 

756: \bibitem[Arendt(1989)]{1989ApJS...70..181A} Arendt, R.~G.\ 1989, \apjs, 70, 

757: 181 

758: 

759: \bibitem[Benjamin et al. (2003)]{2003PASP..115..953B} Benjamin, R.~A. et al.\

760: 2003, \pasp, 115, 953

761: 

762: \bibitem[Becker \& Helfand(1984)]{1984ApJ...283..154B} Becker, R.~H., \&

763: Helfand, D.~J.\ 1984, \apj, 283, 154

764: 

765: \bibitem[Becker \& Kundu(1976)]{1976ApJ...204..427B} Becker, R.~H., \&

766: Kundu, M.~R.\ 1976, \apj, 204, 427

767: 

768: \bibitem[Blanton \& Helfand(1996)]{1996ApJ...470..961B}Blanton, E.~L. \&

769: Helfand, D.~J.\ 1996, \apj, 470, 961

770: 

771: \bibitem[Bock \& Gaensler(2005)]{2005ApJ...626..343B} Bock, D.~C.-J., \&

772: Gaensler, B.~M.\ 2005, \apj, 626, 343

773: 

774: \bibitem[Borkowski et al.(2006)]{2006ApJ...642L.141B} Borkowski, K.~J., et 

775: al.\ 2006, \apjl, 642, L141 (B06)

776: 

777: \bibitem[Chevalier(2005)]{2005ApJ...619..839C} Chevalier, R.~A.\ 2005,

778: \apj, 619, 839

779: 

780: \bibitem[Chiar \& Tielens(2006)]{chairtielens06}

781: Chiar, J. E., \& Tielens, A. G. G. M. 2006, \apj, 637, 774.

782: 

783: \bibitem[Dwek \& Arendt(1992)]{1992ARA&A..30...11D} Dwek, E., \& Arendt, 

784: R.~G.\ 1992, \araa, 30, 11 

785: 

786: \bibitem[Gotthelf et al.(2000)]{2000ApJ...542L..37G} Gotthelf, E. V., 

787: Vasisht, G., Boylan-Kolchin, M., \& Torii, K.\ 2000, ApJ, 542, L37

788: 

789: \bibitem[Helfand et al.(2003)]{2003ApJ...582..783H} Helfand, D.~J.,

790: Collins, B.~F., \& Gotthelf, E.~V.\ 2003, \apj, 582, 783 (H03)

791: 

792: \bibitem[Helfand et al.(2006)]{2006AJ....131.2525H} Helfand, D.~J.,

793: Becker, R. H., White, R. L., Fallon, A., and Tuttle, S.\ 2006, AJ, 131, 2525

794: 

795: \bibitem[Indebetouw et al.(2006)]{2006ApJ...619..931I} Indebetouw, R.

796: et al. 2005, \apj, 619, 931

797: 

798: \bibitem[Lee(2005)]{2005JKAS...38..385L} Lee, H.-G.\ 2005, Journal of 

799: Korean Astronomical Society, 38, 385 

800: 

801: \bibitem[Livingstone et al.(2006)]{2006ApJ...647.1286L} Livingstone, M.~A.,

802: Kaspi, V.~M., Gotthelf, E.~V., \& Kuiper, L.\ 2006, ApJ, 647, 1286

803: 

804: \bibitem[Reach et al.(2006)]{2006AJ....131.1479R} Reach, W.~T., et al.\

805: 2006, \aj, 131, 1479

806: 

807: \bibitem[Saken et al.(1992)]{1992ApJS...81..715S} Saken, J.~M., Fesen, 

808: R.~A., \& Shull, J.~M.\ 1992, \apjs, 81, 715 

809: 

810: \bibitem[Salter et al.(1989)]{1989ApJ...338..171S} Salter, C.~J., Reynolds,

811: S.~P., Hogg, D.~E., Payne, J.~M., \& Rhodes, P.~J.\ 1989, \apj, 338,

812: 171

813: 

814: \bibitem[Stanimirovi\'{c} et al. (2001)]{2005ApJ...632L.103S}

815: Stanimirovi\'{c}, S. et al. 2005, \apj, 632, L103

816: 

817: \bibitem[Weingartner \& Draine(2001)]{2001ApJS..134..263W} Weingartner, 

818: J.~C., \& Draine, B.~T.\ 2001, \apjs, 134, 263 

819: 

820: \bibitem[Williams et al. (2006a)]{2006ApJ...652L..33W} Williams, B.~J.

821: et al. 2006, \apjl, 652, L33 (W06)

822: 

823: \bibitem[Williams et al. (2006b)]{2006AJ....132.1877W} Williams, R.~M.

824: et al. 2006, \aj, 132, 1877

825: 

826: \end{thebibliography}

827: 

828: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

829: 

830: \end{document}

831: