1:
2: \documentclass[12pt,preprint]{aastex}
3:
4: %% manuscript produces a one-column, double-spaced document:
5:
6: %%\documentclass[manuscript]{aastex}
7:
8: \newcommand{\vdag}{(v)^\dagger}
9: \newcommand{\myemail}{yxy7181@cis.rit.edu}
10:
11: \shorttitle{X-ray spectroscopy of BD+30$^{\circ}$3639}
12: \shortauthors{YU et al.}
13:
14: %% This is the end of the preamble. Indicate the beginning of the
15: %% paper itself with \begin{document}.
16:
17: \begin{document}
18:
19: %% LaTeX will automatically break titles if they run longer than
20: %% one line. However, you may use \\ to force a line break if
21: %% you desire.
22:
23: \title{The X-Ray Spectrum of a Planetary Nebula at High Resolution:
24: Chandra Gratings Spectroscopy of BD+30$^{\circ}$3639}
25:
26: %% Use \author, \affil, and the \and command to format
27: %% author and affiliation information.
28: %% Note that \email has replaced the old \authoremail command
29: %% from AASTeX v4.0. You can use \email to mark an email address
30: %% anywhere in the paper, not just in the front matter.
31: %% As in the title, use \\ to force line breaks.
32:
33: \author{Young Sam Yu\altaffilmark{1}, Raanan Nordon\altaffilmark{2}, Joel H. Kastner\altaffilmark{1,3}, John Houck\altaffilmark{4}, Ehud Behar\altaffilmark{2,5}, Noam Soker\altaffilmark{2}}
34:
35: \affil{1. Center for Imaging Science, Rochester Institute of Technology, Rochester, NY 14623-5604}
36: \affil{2. Department of Physics, Technion-Israel Institute of Technology, Haifa 32000, Israel}
37: \affil{3. Laboratoire d'Astrophysique de Grenoble, Universit\'e Joseph Fourier --- CNRS, BP 53, 38041 Grenoble Cedex, France}
38: \affil{4. Kavli Institute, Massachusetts Institute of Technology, Cambridge, MA 02139}
39: \affil{5. Senior NPP Fellow, Code 662, NASA/Goddard Space Flight Center, Greenbelt, MD 20771}
40:
41:
42:
43: \begin{abstract}
44: We present the results of the first X-ray gratings spectroscopy
45: observations of a planetary nebula (PN), the X-ray-bright, young
46: BD+30$^{\circ}$3639. We observed BD+30$^{\circ}$3639 for a total of
47: 300 ks with the Chandra X-ray Observatory's Low Energy Transmission
48: Gratings in combination with its Advanced CCD Imaging Spectrometer
49: (LETG/ACIS-S). The LETG/ACIS-S spectrum of BD+30$^{\circ}$3639 is
50: dominated by H-like resonance lines of O {\sc viii} and C {\sc vi}
51: and the He-like triplet line complexes of Ne {\sc ix} and O {\sc
52: vii}. Other H-like resonance lines, such as N {\sc vii}, as well
53: as lines of highly ionized Fe, are weak or absent. Continuum
54: emission is evident over the range 6--18 \AA. Spectral modeling
55: indicates the presence of a range of plasma
56: temperatures from T$_{x}$ ${\sim}$ 1.7 ${\times}$ 10$^{6}$ K to 2.9
57: ${\times}$ 10$^{6}$ K and an intervening absorbing column
58: $N_H\sim2.4\times10^{21}$ cm$^{-2}$. The same modeling conclusively
59: demonstrates that C and Ne are highly enhanced, with abundance
60: ratios of C/O ${\sim}$15--45 and Ne/O ${\sim}$3.3--5.0 (90\%
61: confidence ranges, relative to the solar ratios), while N and Fe are
62: depleted, N/O ${\sim}$0.0--1.0 and Fe/O ${\sim}$0.1--0.4. The
63: intrinsic luminosity of the X-ray source determined from the
64: modeling and the measured flux ($F_X = 4.1\times10^{-13}$ ergs
65: cm$^{-2}$ s$^{-1}$) is $L_X\sim8.6\times10^{32}$ erg s$^{-1}$
66: (assuming $D = 1.2$ kpc).
67:
68: These gratings spectroscopy results are generally consistent with
69: earlier results obtained from X-ray CCD imaging spectroscopy of
70: BD+30$^{\circ}$3639, but are far more precise. Hence the
71: Chandra/LETGS results for BD+30$^{\circ}$3639 place severe new
72: constraints on models of PN wind-wind interactions in which
73: X-ray emitting gas within PNs is generated via shocks and the
74: plasma temperature is moderated by effects such as heat conduction
75: or rapid evolution of the fast wind. The tight constraints placed
76: on the (nonsolar) abundances directly implicate the present-day
77: central star --- hence, ultimately, the intershell region of the
78: progenitor asymptotic giant branch star --- as the origin of the
79: shocked plasma now emitting in X-rays.
80:
81: \end{abstract}
82:
83:
84: \keywords{planetary nebulae: general---planetary nebulae:
85: individual(BD+303639)---stars: winds, outflows---stars:mass loss, post
86: AGB---stars: Wolf-Rayet---X-rays: ISM}
87:
88:
89:
90: \section{Introduction}\label{sec1}
91: Planetary nebulae (PNe) are the last stages of evolution for
92: intermediate-mass stars (1-8 M$_{\odot}$). The central star that
93: generates a PN terminates its evolution as a cool asymptotic giant
94: branch (AGB) star by ejecting its outer envelope. The UV radiation of
95: the newly exposed hot core --- a future white dwarf (WD) --- then
96: illuminates and ionizes the ejected envelope. At about the same time,
97: a ``hot bubble'' may be produced by collisions between the slowly
98: expanding, ambient AGB gas and the newly-initiated fast central star
99: wind (Kwok et al. 1978). Theories describing such wind
100: interactions within PNe predict that the plasma temperature within
101: the hot bubble should be high enough for the generation of soft
102: ($\stackrel{<}{\sim}$ 1 keV) X-ray emission, and that the dimensions
103: of the hot bubble X-ray source should be smaller than that of the
104: optically-emitting, ionized nebula (e.g., Zhekov \& Perinotto 1996;
105: Soker \& Kastner 2003; and references therein). ROSAT X-ray
106: observations of PNe appeared to offered early support for these
107: predictions (Kreysing et al. 1992; Guerrero et al. 2000; however see
108: also Chu et al. 1993).
109:
110: The advent of Chandra and XMM-Newton has
111: provided far more convincing evidence for the presence of wind-blown
112: hot bubbles within PNe (Kastner et al.\ 2008 and refs.\ therein;
113: hereafter, K08).
114: However, certain puzzling aspects of the wind-collision-generated hot
115: bubbles within PNe remain to be explained. In particular, contrary to
116: the expectations of simple wind-collision models, the temperature of
117: the X-ray emitting hot bubbles in PNe does not appear to depend on the
118: present central star wind velocity (K08). Furthermore, the
119: optical and X-ray
120: emitting regions of the same object can display sharp differences in
121: abundances (e.g., Maness et al. 2003 and refs.\ therein). These
122: observations raise fundamental questions, e.g.: what heating and
123: cooling mechanisms govern the temperature of the X-ray emitting
124: plasma? Does the X-ray emission emanate primarily from the former AGB
125: star wind, the present central star wind, or some mixture of the two?
126:
127: BD+30$^{\circ}$3639 (``Campbell's Star'') is a young planetary nebula
128: with a carbon Wolf-Rayet([WC]-type) central star; it has been studied
129: at a wide variety of wavelengths (e.g., Li et al.\ 2002 and references
130: therein). The nearby (distance $\sim1.2$ kpc; Li et al. 2002)
131: BD+30$^{\circ}$3639 has a young dynamical age ${\sim}$ (700 yr;
132: Leuenhagen et al.\ 1996), and its fast wind speed is ${\sim}$ 700 km
133: s$^{-1}$ (Leuenhagen et al. 1996; Marcolino et al.\ 2007). It is an
134: excellent target for X-ray observations, due to its unusually large
135: soft X-ray flux at earth (F$_{X}$ ${\sim}$ 4${\times}$10$^{-13}$ erg
136: cm$^{-2}$ s$^{-1}$, 0.1-2.0 keV; Kreysing et al. 1992; Arnaud et
137: al. 1996). Kreysing et al.\ (1992) first detected X-rays from
138: BD+30$^{\circ}$3639 with ROSAT, and estimated the hydrogen column
139: density toward and plasma temperature within the X-ray nebula ($N_{H}$
140: ${\sim}$ 1.4 ${\times}$ 10$^{21}$ cm$^{-2}$ and $T_{x}$ ${\sim}$ 2.5
141: ${\times}$ 10$^{6}$ K, respectively). The latter result ruled
142: out, e.g., a hot companion to the $\sim30$ kK central star as the
143: X-ray source and suggested the presence of a wind-shock-generated
144: hot bubble within BD+30$^{\circ}$3639.
145:
146: Using the Chandra Advanced CCD Imaging Spectrometer (ACIS), Kastner
147: et al. (2000) demonstrated that X-ray emission from
148: BD+30$^{\circ}$3639 was spatially extended, and that the X-ray
149: emitting gas is fully confined within the ${\sim}5''$ diameter
150: elliptical ring of photoionized gas seen in optical and IR images.
151: While these results supported the existence of a
152: ``classical'' hot bubble within BD+30$^{\circ}$3639, one could
153: not rule out other possibilities, such as jets resulting from binary
154: interactions (Bachiller et al. 2000; Kastner et al. 2001, 2002;
155: Soker \& Kastner 2003; Akashi et al. 2008). Whether the X-ray
156: emitting gas comes from jets or a fast spherical wind, there remains
157: the question of the role, if any, of heat conduction between the
158: X-ray emitting gas and the visible shell (Soker 1994; Zhekov \&
159: Perinotto 1996; Zhekov \& Myasnikov 1998, 2000), and/or mixing
160: of the two media (Chu et al. 1997) enhanced by instabilities
161: (Steffen et al. 2005, 2008; Stute \& Sahai 2006; Sch\"{o}nberner et al.\
162: 2006), in moderating the X-ray temperature to levels well below that
163: expected from collisions between the present-day 700 km s$^{-1}$
164: stellar wind and the previously ejected AGB star envelope.
165:
166: The source of the X-ray-emitting gas in BD+30$^{\circ}$3639 also
167: remains to be determined, even though it has been a favorite subject
168: of X-ray CCD spectroscopy. Arnaud et al. (1996) obtained estimates of
169: X-ray plasma abundances using ASCA CCD imaging spectrometer data,
170: finding C, N and Ne to be significantly overabundant and Fe to be
171: significantly depleted. Similar results were obtained via analysis
172: of Chandra and Suzaku CCD imaging spectroscopy (Kastner et al.\
173: 2000; Maness et al.\ 2003; Murashima et al.\ 2006). However, these
174: results are somewhat at odds with those obtained from optical/IR
175: wavelengths that show, e.g., depleted Ne in the bright shell of the
176: PN. While the presence of an enhanced Ne abundance in the
177: X-ray-emitting plasma seems reasonably secure, the degree of Ne
178: overabundance as well as other abundance
179: anomalies --- such as highly enhanced C and highly depleted Fe,
180: also inferred on the basis of X-ray CCD spectral modeling ---
181: remain quite uncertain. Indeed, Georgiev et al. (2006) argued that
182: X-ray CCD spectra cannot provide definitive constraints on the
183: plasma abundances in PNe.
184:
185: To make progress on these and other problems concerning the nature
186: and origin of the X-ray-emitting plasma within PNs requires X-ray
187: observations at high spectral resolution, from which we
188: unambiguously infer, with improved precision, the temperature and
189: composition of the X-ray emitting plasma. With this motivation, we
190: obtained a deep observation of BD+30$^{\circ}$3639 using
191: Chandra's Low Energy Transmission Gratings spectrometer in
192: combination with its Advanced CCD Imaging Spectrometer
193: (LETG/ACIS-S). We selected LETG/ACIS-S (as opposed to HETG/ACIS-S or
194: LETG/HRC-S) for Chandra gratings observations of BD+30$^{\circ}$3639
195: on the basis of the superior background rejection, soft X-ray
196: sensitivity, and order-sorting capabilities of this
197: configuration. In addition to the dispersed spectrum of
198: BD+30$^{\circ}$3639 (for which preliminary results were presented in
199: Kastner et al.\ 2006), the LETG/ACIS-S observations produced a
200: highly sensitive, undispersed 0th-order image of the PN. In this
201: paper, we present a comprehensive analysis of the dispersed
202: LETG/ACIS-S spectrum of BD+30$^{\circ}$3639. In a subsequent paper
203: (Yu et al.\ 2008, in prep.) we present a spatial/spectral analysis
204: of the LETG/ACIS-S zeroth-order image of BD+30$^{\circ}$3639 and
205: compare this image with the direct Chandra/ACIS-S3
206: image obtained in Cycle 1 observations.
207:
208: \section{Observations and Data Reduction}\label{sec2}
209:
210:
211: We obtained observations of BD+30$^{\circ}$3639 totaling 300 ks
212: exposure time with LETG/ACIS-S in 2006 February (85.4 ks), March (61.8
213: ks) and December (150 ks). The last observation was obtained in three
214: consecutive blocks at the same roll angle. The event data were
215: subject to standard pipeline processing (using Chandra X-ray Center
216: pipeline versions 7.6.7 for the 2006 February and March data and
217: 7.6.9 for the 2006 December data). In
218: Table~\ref{tbl-1}, we list exposure times and total photon counts
219: within the source and background regions. To determine total
220: counts in the 0th-order images, we selected a $10''$ circular region
221: centered on the source and a surrounding annular region (with inner
222: and outer radii of 13$''$ and 20$''$, respectively) for background.
223: The total 1st-order counts were determined directly from the extracted
224: source and background spectra (Sec.~\ref{sec3}). We also
225: extracted light curves from the source regions of the 2000 image and
226: the 0th-order images obtained in 2006; no measurable variability was
227: found, as expected given the diffuse nature of the source.
228:
229: Because the first two observations in 2006 (ObsIDs 5409 \& 7278) were
230: obtained at different spacecraft roll angles and aimpoints with
231: respect to each other and the last three (Dec.) 2006 observations, we
232: cannot generate a merged spectral image for the full 300 ks
233: exposure. However, the second (Dec.\ 2006) half of the 300 ks exposure
234: (ObsIDs 5410, 8495 \& 8498) was obtained at constant roll angle, so we
235: generated a single, 150-ks-exposure dispersed spectral image from
236: these data by merging the three Level 2 event (\verb+evt2+) files. The
237: full spectral images obtained from the resulting combined
238: \verb+evt2+ files are shown in Figures~\ref{dispNegaImage} and
239: ~\ref{dispPosiImage}. These dispersed images demonstrate that the
240: LETG/ACIS-S X-ray counts spectrum of BD+30$^{\circ}$3639 is dominated
241: by emission lines of highly ionized (He-like and H-like) ions of
242: oxygen and neon.
243:
244: For analysis of the dispersed spectrum, further processing involved
245: removal of artifacts on ACIS-S4 (using the CIAO tool ``destreak'') and
246: applying updated calibrations. Standard gratings point source
247: spectral extraction threads available in
248: CIAO\footnote{http://cxc.harvard.edu/ciao/threads/spectra\_letgacis/}
249: were used to generate spectrum pulse height amplitude (PHA) files and
250: corresponding redistribution matrix files (RMFs) and auxiliary
251: response files (ARFs). These threads implicitly ignore the
252: (non-negligible, $\sim5''$) spatial extension of the
253: BD+30$^{\circ}$3639 X-ray source, resulting in artifically
254: ``broadened'' emission line features (we discuss this effect further
255: in Secs.\ 3.1 and 3.2, respectively). We extracted positive ($m=
256: +1$) and negative ($m= -1$) first-order LETG spectra for each of the
257: five observations (Table~\ref{tbl-1}). To enhance the signal to noise
258: ratio, each LETG spectrum was rebinned by a factor 8, resulting in a
259: wavelength dispersion of 0.1 \AA/bin. We then merged the resulting
260: spectra into a single spectrum. In parallel, the corresponding ARFs
261: were averaged and weighted by the relative exposure times, then merged
262: into a single ARF. The various RMFs obtained from the individual
263: observations for a given order are identical to within the calibration
264: uncertainties. Hence, for the spectral analysis described below, we
265: used the same, representative RMF for each LETG dispersion arm.
266:
267: \section{Analysis and Results}\label{sec3}
268:
269: The merged, 300 ks exposure time spectrum resulting from the LETG/ACIS
270: spectral image data reduction procedure described in \S 2 is displayed
271: in Fig.~\ref{combFirstorder}a. The spectrum displays strong emission
272: lines superposed on a weak continuum, with an abrupt rise in the
273: continuum at $\sim30$ \AA\ that is likely due to background events. To
274: account for this apparent residual background, we identified a region
275: devoid of bright X-ray sources, displaced 215 arcsec from the source
276: position along the detector y direction, and extracted first-order
277: background spectra at this position from each of the 5
278: observations. The corresponding RMFs and ARFs of background spectra
279: were also generated. We rebinned and merged these individual
280: background spectra into a single spectrum (seen superimposed on the
281: source spectrum in Fig.~\ref{combFirstorder}a).
282: Figure~\ref{combFirstorder}b shows the resulting combined, 1st-order,
283: background-subtracted spectrum of BD+30$^{\circ}$3639 in the
284: wavelength range of 5 - 40 \AA. The apparent ``continuum'' in the
285: region 30 - 40 {\AA} is effectively removed from this
286: background-subtracted spectrum.
287:
288: The brightest lines in the background-subtracted spectrum of
289: BD+30$^{\circ}$3639 are due to highly-ionized (H- and He-like)
290: species of C, O, and Ne;
291: the resonance lines of H-like O {\sc viii}
292: (${\lambda}$ 18.97) and C {\sc vi} (${\lambda}$ 33.6) and the He-like
293: triplet line complexes of Ne {\sc ix} (${\lambda}$ 13.45, 13.55, 13.7) and O
294: {\sc vii} (${\lambda}$ 21.60, 21.80, 22.10) are especially
295: prominent. Other H-like resonance lines,
296: such as N {\sc vii} (${\lambda}$ 24.78) and lines of highly ionized Fe, are
297: weak or absent. Continuum emission is evident over the range 6--18
298: \AA. The excess emission near 25 {\AA} can likely be attributed to enhanced
299: high-order lines of C {\sc vi} within its H-like recombination
300: line spectrum (Nordon et al.\ 2008, in prep.).
301:
302: \subsection{Line identifications, fluxes, and source angular sizes}\label{sec3.1}
303:
304: We measured line fluxes with the Interactive Spectral Interpretation
305: System (ISIS\footnote{http://space.mit.edu/CXC/ISIS/}; Houck \&
306: Denicola 2000). Given the relatively symmetric appearances of the
307: profiles of prominent emission lines, we used a fit function
308: consisting of a constant local continuum (polynomial) plus one or more
309: Gaussian functions. In fitting the strong He-like triplet lines (Ne
310: {\sc ix} and O {\sc vii}), the ratios of line strengths and widths of
311: the triplet components were fixed using the values given by the
312: Chandra atomic database (ATOMDB ver. 1.3) for the case of
313: low-density plasma ($n_e << 10^{10}$ cm$^{-3}$). Thus, the free
314: parameters were the intensities of triplets, the line-center
315: wavelength and FWHM of one of the triplets, and the coefficients of
316: the polynomial representing the local continuum.
317:
318: For those emission lines that could be measured with acceptable
319: statistics, Table~\ref{tbl-2} lists the line identifications, fluxes,
320: and widths. The line-fitting procedure thereby confirms the
321: identification of at least 15 lines and line complexes in the LETG
322: spectrum of BD+30$^\circ$3639, ranging from the very strong C {\sc
323: vi} L$\alpha$ line to weak Mg {\sc xii} and Si {\sc xiii}
324: lines. The Table also lists upper limits on the fluxes of the lines
325: of other ions that are of similar ionization potential to the
326: well-detected species such as Ne {\sc ix}, Ne {\sc x}, O {\sc vii}, O
327: {\sc viii}, and C {\sc vi}. Lines of these important species ---
328: e.g., Fe {\sc xvii} and N {\sc vii} --- should be prominent
329: in the spectrum of a solar-abundance plasma at the
330: approximate temperature implied by the well-detected lines ($T_x
331: \sim 2\times10^6$ K).
332:
333: Since the X-ray line widths reflect the extended nature of the
334: source rather than, e.g., plasma turbulence (see below) or
335: kinematics, the width
336: measurements for well detected lines ($\Delta \lambda$) are
337: expressed in terms of
338: the corresponding angular FWHM in arcsec, assuming a dispersion for
339: LETG of 18.02 arcsec {\AA}$^{-1}$ (Dewey, 2002). The resulting angular
340: FWHMs are consistent with each other and with the ($\sim5''$) source
341: angular extent in the zeroth-order image, within the
342: uncertainties. Furthermore there are
343: no discernable systematic redshifts or blueshifts measured for the
344: emission line centers. Hence, for purposes of the plasma modeling
345: described here, all of the emission lines can be considered to arise from
346: the same region within the nebula.
347:
348: \subsection{Global spectral fitting}\label{sec3.2}
349:
350: Global X-ray spectral model fitting of lines and underlying continuum
351: is necessary to simultaneously constrain relative plasma elemental
352: abundances and temperatures. In adopting this approach to fit the LETG
353: spectrum of BD+30$^{\circ}$3639, we selected the {\it ``Cash''} method
354: (Cash, 1979) as the fit statistic to treat low count data. For the
355: model fitting, we specified the displaced background spectrum (see \S
356: 3.1) along with the merged source spectrum. To investigate plasma
357: physical conditions, we used ISIS to construct Astrophysical Plasma
358: Emission Database (APED; Smith et al. 2001) models, varying plasma
359: metal abundances such as Fe, Ne, O, C, N and Mg, and leaving all
360: other abundances, including H and He, fixed at solar (Anders \&
361: Grevesse 1989). The latter assumption is, of course, unlikely to
362: be valid if the X-ray-emitting plasma is dominated by the
363: H-depleted, He-enriched wind characteristic of the present-day
364: central star (e.g., Marcolino et al.\ 2007) --- a hypothesis that is
365: indeed supported by our modeling results (see \S\S 3.2.2, 4.1). We
366: did not consider higher Lyman series C {\sc vi} lines that lie in the
367: region $\sim$ 25--29 {\AA} in the spectral model because APED only
368: includes the C {\sc vi} Lyman series up to the $\delta$ line. Analysis
369: of these higher-energy C {\sc vi} transitions is described in Nordon
370: et al.\ 2008 (in prep.).
371:
372: \subsubsection{Isothermal vs.\ two-component APED models}
373:
374: To reproduce the merged LETG 1st-order spectrum of BD +30$^\circ$3639,
375: we attempted fits with both single-component and two-component APED
376: plasma models. Note that the spectral resolving power of LETG for
377: a point source ranges from $E/{\bigtriangleup}E \sim$130 at 5 {\AA}
378: to $E/{\bigtriangleup}E \sim$1000 at 40 {\AA} (Dewey 2002
379: ). The spatial extension of BD+30$^{\circ}$3639
380: (FWHM ${\sim}$3 arcsec; Table 2), combined with use of ACIS-S rather
381: than HRC as the detector, then degrades the effective resolving
382: power of LETG by a factor $\sim3$ (Dewey 2002). Source spatial
383: extent --- rather than, e.g., gas turbulence or thermal broadening
384: --- therefore determines the line widths. To emulate this artificial
385: line ``broadening'', we used the turbulent velocity parameter
386: ($V_{turb}$) available in the APED model. We find that two turbulent
387: velocity components --- with best-fit parameter values of $\sim1700$
388: km s$^{-1}$ and $\sim900$ km s$^{-1}$ for the short- and
389: long-wavelength spectral regions, respectively, consistent with the
390: mean FWHM measured for the lines --- are sufficient to reproduce the
391: ``broadening'' caused by the extended X-ray source within BD
392: +30$^\circ$3639. We emphasize that these results do not actually
393: represent measurements of gas turbulence; rather, like emission line
394: FWHM (Table 2), the best-fit $V_{turb}$ values serve as an
395: indication of source spatial extent.
396:
397: Each of the models was assumed to undergo absorption due to
398: intervening neutral material characterized in terms of the column
399: density of neutral H, $N_H$,
400: using the standard, solar-abundance (\verb+wabs+) model (Morrison \&
401: McCammon 1983). However note that, given
402: the close correspondence
403: between the visual extinction and X-ray surface brightness
404: distributions of BD+30$^\circ$3639 (Kastner et al.\ 2002), the
405: absorption is in fact best
406: attributed to the nebula itself rather than
407: intervening interstellar medium. The X-ray-absorbing material may be the
408: extended molecular envelope of BD+30$^\circ$3639, pockets of cold,
409: dense gas embedded in the ionized nebula, or some combination of
410: these contributions; as a result, the composition of the absorbing
411: material may differ significantly from that assumed in the
412: \verb+wabs+ model. The implications for the results for $N_H$, as
413: well as for model plasma abundances, are discussed below.
414:
415: Figure~\ref{One_2Vturb_model}a shows the merged, 300 ks exposure, 1st-order
416: LETG/ACIS-S counts spectrum of BD+30$^{\circ}$3639 overlaid with the
417: single-component model spectrum obtained from the APED. The comparison
418: of the flux-calibrated spectrum with the model makes apparent the
419: strength of C {\sc vi} (${\lambda}$ 33.6) relative to the other strong lines
420: (Ne {\sc ix}, O {\sc vii} and O {\sc viii};
421: Figure ~\ref{One_2Vturb_model}b). Other features
422: apparent in the long wavelength (${\lambda}$ $>$ 30) region of the
423: flux-calibrated spectrum are likely artifacts of poor
424: photon counting statistics combined with
425: the very low net effective area of LETG/ACIS in this region.
426: However, we find this single-component APED model is
427: insufficient to adequately fit the spectrum. In particular, while the
428: O {\sc vii} to O {\sc viii} line ratios are reasonably well
429: reproduced, the model cannot simultaneously fit the Ne {\sc ix} and Ne {\sc x}
430: lines in the 12 \AA\ region.
431:
432: In contrast, the two-component APED model well reproduces the
433: intensities of all strong emission lines and better matches the
434: 6--18 \AA\ continuum (Fig.~\ref{Two_comp_model}). The high-temperature
435: boundary is constrained in part by the nondetection
436: of the Mg {\sc xii} line at 8.42 {\AA} contrasted with
437: the weak but clear Mg {\sc xi} line at 9.12 {\AA}. The
438: low-temperature boundary is harder to constrain, as it
439: relies in part on the relative intensities of the (somewhat noisy) C
440: {\sc vi} lines at 28.46 {\AA} and 33.73 {\AA}. Since the plasma
441: temperatures are therefore mainly governed by two indicators --- i.e.,
442: the line ratios of the H-like to He-like O and Ne --- introducing more
443: than 2 temperature components into the model would make the fit result
444: degenerate.
445:
446: \subsubsection{Best-fit parameters: results and confidence ranges}
447:
448: The best-fit temperature of the single-component APED model indicates
449: that, under the isothermal plasma approximation, the characteristic
450: plasma temperature lies in the range 2.2--2.4$\times10^6$ K. It is
451: therefore not surprising that the values of $T_X$ obtained from the
452: two-temperature-component model ($T_1 = 2.9\times10^6$ K and $T_2 =
453: 1.7\times10^6$ K) brace this range. Table 3 demonstrates that the two
454: preceding models also yield consistent results where plasma abundances
455: are concerned but that the two-component model, in addition to
456: providing a superior fit to the LETG spectrum, yields these abundance
457: results to greater precision.
458:
459: We reiterate that the best-fit absolute abundances listed in
460: Table 3 (as number ratios relative to solar) were obtained under the
461: assumption of solar H and He abundance (Anders \& Grevesse 1989) and
462: are therefore subject to large but unknown systematic
463: uncertainties. Specifically, no diagnostics of H and He abundances
464: are available in the X-ray regime, yet it is likely that the
465: X-ray-emitting plasma is dominated by present-day (H-depleted,
466: He-enriched) [WC] stellar wind material (\S 4.1). In this case of a
467: strongly H-depleted stellar wind, the abundance normalization to H
468: becomes irrelevant. One could, however, re-normalize the Table 3 results
469: to the solar He abundance under the assumption that ratio of the C to He
470: abundances is identical to the [WC] stellar wind value, i.e.,
471: C/He $\sim$ 0.4 by number (Marcolino et al. 2007), or $\sim$108 times
472: the solar ratio. The values listed in Table 3 could then be
473: increased by a factor 108/28.3 = 3.8, where 28.3 is the C abundance obtained
474: from the two-component model fit,
475: so as to yield elemental
476: abundances relative to He in
477: solar units. The results for emission measure would then also have
478: to be redefined to refer to the number density of He nuclei (rather
479: than H). In the context of the APEC model, this is equivalent to
480: reducing the model normalization by the same factor as that of the
481: increase in elemental abundances.
482:
483: Given the foregoing problems in attempts to determine absolute
484: abundances, all subsequent analysis in this paper is based on the
485: modeling results for abundances as number ratios relative to O. In
486: this regard, the results listed in Table 3 then indicate
487: that C is very overabundant (C/O $\sim30$, relative to the solar
488: ratio), Ne is overabundant (Ne/O $\sim3.8$), and both Fe and N
489: likely are depleted (Fe/O $\sim0.2$ and N/O $\sim0.4$).
490:
491: Because the LETG response matrix is nearly diagonal, the observed
492: (absorbed) source X-ray flux deduced from the model fitting is similar
493: for the two models ($F_X= 4.4\times10^{-13}$ ergs cm$^{-2}$
494: s$^{-1}$). The inferred intrinsic (unabsorbed) X-ray source luminosity
495: is then constrained by the model fitting to lie in the range
496: $7.4\times10^{32}$ erg s$^{-1}$ to $8.6\times10^{32}$ erg s$^{-1}$. The
497: low-$T$ and high-$T$ components of the latter
498: model account for $\sim$25\% and $\sim$75\% of the total source
499: luminosity, respectively.
500:
501: In Figs.~\ref{confidence_contour_vs_nH}-\ref{confidence_contour_vs_T}
502: we present plots of confidence contours obtained from the
503: two-component APED model for various parameter combinations.
504: Figure~\ref{confidence_contour_vs_nH} demonstrates that N$_H$, which
505: is constrained mainly by the relative strengths of metal lines in the
506: softer part of spectrum, is (in principle) very well determined, and
507: that this parameter is essentially insensitive to the plasma
508: temperature and C and O abundances obtained from the model fitting.
509: Given a standard (ISM) gas-to-dust (hence $N_H/A_V$) ratio, the
510: best-fit value of $N_H = 2.4\times10^{21}$ cm$^{-2}$ obtained from the
511: two-component model is roughly consistent with the typical visual
512: (dust) extinction measured toward the regions of the nebula from which
513: X-rays are detected ($A_V \sim$ 1--2; Kastner et al. 2002). Adopting
514: the (overly simplistic) assumption that the absorption arises in a
515: spherically symmetric envelope surrounding the X-ray-emitting plasma,
516: the best-fit value of $N_H$ in the two-component model implies a
517: neutral envelope mass of $\sim0.03 M_\odot$ --- in reasonable
518: agreement with available estimates (e.g., Bachiller et al.\ 1991). We
519: conclude that $N_H$ is unlikely to be overestimated due to the
520: assumption of solar abundances in the absorbing column (implicit via
521: use of the \verb+wabs+ model).
522:
523: In addition, the confidence contour plots demonstrate that the C and
524: Ne abundance parameters are well correlated with O, indicating that
525: the C/O and Ne/O ratios (as well as other key abundance ratios, such
526: as Fe/O) are very well constrained by the LETG spectrum
527: (Fig.~\ref{confidence_contour_vs_O}) --- even though the individual
528: absolute abundances remain uncertain, due to the lack of constraints
529: on the plasma H abundance. Specifically, we find C/O ${\sim}$15--45,
530: Ne/O ${\sim}$3.3--5, N/O ${\sim}$0--1.0, and Fe/O ${\sim}$0.1--0.4,
531: relative to the solar ratios. The existence of a large C overabundance
532: (relative to solar) is supported by the tight constraints placed on
533: N$_{H}$ and the lack of correlation between inferred C abundance and
534: N$_H$ (Fig.~\ref{confidence_contour_vs_nH}). That is, it is unlikely
535: that the inferred high C abundance can be attributed to an
536: overestimate of the intervening absorption due to neutral metals.
537: On the other hand, as just discussed, it is also unlikely that the
538: abundance of C has been vastly underestimated due to the assumption
539: of standard (solar) metal abundances in the absorbing material.
540: Fig.~\ref{confidence_contour_vs_T} futhermore makes clear that the
541: inferred C, O, and Ne abundances are relatively insensitive to the
542: best-fit plasma temperatures.
543:
544: \section{Discussion}
545:
546: \subsection{Plasma modeling: comparison with previous results}
547:
548: In Table 4 we compare previous results from X-ray CCD
549: spectroscopy of BD+30$^\circ$3639 with those obtained from modeling
550: its LETG/ACIS spectrum (Sec.\ 3.2).
551: Whereas the value of $T_X$ obtained from the isothermal model
552: (Table 3) is on the low side of the range found in
553: the CCD-based work --- likely reflecting the spectral dominance of
554: the longer-wavelength lines of C and O --- the temperatures
555: determined from the two-component model fit brace the range of
556: values previously determined from CCD spectra. Although the LETGS
557: modeling definitively demonstrates that isothermal models are not
558: adequate to match the line spectrum in detail (\S 3.2.1), the
559: comparison to previous work, as well as to the isothermal model
560: explored here, suggests that single-component plasma models
561: that are based on X-ray CCD spectra are capable of
562: recovering the characteristic temperature (as opposed to the
563: temperature extremes) of the superheated plasma in PNs.
564:
565: The LETG/ACIS spectral modeling confirms the Ne and C enrichment and
566: Fe underabundance in the X-ray-emitting plasma of BD+30$^\circ$3639
567: that were previously inferred from X-ray CCD spectra, albeit with
568: much improved constraints on the degree of these abundance anomalies
569: (relative to O). In particular, the LETG/ACIS spectral modeling ---
570: while confirming that C is highly enriched in the X-ray emitting
571: plasma -- definitively precludes a plasma C/O ratio larger
572: than $\sim45$, relative to solar. Indeed, our lower limit on the C/O
573: ratio in the diffuse X-ray emission (C/O $\sim15$) is much
574: more consistent with the ratio at the central star, as derived
575: from optical/UV spectroscopy (C/O $\sim12$; Leuenhagen et al.\ 1996;
576: Marcolino et al.\ 2007), than with the ratio characteristic of
577: the nebular gas (C/O $\sim1.6$; Pwa et al.\ 1986).
578:
579: In contrast to the case of C, whose large overabundance is
580: evident from the strength of the resonance line of C {\sc vi},
581: there is no unambiguous
582: evidence for Fe emission lines in the entire LETG/ACIS spectrum of
583: BD+30$^\circ$3639. Although a line is present at $\sim15.15$ \AA\
584: that may be partly due to 15.014 \AA\ line emission from Fe {\sc
585: xvii}, it is likely that the O {\sc viii} line at 15.1670 {\AA}
586: contributes to (even dominates) the flux measured for the
587: $\sim15.15$ \AA\ line. In
588: addition, we do not clearly detect any Fe {\sc xvii} emission lines
589: around 17\AA; these lines
590: are expected to be bright at the relatively low plasma temperatures
591: found here (Doron \& Behar 2002). Furthermore there is no evidence
592: for lines of Fe {\sc xvi} (or Fe {\sc xviii}), such as might be
593: expected if the temperature were too low (or too high) for efficient
594: formation of Fe {\sc xvii} via ionization equlibrium. This lack of
595: Fe lines results in a firm upper limit on Fe abundance of $\sim0.3$
596: relative to solar (Table 3), consistent with the optical/UV
597: results of Georgiev
598: et al. (2006) for the nebula and Marcolino et al. (2007) for the
599: central star.
600:
601: In addition, the 24.8 \AA\ resonance line of N {\sc vii} is not
602: detected, although the line flux is not as well constrained as that
603: of Fe {\sc xvii} due to the low sensitivity of LETG/ACIS in the
604: former wavelength regime. As a consequence of this nondetection, the
605: N abundance obtained from the LETG/ACIS spectral modeling
606: demonstrates that N is, if anything, underabundant relative to
607: solar, contradicting several previous X-ray abundance studies. Our
608: gratings-based result of an N underabundance, which is consistent with
609: optical/UV results for the central star (Marcolino et al. 2007),
610: indicates that modeling of previous (CCD) X-ray spectra has
611: confused N with C. We also find that previous indications of
612: depleted Mg obtained from CCD spectra are not supported by the LETG
613: data.
614:
615: The absorbing column and (thus) intrinsic X-ray luminosity
616: inferred from the LETG/ACIS modeling are on the high side of the
617: range of values previously determined from X-ray CCD spectra
618: (Table 4). However, the LETG spectral modeling has significantly
619: decreased the uncertainty in the inferred value of N$_{H}$ and,
620: hence, $L_X$.
621:
622: \subsection{Plasma abundances: constraints on the origin of the
623: X-ray-emitting gas}
624:
625: The similarity of the large overabundance of C and underabundance of
626: Fe we determine from the LETG/ACIS spectral modeling to these same
627: anomalies as determined for the central star (Marcolino et al.\ 2007
628: and references therein) traces the X-ray-emitting gas directly back to
629: the present-day central star. Indeed, the mass fractions one
630: would deduce from Table 3 under the assumption that the C/He number
631: ratio is identical to that of the central [WC] star --- 41.5\% He,
632: 49.7\% C, 4.8\% O, 3.3\% Ne, $<0.24$\% N, and $<0.2$\% Fe --- agree
633: remarkably well with those found by Marcolino et al.\ (2007) for the
634: [WC] stellar wind. Furthermore, our robust determination of a Ne
635: overabundance of $\sim3$ to $\sim6$ relative to solar is consistent
636: with the predictions of models describing H-deficient central stars of
637: PNe (Werner \& Herwig 2006). Hence, as also pointed out by Murashima
638: et al.\ (2006) and Kastner et al.\ (2006), the nonsolar composition of
639: the X-ray-emitting plasma in BD+30$^{\circ}$3639 appears to be the
640: direct result of nucleosynthesis processes in the precursor AGB star
641: (e.g., Herwig 2005 and references therein).
642:
643: These results indicate that the shocked plasma now seen in X-rays
644: originated deep within the AGB star, in the ``intershell'' region;
645: He-shell burning just below this region is responsible for the C
646: generation. Meanwhile, the observed enhanced Ne/O and low Fe/O, N/O,
647: and Mg/O abundance ratios can be explained as a natural consequence of
648: the s-process within the ``pulse driven convection zone'' (Herwig
649: 2005). The Ne may be predominantly $^{22}$Ne, which can be readily
650: generated --- at the expense of $^{14}$N --- within the He burning
651: shell. The $^{22}$Ne can then serve as an iron-depleting neutron
652: source during the s-process. In such a scenario, one therefore expects
653: Ne to be enhanced, while Fe, N and Mg are depleted relative to O
654: --- as observed (Table 3).
655:
656: \subsection{The temperature of X-ray-emitting plasma within BD
657: +30$^\circ$3639}
658:
659: The two plasma temperatures obtained from our best-fit,
660: two-component APED model --- $1.7\times10^6$ K and $2.9\times10^6$ K
661: --- likely represent the extremes of a continuous range of
662: temperature within BD$+30^\circ$3639. Hence the spectral diagnostics
663: available in the LETG/ACIS X-ray spectrum of BD$+30^\circ$3639
664: demonstrate conclusively, for the first time, the presence of a
665: temperature gradient within the X-ray-emitting region of a PN.
666: However, even the higher of these two temperatures, obtained from
667: those line ratios diagnostic of the hottest plasma present in the
668: LETG/ACIS-S 1st-order spectrum, is far lower than that expected from
669: simple adiabatic shock models, given the present-day central star fast
670: wind speed (700 km s$^{-1}$; Leuenhagen et al.\ 1996; Marcolino et
671: al.\ 2007). This discrepancy between observed and predicted hot
672: bubble plasma temperatures has been noted by many investigators over
673: the past decade (e.g., Arnaud et al. 1996; Chu et al.\ 2001; Soker \&
674: Kastner 2003; and references therein), and was recently discussed by
675: Kastner et al.\ (2008) in their analysis of the collective Chandra and
676: XMM-Newton data compiled to date for PN hot bubbles. The temperature
677: discrepancy has previously been explained as indicative of heat
678: conduction from the tenuous hot bubble to the dense, relatively cool
679: swept-up shell (Stute \& Sahai 2007; Sch\"{o}nberner et al. 2006;
680: Steffen et al. 2008) or mixing of the two media (Chu et al 1997; Stute
681: \& Sahai 2006).
682:
683: However, such models of mixing and heat conduction predict that the
684: bulk of the X-ray emission arises in the same gas that is responsible
685: for the visible-light nebula (Steffen et al. 2008). Our robust
686: determination of nonsolar abundances (in particular, greatly enhanced
687: C and Ne) in the X-ray-emitting plasma within BD$+30^\circ$3639
688: therefore indicates that the shocked gas is predominantly present-day
689: stellar wind or jet, as opposed to nebular gas. Hence the heat
690: conduction and mixing mechanisms, though certainly viable in the
691: general case, do not play a major role in determining the X-ray
692: temperature of this particular PN. We note that this argument likely
693: holds even at early stages of the heat conduction process. This is
694: because at early times the conduction goes through the ``evaporation''
695: stage, where the intermediate-temperature gas comes from the cold
696: phase (Borkowski et al. 1990). In PNs the cold phase is the visible
697: nebular gas, and the abundance results obtained here preclude such an
698: origin for the X-ray-emitting plasma.
699:
700: There remains the possibility that the shocked wind presently seen in
701: X-rays was ejected at an earlier epoch when the fast wind speed was
702: ${\sim}$ 300--400 km s$^{-1}$, a velocity regime more consistent with
703: the measured range of $T_X$ (Arnaud et al.\ 1996; Akashi et al.\ 2006,
704: 2007). Alternatively, collimated jets, perhaps associated with
705: molecular ``bullets'' detected in mm-wave interferometric imaging
706: (Bachiller et al.\ 2000), may be responsible for the X-ray
707: emission. Such a scenario would be similar to that proposed for NGC
708: 7027 (Kastner et al.\ 2002; Cox et al.\ 2002) and would be
709: consistent with the possibility that BD$+30^\circ$3639 is a bipolar
710: nebula viewed nearly pole-on (Kastner et al.\ 2002; Lee \& Kwok 2005).
711: %Indeed, analysis of ACIS imaging --- and, in
712: %particular, comparison of the zeroth-order LETG/ACIS image obtained
713: %here with the ACIS-S image obtained $\sim6$ years earlier --- suggests
714: %that the X-ray emission from BD$+30^\circ$3639 may arise from
715: %collimated outflows or jets expanding at projected velocities
716: %$\sim200$ km s$^{-1}$, rather than from (or perhaps in addition to) a
717: %hot bubble generated by the temporally evolving, quasi-spherical
718: %central star wind.
719: This possibility will be pursued in our forthcoming
720: paper concerned with spatial analysis of the LETG/ACIS data (Yu et
721: al.\ 2008, in preparation).
722:
723: \section{Conclusions}
724:
725: Using Chandra's LETG/ACIS-S spectrometer, we have obtained the first
726: X-ray gratings spectrum of a planetary nebula. The LETG/ACIS-S
727: spectrum of the young, rapidly evolving BD$+30^\circ$3639 displays
728: strong emission in the H-like resonance lines of O {\sc viii} and C
729: {\sc vi} and He-like triplet line complexes of Ne {\sc ix} and O {\sc
730: vii}, and appears devoid of lines of highly ionized Fe and N. Our
731: spectral modeling, consisting of fits of variable-abundance APED
732: plasma models with one and two temperature components, demonstrates
733: that an isothermal plasma is unable to simultaneously
734: reproduce key spectral features such as the O {\sc vii} to O {\sc
735: viii} and Ne {\sc ix} to Ne {\sc x} line ratios and the 6--18 \AA\
736: continuum. The best-fit two-component plasma model, which is able to
737: well match these same features, indicates that the X-ray emission line
738: spectrum is representative of a range of temperatures from $\sim1.7$
739: MK to $\sim2.9$ MK. These results constitute the first case in
740: which a temperature gradient has been inferred within the
741: X-ray-emitting region of a PN.
742: %%The inferred absorbing column ($N_H \sim
743: %%2.4\times10^{21}$ cm$^{-2}$) is consistent with the typical visual extinction
744: %%toward the X-ray-emitting regions of BD$+30^\circ$3639.
745:
746: The spectral modeling places tight constraints on the degree
747: of abundance anomalies present in the X-ray-emitting plasma within
748: BD$+30^\circ$3639, convincingly demonstrating that Fe is highly deficient
749: (best-fit Fe/O $\sim0.3$, relative to the solar ratio) and that C and Ne
750: are highly enhanced in abundance (best-fit ratios C/O $\sim30$ and
751: Ne/O $\sim4$). This C overabundance, although very large, is not as
752: pronounced as deduced previously on the basis of X-ray CCD
753: spectroscopy. In addition, based on the LETG/ACIS-S line spectrum, we
754: find no evidence for enhanced N and depleted Mg, as inferred
755: previously from CCD spectra; indeed, if anything, N is somewhat
756: underabundant in the X-ray-emitting gas.
757:
758: The sharply nonsolar composition of the X-ray-emitting plasma is
759: similar to that determined for the present-day central star of
760: BD$+30^\circ$3639 via optical/UV spectroscopy. We conclude that the
761: plasma consists predominantly of very recently ejected gas originating
762: from nucleosynthesis processes that occurred deep within the
763: progenitor AGB star. The ``pristine'' state of this C- and Ne-enriched
764: (and Fe-depleted) plasma suggests processes such as heat conduction
765: and/or mixing between the superheated plasma and cooler, denser
766: nebular gas may not suffice to explain the fact that the inferred
767: range of X-ray emission temperatures is well below that expected for
768: shocks generated by the present-day, 700 km s$^{-1}$ central star
769: wind. Instead, it appears that the shocks detected via X-rays likely
770: result from lower-speed (300--400 km s$^{-1}$) ejections, perhaps in
771: the form of collimated jets and/or reflecting the rapid evolution of
772: the central star wind.
773:
774: The intrinsic X-ray luminosity we deduce from the modeling,
775: $\sim10^{33}$ erg s$^{-1}$, is appoximately an order of magnitude
776: larger than most previous, CCD-based estimates. This luminosity lies
777: at the very highest end of the range of $L_X$ predicted by models
778: describing either spherically-symmetric PN hot bubbles or pulsed
779: jets in symbiotic stars (e.g., Stute \& Sahai 2006, 2007),
780: suggesting that the wind collisions in BD$+30^\circ$3639 are very
781: strong indeed. Such strong wind interactions may be commonplace in
782: PN with [WC] central stars, however (K08). In these and other
783: respects, the plasma abundances, temperatures, and luminosity
784: determined from the dispersed X-ray
785: spectrum of BD+30$^{\circ}$3639 should serve both to constrain models
786: of stellar evolution and to guide the development of sophisticated
787: models of the wind interactions responsible for the superheated gas
788: within PNs.
789:
790: \vspace{0.1in}
791:
792: \acknowledgments {\it This research was supported by
793: NASA through Chandra award GO5--6008X issued to Rochester Institute
794: of Technology by the Chandra X-ray Observatory Center, which is
795: operated by Smithsonian Astrophysical Observatory for and on behalf
796: of NASA under contract NAS8--03060.}
797: % E.B. and R.N. were supported by
798: % the Israel Science Foundation, grant 28/03, and by the Asher Fund
799: % for Space Research at the Technion.}
800:
801:
802:
803: %% The reference list follows the main body and any appendices.
804: %% Use LaTeX's thebibliography environment to mark up your reference list.
805: %% Note \begin{thebibliography} is followed by an empty set of
806: %% curly braces. If you forget this, LaTeX will generate the error
807: %% "Perhaps a missing \item?".
808: %%
809: %% thebibliography produces citations in the text using \bibitem-\cite
810: %% cross-referencing. Each reference is preceded by a
811: %% \bibitem command that defines in curly braces the KEY that corresponds
812: %% to the KEY in the \cite commands (see the first section above).
813: %% Make sure that you provide a unique KEY for every \bibitem or else the
814: %% paper will not LaTeX. The square brackets should contain
815: %% the citation text that LaTeX will insert in
816: %% place of the \cite commands.
817:
818: %% We have used macros to produce journal name abbreviations.
819: %% AASTeX provides a number of these for the more frequently-cited journals.
820: %% See the Author Guide for a list of them.
821:
822: %% Note that the style of the \bibitem labels (in []) is slightly
823: %% different from previous examples. The natbib system solves a host
824: %% of citation expression problems, but it is necessary to clearly
825: %% delimit the year from the author name used in the citation.
826: %% See the natbib documentation for more details and options.
827:
828: \begin{thebibliography}{}
829:
830: \bibitem{journal}Akashi, M., Soker, N., \& Behar, E., 2006, \mnras, 368, 1706
831: \bibitem{journal}Akashi, M., Soker, N., Behar, E., \& Blondin, J., 2007, \mnras, 375, 137
832: \bibitem{journal}Akashi, M., Meiron, Y., \& Soker, N. 2008, in press, arXiv:0711.3265
833: \bibitem{journal}Anders, E., \& Grevesse, N., 1989, Geochim. Cosmochim. Acta, 53, 197
834: \bibitem{journal}Arnaud, K., Borkowski, K. J., \& Harrington,J.P., 1996, \apj, 462, L75
835: \bibitem{journal}Bachiller, R., Huggins, P. J., Cox, P., \& Forveille, T., 1991, \aap, 247, 525
836: \bibitem{journal}Bachiller, R., Forveille, T., Huggins, P. J., Cox, P., \& Maillard, J. P., 2000, \aap, 353, L5
837: \bibitem{journal}Borkowski, K. J., Balbus, S. A., \& Fristrom, C. C., 1990, \apj, 355, 501
838: \bibitem{journal}Cash, W., 1979, \apj, 228, 939
839: \bibitem{journal}Chu, Y.-H., Kwitter, K., \& Kaler, J. B. 1993, \aj, 106, 650
840: \bibitem{journal}Chu, Y.-H., Chang, T. H., \& Conway, G. M. 1997, ApJ, 482, 891
841: \bibitem{journal}Chu, Y.-H., Guerrero, M. A., Gruendl, R. A., Williams, R. M., \& Kaler, J. B., 2001, \apj, 553, L69
842: %\bibitem{journal}Corradi, R. L. M., Steffen, M., Schönberner, D., \& Jacob, R., 2007, \aap, 474, 529
843: \bibitem{journal}Cox, P., Huggins, P. J., Maillard, J.-P., Habart, E., Morisset, C., Bachiller, R., \& Forveille, T., 2002, \aap, 384, 603
844: \bibitem{journal}Dewey, D. 2002, in ``High-Resolution X-ray Spectroscopy with XMM-Newton
845: and Chandra'', ed. G. Branduardi-Raymont, Electronic Abstract 14
846: \bibitem{journal}Doron, R., \& Behar. E., 2002, \apj, 574, 518
847: \bibitem{journal}Georgiev, L. N., Richer, M. G., Arrieta, A., \& Zhekov, S. A., 2006, \apj, 639, 185
848: \bibitem{journal}Guerrero, M. A., Chu, Y.-H., \& Gruendl, R. A., 2000, ApJS, 129, 295
849: \bibitem{journal}Herwig, F., 2005, \araa, 43, 435
850: \bibitem{journal}Houck, J. C., \& Denicola, L. A., 2000, ASPC, 216, 591
851: \bibitem{journal}Kastner, J. H., Soker, N., Vrtilek, S. D., \& Dgani, R., 2000, \apj, 545, L57
852: \bibitem{journal}Kastner, J. H., Vrtilek, S. D., \& Soker, N., 2001, \apj, 550, L189
853: \bibitem{journal}Kastner, J. H., Li, J., Vrtilek, S. D., Gately, I., Merrill, K. M., \& Soker, N., 2002, \apj, 581, 1225
854: \bibitem{journal}Kastner, J. H., YU, Y. S., Houck, J., Behar. E., Nordon, R., \& Soker, N., 2006, IAUS, 234, 169
855: %%%\bibitem{journal}Kastner, J. H. 2007, in Proc. of APNIV, in press
856: \bibitem{journal}Kastner, J. H., Montez, R., Balick, B., \& De Marco, O. 2008, \apj, 672, 957
857: \bibitem{journal}Kreysing, H. C., Diesch, C., Zweigle, J., Staubert, R., Grewing, M., \& Hasinger, G. 1992, \aap, 264, 623
858: \bibitem{journal}Kwok, S., Purton, C. R., \& Fitzgerald, P. M., 1978, \apj, 219, L125
859: \bibitem{journal}Leuenhagen, U., Hamann, W. R., \& Jeffery, C. S., 1996, \aap, 312, 167
860: %\bibitem{journal}Li, J., Kastner, J. H., Prigozhin, G. Y., \& Schulz, N. S., 2003, \apj, 590, 586
861: \bibitem{journal}Lee, T.-H., \& Kwok, S., 2005, \apj, 632, 340
862: \bibitem{journal}Li, J., Harrington, J. P., \& Borkowski, K. J., 2002, \aj, 123, 2676
863: %\bibitem{journal}Lucy, L. B., 1974, \aj, 79, 745
864: \bibitem{journal}Maness, H. L., Vrtilek, S. D., Kastner, J. H., \& Soker, N., 2003, \apj, 589, 439
865: \bibitem{journal}Marcolino, W. L. F., Hillier, D. J., de Araujo, F. X., \& Pereira, C. B., 2007, \apj, 654, 1068
866: \bibitem{journal}Morrison, R., \& McCammon, D., 1983, \apj, 270, 119
867: \bibitem{journal}Murashima, M., Kokubun, M., Makishima, K., Kotoku, J., Murakami, H., Matsushita, K., Hayashida, K., Arnaud, K., Hamaguchi, K., \& Matsumoto, H. 2006, \apj, 647, L131
868: \bibitem{journal}Pwa, T. H., Pottasch, S. R., \& Mo, J. E. 1986, \aap, 164, 184
869: %\bibitem{journal}Richardson, W. H. 1972, J. Opt. Soc. Am., 62, 55
870: \bibitem{journal} Sch\H{o}nberner, D., Steffen, M., \& Warmuth, A., 2006, in Planetary Nebulae in Our Galaxy and Beyond, ed. M. J. Barlow, \& R. H. Mendez, IAU Symp., 234, 161
871: \bibitem{journal}Soker, N., 1994, MNRAS, 270, 774
872: \bibitem{journal}Soker, N., \& Kastner, J. H., 2003, \apj, 583, 368
873: \bibitem{journal}Smith, R. K., Brickhouse, N. S., Liedahl, D. A., \& Raymond, J. C., 2001, \apj, 556, L91
874: \bibitem{journal}Steffen, M., Sch\H{o}nberner, D., Warmuth A., Schwope A., Landi E., Perinotto M., \& Bucciantini N., 2005, in Szczerba R., Stasi\'{n}ska G., G´orny S. K., eds, Planetary Nebulae as Astronomical Tools, AIP Conf. Proc. Vol. 804, Melville, New York, p. 161
875: \bibitem{journal}Steffen, M., Sch\"onberner, D., \& Warmuth, A., 2008, \aap, in press
876: \bibitem{journal}Stute, M., \& Sahai, R., 2006, \apj, 651, 882
877: \bibitem{journal}Stute, M., \& Sahai, R., 2007, \apj, 665, 698
878: %\bibitem{journal}Vuong, M. H., Montmerle, T., Grosso, N., Feigelson, E. D., Verstraete, L., \& Ozawa, H., 2003, \aap, 408, 581
879: \bibitem{journal}Werner, K., \& Herwig, F., 2006, \pasp, 118, 183
880: %\bibitem{journal}Yu et al. in preparation.
881: \bibitem{journal}Zhekov, S. A., \& Perinotto, M., 1996, \aap, 309, 648
882: \bibitem{journal}Zhekov, S. A., \& Myasnikov, A. V., 1998, NewA, 3, 57
883: \bibitem{journal}Zhekov, S. A., \& Myasnikov, A. V., 2000, \apj, 543, L53
884: \end{thebibliography}
885: \clearpage
886:
887:
888:
889: %% Use the figure environment and \plotone or \plottwo to include
890: %% figures and captions in your electronic submission.
891: %% To embed the sample graphics in
892: %% the file, uncomment the \plotone, \plottwo, and
893: %% \includegraphics commands
894: %%
895: %% If you need a layout that cannot be achieved with \plotone or
896: %% \plottwo, you can invoke the graphicx package directly with the
897: %% \includegraphics command or use \plotfiddle. For more information,
898: %% please see the tutorial on "Using Electronic Art with AASTeX" in the
899: %% documentation section at the AASTeX Web site,
900: %% http://www.journals.uchicago.edu/AAS/AASTeX.
901: %%
902: %% The examples below also include sample markup for submission of
903: %% supplemental electronic materials. As always, be sure to check
904: %% the instructions to authors for the journal you are submitting to
905: %% for specific submissions guidelines as they vary from
906: %% journal to journal.
907:
908: %% This example uses \plotone to include an EPS file scaled to
909: %% 80% of its natural size with \epsscale. Its caption
910: %% has been written to indicate that additional figure parts will be
911: %% available in the electronic journal.
912:
913:
914:
915: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
916: %Figure list
917: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
918:
919: \begin{figure}
920: \epsscale{.80}
921: \centering
922: \includegraphics [scale=0.425, angle=90]{f1a.eps}
923: \vskip 0.3in
924: \includegraphics [scale=0.500, angle=90]{f1b.eps}
925: \vskip 0.1in
926: \includegraphics [scale=0.500, angle=90]{f1c.eps}
927: \caption{Dispersed spectral images of BD +30$^\circ$3639 for negative
928: LETG orders, with wavelength scales (in \AA) overlaid. Top: full range of
929: negative orders. Middle and bottom: closeups of dispersed spectral
930: images over the wavelength ranges 5-20 \AA\ and 20-35 \AA,
931: respectively. \label{dispNegaImage}}
932: \end{figure}
933: \clearpage
934:
935:
936: \begin{figure}
937: \epsscale{.80}
938: \centering
939: \includegraphics [scale=0.425, angle=90]{f2a.eps}
940: \vskip 0.3in
941: \includegraphics [scale=0.500, angle=90]{f2b.eps}
942: \vskip 0.1in
943: \includegraphics [scale=0.500, angle=90]{f2c.eps}
944: \caption{As in Fig.\ 1, for positive LETG orders. \label{dispPosiImage}}
945: \end{figure}
946: \clearpage
947:
948:
949: \begin{figure}
950: \epsscale{.80}
951: \centering
952: \includegraphics [scale=0.55, angle=270]{f3a.eps}
953: \vskip 0.2in
954: \includegraphics [scale=0.55, angle=270]{f3b.eps}
955: \caption{(a) Combined positive and negative first order LETG/ACIS-S counts
956: spectra of BD+30$^{\circ}$3639 (black) and background (red). (b)
957: Background-subtraction counts spectrum of
958: BD+30$^{\circ}$3639.\label{combFirstorder}}
959: \end{figure}
960: \clearpage
961:
962:
963: \begin{figure}
964: \epsscale{.80}
965: \centering
966: \includegraphics [scale=0.55, angle=270]{f4a.eps}
967: \vskip 0.2in
968: \includegraphics [scale=0.55, angle=270]{f4b.eps}
969: \caption{(a) Combined positive and negative first-order LETG/ACIS-S
970: counts spectrum of BD+30$^{\circ}$3639 (including background)
971: overlaid with the best-fit single-component APED model. (b) Flux-calibrated
972: first-order LETG/ACIS-S spectrum, overlaid with the same model. In
973: each panel, black shows the source spectrum and red indicates the
974: model. \label{One_2Vturb_model}}
975: \end{figure}
976: \clearpage
977:
978:
979: \begin{figure}
980: \epsscale{.80}
981: \centering
982: \includegraphics [scale=0.55, angle=270]{f5a.eps}
983: \vskip 0.2in
984: \includegraphics [scale=0.55, angle=270]{f5b.eps}
985: \caption{As in Fig.~\ref{One_2Vturb_model}, but for the best-fit
986: two-component APED model. \label{Two_comp_model}}
987: \end{figure}
988: \clearpage
989:
990:
991: \begin{figure}
992: \epsscale{.80}
993: \centering
994: \includegraphics [scale=0.35, angle=270]{f6a.eps}
995: \vskip 0.1in
996: \includegraphics [scale=0.35, angle=270]{f6b.eps}
997: \vskip 0.1in
998: \includegraphics [scale=0.35, angle=270]{f6c.eps}
999: \caption{Plots of best-fit confidence contours (68\%, dotted; 90\%, solid; and
1000: 99\%, dashed), as obtained from the two-component
1001: APED model fitting, for the column density parameter $N_{H}$ vs. the C
1002: abundance parameter (top), O abundance parameter (middle),
1003: and the lower of the two temperatures (bottom).
1004: \label{confidence_contour_vs_nH}}
1005: \end{figure}
1006:
1007:
1008: \begin{figure}
1009: \epsscale{.80}
1010: \centering
1011: \includegraphics [scale=0.22, angle=270]{f7a.eps}
1012: \vskip 0.1in
1013: \includegraphics [scale=0.22, angle=270]{f7b.eps}
1014: \vskip 0.1in
1015: \includegraphics [scale=0.22, angle=270]{f7c.eps}
1016: \vskip 0.1in
1017: \includegraphics [scale=0.22, angle=270]{f7d.eps}
1018: \vskip 0.1in
1019: \includegraphics [scale=0.22, angle=270]{f7e.eps}
1020: \vskip 0.1in
1021: \caption{As in Fig.~\ref{confidence_contour_vs_nH} for the C, Ne,
1022: N, Fe, and Mg abundance parameters (top to bottom, respectively) vs. the
1023: O abundance parameter.
1024: \label{confidence_contour_vs_O}}
1025: \end{figure}
1026:
1027:
1028: \begin{figure}
1029: \epsscale{.80}
1030: \centering
1031: \includegraphics [scale=0.35, angle=270]{f8a.eps}
1032: \vskip 0.1in
1033: \includegraphics [scale=0.35, angle=270]{f8b.eps}
1034: \vskip 0.1in
1035: \includegraphics [scale=0.35, angle=270]{f8c.eps}
1036: \caption{As in Fig.~\ref{confidence_contour_vs_nH} for C and O
1037: abundances vs.\ the lower of
1038: the two temperatures (top and middle panels) and the Ne abundance
1039: vs. the higher of the two temperatures (bottom).
1040: \label{confidence_contour_vs_T}}
1041: \end{figure}
1042:
1043: % \label{confidence_contour}}
1044:
1045: \clearpage
1046:
1047:
1048:
1049: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1050: %Table list
1051: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1052:
1053: \clearpage
1054: \input{tab1.tex}
1055: \clearpage
1056: \input{tab2.tex}
1057: \clearpage
1058: \input{tab3.tex}
1059: \clearpage
1060: \input{tab4.tex}
1061:
1062: \end{document}
1063:
1064: