1: %*******************************************************************************
2: %*
3: %* PAH BAND RATIO PAPER
4: %*
5: %*******************************************************************************
6:
7:
8: \documentclass[12pt,preprint]{aastex}
9: %\documentclass{emulateapj}
10:
11: % Top of the page
12: \slugcomment{\today\ (submitted to ApJ November 28, 2007)}
13: \shorttitle{PAH Variations Inside and Among Galaxies}
14: \shortauthors{Galliano et al.}
15:
16: % Aliases
17: %--------
18: % My aliases
19: \input{alias_general}
20: \input{alias_expressions}
21: \input{alias_physics}
22: \input{alias_objects}
23: \input{alias_instruments}
24:
25: % Methods
26: \newcommand{\frmet}{{\it Lorentzian method}}
27: \newcommand{\nlmet}{{\it Spline method}}
28: \newcommand{\usmet}{{\it Template method}}
29:
30: % Measured quantities
31: \newcommand{\ipah}[1]{{\rm I}_{#1}}
32: \newcommand{\iPah}{{\rm I}_\sms{PAH}}
33: \newcommand{\icont}{{\rm I}_\sms{cont}}
34:
35: \graphicspath{{.}}
36: \DeclareGraphicsExtensions{.eps}
37:
38: \newcommand{\modif}[1]{{\bf #1}}
39:
40: \begin{document}
41:
42:
43: % Headings
44: %---------
45: \title{VARIATIONS OF THE MID-IR AROMATIC FEATURES
46: INSIDE AND AMONG GALAXIES}
47:
48: \author{Fr\'ed\'eric Galliano}
49: \affil{Observational Cosmology Lab., Code 665,
50: NASA Goddard Space Flight Center,
51: Greenbelt MD 20910, USA \\
52: Department of Astronomy, University of Maryland,
53: College Park, MD 20742, USA}
54: \email{galliano@astro.umd.edu}
55: \author{Suzanne C. Madden}
56: \affil{Service d'Astrophysique, L'Orme des Merisiers, CEA/Saclay,
57: 91191 Gif-sur-Yvette, France}
58: \author{Alexander G. G. M. Tielens}
59: \affil{NASA Ames Research Center, Mail Stop 245-3, Moffett Field,
60: CA 94035, USA}
61: \author{Els Peeters}
62: \affil{NASA Ames Research Center, Mail Stop 245-6, Moffett Field,
63: CA 94035, USA \\
64: SETI Intstitute, 515 N. Whisman Rd, Mountain View, CA 94043, USA \\
65: and Physics \& Astronomy Dept., University of Western Ontario,
66: PAB 213, London ON N6A 3K7, Canada}
67: \and
68: \author{Anthony P. Jones}
69: \affil{Institut d'Astrophysique Spatiale, Universit\'e de Paris XI,
70: 91405 Orsay, France}
71:
72: \begin{abstract}
73: We present the results of a systematic study of mid-IR spectra of
74: Galactic regions, Magellanic \hii\ regions, and galaxies of various
75: types (dwarf, spiral, starburst), observed by the satellites \iso\ and
76: \spitz.
77: We study the relative variations of the 6.2, 7.7, 8.6 and $11.3\mic$
78: features inside spatially resolved objects (such as \M{82}, \M{51}, \xxxdor,
79: \M{17} and the \orb), as well as among 90 integrated spectra of 50 objects.
80: Our main results are that the 6.2, 7.7 and $8.6\mic$ bands are
81: essentially tied together, while the ratios between these bands and the
82: $11.3\mic$ band varies by one order of magnitude.
83: This implies that the properties of the PAHs are remarkably universal
84: throughout our sample, and that the relative variations of the band ratios
85: are mainly controled by the fraction of ionized PAHs.
86: In particular, we show that we can rule out both the modification
87: of the PAH size distribution, and the mid-infrared extinction, as an
88: explanation of these variations.
89: Using a few well-studied Galactic regions (including the spectral
90: image of the \orb), we give an empirical relation between the
91: $\ipah{6.2}/\ipah{11.3}$ ratio and the ionization/recombination ratio
92: $G_0/n_e\times\sqrt{T_\sms{gas}}$, therefore providing a useful
93: quantitative diagnostic tool of the physical conditions in the regions
94: where the PAH emission originates.
95: Finally, we discuss the physical interpretation of the
96: $\ipah{6.2}/\ipah{11.3}$ ratio, on galactic size scales.
97: \end{abstract}
98:
99: \keywords{dust --
100: HII regions --
101: ISM: structure --
102: galaxies: dwarf, starburst --
103: infrared: general
104: }
105:
106:
107: %===========================================================================
108:
109:
110: % Intro
111: %------
112: \section{INTRODUCTION}
113: \label{sec:intro}
114:
115: % Features or no features
116: The reprocessing of stellar light by dust in the infrared (IR) is widely used
117: to probe embedded star formation.
118: However, in the absence of other constraints, the physical properties which
119: can usually be derived from an
120: almost-featureless grain continuum emission is limited to global quantities,
121: such as the dust mass and its average temperature.
122: The ubiquity of numerous mid-IR aromatic features, in a wide variety of
123: astrophysical objects and environments, potentially provides more articulate
124: diagnostics of the physical conditions.
125: Indeed, these features dominate the mid-IR spectra of evolved stars
126: \citep[{e.g.}][]{blommaert05,kraemer06}, the cool ISM
127: \citep[{e.g.}][]{abergel05,flagey06,povich07},
128: as well as whole galaxies \citep[{e.g.}][]{verma05,smith07}, that have been
129: extensively observed by the \isoST\ (\iso), and are currently investigated
130: with a higher sensitivity by the \spitzST.
131: In our Galaxy, one third of the stellar light is reprocessed by dust,
132: while this fraction can go up to $99\%$ and higher, in starburst
133: galaxies.
134: At solar metallicity, roughly $15\%$ of the cooling is radiated through the
135: most powerful mid-IR bands, centered at 3.3, 6.2, 7.7, 11.3 and 12.7$\mic$.
136:
137: % Heating and modeling
138: Historically, these emission features were attributed to very small grains
139: ($\simeq10\,$\AA),
140: transiently heated by single photon absorption, in order to account for
141: the independence of the color temperature with the distance from the
142: illuminating star, in several reflection nebulae \citep{sellgren84}.
143: In parallel, the central wavelengths of these bands were recognized to coincide
144: with the vibrational modes of aromatic material \citep{duley81}.
145: These features are now commonly attributed to the molecular modes of
146: Polycyclic Aromatic Hydrocarbons \citep[hereafter PAHs;][]{leger84,allamandola85,allamandola89}, which are planar molecules made of
147: $\simeq10$ to 1000 carbon atoms, excited primarily by ultraviolet (UV)
148: photons.
149: With silicate and carbon grains, PAHs are a main component of dust models
150: \citep{desert90,dwek97,draine01,zubko04}.
151: Their absorption efficiency has been modeled using astrophysical
152: observations, laboratory measurements and quantum theory
153: \citep[in particular][]{desert90,joblin92,verstraete01,li01,mattioda05,mattioda05b,draine07,malloci07}.
154: In addition to being major radiative coolants of the interstellar medium (ISM),
155: PAHs are responsible for most of the photoelectric heating of
156: the gas in photodissociation regions (hereafter PDRs) and the neutral
157: interstellar medium, due to their high cumulative surface area
158: \citep[{e.g.}][]{bakes94,hollenbach97}.
159: For the same reason, they probably play an important role in grain surface
160: chemistry \citep[{e.g.}][]{tielens87}.
161: In our Galaxy, they contain $\simeq15-20\%$ of the depleted carbon
162: \citep[with solar abundance constraints]{zubko04}.
163: As a consequence, they are believed to be part of the interstellar carbon
164: condensation chain \citep{cherchneff00,dartois05}.
165:
166: % Details of the features
167: From an extragalactic point of view, the luminosity of the $6.2\mic$ feature
168: can be used as a tracer of star formation \citep{peeters04}.
169: However, this tracer is biased by global parameters such as the ISM metallicity.
170: Indeed, PAHs are underabundant in low-metallicity galaxies
171: \citep[{e.g.}][]{galliano03,galliano05,galliano08a,draine07b}.
172: There is a general correlation between the PAH-to-continuum intensity ratio
173: and the ISM metallicity \citep{madden06,wu06,ohalloran06,smith07}, and
174: consequently between the \IRACiv/\MIPSi\ broadband ratio and the metallicity
175: \citep{engelbracht05}.
176: The origin of this trend has been attributed to radiative and mechanical
177: destruction mechanisms by \citet{madden06} and \citet{ohalloran06} respectively.
178: Conversely, from the detailed modeling of the spectral energy distribution
179: (SED) of nearby galaxies, \citet{galliano08a} showed that the PAH-to-gas mass
180: ratio at different metallicities coincides with the relative amount of
181: carbon dust condensed in the envelopes of low-mass stars, during the
182: Asymptotic Giant Branch phase (AGB).
183: This study suggests that PAHs are injected into the ISM by their
184: progenitors, the AGB stars, several hundreds of Myr after the beginning of
185: the star formation, when the gas has already been enriched by more massive
186: stars.
187: This delay corresponds to the time needed for AGB stars to evolve off the
188: main sequence.
189: Therefore, the delayed injection of AGB-condensed carbon dust into the
190: ISM offers a natural explanation for the paucity of PAHs in low-metallicity
191: environments.
192:
193: % Cosmology
194: From a cosmological point of view, the large luminosity in these IR emission
195: features coupled with the high sensitivity of \spitz\ has allowed
196: the detection of PAHs in distant luminous infrared galaxies
197: out to redshift $z\simeq2$ \cite[{e.g.}][]{elbaz05,yan05,houck05}.
198: Hence, understanding what controls the properties of the aromatic features
199: on large scales is required, in order to properly interpret broadband surveys.
200:
201: % Shape of the spectrum
202: The detailed characteristics of the mid-IR features, such as their shape, their
203: central wavelength or the intensity ratio between the different bands are
204: known to vary \citep[see][]{peeters04b}.
205: Such variations are essentially due to modifications of the molecular structure
206: of the PAHs, in different astrophysical environments.
207: In particular, since each feature is attributed to a given vibrational
208: mode, the ratio between these features will vary with quantities such as the
209: charge, the hydrogenation or the size and shape of the molecule.
210: The $3.3\mic$ PAH band arises from the radiative relaxation of CH
211: stretching modes, while the $11.3\mic$ feature originates in the CH
212: out-of-plane bending modes;
213: CC stretching modes are responsible for the features between 6 and $9\mic$;
214: CH in plane bending excitation produces part of the $8.6\mic$ band.
215: Now, laboratory studies and quantum calculations shows that the CC modes are
216: instrinsically weak in neutral PAHs, and become stronger when the PAHs are
217: ionized \citep{langhoff96,allamandola99,bauschlicher02,kim02}.
218: Therefore, the 6 to $9\mic$ bands will be much more intense for a PAH$^+$
219: than for a PAH$^0$, while it will be the opposite for the 3.3 and
220: $11.3\mic$.
221: Consequently, the ratios between the CC and the CH feature intensities depend
222: on the charge of the PAHs, which is directly related to the physical
223: conditions (e.g.\ intensity of the ionizing radiation field, electron
224: density, etc.) in the environment where the emission is originating.
225:
226: % Variations of the bands
227: Evidence of variations between features in different astrophysical environments
228: have been reported by many authors.
229: For example, \citet{joblin96} showed that the $\ipah{8.6}/\ipah{11.3}$ ratio
230: decreases with increasing distance from the exciting star, in the
231: reflection nebulae \ngc{1333}
232: --~where $\ipah{\lambda}$ is the integrated intensity of the feature centered
233: at $\lambda\mic$.
234: \citet{hony01} found a good correlation between the 3.3 and $11.3\mic$ CH bands,
235: in a sample of Galactic \hii\ regions, YSOs, and evolved stars,
236: while they reported variations of $\ipah{6.2}/\ipah{11.3}$ by a factor of 5.
237: The observations of Galactic and Magellanic \hii\ regions, presented by
238: \citet{vermeij02}, indicate that the ratios $\ipah{6.2}/\ipah{11.3}$,
239: $\ipah{7.7}/\ipah{11.3}$ and $\ipah{8.6}/\ipah{11.3}$ are correlated.
240: Furthermore, they suggest a segregation between the values of these ratios
241: in the Milky Way and those in the Magellanic Clouds.
242: \citet{bregman05} studied the variation of $\ipah{7.7}/\ipah{11.3}$ in three
243: reflection nebulae.
244: Assuming that this variation is controlled by the charge of the PAHs, they could
245: relate this band ratio to the ratio $G_0/n_e$ between the integrated intensity
246: of the UV field, $G_0$, and the electron density, $n_e$.
247: Similarly, \citet{compiegne07}, studying the detailed variations
248: of the mid-IR spectrum in the Horsehead nebula, attributed the high relative
249: strength of the $\ipah{11.3}$ feature to a high fraction of neutral
250: PAHs, due
251: to the high ambient electron density.
252: On the contrary, \citet{smith07} studied the variation of
253: $\ipah{7.7}/\ipah{11.3}$ coming from the nuclear regions of the SINGS legacy
254: program galaxies.
255: They find that this ratio is relatively constant among pure starbursts, but
256: varies by a factor of 5 among galaxies having a weak AGN.
257: They interpret this effect as a selective destruction of the smallest PAHs
258: by the hard radiation arising from the accretion disk, ruling out the
259: explanation in terms of ionization of the molecules, in these particular
260: environments.
261: This interpretation is also supported by the high $11.3\mic$ and the
262: weakness of the $3.3\mic$ band in the \akari\ spectrum of the giant
263: elliptical galaxy \ngc{1316} \citep{kaneda07}.
264:
265: % The aim of the paper
266: The previous considerations stress the diversity of the possible interpretation
267: of the mid-IR feature variations in galaxies.
268: We need to identify the main physical processes controling the PAH bands, if
269: we are to use them as diagnostic tools.
270: This is the aim of this paper.
271: It presents a quantitative analysis of mid-IR spectra (\iso\ and \spitz) of
272: Galactic regions, low-metallicity dwarf galaxies, quiescent spirals, starbursts
273: and AGNs.
274: To achieve our goal, we focus on identifying the main trends between the
275: various PAH features, at different spatial scales and in different environments.
276: We then link these variations to the physical conditions inside the studied
277: region.
278: Preliminary results of this study were published in
279: \citet{galliano04,galliano07c}.
280:
281: % Content
282: The paper is organized as follows.
283: In \refS{sec:obs}, we present our sample and the data reduction.
284: We discuss the spectral decomposition of the mid-IR spectrum, in
285: \refS{sec:decomp}.
286: The results of this decomposition are presented in \refS{sec:correl};
287: we study the various trends between the band ratios within galaxies, and among
288: different types of environment.
289: Then, in \refS{sec:explanation}, we provide a physical interpretation of these
290: trends, when the structure of the ISM is resolved, and when it is not.
291: Finally, we summarize our conclusions in \refS{sec:concl}.
292:
293:
294: %===========================================================================
295:
296:
297: % Observations
298: %-------------
299: \section{OBSERVATIONS AND DATA REDUCTION}
300: \label{sec:obs}
301:
302: % Sources
303: %--------
304: \subsection{The Sample}
305: \label{sec:source}
306:
307: In order to systematically study the properties of the mid-IR aromatic
308: features, we include in our sample a wide variety of Galactic regions
309: and galaxies, covering a large range of metallicities and star formation
310: activities.
311: We merge the \iso\ samples of starbursts and AGNs presented by
312: \citet{laurent00}, spirals by \citet{roussel01},
313: Magellanic regions and dwarfs by \citet{madden06},
314: and complement them by low-metallicity sources observed with \spitz.
315: We add to this sample, the \isocam\ spectra of several Galactic regions, like
316: \M{17} \citep{cesarsky96b}, \ngc{2023} \citep{abergel02}, \ngc{7027}
317: \citep{persi99} and the \orb\ \citep{cesarsky00}, and the \isosws\ spectra
318: ($2.5-45\mic$)
319: of several compact \hii\ regions published by \citet{peeters02}.
320: The \smcb\ spectrum is the one presented by \citet{reach00}.
321:
322: The global properties of the selected sources are presented in
323: \reftab{tab:source}.
324: If relevant, the distances were homogenised to
325: $H_\sms{0}=71\; h^{-2}\,\rm km\,s^{-1}\,Mpc^{-1}$.
326: % Source
327: %-------
328: \begin{deluxetable}{lrrrrrrrr}
329: % Head
330: %-----
331: \tabletypesize{\footnotesize}
332: % \rotate
333: \tablecolumns{9}
334: \tablewidth{0pc}
335: \tablecaption{General properties of the sample.}
336: \tablehead{
337: \colhead{Name} & \colhead{R.A.} & \colhead{Dec.} &
338: \colhead{Aperture} & \colhead{Mid-IR} & \colhead{Distance} &
339: \multicolumn{2}{c}{$12+\log({\rm O/H})$} & \colhead{Category} \\
340: \colhead{} & \colhead{(J2000)} & \colhead{(J2000)} &
341: \colhead{} & \colhead{spectrograph} & \colhead{(Mpc)} &
342: & \colhead{[ref.]} & }
343: % Data
344: %-----
345: \startdata
346: % Haro 11
347: \haro{11} & $00^h36^m52\fs5$ & $-33\degr33\arcmin19\arcsec$ &
348: $10\arcsec$ & \irs & 92 &
349: 7.9 & \obsref{ZHaro11} & Dwarf \\
350: % SMC B1#1
351: \smcb & $00^h45^m33\fs0$ & $-73\degr18\arcmin46\arcsec$ &
352: $84\arcsec-36\arcsec$ & \isocam & 0.06 &
353: 8.0 & \obsref{ZMC} & Magellanic \\
354: % NGC 253
355: \ngc{253} & $00^h47^m32\fs9$ & $-25\degr17\arcmin18\arcsec$ &
356: $40\arcsec$ & \isocam & 3.3 &
357: 9.0 & \obsref{ZN253} & SB/AGN \\
358: \ngc{253}~p & $00^h47^m32\fs9$ & $-25\degr17\arcmin18\arcsec$ &
359: $10\arcsec$ & \isocam & 3.3 &
360: 9.0 & \refobs{ZN253} & SB/AGN \\
361: \ngc{253}~e & $00^h47^m32\fs9$ & $-25\degr17\arcmin18\arcsec$ &
362: $40\arcsec-10\arcsec$ & \isocam & 3.3 &
363: 9.0 & \refobs{ZN253} & SB/AGN \\
364: % SMC N 66
365: \smcn & $00^h59^m02\fs0$ & $-72\degr10\arcmin36\arcsec$ &
366: $120\arcsec$ & \isocam & 0.06 &
367: 8.0 & \refobs{ZMC} & Magellanic \\
368: % NGC 520
369: \ngc{520} & $01^h24^m34\fs9$ & $+03\degr47\arcmin31\arcsec$ &
370: $30\arcsec$ & \isocam & 27 &
371: \nodata & & Dwarf \\
372: % NGC 613
373: \ngc{613} & $01^h34^m17\fs5$ & $-29\degr24\arcmin58\arcsec$ &
374: $60\arcsec$ & \isocam & 19 &
375: 9.2 & \obsref{ZN613} & Spiral \\
376: \ngc{613}~p & $01^h34^m17\fs5$ & $-29\degr24\arcmin58\arcsec$ &
377: $20\arcsec$ & \isocam & 19 &
378: 9.2 & \refobs{ZN613} & Spiral \\
379: % NGC 891
380: \ngc{891} & $02^h22^m33\fs4$ & $+42\degr20\arcmin57\arcsec$ &
381: $200\arcsec$ & \isocam & 9.6 &
382: 8.9 & \obsref{ZN891} & Spiral \\
383: % NGC 1068
384: \ngc{1068} & $02^h42^m40\fs6$ & $-00\degr00\arcmin47\arcsec$ &
385: $60\arcsec$ & \isocam & 15 &
386: 9.0 & \obsref{ZN1068} & SB/AGN \\
387: \ngc{1068}~p & $02^h42^m40\fs6$ & $-00\degr00\arcmin47\arcsec$ &
388: $20\arcsec$ & \isocam & 15 &
389: 9.0 & \refobs{ZN1068} & SB/AGN \\
390: \ngc{1068}~e & $02^h42^m40\fs6$ & $-00\degr00\arcmin47\arcsec$ &
391: $60\arcsec-20\arcsec$ & \isocam & 15 &
392: 9.0 & \refobs{ZN1068} & SB/AGN \\
393: % NGC 1097
394: \ngc{1097} & $02^h46^m19\fs1$ & $-30\degr16\arcmin28\arcsec$ &
395: $100\arcsec$ & \isocam & 12 &
396: 9.0 & \obsref{ZN1097} & Spiral \\
397: \ngc{1097}~p & $02^h46^m19\fs1$ & $-30\degr16\arcmin28\arcsec$ &
398: $40\arcsec$ & \isocam & 12 &
399: 9.0 & \refobs{ZN1097} & Spiral \\
400: % NGC 1140
401: \ngc{1140} & $02^h54^m33\fs5$ & $-10\degr01\arcmin44\arcsec$ &
402: $20\arcsec$ & \isocam & 25 &
403: 8.0 & \obsref{heckman} & Dwarf \\
404: % NGC 1365
405: \ngc{1365} & $03^h33^m35\fs6$ & $-36\degr08\arcmin23\arcsec$ &
406: $100\arcsec$ & \isocam & 19 &
407: 9.1 & \obsref{ZN1365} & Spiral \\
408: \ngc{1365}~p & $03^h33^m35\fs6$ & $-36\degr08\arcmin23\arcsec$ &
409: $40\arcsec$ & \isocam & 19 &
410: 9.1 & \refobs{ZN1365} & Spiral \\
411: \ngc{1365}~e & $03^h33^m35\fs6$ & $-36\degr08\arcmin23\arcsec$ &
412: $100\arcsec-40\arcsec$ & \isocam & 19 &
413: 9.1 & \refobs{ZN1365} & Spiral \\
414: % IC 342
415: \IC{342} & $03^h46^m49\fs7$ & $+68\degr05\arcmin45\arcsec$ &
416: $40\arcsec$ & \isocam & 3.8 &
417: 8.9 & \obsref{Zic} & SB/AGN \\
418: \IC{342}~p & $03^h46^m49\fs7$ & $+68\degr05\arcmin45\arcsec$ &
419: $12\arcsec$ & \isocam & 3.8 &
420: 8.9 & \refobs{Zic} & SB/AGN \\
421: \IC{342}~e & $03^h46^m49\fs7$ & $+68\degr05\arcmin45\arcsec$ &
422: $40\arcsec-12\arcsec$ & \isocam & 3.8 &
423: 8.9 & \refobs{Zic} & SB/AGN \\
424: \IC{342}~map & $03^h46^m49\fs7$ & $+68\degr05\arcmin45\arcsec$ &
425: $3\arcsec$ & \isocam & 3.8 &
426: 8.9 & \refobs{Zic} & SB/AGN \\
427: % NGC 1569
428: \ngc{1569} & $04^h30^m49\fs1$ & $+64\degr50\arcmin52\arcsec$ &
429: $120\arcsec$ & \isocam & 2.2 &
430: 8.2 & \obsref{ZN1569} & Dwarf \\
431: \ngc{1569}~e & $04^h30^m49\fs1$ & $+64\degr50\arcmin52\arcsec$ &
432: $120\arcsec-12\arcsec$ & \isocam & 2.2 &
433: 8.2 & \refobs{ZN1569} & Dwarf \\
434: % NGC 1808
435: \ngc{1808} & $05^h07^m42\fs3$ & $-37\degr30\arcmin47\arcsec$ &
436: $50\arcsec$ & \isocam & 11 &
437: 9.1 & \obsref{ZN1808} & SB/AGN \\
438: % Orion bar SWS
439: \orb~D8 & $05^h35^m18\fs2$ & $-05\degr24\arcmin40\arcsec$ &
440: $14\arcsec\times20\arcsec$ & \isosws & 475 pc &
441: \nodata & & \hii\ region \\
442: \orb~D5 & $05^h35^m19\fs8$ & $-05\degr25\arcmin10\arcsec$ &
443: $14\arcsec\times20\arcsec$ & \isosws & 475 pc &
444: \nodata & & \hii\ region \\
445: \orb~D2 & $05^h35^m21\fs4$ & $-05\degr25\arcmin40\arcsec$ &
446: $14\arcsec\times20\arcsec$ & \isosws & 475 pc &
447: \nodata & & \hii\ region \\
448: % Orion bar
449: \orb & $5^h35^m20\fs0$ & $-05\degr25\arcmin20\arcsec$ &
450: $6\arcsec$ & \isocam & 475 pc &
451: \nodata & & PDR \\
452: % 30 Dor
453: \xxxdor & $05^h38^m34\fs0$ & $-69\degr05\arcmin57\arcsec$ &
454: $120\arcsec$ & \isocam & 0.05 &
455: 8.4 & \refobs{ZMC} & Magellanic \\
456: \xxxdor~p & $05^h38^m34\fs0$ & $-69\degr05\arcmin57\arcsec$ &
457: $40\arcsec$ & \isocam & 0.05 &
458: 8.4 & \refobs{ZMC} & Magellanic \\
459: \xxxdor~e & $05^h38^m34\fs0$ & $-69\degr05\arcmin57\arcsec$ &
460: $120\arcsec-40\arcsec$ & \isocam & 0.05 &
461: 8.4 & \refobs{ZMC} & Magellanic \\
462: \xxxdor~map & $05^h38^m34\fs0$ & $-69\degr05\arcmin57\arcsec$ &
463: $6\arcsec$ & \isocam & 0.05 &
464: 8.4 & \refobs{ZMC} & Magellanic \\
465: % NGC 2023
466: \ngc{2023} & $05^h41^m38\fs3$ & $-02\degr16\arcmin33\arcsec$ &
467: $14\arcsec\times20\arcsec$ & \isosws & 475 pc &
468: \nodata & & PDR \\
469: % II Zw 40
470: \iizw & $05^h55^m42\fs7$ & $+03\degr23\arcmin29\arcsec$ &
471: $24\arcsec$ & \isocam & 10 &
472: 8.1 & \obsref{Ziizw} & Dwarf \\
473: % He2-10
474: \hen & $08^h36^m15\fs2$ & $-26\degr24\arcmin34\arcsec$ &
475: $10\arcsec$ & \irs & 8.7 &
476: 8.9 & \refobs{ZN1569} & Dwarf \\
477: % M 82
478: \M{82} & $09^h55^m51\fs8$ & $+69\degr40\arcmin46\arcsec$ &
479: $90\arcsec$ & \isocam & 3.6 &
480: 9.0 & \obsref{ZM82} & SB/AGN \\
481: \M{82}~p & $09^h55^m51\fs8$ & $+69\degr40\arcmin46\arcsec$ &
482: $24\arcsec$ & \isocam & 3.6 &
483: 9.0 & \refobs{ZM82} & SB/AGN \\
484: \M{82}~e & $09^h55^m51\fs8$ & $+69\degr40\arcmin46\arcsec$ &
485: $90\arcsec-24\arcsec$ & \isocam & 3.6 &
486: 9.0 & \refobs{ZM82} & SB/AGN \\
487: \M{82}~map & $09^h55^m51\fs8$ & $+69\degr40\arcmin46\arcsec$ &
488: $3\arcsec$ & \isocam & 3.6 &
489: 9.0 & \refobs{ZM82} & SB/AGN \\
490: % NGC 3256
491: \ngc{3256} & $10^h27^m51\fs1$ & $-43\degr54\arcmin17\arcsec$ &
492: $24\arcsec$ & \isocam & 37 &
493: 8.9 & \obsref{ZN3256} & SB/AGN \\
494: \ngc{3256}~p & $10^h27^m51\fs1$ & $-43\degr54\arcmin17\arcsec$ &
495: $10\arcsec$ & \isocam & 37 &
496: 8.9 & \refobs{ZN3256} & SB/AGN \\
497: \ngc{3256}~e & $10^h27^m51\fs1$ & $-43\degr54\arcmin17\arcsec$ &
498: $24\arcsec-10\arcsec$ & \isocam & 37 &
499: 8.9 & \refobs{ZN3256} & SB/AGN \\
500: % Mrk 33
501: \mrk{33} & $10^h32^m31\fs9$ & $+54\degr24\arcmin04\arcsec$ &
502: $10\arcsec$ & \irs & 20 &
503: 8.4 & \refobs{ZN3256} & Dwarf \\
504: % Arp 299
505: \arp{299} & $11^h28^m31\fs0$ & $+58\degr33\arcmin39\arcsec$ &
506: $24\arcsec$ & \isocam & 41 &
507: \nodata & & SB/AGN \\
508: % UM 448
509: \um{448} & $11^h42^m12\fs4$ & $+00\degr20\arcmin03\arcsec$ &
510: $10\arcsec$ & \irs & 70 &
511: 8.0 & \obsref{ZUM448} & Dwarf \\
512: % IR 12331
513: \IR{12331} & $12^h36^m01\fs9$ & $-61\degr51\arcmin04\arcsec$ &
514: $14\arcsec\times20\arcsec$ & \isosws & 4.5 kpc &
515: \nodata & & \hii\ region \\
516: % NGC 4945
517: \ngc{4945} & $13^h05^m26\fs2$ & $-49\degr28\arcmin15\arcsec$ &
518: $60\arcsec$ & \isocam & 3.9 &
519: \nodata & & SB/AGN \\
520: \ngc{4945}~p & $13^h05^m26\fs2$ & $-49\degr28\arcmin15\arcsec$ &
521: $20\arcsec$ & \isocam & 3.9 &
522: \nodata & & SB/AGN \\
523: \ngc{4945}~e & $13^h05^m26\fs2$ & $-49\degr28\arcmin15\arcsec$ &
524: $60\arcsec-20\arcsec$ & \isocam & 3.9 &
525: \nodata & & SB/AGN \\
526: % Centaurus A
527: \cenA & $13^h25^m28\fs0$ & $-43\degr01\arcmin06\arcsec$ &
528: $30\arcsec$ & \isocam & 3.8 &
529: $\sim9$ & \obsref{ZcenA} & SB/AGN \\
530: \cenA~e & $13^h25^m28\fs0$ & $-43\degr01\arcmin06\arcsec$ &
531: $10\arcsec$ & \isocam & 3.8 &
532: $\sim9$ & \refobs{ZcenA} & SB/AGN \\
533: % M51
534: \M{51} & $13^h29^m52\fs7$ & $+47\degr11\arcmin43\arcsec$ &
535: $140\arcsec$ & \isocam & 8.4 &
536: 8.7 & \obsref{ZM51} & Spiral \\
537: \M{51}~p & $13^h29^m52\fs7$ & $+47\degr11\arcmin43\arcsec$ &
538: $60\arcsec$ & \isocam & 8.4 &
539: 8.7 & \refobs{ZM51} & Spiral \\
540: \M{51}~e & $13^h29^m52\fs7$ & $+47\degr11\arcmin43\arcsec$ &
541: $140\arcsec-60\arcsec$ & \isocam & 8.4 &
542: 8.7 & \refobs{ZM51} & Spiral \\
543: \M{51}~map & $13^h29^m52\fs7$ & $+47\degr11\arcmin43\arcsec$ &
544: $6\arcsec$ & \isocam & 8.4 &
545: 8.7 & \refobs{ZM51} & Spiral \\
546: % M 83
547: \M{83} & $13^h37^m00\fs7$ & $-29\degr51\arcmin58\arcsec$ &
548: $200\arcsec$ & \isocam & 4.5 &
549: 9.2 & \obsref{ZM83} & Spiral \\
550: \M{83}~p & $13^h37^m00\fs7$ & $-29\degr51\arcmin58\arcsec$ &
551: $50\arcsec$ & \isocam & 4.5 &
552: 9.2 & \refobs{ZM83} & Spiral \\
553: \M{83}~e & $13^h37^m00\fs7$ & $-29\degr51\arcmin58\arcsec$ &
554: $200\arcsec-50\arcsec$ & \isocam & 4.5 &
555: 9.2 & \refobs{ZM83} & Spiral \\
556: \M{83}~map & $13^h37^m00\fs7$ & $-29\degr51\arcmin58\arcsec$ &
557: $6\arcsec$ & \isocam & 4.5 &
558: 9.2 & \refobs{ZM83} & Spiral \\
559: % Cicinus
560: Circinus & $14^h13^m09\fs6$ & $-65\degr20\arcmin21\arcsec$ &
561: $40\arcsec$ & \isocam & 4.0 &
562: \nodata & & SB/AGN \\
563: Circinus~e & $14^h13^m09\fs6$ & $-65\degr20\arcmin21\arcsec$ &
564: $40\arcsec-10\arcsec$ & \isocam & 4.0 &
565: \nodata & & SB/AGN \\
566: % Arp 220
567: \arp{220} & $15^h34^m57\fs2$ & $+23\degr30\arcmin11\arcsec$ &
568: $20\arcsec$ & \isocam & 73 &
569: \nodata & & ULIRG \\
570: \arp{220}~p & $15^h34^m57\fs2$ & $+23\degr30\arcmin11\arcsec$ &
571: $10\arcsec$ & \isocam & 73 &
572: \nodata & & ULIRG \\
573: % IR 15384
574: \IR{15384} & $15^h42^m17\fs1$ & $-53\degr58\arcmin31\arcsec$ &
575: $14\arcsec\times20\arcsec$ & \isosws & 2.7 kpc &
576: \nodata & & \hii\ region \\
577: % NGC 6240
578: \ngc{6240} & $16^h52^m58\fs8$ & $+02\degr24\arcmin06\arcsec$ &
579: $20\arcsec$ & \isocam & 98 &
580: \nodata & & SB/AGN \\
581: \ngc{6240}~p & $16^h52^m58\fs8$ & $+02\degr24\arcmin06\arcsec$ &
582: $10\arcsec$ & \isocam & 98 &
583: \nodata & & SB/AGN \\
584: % M 17
585: \M{17}~map & $18^h20^m22\fs0$ & $-16\degr12\arcmin40\arcsec$ &
586: $6\arcsec$ & \isocam & 1.5 kpc &
587: \nodata & & PDR \\
588: % IR 18317
589: \IR{18317} & $18^h34^m24\fs9$ & $-07\degr54\arcmin47\arcsec$ &
590: $14\arcsec\times20\arcsec$ & \isosws & 4.9 kpc &
591: \nodata & & \hii\ region \\
592: % NGC 6946
593: \ngc{6946} & $20^h34^m51\fs2$ & $+60\degr09\arcmin17\arcsec$ &
594: $140\arcsec$ & \isocam & 5.5 &
595: 9.1 & \obsref{ZN6946} & Spiral \\
596: \ngc{6946}~p & $20^h34^m51\fs2$ & $+60\degr09\arcmin17\arcsec$ &
597: $40\arcsec$ & \isocam & 5.5 &
598: 9.1 & \refobs{ZN6946} & Spiral \\
599: \ngc{6946}~e & $20^h34^m51\fs2$ & $+60\degr09\arcmin17\arcsec$ &
600: $140\arcsec-40\arcsec$ & \isocam & 5.5 &
601: 9.1 & \refobs{ZN6946} & Spiral \\
602: % NGC 7027
603: \ngc{7027} & $21^h07^m01\fs7$ & $+42\degr14\arcmin09\arcsec$ &
604: $14\arcsec\times20\arcsec$ & \isosws & 700 pc &
605: \nodata & & PDR \\
606: % IR 22308
607: \IR{22308} & $22^h32^m45\fs9$ & $+58\degr28\arcmin21\arcsec$ &
608: $14\arcsec\times20\arcsec$ & \isosws & 5.5 kpc &
609: \nodata & & \hii\ region \\
610: % IR 23030
611: \IR{23030} & $23^h05^m10\fs6$ & $+60\degr14\arcmin41\arcsec$ &
612: $14\arcsec\times20\arcsec$ & \isosws & 5.2 kpc &
613: \nodata & & \hii\ region \\
614: % IR 23133
615: \IR{23133} & $23^h15^m31\fs4$ & $+61\degr07\arcmin08\arcsec$ &
616: $14\arcsec\times20\arcsec$ & \isosws & 5.5 kpc &
617: \nodata & & \hii\ region \\
618: % IR 23128
619: \IR{23128} & $23^h15^m46\fs0$ & $-59\degr03\arcmin17\arcsec$ &
620: $20\arcsec$ & \isocam & 180 &
621: \nodata & & SB/AGN \\
622: \IR{23128}~p & $23^h15^m46\fs0$ & $-59\degr03\arcmin17\arcsec$ &
623: $10\arcsec$ & \isocam & 180 &
624: \nodata & & SB/AGN \\
625: % NGC 7714
626: \ngc{7714} & $23^h36^m14\fs1$ & $+02\degr09\arcmin19\arcsec$ &
627: $10\arcsec$ & \irs & 37 &
628: 8.5 & \obsref{ZN7714} & Dwarf \\
629: \enddata
630: \label{tab:source}
631: \tablerefs{\refobs{ZHaro11}~\citet{bergvall00};
632: \refobs{ZMC}~\citet{dufour82};
633: \refobs{ZN253}~\citet{zaritsky94};
634: \refobs{ZN613}~\citet{alloin79};
635: \refobs{ZN891}~\citet{otte01};
636: \refobs{ZN1068}~\citet{dutil99};
637: \refobs{ZN1097}~\citet{storchi-bergmann95};
638: \refobs{heckman}~\citet{heckman98};
639: \refobs{ZN1365}~\citet{roy97};
640: \refobs{Zic}~\citet{pilyugin04};
641: \refobs{ZN1569}~\citet{kobulnicky97};
642: \refobs{ZN1808}~\citet{ravindranath01};
643: \refobs{Ziizw}~\citet{perez-montero03};
644: \refobs{ZM82}~\citet{boselli02};
645: \refobs{ZN3256}~\citet{mas-hesse99};
646: \refobs{ZUM448}~\citet{izotov98};
647: \refobs{ZcenA}~\citet{schaerer00};
648: \refobs{ZM51}~\citet{bresolin04};
649: \refobs{ZM83}~\citet{webster83};
650: \refobs{ZN6946}~\citet{kobulnicky99b};
651: \refobs{ZN7714}~\citet{gonzalez-delgado95}.}
652: \tablecomments{The sources are ordered according to their right ascension.
653: For \isocam\ sources, the apertures are circular, centered
654: on the coordinates listed in the second and third columns.
655: For \isosws\ sources, the beam is $14\arcsec\times20\arcsec$
656: (bands 1 and 2).
657: The diameter of the aperture is given in the fourth column.
658: When the diameter has the form
659: $\theta_1\arcsec-\theta_2\arcsec$, it means that
660: the we subtracted from the flux in the circular
661: aperture
662: of diameter $\theta_1\arcsec$,
663: the flux in a concentric circular aperture of diameter
664: $\theta_2\arcsec<\theta_1\arcsec$.
665: The sources which are followed by ``map'' are those for which
666: the signal-to-noise of the spectral-image is good;
667: in this case, the aperture refers to the pixel field of view.
668: The letter after the name of the source designates the type
669: of aperture:
670: when there is no letter, most of the emission of the object
671: is encompassed in the aperture;
672: ``p'' is for peak emission; ``e'' for extended emission.}
673: \end{deluxetable}
674: \clearpage
675:
676: % CAM
677: %----
678: \subsection{\isocam\ Data Reduction}
679: \label{sec:cam}
680:
681: Most of the sources in \reftab{tab:source} were observed with \isocam\
682: \citep{cesarsky96cam} on board the \iso\ satellite \citep{kessler96}.
683: We refer to the work of \citet{madden06} for a detailed description of the
684: data reduction applied to all of the \isocam\ data used in this study.
685: The CVF performed spectral imaging using a $32\times 32$ detector
686: array, with a sampling of $3\upfov$ or $6\upfov$ in our cases, from
687: $\lambda=5\mic$ to $16.5\mic$ with one pointing of two CVFs, from $\lambda=5$
688: to $9.5\mic$ and from $\lambda=9.0$ to $16.5\mic$.
689: The spectral resolution increases from $\lambda/\Delta\lambda=35$ to 51 across
690: the full spectra.
691:
692: For the \iso\ data treatment, we used the CAM Interactive Reduction
693: \citep[CIR, version AUG01;][]{chanial03}.
694: % Darks
695: The subtraction of the dark currents was performed using the \citet{biviano98}
696: model which predicts the time evolution for each row of the detector, taking
697: into account drifts along each orbit and each revolution.
698: % Glitches
699: We masked the glitches using multi-resolution median filtering
700: \citep{starck99} on each block of data after slicing the cube.
701: Additional deglitching was performed manually, examining the temporal cut for
702: each pixel.
703: % Transients
704: We corrected the systematic memory effects using the Fouks-Schubert method
705: \citep{coulais00}.
706: % Flats
707: We computed a hybrid flat-field image placing a mask on the source and
708: computing a flat field outside this mask from the median of the temporal cut
709: for each pixel.
710: For the pixels which were on-source, the flat-field response was set to the
711: corresponding calibration flat-field.
712: % Calibration
713: The conversion from Analog to Digital Units to mJy/pixel was performed using the
714: standard in-flight calibration data base.
715: % Dezodiaqualisation
716: To remove the sky contribution, sources smaller than the array were masked
717: and, for a given wavelength, the median of the pixels which are off-source
718: were subtracted from each pixel.
719: For the more extended sources, we subtracted an independently observed
720: zodiacal spectrum.
721: The contribution of this spectrum was a free parameter varied in order to match
722: the properly sky subtracted fluxes in the LW2 ($6.7\mic$) and LW3 ($14.3\mic$)
723: broadbands.
724: % Final
725: The final product is a 3D spectral-image.
726: We integrated the spectrum using an aperture encompassing the entire
727: galaxy, and obtained a 1D spectrum.
728: When the angular size of the source was larger than the array, we
729: scaled the spectrum by matching \irasi\ broadband flux with that derived
730: from convolving the spectrum with the \irasi\ bandpass.
731:
732:
733: % IRS
734: %----
735: \subsection{\irs\ Spectrum Extraction}
736: \label{sec:irs}
737:
738: Several of the low-metallicity sources in \reftab{tab:source} were not
739: observed by \isocam, thus we complemented our database with mid-IR spectra
740: from the \irs\ spectrometer \citep{houck04irs} on board the \spitzST\
741: \citep{werner04}, when these data were released.
742: Among these galaxies, the spectrum of \sbs\ has been published by
743: \citet{houck04}, \ngc{7714} by \citet{brandl04},
744: \ngc{5253} by \citet{beirao06}, \haro{11} and \izw\ by \citet{wu06}.
745: We refer to these studies for a detailed presentation of their mid-IR
746: properties.
747: We considered only low-resolution data, taken with
748: the SL (Short-Low) module, from $\lambda=5.2\mic$ to $14.5\mic$,
749: and the LL (Long-Low) module, from $\lambda=14.0\mic$ to $38.0\mic$, both
750: with a spectral resolution of $\lambda/\Delta\lambda\simeq64-128$.
751:
752: We retrieved the Basic Calibrated Data (BCD) that have been preprocessed by
753: the Spitzer Science Center (SSC) data reduction pipeline.
754: They have been converted to flux density, corrected for stray light and
755: flatfielded.
756: The extraction of the spectra from the 2D space/wavelength images was
757: performed with the Spectral Modeling, Analysis and Reduction Tool
758: \citep[SMART, version 5.5.6;][]{higdon04}.
759: % 1) Campaign masks
760: We first inspected the BCD images and identified the hot pixels which had not
761: been masked out by the SSC.
762: We replaced them by the median of their neighbors.
763: % 2) Sky subtraction
764: For each module, one order is on-source and the other is off-source, at
765: a time.
766: Then the positions are switched.
767: We subtracted the off-source spectrum from that of the source, in order to
768: remove the sky emission.
769: % 3) Extraction
770: The extraction of the 1D spectrum was performed inside a column whose width
771: varies with wavelength.
772: We have excluded the bonus order.
773: % 4) Coaddition
774: Then, the various frames, for each nod position, were coadded.
775: % 5) Scaling of the two modules
776: Since the long wavelength end of the SL module and the short wavelength end
777: of the LL module were not systematically overlapping, we finally scaled the SL
778: module, in order to obtain a continuous spectrum.
779: This scaling factor can be as large as $50\%$.
780: % 6) Errors
781: At the time of this publication, the \irs\ data handbook recommends
782: that the signal to noise ratio not be derived from the uncertainties generated
783: by the pipeline.
784: Instead, we adopted the recommended systematic error of $20\%$.
785: To take into account these statistical variations, we smoothed the spectra
786: into a $\Delta\lambda\simeq 0.2\mic$ window (4 points), and took the standard
787: deviation inside this window as the error.
788: % Final
789: As we did with the \isocam\ spectra (\refS{sec:cam}), in order
790: to compensate for the fact that we may be overlooking some extended emission,
791: we scaled the spectrum to match the \irasi\ and \irasii.
792: In the case of \iizw, the \irs\ spectrum does not exhibit the PAH
793: features that \citet{madden06} detected in the extended emission, since
794: the \irs\ observations were conducted in staring mode.
795: Indeed, this \irs\ observation does not encompass the region where the
796: tail was detected.
797: Thus we prefer to use the \isocam\ data for this galaxy.
798:
799:
800: %===========================================================================
801:
802:
803: % Methods
804: %--------
805: \section{THE SPECTRAL DECOMPOSITION METHODS}
806: \label{sec:decomp}
807:
808: % Components
809: %-----------
810: \subsection{Inventory of the Physical Components}
811: \label{sec:descMIR}
812:
813: We restrain our study to the spectral range $5-16\mic$, since it
814: is the overlap between the various data sets we have compiled.
815: The main physical components contributing at these wavelengths are the
816: following.
817: \begin{itemize}
818: \item The ionic gas emits strong fine-structure ionic lines.
819: The \neiiiline, \neiiline, \sivline, \ariiiline\ and \ariiline\ are
820: the brightest lines.
821: \item Several ubiquitous broad features, originating from the molecular
822: modes of stochastically heated PAH molecules.
823: The most prominent of them are centered at 6.2, 7.7, 8.6, 11.3 and
824: $12.7\mic$.
825: The \neiiline\ line is blended with the $12.7\mic$ feature.
826: \item Very small grains (VSG), fluctuating in temperature
827: around a few hundred degrees Kelvin, produce a continuum emission.
828: In addition, hot grains in \hii\ regions, at thermal equilibrium with
829: the radiation field, may contribute to this continuum.
830: \item The extinction feature at $9.7\mic$, attributed to amorphous
831: silicates, can be seen toward the most embedded sources.
832: \item Molecular lines are also present, especially the \hmolooiii\ line.
833: However, they are much weaker than the other components.
834: \end{itemize}
835:
836: \begin{figure}[htbp]
837: \centering
838: \includegraphics[width=0.48\textwidth]{f1}
839: \caption{Various spectra normalized to the $11.3\mic$ feature.
840: The top panel shows the total integrated spectra of a low-metallicity
841: galaxy (\um{448}), a normal spiral (\M{51}), and a LIRG
842: (\ngc{4945}).
843: The middle panel shows the spectra of three different regions inside
844: the starburst galaxy \M{82}.
845: The bottom panel shows, for comparison, the absorption
846: coefficient of neutral and cationic PAHs measured in laboratory by
847: \citet{allamandola99}.
848: The solid vertical lines mark the wavelengths of the brightest lines,
849: and the dashed vertical lines mark the wavelengths of the major
850: bands.
851: In this figure and in what follows, $F_\nu(\lambda)$ is the
852: monochromatic flux density at wavelength $\lambda\mic$.}
853: \label{fig:spec}
854: \end{figure}
855: \reffig{fig:spec} shows the variations of the aromatic feature spectrum among
856: galaxies (top panel), inside one galaxy (middle panel) and from a theoretical
857: point of view (bottom panel).
858: This figure shows that the 6.2, 7.7 and $8.6\mic$ features are
859: qualitatively tied together, and that most of the variations in the
860: mid-IR aromatic spectrum is a variation of these three features relative
861: to the $11.3\mic$ band.
862: The lower panel of \reffig{fig:spec} suggests that most of these variations
863: could be explained by a variation of the neutral-to-cationic-PAH ratio.
864: In what follows, we quantify these variations in our sample, in order to
865: explore the validity of this explanation.
866:
867:
868: % Decomposition
869: %--------------
870: \subsection{Measuring the Intensity of the Aromatic Features}
871: \label{sec:method}
872:
873: % PAH profile
874: Measuring the intensity of the aromatic features is uncertain, due to the
875: intrinsic width ($\Delta\lambda\simeq1\mic$) and the complexity of the band
876: profiles.
877: Indeed, contrary to gas lines, a large fraction of the energy of the PAH bands is radiated in the {\it wings}.
878: These wings extend far outside of the central wavelength and can be difficult
879: to reliably disentangle from the underlying continuum emission.
880: Therefore, a proper extraction of the feature requires an assumed band profile.
881: However, the actual profile of each band is not known.
882: It is asymmetric and varies as a function of environment
883: \citep[{e.g.}][]{peeters02b,van-diedenhoven04}.
884: Consequently, several profiles have been used in the literature, motivated by
885: different physical arguments.
886: For example, \citet{boulanger98} proposed that the width of the band is a
887: consequence of the continuous redistribution of intramolecular vibrational
888: energy, between different excitation levels, at high temperature.
889: They estimated the transition timescales to be $\lesssim 10^{-13}$~s, and
890: showed that the PAH bands were well represented by Lorentzian profiles.
891: Conversely, a Drude profile was used by \citet{li01}.
892: Such a profile describes the electric permeability in a solid or a large
893: molecule.
894: Finally, numerous studies simply measure the tip of the aromatic band, and
895: assumes that the total flux radiated in the band scales with this quantity.
896:
897: % Motivations
898: \begin{figure*}[htbp]
899: \centering
900: \begin{tabular}{cc}
901: \includegraphics[width=0.48\textwidth]{f2a} &
902: \includegraphics[width=0.48\textwidth]{f2b} \\
903: \end{tabular}
904: \caption{Demonstration of the two spectral decomposition methods for
905: the total spectrum of \M{82} and \xxxdor.}
906: \label{fig:methods}
907: \end{figure*}
908: In order to take into account the uncertainty of the actual shape of the
909: aromatic bands, we systematically analysed our spectra using two different
910: spectral decomposition methods.
911: These two methods differ in the way the continuum and the aromatic bands are fitted.
912: They are demonstrated in \reffig{fig:methods}, on a PAH-rich spectrum (\M{82}), and PAH-poor one (\xxxdor).
913: \begin{description}
914: % Dutch
915: \item[The \nlmet] (hereafter identified by the letter $\mathcal{S}$),
916: used e.g.\ by \citet{hony01} and \citet{vermeij02}, takes into account
917: only the tip of the aromatic bands when computing their intensities,
918: as shown in left panels of \reffig{fig:methods}.
919: The continuum (grey filled curve) is fitted with a spline function,
920: constrained at the wavelengths $\lambda=5.04$, 5.47, 5.84, 9.18, 10.85,
921: 13.82, 14.60, 15.08, 16.00, $16.15\mic$.
922: A second spline function is fitted to the spectrum, corresponding to the
923: previous continuum, plus the ``{\it plateau}'' under the 6.2, 7.7 and
924: $8.6\mic$ bands, and the plateau under the 11.3 and $12.7\mic$ bands.
925: In addition to the previous wavelengths, this spline function is also
926: constrained at $\lambda=6.61$, 6.80, 7.13, 8.24, 8.80, 11.85, 12.20,
927: $13.22\mic$.
928: The difference between the total spectrum and this continuum+plateau
929: component defines the lines and features (line filled areas of the left
930: panels of \reffig{fig:methods}).
931: This method is relatively robust.
932: However, it does not allow us to separate the \neiiline\ from the
933: $12.7\mic$ band (most of our spectra have a low spectral resolution).
934: % French
935: \item[The \frmet] (hereafter identified by the letter $\mathcal{L}$), used
936: e.g.\ by \citet{boulanger98}, \citet{laurent00} and \citet{verstraete01},
937: assumes that the aromatic bands have lorentz profiles.
938: The wings of these lorentzians account for the underlying plateau.
939: A variation of this method, implemented by \citet{smith07}, consists
940: of modeling the PAH features with Drude profiles.
941: We proceed as following.
942: \begin{enumerate}
943: \item The \neiiiline, \neiiline, \sivline, \ariiiline, and \ariiline\
944: ionic lines are fitted with gaussian functions (diagonal line filling
945: in the right panels of \reffig{fig:methods}).
946: For each (number $i$) of the $N_\sms{line}$ lines, the central
947: frequency, $\nu_i^\sms{line}$, is fixed,
948: the width is fixed by the spectral resolution of the instrument,
949: $\Delta\nu_\sms{spectro}$, and the total flux of each line,
950: $F_i^\sms{line}$, is the free parameter.
951: \item The PAH bands are fitted with lorentzian functions (horizontal line
952: filling in the right panels of \reffig{fig:methods}).
953: The considered features are centered at:
954: $\lambda=5.3$, 5.7, 6.2, 7.7,
955: 8.6, 11.3, 12.0, 12.7, 13.6 and $14.3\mic$.
956: For each (number $i$) of the $N_\sms{band}$ band,
957: the central frequency, $\nu_i^\sms{band}$,
958: the width, $\Delta\nu_i^\sms{band}$, and the total flux, $F_i^\sms{band}$,
959: are the free parameters.
960: We emphasize the fact that we fit only one band around $7.7\mic$, where
961: there are actually two bands centered at $7.6\mic$ and $7.8\mic$.
962: We proceed this way, in order to keep the number of parameters reasonably
963: low.
964: We will use the variation of the centroid of the $7.7\mic$ feature,
965: noted $\lambda_{7.7}$, as an indicator of the $7.6\mic$ to
966: $7.8\mic$ ratio.
967: \item The VSG continuum is fitted with the sum of $N_\sms{cont}$
968: modified black-bodies having the absorption efficiency,
969: $Q_\sms{abs}(\nu)$, of graphite in the Rayleigh approximation
970: \citep[grey filled area on the right panels of \reffig{fig:methods};][]
971: {laor93}.
972: The temperatures, $T_i^\sms{cont}$, and the total fluxes,
973: $F_i^\sms{cont}$, are the free parameters.
974: We adopt $N_\sms{cont}=2$.
975: \item The sum of the previous components is multiplied
976: by $\exp (-\tau)$, where $\tau(\nu)$ is the \citet{mathis90}
977: extinction law, with the \citet{dudley97} silicate features.
978: The column density is the only free parameter.
979: \end{enumerate}
980: All these components are fitted simultaneously as part of one mathematical
981: function:
982: \begin{equation}
983: F_\nu (\nu) =
984: \left(F_\nu^\sms{line}(\nu)+F_\nu^\sms{band}(\nu)+F_\nu^\sms{cont}(\nu)
985: \right)\times\exp\left(-\tau(\nu)\right),
986: \end{equation}
987: with:
988: \begin{equation}
989: \left\{
990: \begin{array}{rcl}
991: F_\nu^\sms{line}(\nu) & = & \displaystyle\sum_{i=1}^{N_{\rm line}}
992: F_i^\sms{line} \,\sqrt{\frac{2}{\pi\Delta\nu_\sms{spectro}^2}}
993: \,\exp\left(-2\frac{(\nu-\nu_i^\sms{line})^2}
994: {\Delta\nu_\sms{spectro}^2}\right) \\
995: &&\\
996: F_\nu^\sms{band}(\nu) & = & \displaystyle\sum_{i=1}^{N_{\rm band}}
997: F_i^\sms{band} \,\frac{1}{2\pi}
998: \,\frac{\Delta\nu_i^\sms{band}}{(\nu-\nu_i^\sms{band})^2
999: +(\Delta\nu_i^\sms{band}/2)^2} \\
1000: &&\\
1001: F_\nu^\sms{cont}(\nu) & = & \displaystyle\sum_{i=1}^{N_{\rm cont}}
1002: F_i^\sms{cont}
1003: \,\frac{B_\nu (T_i^\sms{cont},\nu)\,Q_\sms{abs}(\nu)}
1004: {\displaystyle\int_{c/(16\mic)}^{c/(10\mic)}
1005: B_\nu (T_i^\sms{cont},\nu)\,Q_\sms{abs}(\nu)\ddiff\nu},
1006: \end{array}
1007: \right.
1008: \end{equation}
1009: the various $F_\nu$ being the monochromatic flux densities.
1010: An advantage of the \frmet\ over the \nlmet\ is that it allows us to
1011: separate the \neiiline\ line and the $12.7\mic$ feature, and to study the
1012: variations of the centroids of the bands.
1013: However, when the PAH-to-VSG ratio is very low, as in the case of \xxxdor\
1014: (\reffig{fig:methods}), the width of the features is more uncertain.
1015: \end{description}
1016: In both cases, $\icont$ is defined as the integrated intensity of the continuum
1017: between 10 and $16\mic$, and $\iPah$, the sum of the intensities of all the
1018: bands between 5 and $16\mic$.
1019: We emphasize that our methods work automatically, without any {\it by eye}
1020: adjustment.
1021: This condition prevents systematic effects that could result from arbitrary
1022: choices of parameters.
1023: These two methods are orthogonal but give similar trends \citep[see
1024: \refS{sec:correl} and][]{smith07}.
1025: The \nlmet\ is clearly continuum-biased in its assumptions whereas the
1026: \frmet\ is a line-biased assumption method.
1027:
1028: % Application
1029: Our two methods can be applied to an integrated spectrum, or to each pixel
1030: of a spectro-image, provided that the signal-to-noise ratio is sufficient.
1031: We have systematically applied the two methods to all the spectra presented in
1032: \reftab{tab:source}.
1033: In our analysis, we will systematically consider the results of the two methods
1034: before drawing conclusions on the PAH properties.
1035:
1036:
1037: %===========================================================================
1038:
1039:
1040: % Correlations
1041: %-------------
1042: \section{SYSTEMATIC ANALYSIS OF THE MAJOR BAND RATIOS}
1043: \label{sec:correl}
1044:
1045: % Global measurements
1046: \subsection{Correlations Exhibited Among Integrated Spectra}
1047: \label{sec:corglo}
1048:
1049: \reffigs{fig:corglo1}-\ref{fig:corglo2} show select correlations between
1050: band ratios of the
1051: integrated spectra of our sample (\reftab{tab:source}), obtained with the
1052: two methods presented in \refS{sec:method}.
1053: We focus on the four brightest bands at 6.2, 7.7, 8.6 and $11.3\mic$, and
1054: study the correlations between the various ratios.
1055: The intensities of the features are reported in \refapp{sec:tabfig}
1056: (\reftab{tab:intens}), and
1057: the parameters of the various correlations are given in \reftab{tab:correl}.
1058: As mentioned in \refS{sec:method}, the two methods are uncertain for
1059: very low values of the PAH-to-VSG ratio.
1060: Thus we define two subgroups of data.
1061: \begin{enumerate}
1062: \item The data that we consider to be reliable are marked with black symbols
1063: in \reffigs{fig:corglo1}-\ref{fig:corglo2}.
1064: We define them as the measurements which have a signal-to-noise ratio,
1065: at $\lambda=7.7\mic$, larger than 6, and
1066: $\iPah/\icont \geq 0.5$ with the \frmet, and $\iPah/\icont \geq 0.3$
1067: with the \nlmet.
1068: \item The fits that we consider to be less certain are the complementary data
1069: (grey symbols in \reffigs{fig:corglo1}-\ref{fig:corglo2}), having lower
1070: signal-to-noise and PAH-to-VSG ratios.
1071: \end{enumerate}
1072:
1073: \reffig{fig:corglo1} shows that there is an excellent linear correlation between the ratios $\ipah{6.2}/\ipah{11.3}$, $\ipah{7.7}/\ipah{11.3}$ and
1074: $\ipah{8.6}/\ipah{11.3}$, using either of the two methods.
1075: The variations of these ratios spread roughly over one order of magnitude, while
1076: the ratios $\ipah{6.2}/\ipah{7.7}$, and $\ipah{7.7}/\ipah{8.6}$ are roughly
1077: constant within the error bars (\reffig{fig:corglo2}).
1078: In general, the measure of the intensity of the $8.6\mic$ band is
1079: less accurate than for the other main bands.
1080: Indeed, it is less intense than the 6.2 and $7.7\mic$ features.
1081: In addition, it is merged with the long wavelength wing of the $7.7\mic$ band.
1082: Finally, this particular feature is significantly affected by the silicate
1083: extinction feature around $9.7\mic$, when the source is deeply embedded.
1084: This extinction is not corrected in the case of the \nlmet, and is
1085: corrected very simply in the case of the \frmet.
1086: That is the reason why correlations involving the $8.6\mic$ feature are
1087: always more dispersed than the others.
1088: Quantitatively, the correlation coefficients are around $0.8$ for the
1089: four panels of \reffig{fig:corglo1}, except for that involving the
1090: $\ipah{8.6}$ feature, with the \nlmet.
1091: Most of the outsiders are based on fits that we consider to be
1092: uncertain.
1093:
1094: \reffig{fig:compglo} shows the consistency between the two methods.
1095: It demonstrates that, whatever method we use to measure the band ratios,
1096: the order of the various measures is conserved; i.e.\ a spectrum
1097: $\mathcal{A}$ having a lower $\ipah{6.2}/\ipah{11.3}$ ratio than a spectrum
1098: $\mathcal{B}$, using the \nlmet, will also have a lower ratio than
1099: $\mathcal{B}$ using the \frmet.
1100: Moreover, the fact that both methods give similar trends means that
1101: we did not artificially incorporate part of the continuum flux within the
1102: aromatic band intensities.
1103: Indeed, it is possible to accidentally account for a fraction of the continuum
1104: intensity within the wings of the PAH profile, with the \frmet,
1105: especially within the $11.3\mic$ feature.
1106: However, this bias is not possible with the \nlmet, since it integrates only
1107: the tip of the band.
1108: Therefore, the agreement between the two methods allow us to rule out this bias
1109: and claim that the trends of \reffig{fig:corglo1} are not induced by the fitting
1110: methods.
1111:
1112: These first relations, established on integrated spectra, indicate that the
1113: properties of the PAHs throughout different types of galaxies and Galactic
1114: regions are remarkably homogeneous.
1115: They are consistent with a significant variation of the 6.2, 7.7 and $8.6\mic$
1116: features relative to the $11.3\mic$ band, coupled with an absence of significant variations among the 6.2, 7.7 and $8.6\mic$ bands.
1117: \begin{figure*}[htbp]
1118: \centering
1119: \includegraphics[width=\textwidth]{f3a} \\
1120: \includegraphics[width=\textwidth]{f3b}
1121: \caption{PAH properties for integrated spectra.
1122: The black error bars are for the measurements that we consider to be
1123: reliable, while the grey are more uncertain.
1124: In each panel representing a relation
1125: $Y$ versus $X$, the grey filled area is the domain between
1126: $Y = a_{Y/X}\times X - 1\sigma_{Y/X}$, and
1127: $Y = a_{Y/X}\times X + 1\sigma_{Y/X}$,
1128: $Y = a_{Y/X}\times X$ being the linear correlation of the data, and
1129: $\sigma_{Y/X}$, the dispersion of the data around this correlation.}
1130: \label{fig:corglo1}
1131: \end{figure*}
1132: \begin{figure*}[htbp]
1133: \centering
1134: \begin{tabular}{cc}
1135: \includegraphics[width=0.48\textwidth]{f4a} &
1136: \includegraphics[width=0.48\textwidth]{f4b} \\
1137: \end{tabular}
1138: \caption{PAH properties for integrated spectra (continued).
1139: The same symbol conventions are adopted as in \reffig{fig:corglo1}.}
1140: \label{fig:corglo2}
1141: \end{figure*}
1142: \begin{figure}[htbp]
1143: \centering
1144: \includegraphics[width=0.48\textwidth]{f5}
1145: \caption{Comparison between the two methods.
1146: The same symbol conventions are adopted as in \reffig{fig:corglo1}.}
1147: \label{fig:compglo}
1148: \end{figure}
1149:
1150: Finally, as a consistency check, \reffig{fig:cordist} shows the
1151: variations of the band ratios as a function of distance.
1152: The correlation coefficient is 0.23 for the \frmet, and 0.20 for the \nlmet.
1153: This absence of correlation shows the uniformity of the sample and rules out
1154: the possibility of variations induced by aperture effects.
1155: \begin{figure}[htbp]
1156: \centering
1157: \includegraphics[width=0.48\textwidth]{f6}
1158: \caption{Correlation of the band ratio with the distance.
1159: The absence of correlation shows the uniformity of the sample.
1160: The same symbol conventions are adopted as in \reffig{fig:corglo1}.}
1161: \label{fig:cordist}
1162: \end{figure}
1163:
1164:
1165: % Spatial measurements
1166: \subsection{Spatial Variations of the Band Properties}
1167: \label{sec:cormap}
1168:
1169: We now study the spatial variations of the properties of the mid-IR features,
1170: by analyzing the spectral maps of a sub-sample of sources:
1171: \IC{342}, \M{17}, \M{51}, \M{82}, \M{83}, \xxxdor, and the \orb.
1172: These objects are those which are spatially resolved, satisfy the two
1173: PAH-to-continuum and signal-to-noise ratio thresholds listed in
1174: \refS{sec:corglo}, and show significant variations of the band ratios.
1175: Similar to \refS{sec:corglo}, we systematically compare the results of the
1176: two methods, but instead of applying them to integrated spectra, we fit the
1177: spectrum of each pixel of the spectral maps.
1178: We then degrade the images of each spectral component to the spatial resolution
1179: of the longest wavelength of the spectrum (${\rm FWHM} \simeq 9''$).
1180: We achieve this by convolving the image of a given component
1181: (e.g.\ $\ipah{6.2}$), with the PSF at the longest wavelength
1182: (i.e.\ $\lambda=16\mic$) deconvolved beforehand by the PSF of the component
1183: (i.e.\ $\lambda=6.2\mic$).
1184: \reffigs{fig:cormap1m82} to \ref{fig:compmapm51} (as well as
1185: \reffigs{fig:cormap1ic342} to \ref{fig:compmapm17} in \refapp{sec:tabfig}) show
1186: the same correlations as in \reffigs{fig:corglo1}-\ref{fig:compglo},
1187: obtained inside resolved sources, instead of integrated spectra.
1188: Each data point represents the fit of the spectrum of one pixel.
1189: We selected the points, according to the criterion defined in
1190: \refS{sec:corglo}:
1191: the signal-to-noise ratio at $\lambda=7.7\mic$ must be larger than 6, and
1192: $\iPah/\icont \geq 0.5$ with the \frmet, and $\iPah/\icont \geq 0.3$
1193: with the \nlmet.
1194: The pixels which do not fall into this category have not been considered here.
1195: The parameters and statistical quantities relative to these correlations
1196: are given in \reftab{tab:correl}.
1197: \reffigs{fig:imm82} to \ref{fig:imm51} (as well as \reffigs{fig:imic342}
1198: to \ref{fig:imm17} in \refapp{sec:tabfig})
1199: show the spatial distributions of the components.
1200:
1201: In general, the correlations presented in
1202: \reffigs{fig:corglo1}-\ref{fig:compglo} still hold inside individual
1203: objects as well.
1204: Therefore, the variations of the PAH band ratios are independent of the
1205: spatial resolution.
1206: For example, the pixel size is $\simeq0.1$~pc in \M{17} and the \orb, while
1207: it is $\simeq1$~kpc in external galaxies.
1208: From the images, we notice that the ratio $\ipah{6.2}/\ipah{11.3}$
1209: (as well as the ratios $\ipah{7.7}/\ipah{11.3}$ and $\ipah{8.6}/\ipah{11.3}$)
1210: is roughly correlated with the intensity of the PAH emission, the highest
1211: ratios being found in the brightest regions (spiral arms, starburst region,
1212: etc.), while the lowest ratios are generally found in the extended,
1213: low-luminosity regions.
1214: We do not detect any significant variations of the centroid of the $7.7\mic$
1215: feature inside each source.
1216: However, the average centroid varies from one source to another.
1217:
1218: \refapp{sec:tabfig} comments on each individual source.
1219: Here we will illustrate the systematic variations within the individual sources
1220: by focusing on two extragalactic sources: the edge-on irregular
1221: starburst galaxy \M{82} and the face-on spiral galaxy \M{51}.
1222: For \M{82} (\reffig{fig:imm82}), the maximum values of the
1223: $\ipah{7.7}/\ipah{11.3}$ ratio are found along the disc, in the star forming
1224: region, where the infrared emission is the highest.
1225: The value of the band ratio drops by a factor of $\simeq3$ at $\simeq200$~pc
1226: above and below the disc, in the halo.
1227: For \M{51} (\reffig{fig:imm51}), the $\ipah{7.7}/\ipah{11.3}$ ratio is maximum
1228: along the circumnuclear ring.
1229: It is somewhat lower in the $10''$ nuclear region.
1230: The band ratio is also high along the spiral arms, coinciding with the infrared
1231: bright blobs, but can exhibit variations by a factor of $\simeq2$.
1232: Finally, the ratio drops in the interarm region.
1233:
1234: The highest values of the $\ipah{7.7}/\ipah{11.3}\simeq7$ ratio
1235: (\frmet), found in the nuclear star forming regions of \M{82} and \M{51}, are similar to the values found in the compact Galactic \hii\ regions \IR{15384} and
1236: \IR{18317} (\reftab{tab:intens}).
1237: Intermediate values of $\ipah{7.7}/\ipah{11.3}\simeq3$ (\frmet), similar to the
1238: Galactic reflection nebula \ngc{2023} are found in the halo of \M{82} and along
1239: spiral arms of \M{51}.
1240: The lowest values of $\ipah{7.7}/\ipah{11.3}\lesssim 1$ (\frmet), seen in the planetary nebula \ngc{7027}, correspond to the outermost regions of the two
1241: galaxies.
1242:
1243:
1244: % Spatial correlations
1245: %---------------------
1246: % M 82
1247: \begin{figure*}[htbp]
1248: \centering
1249: \includegraphics[width=\textwidth]{f7a} \\
1250: \includegraphics[width=\textwidth]{f7b}
1251: \caption{PAH band ratios within \M{82}.
1252: The grey filled areas are the correlations obtained for the
1253: integrated spectra (\reffig{fig:corglo1}).
1254: The correlations obtained inside \M{82} are represented by the
1255: hatched region.
1256: The white symbol is the value of the global measurement over
1257: the entire galaxy.}
1258: \label{fig:cormap1m82}
1259: \end{figure*}
1260: \begin{figure*}[htbp]
1261: \centering
1262: \begin{tabular}{cc}
1263: \includegraphics[width=0.48\textwidth]{f8a} &
1264: \includegraphics[width=0.48\textwidth]{f8b} \\
1265: \end{tabular}
1266: \caption{PAH band ratios within \M{82} (continued).}
1267: \label{fig:cormap2m82}
1268: \end{figure*}
1269: \begin{figure}[htbp]
1270: \centering
1271: \includegraphics[width=0.48\textwidth]{f9}
1272: \caption{Comparison between the two methods in \M{82}.
1273: The same symbol conventions are adopted as in
1274: \reffig{fig:cormap1m82}.}
1275: \label{fig:compmapm82}
1276: \end{figure}
1277: \clearpage
1278: % M 51
1279: \begin{figure*}[htbp]
1280: \centering
1281: \includegraphics[width=\textwidth]{f10a} \\
1282: \includegraphics[width=\textwidth]{f10b}
1283: \caption{PAH band ratios within \M{51}.
1284: The same symbol conventions are adopted as in
1285: \reffig{fig:cormap1m82}.}
1286: \label{fig:cormap1m51}
1287: \end{figure*}
1288: \begin{figure*}[htbp]
1289: \centering
1290: \begin{tabular}{cc}
1291: \includegraphics[width=0.48\textwidth]{f11a} &
1292: \includegraphics[width=0.48\textwidth]{f11b} \\
1293: \end{tabular}
1294: \caption{PAH band ratios within \M{51} (continued).}
1295: \label{fig:cormap2m51}
1296: \end{figure*}
1297: \begin{figure}[htbp]
1298: \centering
1299: \includegraphics[width=0.48\textwidth]{f12}
1300: \caption{Comparison between the two methods in \M{51}.
1301: The same symbol conventions are adopted as in
1302: \reffig{fig:cormap1m82}.}
1303: \label{fig:compmapm51}
1304: \end{figure}
1305: \clearpage
1306:
1307: % Images
1308: %-------
1309: \begin{figure*}[htbp]
1310: % M 82
1311: \centering
1312: \plottwo{f13a}{f13b}
1313: \caption{Spatial distribution of the PAHs in \M{82}.
1314: For each method, the image is the total PAH intensity, and the
1315: contours are the $\ipah{6.2}/\ipah{11.3}$ ratio.
1316: The shaded circle indicates the beam size.}
1317: \label{fig:imm82}
1318: % M 51
1319: \centering
1320: \plottwo{f14a}{f14b}
1321: \caption{Spatial distribution of the PAHs in \M{51}.
1322: For each method, the image is the total PAH intensity, and the
1323: contours are the $\ipah{6.2}/\ipah{11.3}$ ratio.
1324: The shaded circle indicates the beam size.}
1325: \label{fig:imm51}
1326: \end{figure*}
1327: \clearpage
1328:
1329: % Statistics
1330: %-----------
1331: \begin{deluxetable}{llllrrrrrrrl}
1332: % Head
1333: %-----
1334: \tabletypesize{\footnotesize}
1335: \rotate
1336: \tablecolumns{8}
1337: \tablewidth{0pc}
1338: \tablecaption{Parameters of the various correlations.}
1339: \tablehead{
1340: \colhead{Source} & \colhead{$X$} & \colhead{$Y$} & \colhead{Method}
1341: & \colhead{$a_{Y/X}$} & \colhead{$b_{Y/X}$} & \colhead{$\sigma_{Y/X}$}
1342: & \colhead{$\rho_{Y/X}$} }
1343: % Data
1344: %-----
1345: \startdata
1346: % Global
1347: Integrated spectra & $\ipah{6.2}/\ipah{11.3}$
1348: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1349: & $1.78$ & $0$ & $0.69$
1350: & $0.89$ \\
1351: (\reffig{fig:corglo1})
1352: & & & $\mathcal{L}$
1353: & $2.70$ & $0$ & $0.64$
1354: & $0.85$ \\
1355: %
1356: & $\ipah{8.6}/\ipah{11.3}$
1357: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1358: & $8.21$ & $0$ & $1.09$
1359: & $0.44$ \\
1360: & & & $\mathcal{L}$
1361: & $4.43$ & $0$ & $0.79$
1362: & $0.78$ \\
1363: %
1364: & $\ipah{6.2}/\ipah{11.3}$
1365: & $\ipah{6.2}/\ipah{11.3}$ & $\mathcal{S}/\mathcal{L}$
1366: & $1.03$ & $0$ & $0.46$
1367: & $0.45$ \\
1368: %
1369: % (\reffig{fig:cent77})
1370: % & $\log(Z_\sms{ISM}/Z_\odot)$ &
1371: % $\lambda_{7.7}\;[\mu m]$ &
1372: % $\mathcal{L}$
1373: % & $0.050$ & $7.69$ & $0.024$
1374: % & \nodata \\
1375: \hline
1376: % M 82
1377: \M{82} & $\ipah{6.2}/\ipah{11.3}$
1378: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1379: & $1.86$ & $0$ & $0.32$
1380: & $0.97$ \\
1381: (\reffig{fig:cormap1m82})
1382: & & & $\mathcal{L}$
1383: & $2.70$ & $0$ & $0.32$
1384: & $0.97$ \\
1385: %
1386: & $\ipah{8.6}/\ipah{11.3}$
1387: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1388: & $8.06$ & $0$ & $0.53$
1389: & $0.87$ \\
1390: & & & $\mathcal{L}$
1391: & $4.04$ & $0$ & $0.39$
1392: & $0.95$ \\
1393: %
1394: & $\ipah{6.2}/\ipah{11.3}$
1395: & $\ipah{6.2}/\ipah{11.3}$ & $\mathcal{S}/\mathcal{L}$
1396: & $1.07$ & $0$ & $0.23$
1397: & $0.88$ \\
1398: %
1399: & $\ipah{7.7}/\ipah{11.3}$
1400: & $\lambda_{7.7}\;[\mu m]$ & $\mathcal{L}$
1401: & $0$ & $7.68$ & $0.012$
1402: & \nodata \\
1403: \hline
1404: % IC 342
1405: \IC{342} & $\ipah{6.2}/\ipah{11.3}$
1406: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1407: & $1.50$ & $0$ & $0.63$
1408: & $0.82$ \\
1409: (\reffig{fig:cormap1ic342})
1410: & & & $\mathcal{L}$
1411: & $2.33$ & $0$ & $1.35$
1412: & $0.79$ \\
1413: %
1414: & $\ipah{8.6}/\ipah{11.3}$
1415: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1416: & $5.71$ & $0$ & $0.69$
1417: & $0.81$ \\
1418: & & & $\mathcal{L}$
1419: & $1.89$ & $0$ & $1.41$
1420: & $0.66$ \\
1421: %
1422: & $\ipah{6.2}/\ipah{11.3}$
1423: & $\ipah{6.2}/\ipah{11.3}$ & $\mathcal{S}/\mathcal{L}$
1424: & $1.12$ & $0$ & $0.40$
1425: & $0.74$ \\
1426: %
1427: & $\ipah{7.7}/\ipah{11.3}$
1428: & $\lambda_{7.7}\;[\mu m]$ & $\mathcal{L}$
1429: & $0$ & $7.68$ & $0.012$
1430: & \nodata \\
1431: \hline
1432: %M 51
1433: \M{51} & $\ipah{6.2}/\ipah{11.3}$
1434: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1435: & $1.49$ & $0$ & $0.50$
1436: & $0.62$ \\
1437: (\reffig{fig:cormap1m51})
1438: & & & $\mathcal{L}$
1439: & $2.59$ & $0$ & $0.58$
1440: & $0.65$ \\
1441: %
1442: & $\ipah{8.6}/\ipah{11.3}$
1443: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1444: & $7.19$ & $0$ & $0.62$
1445: & $0.41$ \\
1446: & & & $\mathcal{L}$
1447: & $4.55$ & $0$ & $0.56$
1448: & $0.65$ \\
1449: %
1450: & $\ipah{6.2}/\ipah{11.3}$
1451: & $\ipah{6.2}/\ipah{11.3}$ & $\mathcal{S}/\mathcal{L}$
1452: & $1.01$ & $0$ & $0.42$
1453: & $0.33$ \\
1454: %
1455: & $\ipah{7.7}/\ipah{11.3}$
1456: & $\lambda_{7.7}\;[\mu m]$ & $\mathcal{L}$
1457: & $0$ & $7.72$ & $0.012$
1458: & \nodata \\
1459: \hline
1460: %M 83
1461: \M{83} & $\ipah{6.2}/\ipah{11.3}$
1462: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1463: & $1.30$ & $0$ & $0.52$
1464: & $0.62$ \\
1465: (\reffig{fig:cormap1m83})
1466: & & & $\mathcal{L}$
1467: & $2.19$ & $0$ & $0.54$
1468: & $0.67$ \\
1469: %
1470: & $\ipah{8.6}/\ipah{11.3}$
1471: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1472: & $7.18$ & $0$ & $0.69$
1473: & $0.43$ \\
1474: & & & $\mathcal{L}$
1475: & $3.52$ & $0$ & $0.70$
1476: & $0.66$ \\
1477: %
1478: & $\ipah{6.2}/\ipah{11.3}$
1479: & $\ipah{6.2}/\ipah{11.3}$ & $\mathcal{S}/\mathcal{L}$
1480: & $1.07$ & $0$ & $0.41$
1481: & $0.39$ \\
1482: %
1483: & $\ipah{7.7}/\ipah{11.3}$
1484: & $\lambda_{7.7}\;[\mu m]$ & $\mathcal{L}$
1485: & $0$ & $7.71$ & $0.016$
1486: & \nodata \\
1487: \hline
1488: % 30 Dor
1489: \xxxdor & $\ipah{6.2}/\ipah{11.3}$
1490: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1491: & $1.43$ & $0$ & $0.79$
1492: & $0.78$ \\
1493: (\reffig{fig:cormap130dor})
1494: & & & $\mathcal{L}$
1495: & $2.31$ & $0$ & $0.67$
1496: & $0.68$ \\
1497: %
1498: & $\ipah{8.6}/\ipah{11.3}$
1499: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1500: & $11.48$ & $0$ & $1.61$
1501: & $0.35$ \\
1502: & & & $\mathcal{L}$
1503: & $2.10$ & $0$ & $0.62$
1504: & $0.49$ \\
1505: %
1506: & $\ipah{6.2}/\ipah{11.3}$
1507: & $\ipah{6.2}/\ipah{11.3}$ & $\mathcal{S}/\mathcal{L}$
1508: & $2.01$ & $0$ & $0.86$
1509: & $0.44$ \\
1510: %
1511: & $\ipah{7.7}/\ipah{11.3}$
1512: & $\lambda_{7.7}\;[\mu m]$ & $\mathcal{L}$
1513: & $0$ & $7.66$ & $0.016$
1514: & \nodata \\
1515: \hline
1516: % M 17
1517: \M{17} & $\ipah{6.2}/\ipah{11.3}$
1518: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1519: & $1.96$ & $0$ & $0.33$
1520: & $0.81$ \\
1521: (\reffig{fig:cormap1m17})
1522: & & & $\mathcal{L}$
1523: & $2.41$ & $0$ & $0.45$
1524: & $0.93$ \\
1525: %
1526: & $\ipah{8.6}/\ipah{11.3}$
1527: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1528: & $11.91$ & $0$ & $0.81$
1529: & $0.33$ \\
1530: & & & $\mathcal{L}$
1531: & $4.04$ & $0$ & $0.84$
1532: & $0.80$ \\
1533: %
1534: & $\ipah{6.2}/\ipah{11.3}$
1535: & $\ipah{6.2}/\ipah{11.3}$ & $\mathcal{S}/\mathcal{L}$
1536: & $1.01$ & $0$ & $0.25$
1537: & $0.61$ \\
1538: %
1539: & $\ipah{7.7}/\ipah{11.3}$
1540: & $\lambda_{7.7}\;[\mu m]$ & $\mathcal{L}$
1541: & $0$ & $7.69$ & $0.012$
1542: & \nodata \\
1543: \hline
1544: % Orion Bar
1545: \orb & $\ipah{6.2}/\ipah{11.3}$
1546: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1547: & $1.77$ & $0$ & $0.32$
1548: & $0.83$ \\
1549: (\reffig{fig:cormap1orionBar})
1550: & & & $\mathcal{L}$
1551: & $2.32$ & $0$ & $0.21$
1552: & $0.95$ \\
1553: %
1554: & $\ipah{8.6}/\ipah{11.3}$
1555: & $\ipah{7.7}/\ipah{11.3}$ & $\mathcal{S}$
1556: & $8.38$ & $0$ & $0.38$
1557: & $0.91$ \\
1558: & & & $\mathcal{L}$
1559: & $3.72$ & $0$ & $0.52$
1560: & $0.86$ \\
1561: %
1562: & $\ipah{6.2}/\ipah{11.3}$
1563: & $\ipah{6.2}/\ipah{11.3}$ & $\mathcal{S}/\mathcal{L}$
1564: & $1.01$ & $0$ & $0.29$
1565: & $0.40$ \\
1566: %
1567: & $\ipah{7.7}/\ipah{11.3}$
1568: & $\lambda_{7.7}\;[\mu m]$ & $\mathcal{L}$
1569: & $0$ & $7.68$ & $0.011$
1570: & \nodata \\
1571: \enddata
1572: \label{tab:correl}
1573: % \tablerefs{}
1574: \tablecomments{This table gives, for two given measured quantities $X$ and
1575: $Y$, the parameters corresponding to the fit of $Y$ by
1576: $a_{Y/X}\times X + b_{Y/X}$.
1577: The dispersion of the measures around this correlation is
1578: $\sigma_{Y/X}
1579: =\sqrt{\langle(Y-a_{Y/X}\times X+ b_{Y/X})^2\rangle}$.
1580: In the case where $b_{Y/X}=0$, $\rho_{Y/X}$ is the linear
1581: correlation coefficient.
1582: }
1583: \end{deluxetable}
1584: \clearpage
1585:
1586:
1587: %===========================================================================
1588:
1589:
1590: % Interpretation
1591: %---------------
1592: \section{INTERPRETATION OF THE AROMATIC FEATURE VARIATIONS}
1593: \label{sec:explanation}
1594:
1595: % Band ratios
1596: \subsection{Origin of the Band Ratio Variations}
1597: \label{sec:pahU}
1598:
1599: The main conclusions of \refS{sec:correl} are
1600: that\textlist{\thetextlist~the ratios $\ipah{6.2}/\ipah{11.3}$,
1601: $\ipah{7.7}/\ipah{11.3}$ and $\ipah{8.6}/\ipah{11.3}$
1602: are correlated and span one order of magnitude throughout our sample, and
1603: \thetextlist~the ratios $\ipah{7.7}/\ipah{6.2}$ and $\ipah{8.6}/\ipah{6.2}$
1604: are not significantly anticorrelated with the
1605: $\ipah{7.7}/\ipah{11.3}$ ratio, and do not show significant variations
1606: within the sources which have the best signal-to-noise ratios (especially
1607: \M{82}).}As
1608: explained in \refS{sec:intro} and demonstrated in \reffig{fig:spec},
1609: the intensity of the features between 6 and $9\mic$ relative to the
1610: $11.3\mic$ band are one order of magnitude higher for PAH$^+$ than for PAH$^0$.
1611: Therefore, the universal correlations displayed from \reffig{fig:corglo1} to
1612: \reffig{fig:compmaporionBar} can be attributed to variations of the charge of
1613: the carriers of the aromatic features.
1614: However, before concluding, we first need to explore the ability of other
1615: physical processes to reproduce these trends.
1616:
1617: First, dehydrogenation of the PAHs have an effect similar to ionization,
1618: on the mid-IR spectrum, as shown by several laboratory studies
1619: \citep{allamandola89,allain96,jochims99}.
1620: However, as reported by \citet{allain96}, only PAHs containing less than
1621: $N_\sms{C}\simeq 50$ carbon atoms can be considerably dehydrogenated.
1622: At the same time, this threshold corresponds to the minimum size of PAHs
1623: that can survive in most PDRs \citep{allain96}.
1624: For comparison, the PAH size distribution derived by \citet{zubko04}, for the
1625: diffuse Galactic ISM, has a lower cut-off of $N_\sms{C}^\sms{min}=20$.
1626: Their emission is dominated by PAHs of $N_\sms{C}\simeq50-500$, depending
1627: on their excitation rate.
1628: This conclusion is supported by the observed linear correlation of the
1629: 3.3$\mic$ band with the 11.3$\mic$ band in Galactic sources \citep{hony01}.
1630: If dehydrogenation were important then a non-linear relation would be expected as duo and trio groups (with bands longwards of 11.3$\mic$) were converted into
1631: solo groups \citep[with a 11.3$\mic$ band;][]{schutte93,hony01}.
1632: Likewise, if dehydrogenation were important, then the decrease in the
1633: $\ipah{11.3}/\ipah{6.2}$ ratio (decreasing H coverage) would be expected to
1634: be accompanied by a decreasing $\ipah{12.7}/\ipah{11.3}$ ratio (conversion of
1635: duo/trio's into solo's) in contrast to the observations \citep{hony01}.
1636: Therefore, we can neglect the effect of dehydrogenation on our mid-IR spectra.
1637:
1638: Second, deep extinction by the silicate feature at $9.7\mic$ can
1639: cause variations of the band ratios, as proposed by \citet{peeters02},
1640: \citet{spoon02} or \citet{brandl06}.
1641: Indeed, the $11.3\mic$ feature is more absorbed than the $7.7\mic$ band.
1642: This effect would also cause the $8.6\mic$ feature to suffer from the
1643: same amount of extinction as the $11.3\mic$ band, and these two bands should
1644: be correlated.
1645: However, this is not what \reffigs{fig:corglo1} to \reffig{fig:cormap2m17}
1646: show.
1647: Moreover, the \frmet\ corrects the bands for extinction (\refS{sec:decomp}).
1648: The good agreement between the \frmet\ and the \nlmet\ tells us that extinction
1649: can not explain the majority of the observed band ratio variations.
1650: Finally, there is no evidence for silicate absorption in the spectra of
1651: the \orb.
1652: Indeed, the total gas and dust column along the line of sight for this source is
1653: insufficient to produce noticeable 10$\mic$ absorption.
1654:
1655: \begin{figure*}[htbp]
1656: \centering
1657: \includegraphics[width=\textwidth]{f15}
1658: \caption{Effects of the temperature distribution on the PAH spectra.
1659: The various curves are the theoretical emission spectra, varying the
1660: shape (Galactic or young starburst) and
1661: the intensity of the ISRF ($\chi_\sms{ISRF}$), as well as the lower
1662: cut-off of the size distribution ($N_\sms{C}^\sms{min}$).
1663: The size distribution is the BARE-GR-S of \citet{zubko04}.
1664: The absorption efficiencies are taken from \citet{draine07}.}
1665: \label{fig:pah_th}
1666: \end{figure*}
1667: % PAH emission modeling
1668: Third, a modification of the temperature distribution of the PAHs
1669: can affect the relative weight of the aromatic features.
1670: For example, truncating the highest temperatures would have the effect of
1671: decreasing the intensity of the short wavelength bands relative to the long
1672: wavelength ones.
1673: There are essentially two ways of modifying the temperature distribution of the
1674: molecules:\textlist{\thetextlist~by varying the interstellar
1675: radiation field (ISRF) intensity or shape, and
1676: \thetextlist~by changing their size
1677: distribution.}The latter effect could happen in the vicinity of a very hard
1678: radiation source, like an AGN, where the smallest PAHs could undergo
1679: photosublimation, as proposed by \citet{smith07}.
1680:
1681: We performed theoretical modeling of the PAH emission spectra,
1682: in order to study the latter effects.
1683: These spectra, which take into account the stochastic heating of the particles,
1684: are shown in \reffig{fig:pah_th}.
1685: We adopted the PAH size distribution by \citet[][bare grains, solar abundance constraints; BARE-GR-S]{zubko04}, and the absorption efficiencies by
1686: \citet{draine07}, with an ionization fraction of $50\%$.
1687: We consider successively two classes of ISRF, in order to explore the effect
1688: of the hardness of the radiation on the PAH
1689: emission:\textlist{\thetextlist~the Galactic ISRF of \citet{mathis83} and
1690: \thetextlist~a very hard ISRF, corresponding to an instantaneous burst
1691: of star formation with a Salpeter IMF, synthesized with the stellar
1692: evolutionary model \peg\ \citep{fioc97}.}The
1693: latter is normalised to the intensity of the \citet{mathis83} ISRF.
1694: We also vary the intensity of each ISRF, by multiplying them
1695: by a factor $\chi_\sms{ISRF}=1-10^5$.
1696: In addition, we explore the effect of a possible small PAH segregation, by
1697: varying the value of the lower cut-off of the PAH size distribution.
1698: This lower cut-off is $N_\sms{C}^\sms{min}=20$ carbon atoms, for the
1699: \citet{zubko04} model.
1700: We vary it up to $N_\sms{C}^\sms{min}=10^3$.
1701: Finally, we apply a screen extinction to the modeled spectra
1702: by multiplying them by a factor $\exp[-\tau(\lambda,A_V)]$.
1703: We adopt the \citet{zubko04} optical depth, $\tau(\lambda,A_V)$, and
1704: plot the results using two values of the V band attenuation: $A_V=0$
1705: (no extinction) and $A_V=10$.
1706:
1707: % Simulation
1708: \begin{figure*}[htbp]
1709: \centering
1710: \includegraphics[width=0.48\textwidth]{f16}
1711: \caption{Theoretical modeling of the effect of the PAH temperature
1712: distribution on their band ratios, with a fixed
1713: ionization fraction ($50\%$).
1714: No extinction is applied.}
1715: \label{fig:pahU1}
1716: \end{figure*}
1717: \begin{figure*}[htbp]
1718: \centering
1719: \includegraphics[width=0.48\textwidth]{f17}
1720: \caption{The two panels present the results of the same modeling as in
1721: \reffig{fig:pahU1}, the only difference being that the top panel
1722: shows the resulting
1723: $\ipah{7.7}/\ipah{6.2}$, while the bottom panel shows the
1724: $\ipah{8.6}/\ipah{6.2}$.
1725: To read this figure, first look at the black empty
1726: polygon with the circled corners.
1727: This polygon shows the effect of varying both the minimum cut-off
1728: PAH size, $N_\sms{C}^\sms{min}$, and the ISRF intensity,
1729: $\chi_\sms{ISRF}$, with the Galactic ISRF (labeled {\it Galaxy})
1730: and with no extinction.
1731: The values between parenthesis, close to each corners are the values
1732: of these two parameters at the corner.
1733: Second, the grey empty polygon is the analog of the previous one,
1734: except that the ISRF is now the Starburst one (labeled {\it Burst}).
1735: Finally, the two grey filled polygons are the analog of the two
1736: previous ones, except that we applied a screen extinction with
1737: $A_V=10$.
1738: As a summary, the various arrows show the sense of variation of the
1739: band ratios with each parameter, including the ionization degree of
1740: the PAHs.}
1741: \label{fig:pahU2}
1742: \end{figure*}
1743: % Exploration of the temperature distribution effects
1744: \reffigs{fig:pahU1}-\ref{fig:pahU2} synthesize the results of this modeling.
1745: \reffig{fig:pahU1} demonstrates that effect of the
1746: radiation field intensity on the $\ipah{7.7}/\ipah{11.3}$ band ratio, for
1747: the two types of radiation fields and the two extreme PAH size cut-offs.
1748: This figure shows that, for a given size distribution and ISRF type, the
1749: $\ipah{7.7}/\ipah{11.3}$ ratio is independent of the intensity of the ISRF,
1750: up to $\chi_\sms{ISRF}\simeq10^4$.
1751: Indeed, the PAHs are stochastically heated below $\chi_\sms{ISRF}\simeq10^4$,
1752: therefore the shape of their emission spectrum is independent of the ISRF.
1753: However, when $\chi_\sms{ISRF}\gtrsim10^4$, the largest PAHs reach thermal
1754: equilibrium, and the temperature fluctuation range shrinks significantly enough
1755: to affect the ratio between the bands.
1756: % The key point
1757: \reffig{fig:pahU2} shows how every effect considered here
1758: affects the relation between $\ipah{7.7}/\ipah{11.3}$ and
1759: $\ipah{7.7}/\ipah{6.2}$ and $\ipah{8.6}/\ipah{6.2}$.
1760: The arrows indicate the direction toward which these parameters affect the
1761: relation.
1762: This figure shows that the combination of all the effects affecting the
1763: PAH temperature distribution (ISRF intensity, ISRF hardness and size distribution) can not explain
1764: a variation of the $\ipah{7.7}/\ipah{11.3}$ ratio by more than a factor
1765: $\simeq3$.
1766: Moreover, it shows that if these effects were responsible for the
1767: variation of the $\ipah{7.7}/\ipah{11.3}$ ratio, then the
1768: $\ipah{7.7}/\ipah{6.2}$ ratio would also be anticorrelated with it.
1769: The ratio $\ipah{7.7}/\ipah{6.2}$ is particularly very sensitive to
1770: $N_\sms{C}^\sms{min}$.
1771: According to \reffig{fig:pahU2}, an increase of the
1772: $\ipah{7.7}/\ipah{11.3}$ ratio by a factor $\simeq3$ would imply a
1773: decrease of the $\ipah{7.7}/\ipah{6.2}$ ratio by a factor of $\simeq2$.
1774:
1775: One of the strong results of \refS{sec:corglo} and \refS{sec:cormap}
1776: is that the ratios $\ipah{7.7}/\ipah{6.2}$ do not show any trend with
1777: $\ipah{7.7}/\ipah{11.3}$.
1778: Therefore, the correlations discussed in \refS{sec:correl} can not be attributed
1779: to a modification of the temperature distribution.
1780: In particular, we can rule out an interpretation of the variation of the
1781: $\ipah{7.7}/\ipah{11.3}$ in terms of destruction of the smallest PAHs.
1782: These results are in good agreement with those by \citet{hony01}, who
1783: studied a wide sample of Galactic sources and showed that the
1784: $\ipah{3.3}/\ipah{11.3}$ ratio was not varying significantly, while the
1785: $\ipah{6.2}/\ipah{11.3}$ ratio was varying by a factor of 5.
1786: In addition, \reffig{fig:pahU2} shows that the effect of extinction
1787: do not account for our observed trends.
1788: Indeed, the bottom panel of \reffig{fig:pahU2} shows that the $8.6\mic$ and
1789: $11.3\mic$ bands would be correlated if extinction was the main cause of
1790: variation of the aromatic bands.
1791: Consequently, our work shows that the variation of the PAH band ratios,
1792: throughout different environments, and at different spatial scales, is
1793: primarily controled by the ionization fraction of the molecules.
1794: In addition, it appears that the mixture of PAH molecules, in all these
1795: environments, is remarkably universal.
1796:
1797: We emphasize that our sample contains only a few AGNs.
1798: Therefore, our results reflect primarily the properties of star forming
1799: regions.
1800: Observations of AGN environments by \citet{smith07} and \citet{kaneda07}
1801: show PAH properties that differ from our trends, suggesting that the AGN
1802: could alter the PAH composition.
1803:
1804:
1805: % Calibration
1806: \subsection{Relating the Band Ratios to the Physical Conditions}
1807: \label{sec:calibration}
1808:
1809: In \refS{sec:pahU}, we showed that the band ratio variations were
1810: essentially due to a variation of the fraction of ionized PAHs in the beam.
1811: The mid-IR feature ratios are therefore related to the physical conditions
1812: where the emission is originating.
1813: Indeed, the fraction of ionized PAHs, in a given region, depends on the quantity
1814: $G_0/n_e\times\sqrt{T_\sms{gas}}$ \citep[][for a review]{tielens05}, where
1815: $G_0$, is the UV radiation field density, $n_e$,
1816: the electron density, and $T_\sms{gas}$, the gas temperature.
1817: The UV field density is the integration between $\lambda=0.09\mic$ and
1818: $\lambda=0.2\mic$ of the monochromatic mean intensity,
1819: $J_\lambda$, and is normalised to the solar neighborhood value:
1820: \begin{equation}
1821: G_0 = \frac{\displaystyle
1822: \int_{0.09\mic}^{0.2\mic} 4\pi\,J_\lambda\ddiff\lambda}
1823: {1.6\E{-6}\;\rm W\,m^{-2}}.
1824: \label{eq:G0}
1825: \end{equation}
1826: We emphasize that, although \S5.1 showed that the intensity of the ISRF
1827: was not responsible for the observed variations of the aromatic band ratios,
1828: the intensity of the ISRF is likely to vary significantly within and among our
1829: sources.
1830: In this section, we use the most well-studied Galactic regions of our sample, in order to derive an empirical relation between the $\ipah{6.2}/\ipah{11.3}$ band
1831: ratio and the quantity $G_0/n_e\times\sqrt{T_\sms{gas}}$.
1832:
1833: % G0/ne
1834: \begin{table}[htbp]
1835: \centering
1836: \begin{tabular}{lrrrr}
1837: \hline\hline
1838: Region & $G_0$ & $T_\sms{gas}$ [K]
1839: & $n_\sms{H}\;\rm [cm^{-3}]$ & Reference \\
1840: \hline
1841: \ngc{2023} & $1.5\E{4}$ & 750
1842: & $10^5$ & \citet{steiman-cameron97} \\
1843: \ngc{7027} & $6\E{5}$ & 2000
1844: & $10^7$ & \citet{justtanont97} \\
1845: \orb\ (position 4) & $4\E{4}$ & 500
1846: & $5\E{4}$ & \citet{tauber94} \\
1847: \hline
1848: \end{tabular}
1849: \caption{Physical conditions in select regions.
1850: The last columns refers to the study where these quantities have
1851: been derived.}
1852: \label{tab:g0ne}
1853: \end{table}
1854: \citet{bregman05} gave such a calibration, from the spectro-images of three
1855: reflection nebulae.
1856: In order to derive the physical conditions in their nebulae, they made
1857: several hypotheses.
1858: They\textlist{\thetextlist~assumed a constant electronic density throughout
1859: their regions,
1860: \thetextlist~they neglected dust extinction, and
1861: \thetextlist~they derived the relative geometry of the star and nebula
1862: from the profile of the scattered light.}In
1863: our analysis, we have adopted a slightly different strategy.
1864: We have made the conscious decision to include only the sources for which
1865: the physical conditions have been reliably determined, using PDR models,
1866: constrained by several gas lines.
1867: The sources of our sample which have been modeled in detail are
1868: \ngc{2023}, \ngc{7027} and position~4 of the \orb\ \citep{roche89}.
1869: Their properties are summarized in \reftab{tab:g0ne}.
1870:
1871: % VSGs
1872: \begin{figure*}[htbp]
1873: \centering
1874: \begin{tabular}{cc}
1875: \includegraphics[width=0.48\textwidth]{f18a} &
1876: \includegraphics[width=0.48\textwidth]{f18b} \\
1877: \end{tabular}
1878: \caption{{\bf Top panels:} variation of the color of the VSG continuum
1879: (ratio of the flux at 8 and $14\mic$) as a function of the intensity
1880: of the VSG continuum, inside the \orb.
1881: These figures show that the color of the VSG continuum is virtually
1882: independent of its intensity, since these grains are stochastically
1883: heated.
1884: {\bf Bottom panels:} variation of the PAH-to-VSG ratio
1885: as a function of the intensity.
1886: These figures show that the relative strength of the PAHs in the PDR
1887: decreases slightly toward the \hii\ region (when $\icont$ increases),
1888: as an effect of their photo-depletion.}
1889: \label{fig:colorVSG}
1890: \end{figure*}
1891: \begin{figure}[htbp]
1892: \centering
1893: \includegraphics[width=0.48\textwidth]{f19}
1894: \caption{Theoretical modeling of the continuum intensity as a function
1895: of the UV radiation field, $G_0$, in black.
1896: $L_\sms{cont}/M_\sms{dust}$ is the luminosity of the
1897: graphite component of \citet{zubko04}, integrated between 10 and
1898: $16\mic$, and normalized to the dust mass.
1899: The dotted line varies as $G_0$.
1900: The dashed-dotted line is a power-law fit to the black line,
1901: in the range $10^2<G_0<10^4$.}
1902: \label{fig:contU}
1903: \end{figure}
1904: % Actual calibration
1905: We adopt the values given in \reftab{tab:g0ne} and
1906: assume that the electrons result from photo-ionization of carbon and that all
1907: the gas phase carbon is ionized in PDRs, so that:
1908: \begin{equation}
1909: n_e \simeq \displaystyle\left(\frac{\rm C}{\rm H}\right) \times n_\sms{H}
1910: \simeq 1.6\E{-4}\, n_\sms{H},
1911: \end{equation}
1912: where we have adopted the carbon abundance by \citet{sofia04}.
1913: For \ngc{2023} and \ngc{7027}, we do not have reliable spatial information,
1914: hence we will only use the global values of the ratios.
1915: In the \orb, we study the spatial variations by proceeding as follows.
1916: \begin{enumerate}
1917: \item We exclude the region north of the bar (see \reffig{fig:imorionBar}),
1918: which is the location of the \hii\ region cavity.
1919: Indeed, the PAH-to-continuum ratio is very low there, thus the fits are
1920: uncertain.
1921: Moreover, the spherical geometry of the region causes confusion along the
1922: lines of sight.
1923: To the contrary, the region south of the bar, is an edge-on PDR, where
1924: the conditions are believed to be roughly homogeneous along each line
1925: of sight.
1926: Indeed, it is observed to be well stratified in the main PDR tracers
1927: (CO data as well as \hmol, \ci, \cii\ and PAH data;
1928: \citealt{tielens93,tauber94,wyrowski97}; and HCN
1929: data by \citealt{fuente96}).
1930: \item We assume that the column density is homogeneous in the edge-on PDR,
1931: so that we can use the intensity of the VSG continuum, in order to scale
1932: the intensity of the UV radiation field, $G_0$.
1933: Indeed, the top panels of \reffig{fig:colorVSG} show that the
1934: VSGs are stochastically heated.
1935: Therefore, their intensity scales with the radiation field density.
1936: Moreover, the intensity of the VSG continuum is probably a better tracer
1937: of the radiation density than the PAHs, since the latter are gradually
1938: depleted in direction of the \hii\ region (bottom panels of
1939: \reffig{fig:colorVSG}).
1940: \reffig{fig:contU} illustrates this fact by showing the theoretical
1941: variation of the integrated intensity of the VSG continuum between 10 and
1942: $16\mic$, as a function of $G_0$.
1943: To perform this simulation, we used the radiation field defined in
1944: \refeq{eq:G0}, and the Galactic graphite size distribution of
1945: \citet[with solar abundance constraints]{zubko04}.
1946: For $G_0\lesssim10^3$, the mid-IR continuum intensity scales
1947: perfectly with $G_0$, since the grains are out of thermal equilibrium.
1948: At higher $G_0$, the relation between $G_0$ and $I_\sms{cont}$
1949: becomes non-linear due to temperature effects.
1950: In the range $10^2\lesssim G_0\lesssim10^4$, which are the conditions in the
1951: \orb, the UV radiation field is very well approximated by
1952: $G_0\propto I_\sms{cont}^{1/1.3}$.
1953: This method has the advantage of taking into account the actual
1954: variations of $G_0$, with both the distance from the star cluster, and
1955: the dust extinction.
1956: \item In order to improve the signal-to-noise ratio of our spectra, we
1957: average the pixels into 30 bins of energy ($I_\sms{cont}$).
1958: The two top panels of \reffig{fig:g0ne} show the variation of the
1959: $\ipah{6.2}/\ipah{11.3}$ ratio, as a function of the continuum intensity,
1960: with the two methods.
1961: \item Finally, we assume a constant electron density and temperature
1962: throughout the PDR.
1963: \end{enumerate}
1964: The two bottom panels of \reffig{fig:g0ne} shows the final calibration of the
1965: band ratio, with the two methods.
1966: As found by \citet{bregman05}, the ratio $\ipah{6.2}/\ipah{11.3}$
1967: (or $\ipah{7.7}/\ipah{11.3}$) increases in the range
1968: $10^2\lesssim G_0/n_e\lesssim 10^3$.
1969: The grey stripes are the following linear fits:
1970: \begin{equation}
1971: \left\{
1972: \begin{array}{rcll}
1973: \displaystyle \frac{\ipah{6.2}}{\ipah{11.3}} & \simeq &
1974: \left(\displaystyle\frac{G_0}{n_e\;\rm[cm^{-3}]}
1975: \sqrt{\displaystyle\frac{T_\sms{gas}}{10^3\;\rm K}}\right)/3040
1976: + 0.53 \pm 0.10 & \mbox{ (\nlmet),} \\
1977: & & \\
1978: \displaystyle \frac{\ipah{6.2}}{\ipah{11.3}} & \simeq &
1979: \left(\displaystyle\frac{G_0}{n_e\;\rm[cm^{-3}]}
1980: \sqrt{\displaystyle\frac{T_\sms{gas}}{10^3\;\rm K}}\right)/1990
1981: + 0.26 \pm 0.16 & \mbox{ (\frmet),} \\
1982: \end{array} \\
1983: \right.
1984: \label{eq:calib}
1985: \end{equation}
1986: valid in the range $400\lesssim
1987: \displaystyle\frac{G_0}{n_e\;\rm[cm^{-3}]}
1988: \sqrt{\displaystyle\frac{T_\sms{gas}}{10^3\;\rm K}} \lesssim 4000$.
1989: % G0/ne
1990: \begin{figure*}[htbp]
1991: \centering
1992: \begin{tabular}{cc}
1993: \includegraphics[width=0.48\textwidth]{f20a} &
1994: \includegraphics[width=0.48\textwidth]{f20b} \\
1995: \includegraphics[width=0.48\textwidth]{f20c} &
1996: \includegraphics[width=0.48\textwidth]{f20d} \\
1997: \end{tabular}
1998: \caption{Empirical calibration of the band ratio.
1999: The two upper panels show the variation of the band ratio inside
2000: the \orb, as a function of the VSG intensity.
2001: The black error bars correspond to the edge-on PDR,
2002: and the grey error bars, to the \hii\ region cavity.
2003: The two lower panels show the final calibration, including
2004: \ngc{2023} and \ngc{7027}.
2005: We kept only the edge-on PDR measurements, in the \orb.}
2006: \label{fig:g0ne}
2007: \end{figure*}
2008:
2009:
2010: % ISM morphology
2011: \subsection{Diagnosing Galaxies Using PAH Band Ratios}
2012:
2013: In principle, \refeq{eq:calib} can be used to derive the ratio
2014: $G_0/n_e\times\sqrt{T_\sms{gas}}$ from any mid-IR spectrum.
2015: However, if several regions with different physical conditions are encompassed
2016: within the spectrograph's beam, the situation is more complex.
2017: The band ratio is then averaged over these various environments and its global
2018: value does not reflect a single $G_0/n_e\times\sqrt{T_\sms{gas}}$.
2019:
2020: \begin{figure*}[htbp]
2021: \centering
2022: \begin{tabular}{cc}
2023: \includegraphics[width=0.48\textwidth]{f21a} &
2024: \includegraphics[width=0.48\textwidth]{f21b} \\
2025: \end{tabular}
2026: \caption{Determination of $G_0/n_e\times\sqrt{T_\sms{gas}}$ for the
2027: sources in our sample.
2028: These figures are similar to \reffig{fig:corglo1}, where the
2029: $\ipah{6.2}/\ipah{11.3}$ has been converted into
2030: $G_0/n_e\times\sqrt{T_\sms{gas}}$ using \refeq{eq:calib}.
2031: The same symbol conventions are adopted as in \reffig{fig:corglo1}.
2032: The small additional symbols are the spatial variations within
2033: \M{82}.}
2034: \label{fig:final}
2035: \end{figure*}
2036: In the case of the global spectrum of a star forming region, we can
2037: expect the $\ipah{7.7}/\ipah{11.3}$ ratio to depend on the age and initial
2038: mass function of the stellar cluster, that will determine the $G_0$ at the
2039: edge of the PDR, as well as on the geometry of the ISM (clumpiness, densities),
2040: that will determine the transfer of the UV light into the PDR, and the electron
2041: density.
2042: If this hypothesis is correct, the value of $\ipah{7.7}/\ipah{11.3}$ can
2043: potentially be used in combination with other PDR tracers, like the
2044: \ciiline, \oiline, \siIIline\ and \hmol\ mid-IR rotational lines, to
2045: determine the geometry and physical conditions of an unresolved
2046: star forming region.
2047: \reffig{fig:final} demonstrates the relation between the observed band
2048: ratios and the averaged $G_0/n_e\times\sqrt{T_\sms{gas}}$ quantity.
2049:
2050: It has long been surmised that the IR emission features provide a clear
2051: mid-IR signature of the interaction of FUV photons with cloud surfaces and
2052: hence a probe of the importance of (massive) star formation in a region
2053: \citep{genzel98,peeters04}.
2054: The present study has extended this by developing the observed 6.2 to
2055: 11.3$\mic$ band ratio as a quantitative tool to probe the physical conditions
2056: ({e.g.} $G_0/n_e\times\sqrt{T_\sms{gas}}$) in the emitting regions.
2057: We expect that this study will be of fundamental value for the interpretation
2058: of \spitz\ data as well as future \sofia\ and \jwst\ observations of galaxies in
2059: the nearby and early universe.
2060:
2061:
2062: %===========================================================================
2063:
2064:
2065: % Conclusion
2066: %-----------
2067: \section{SUMMARY AND CONCLUSION}
2068: \label{sec:concl}
2069:
2070: We presented the results of a systematic study
2071: aimed at understanding the main properties of the mid-IR features in
2072: different astrophysical environments.
2073: It is based on observations of Galactic regions and galaxies of various types,
2074: observed by the satellites \iso\ and \spitz.
2075: We have developed two distinct methods of spectral decomposition with
2076: different hypotheses, in order to test the robustness of our trends, and
2077: overcome eventual biases.
2078: These two methods have shown similar trends between the physical quantities
2079: that we have studied.
2080: We explored the variations of the different mid-IR features between the
2081: wavelengths 5 and $16\mic$, among integrated spectra of galaxies, and
2082: inside galaxies and Galactic regions.
2083: Our main results are the followings.
2084: \begin{enumerate}
2085: \item We find that the 6.2, 7.7, $8.6\mic$ features are essentially tied
2086: together,
2087: while the ratio of these bands to the $11.3\mic$ feature can vary by one
2088: order of magnitude, in our sample.
2089: These variations are seen both inside individual sources (like \M{82},
2090: \M{17}, the \orb, \xxxdor, etc.), as well as among integrated spectra.
2091: In general, the $\ipah{6.2}/\ipah{11.3}$ ratio is spatially correlated
2092: with the power radiated by the PAHs.
2093: It indicates that the ratio $\ipah{6.2}/\ipah{11.3}$ is higher in regions
2094: of intense star formation.
2095: \item With the help of a stochastic heating model and realistic absorption
2096: efficiencies, we show that the variations of the mid-IR features are
2097: essentially due to the variation of the fraction of ionized PAHs.
2098: We conclude that the properties of the PAHs, throughout our sample,
2099: are remarkably universal.
2100: In particular, we rule out both the modification of the grain size
2101: distribution, and the extinction by the $9.7\mic$ silicate feature,
2102: as an explanation of these variations.
2103: Indeed, we show that a modification of the size distribution could explain
2104: the observed variation of the $\ipah{6.2}/\ipah{11.3}$ ratio.
2105: However, it would cause the $\ipah{6.2}/\ipah{7.7}$ and
2106: $\ipah{6.2}/\ipah{8.6}$ (and $\ipah{3.3}/\ipah{11.3}$) ratios to vary
2107: significantly and to be correlated with the $\ipah{6.2}/\ipah{11.3}$,
2108: contradicting our observations.
2109: Similarly, a deep absorption by the silicate feature would vary the
2110: $\ipah{6.2}/\ipah{11.3}$ ratio, but it would decouple the $8.6\mic$
2111: from the 6.2 and $7.7\mic$ features at the same time.
2112: \item The universality of the properties of the PAHs and the fact that the
2113: band ratios are mainly sensitive to the charge of the molecules allow us
2114: to use these band ratios as a tracer of the physical conditions inside the
2115: emitting region.
2116: Using a few well-studied Galactic regions (including the spectral image
2117: of the \orb), we give an empirical relation between $\ipah{6.2}/\ipah{11.3}$
2118: and the ratio $G_0/n_e\times\sqrt{T_\sms{gas}}$.
2119: \item In the case where several regions with different physical conditions
2120: are integrated, the band ratios are dependent on the morphology of the ISM,
2121: as well as on the evolutionary properties of the illuminating
2122: stellar cluster.
2123: Thus, we find that the $\ipah{6.2}/\ipah{11.3}$ band ratios in the
2124: star forming nuclei of \M{51} and \M{82} are very similar to those
2125: in the galactic compact \hii\ regions, \IR{15384} and \IR{18317},
2126: while the halo of \M{82} and the spiral arms of \M{51} are similar
2127: to those in the Galactic reflection nebula \ngc{2023}.
2128: These differences in band ratio reflect differences in the
2129: ionization over recombination rate and hence trace back to
2130: variations in the ratio of the ionizing radiation field to the
2131: electron density.
2132: \end{enumerate}
2133:
2134:
2135: %===========================================================================
2136:
2137:
2138: % Appendix
2139: %---------
2140: \appendix
2141: \section{ADDITIONAL TABLE AND FIGURES}
2142: \label{sec:tabfig}
2143:
2144: % Individual sources
2145: \subsubsection*{Notes on Individual Sources}
2146:
2147: \begin{description}
2148: \item[\M{82}.]
2149: (\reffigs{fig:cormap1m82}-\ref{fig:cormap2m82})
2150: \reffig{fig:cormap1m82} shows that the correlations inside \M{82} are
2151: remarkably well articulated.
2152: The correlation coefficients (\reftab{tab:correl}) are $0.97$, for the
2153: correlations between $\ipah{6.2}/\ipah{11.3}$ and $\ipah{7.7}/\ipah{11.3}$.
2154: The band ratios inside \M{82} follow the general trends
2155: observed for the integrated spectra very well.
2156: The centroid of the $7.7\mic$ feature is essentially fixed throughout \M{82}
2157: (\reftab{tab:correl}).
2158: The ratio $\ipah{6.2}/\ipah{11.3}$ follows the spatial distribution of the
2159: PAH emission very well, which is maximum in the central bar,
2160: where the star formation occurs.
2161: With good confidence ($\rm S/N>6$), we detect low ratios in the
2162: outer regions.
2163: \item[\IC{342}.]
2164: (\reffigs{fig:cormap1ic342}-\ref{fig:cormap2ic342})
2165: The morphology of the PAH emission (\reffig{fig:imic342}) follows the
2166: \COio\ emission which traces
2167: out the nuclear ring and ridge associated with the stellar bar
2168: \citet{sakamoto99}.
2169: The spectra of this galaxy are much noisier than those of \M{82}.
2170: The correlations are very well defined, with a linear correlation
2171: coefficient of $\simeq0.8$ (\reftab{tab:correl}).
2172: The deviation of the relation involving the $8.6\mic$ feature with
2173: the \frmet\ (middle right panel of \reffig{fig:cormap1ic342}), compared
2174: to the general correlation, is due to a systematic over-estimation of the
2175: attenuation, in the central region.
2176: The average centroid of the $7.7\mic$ feature is identical for \IC{342} and
2177: \M{82} ($\langle\lambda_{7.7}\rangle\simeq 7.68\mic$;
2178: \reftab{tab:correl}).
2179: \item[\M{51}.]
2180: (\reffigs{fig:cormap1m51}-\ref{fig:cormap2m51})
2181: This galaxy is more quiescent than \M{82} and \IC{342}.
2182: Its band ratios show less variation (\reffig{fig:cormap1m51})
2183: throughout the galaxy, although they are in good agreement with the
2184: general correlations of \reffig{fig:corglo1}.
2185: The correlation coefficients are $\simeq0.6$ (\reftab{tab:correl})
2186: slightly lower than those for \M{82} and \IC{342}, essentially because of
2187: the lower signal-to-noise ratio of the spectra.
2188: The spatial distribution of the $\ipah{6.2}/\ipah{11.3}$ ratio
2189: (\reffig{fig:imm51}) follows the PAH emission remarkably well enhanced
2190: along the spiral arms and the bulge.
2191: This band ratio reaches its maximum on the central ring and decreases
2192: somewhat in the center.
2193: \item[\M{83}.]
2194: (\reffigs{fig:cormap1m83}-\ref{fig:cormap2m83})
2195: \M{83} is a spiral galaxy which is similar in several aspects to \M{51}.
2196: Its band ratios (\reffig{fig:cormap1m83}) and the spatial distribution
2197: of its PAH emission (\reffig{fig:imm83}) have the same
2198: properties qualitatively.
2199: The average centroid of the $7.7\mic$ is identical to that for \M{51} and
2200: \M{83} ($\langle\lambda_{7.7}\rangle\simeq 7.72\mic$;
2201: \reftab{tab:correl}).
2202: \item[\xxxdor.]
2203: (\reffigs{fig:cormap130dor}-\ref{fig:cormap230dor})
2204: The band ratio correlations measured inside \xxxdor\ are significantly
2205: dispersed (\reffig{fig:cormap130dor}).
2206: The results of the 2 different feature extraction methods are not
2207: particularly consistent.
2208: This is certainly due to the very low PAH-to-continuum ratio.
2209: Indeed, \reffig{fig:im30dor} shows that the spatial distribution of the
2210: PAH emission, obtained with the \frmet, is different from that obtained
2211: with the \nlmet.
2212: In particular, the right panel of \reffig{fig:im30dor} shows 3 bright
2213: spots along the bar of \xxxdor, that are not seen on the left panel.
2214: After verification, these are fits of such low PAH-to-continuum ratio
2215: pixels that the widths of the PAH bands are not constrained at all.
2216: The \frmet\ determines very wide bands whose integrated intensities
2217: are significantly overestimated.
2218: For the same reason, the highest values of the $\ipah{6.2}/\ipah{11.3}$
2219: ratio, with the \nlmet\ ($\simeq2$; \reffig{fig:im30dor}),
2220: correspond to low values, with the \frmet\ ($\lesssim0.3$).
2221: \item[\M{17}.]
2222: (\reffigs{fig:cormap1m17}-\ref{fig:cormap2m17})
2223: We encounter the same trouble in the \hii\ region part of \M{17}
2224: (upper left corner of the images in \reffig{fig:imm17}), as in
2225: \xxxdor.
2226: For the same reasons, the $\ipah{6.2}/\ipah{11.3}$ ratio is maximum in
2227: this region
2228: with the \nlmet\ ($\simeq2.5$; \reffig{fig:imm17}), while it is minimum
2229: with the \frmet\ ($\lesssim0.7$).
2230: However, the PDR part has relatively high PAH-to-continuum ratios.
2231: The band ratio correlations are well defined inside \M{17}, except for
2232: $8.6\mic$, with the \nlmet\ (\reffig{fig:cormap1m17}).
2233: \item[The \orb.]
2234: (\reffigs{fig:cormap1orionBar}-\ref{fig:cormap2orionBar})
2235: The PAH-to-continuum ratio is on average higher in this region than in
2236: \M{17}.
2237: Thus the fits are better, even along the ionization front.
2238: The correlation between band ratios (\reffig{fig:cormap1orionBar}) are
2239: relatively good with linear correlation coefficients around 0.9
2240: (\reftab{tab:correl}).
2241: The plot of the consistency between the two methods
2242: (\reffig{fig:compmaporionBar}) shows a few outlying points.
2243: These fits are those of the pixels located inside the \hii\ region
2244: (upper part of the images in \reffig{fig:imorionBar}).
2245: The spatial distribution of the $\ipah{6.2}/\ipah{11.3}$ band ratio
2246: decreases on average from the north, where the ionizing stars are located,
2247: to the south with the two methods.
2248: The average centroid of the $7.7\mic$ feature is similar in \M{17}
2249: and the \orb\ ($\langle\lambda_{7.7}\rangle\simeq 7.68\mic$;
2250: \reftab{tab:correl}).
2251: \end{description}
2252:
2253: % Line intensities
2254: %-----------------
2255: \begin{deluxetable}{llrrrrrrrr}
2256: % Head
2257: %-----
2258: \tabletypesize{\tiny}
2259: \rotate
2260: \tablecolumns{10}
2261: \tablewidth{0pc}
2262: \tablecaption{Results of the spectral decomposition.}
2263: \tablehead{
2264: \colhead{Name} & \colhead{} & \colhead{$\ipah{6.2}$}
2265: & \colhead{$\ipah{7.7}$}
2266: & \colhead{$\ipah{8.6}$} & \colhead{$\ipah{11.3}$}
2267: & \colhead{$\rm I_{PAH}$} & \colhead{$\rm I_{cont}$}
2268: & \colhead{$\lambda_{7.7}$} & \colhead{Relia-} \\
2269: \colhead{} & \colhead{} & \colhead{[$10^{-15}\;\rm W\,m^{-2}$]}
2270: & \colhead{[$10^{-15}\;\rm W\,m^{-2}$]}
2271: & \colhead{[$10^{-15}\;\rm W\,m^{-2}$]}
2272: & \colhead{[$10^{-15}\;\rm W\,m^{-2}$]}
2273: & \colhead{[$10^{-15}\;\rm W\,m^{-2}$]}
2274: & \colhead{[$10^{-15}\;\rm W\,m^{-2}$]}
2275: & \colhead{[$\mu m$]} & \colhead{bility} \\}
2276: % Data
2277: %-----
2278: \startdata
2279: % Haro 11
2280: \haro{11} & $\mathcal{S}$ & $0.90\pm0.09$
2281: & $1.81\pm0.18$ & $0.23\pm0.02$ & $3.78\pm2.15$
2282: & $3.84\pm2.15$ & $51.1\pm5.1$
2283: & \nodata & Low \\
2284: & $\mathcal{L}$ & $1.59\pm0.16$
2285: & $5.23\pm0.52$ & $0.32\pm0.03$ & $1.28\pm0.13$
2286: & $11.2\pm0.6$ & $59.8\pm6.0$
2287: & $7.634\pm0.025$ & Low \\
2288: % SMC B1
2289: \smcb & $\mathcal{S}$ & $16.2\pm2.4$
2290: & $24.1\pm3.6$ & $3.09\pm0.46$ & $16.0\pm2.4$
2291: & $73.8\pm35.7$ & $76.8\pm11.5$
2292: & \nodata & High \\
2293: & $\mathcal{L}$ & $33.8\pm5.1$
2294: & $70.2\pm1.1$ & $16.1\pm2.4$ & $51.0\pm7.6$
2295: & $205\pm14$ & $15.1\pm2.3$
2296: & $7.647\pm0.025$ & High \\
2297: % NGC 253
2298: \ngc{253} & $\mathcal{S}$ & $148\pm4$
2299: & $293\pm11$ & $28.5\pm3.7$ & $105\pm4$
2300: & $604\pm346$ & $(1.74\pm0.09)\E{3}$
2301: & \nodata & High \\
2302: & $\mathcal{L}$ & $532\pm28$
2303: & $(1.54\pm0.06)\E{3}$ & $36.4\pm1.7$ & $313\pm13$
2304: & $(3.42\pm0.08)\E{3}$ & $(1.49\pm0.07)\E{3}$
2305: & $7.707\pm0.025$ & High \\
2306: \ngc{253}~p & $\mathcal{S}$ & $48.5\pm2.2$
2307: & $98.0\pm6.3$ & $10.2\pm0.2$ & $29.1\pm1.9$
2308: & $195\pm114$ & $823\pm44$
2309: & \nodata & Low \\
2310: & $\mathcal{L}$ & $15.1\pm8.3$
2311: & $481\pm22$ & $113\pm6$ & $114\pm6$
2312: & $(1.09\pm0.03)\E{3}$ & $(1.00\pm0.05)\E{3}$
2313: & $7.702\pm0.025$ & High \\
2314: \ngc{253}~e & $\mathcal{S}$ & $102\pm3$
2315: & $199\pm8$ & $19.1\pm0.6$ & $78.6\pm2.6$
2316: & $420\pm238$ & $974\pm66$
2317: & \nodata & High \\
2318: & $\mathcal{L}$ & $324\pm22$
2319: & $(1.00\pm0.05)\E{3}$ & $228\pm13$ & $225\pm11$
2320: & $(2.21\pm0.06)\E{3}$ & $689\pm40$
2321: & $7.711\pm0.025$ & High \\
2322: % SMC N66
2323: \smcn & $\mathcal{S}$ & $10.3\pm0.5$
2324: & $14.0\pm0.6$ & $0.34\pm0.05$ & $6.09\pm0.11$
2325: & $37.0\pm19.1$ & $290\pm41$
2326: & \nodata & Low \\
2327: & $\mathcal{L}$ & $82.7\pm20.6$
2328: & $99.7\pm19.1$ & $57.5\pm12.9$ & $64.5\pm8.0$
2329: & $340\pm34$ & $203\pm28$
2330: & $7.619\pm0.025$ & Low \\
2331: % NGC 520
2332: \ngc{520} & $\mathcal{S}$ & $8.38\pm0.37$
2333: & $19.6\pm0.7$ & $1.00\pm0.05$ & $4.50\pm0.10$
2334: & $36.6\pm22.0$ & $45.4\pm4.3$
2335: & \nodata & High \\
2336: & $\mathcal{L}$ & $38.9\pm2.9$
2337: & $107\pm5$ & $29.2\pm1.9$ & $26.3\pm1.7$
2338: & $240\pm7$ & $72.3\pm5.2$
2339: & $7.748\pm0.025$ & High \\
2340: % NGC 613
2341: \ngc{613} & $\mathcal{S}$ & $7.92\pm0.26$
2342: & $13.5\pm1.0$ & $1.49\pm0.01$ & $4.89\pm0.05$
2343: & $30.2\pm16.5$ & $74.0\pm31.9$
2344: & \nodata & Low \\
2345: & $\mathcal{L}$ & $19.3\pm9.2$
2346: & $40.4\pm10.4$ & $7.41\pm3.16$ & $15.9\pm3.6$
2347: & $112\pm18$ & $55.4\pm26.4$
2348: & $7.693\pm0.025$ & Low \\
2349: \ngc{613}~p & $\mathcal{S}$ & $4.03\pm0.12$
2350: & $8.35\pm0.48$ & $0.72\pm0.05$ & $3.88\pm0.20$
2351: & $18.3\pm10.1$ & $35.9\pm10.0$
2352: & \nodata & High \\
2353: & $\mathcal{L}$ & $12.4\pm3.5$
2354: & $32.4\pm5.2$ & $6.94\pm1.67$ & $12.7\pm1.6$
2355: & $79.8\pm7.5$ & $33.9\pm10.8$
2356: & $7.703\pm0.025$ & High \\
2357: % NGC 891
2358: \ngc{891} & $\mathcal{S}$ & $93.1\pm9.3$
2359: & $181\pm18$ & $17.0\pm1.7$ & $55.0\pm5.5$
2360: & $361\pm212$ & $304\pm30$
2361: & \nodata & High \\
2362: & $\mathcal{L}$ & $207\pm21$
2363: & $605\pm61$ & $124\pm12$ & $148\pm15$
2364: & $(1.34\pm0.07)\E{3}$ & $87.1\pm8.7$
2365: & $7.705\pm0.025$ & High \\
2366: % NGC 1068
2367: \ngc{1068} & $\mathcal{S}$ & $55.5\pm5.9$
2368: & $126\pm35$ & $20.1\pm6.1$ & $59.3\pm33.6$
2369: & $287\pm152$ & $(4.80\pm0.37)\E{3}$
2370: & \nodata & Low \\
2371: & $\mathcal{L}$ & $181\pm10$
2372: & $586\pm85$ & $131\pm16$ & $216\pm11$
2373: & $(1.33\pm0.09)\E{3}$ & $(4.75\pm0.23)\E{3}$
2374: & $7.706\pm0.025$ & Low \\
2375: \ngc{1068}~p & $\mathcal{S}$ & $\lesssim10.2$
2376: & $\lesssim45$ & $5.1\pm4.8$ & $\lesssim31$
2377: & $573\pm274$ & $(3.59\pm0.34)\E{3}$
2378: & \nodata & Low \\
2379: & $\mathcal{L}$ & $209\pm15$
2380: & $352\pm80$ & $86.0\pm15.0$ & $97.3\pm6.0$
2381: & $902\pm83$ & $(3.52\pm0.19)\E{3}$
2382: & $7.699\pm0.025$ & Low \\
2383: \ngc{1068}~e & $\mathcal{S}$ & $48.9\pm7.1$
2384: & $101\pm35$ & $14.3\pm6.0$ & $51.1\pm36.4$
2385: & $237\pm125$ & $(1.01\pm0.30)\E{3}$
2386: & \nodata & Low \\
2387: & $\mathcal{L}$ & $313\pm43$
2388: & $537\pm202$ & $108\pm40$ & $115\pm13$
2389: & $(1.28\pm0.21)\E{3}$ & $859\pm136$
2390: & $7.699\pm0.025$ & High \\
2391: % NGC 1097
2392: \ngc{1097} & $\mathcal{S}$ & $27.7\pm0.1$
2393: & $50.6\pm0.01$ & $5.25\pm0.15$ & $27.0\pm0.8$
2394: & $116\pm64$ & $229\pm32$
2395: & \nodata & High \\
2396: & $\mathcal{L}$ & $56.4\pm6.9$
2397: & $166\pm12$ & $31.0\pm3.1$ & $61.7\pm4.7$
2398: & $410\pm18$ & $127\pm18$
2399: & $7.707\pm0.025$ & High \\
2400: \ngc{1097}~p & $\mathcal{S}$ & $18.7\pm0.15$
2401: & $37.7\pm1.2$ & $4.25\pm0.32$ & $19.1\pm1.0$
2402: & $82.8\pm46.4$ & $156\pm13.8$
2403: & \nodata & High \\
2404: & $\mathcal{L}$ & $42.8\pm3.2$
2405: & $134\pm7$ & $29.5\pm2.0$ & $48.4\pm3.1$
2406: & $319\pm9$ & $88.9\pm8.0$
2407: & $7.703\pm0.025$ & High \\
2408: % NGC 1140
2409: \ngc{1140} & $\mathcal{S}$ & $0.92\pm0.05$
2410: & $1.31\pm0.01$ & $0.07\pm0.02$ & $0.62\pm0.01$
2411: & $3.34\pm1.73$ & $6.96\pm2.16$
2412: & \nodata & Low \\
2413: & $\mathcal{L}$ & $1.63\pm0.66$
2414: & $3.65\pm0.81$ & $0.45\pm0.13$ & $1.45\pm0.17$
2415: & $8.29\pm1.17$ & $5.57\pm1.71$
2416: & $7.684\pm0.025$ & Low \\
2417: % NGC 1365
2418: \ngc{1365} & $\mathcal{S}$ & $38.0\pm0.5$
2419: & $76.5\pm2.4$ & $9.21\pm0.27$ & $34.9\pm1.3$
2420: & $177\pm94$ & $367\pm31$
2421: & \nodata & High \\
2422: & $\mathcal{L}$ & $89.3\pm7.1$
2423: & $253\pm13$ & $52.7\pm3.8$ & $79.5\pm4.4$
2424: & $566\pm16$ & $242\pm20$
2425: & $7.710\pm0.025$ & High \\
2426: \ngc{1365}~p & $\mathcal{S}$ & $26.7\pm1.1$
2427: & $54.1\pm1.9$ & $6.22\pm0.21$ & $22.3\pm1.2$
2428: & $117\pm65$ & $253\pm15$
2429: & \nodata & High \\
2430: & $\mathcal{L}$ & $57.7\pm3.1$
2431: & $182\pm8$ & $42.3\pm2.2$ & $53.2\pm2.5$
2432: & $148\pm10$ & $71.1\pm16.2$
2433: & $7.713\pm0.025$ & High \\
2434: \ngc{1365}~e & $\mathcal{S}$ & $11.4\pm0.02$
2435: & $22.4\pm1.0$ & $2.98\pm0.10$ & $12.6\pm0.7$
2436: & $60.5\pm29.6$ & $113\pm24$
2437: & \nodata & High \\
2438: & $\mathcal{L}$ & $18.3\pm4.9$
2439: & $61.2\pm7.6$ & $11.7\pm2.4$ & $30.8\pm3.1$
2440: & $148\pm10$ & $71.1\pm16.2$
2441: & $7.708\pm0.025$ & High \\
2442: % IC 342
2443: \IC{342} & $\mathcal{S}$ & $36.8\pm0.7$
2444: & $62.1\pm1.6$ & $9.69\pm0.27$ & $26.7\pm0.8$
2445: & $141\pm78$ & $270\pm15$
2446: & \nodata & High \\
2447: & $\mathcal{L}$ & $62.2\pm3.9$
2448: & $182\pm8$ & $48.9\pm2.6$ & $60.7\pm2.6$
2449: & $441\pm10$ & $158\pm8$
2450: & $7.686\pm0.025$ & High \\
2451: \IC{342}~p & $\mathcal{S}$ & $16.5\pm0.6$
2452: & $28.3\pm1.1$ & $3.90\pm0.16$ & $11.3\pm0.5$
2453: & $62.9\pm34.9$ & $135\pm6$
2454: & \nodata & High \\
2455: & $\mathcal{L}$ & $30.4\pm1.6$
2456: & $78.0\pm3.2$ & $15.2\pm0.8$ & $26.9\pm1.2$
2457: & $188\pm4$ & $87.5\pm3.9$
2458: & $7.678\pm0.025$ & High \\
2459: \IC{342}~e & $\mathcal{S}$ & $20.2\pm0.3$
2460: & $33.8\pm0.7$ & $5.79\pm0.14$ & $15.4\pm0.4$
2461: & $78.4\pm42.7$ & $135\pm12$
2462: & \nodata & High \\
2463: & $\mathcal{L}$ & $32.8\pm3.3$
2464: & $105\pm6$ & $34.4\pm2.5$ & $33.5\pm1.8$
2465: & $257\pm8$ & $70.1\pm5.4$
2466: & $7.692\pm0.025$ & High \\
2467: % NGC 1569
2468: \ngc{1569} & $\mathcal{S}$ & $6.82\pm0.03$
2469: & $13.2\pm0.1$ & $0.50\pm0.08$ & $5.38\pm0.12$
2470: & $28.1\pm15.8$ & $114\pm24$
2471: & \nodata & Low \\
2472: & $\mathcal{L}$ & $20.2\pm9.8$
2473: & $35.7\pm10.3$ & $3.21\pm1.35$ & $13.9\pm2.2$
2474: & $102\pm29$ & $91.4\pm14.9$
2475: & $7.669\pm0.025$ & Low \\
2476: \ngc{1569}~e & $\mathcal{S}$ & $5.43\pm0.03$
2477: & $10.7\pm0.1$ & $0.33\pm0.08$ & $4.02\pm0.10$
2478: & $22.5\pm12.7$ & $55.8\pm23.9$
2479: & \nodata & Low \\
2480: & $\mathcal{L}$ & $13.3\pm9.6$
2481: & $29.5\pm12.3$ & $3.06\pm1.98$ & $10.3\pm2.5$
2482: & $69.7\pm38.2$ & $46.2\pm15.2$
2483: & $7.679\pm0.025$ & Low \\
2484: % NGC 1808
2485: \ngc{1808} & $\mathcal{S}$ & $86.8\pm8.7$
2486: & $147\pm15$ & $15.0\pm1.5$ & $59.1\pm5.9$
2487: & $321\pm181$ & $497\pm50$
2488: & \nodata & High \\
2489: & $\mathcal{L}$ & $180\pm18$
2490: & $487\pm49$ & $102\pm10$ & $185\pm19$
2491: & $(1.24\pm0.06)\E{3}$ & $274\pm27$
2492: & $7.710\pm0.025$ & High \\
2493: % Orion 3
2494: \orb~D8 & $\mathcal{S}$ & $338\pm3$
2495: & $632\pm7$ & $77.8\pm4.4$ & $301\pm12$
2496: & $(1.45\pm0.78)\E{3}$ & $(9.80\pm0.11)\E{3}$
2497: & \nodata & Low \\
2498: & $\mathcal{L}$ & $590\pm13$
2499: & $912\pm5$ & $73.8\pm0.8$ & $501\pm12$
2500: & $(2.79\pm0.03)\E{3}$ & $(9.25\pm0.08)\E{3}$
2501: & $7.645\pm0.025$ & Low \\
2502: % Orion 2
2503: \orb~D5 & $\mathcal{S}$ & $532\pm3$
2504: & $(1.05\pm0.01)\E{3}$ & $131\pm2$ & $470\pm3$
2505: & $(2.33\pm1.28)\E{3}$ & $(5.84\pm0.13)\E{3}$
2506: & \nodata & High \\
2507: & $\mathcal{L}$ & $981\pm18$
2508: & $(2.49\pm0.02)\E{3}$ & $580\pm7$ & $737\pm7$
2509: & $(5.97\pm0.04)\E{3}$ & $(4.94\pm0.07)\E{3}$
2510: & $7.670\pm0.025$ & High \\
2511: % Orion 1
2512: \orb~D2 & $\mathcal{S}$ & $107\pm10$
2513: & $186\pm1$ & $21.0\pm0.1$ & $102\pm8$
2514: & $468\pm240$ & $(2.67\pm0.12)\E{3}$
2515: & \nodata & Low \\
2516: & $\mathcal{L}$ & $233\pm10$
2517: & $491\pm8$ & $70.1\pm1.6$ & $182\pm5$
2518: & $(1.58\pm0.03)\E{3}$ & $(2.18\pm0.10)\E{3}$
2519: & $7.680\pm0.025$ & High \\
2520: % 30 Dor
2521: \xxxdor & $\mathcal{S}$ & $205\pm1$
2522: & $405\pm8$ & $12.4\pm0.3$ & $98.2\pm3.5$
2523: & $78.4\pm46.6$ & $(1.39\pm0.06)\E{4}$
2524: & \nodata & Low \\
2525: & $\mathcal{L}$ & $(1.12\pm0.07)\E{3}$
2526: & $(1.68\pm0.07)\E{3}$ & $781\pm36$ & $(1.04\pm0.04)\E{3}$
2527: & $(5.73\pm0.13)\E{3}$ & $(1.23\pm0.05)\E{4}$
2528: & $7.616\pm0.025$ & Low \\
2529: \xxxdor~p & $\mathcal{S}$ & $25.3\pm1.2$
2530: & $56.0\pm0.1$ & $0.73\pm0.16$ & $11.0\pm0.4$
2531: & $99.7\pm62.6$ & $(2.84\pm0.11)\E{3}$
2532: & \nodata & Low \\
2533: & $\mathcal{L}$ & $184\pm23$
2534: & $264\pm18$ & $122\pm8$ & $134\pm6$
2535: & $873\pm36$ & $(2.61\pm0.10)\E{3}$
2536: & $7.711\pm0.025$ & Low \\
2537: \xxxdor~e & $\mathcal{S}$ & $164\pm1$
2538: & $315\pm4$ & $11.7\pm0.2$ & $81.2\pm2.9$
2539: & $624\pm366$ & $(9.62\pm0.40)\E{3}$
2540: & \nodata & Low \\
2541: & $\mathcal{L}$ & $839\pm61$
2542: & $(1.34\pm0.07)\E{3}$ & $479\pm26$ & $798\pm33$
2543: & $(4.30\pm0.12)\E{3}$ & $(8.49\pm0.35)\E{3}$
2544: & $7.624\pm0.025$ & High \\
2545: % NGC 2023
2546: \ngc{2023} & $\mathcal{S}$ & $54.1\pm2.3$
2547: & $125\pm5$ & $8.25\pm0.11$ & $58.2\pm0.5$
2548: & $263\pm149$ & $263\pm55$
2549: & \nodata & High \\
2550: & $\mathcal{L}$ & $120\pm9$
2551: & $284\pm8$ & $60.9\pm3.7$ & $106\pm7$
2552: & $705\pm19$ & $99.7\pm14$
2553: & $7.743\pm0.025$ & High \\
2554: % He 2-10
2555: \hen & $\mathcal{S}$ & $6.74\pm0.67$
2556: & $14.7\pm1.5$ & $2.01\pm0.20$ & $5.90\pm0.59$
2557: & $31.1\pm17.4$ & $109\pm11$
2558: & \nodata & Low \\
2559: & $\mathcal{L}$ & $14.4\pm1.4$
2560: & $36.1\pm3.6$ & $5.50\pm0.55$ & $14.2\pm1.4$
2561: & $85.9\pm4.3$ & $125\pm13$
2562: & $7.662\pm0.025$ & High \\
2563: % M 82
2564: \M{82} & $\mathcal{S}$ & $992\pm36$
2565: & $(2.01\pm0.07)\E{3}$ & $231\pm8$ & $602\pm22$
2566: & $(4.09\pm2.34\E{3}$ & $(6.74\pm0.29)\E{3}$
2567: & \nodata & High \\
2568: & $\mathcal{L}$ & $(2.68\pm0.10)\E{3}$
2569: & $(7.62\pm0.27)\E{3}$ & $(1.80\pm0.07)\E{3}$ & $(2.23\pm0.09)\E{3}$
2570: & $(1.72\pm0.03)\E{4}$ & $(5.82\pm0.24)\E{3}$
2571: & $7.697\pm0.025$ & High \\
2572: \M{82}~p & $\mathcal{S}$ & $369\pm18$
2573: & $790\pm33$ & $88.4\pm3.5$ & $195\pm11$
2574: & $(1.53\pm0.90\E{3}$ & $(3.16\pm0.12)\E{3}$
2575: & \nodata & High \\
2576: & $\mathcal{L}$ & $983\pm42$
2577: & $(2.92\pm0.11)\E{3}$ & $697\pm26$ & $701\pm27$
2578: & $(6.35\pm0.12)\E{3}$ & $(3.02\pm0.11)\E{3}$
2579: & $7.695\pm0.025$ & High \\
2580: \M{82}~e & $\mathcal{S}$ & $819\pm30$
2581: & $(1.64\pm0.06)\E{3}$ & $192\pm7$ & $527\pm20$
2582: & $(3.39\pm1.93\E{3}$ & $(5.22\pm0.24)\E{3}$
2583: & \nodata & High \\
2584: & $\mathcal{L}$ & $(1.99\pm0.08)\E{3}$
2585: & $(5.71\pm0.21)\E{3}$ & $(1.21\pm0.04)\E{3}$ & $(1.26\pm0.05)\E{3}$
2586: & $(1.21\pm0.02)\E{4}$ & $(2.82\pm0.13)\E{3}$
2587: & $7.701\pm0.025$ & High \\
2588: % NGC 3256
2589: \ngc{3256} & $\mathcal{S}$ & $24.7\pm0.3$
2590: & $45.5\pm4.2$ & $5.29\pm0.02$ & $19.4\pm0.2$
2591: & $97.4\pm55.6$ & $193\pm27$
2592: & \nodata & High \\
2593: & $\mathcal{L}$ & $65.0\pm8.9$
2594: & $183\pm14$ & $48.6\pm5.9$ & $57.9\pm4.5$
2595: & $411\pm19$ & $181\pm20$
2596: & $7.701\pm0.025$ & High \\
2597: \ngc{3256}~p & $\mathcal{S}$ & $6.16\pm0.40$
2598: & $11.0\pm0.3$ & $1.05\pm0.02$ & $4.37\pm0.25$
2599: & $23.1\pm13.4$ & $52.1\pm10.2$
2600: & \nodata & High \\
2601: & $\mathcal{L}$ & $18.3\pm3.9$
2602: & $45.4\pm5.8$ & $9.88\pm2.31$ & $13.6\pm1.7$
2603: & $105\pm10$ & $51.7\pm7.3$
2604: & $7.703\pm0.025$ & High \\
2605: \ngc{3256}~e & $\mathcal{S}$ & $14.5\pm1.5$
2606: & $24.5\pm0.1$ & $3.19\pm0.01$ & $12.3\pm0.1$
2607: & $55.7\pm32.1$ & $101\pm26$
2608: & \nodata & High \\
2609: & $\mathcal{L}$ & $38.0\pm9.3$
2610: & $104\pm13$ & $29.2\pm6.0$ & $37.1\pm4.6$
2611: & $245\pm19$ & $89.8\pm18.8$
2612: & $7.699\pm0.025$ & High \\
2613: % Mrk 33
2614: \mrk{33} & $\mathcal{S}$ & $2.51\pm0.25$
2615: & $5.12\pm0.51$ & $0.31\pm0.03$ & $3.14\pm0.31$
2616: & $11.6\pm6.5$ & $27.8\pm2.8$
2617: & \nodata & High \\
2618: & $\mathcal{L}$ & $3.88\pm0.39$
2619: & $14.2\pm0.14$ & $2.35\pm0.24$ & $4.94\pm0.49$
2620: & $30.4\pm1.6$ & $25.4\pm2.5$
2621: & $7.721\pm0.025$ & High \\
2622: % Arp 299
2623: \arp{299} & $\mathcal{S}$ & $9.42\pm0.72$
2624: & $16.8\pm1.0$ & $1.93\pm0.01$ & $6.10\pm0.20$
2625: & $37.2\pm20.4$ & $223\pm27$
2626: & \nodata & Low \\
2627: & $\mathcal{L}$ & $28.0\pm6.6$
2628: & $43.7\pm5.7$ & $4.99\pm0.82$ & $13.2\pm1.6$
2629: & $147\pm18$ & $413\pm36$
2630: & $7.669\pm0.025$ & Low \\
2631: % UM 448
2632: \um{448} & $\mathcal{S}$ & $0.83\pm0.08$
2633: & $1.39\pm0.14$ & $0.27\pm0.03$ & $0.83\pm0.08$
2634: & $3.59\pm1.85$ & $11.4\pm1.1$
2635: & \nodata & High \\
2636: & $\mathcal{L}$ & $1.34\pm0.13$
2637: & $4.92\pm0.49$ & $0.90\pm0.09$ & $1.52\pm0.15$
2638: & $10.0\pm0.6$ & $9.95\pm1.0$
2639: & $7.670\pm0.025$ & High \\
2640: % IR 12331
2641: \IR{12331} & $\mathcal{S}$ & $38.0\pm4.2$
2642: & $69.4\pm1.9$ & $7.11\pm1.20$ & $23.2\pm0.7$
2643: & $160\pm84$ & $869\pm62$
2644: & \nodata & Low \\
2645: & $\mathcal{L}$ & $115\pm7$
2646: & $247\pm6$ & $34.5\pm3.2$ & $38.5\pm4.1$
2647: & $624\pm20$ & $765\pm28$
2648: & $7.696\pm0.025$ & High \\
2649: % NGC{4945}
2650: \ngc{4945} & $\mathcal{S}$ & $65.7\pm1.3$
2651: & $182\pm7$ & $11.6\pm0.3$ & $40.7\pm0.3$
2652: & $310\pm198$ & $393\pm35$
2653: & \nodata & High \\
2654: & $\mathcal{L}$ & $260\pm16$
2655: & $(1.15\pm0.05)\E{3}$ & $186\pm11$ & $222\pm12$
2656: & $(2.15\pm0.05)\E{3}$ & $389\pm29$
2657: & $7.713\pm0.025$ & High \\
2658: \ngc{4945}~p & $\mathcal{S}$ & $29.6\pm0.1$
2659: & $94.2\pm3.9$ & $3.72\pm0.03$ & $10.0\pm0.1$
2660: & $149\pm100$ & $156\pm12$
2661: & \nodata & High \\
2662: & $\mathcal{L}$ & $322\pm15$
2663: & $(1.15\pm0.05)\E{3}$ & $195\pm11$ & $208\pm13$
2664: & $(2.22\pm0.05)\E{3}$ & $370\pm23$
2665: & $7.709\pm0.025$ & High \\
2666: \ngc{4945}~e & $\mathcal{S}$ & $36.5\pm0.7$
2667: & $87.7\pm3.6$ & $7.89\pm0.19$ & $30.7\pm0.1$
2668: & $164\pm100$ & $237\pm31$
2669: & \nodata & High \\
2670: & $\mathcal{L}$ & $90.1\pm9.6$
2671: & $413\pm23$ & $87.5\pm7.7$ & $99.1\pm6.4$
2672: & $826\pm28$ & $136\pm16$
2673: & $7.713\pm0.025$ & High \\
2674: % Cen a
2675: \cenA & $\mathcal{S}$ & $17.0\pm0.2$
2676: & $26.0\pm0.1$ & $2.66\pm0.14$ & $13.6\pm0.2$
2677: & $61.1\pm34.0$ & $219\pm33$
2678: & \nodata & Low \\
2679: & $\mathcal{L}$ & $30.2\pm5.8$
2680: & $112\pm13$ & $27.2\pm1.3$ & $50.3\pm5.0$
2681: & $282\pm16$ & $219\pm30$
2682: & $7.696\pm0.025$ & High \\
2683: \cenA~e & $\mathcal{S}$ & $14.3\pm0.1$
2684: & $22.7\pm0.1$ & $2.07\pm0.12$ & $10.7\pm0.1$
2685: & $51.3\pm29.0$ & $97.2\pm32.0$
2686: & \nodata & Low \\
2687: & $\mathcal{L}$ & $23.8\pm7.4$
2688: & $87.3\pm14.3$ & $21.8\pm5.8$ & $36.7\pm5.7$
2689: & $219\pm19$ & $64.3\pm20.4$
2690: & $7.695\pm0.025$ & Low \\
2691: % M 51
2692: \M{51} & $\mathcal{S}$ & $64.1\pm0.9$
2693: & $123\pm3$ & $12.9\pm0.4$ & $57.2\pm1.7$
2694: & $264\pm151$ & $435\pm49$
2695: & \nodata & High \\
2696: & $\mathcal{L}$ & $125\pm10$
2697: & $398\pm20$ & $76.0\pm5.1$ & $131\pm8$
2698: & $843\pm25$ & $251\pm35$
2699: & $7.716\pm0.025$ & High \\
2700: \M{51}~p & $\mathcal{S}$ & $22.6\pm0.5$
2701: & $42.7\pm0.8$ & $4.93\pm0.1$ & $21.0\pm0.5$
2702: & $93.6\pm52.9$ & $161\pm17$
2703: & \nodata & High \\
2704: & $\mathcal{L}$ & $46.0\pm3.9$
2705: & $145\pm8$ & $28.8\pm2.1$ & $49.5\pm2.7$
2706: & $326\pm10$ & $82.4\pm10.6$
2707: & $7.721\pm0.025$ & High \\
2708: \M{51}~e & $\mathcal{S}$ & $34.9\pm0.4$
2709: & $68.4\pm1.0$ & $6.76\pm0.19$ & $30.7\pm0.8$
2710: & $145\pm83$ & $231\pm41$
2711: & \nodata & High \\
2712: & $\mathcal{L}$ & $61.6\pm8.4$
2713: & $212\pm15$ & $41.7\pm4.4$ & $71.5\pm5.4$
2714: & $436\pm19$ & $141\pm32$
2715: & $7.713\pm0.025$ & High \\
2716: % M 83
2717: \M{83} & $\mathcal{S}$ & $152\pm3$
2718: & $291\pm9$ & $33.9\pm1.4$ & $130\pm5$
2719: & $656\pm357$ & $901\pm174$
2720: & \nodata & High \\
2721: & $\mathcal{L}$ & $304\pm22$
2722: & $810\pm48$ & $147\pm14$ & $343\pm30$
2723: & $(1.79\pm0.06)\E{3}$ & $622\pm154$
2724: & $7.708\pm0.025$ & High \\
2725: \M{83}~p & $\mathcal{S}$ & $60.4\pm2.8$
2726: & $117\pm5$ & $14.2\pm0.5$ & $52.9\pm2.7$
2727: & $260\pm143$ & $458\pm28$
2728: & \nodata & High \\
2729: & $\mathcal{L}$ & $163\pm8$
2730: & $430\pm18$ & $103\pm5$ & $119\pm5$
2731: & $931\pm22$ & $311\pm19$
2732: & $7.707\pm0.025$ & High \\
2733: \M{83}~e & $\mathcal{S}$ & $99.4\pm1.9$
2734: & $187\pm6$ & $21.1\pm0.9$ & $84.3\pm3.2$
2735: & $426\pm230$ & $501\pm156$
2736: & \nodata & High \\
2737: & $\mathcal{L}$ & $163\pm16$
2738: & $540\pm40$ & $105\pm14$ & $247\pm28$
2739: & $(1.21\pm0.06)\E{3}$ & $180\pm94$
2740: & $7.712\pm0.025$ & High \\
2741: % Circinus
2742: Circinus & $\mathcal{S}$ & $79.1\pm0.5$
2743: & $211\pm6$ & $24.8\pm0.6$ & $89.4\pm5.5$
2744: & $421\pm244$ & $(22.8\pm1.0)\E{3}$
2745: & \nodata & Low \\
2746: & $\mathcal{L}$ & $347\pm18$
2747: & $(1.05\pm0.04)\E{3}$ & $173\pm8$ & $494\pm20$
2748: & $(2.55\pm0.05)\E{3}$ & $(3.58\pm0.14)\E{3}$
2749: & $7.684\pm0.025$ & High \\
2750: Circinus~e & $\mathcal{S}$ & $64.2\pm0.4$
2751: & $161\pm4$ & $20.4\pm0.4$ & $70.5\pm5.1$
2752: & $330\pm188$ & $945\pm67$
2753: & \nodata & Low \\
2754: & $\mathcal{L}$ & $261\pm19$
2755: & $774\pm38$ & $187\pm11$ & $385\pm19$
2756: & $(1.95\pm0.05)\E{3}$ & $(1.13\pm0.06)\E{3}$
2757: & $7.688\pm0.025$ & High \\
2758: % Arp 220
2759: \arp{220} & $\mathcal{S}$ & $0.48\pm0.03$
2760: & $6.33\pm0.37$ & $(3.63\pm1.72)\E{-2}$ & $1.50\pm0.02$
2761: & $11.5\pm7.1$ & $531\pm214$
2762: & \nodata & Low \\
2763: & $\mathcal{L}$ & $\lesssim59$
2764: & $68.5\pm23.7$ & $43.7\pm32.4$ & $36.2\pm21.1$
2765: & $259\pm76$ & $258\pm62$
2766: & $7.704\pm0.025$ & Low \\
2767: \arp{220}~p & $\mathcal{S}$ & $0.63\pm0.14$
2768: & $4.00\pm0.05$ & $\lesssim4.4\E{-2}$ & $0.90\pm0.01$
2769: & $6.64\pm4.23$ & $34.9\pm9.9$
2770: & \nodata & Low \\
2771: & $\mathcal{L}$ & $31.6\pm21.7$
2772: & $68.7\pm17.6$ & $78.3\pm38.7$ & $57.0\pm28.8$
2773: & $334\pm73$ & $247\pm41$
2774: & $7.715\pm0.025$ & Low \\
2775: % IR 15384
2776: \IR{15384} & $\mathcal{S}$ & $315\pm2$
2777: & $608\pm6$ & $51.4\pm2.8$ & $142\pm3$
2778: & $(1.22\pm0.70)\E{3}$ & $(3.60\pm0.07)\E{3}$
2779: & \nodata & High \\
2780: & $\mathcal{L}$ & $711\pm13$
2781: & $(2.13\pm0.02)\E{3}$ & $304\pm0.07$ & $305\pm4$
2782: & $(4.32\pm0.03)\E{3}$ & $(2.89\pm0.03)\E{3}$
2783: & $7.674\pm0.025$ & High \\
2784: % NGC 6240
2785: \ngc{6240} & $\mathcal{S}$ & $4.22\pm0.06$
2786: & $9.05\pm0.01$ & $0.91\pm0.05$ & $3.33\pm0.04$
2787: & $18.2\pm10.6$ & $51.2\pm6.1$
2788: & \nodata & High \\
2789: & $\mathcal{L}$ & $14.1\pm2.6$
2790: & $43.2\pm4.0$ & $6.42\pm1.05$ & $17.6\pm1.5$
2791: & $104\pm6$ & $96.0\pm8.5$
2792: & $7.725\pm0.025$ & High \\
2793: \ngc{6240}~p & $\mathcal{S}$ & $2.88\pm0.13$
2794: & $5.78\pm0.17$ & $0.49\pm0.05$ & $1.94\pm0.02$
2795: & $11.4\pm6.8$ & $32.6\pm3.0$
2796: & \nodata & High \\
2797: & $\mathcal{L}$ & $10.3\pm1.3$
2798: & $30.1\pm2.0$ & $3.78\pm0.43$ & $9.21\pm0.63$
2799: & $66.3\pm2.8$ & $59.4\pm4.4$
2800: & $7.722\pm0.025$ & High \\
2801: % IR 18317
2802: \IR{18317} & $\mathcal{S}$ & $292\pm17$
2803: & $527\pm4$ & $44.7\pm0.1$ & $122\pm5$
2804: & $(1.07\pm0.62)\E{3}$ & $(3.70\pm0.87)\E{3}$
2805: & \nodata & Low \\
2806: & $\mathcal{L}$ & $671\pm14$
2807: & $(1.99\pm0.02)\E{3}$ & $256\pm3$ & $275\pm4$
2808: & $(4.18\pm0.03)\E{3}$ & $(2.93\pm0.05)\E{3}$
2809: & $7.681\pm0.025$ & High \\
2810: % NGC 6946
2811: \ngc{6946} & $\mathcal{S}$ & $81.4\pm2.8$
2812: & $156\pm6$ & $16.0\pm0.5$ & $62.1\pm2.4$
2813: & $322\pm188$ & $433\pm47$
2814: & \nodata & High \\
2815: & $\mathcal{L}$ & $173\pm9$
2816: & $527\pm24$ & $88.7\pm5.6$ & $165\pm11$
2817: & $(1.10\pm0.03)\E{3}$ & $192\pm24$
2818: & $7.713\pm0.025$ & High \\
2819: \ngc{6946}~p & $\mathcal{S}$ & $24.8\pm1.8$
2820: & $51.7\pm3.3$ & $4.62\pm0.04$ & $17.6\pm1.3$
2821: & $100\pm60$ & $162\pm10$
2822: & \nodata & High \\
2823: & $\mathcal{L}$ & $77.9\pm4.1$
2824: & $211\pm8$ & $42.3\pm2.1$ & $58.4\pm4.3$
2825: & $462\pm11$ & $85.5\pm4.6$
2826: & $7.714\pm0.025$ & High \\
2827: \ngc{6946}~e & $\mathcal{S}$ & $56.8\pm1.8$
2828: & $105\pm4$ & $11.4\pm0.4$ & $44.6\pm1.8$
2829: & $222\pm128$ & $273\pm39$
2830: & \nodata & High \\
2831: & $\mathcal{L}$ & $105\pm6$
2832: & $327\pm17$ & $52.6\pm4.0$ & $119\pm9$
2833: & $687\pm21$ & $122\pm22$
2834: & $7.712\pm0.025$ & High \\
2835: % NGC 7027
2836: \ngc{7027} & $\mathcal{S}$ & $1.00\pm0.10$
2837: & $1.82\pm0.18$ & $0.14\pm0.02$ & $1.52\pm0.15$
2838: & $6.84\pm0.68$ & $33.4\pm3.3$
2839: & \nodata & Low \\
2840: & $\mathcal{L}$ & $1.38\pm0.14$
2841: & $4.05\pm0.41$ & $0.48\pm0.05$ & $4.34\pm0.43$
2842: & $17.0\pm0.7$ & $27.0\pm2.7$
2843: & $7.774\pm0.025$ & High \\
2844: % IR 22308
2845: \IR{22308} & $\mathcal{S}$ & $102\pm2$
2846: & $236\pm3$ & $25.0\pm1.0$ & $85.1\pm1.3$
2847: & $485\pm273$ & $906\pm75$
2848: & \nodata & High \\
2849: & $\mathcal{L}$ & $191\pm5$
2850: & $572\pm10.9$ & $100\pm2$ & $153\pm10$
2851: & $(1.31\pm0.02)\E{3}$ & $629\pm26$
2852: & $7.684\pm0.025$ & High \\
2853: % IR 23030
2854: \IR{23030} & $\mathcal{S}$ & $75.4\pm8.7$
2855: & $136\pm3$ & $17.4\pm0.4$ & $54.0\pm2.3$
2856: & $301\pm166$ & $(1.36\pm0.06)\E{3}$
2857: & \nodata & Low \\
2858: & $\mathcal{L}$ & $135\pm5$
2859: & $188\pm5$ & $21.5\pm0.6$ & $129\pm5$
2860: & $736\pm15$ & $(1.13\pm0.03)\E{3}$
2861: & $7.649\pm0.025$ & High \\
2862: % IR 23133
2863: \IR{23133} & $\mathcal{S}$ & $233\pm6$
2864: & $496\pm2$ & $66.8\pm0.4$ & $158\pm1$
2865: & $997\pm575$ & $(2.88\pm0.10)\E{3}$
2866: & \nodata & High \\
2867: & $\mathcal{L}$ & $493\pm16$
2868: & $(1.18\pm0.02)\E{3}$ & $216\pm2$ & $(2.71\pm0.03)\E{3}$
2869: & $(2.71\pm0.03)\E{3}$ & $(2.46\pm0.04)\E{3}$
2870: & $7.674\pm0.025$ & High \\
2871: % IR 23128
2872: \IR{23128} & $\mathcal{S}$ & $0.82\pm0.18$
2873: & $2.43\pm0.12$ & $0.27\pm0.09$ & $1.14\pm0.02$
2874: & $6.35\pm3.09$ & $26.4\pm5.5$
2875: & \nodata & Low \\
2876: & $\mathcal{L}$ & $5.24\pm1.95$
2877: & $13.1\pm2.6$ & $2.24\pm0.63$ & $3.38\pm0.54$
2878: & $35.7\pm5.2$ & $35.4\pm5.5$
2879: & $7.681\pm0.025$ & Low \\
2880: \IR{23128}~p & $\mathcal{S}$ & $0.79\pm0.02$
2881: & $1.69\pm0.07$ & $0.25\pm0.01$ & $0.74\pm0.01$
2882: & $4.23\pm2.10$ & $11.9\pm2.6$
2883: & \nodata & Low \\
2884: & $\mathcal{L}$ & $1.91\pm0.56$
2885: & $7.23\pm1.06$ & $2.57\pm0.56$ & $1.47\pm0.20$
2886: & $20.8\pm1.6$ & $6.66\pm0.90$
2887: & $7.688\pm0.025$ & Low \\
2888: % NGC 7714
2889: \ngc{7714} & $\mathcal{S}$ & $3.17\pm0.32$
2890: & $6.00\pm0.60$ & $0.90\pm0.09$ & $2.78\pm0.28$
2891: & $13.8\pm7.4$ & $43.5\pm4.4$
2892: & \nodata & High \\
2893: & $\mathcal{L}$ & $5.15\pm0.51$
2894: & $14.9\pm1.5$ & $2.75\pm0.28$ & $5.34\pm0.53$
2895: & $34.8\pm1.7$ & $36.4\pm3.6$
2896: & $7.681\pm0.025$ & High \\
2897: \enddata
2898: \label{tab:intens}
2899: % \tablerefs{}
2900: \tablecomments{We consider that a given fit has a ``high'' reliability, if
2901: $S/N\geq6$ at $\lambda=7.7\mic$,
2902: and $\rm I_{PAH}/I_{cont}\geq0.3$ with the \nlmet,
2903: and $\rm I_{PAH}/I_{cont}\geq0.5$ with the \frmet.
2904: Otherwise, the reliability is ``low''.}
2905: \end{deluxetable}
2906: \clearpage
2907:
2908: % IC 342
2909: \begin{figure*}[htbp]
2910: \centering
2911: \includegraphics[width=\textwidth]{f22a} \\
2912: \includegraphics[width=\textwidth]{f22b}
2913: \caption{PAH band ratios within \IC{342}.
2914: The same symbol conventions are adopted as in
2915: \reffig{fig:cormap1m82}.}
2916: \label{fig:cormap1ic342}
2917: \end{figure*}
2918: \begin{figure*}[htbp]
2919: \centering
2920: \begin{tabular}{cc}
2921: \includegraphics[width=0.48\textwidth]{f23a} &
2922: \includegraphics[width=0.48\textwidth]{f23b} \\
2923: \end{tabular}
2924: \caption{PAH band ratios within \IC{342} (continued).}
2925: \label{fig:cormap2ic342}
2926: \end{figure*}
2927: \begin{figure}[htbp]
2928: \centering
2929: \includegraphics[width=0.48\textwidth]{f24}
2930: \caption{Comparison between the two methods in \IC{342}.
2931: The same symbol conventions are adopted as in
2932: \reffig{fig:cormap1m82}.}
2933: \label{fig:compmapic342}
2934: \end{figure}
2935: \clearpage
2936: % M 83
2937: \begin{figure*}[htbp]
2938: \centering
2939: \includegraphics[width=\textwidth]{f25a} \\
2940: \includegraphics[width=\textwidth]{f25b}
2941: \caption{PAH band ratios within \M{83}.
2942: The same symbol conventions are adopted as in
2943: \reffig{fig:cormap1m82}.}
2944: \label{fig:cormap1m83}
2945: \end{figure*}
2946: \begin{figure*}[htbp]
2947: \centering
2948: \begin{tabular}{cc}
2949: \includegraphics[width=0.48\textwidth]{f26a} &
2950: \includegraphics[width=0.48\textwidth]{f26b} \\
2951: \end{tabular}
2952: \caption{PAH band ratios within \M{83} (continued).}
2953: \label{fig:cormap2m83}
2954: \end{figure*}
2955: \begin{figure}[htbp]
2956: \centering
2957: \includegraphics[width=0.48\textwidth]{f27}
2958: \caption{Comparison between the two methods in \M{83}.
2959: The same symbol conventions are adopted as in
2960: \reffig{fig:cormap1m82}.}
2961: \label{fig:compmapm83}
2962: \end{figure}
2963: \clearpage
2964: % 30 Dor
2965: \begin{figure*}[htbp]
2966: \centering
2967: \includegraphics[width=\textwidth]{f28a} \\
2968: \includegraphics[width=\textwidth]{f28b}
2969: \caption{PAH band ratios within \xxxdor.
2970: The same symbol conventions are adopted as in
2971: \reffig{fig:cormap1m82}.}
2972: \label{fig:cormap130dor}
2973: \end{figure*}
2974: \begin{figure*}[htbp]
2975: \centering
2976: \begin{tabular}{cc}
2977: \includegraphics[width=0.48\textwidth]{f29a} &
2978: \includegraphics[width=0.48\textwidth]{f29b} \\
2979: \end{tabular}
2980: \caption{PAH band ratios within \xxxdor\ (continued).}
2981: \label{fig:cormap230dor}
2982: \end{figure*}
2983: \begin{figure}[htbp]
2984: \centering
2985: \includegraphics[width=0.48\textwidth]{f30}
2986: \caption{Comparison between the two methods in \xxxdor.
2987: The same symbol conventions are adopted as in
2988: \reffig{fig:cormap1m82}.}
2989: \label{fig:compmap30dor}
2990: \end{figure}
2991: \clearpage
2992: % Orion Bar
2993: \begin{figure*}[htbp]
2994: \centering
2995: \includegraphics[width=\textwidth]{f31a} \\
2996: \includegraphics[width=\textwidth]{f31b}
2997: \caption{PAH band ratios within the \orb.
2998: The same symbol conventions are adopted as in
2999: \reffig{fig:cormap1m82}.}
3000: \label{fig:cormap1orionBar}
3001: \end{figure*}
3002: \begin{figure*}[htbp]
3003: \centering
3004: \begin{tabular}{cc}
3005: \includegraphics[width=0.48\textwidth]{f32a} &
3006: \includegraphics[width=0.48\textwidth]{f32b} \\
3007: \end{tabular}
3008: \caption{PAH band ratios within \orb\ (continued).}
3009: \label{fig:cormap2orionBar}
3010: \end{figure*}
3011: \begin{figure}[htbp]
3012: \centering
3013: \includegraphics[width=0.48\textwidth]{f33}
3014: \caption{Comparison between the two methods in \orb.
3015: The same symbol conventions are adopted as in
3016: \reffig{fig:cormap1m82}.}
3017: \label{fig:compmaporionBar}
3018: \end{figure}
3019: \clearpage
3020: % M 17
3021: \begin{figure*}[htbp]
3022: \centering
3023: \includegraphics[width=\textwidth]{f34a} \\
3024: \includegraphics[width=\textwidth]{f34b}
3025: \caption{PAH band ratios within \M{17}.
3026: The same symbol conventions are adopted as in
3027: \reffig{fig:cormap1m82}.}
3028: \label{fig:cormap1m17}
3029: \end{figure*}
3030: \begin{figure*}[htbp]
3031: \centering
3032: \begin{tabular}{cc}
3033: \includegraphics[width=0.48\textwidth]{f35a} &
3034: \includegraphics[width=0.48\textwidth]{f35b} \\
3035: \end{tabular}
3036: \caption{PAH band ratios within \M{17} (continued).}
3037: \label{fig:cormap2m17}
3038: \end{figure*}
3039: \begin{figure}[htbp]
3040: \centering
3041: \includegraphics[width=0.48\textwidth]{f36}
3042: \caption{Comparison between the two methods in \M{17}.
3043: The same symbol conventions are adopted as in
3044: \reffig{fig:cormap1m82}.}
3045: \label{fig:compmapm17}
3046: \end{figure}
3047: \clearpage
3048:
3049: % IC 342
3050: \begin{figure*}[htbp]
3051: \centering
3052: \plottwo{f37a}{f37b}
3053: \caption{Spatial distribution of the PAHs in \IC{342}.
3054: For each method, the image is the total PAH intensity, and the
3055: contours are the $\ipah{6.2}/\ipah{11.3}$ ratio.
3056: The shaded circle indicates the beam size.}
3057: \label{fig:imic342}
3058: \end{figure*}
3059: % M 83
3060: \begin{figure*}[htbp]
3061: \centering
3062: \plottwo{f38a}{f38b}
3063: \caption{Spatial distribution of the PAHs in \M{83}.
3064: For each method, the image is the total PAH intensity, and the
3065: contours are the $\ipah{6.2}/\ipah{11.3}$ ratio.
3066: The shaded circle indicates the beam size.}
3067: \label{fig:imm83}
3068: \end{figure*}
3069: % 30 Dor
3070: \begin{figure*}[htbp]
3071: \centering
3072: \plottwo{f39a}{f39b}
3073: \caption{Spatial distribution of the PAHs in \xxxdor.
3074: For each method, the image is the total PAH intensity, and the
3075: contours are the $\ipah{6.2}/\ipah{11.3}$ ratio.
3076: The shaded circle indicates the beam size.}
3077: \label{fig:im30dor}
3078: \end{figure*}
3079: % Orion bar
3080: \begin{figure*}[htbp]
3081: \centering
3082: \plottwo{f40a}{f40b}
3083: \caption{Spatial distribution of the PAHs in \orb.
3084: For each method, the image is the total PAH intensity, and the
3085: contours are the $\ipah{6.2}/\ipah{11.3}$ ratio.
3086: The shaded circle indicates the beam size.}
3087: \label{fig:imorionBar}
3088: \end{figure*}
3089: % M 17
3090: \begin{figure*}[htbp]
3091: \centering
3092: \plottwo{f41a}{f41b}
3093: \caption{Spatial distribution of the PAHs in \M{17}.
3094: For each method, the image is the total PAH intensity, and the
3095: contours are the $\ipah{6.2}/\ipah{11.3}$ ratio.
3096: The shaded circle indicates the beam size.}
3097: \label{fig:imm17}
3098: \end{figure*}
3099:
3100:
3101: % Acknowledgements
3102: %-----------------
3103: \acknowledgments
3104:
3105: We are grateful to Lou Allamandola and Henrik Spoon for in-depth useful
3106: discussions and detailed comments on our work.
3107: We thank the anonymous referee of this paper for comments that improved its
3108: quality.
3109: This work was performed, while F.~G. held a National Research
3110: Council research associateship award at NASA GSFC,
3111: and later a NASA Postdoctoral Program fellowship at NASA GSFC.
3112: This study is based essentially on observations made with ISO, an ESA project
3113: with instruments funded by ESA Member States (especially the PI countries:
3114: France, Germany, the Netherlands and the United Kingdom) and with the participation of ISAS and NASA.
3115: It is also based in part on observations made with the \spitzST,
3116: which is operated by the Jet Propulsion Laboratory, California
3117: Institute of Technology under a contract with NASA.
3118:
3119:
3120: % Facilities
3121: %-----------
3122: {\it Facilities:} \facility{ISO (CAM)},
3123: \facility{ISO (SWS)},
3124: \facility{Spitzer (IRS)}.
3125:
3126: % Bibliography
3127: %-------------
3128: \bibliographystyle{/Users/Fred/Documents/Astro/TeXstyle/Packages_AAS/aas}
3129: \bibliography{/Users/Fred/Documents/Astro/TeXstyle/references}
3130:
3131:
3132: \end{document}
3133: