1: %%
2: %% Beginning of file 'sample.tex'
3: %%
4: %% Modified 2005 December 5
5: %%
6: %%\documentclass[12pt,preprint]{aastex}
7:
8: %% manuscript produces a one-column, double-spaced document:
9:
10: \documentclass[12pt,manuscript]{aastex}
11:
12: %% preprint2 produces a double-column, single-spaced document:
13:
14: %\documentclass[preprint2]{aastex}
15:
16: %% Sometimes a paper's abstract is too long to fit on the
17: %% title page in preprint2 mode. When that is the case,
18: %% use the longabstract style option.
19:
20: %% \documentclass[preprint2,longabstract]{aastex}
21:
22: %% If you want to create your own macros, you can do so
23: %% using \newcommand. Your macros should appear before
24: %% the \begin{document} command.
25: %%
26: %% If you are submitting to a journal that translates manuscripts
27: %% into SGML, you need to follow certain guidelines when preparing
28: %% your macros. See the AASTeX v5.x Author Guide
29: %% for information.
30:
31: \newcommand{\1}{{~\sc i}}
32: \newcommand{\2}{{~\sc ii}}
33: \newcommand{\3}{{~\sc iii}}
34: \newcommand{\4}{{~\sc iv}}
35: \newcommand{\vdag}{(v)^\dagger}
36: \newcommand{\myemail}{vianney@isc.astro.cornell.edu}
37:
38: \usepackage{eucal}
39:
40: %% You can insert a short comment on the title page using the command below.
41:
42: %\slugcomment{Not to appear in Nonlearned J., 45.}
43:
44: %% If you wish, you may supply running head information, although
45: %% this information may be modified by the editorial offices.
46: %% The left head contains a list of authors,
47: %% usually a maximum of three (otherwise use et al.). The right
48: %% head is a modified title of up to roughly 44 characters.
49: %% Running heads will not print in the manuscript style.
50:
51: \shorttitle{PAH Strength and the Interstellar Radiation Field around the Massive Young Cluster NGC\,3603}
52: \shortauthors{Lebouteiller et al.}
53:
54: %% This is the end of the preamble. Indicate the beginning of the
55: %% paper itself with \begin{document}.
56:
57: \begin{document}
58:
59: %% LaTeX will automatically break titles if they run longer than
60: %% one line. However, you may use \\ to force a line break if
61: %% you desire.
62:
63: \title{PAH Strength and the Interstellar Radiation Field around the Massive Young Cluster NGC\,3603}
64:
65: %% Use \author, \affil, and the \and command to format
66: %% author and affiliation information.
67: %% Note that \email has replaced the old \authoremail command
68: %% from AASTeX v4.0. You can use \email to mark an email address
69: %% anywhere in the paper, not just in the front matter.
70: %% As in the title, use \\ to force line breaks.
71:
72: \author{V. Lebouteiller}
73: \affil{Center for Radiophysics and Space Research, Cornell University, Space Sciences Building, Ithaca, NY 14853-6801, USA}
74: \email{vianney@isc.astro.cornell.edu}
75:
76: \author{B. Brandl}
77: \affil{Leiden Observatory, University of Leiden, P.O. Box 9513, 2300 RA Leiden, Netherlands}
78: %\email{brandl@strw.leidenuniv.nl}
79:
80: \and
81:
82: \author{J. Bernard-Salas, D. Devost, and J.~R. Houck}
83: \affil{Center for Radiophysics and Space Research, Cornell University, Space Sciences Building, Ithaca, NY 14853-6801, USA}
84:
85: %% Notice that each of these authors has alternate affiliations, which
86: %% are identified by the \altaffilmark after each name. Specify alternate
87: %% affiliation information with \altaffiltext, with one command per each
88: %% affiliation.
89: %\altaffiltext{1}{Visiting Astronomer, Cerro Tololo Inter-American Observatory.
90: %CTIO is operated by AURA, Inc.\ under contract to the National Science
91: %Foundation.}
92:
93: %% Mark off your abstract in the ``abstract'' environment. In the manuscript
94: %% style, abstract will output a Received/Accepted line after the
95: %% title and affiliation information. No date will appear since the author
96: %% does not have this information. The dates will be filled in by the
97: %% editorial office after submission.
98:
99: \begin{abstract}
100: We present spatial distribution of polycyclic aromatic hydrocarbons and ionized gas
101: within the Galactic giant H\2\ region NGC\,3603. Using the \textit{IRS}
102: instrument on board the Spitzer Space Telescope, we study in particular the PAH emission
103: features at $\sim$5.7, 6.2, 7.7, 8.6, and 11.3\,$\mu$m, and the [Ar\2] 6.99\,$\mu$m,
104: [Ne\2] 12.81\,$\mu$m, [Ar\3] 8.99\,$\mu$m, and [S\4] 10.51\,$\mu$m forbidden emission
105: lines. The observations probe both ionized regions and photodissociation regions.
106: Silicate emission is detected close to the central cluster while silicate absorption is
107: seen further away. We find no significant variation of the PAH ionization fraction across
108: the whole region. The emission of very small grains lies closer to the central stellar
109: cluster than emission of PAHs. The PAH/VSG ratio anticorrelates with the hardness of the
110: interstellar radiation field suggesting a destruction mechanism of the molecules within
111: the ionized gas, as shown for low-metallicity galaxies by Madden et al.\ (2006).
112: \end{abstract}
113:
114: %% Keywords should appear after the \end{abstract} command. The uncommented
115: %% example has been keyed in ApJ style. See the instructions to authors
116: %% for the journal to which you are submitting your paper to determine
117: %% what keyword punctuation is appropriate.
118:
119: \keywords{HII regions, ISM: individual: NGC 3603, ISM: dust, ISM: atoms, ISM: molecules,
120: infrared: ISM, Telescopes: \textit{Spitzer}}
121:
122:
123: \section{Introduction}
124:
125: The investigation of star-formation feedback on the interstellar medium (ISM) in low-metallicity environments is of prime
126: importance to understand and constrain the evolution of primordial systems.
127: The most relevant diagnostics are drawn from the study of the various ISM components as a function of the physical
128: conditions in starburst-dominated objects such as dwarf star-forming galaxies and giant H\2\ regions.
129: In this perspective, the mid-infared (MIR) domain, which gives access to both gas and dust components,
130: provides one of the most powerful tools.
131:
132: Star-forming objects are characterized by strong MIR emission features attributed to polycyclic
133: aromatic hydrocarbons (PAHs)
134: in the photodissociation envelopes surrounding massive star clusters (L{\'e}ger \& Puget 1984;
135: Allamandola et al.\ 1985, 1989; Puget \& L{\'e}ger 1989; Tielens et al.\ 1999).
136: The PAHs emitting in the $3$-$13$\,$\mu$m range contain from several tens up to several hundreds carbon atoms
137: (see, e.g., Schutte et al.\ 1993). PAH molecules are mainly excited by far-UV radiation.
138: The MIR emission features are due to the subsequent fluorescence of aromatic C$-$C (especially
139: dominant at 6.2 and 7.7\,$\mu$m) and peripheral C$-$H
140: (3.3, 8.6, and 11.3\,$\mu$m) fundamental vibrational and bending modes.
141:
142:
143: The various PAH emission features are diffe\-ren\-tly affected by the local physical
144: conditions such as the hardness of the interstellar radiation field (ISRF), the dust
145: temperature, or the PAH mixture (Hony et al.\ 2001; Peeters et al.\ 2002). It has been
146: suggested that strong UV radiation is able to ionize the PAHs, while for weaker
147: radiation, PAHs can be neutral or even negatively charged by accretion of a single
148: electron (Bakes \& Tielens 1994, 1998; Salama et al.\ 1996; Dartois \& d'Hendecourt
149: 1997). Laboratory experiments indicate that neutral PAHs show stronger C$-$H mode emission
150: relatively to C$-$C modes while the inverse is true for ionized PAHs (Szczepanski \&
151: Vala 1993; Langhoff 1996; Kim et al.\ 2001; Hudgins \& Allamandolla 1999). As a result,
152: the emission features at 3.3 and 11.3\,$\mu$m are thought to mainly originate from
153: neutral PAHs while the emission features between 6 and 9\,$\mu$m are due to ionized PAHs.
154: Recently, it has been demonstrated that not only the UV radiation but also the
155: metallicity and the dust extinction are able to significantly influence the PAH
156: ionization fraction (Cox~\&~Spaans 2006).
157:
158:
159: Because of their chemical composition, PAH molecules are expected to be less abundant in low-metallicity environments.
160: The PAH intensity in galaxies has been found to correlate with the ISM metallicity (see e.g., Madden et al.\ 2006;
161: Engelbracht et al.\ 2005; Wu et al.\ 2006).
162: This correlation can be due either to the low carbon abundance, to the harder ISRF from
163: low-metallicity hot stellar atmospheres (see e.g., Schaller et al.\ 1992;
164: Schaerer \& Maeder 1992;Schaerer et al.\ 1993; Charbonnel et al.\ 1993), or to a combination of these two effects.
165: Madden et al.\ (2006) showed that the radiation field itself has an impact on the PAH survival
166: in various metallicity dwarf galaxies.
167:
168: The PAH spectrum is seen to vary not only from one object to another but also within a single
169: object. Hence it is possible to investigate the variations of the PAH spectrum as a
170: function of the physical conditions across a given region (see e.g., Joblin et al.\ 2000;
171: Vermeij et al.\ 2002; Bregman \& Temi 2005; Kassis et al.\ 2006). In a few objects, the
172: PAH intensity has been found to decrease when the ISRF hardens (Verstraete et al.\
173: 1996; Madden et al.\ 2006; Beir{\~a}o et al.\ 2006), suggesting that PAH molecules are
174: destroyed by high-energy photons.
175:
176: The \textit{Infrared Spectrograph} (\textit{IRS}; Houck et al.\ 2004) on board the Spitzer Space Telescope
177: (Werner et al.\ 2004a) opened a new perspective in the extraction of
178: small-scale regions within extended source. The wavelength range covered by the \textit{IRS} gives the possibility to
179: investigate the spatial distribution of the ionized gas, the molecular hydrogen gas, the PAHs, the silicate dust, and
180: the very small grains (VSGs). One of the most interesting application enabled by the \textit{IRS} is to understand the
181: influence of the local physical conditions, such as the hardness of the ISRF, on the molecular content.
182:
183:
184: Galactic giant H\2\ regions are ideal benchmarks for such a study. NGC\,3603 is a giant
185: H\2\ region located $\sim$7\,kpc from the Sun along the Galactic plane (Moffat 1983;
186: Moffat et al.\ 1994; Drissen et al.\ 1995; Brandl et al.\ 1999). Oxygen abundance measurements
187: range from $12 + \log (\mathrm{O}/\mathrm{H}) = 8.39$ to $8.52$ (Melnick et al.\ 1989; Tapia et al.\ 2001; Garc{\'{\i}}a-Rojas et al.\ 2006) and imply
188: a metallicity close to solar. More than 50
189: O and WR stars (Moffat et al.\ 1994) produce a Lyman continuum flux of
190: $10^{51}$\,s$^{-1}$ (Kennicutt 1984; Drissen et al.\ 1995), which is about 100 times the
191: ionizing power of the Orion Trapezium cluster. Within the core of the cluster, the system
192: HD\,97950 contains several WR, O3, and many late O stars (Melnick et al.\ 1989; Hofmann
193: et al.\ 1995). The massive stars influence the surrounding ISM morphology, notably by
194: compressing the molecular clouds through stellar winds (N{\"u}rnberger \& Stanke 2002). For
195: this reason, the geometry consists in a complex arrangement of numerous, localized H\2\
196: region $-$ PDR transitions. Although its bolometric luminosity of $L_{\rm{bol}} \sim
197: 10^7$\,L$_{\odot}$ is only about one tenth of the luminosity of 30\,Doradus, it looks
198: remarkably similar to R136, the stellar core of 30\,Doradus (Moffat et al.\ 1994).
199: NGC\,3603 has often been referenced as the most massive, optically visible H\2\ region in
200: our Galaxy, due to its high luminosity and moderately low extinction of $A_{\rm{v}} \sim
201: 4.5$ (Eisenhauer et al.\ 1998), most of which is Galactic foreground extinction.
202:
203: In this paper, we investigate with the \textit{IRS} the spatial variations of the MIR features across NGC\,3603.
204: After introducing the observations in \S\ref{sec:observations},
205: we describe the data analysis in \S\ref{sec:method}.
206: The MIR morphology is then investigated in \S\ref{sec:irac}.
207: The gas distribution and in particular the ISRF hardness is derived in \S\ref{sec:isrf}.
208: Distribution of dust and molecular features is discussed in \S\ref{sec:solid}.
209: Finally, we question the PAH survival in \S\ref{sec:pahsurvival}, and apply our results to usual MIR diagnostic
210: diagrams in \S\ref{sec:diagnostic}.
211:
212:
213: \section{Observations}\label{sec:observations}
214:
215: \begin{figure*}
216: \includegraphics[angle=0,scale=.51,clip=true]{f1a.eps}\vspace{0cm}
217: \hspace{0cm}
218: \includegraphics[angle=0,scale=.51,clip=true]{f1b.eps}\vspace{0cm}
219: \figcaption{Slit positions are superimposed on an \textit{IRAC} composite image. The colors
220: are defined by the \textit{ch1} band ($[$3.17-3.93$]$\,$\mu$m) for the red, \textit{ch3}
221: ($[$5.02-6.44$]$\,$\mu$m) for the green, and \textit{ch4} ($[$6.42-9.32$]$\,$\mu$m) for
222: the blue. For a given position, the northern slits observe only the second spectral order
223: (SL2), the southern slits observe only the first spectral order (SL1), and the middle slit
224: observations give access to both orders.
225: The local PAH emission maxima (small circles) are plotted on the left panel. The size of the circles
226: is proportional to the PAH peak intensity (sum of all bands). We also show the regions (arrows) showing an anticorrelation between
227: PAH intensity and both [Ar\3] and [S\4] line intensities.
228: \label{fig:slits}}
229: \end{figure*}
230:
231: NGC\,3603 was observed as part of the \textit{IRS} GTO program to study massive star
232: formation in Local Group giant H\2\ regions (PID\,63). Extensive imaging using the
233: \textit{Infrared Array Camera} (\textit{IRAC}; Fazio et al.\ 2004) was also peformed
234: (Brandl et al.\ 2007, in preparation), and a total of nine infrared bright sources (clusters, shock
235: fronts, protostars, etc...) have been selected for follow-up spectroscopy with the high-
236: and the low-resolution modules of the \textit{IRS} (AORKEY 12080384). The complete spectral analysis will
237: be discussed in Lebouteiller et al.\ (2007, in preparation). The long slits of the \textit{IRS}
238: low-resolution modules, however, cover a significant area of NGC\,3603 and provide
239: information on the spatial variations of spectral features across the region.
240:
241: We present an analysis of the positions \#3 to \#8 from the original follow-up
242: observations (see the coordinates in Table~\ref{tab:positions} and the image of
243: Figure\,\ref{fig:slits}). Position \#1 is an offset observation and was used for sky
244: substraction. Positions \#2 and \#9 (the latter corresponding to the IR luminous source IRS-9)
245: gave corrupted and saturated data respectively. The position \#6 is centered on the
246: central stellar cluster while the other positions are centered on bright MIR knots
247: located from $\sim$1.6 to $\sim$6\,pc away from the cluster (we use hereafter the projected distance from the
248: central stellar cluster, which represents the smallest possible distance).
249:
250: \begin{deluxetable}{llll}
251: \tablecolumns{4}
252: \tablecaption{Requested coordinates of each position.\label{tab:positions}}
253: \startdata
254: \tableline\tableline
255: \# & $\alpha$ (J2000) & $\delta$ (J2000) & PA\tablenotemark{a} \\
256: \tableline
257: 1\tablenotemark{b} &$11^\mathrm{h}14^\mathrm{m}21^\mathrm{s}.80$ & $-61^\circ24'20''.0$ & $+24^\circ.62$ \\
258: 2\tablenotemark{c} &$11^\mathrm{h}15^\mathrm{m}17^\mathrm{s}.00$ & $-61^\circ19'10''.0$ & $+24^\circ.42$ \\
259: \tableline
260: 3 & $11^\mathrm{h}14^\mathrm{m}49^\mathrm{s}.06$ & $-61^\circ17'09''.1$ & $+24^\circ.55$ \\
261: 4 & $11^\mathrm{h}14^\mathrm{m}56^\mathrm{s}.71$ & $-61^\circ12'56''.6$ & $+24^\circ.52$ \\
262: 5 & $11^\mathrm{h}14^\mathrm{m}52^\mathrm{s}.40$ & $-61^\circ15'46''.3$ & $+24^\circ.54$\\
263: 6 & $11^\mathrm{h}15^\mathrm{m}07^\mathrm{s}.40$ & $-61^\circ15'39''.2$ & $+24^\circ.48$ \\
264: 7 & $11^\mathrm{h}15^\mathrm{m}08^\mathrm{s}.03$ & $-61^\circ16'40''.2$ & $+24^\circ.48$\\
265: 8 & $11^\mathrm{h}15^\mathrm{m}02^\mathrm{s}.88$ & $-61^\circ15'51''.6$ & $+24^\circ.50$\\
266: \tableline
267: 9\tablenotemark{c} & $11^\mathrm{h}15^\mathrm{m}03^\mathrm{s}.30$ & $-61^\circ21'25''.0$ & $+24^\circ.47$ \\
268: \tableline
269: \enddata
270: \tablenotetext{a}{Position angle.}
271: \tablenotetext{b}{Position \#1 was used for sky substraction.}
272: \tablenotetext{c}{Position \#2 and \#9 gave unusable data (see text).}
273: \end{deluxetable}
274:
275: The low-resolution spectra ($\lambda/\Delta\lambda \sim 65$-$130$)
276: from the short-low (SL) module cover the spectral region 5.2-14.5\,$\mu$m.
277: The SL module contains two slits, one for each spectral order. The SL1 slit corresponds to the
278: first order (7.4-14.5\,$\mu$m) and has a size of $3.7''\times57''$ ($\leftrightarrow0.13\times1.93$\,pc$^2$ at
279: a distance of 7\,kpc). The second order (5.2-7.7\,$\mu$m) is observed through the
280: SL2 slit ($3.6''\times57''\leftrightarrow0.12\times1.93$\,pc$^2$).
281: The observations were done in staring mode\footnote{See the \textit{Spitzer/IRS} observer's manual at
282: \textit{http://ssc.spitzer.caltech.edu/documents/som/}}, consisting in two subsequent observations centered at the 1/3 (nod 1)
283: and 2/3 (nod 2) positions along the slits.
284: When the source is being observed in a given slit (\textit{nominal observation}), the other slit, corresponding to
285: the other diffraction order, performs an \textit{offset observation}.
286: Since we are dealing with sources being more extended than the slit length, we take the
287: opportunity given by the offset observations to extend our measures to larger spatial
288: scales, $\sim 2'$.
289:
290:
291:
292:
293: \section{Data analysis}\label{sec:method}
294:
295:
296: \subsection{Detector image reduction}\label{sec:reduction}
297:
298: \begin{figure*}
299: \hspace{2cm}\vspace*{1cm}\includegraphics[angle=0,width=5in,height=3.5in,clip=true]{f2.eps}
300: \figcaption{In the top panel, the full slit spectra of each position (nominal observations) are sorted as a function of the projected distance from the slit center
301: to the central stellar cluster. In the bottom panel, we show the sum of these spectra.\label{fig:fullslit}}
302: \end{figure*}
303:
304: The data were processed by the Spitzer Science Center, using the S13.2 pipeline.
305: We used the basic calibrated data (BCD) product. The background was calculated using
306: the nominal and offset observations of position \#1, and substracted from the other positions.
307:
308: Spectra were extracted from the two-dimensional flat-fielded detector image using scripts
309: within the Cornell IRS Spectroscopy Modelling Analysis and Reduction Tool environment
310: (SMART; Higdon et al.\ 2004). The two order spectra scale relatively well, and no
311: corrections were needed. The full slit spectra of the various positions and the global
312: spectrum of NGC\,3603 (simple sum of the spectra of all positions) are shown in
313: Figure\,\ref{fig:fullslit}. The final resolution is $\Delta\lambda=0.06$\,$\mu$m for SL2
314: and $0.12$\,$\mu$m for SL1.
315:
316: \begin{figure*}\hspace{1cm}
317: \includegraphics[angle=0,scale=0.7,clip=true]{f3.eps}\vspace{0.1cm}
318: \figcaption{Spectral image of the nominal observation of position \#3. The flux (light scale) is plotted as a function of the spatial position
319: within the slit (y-axis) and of the wavelength (x-axis). The segment on the right side indicates the size
320: of the extraction windows.\label{fig:spectral_image}}
321: \end{figure*}
322:
323:
324: \subsection{Extraction of small-scale regions}\label{sec:smallscale}
325:
326: Spatial variations can be observed within the long slits (see
327: Figure\,\ref{fig:spectral_image}), with regions emitting strong forbidden lines and no
328: PAHs, and regions showing intense PAH features together with relatively weak lines. To
329: investigate these spatial variations, we extracted columns, fixed to 4 pixels wide,
330: within the long slit on the basis that the full width at half maximum of the point spread
331: function (PSF) is $\approx$2 pixels. MIR sources above or below the slit, or on each side
332: of the extraction window within the slit, can contribute (because of the PSF) to the
333: total flux in a given chosen extraction window. Hence the spatial column actually
334: extracted corresponds to a region somewhat larger than 0.24$\times$0.12\,pc$^2$ $-$ which
335: would be the size of the extracted spatial region if the PSF width was null. For this
336: reason, it is particularly difficult to estimate the related absolute uncertainty on the flux
337: calibration. Given the fact that our study is focused on relative spatial variations, we
338: consider this caveat as being of little importance as far as the interpretations are
339: concerned. Comparing the fluxes of same spatial positions observed at different locations
340: within the slit, we find relative errors ranging from $\pm$7\% (position \#3, lowest integrated flux)
341: to $\pm$4\% (position \#7, largest flux).
342:
343: We extracted overlapping extraction windows along the slits, shifting by one pixel
344: between each other (which make a total of 37 different columns, and as many data points,
345: for a given nominal or offset observation). This is not an actual oversampling of the
346: spatial PSF. Indeed, whatever the sampling is, one is limited by the PSF convolution. The
347: spatial profile of spectral feature emission along the slit are smoothed by the PSF so
348: that the choice of the number of extracting windows only results in different
349: samplings of the $-$ same $-$ eventually smoothed spatial profile. Finally, it must be
350: added that variations of features in the observed (PSF convolved) spatial profile imply
351: in reality even larger variations.
352:
353:
354:
355: \subsection{Measurements}\label{sec:measurements}
356:
357: We observe several PAH features and forbidden emission lines superimposed on a spectral continuum which in our case is
358: dominated by thermal emission of dust and broad silicate emission/absorption.
359: Several windows in positions \#3, \#4, \#5, and \#7 show silicate absorption around 10\,$\mu$m
360: while positions \#6 and \#8 show silicate emission in the same range (see Fig.\,\ref{fig:example}).
361: Note that the silicate features appear less prominent in the full slit spectra of Figure\,\ref{fig:fullslit} due
362: to the contribution of several distinct physical regions within the slit.
363: Only the position \#6 shows signs of a stellar emission continuum rising toward wavelengths shorter
364: than $\sim$8\,$\mu$m (see Fig.\,\ref{fig:fullslit}).
365:
366: \begin{figure}[b!]\vspace*{0.2cm}
367: \includegraphics[angle=0,scale=.47,clip=true]{f4a.eps}\\
368: \includegraphics[angle=0,scale=.47,clip=true]{f4b.eps}
369: \figcaption{Examples of the spectral fit of a 4\,pixels window in positions \#3 (top) and \#8 (bottom).
370: The various dust continua, PAH components, and forbidden emission lines can be identified.
371: We also show the coverage of the \textit{IRAC}/\textit{ch3} and \textit{ch4} bands.
372: The \textit{ch1} and \textit{ch2} bands observe at wavelengths shorter than 5\,$\mu$m.\label{fig:example}}
373: \end{figure}
374:
375: PAH features are detected at $\sim$5.7, $\sim$6.2, $\sim$7.7, $\sim$8.6, $\sim$11.3, and
376: $\sim$12.7\,$\mu$m. A weak feature at $\sim$7.9\,$\mu$m can be observed for some
377: positions as a shoulder of the 7.7\,$\mu$m PAH.
378:
379: The following forbidden emission lines are detected, [Ne\2] (21.6\,eV) at 12.81\,$\mu$m,
380: [Ar\2] (15.8\,eV) at 6.99\,$\mu$m, [Ar\3] (27.6\,eV) at 8.99\,$\mu$m, and [S\4]
381: (34.8\,eV) at 10.51\,$\mu$m. The [Ne\2] line is blended with the PAH feature at
382: 12.7\,$\mu$m. The [Ar\2] line could possibly be blended with a weak PAH feature at
383: 7.0\,$\mu$m. This contamination is probably not larger than $\approx$20\% of the total
384: line flux (see the measurements of Sturm et al.\ 2000; F{\"o}rster Schreiber et al.\
385: 2001).
386:
387: \begin{figure*}\vspace{-0.5cm}
388: \includegraphics[angle=0,width=3.1in,height=2.2in,clip=true]{f5a.eps}\hspace{0.5cm}
389: \includegraphics[angle=0,width=3.1in,height=2.2in,clip=true]{f5b.eps}\vspace{0.5cm}\newline
390: \includegraphics[angle=0,width=3.1in,height=2.2in,clip=true]{f5c.eps}\hspace{0.5cm}
391: \includegraphics[angle=0,width=3.1in,height=2.2in,clip=true]{f5d.eps}\vspace{0.5cm}\newline
392: \includegraphics[angle=0,width=3.1in,height=2.2in,clip=true]{f5e.eps}\hspace{0.5cm}
393: \includegraphics[angle=0,width=3.1in,height=2.2in,clip=true]{f5f.eps}\vspace{0.2cm}
394: \figcaption{Spatial variations of the spectral feature intensities. Integrated flux is
395: plotted against the declination for each slit position. Data on the the left side of each
396: plot correspond to the southern offset observation (first spectral order only,
397: 5.2-7.7\,$\mu$m), and data on the right to the northern observation (second spectral
398: order only, 7.4-14.5\,$\mu$m). Legend is displayed in position \#7. The bottom panels show the relative errors.
399: The black rectangles along the declination axis indicate regions where PAH intensities spatially anticorrelate
400: with [Ar\3] and [S\4] line intensities. The large vertical gray rectangles indicate zones
401: with no coverage.
402: \label{fig:pro3}}
403: \end{figure*}
404:
405: The intensity measurements were done using the spectral fitting algorithm PAHFIT (Smith et al.\ 2007).
406: This procedure is especially suited for mid-IR low-resolution spectra of star-forming regions dominated by PAH emission.
407: The modelled spectral continuum is computed from a combination of starlight and 8 preset
408: (fixed temperature blackbodies) thermal dust continua,
409: whose strength is allowed to vary.
410: The unresolved forbidden emission lines are represented by Gaussian profiles, while the PAH features are decomposed in individual
411: blended Drude profiles. The main components we consider in this study are the PAH bands at
412: 5.70\,$\mu$m, 6.22\,$\mu$m, 7.60+7.85\,$\mu$m (which we will refer from now on as the 7.7\,$\mu$m feature), 8.61\,$\mu$m, and 11.23+11.33\,$\mu$m (11.3\,$\mu$m feature).
413: The absorption by silicate dust, as opposed to its emission, is accounted for in the PAHFIT
414: calculations. We decided to model the silicate emission in positions \#6 and \#8 by including two
415: broad Gaussian components, centered at 9.5 and 10.5\,$\mu$m. This is an arbitrary choice
416: whose main purpose is to be able to measure the superimposed features with the smallest
417: uncertainty possible. We integrated the PAHFIT algorithm in an automatic script able to
418: handle the large number of spectra ($>$600) of our dataset. Examples of spectral fittings
419: are presented in Figure\,\ref{fig:example}. The small-scale variations of the MIR
420: spectral feature intensities across the slit positions are given in
421: Table~\ref{tab:measures} and plotted in Figure\,\ref{fig:pro3}.
422:
423: Finally, in order to trace the presence of VSGs, which are expected to dominate the
424: continuum emission longward of $\lambda\sim10$\,$\mu$m (D{\'e}sert et al.\ 1990), we
425: measure the average flux between 13.6 and 14.0\,$\mu$m,
426: where there is no contamination by line emission and PAH features.
427:
428:
429:
430: \begin{deluxetable}{llllllll}
431: \tablecolumns{4}
432: \tablecaption{Flux of the lines and the PAH features.\label{tab:measures}}
433: \startdata
434: \tableline\tableline
435: \# & $\alpha$, $\delta$ (J2000) & d\tablenotemark{a} &5.7\,$\mu$m & 6.2\,$\mu$m & 7.7\,$\mu$m & 8.6\,$\mu$m & 11.3\,$\mu$m \\
436: & (deg) & (pc) & & & & & \\
437: \tableline
438: 3 & $168.70210,-61.2883 $ & $ 6.000$ &$ 0.2831$ &$ 1.9817 $ & $ 4.2977 $ & $ 1.0332 $ & $ 1.5536 $\\
439: & & & (0.0516) & (0.0739) & (0.1505) & (0.0487) & (0.1001) \\
440: 3 & $168.70244,-61.2880 $ & $ 5.958$ &$ 0.3213$ &$ 2.1403 $ & $ 4.6732 $ & $ 1.2015 $ & $ 1.8042 $\\
441: & & & (0.0531) & (0.0677) & (0.1391) & (0.0456) & (0.0970) \\
442: 3 & $168.70644,-61.2838 $ & $ 5.466$ &$ 0.1318$ &$ 1.0722 $ & $ 1.7679 $ & $ 0.3879 $ & $ 0.8580 $\\
443: & & & (0.0495) & (0.0395) & (0.1166) & (0.0420) & (0.0480) \\
444: %3 & $168.70166,-61.2888 $ & $ 6.055$ &$ 0.2618$ &$ 1.8428 $ & $ 3.9170 $ & $ 0.9376 $ & $ 1.3967 $\\
445: % & & & (0.0401) & (0.0562) & (0.1310) & (0.0482) & (0.0769) \\
446: \multicolumn{8}{c}{\textit{(truncated)}} \\
447: % & ... & ... & ... & ... & ... & ... & ... \\
448: \tableline\tableline
449: \# & $\alpha$, $\delta$ (J2000) & d\tablenotemark{a} & [Ar\2] & [Ne\2] & [Ar\3] & [S\4] & \\ % cont. @14\,$\mu$m \\
450: & (deg) & (pc) & & & & \\
451: \tableline
452: 3 & $ 168.70210,-61.2883 $ & $ 6.000 $ & $ 0.0350 $ & $ 0.4403 $ & $ 0.1572 $ & $ 0.2307 $ \\
453: & & & (0.0085) & (0.0103) & (0.0214) & (0.0247) \\
454: 3 & $ 168.70244,-61.2880 $ & $ 5.958 $ & $ 0.0371 $ & $ 0.5342 $ & $ 0.1861 $ & $ 0.2610 $ \\
455: & & & (0.0074) & (0.0098) & (0.0155) & (0.0225) \\
456: 3 & $ 168.70644,-61.2838 $ & $ 5.466 $ & $ 0.0333 $ & $ 0.6071 $ & $ 0.2992 $ & $ 0.5256 $\\
457: & & & (0.0071) & (0.0063) & (0.0150) & (0.0220) \\
458: %3 & $ 168.70166,-61.2888 $ & $ 6.055 $ & $ 0.0271 $ & $ 0.3811 $ & $ 0.1396 $ & $ 0.2221 $ \\
459: % & & & (0.0067) & (0.0085) & (0.0135) & (0.0214) \\
460: \multicolumn{8}{c}{\textit{(truncated)}} \\
461: % & ... & ... & ... & ... & ... & ... & \\
462: \tableline
463: \enddata
464: \tablenotetext{a}{Distance from the central stellar cluster ($\alpha=11^\mathrm{h}15^\mathrm{m}07^\mathrm{s}.966$, $\delta=-61^\circ15'30''.348$).}
465: \tablecomments{Fluxes are expressed in $\times10^{20}$~W\,cm$^{-2}$. Numbers between brackets represent the uncertainties.}
466: \end{deluxetable}
467:
468:
469: \section{Interpretation of the \textit{IRAC} image}\label{sec:irac}
470:
471: The \textit{IRAC} image of NGC\,3603 (Fig.\,\ref{fig:slits}) reveals a complex MIR
472: morphology that the spectroscopic results of the \textit{IRS} in Figure\,\ref{fig:pro3}
473: can help us to understand.
474:
475: We expect the \textit{IRAC/ch1} band to be dominated by stellar continuum emission and
476: by the 3.3\,$\mu$m PAH feature, which is seen to scale with the 11.3\,$\mu$m feature
477: in various objects (Hony et al.\ 2001). As an illustration, the
478: full slit spectrum of the position \#6, which is centered on the central stellar cluster,
479: is strongly dominated by stellar emission at wavelengths shorter than 8\,$\mu$m
480: (Fig.\,\ref{fig:fullslit}). Furthermore, Figure\,\ref{fig:slits} shows that
481: stars indeed emit mostly in the \textit{ch1} band.
482:
483: The regions where PAH intensity is the largest are bright in both \textit{ch1} and \textit{ch3} bands.
484: This is because of the presence of PAH features (3.3\,$\mu$m and 6.2\,$\mu$m) within these bands.
485:
486: The \textit{ch4} band is sensitive to the presence of the PAH features at 7.7 and
487: 8.6\,$\mu$m, and to the forbidden emission lines [Ar\2] and [Ar\3]. By comparing the \textit{IRAC}
488: image with the results of the \textit{IRS}, we notice that the regions showing a flux
489: excess in the \textit{ch4} band as compared to the other bands are also those showing
490: a relatively intense [Ar\3] line in the \textit{IRS} spectra.
491:
492:
493:
494:
495: \section{Gas distribution}\label{sec:isrf}
496:
497:
498: The forbidden line intensities vary significantly across each observation.
499: We assume on first approximation that the metallicity is uniform within the region and that it does not
500: affect the spatial variations we observe. Preliminary results on abundance determinations in NGC3603
501: confirm the validity of this assumption (Lebouteiller et al.\ 2007, in preparation). The variations are neither due to depletion
502: effects, at least for argon and neon which are not expected to be
503: depleted onto dust grains. The depletion of sulfur is more uncertain. Abundance
504: determinations in H\2\ regions and planetary nebul\ae\ suggest that sulfur depletion
505: ranges from null to $-0.5$\,dex with respect to the solar abundance (Pottasch \&
506: Bernard-Salas 2006; Henry et al.\ 2004; Marigo et al.\ 2003; Mart{\'i}n-Hern{\'a}ndez et
507: al.\ 2002). On the other hand, the line intensity spatial variations and, to an even larger extent, the line ratios, are sensitive to the ionization
508: structure. Since both high-ionization lines [Ar\3] and [S\4] follow the same trend, the depletion of sulfur should not
509: dominate the spatial variations of [S\4] line intensity.
510:
511: Because the line ratios implying species with large differences in their
512: ionization potentials can actually probe different regions along the sightline, we expect
513: \textit{a priori} the ratios [S\4]/[Ne\2], [S\4]/[Ar\3], and [Ar\3]/[Ne\2] to be our most
514: reliable tracers of the transversal spatial profile of the ISRF hardness.
515: The usual MIR estimators used in the literature are [Ne\3]/[Ne\2] and [S\4]/[S\3], which
516: we cannot measure with our dataset. In order to compare the reliability of our
517: estimators, we used the results of the stellar population and photoionization code
518: developed by Guiles et al. (2004). They used Starburst\,99 (Leitherer et al.\ 1999) and
519: MAPPINGS IIIq (Dopita et al.\ 2002; Sutherland et al.\ 2002) to derive the
520: variation of the MIR photoionization lines as a function of the characteristics of the
521: stellar population. We computed the line ratios in environments with
522: metallicities between 1/3 and twice solar. The stellar population in the models was
523: defined by an instantaneous burst, with a Salpeter initial mass function (Salpeter 1955),
524: and an upper mass cut-off of 100\,$M_\odot$.
525:
526: The models predict that the [S\4]/[S\3] ratio traces [Ne\3]/[Ne\2], as expected, and that the [S\4]/[Ne\2],
527: [S\4]/[Ar\3], and [Ar\3]/[Ne\2] ratios correlate as tightly with [Ne\3]/[Ne\2]. They can
528: be thus considered reliable tracers of the ISRF hardness in the environments considered.
529: From now on, we will make use of [S\4]/[Ne\2], principally because this ratio involves
530: the two most prominent forbidden emission lines in our spectra.
531:
532:
533: \begin{figure}[b!]
534: \includegraphics[angle=0,scale=0.39]{f6a.eps}\vspace*{-0.2cm}\\
535: \includegraphics[angle=0,scale=0.39,]{f6b.eps}\vspace*{-0.2cm}\\
536: \includegraphics[angle=0,scale=0.39]{f6c.eps}
537: \figcaption{[S\4] (top), [Ne\2] (middle) intensities, and the [S\4]/[Ne\2] ratio (bottom)
538: are plotted as a function of the projected distance from the central cluster. Intensities are expressed in $\times10^{20}$~W\,m$^{-2}$.
539: In order to minimize the projection effects, results in each distance bin of 0.5\,pc were error-weighted and averaged.
540: The vertical dotted lines refer to the distance marks plotted in Figure\,\ref{fig:slits}.
541: \label{fig:sivneii_dist}}
542: \end{figure}
543:
544: We plot in Figure\,\ref{fig:sivneii_dist} the distribution of [S\4], [Ne\2], and their
545: ratio as a function of the projected distance from the central cluster.
546: It can be seen that $I$([S\4]) shows a sharp increase between 0 and 1.5\,pc, and then decreases
547: progressively until distances larger than 8\,pc. The lack of emission at very small distances from the cluster is
548: probably a consequence of the stellar winds that washed out the surrounding ISM to create a cavity.
549: The image of Figure\,\ref{fig:slits} shows indeed the total absence of interstellar MIR emission in the immediate vicinity
550: of the central cluster. Similarly to [S\4], the [Ne\2] line intensity is relatively weak at distances $\lesssim1$\,pc but
551: then shows a fairly flat emission until $\sim$6\,pc.
552: The shallow distribution of $I$([Ne\2]) suggests that [Ne\2] emission is more extended than that of [S\4]. The [S\4]/[Ne\2]
553: ratio shows roughly the same behavior as the [S\4] line intensity, and implies that the ISRF hardness decreases by a factor $\sim$20
554: toward the outer parts of the giant H\2\ region.
555:
556:
557:
558:
559: \section{Dust and molecule distribution}\label{sec:solid}
560:
561: \subsection{Silicate dust}\label{sec:silicate}
562:
563: Amorphous silicate dust shows a broad spectral feature centered on 9.7\,$\mu$m, originating from the
564: stretching mode of the Si$-$O bond (see, e.g., Knacke \& Thomson 1973).
565: While it is mostly observed in absorption in astrophysical objects, silicate emission has also been detected
566: in a few H\2\ regions, including the Orion nebula, and is thought to be due to grains with size $\gtrsim$0.01\,$\mu$m
567: heated to $\sim$100\,K (Cesarsky et al.\ 2000; Contursi et al.\ 2000).
568: In NGC\,3603, we detect silicate dust in absorption in the spectra of positions \#3, \#4, \#5, and \#7.
569: Silicate in emission is observed in the spectra of the other positions, \#6 and \#8.
570: Examples of various silicate profiles across the region are presented in Figure\,\ref{fig:si_compa}.
571:
572: \begin{figure}[b!]
573: \includegraphics[scale=0.41,clip=true]{f7.eps}
574: \figcaption{The 9.7\,$\mu$m is detected in emission and in absorption. The two bottom spectra are from position \#3, the two on the
575: top from the position \#8, and the middle one is from position \#7.\label{fig:si_compa}}
576: \end{figure}
577:
578:
579:
580: We measured the apparent strength of the 9.7\,$\mu$m feature using the method of Spoon et al.\ (2007).
581: This method requires to choose a local MIR continuum, and to calculate the ratio between the observed flux at 9.7\,$\mu$m
582: and the continuum value. The silicate strength, as defined by the equation (1) of Spoon et al.\ (2007), is positive for
583: silicate emission, and negative for silicate absorption.
584: The spatial distribution of the silicate strength in NGC\,3603 is presented in
585: Figure\,\ref{fig:silicate}. Silicate dust is seen in emission relatively close to the
586: central cluster, while it is detected in absorption further away.
587:
588: It is interesting to notice that the silicate emission is observed around the same distances to the cluster as the VSG emission (\S\ref{sec:vsgs}).
589: The transition zone between silicate emission and silicate absorption is
590: located between 2 and 3.5\,pc away from the cluster. This corresponds to the region where the ISRF hardness decreases
591: significantly (\S\ref{sec:isrf}) and where the PAH emission begins to dominate the spectra.
592: Finally, we observe that the most prominent silicate absorption features seem to correspond to bright PAH emission regions.
593:
594: \begin{figure}[b!]
595: \includegraphics[scale=0.31,clip=true]{f8.eps}
596: \figcaption{Silicate strength is plotted as a function of the distance from the central
597: cluster. See Figure\,\ref{fig:sivneii_dist} for the plot description.\label{fig:silicate}}
598: \end{figure}
599:
600:
601:
602: \subsection{PAHs}\label{sec:pahs}
603:
604: \begin{figure}[b!]
605: \includegraphics[scale=0.39,angle=0,clip=true]{f9a.eps}\\
606: \includegraphics[scale=0.39,angle=0,clip=true]{f9b.eps}
607: \figcaption{(a) Intensity of the 11.3\,$\mu$m feature (neutral PAHs) is plotted against the sum of
608: the 6.2, 7.7, and 8.6\,$\mu$m (cations) feature intensities. (b) The ratio 11.3/(6.2+7.7+8.6) is plotted against the
609: distance to the stellar cluster. See Figure\,\ref{fig:sivneii_dist} for the plot description.\label{fig:pahpah1}}
610: \end{figure}
611:
612: %?
613: For the following analysis, we consider hereafter that extinction has a negligible effect on the PAH spectrum from one position
614: to another, or across a given observation (although it certainly has an effect on a single spectrum,
615: see e.g., Peeters et al.\ 2002).
616:
617: The peak wavelength of the PAH features at 6.2, 7.7, and 8.6\,$\mu$m is expected to possibly shift, depending mostly on the molecule properties
618: (size, symmetry, ...) as opposed to the 11.3\,$\mu$m profile which is found invariably peaking at the same wavelength
619: (see, e.g., Hony et al.\ 2001).
620: However, we find that the peak of all the PAH profiles does not vary by more
621: than one resolution element across the observations of NGC\,3603 (0.06\,$\mu$m for the 5.7 and 6.2\,$\mu$m features, 0.12\,$\mu$m for
622: the 8.6 and 11.3\,$\mu$m features), the only exception being the PAH feature at 7.7\,$\mu$m which shows a somewhat larger distribution,
623: centered on 7.64$\pm$0.24\,$\mu$m. %7.58 +/- 0.22 std dev
624: These results imply that the molecule properties do not change significantly across the region.
625: In particular, the invariance of the 6.2\,$\mu$m profile peak, centered on
626: 6.23$\pm$0.06\,$\mu$m, suggests the presence of complexed and/or substituted carriers
627: (by introduction of an hetero atom in place of a carbon atom in the skeleton; Hudgins et al.\ 2005; Peeters et al.\ 2002). %6.23 +/- 0.03 std dev
628: %11.26 +/- 0.075
629: The spectra across NGC\,3603 identify with the "class $\mathcal A$" in the classification of Peeters et al.\ (2002), i.e., where
630: the PAH features actually peak at $\approx$6.22, $\approx$7.6, and $\approx$8.6\,$\mu$m. This class is populated by H\2\ regions, reflection nebul\ae\ and
631: most extragalactic sources.
632:
633:
634: \begin{figure}[b!]
635: \includegraphics[scale=0.39,angle=0,clip=true]{f10a.eps}\\
636: \includegraphics[scale=0.39,angle=0,clip=true]{f10b.eps}
637: \figcaption{(a) The summed intensity of the 6.2 and 7.7\,$\mu$m PAH features (C$-$C mode) is plotted
638: against the sum of the 8.6 and 11.3\,$\mu$m feature (C$-$H mode) intensities. (b) The ratio (6.2+7.7)/(8.6+11.3) is plotted against the distance to the stellar cluster.
639: See Figure\,\ref{fig:sivneii_dist} for the plot description.\label{fig:pahpah2}}
640: \end{figure}
641:
642: On first approximation, the PAH spectra across the region look identical, with the
643: intensities of the various features scaling together. This suggests, together with the
644: presence of both neutral (11.3\,$\mu$m) and ionized (6.2, 7.7, 8.6\,$\mu$m) features in the spectra, that the PAH
645: ionization fraction is relatively constant. In order to investigate in more details the influence of
646: ionization on the PAH spectra, we compare in Figure\,\ref{fig:pahpah1}a the intensity
647: $I$(11.3) with $I$(6.2)+$I$(7.7)+$I$(8.6). There is a tight correlation, implying that
648: the neutral/ionized mixture does not vary significantly across the region.
649: More particularly, the ratio is essentially constant as a function of the distance to the cluster (Fig.\,\ref{fig:pahpah1}b).
650: We attribute the constant ionization fraction to the electron recombination rate in the ionized region which
651: balances the PAH ionization rate.
652:
653: Figure\,\ref{fig:pahpah2}a shows that emission bands dominated by C$-$C modes and C$-$H modes correlate with each other.
654: This correlation hold for the range of [S\4]/[Ne\2] ratios probed across the region (\S\ref{sec:isrf}). There is no correlation with the distance to
655: the cluster (Fig.\,\ref{fig:pahpah2}b). These findings are consistent with the fact that the PAH ionization fraction
656: is constant across NGC\,3603, since ionized PAHs should show enhanced C$-$C mode emission
657: (see introduction).
658:
659:
660:
661:
662:
663: \subsection{Very small grains}\label{sec:vsgs}
664:
665: While PAH emission originates in PDRs (Tielens 1993; Sellgren et al.\ 1990), VSG
666: emission is seen to peak in ionized regions (Cesarsky et al.\ 1996; Verstraete et al.\
667: 1996). In order to check whether these results hold for NGC\,3603, we estimated the VSG
668: emission by measuring the continuum flux at 14\,$\mu$m. The VSG emission spans a wide range
669: of values across the region (Fig.\,\ref{fig:vsg_compa}).
670:
671: \begin{figure}[b!]
672: \includegraphics[scale=0.41,clip=true]{f11.eps}
673: \figcaption{Example of various spectra across NGC\,3603, showing the diversity of the VSG continuum.\label{fig:vsg_compa}}
674: \end{figure}
675:
676: We find that the VSG continuum intensity scales tightly with [S\4] line intensity (Fig.\,\ref{fig:nice}).
677: Since [S\4] line intensity is seen to peak close
678: to the central cluster (Fig.\,\ref{fig:sivneii_dist}), we conclude that VSGs also emit mostly in these regions
679: and that they are the dominant dust component in the ionized region.
680:
681: \begin{figure}[b!]
682: \includegraphics[angle=0,scale=0.39,clip=true]{f12.eps}
683: \figcaption{$I$([S\4]), expressed in $\times10^{-20}$~W\,m$^{-2}$, is plotted against the VSG emission (in Jy).
684: \label{fig:nice}}
685: \end{figure}
686:
687:
688:
689:
690:
691: \section{PAH survival}\label{sec:pahsurvival}
692:
693: \subsection{Comparison with VSG emission}\label{sec:pahvsg}
694:
695: The difference between the spatial emission of PAH and VSG can be seen in Figure\,\ref{fig:pahvsg_dist}, where we plot the
696: PAH/VSG ratio as a function of the distance to the cluster.
697:
698: \begin{figure}[b!]
699: \includegraphics[angle=0,scale=0.39,clip=true]{f13.eps}
700: \figcaption{The PAH/VSG ratio is plotted against the distance to the central cluster. See Figure\,\ref{fig:sivneii_dist} for the plot description.
701: We use the PAH features at 8.6 and 11.3\,$\mu$m only since we plot here the results of the nominal and SL1 offset observations.
702: \label{fig:pahvsg_dist}}
703: \end{figure}
704:
705: VSG and PAH emission do not coexist spatially, the VSGs emitting
706: mostly in the ionized region (although they can be present elsewhere but without being
707: excited). The photons exciting VSGs also illuminate PAHs, thus the PAH molecules may not
708: survive the physical conditions required to heat VSGs (Madden et al.\ 2006). In order to
709: investigate the relation between PAH emission and the hardness of the ISRF in more
710: details, we use a similar approach to that of Madden et al.\ (2006) by comparing the PAH
711: emission with VSG emission. We show in Figure\,\ref{fig:pahnom} the variation of the
712: PAH/VSG intensity ratio as a function of [S\4]/[Ne\2]. It can be seen that PAH emission
713: becomes globally weaker when the ISRF becomes harder.
714:
715: The PAH features dominated by C-C or C-H emission modes do not show significant
716: differences across the region (\S\ref{sec:pahs}). Thus, dehydrogenation (rupture of the
717: C$-$H bonds by absorption of UV photons) is unlikely to be responsible for the lack of
718: PAHs relative to dust in regions where the ISRF is the hardest. On the other hand, PAH
719: molecules could be destroyed by high-energy photons.
720:
721: \begin{figure}[b!]
722: \includegraphics[angle=0,scale=0.39,clip=true]{f14.eps}
723: \figcaption{The PAH intensity (sum of the 8.6, and 11.3\,$\mu$m features) is plotted against the ISRF hardness probed by the [S\4]/[Ne\2] ratio.
724: \label{fig:pahnom}}
725: \end{figure}
726:
727:
728: \subsection{Presence of PDRs}
729:
730: We find that [Ar\2] and [Ne\2] emission peaks always correlates with PAH emission peaks.
731: We find no regions where PAH emission anticorrelates with both $I$([Ar\2]) and $I$([Ne\2]).
732: In contrast, some regions show an anticorrelation between PAH intensity and $I$([Ar\3]) and $I$([S\4]) (see their identification in Fig.\,\ref{fig:pro3} and their corresponding locations in Fig.\,\ref{fig:slits}).
733: Within these regions, the location of the ion peak emission always follows a clear structure, being correlated
734: with the ionization potential, and implying a sharp variation of the hardness of the ISRF.
735:
736: We identify these transitions as being interfaces between ionized region and PDRs.
737: The interfaces toward the ionized region regime are all located in the direction the central cluster
738: (Fig.\,\ref{fig:slits}). The size of the interfaces (between the maximal and minimal PAH
739: intensities) is respectively 0.48, 0.72, 0.57, 0.66, and 0.40\,pc (note that this should
740: not be confused with the size of the PDR).
741:
742:
743: \subsection{Photodestruction}\label{sec:pahdestruction}
744:
745: The energy deposited via photon heating is potentially large enough to dissociate bonds
746: within PAHs. This has been the explanation to the PAH variations in the Orion bar and in
747: the Red Rectangle nebula (Geballe et al.\ 1989). The Ar\2\ ion exists for energies larger
748: than 15.8\,eV and it is the dominant ionization stage in regions where PAH intensity is
749: maximal. This means that the far-UV radiation responsible for the dominant presence of
750: Ar\2\ (from 15.8 to 27.6\,eV which is the ionization potential of Ar\2) is not able to
751: destroy efficiently the PAHs, at least for the intensity of the ISRF at this energy.
752: Although the energy required to break C$-$H or C$-$C bonds is 4.8\,eV and $\sim$8.5\,eV
753: respectively (Omont 1986), photons with higher energy may be needed to dissociate bonds
754: in a large molecule. Using the models of Omont (1986), we find that bonds can be broken
755: by $\lesssim27.6$\,eV photons only in a molecule smaller than $\sim$25-50 C-atoms,
756: depending on the bond type. Note that the threshold size varies from one model to
757: another, and the values we derive should only give a first order approximation. Hence,
758: one possibility is that PAH molecules are large enough to prevent dissociation to occur.
759: Another possibility requires that PAH molecules are small, and that the energy density is
760: relatively weak.
761:
762: Can we infer a maximal size for the PAHs in NGC\,3603? The UV radiation required for the
763: dominant presence of Ar\3\ (27.6-40.7\,eV) and S\4\ (34.8-47.3\,eV) ions could be
764: responsible for PAH molecule destruction. Single photons with energies 40.7\,eV
765: (47.3\,eV) are able to break bonds in molecules smaller than $\sim$40-75 ($\sim$50-85)
766: C-atoms, depending on the bond type. Furthermore, it also implies that the energy density
767: is large enough. Although even larger PAHs may be present, it is difficult to set an
768: upper limit on their size. However, since PAH emission is almost zero in regions where
769: Ar\3\ and S\4\ are dominant, it is likely that most PAHs have sizes smaller than $\sim$85
770: C-atoms, at least in regions where the ISRF is the hardest.
771:
772:
773:
774: \section{MIR diagnostic diagram}\label{sec:diagnostic}
775:
776:
777: \begin{figure*}\hspace{0.5cm}
778: \includegraphics[angle=0,scale=.6,clip=true]{f15.eps}\vspace{0.2cm}
779: \figcaption{MIR diagnostic diagram.
780: The dashed and dotted curves indicate
781: 50-50\% and 25-75\% contributions respectively. Small filled circles correspond to individual column spectra.
782: The stars refer to the columns within position \#6 which are centered on the central stellar cluster.
783: Large filled circles represent the measure using full slit extractions. The diamond stands for the measure of the global
784: summed spectrum of NGC\,3603.
785: \label{fig:laurent}}
786: \end{figure*}
787:
788: The small-scale variations of the spectral feature intensities across NGC\,3603 show the
789: transition between regions dominated by PAH emission and regions dominated by
790: high-ionization lines. Our results should fit in MIR diagnostic diagrams used to
791: distinguish between active galactic nuclei (AGN), PDR, and ionized region behaviors.
792:
793: In Figure\,\ref{fig:laurent}, we present a diagram, based on Laurent et al.\ (2000) and
794: Peeters et al.\ (2004). The templates we use are the Galactic reflection nebula NGC7023
795: (Werner et al.\ 2004b; PDR excited by optical photons), the H\2\ region M17 - position
796: \#2 (Peeters et al.\ 2004), and the quasar 3C273. Accordingly to the approach of Laurent
797: et al.\ (2000) and Peeters et al.\ (2004), we plot the continuum ratio between
798: $\sim$13.8\,$\mu$m (average flux between 13.6 and 14.0\,$\mu$m) and $\sim$6.0\,$\mu$m
799: (average between 5.9 and 6.1\,$\mu$m) against the ratio of the 6.2\,$\mu$m PAH intensity
800: over the continuum level at $\sim$6.0\,$\mu$m.
801:
802: Our data points form a relatively narrow stripe, probing observationally the ionized region - PDR transition (Fig.\,\ref{fig:laurent}).
803: The results of positions \#7 and \#8 lie around the ionized region template, while the results of positions
804: \#3, \#4, and \#5 range from 100\% ionized region-contribution to 75\% PDR-contribution.
805: There is no region within our observed field in NGC\,3603 similar to an isolated "exposed" PDRs such as NGC7023.
806: This is likely due to a geometry effect because dust emission at 14\,$\mu$m could lie in the foreground or background, resulting in
807: an overestimation of the 14.0/6.0 continuum ratio.
808:
809: The only points showing more than 50\% ``AGN-like'' behavior are those from position \#6 that are probing the central stellar cluster.
810: This is because for these positions, the continuum measured at 6\,$\mu$m is dominated by stellar emission. Thus, if a given starburst galaxy spectrum is
811: dominated by stellar cluster emission, it could \textit{a priori} be confused with an AGN when using MIR diagnostics. However, the global
812: spectrum of NGC\,3603, obtained using the sum of the full slit spectra of all positions (diamond in the Fig.\,\ref{fig:laurent}), does not show
813: any sign of stellar contamination. In this case, the regions characterized by dust emission dominate the global spectrum.
814: Other than the small-scale results of position \#6, none of the points probing purely ISM material show signs of AGN regime. Using the
815: observations in NGC\,3603, we successfully test this diagram as a diagnostic tool to distinguish regimes in a single
816: objects, as long as the stellar emission is not significant.
817:
818: We also show in Figure\,\ref{fig:laurent} the results using full slit extraction of each
819: position (large squares), that were obtained by integrating the MIR emission of all the
820: regions within the slit. The corresponding data points of each position lie in the middle
821: of the small-scale results, implying that the full slit extraction does not give any
822: systematic error on the diagnostic.
823:
824:
825: \section{Conclusions}
826:
827: We have investigated the spatial variations of the MIR spectral feature intensities (both
828: ionic and solid state) in the Galactic H\2\ region complex NGC\,3603, and obtained the
829: following results:
830:
831: \begin{itemize}
832: \item[1.] On first approximation, the various PAH emission features have identical spatial distribution and the PAH ionization fraction
833: is constant.
834:
835: \item[2.] The ISRF hardness, as probed by the [S\4]/[Ne\2] ratio, decreases as a
836: function of the distance from the central stellar cluster.
837:
838: \item[3.] Silicate is detected in emission close to the cluster while it is detected in
839: absorption further away. Local absorption maxima of the 9.7\,$\mu$m feature seem
840: to identify with bright PAH emission knots.
841:
842: \item[5.] PAH emission lies at larger distances than VSG emission and becomes weaker
843: when the ISRF becomes harder. It has been shown
844: for low-metallicity galaxies by Madden et al.\ (2006).
845: It seems that PAH molecules are not able to survive the same physical conditions as VSGs. The simplest
846: explanation is the photodestruction of the molecules.
847:
848: \item[6.] Small-scale results within NGC\,3603 allowed us to probe observationnally the transition between ionized
849: region and PDR regimes in the MIR diagnostic diagram of Laurent et al.\ (2000). In this diagram,
850: the measurements using individual column spectra form a relatively narrow stripe.
851:
852: \end{itemize}
853:
854: \acknowledgments We thank E.\,Peeters for providing data for the diagnostic diagram
855: (Fig.\,\ref{fig:laurent}), S.\,Madden, G.\,Sloan, and H.\ Spoon for their useful comments. This work
856: is based on observations made with the {\it Spitzer Space Telescope}, which is operated
857: by the Jet Propulsion Laboratory, California Institute of Technology, under NASA contract
858: 1047. Support for this work was provided by NASA through contract 1257184 issued by
859: JPL/Caltech.
860: %\email{sda@sda.org}.
861:
862:
863: %{\it Facilities:} \facility{SPITZER (IRS, IRAC)}.
864:
865: \begin{thebibliography}{}
866:
867: %\bibitem[Allamandola et al.(1999)]{1999ApJ...511L.115A} Allamandola, L.~J.,
868: %Hudgins, D.~M., \& Sandford, S.~A.\ 1999, \apjl, 511, L115
869:
870: \bibitem[Allamandola et al.(1989)]{1989ApJS...71..733A} Allamandola, L.~J.,
871: Tielens, G.~G.~M., \& Barker, J.~R.\ 1989, \apjs, 71, 733
872:
873: \bibitem[Allamandola et al.(1985)]{1985ApJ...290L..25A} Allamandola, L.~J.,
874: Tielens, A.~G.~G.~M., \& Barker, J.~R.\ 1985, \apjl, 290, L25
875:
876: %\bibitem[Bakes et al.(2001)]{2001ApJ...560..261B} Bakes, E.~L.~O., Tielens,
877: %A.~G.~G.~M., Bauschlicher, C.~W., Jr., Hudgins, D.~M., \& Allamandola,
878: %L.~J.\ 2001, \apj, 560, 261
879:
880: \bibitem[Bakes \& Tielens(1998)]{1998ApJ...499..258B} Bakes, E.~L.~O., \&
881: Tielens, A.~G.~G.~M.\ 1998, \apj, 499, 258
882:
883: \bibitem[Bakes \& Tielens(1994)]{1994ApJ...427..822B} Bakes, E.~L.~O., \&
884: Tielens, A.~G.~G.~M.\ 1994, \apj, 427, 822
885:
886: \bibitem[Beir{\~a}o et al.(2006)]{2006ApJ...643L...1B} Beir{\~a}o, P.,
887: Brandl, B.~R., Devost, D., Smith, J.~D., Hao, L., \& Houck, J.~R.\ 2006,
888: \apjl, 643, L1
889:
890: %\bibitem[Boulanger et al.(1996)]{1996A&A...315L.325B} Boulanger, F., et
891: %al.\ 1996, \aap, 315, L325
892:
893: %\bibitem[Brandl et al.(2004)]{2004ApJS..154..188B} Brandl, B.~R., et al.\
894: %2004, \apjs, 154, 188
895:
896: \bibitem[Brandl et al.(1999)]{1999A&A...352L..69B} Brandl, B., Brandner,
897: W., Eisenhauer, F., Moffat, A.~F.~J., Palla, F., \& Zinnecker, H.\ 1999,
898: \aap, 352, L69
899:
900: \bibitem[Bregman \& Temi(2005)]{2005ApJ...621..831B} Bregman, J., \& Temi,
901: P.\ 2005, \apj, 621, 831
902:
903: %\bibitem[Bregman et al.(1989)]{1989ApJ...344..791B} Bregman, J.~D.,
904: %Allamandola, L.~J., Witteborn, F.~C., Tielens, A.~G.~G.~M., \& Geballe,
905: %T.~R.\ 1989, \apj, 344, 791
906:
907: \bibitem[Cesarsky et al.(2000)]{2000A&A...358..708C} Cesarsky, D., Jones,
908: A.~P., Lequeux, J., \& Verstraete, L.\ 2000, \aap, 358, 708
909:
910: \bibitem[Cesarsky et al.(1996)]{1996A&A...315L.309C} Cesarsky, D., Lequeux,
911: J., Abergel, A., Perault, M., Palazzi, E., Madden, S., \& Tran, D.\ 1996,
912: \aap, 315, L309
913:
914: \bibitem[Charbonnel et al.(1993)]{1993A&AS..101..415C} Charbonnel, C.,
915: Meynet, G., Maeder, A., Schaller, G., \& Schaerer, D.\ 1993, \aaps, 101,
916: 415
917:
918: \bibitem[Contursi et al.(2000)]{2000A&A...362..310C} Contursi, A., et al.\
919: 2000, \aap, 362, 310
920:
921: \bibitem[Cox \& Spaans(2006)]{2006A&A...451..973C} Cox, N.~L.~J., \&
922: Spaans, M.\ 2006, \aap, 451, 973
923:
924: \bibitem[Dartois \& D'Hendecourt(1997)]{1997A&A...323..534D} Dartois, E.,
925: \& D'Hendecourt, L.\ 1997, \aap, 323, 534
926:
927: %\bibitem[de Frees et al.(1993)]{1993ApJ...408..530D} de Frees, D.~J.,
928: %Miller, M.~D., Talbi, D., Pauzat, F., \& Ellinger, Y.\ 1993, \apj, 408, 530
929:
930: \bibitem[Desert et al.(1990)]{1990A&A...237..215D} Desert, F.-X.,
931: Boulanger, F., \& Puget, J.~L.\ 1990, \aap, 237, 215
932:
933: \bibitem[Dopita et al.(2002)]{} Dopita, M. A., Groves, B. A., Sutherland, R. S., Binette, L., \& Cecil, G. 2002, ApJ, 572,
934: 753
935:
936: \bibitem[Drissen et al.(1995)]{1995AJ....110.2235D} Drissen, L., Moffat,
937: A.~F.~J., Walborn, N.~R., \& Shara, M.~M.\ 1995, \aj, 110, 2235
938:
939: \bibitem[Eisenhauer et al.(1998)]{1998ApJ...498..278E} Eisenhauer, F.,
940: Quirrenbach, A., Zinnecker, H., \& Genzel, R.\ 1998, \apj, 498, 278
941:
942: \bibitem[Engelbracht et al.(2005)]{2005ApJ...628L..29E} Engelbracht, C.~W.,
943: Gordon, K.~D., Rieke, G.~H., Werner, M.~W., Dale, D.~A., \& Latter, W.~B.\
944: 2005, \apjl, 628, L29
945:
946: \bibitem[Fazio et al.(2004)]{2004ApJS..154...10F} Fazio, G.~G., et al.\
947: 2004, \apjs, 154, 10
948:
949: \bibitem[F{\"o}rster Schreiber et al.(2001)]{2001ApJ...552..544F}
950: F{\"o}rster Schreiber, N.~M., Genzel, R., Lutz, D., Kunze, D., \&
951: Sternberg, A.\ 2001, \apj, 552, 544
952:
953: %\bibitem[Galliano (2006)]{2006astro.ph.10852G} Galliano, F.\ 2006, ArXiv
954: %Astrophysics e-prints, arXiv:astro-ph/0610852
955:
956: %\bibitem[Galliano et al.(2006)]{} Galliano, F., Madden, S.~C., Tielens, A.~G.~G.~M.,
957: %Jones, A.~P., Peeters, E., \& Allamandola, L.~J.\ 2006, \apj, in preparation
958:
959: \bibitem[Garc{\'{\i}}a-Rojas et al.(2006)]{2006MNRAS.368..253G}
960: Garc{\'{\i}}a-Rojas, J., Esteban, C., Peimbert, M., Costado, M.~T.,
961: Rodr{\'{\i}}guez, M., Peimbert, A., \& Ruiz, M.~T.\ 2006, \mnras, 368, 253
962:
963: \bibitem[Geballe et al.(1989)]{1989ApJ...341..278G} Geballe, T.~R.,
964: Tielens, A.~G.~G.~M., Allamandola, L.~J., Moorhouse, A., \& Brand,
965: P.~W.~J.~L.\ 1989, \apj, 341, 278
966:
967: \bibitem[Guiles et al.(2004)]{2004AAS...205.2601G} Guiles, S., Devost, D.,
968: \& Houck, J.~R.\ 2004, Bulletin of the American Astronomical Society, 36, 1386
969:
970: \bibitem[Henry et al.(2004)]{2004AJ....127.2284H} Henry, R.~B.~C., Kwitter,
971: K.~B., \& Balick, B.\ 2004, \aj, 127, 2284
972:
973: \bibitem[Higdon et al.(2004)]{2004PASP..116..975H} Higdon, S.~J.~U., et
974: al.\ 2004, \pasp, 116, 975
975:
976: \bibitem[Hofmann et al.(1995)]{1995A&A...300..403H} Hofmann, K.-H.,
977: Seggewiss, W., \& Weigelt, G.\ 1995, \aap, 300, 403
978:
979: \bibitem[Hony et al.(2001)]{2001A&A...370.1030H} Hony, S., Van Kerckhoven,
980: C., Peeters, E., Tielens, A.~G.~G.~M., Hudgins, D.~M., \& Allamandola,
981: L.~J.\ 2001, \aap, 370, 1030
982:
983: \bibitem[Houck et al.(2004)]{2004ApJS..154...18H} Houck, J.~R., et al.\
984: 2004, \apjs, 154, 18
985:
986: %\bibitem[Houck et al.(2004b)]{2004ApJS..154..211H} Houck, J.~R., et al.\
987: %2004b, \apjs, 154, 211
988:
989: \bibitem[Hudgins et al.(2005)]{2005ApJ...632..316H} Hudgins, D.~M.,
990: Bauschlicher, C.~W., Jr., \& Allamandola, L.~J.\ 2005, \apj, 632, 316
991:
992: \bibitem[Hudgins \& Allamandola(1999)]{1999ApJ...516L..41H} Hudgins, D.~M.,
993: \& Allamandola, L.~J.\ 1999, \apjl, 516, L41
994:
995: \bibitem[Joblin et al.(2000)]{2000ibp..conf...49J} Joblin, C., Abergel, A.,
996: Bregman, J., D'Hendecourt, L., Heras, A.~M., Jourdain de Muizon, M., Pech,
997: C., \& Tielens, A.~G.~G.~M.\ 2000, ESA SP-456: ISO Beyond the Peaks: The
998: 2nd ISO Workshop on Analytical Spectroscopy, 49
999:
1000: \bibitem[Kassis et al.(2006)]{2006ApJ...637..823K} Kassis, M., Adams,
1001: J.~D., Campbell, M.~F., Deutsch, L.~K., Hora, J.~L., Jackson, J.~M., \&
1002: Tollestrup, E.~V.\ 2006, \apj, 637, 823
1003:
1004: \bibitem[Kennicutt(1984)]{1984ApJ...287..116K} Kennicutt, R.~C., Jr.\ 1984,
1005: \apj, 287, 116
1006:
1007: \bibitem[Kim et al.(2001)]{2001PhRvL..86.5691K} Kim, H.-S., Wagner, D.~R.,
1008: \& Saykally, R.~J.\ 2001, Physical Review Letters, 86, 5691
1009:
1010: \bibitem[Knacke \& Thomson(1973)]{1973PASP...85..341K} Knacke, R.~F., \&
1011: Thomson, R.~K.\ 1973, \pasp, 85, 341
1012:
1013: \bibitem[Langhoff(1996)]{} Langhoff, S.~R.\ 1996, J.\ Phys.\ Chem.\ A, 100, 2819
1014:
1015: \bibitem[Laurent et al.(2000)]{2000A&A...359..887L} Laurent, O., Mirabel,
1016: I.~F., Charmandaris, V., Gallais, P., Madden, S.~C., Sauvage, M., Vigroux,
1017: L., \& Cesarsky, C.\ 2000, \aap, 359, 887
1018:
1019: \bibitem[Leger \& Puget(1984)]{1984A&A...137L...5L} Leger, A., \& Puget,
1020: J.~L.\ 1984, \aap, 137, L5
1021:
1022: \bibitem[Leitherer et al.(1999)]{1999ApJS..123....3L} Leitherer, C., et
1023: al.\ 1999, \apjs, 123, 3
1024:
1025: \bibitem[Madden et al.(2006)]{2006A&A...446..877M} Madden, S.~C., Galliano,
1026: F., Jones, A.~P., \& Sauvage, M.\ 2006, \aap, 446, 877
1027:
1028: \bibitem[Marigo et al.(2003)]{2003A&A...409..619M} Marigo, P.,
1029: Bernard-Salas, J., Pottasch, S.~R., Tielens, A.~G.~G.~M., \& Wesselius,
1030: P.~R.\ 2003, \aap, 409, 619
1031:
1032: %\bibitem[Mart{\'{\i}}n-Hern{\'a}ndez et al.(2006)]{2006A&A...455..853M}
1033: %Mart{\'{\i}}n-Hern{\'a}ndez, N.~L., Schaerer, D., Peeters, E., Tielens,
1034: %A.~G.~G.~M., \& Sauvage, M.\ 2006, \aap, 455, 853
1035:
1036: \bibitem[Mart{\'{\i}}n-Hern{\'a}ndez et al.(2002)]{2002A&A...381..606M}
1037: Mart{\'{\i}}n-Hern{\'a}ndez, N.~L., et al.\ 2002, \aap, 381, 606
1038:
1039: \bibitem[Melnick et al.(1989)]{1989A&A...213...89M} Melnick, J., Tapia, M.,
1040: \& Terlevich, R.\ 1989, \aap, 213, 89
1041:
1042: \bibitem[Moffat et al.(1994)]{1994ApJ...436..183M} Moffat, A.~F.~J.,
1043: Drissen, L., \& Shara, M.~M.\ 1994, \apj, 436, 183
1044:
1045: \bibitem[Moffat(1983)]{1983A&A...124..273M} Moffat, A.~F.~J.\ 1983, \aap,
1046: 124, 273
1047:
1048: \bibitem[N{\"u}rnberger \& Stanke(2003)]{2003A&A...400..223N}
1049: N{\"u}rnberger, D.~E.~A., \& Stanke, T.\ 2003, \aap, 400, 223
1050:
1051: \bibitem[Omont(1986)]{1986A&A...164..159O} Omont, A.\ 1986, \aap, 164, 159
1052:
1053: \bibitem[Peeters et al.(2004)]{2004ApJ...613..986P} Peeters, E., Spoon,
1054: H.~W.~W., \& Tielens, A.~G.~G.~M.\ 2004, \apj, 613, 986
1055:
1056: \bibitem[Peeters et al.(2002)]{2002A&A...390.1089P} Peeters, E., Hony, S.,
1057: Van Kerckhoven, C., Tielens, A.~G.~G.~M., Allamandola, L.~J., Hudgins,
1058: D.~M., \& Bauschlicher, C.~W.\ 2002, \aap, 390, 1089
1059:
1060: \bibitem[Pottasch \& Bernard-Salas(2006)]{} Pottasch, S.~R., \& Bernard-Salas, J.,
1061: \ 2006, \aap, 457, 189
1062:
1063: \bibitem[Puget \& Leger(1989)]{1989ARA&A..27..161P} Puget, J.~L., \& Leger,
1064: A.\ 1989, \araa, 27, 161
1065:
1066: \bibitem[Salama et al.(1996)]{1996ApJ...458..621S} Salama, F., Bakes,
1067: E.~L.~O., Allamandola, L.~J., \& Tielens, A.~G.~G.~M.\ 1996, \apj, 458, 621
1068:
1069: \bibitem[Salpeter(1955)]{1955ApJ...121..161S} Salpeter, E.~E.\ 1955, \apj,
1070: 121, 161
1071:
1072: \bibitem[Schaerer et al.(1993)]{1993A&AS...98..523S} Schaerer, D., Meynet,
1073: G., Maeder, A., \& Schaller, G.\ 1993, \aaps, 98, 523
1074:
1075: \bibitem[Schaerer \& Maeder(1992)]{1992A&A...263..129S} Schaerer, D., \&
1076: Maeder, A.\ 1992, \aap, 263, 129
1077:
1078: \bibitem[Schaller et al.(1992)]{1992A&AS...96..269S} Schaller, G.,
1079: Schaerer, D., Meynet, G., \& Maeder, A.\ 1992, \aaps, 96, 269
1080:
1081: \bibitem[Schutte et al.(1993)]{1993ApJ...415..397S} Schutte, W.~A.,
1082: Tielens, A.~G.~G.~M., \& Allamandola, L.~J.\ 1993, \apj, 415, 397
1083:
1084: \bibitem[Sellgren et al.(1990)]{1990ApJ...349..120S} Sellgren, K.,
1085: Tokunaga, A.~T., \& Nakada, Y.\ 1990, \apj, 349, 120
1086:
1087: \bibitem[Smith et al.(2007)]{2007ApJ...656..770S} Smith, J.~D.~T., et al.\
1088: 2007, \apj, 656, 770
1089:
1090: \bibitem[Spoon et al.(2007)]{2007ApJ...654L..49S} Spoon, H.~W.~W.,
1091: Marshall, J.~A., Houck, J.~R., Elitzur, M., Hao, L., Armus, L., Brandl,
1092: B.~R., \& Charmandaris, V.\ 2007, \apjl, 654, L49
1093:
1094: \bibitem[Sturm et al.(2000)]{2000A&A...358..481S} Sturm, E., Lutz, D.,
1095: Tran, D., Feuchtgruber, H., Genzel, R., Kunze, D., Moorwood, A.~F.~M., \&
1096: Thornley, M.~D.\ 2000, \aap, 358, 481
1097:
1098: \bibitem[Sutherland \& Dopita(1993)]{1993ApJS...88..253S} Sutherland,
1099: R.~S., \& Dopita, M.~A.\ 1993, \apjs, 88, 253
1100:
1101: \bibitem[Szczepanski \& Vala(1993)]{1993ApJ...414..646S} Szczepanski, J.,
1102: \& Vala, M.\ 1993, \apj, 414, 646
1103:
1104: \bibitem[Tapia et al.(2001)]{2001RMxAA..37...39T} Tapia, M., Bohigas, J.,
1105: P{\'e}rez, B., Roth, M., \& Ruiz, M.~T.\ 2001, Revista Mexicana de
1106: Astronomia y Astrofisica, 37, 39
1107:
1108: \bibitem[Tielens et al.(1999)]{1999usis.conf..579T} Tielens, A.~G.~G.~M.,
1109: Hony, S., van Kerckhoven, C., \& Peeters, E.\ 1999, ESA SP-427: The
1110: Universe as Seen by ISO, 579
1111:
1112: \bibitem[Tielens(1993)]{1993IAUS..155..155T} Tielens, A.~G.~G.~M.\ 1993,
1113: Planetary Nebulae, 155, 155
1114:
1115: %\bibitem[Uchida et al.(2000)]{2000ApJ...530..817U} Uchida, K.~I., Sellgren,
1116: %K., Werner, M.~W., \& Houdashelt, M.~L.\ 2000, \apj, 530, 817
1117:
1118: \bibitem[Vermeij et al.(2002)]{2002A&A...382.1042V} Vermeij, R., Peeters,
1119: E., Tielens, A.~G.~G.~M., \& van der Hulst, J.~M.\ 2002, \aap, 382, 1042
1120:
1121: %\bibitem[Verstraete et al.(2001)]{2001A&A...372..981V} Verstraete, L., et
1122: %al.\ 2001, \aap, 372, 981
1123:
1124: \bibitem[Verstraete et al.(1996)]{1996A&A...315L.337V} Verstraete, L.,
1125: Puget, J.~L., Falgarone, E., Drapatz, S., Wright, C.~M., \& Timmermann, R.\
1126: 1996, \aap, 315, L337
1127:
1128: \bibitem[Werner et al.(2004)]{2004ApJS..154....1W} Werner, M.~W., et al.\
1129: 2004a, \apjs, 154, 1
1130:
1131: \bibitem[Werner et al.(2004)]{2004ApJS..154..309W} Werner, M.~W., Uchida,
1132: K.~I., Sellgren, K., Marengo, M., Gordon, K.~D., Morris, P.~W., Houck,
1133: J.~R., \& Stansberry, J.~A.\ 2004b, \apjs, 154, 309
1134:
1135: \bibitem[Wu et al.(2006)]{2006ApJ...639..157W} Wu, Y., Charmandaris, V.,
1136: Hao, L., Brandl, B.~R., Bernard-Salas, J., Spoon, H.~W.~W., \& Houck,
1137: J.~R.\ 2006, \apj, 639, 157
1138:
1139: \end{thebibliography}
1140:
1141:
1142: \end{document}
1143:
1144:
1145:
1146: %Hence, if a spectra was taken with no spatial resolution, the region would be listed as a
1147: %pure H\2\ region. It differs a lot from the zone where starburst galaxies are, and asks the question of can we consider Local giant H\2\ regions
1148: %as small-scale templates of starburst galaxies. Distribution of massive stars is more dispersed in SB galaxies wrt the compactness of a giant
1149: %H\2\ region where OB stars are concentrated ?
1150:
1151: %POS3
1152: %y = 0.0025907 + 0.015089 * x
1153: %y = 0.21227 + 0.38688 * x
1154: %y = 0.08036 - 0.059331 * x
1155: %y = 0.2177 - 0.33692 * x
1156: %2864-2829
1157: %a -2
1158:
1159: %POS4
1160: %y = 0.003725 + 0.064195 * x
1161: %y = 0.4079 + 0.72806 * x
1162: %y = 0.1287 - 0.15914 * x
1163: %y = 0.18224 - 0.26588 * x
1164: %2205-2151
1165: %a -8
1166:
1167: %POS5
1168: %y = 0.0023232 + 0.0096654 * x
1169: %y = 0.18747 + 0.19989 * x
1170: %y = 0.081608 - 0.075525 * x
1171: %y = 0.30839 - 0.38259 * x
1172: %2630-2585
1173: %a -0.00 (!!)
1174:
1175: %POS5off
1176: %n
1177: %y = 0.12063 + 0.20699 * x
1178: %y = 0.043962 - 0.04425 * x
1179: %y = 0.13833 - 0.22337 * x
1180: %2626-2603
1181: %a -12.0
1182:
1183: %POS7
1184: %n
1185: %y = 0.33869 + 0.71182 * x
1186: %y = 0.10499 - 0.039609 * x
1187: %y = 0.26012 - 0.39479 * x
1188: %a -11.0
1189: