astro-ph0611918/ms.tex
1: \documentclass{emulateapj}
2: \usepackage{apjfonts}
3: 
4: \slugcomment{Submitted to ApJ Letters on 10/26/2006; accepted on 11/14/2006}
5: 
6: \shorttitle{Mid-IR galaxy classification}
7: \shortauthors{Spoon et al.}
8: 
9: \begin{document}
10: 
11: \title{Mid-IR galaxy classification based on silicate 
12: obscuration and PAH equivalent width}
13: 
14: 
15: \author{H.W.W. Spoon\altaffilmark{1,2}}
16: \email{spoon@astro.cornell.edu}
17: \author{J.A. Marshall\altaffilmark{1}}
18: \author{J.R. Houck\altaffilmark{1}}
19: \author{M. Elitzur\altaffilmark{3}}
20: \author{L. Hao\altaffilmark{1}}
21: \author{L. Armus\altaffilmark{4}}
22: \author{B.R. Brandl\altaffilmark{5}}
23: \author{V. Charmandaris\altaffilmark{6,7}}
24: 
25: 
26: 
27: \altaffiltext{1}{Cornell University, Astronomy Department, Ithaca, NY 14853}
28: \altaffiltext{2}{Spitzer Fellow}
29: \altaffiltext{3}{Department of Physics and Astronomy, University of
30:              Kentucky, Lexington, KY 40506}
31: \altaffiltext{4}{Caltech, Spitzer Science Center, MS 220-6, Pasadena, CA 91125}
32: \altaffiltext{4}{Leiden Observatory, PO Box 9513, 2300 RA Leiden,
33:              The Netherlands}
34: \altaffiltext{6}{Department of Physics, University of Crete,   
35:              GR-71003, Heraklion, Greece}
36: \altaffiltext{7}{IESL / Foundation for Research and Technology-Hellas, 
37:              PO Box 1527, GR-71110, Heraklion, Greece, and Chercheur 
38:              Associ\'e, Observatoire de Paris, F-75014,  Paris, France}
39: 
40: 
41: \begin{abstract}
42: We present a new diagnostic diagram for mid-infrared spectra of 
43: infrared galaxies based on the equivalent width of the 6.2\,$\mu$m PAH
44: emission feature and the strength of the 9.7\,$\mu$m silicate
45: feature. Based on the position in this diagram we classify galaxies 
46: into 9 classes ranging from continuum-dominated AGN hot dust spectra
47: and PAH-dominated starburst spectra to absorption-dominated spectra
48: of deeply obscured galactic nuclei.
49: We find that galaxies are systematically distributed along 
50: two distinct branches: one of AGN and starburst-dominated spectra 
51: and one of deeply obscured nuclei and starburst-dominated spectra. 
52: The separation into two branches likely reflects a fundamental
53: difference in the dust geometry in the two
54: sets of sources: clumpy versus non-clumpy obscuration.
55: Spectra of ULIRGs are found along the full length of both branches, 
56: reflecting the diverse nature of the ULIRG family.
57: \end{abstract}
58: 
59: \keywords{galaxies: ISM --- infrared: galaxies ---
60:  galaxies: active --- galaxies: starburst} 
61: 
62: 
63: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
64: \section{Introduction}
65: 
66: Over the last decade several diagnostic diagrams have been 
67: proposed to quantify the contribution of star formation and
68: AGN activity to the infrared luminosity of infrared galaxies
69: based on mid-infrared (to far-infrared) continuum slope,
70: PAH line-to-continuum ratio, PAH to far-infrared luminosity
71: ratio and the ratio of a high to a low ionization forbidden
72: line such as [Ne{\sc v}]/[Ne{\sc ii}] 
73: \citep{genzel98,lutz98,laurent00,peeters04,dale06,sturm06}.
74: However, none of these diagrams takes into account the effects 
75: of strong obscuration of the nuclear power source.
76: 
77: With the advent of the Infrared Spectrograph \citep[{\it IRS};][]{houck04} 
78: on board the {\it Spitzer} Space Telescope \citep{werner04} 
79: astronomers have been handed a powerful tool 
80: to study the 5--37\,$\mu$m range for a wide range of galaxy types
81: at an unprecedented sensitivity. This enables for the first time
82: a systematic study of a large number of galaxies over the wavelength 
83: range in which amorphous silicate grains have strong opacity peaks 
84: due to the Si--O stretching and the O--Si--O bending modes centered 
85: at 9.7 and 18\,$\mu$m, respectively.
86: 
87: Here we will introduce the strength of the 9.7\,$\mu$m silicate feature
88: as a tool to distinguish between different dust geometries in the 
89: central regions of (ultra-)(luminous) infrared galaxies as part
90: of a new diagnostic diagram and mid-infrared galaxy classification 
91: scheme.
92: 
93: 
94: 
95: 
96: 
97: 
98: 
99: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
100: \section{Observations and data reduction}
101: 
102: 
103: The results presented in this paper are based on both {\it Spitzer IRS}
104: and {\it ISO-SWS} observations. The core sample is formed by the IRS 
105: GTO ULIRG sample (PID 105; J.R.\,Houck PI), which comprises $\sim$100 
106: ULIRGs in the redshift range 0.02 $< z <$ 0.93.
107: This sample is compared to samples of AGN and starburst templates
108: from the IRS GTO programs 14 \citep{brandl06,weedman05} and 96
109: and to selected AGNs from the GTO programs 82 and 86 \citep{shi06}.
110: Additional IRGs, LIRGs and ULIRGs were taken from program 159 
111: \citep[NGC\,1377;][]{roussel06} and from the Spitzer DDT program
112: 1096. Archival {\it ISO-SWS} spectra of starburst nuclei were taken from 
113: \cite{sturm00}.
114: 
115: The {\it Spitzer} observations were made with the Short-Low (SL) and 
116: Long-Low (LL) modules of the {\em IRS}. The spectra were extracted 
117: from DROOP level images provided by the Spitzer Science Center 
118: (34 using pipeline version S11.0.2, 131 using pipeline version
119: S14.0) and background-subtracted 
120: by differencing the first and second order apertures. The spectra
121: were calibrated using the {\em IRS} standard stars HD\,173511 
122: (5.2--19.5\,$\mu$m) and $\xi$\,Dra (19.5--38.5\,$\mu$m).
123: %
124: After flux calibration the orders were stitched to LL order 1, requiring 
125: order-to-order scaling adjustments of typically less than 10\%. 
126: 
127: 
128: 
129: 
130: 
131: %+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
132: \begin{figure*}
133: \epsscale{1.15}
134: \plotone{f1.eps}
135: \caption{Diagnostic plot of the equivalent width of the 6.2\,$\mu$m 
136: PAH emission feature versus the 9.7\,$\mu$m silicate strength.
137: Upper and lower limits are denoted by arrows. The galaxy spectra 
138: are classified into 9 classes (identified by 9 shaded rectangles) 
139: based on their position in this plot. 
140: Colors are used to distinguish the various classes. 
141: From class 1A to 1B, 1C, 2A, 2B, 2C, 3A and 3B the colors used are: 
142: {\it green, cyan, dark blue, yellow, pink, purple, red} and {\it orange}. 
143: %
144: After class assignment the PAH equivalent widths were corrected for
145: the effect of 6\,$\mu$m water ice absorption on the 6.2\,$\mu$m
146: continuum. The extent of the individual corrections are indicated by 
147: dotted horizontal lines.
148: %
149: The two dotted black lines are mixing lines between the spectrum of 
150: the deeply obscured nucleus of NGC\,4418 and the starburst
151: nuclei of M\,82 and NGC\,7714, respectively.
152: %
153: Galaxy types are distinguished by their plotting symbol:
154: {\it filled circles}: ULIRGs and HyLIRGs; {\it filled triangles}: 
155: starburst galaxies; {\it filled squares}: Seyferts and QSOs; 
156: {\it filled diamonds}: other infrared galaxies.
157: \label{fig1}}
158: \end{figure*}
159: %+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
160: 
161: 
162: 
163: 
164: 
165: 
166: 
167: 
168: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
169: \section{Analysis}
170: 
171: For all the spectra in our sample we have measured the equivalent 
172: width of the 6.2\,$\mu$m PAH emission feature as well as the 
173: strength of the 9.7\,$\mu$m silicate feature and plotted the two 
174: quantities in a diagnostic diagram (Fig.\,\ref{fig1}).
175: 
176: The flux in the 6.2\,$\mu$m PAH emission band is measured by
177: integrating the flux above a spline interpolated local continuum 
178: from 5.95--6.55\,$\mu$m. The equivalent
179: width (EW) of the PAH feature is then obtained by dividing the 
180: integrated PAH flux by the interpolated continuum flux density 
181: below the peak ($\sim$ 6.22\,$\mu$m) of the PAH feature.
182: 
183: The apparent strength of the 9.7\,$\mu$m silicate feature
184: is inferred by adopting a local mid-infrared continuum 
185: and evaluating the ratio of observed flux density (f$_{\rm obs}$) to 
186: continuum flux density (f$_{\rm cont}$) at 9.7\,$\mu$m and defining 
187: %
188: \begin{equation}
189: S_{sil} = \ln \frac{f_{\rm obs} (9.7\,\mu m)}{f_{\rm cont} (9.7\,\mu m)}
190: \end{equation}
191: %
192: For sources with a silicate absorption feature $S_{sil}$ can be 
193: interpreted as the negative of the apparent silicate optical depth.
194: 
195: Given the great diversity among our mid-infrared galaxy spectra, 
196: there is no ``one-size-fits-all'' procedure to define the local 
197: continuum in all spectra. We therefore developed separate methods for 
198: three distinctly different types of mid-infrared galaxy spectra: 
199: continuum-dominated spectra, PAH-dominated spectra and 
200: absorption-dominated spectra. These methods are illustrated in 
201: Fig.\,\ref{fig2} and described in its caption.
202: %
203: We are forcing the local continuum to touch down at
204: 14.0--14.5\,$\mu$m, because the ISM dust cross-section 
205: decreases between the two silicate peaks in this region. 
206: Detailed radiative transfer calculations verify that 
207: our interpolation procedure properly reproduces the 
208: emission that would be generated by dust stripped of 
209: its silicate features (Sirocky et al. in preparation).
210: 
211: %+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
212: \begin{figure}
213: \epsscale{1.15}
214: \plotone{f2.eps}
215: \caption{Comparison of several methods used to determine the 
216: apparent 9.7\,$\mu$m silicate optical depth. The three methods
217: differ by the way the local mid-infrared continuum is defined.
218: %
219: For continuum-dominated spectra (top) simple logarithmic 
220: linear fits are made to the 5.0--7.5\,$\mu$m and 26.1--31.5\,$\mu$m
221: Jy-spectrum. These are spline interpolated over the 7.5--26\,$\mu$m
222: range, using a continuum point at 14.0\,$\mu$m to guide the
223: spline in the inter-silicate-feature range. Pivots and fit
224: are shown in {\it green}. 
225: %
226: For PAH-dominated spectra (2nd and 3rd spectrum)
227: the 5.5--14.5\,$\mu$m continuum is estimated from a power law
228: interpolation over this range. From 14.5 to 26.0\,$\mu$m this
229: power law is continued as a spline function with a curvature 
230: determined by continuum pivots at 26.0 and 30.0\,$\mu$m. Pivots 
231: and fit are shown in {\it blue}. 
232: %
233: For absorption-dominated spectra (4th and 5th spectrum)
234: the local 5.6--26\,$\mu$m continuum is determined from a spline
235: interpolation to continuum pivots at 5.2, 5.6, 14.0 and 
236: 26.0\,$\mu$m. An additional continuum pivot at 7.8\,$\mu$m
237: is included for spectra with very little PAH emission (e.g.
238: bottom spectrum). Pivots and fit are shown in {\it red}. 
239: %
240: In all three cases the 9.7\,$\mu$m silicate strength 
241: is computed by evaluating the ratio of observed and interpolated 
242: continuum at its peak. From top to bottom, the five spectra 
243: shown are the IRS low-resolution spectra of PG\,1351+640, NGC\,7714,
244: NGC\,3628, UGC\,5101 and NGC\,4418.
245: \label{fig2}}
246: \end{figure}
247: %+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
248: 
249: 
250: 
251: 
252: \subsection{Galaxy classification}
253: 
254: The galaxy spectra in Fig.\,\ref{fig1} are classified into
255: 9 different classes based on their 6.2\,$\mu$m PAH EW and
256: 9.7\,$\mu$m silicate strength. The parameter space
257: covered by the various classes is indicated by shaded 
258: rectangles in Fig.\,\ref{fig1}. Average mid-infrared 
259: spectra for the eight populated classes are shown in 
260: Fig.\,\ref{fig3}.
261: %
262: The average spectra were constructed by normalizing all spectra 
263: to unity at 14.5\,$\mu$m flux before the averaging process. 
264: In order to maximize the signal-to-noise (S/N) of the average 
265: spectra, low S/N spectra were discarded from the process.
266: Below we describe the 8 average spectra in the order (bottom 
267: to top) they are presented in Fig.\,\ref{fig3}. The ninth
268: class, 3C, is not populated.
269: 
270: The class 1A spectrum is characterized by a nearly featureless
271: hot dust continuum with a very weak silicate absorption feature 
272: at 9.7\,$\mu$m. The class 1B spectrum differs from the 
273: class 1A spectrum by clearly showing the family of PAH emission 
274: features at 6.2, 7.7, 8.6, 11.2, 12.7 and 17.3\,$\mu$m on top 
275: of a hot dust continuum. This hot dust continuum is nearly 
276: absent in the class 1C spectrum, allowing the PAH emission 
277: features to dominate the mid-infrared spectral appearance. 
278: Silicate absorption at
279: 9.7\,$\mu$m becomes noticable in the class 2C spectrum as 
280: an increased depth of the depression between the 6--9 and 
281: 11--13\,$\mu$m PAH emission complexes. Another marked 
282: difference with the class 1C spectrum is the steepening
283: of the 20--30\,$\mu$m continuum and the appearance of an
284: 18\,$\mu$m silicate absorption feature. In the class 2B 
285: spectrum the PAH features appear weaker than in the 
286: class 2C spectrum. In addition, the 
287: spectrum starts to show a 6\,$\mu$m water ice and a 
288: 6.85\,$\mu$m aliphatic hydrocarbon absorption band.
289: In the class 3B spectrum these absorption features reach
290: their maximum depths, while the 9.7 and 18\,$\mu$m silicate 
291: features continue to increase in depth up to class 3A.
292: Absorption features of {\it crystalline} silicates \citep{spoon06}
293: appear in the spectra of the classes 3B and 3A at 16, 19 and 23\,$\mu$m. 
294: Equivalent widths of PAH emission features and emission 
295: lines decrease from class 2B to 3B and 3A. Note especially
296: the change in shape of the 7.7\,$\mu$m PAH feature as
297: it first broadens and then disappears going from class 
298: 2B to 3B and 3A. Finally, the class 2A spectrum differs 
299: from the class 3A spectrum mainly by a clearly lower 
300: apparent depth of the 9.7 and 18\,$\mu$m silicate features.
301: Note that the spectral structure in individual spectra may 
302: differ substantially from the average properties of the
303: classes. In Fig.\,\ref{fig3} this is represented by the
304: 1-$\sigma$ dispersion ranges around the individual average 
305: spectra.
306: 
307: 
308: 
309: 
310: 
311: 
312: 
313: 
314: 
315: 
316: \subsection{Galaxy distribution}
317: 
318: The galaxies shown in Fig.\,\ref{fig1} are color-coded according 
319: to their galaxy classification. However, their positions may differ 
320: from their original assignments, as in Fig.\,\ref{fig1} the 6.2\,$\mu$m 
321: PAH EW has been corrected for the effect of 6\,$\mu$m water ice 
322: absorption on the 6.2\,$\mu$m continuum through substitution of 
323: the observed 6.2\,$\mu$m continuum by the 6.2\,$\mu$m continuum 
324: that was defined to infer the 9.7\,$\mu$m silicate strength. 
325: The ice-correction is justified assuming that the ice is part of 
326: the obscuring medium {\it behind} the PAH emitting region.
327: The correction is applied only to spectra clearly showing the 
328: imprint of water ice absorption, as found among the spectra of 
329: classes 2B, 2C, 3A and 3B. 
330: The extent of the correction for individual spectra is indicated 
331: in Fig.\,\ref{fig1} by horizontal dotted lines. 
332: 
333: The ULIRGs, Seyferts, quasars and starburst galaxies shown in
334: Fig.\,\ref{fig1} are not distributed randomly through the diagram.
335: Instead, the galaxies appear to be distributed along two branches.
336: One extending horizontally from the positions of the prototypical
337: Seyfert-1 nucleus NGC\,4151 \citep{weedman05} to the prototypical 
338: starburst nucleus NGC\,7714 \citep{brandl04}, the other branch 
339: extending diagonally from the 
340: prototypical deeply obscured nucleus of NGC\,4418 \citep{spoon01} 
341: to the starburst nuclei of M\,82 \citep{sturm00} and NGC7714.
342: Very few sources are found above and to the right of the diagonal 
343: branch and in between the two branches (the class 2A domain).
344: 
345: The three extremes distinguish between AGN-heated hot dust 
346: spectra (class 1A), PAH-dominated spectra (classes 1C and 2C)
347: and absorption-dominated spectra (class 3A).  The galaxies lined 
348: up along the two branches in between the extremes show signatures
349: of both extremes to varying proportions. The galaxies along the
350: horizontal branch may be thought of as combinations of AGN and 
351: starburst activity, while the galaxies along the diagonal branch 
352: may be thought of as intermediate stages in between 
353: a fully obscured galactic nucleus and an unobscured nuclear 
354: starburst. The latter is illustrated by the two dotted lines 
355: in Fig.\,\ref{fig1}, which represent mixing lines between 
356: the spectrum of NGC\,4418 on one extreme and either NGC\,7714 or
357: M\,82 on the other. The large majority of galaxies on the diagonal 
358: branch fall in between these mixing lines. One notable exception 
359: is Arp\,220.
360: 
361: Relatively few galaxies are found in the section between the two 
362: branches: the domain of the class 2A sources. Spectroscopically, 
363: there are several simple scenarios which will populate this class 
364: from adjacent classes.
365: %
366: A class 3A galaxy can be turned into a class 2A galaxy by 
367: adding a featureless hot dust continuum to the class 3A absorbed 
368: continuum spectrum. The main effect of this continuum dilution is 
369: the filling in of the deep 9.7\,$\mu$m silicate feature, resulting 
370: in a less pronounced feature. Increasing the dilution
371: not only further decreases the depth of the silicate feature
372: but also increases the continuum at 6.2\,$\mu$m, resulting in 
373: both a less negative silicate strength and a lower 6.2\,$\mu$m 
374: PAH EW. The effect of this type of continuum dilution is illustrated 
375: in Fig.\,\ref{fig1} by a curved arrow.
376: %
377: Alternatively, a class 1A galaxy may be turned into class 2A 
378: galaxy by passing its mid-infrared spectrum through a foreground 
379: cold dust screen. The path that a galaxy will take is straight up, 
380: since the PAH EW is not affected by cold foreground extinction.
381: 
382: The presence of foreground dust or dust mixed in within the PAH
383: emitting region may also explain the difference between class 1C 
384: and 2C starburst galaxies, and the difference between the 
385: classification of the prototypical starburst galaxies NGC\,7714 
386: and M\,82 in particular. In the latter case this may be primarily 
387: an orientation effect, as the nuclear starburst in NGC\,7714 is 
388: observed face-on, whereas the starburst ring in M\,82 is seen under 
389: an inclination of $\sim$80 degrees.
390: 
391: Finally, the complete absence of galaxy spectra classified as 3C must imply 
392: that strong, pure foreground, cold dust obscuration is not common 
393: in starburst nuclei. The obscuring dust instead contributes to the
394: continuum underlying the 6.2\,$\mu$m PAH feature, reducing its EW.
395: 
396: 
397: 
398: %+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
399: \begin{figure}
400: \epsscale{1.15}
401: \plotone{f3.eps}
402: \caption{Average spectra for eight of the classes of galaxy
403: spectra defined in Fig.\,\ref{fig2}, along with the 1-$\sigma$ 
404: dispersion range ({\it grey} shading). Most of the dispersion
405: is due to the overall slope of the spectrum. Spectra at the 
406: bottom of the
407: figure (classes 1A and 1B) are dominated by hot dust emission, 
408: spectra in the middle (classes 1C and 2C) by PAH emission and
409: spectra towards the top by absorption features 
410: (classes 3B, 3A and 2A).
411: Vertical dotted lines denote the central wavelengths 
412: of the family of PAH emission features at 6.2, 7.7, 8.6, 11.2 and 
413: 12.7\,$\mu$m. 
414: \label{fig3}}
415: \end{figure}
416: %+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
417: 
418: 
419: 
420: 
421: 
422: 
423: 
424: 
425: 
426: 
427: 
428: 
429: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
430: 
431: \section{Discussion and conclusions}
432: 
433: We have constructed a diagnostic diagram of two mid-infrared
434: spectral quantities, the equivalent width of the 6.2\,$\mu$m 
435: PAH feature and the apparent 9.7\,$\mu$m silicate strength, for 
436: the purpose of classifying infrared galaxies according to their 
437: mid-infrared spectral shape (Fig.\,\ref{fig1}).
438: 
439: The large majority of galaxies ($>$90\%) are dispersed around 
440: two branches: a horizontal one extending from 
441: continuum-dominated spectra to PAH-dominated spectra and a 
442: diagonal one spanned between absorption-dominated spectra 
443: and PAH-dominated spectra.
444: %
445: Seyfert galaxies and quasars (square plotting symbols in 
446: Fig.\,\ref{fig1}) are found exclusively on the horizontal branch
447: characterized by the absence of pronounced silicate features.
448: %
449: Starburst galaxy spectra (triangular plotting symbols) are 
450: concentrated toward the extreme right 
451: tip of the two branches, the locus of PAH-dominated spectra.
452: %
453: Of the ULIRGs and HyLIRGs in our sample (round plotting symbols) 
454: about a quarter occupy the same
455: space as the AGNs and starbursts in our sample. The large
456: majority, however, are distributed along the diagonal branch,
457: characterized by increasingly apparent silicate absorption
458: and less pronounced PAH emission features, with deeply
459: obscured galactic nuclei such as NGC\,4418 \citep{spoon01} 
460: and NGC\,1377 \citep{roussel06} as the only non-ULIRG, 
461: non-Seyfert end members on this branch.
462: %
463: Note that low-metallicity galaxies (not included in this
464: study) would be distributed along the horizontal branch.
465: 
466: The existence of two distinct regimes of apparent silicate 
467: absorption, as suggested by the two distinct branches in 
468: Fig.\,\ref{fig1}, is intriguing and likely points to clear 
469: differences in the nuclear dust distribution between galaxies 
470: on the two branches. Clumpy distributions produce only 
471: shallow absorption features \citep{nenkova02}, and are 
472: the likely explanation for the horizontal branch. The deep 
473: silicate absorption found on the slanted branch requires the 
474: nuclear source to be deeply embedded in a smooth distribution 
475: of dust that is both geometrically and optically thick 
476: \citep{levenson06}. The transition from clumpy geometry to 
477: one dominated by a smooth dust distribution is a possible 
478: explanation for the sources located between the branches.
479: 
480: The distribution of ULIRG spectra along the full length
481: of both the horizontal and the diagonal branch once again
482: illustrates the diverse nature of the ULIRG family. 
483: It, however, also raises the issue of ULIRG evolution.
484: For instance, what evolutionary path has taken 
485: IRAS\,08572+3915 to its extreme class 3A position in 
486: which the central power source is deeply buried? 
487: Did the interaction start with both nuclei classified
488: as class 1C/2C starburst galaxies, which then gradually 
489: moved up the diagonal branch as the interaction 
490: stengthened and more dust accumulated on the remnant
491: nuclei? And how will IRAS\,08572+3915 evolve from
492: there, once the obscuring screen breaks up and the hidden 
493: power sources are revealed? In case of a hidden starburst
494: the source likely will move back diagonally toward the 
495: starburst locus. However, if the dominant power source 
496: is an AGN, will the source cross into the sparsely 
497: populated class 2A regime, evolving directly toward the 
498: AGN-dominated class 1A regime, or will it first 
499: undergo a starburst before settling somewhere along 
500: the horizontal branch?
501: The sparse population of the class 2A domain either 
502: indicates that the crossing time is brief -- with
503: few ULIRGs caught in transformation -- or that deeply
504: obscured ULIRG nuclei mostly evolve to starbursts 
505: (first).
506: 
507: In a more extensive paper (Spoon et al. in preparation) 
508: we will present correlations of additional parameters 
509: within the diagram (e.g. mid-IR line ratios, L$_{\rm IR}$, 
510: L$_{\rm PAH}$/L$_{\rm IR}$,absorption features, nuclear 
511: separation, optical classification).
512: 
513: 
514: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
515: 
516: 
517: \acknowledgements
518: 
519: The authors wish to thank Jer\'onimo Bernard-Salas, 
520: Nancy Levenson, H\'el\`ene Roussel, Matthew Sirocky,
521: Alexander Tielens and Dan Weedman for discussions, and
522: Vandana Desai for help with the data reduction. 
523: Support for this work was provided by 
524: NASA through Contract Number 1257184 issued by the Jet Propulsion 
525: Laboratory, California Institute of Technology under NASA contract 
526: 1407.  HWWS was supported under this contract through the 
527: Spitzer Space Telescope Fellowship Program. 
528: 
529: 
530: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
531: 
532: \begin{thebibliography}{}
533: 
534: \bibitem[Brandl et al.(2004)Brandl et al.]{brandl04} 
535:          Brandl, B.R., et al., 2004, \apjs 154, 188
536: 
537: \bibitem[Brandl et al.(2006)Brandl et al.]{brandl06} 
538:          Brandl, B.R., et al., 2006, \apj, 653 (astro\_ph/0609024)
539: 
540: \bibitem[Dale et al.(2006)Dale et al.]{dale06} 
541:          Dale, D.A., et al., 2006, \apj, 646, 161
542: 
543: \bibitem[Genzel et al.(1998)Genzel et al.]{genzel98}
544:          Genzel, R., et al., 1998, \apj, 498, 579
545: 
546: \bibitem[Houck et al.(2004)Houck et al.]{houck04} 
547:          Houck, J.R., et al. 2004, \apjs, 154, 18
548: 
549: \bibitem[Laurent et al.(2000)Laurent et al.]{laurent00} 
550:          Laurent, O., et al., 2000, \aap, 359, 887
551: 
552: \bibitem[Levenson et al.(2006)Levenson et al.]{levenson06}
553:          Levenson, N.A., et al., 2006, \apjl, accepted
554: 
555: \bibitem[Lutz et al.(1998)Lutz et al.]{lutz98}
556:          Lutz, D., et al., 1998, \apjl, 505, L103
557: 
558: \bibitem[Nenkova, Ivezic \& Elitzur (2002)Nenkova, Ivezic \& Elitzur]{nenkova02}
559: Nenkova, M., Ivezic, Z., \& Elitzur, M. 2002, \apjl, 570, L9
560: 
561: \bibitem[Peeters et al.(2004)Peeters et al.]{peeters04}
562:          Peeters, E., Spoon, H.W.W., Tielens, A.G.G.M.,
563:          2004, \aap, 613, 986
564: 
565: \bibitem[Roussel et al.(2006)Roussel et al.]{roussel06} 
566:          Roussel, H., et al., 2006, \apj, 646, 841
567: 
568: \bibitem[Shi et al.(2006)Shi et al.]{shi06} 
569:          Shi, Y., et al., 2006, 
570:          \apj, in press (astro\_ph/0608645)
571: 
572: \bibitem[Spoon et al.(2001)Spoon et al.]{spoon01} 
573:          Spoon, H.W.W., et al., 2001, \aap, 365, L353
574: 
575: \bibitem[Spoon et al.(2006)Spoon et al.]{spoon06} 
576:          Spoon, H.W.W., et al., 2006, \apj, 638, 759
577: 
578: \bibitem[Sturm et al.(2000)Sturm et al.]{sturm00}
579:          Sturm, E., et al., 2000, \aap, 358, 481
580: 
581: \bibitem[Sturm et al.(2006)Sturm et al.]{sturm06}
582:          Sturm, E., et al., 2006, \apjl, in press (astro\_ph/0610772)
583: 
584: \bibitem[Weedman et al.(2005)Weedman et al.]{weedman05} 
585:          Weedman, et al., 2005, \apj
586: 
587: \bibitem[Werner et al.(2004)Werner et al.]{werner04} 
588:          Werner, M.W., et al., 2004, \apjs, 154, 1
589: 
590: \end{thebibliography}
591: 
592: \end{document}
593: