1: %% Beginning of file 'sample.tex'

2: %%

3: %% Modified 2005 December 5

4: %%

5: %% This is a sample manuscript marked up using the

6: %% AASTeX v5.x LaTeX 2e macros.

7: 

8: %% The first piece of markup in an AASTeX v5.x document

9: %% is the \documentclass command. LaTeX will ignore

10: %% any data that comes before this command.

11: 

12: %% The command below calls the preprint style

13: %% which will produce a one-column, single-spaced document.

14: %% Examples of commands for other substyles follow. Use

15: %% whichever is most appropriate for your purposes.

16: %%

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

18: 

19: %% manuscript produces a one-column, double-spaced document:

20: 

21: %\documentclass[manuscript]{aastex}

22: 

23: %% preprint2 produces a double-column, single-spaced document:

24: %\documentclass{emulateapj}

25: \usepackage{amsmath}

26: 

27: %\documentclass[preprint2]{aastex}

28: 

29: %% Sometimes a paper's abstract is too long to fit on the

30: %% title page in preprint2 mode. When that is the case,

31: %% use the longabstract style option.

32: 

33: %% \documentclass[preprint2,longabstract]{aastex}

34: 

35: %% If you want to create your own macros, you can do so

36: %% using \newcommand. Your macros should appear before

37: %% the \begin{document} command.

38: %%

39: %% If you are submitting to a journal that translates manuscripts

40: %% into SGML, you need to follow certain guidelines when preparing

41: %% your macros. See the AASTeX v5.x Author Guide

42: %% for information.

43: 

44: \newcommand{\vdag}{(v)^\dagger}

45: \newcommand{\myemail}{quanz@mpia.de}

46: 

47: %% You can insert a short comment on the title page using the command below.

48: 

49: \slugcomment{To appear in ApJ}

50: 

51: %% If you wish, you may supply running head information, although

52: %% this information may be modified by the editorial offices.

53: %% The left head contains a list of authors,

54: %% usually a maximum of three (otherwise use et al.).  The right

55: %% head is a modified title of up to roughly 44 characters.

56: %% Running heads will not print in the manuscript style.

57: 

58: \shorttitle{Dust and ices in MIR spectra of FUors}

59: \shortauthors{Quanz et al.}

60: 

61: %% This is the end of the preamble.  Indicate the beginning of the

62: %% paper itself with \begin{document}.

63: 

64: \begin{document}

65: 

66: %% LaTeX will automatically break titles if they run longer than

67: %% one line. However, you may use \\ to force a line break if

68: %% you desire.

69: 

70: \title{Evolution of dust and ice features around FU Orionis objects\altaffilmark{1}}

71: 

72: %% Use \author, \affil, and the \and command to format

73: %% author and affiliation information.

74: %% Note that \email has replaced the old \authoremail command

75: %% from AASTeX v4.0. You can use \email to mark an email address

76: %% anywhere in the paper, not just in the front matter.

77: %% As in the title, use \\ to force line breaks.

78: 

79: \author{S. P. Quanz, Th. Henning, J. Bouwman, R. van Boekel, A. Juh\'asz, H. Linz,}

80: \affil{Max Planck Institute for Astronomy, K\"onigstuhl 17, Heidelberg,

81:     Germany}

82: \email{\myemail}

83: \author{K.~M. Pontoppidan}

84: \affil{California Institute of Technology, Division of Geological and Planetary Sciences, Pasadena, CA; Hubble Fellow}

85: \and

86: \author{F. Lahuis}

87: \affil{Leiden Observatory, P.O. Box 9513, 2300 RA Leiden, Netherlands}

88: \affil{SRON Netherlands Institute for Space Research, Groningen, Netherlands}

89: 

90: \altaffiltext{1}{Based on observations with ISO, an ESA project with instruments funded by ESA Member States (especially the PI countries: France, Germany, the Netherlands and the United Kingdom) and with the participation of ISAS and NASA. This work is based in part on observations made with the {\sc Spitzer Space Telescope}, which is operated by the Jet Propulsion Laboratory, California Institute of Technology under a contract with NASA.}

91: 

92: %\email{aastex-help@aas.org}

93: 

94: %\and

95: 

96: %\author{R. J. Hanisch\altaffilmark{5}}

97: %\affil{Space Telescope Science Institute, Baltimore, MD 21218}

98: 

99: %% Notice that each of these authors has alternate affiliations, which

100: %% are identified by the \altaffilmark after each name.  Specify alternate

101: %% affiliation information with \altaffiltext, with one command per each

102: %% affiliation.

103: 

104: %\altaffiltext{1}{Visiting Astronomer, Cerro Tololo Inter-American Observatory.

105: %CTIO is operated by AURA, Inc.\ under contract to the National Science

106: %Foundation.}

107: %\altaffiltext{2}{Society of Fellows, Harvard University.}

108: %\altaffiltext{3}{present address: Center for Astrophysics,

109: %    60 Garden Street, Cambridge, MA 02138}

110: %\altaffiltext{4}{Visiting Programmer, Space Telescope Science Institute}

111: %\altaffiltext{5}{Patron, Alonso's Bar and Grill}

112: 

113: %% Mark off your abstract in the ``abstract'' environment. In the manuscript

114: %% style, abstract will output a Received/Accepted line after the

115: %% title and affiliation information. No date will appear since the author

116: %% does not have this information. The dates will be filled in by the

117: %% editorial office after submission.

118: 

119: \begin{abstract}

120: We present spectroscopy data  

121: for a sample of 14 FUors and 2 TTauri stars observed with the 

122: {\sc Spitzer Space Telescope} or with the {\sc Infrared Space Observatory (ISO)}.

123: Based on the appearance of the 10\,$\mu$m silicate feature we define 2 categories of FUors.

124: Objects showing the silicate feature in absorption (Category 1) are still embedded in a dusty and icy

125: envelope. The shape of the 10\,$\mu$m silicate absorption bands is compared to typical dust compositions  

126: of the interstellar medium and found to be in general agreement. Only one object (RNO 1B) appears to be too

127: rich in amorphous pyroxene dust, 

128: but a superposed emission feature can explain the observed shape. We

129: derive optical depths and extinction values from the silicate band and additional ice bands at 

130: 6.0, 6.8 and 15.2\,$\mu$m. In particular the analysis of the CO$_2$ ice band at 15.2\,$\mu$m allows us to 

131: search for evidence for ice processing and constrains whether the absorbing material 

132: is physically linked to the central object or in the foreground. For objects showing the 

133: silicate feature in emission (Category 2), we argue that the emission comes from the surface layer 

134: of accretion disks. Analyzing the dust composition reveals that significant grain growth has already 

135: taken place within the accretion disks, but no clear indications for crystallization are present.

136: We discuss how these observational results can be explained in the picture of a young, and highly active 

137: accretion disk. Finally, a framework is proposed as to how the two categories of FUors can be understood 

138: in a general paradigm of the evolution of young, low-mass stars. Only one object (Parsamian 21) shows

139: PAH emission features. Their shapes, however, are 

140: often seen toward evolved stars and we question the object's status as a FUor and discuss

141: other possible classifications. Two spectra (RNO 1B and L1551 IRS 5) show [Fe II] emission lines

142: which are attributed to hot and dense material located close to the root of an outflow (L1551 IRS 5) or 

143: to shocked material directly within an outflow (RNO 1B).

144: \end{abstract}

145: 

146: %% Keywords should appear after the \end{abstract} command. The uncommented

147: %% example has been keyed in ApJ style. See the instructions to authors

148: %% for the journal to which you are submitting your paper to determine

149: %% what keyword punctuation is appropriate.

150: 

151: \keywords{accretion, accretion disks --- circumstellar matter --- stars: formation --- stars: pre-main sequence ---

152: dust, extinction --- stars: individual (FU Ori)}

153: 

154: %% From the front matter, we move on to the body of the paper.

155: %% In the first two sections, notice the use of the natbib \citep

156: %% and \citet commands to identify citations.  The citations are

157: %% tied to the reference list via symbolic KEYs. The KEY corresponds

158: %% to the KEY in the \bibitem in the reference list below. We have

159: %% chosen the first three characters of the first author's name plus

160: %% the last two numeral of the year of publication as our KEY for

161: %% each reference.

162: 

163: %% Authors who wish to have the most important objects in their paper

164: %% linked in the electronic edition to a data center may do so by tagging

165: %% their objects with \objectname{} or \object{}.  Each macro takes the

166: %% object name as its required argument. The optional, square-bracket 

167: %% argument should be used in cases where the data center identification

168: %% differs from what is to be printed in the paper.  The text appearing 

169: %% in curly braces is what will appear in print in the published paper. 

170: %% If the object name is recognized by the data centers, it will be linked

171: %% in the electronic edition to the object data available at the data centers  

172: %%

173: %% Note that for sources with brackets in their names, e.g. [WEG2004] 14h-090,

174: %% the brackets must be escaped with backslashes when used in the first

175: %% square-bracket argument, for instance, \object[\[WEG2004\] 14h-090]{90}).

176: %%  Otherwise, LaTeX will issue an error. 

177: 

178: 

179: 

180: %______________________________________________________________

181: \section{Introduction}

182: By means of mid-infrared (MIR) spectroscopy gaseous and solid state features have been observed and

183: analyzed in a variety of astronomical environments. While ground-based observations are restricted

184: to certain atmospheric windows (e.g., the N-band around 10\,$\mu$m), spectrographs onboard  

185: space-borne telescopes, such as {\sc ISO} and {\sc Spitzer}, 

186: enable us to study a broad wavelength range from the 

187: near-infrared (NIR) to well beyond 30\,$\mu$m. Those instruments fostered and revolutionized 

188: our understanding in numerous astronomical fields of research. 

189: In particular the star-formation community benefitted from studies based on data from the space telescopes.

190: The composition of dust grains and their evolution (e.g., grain growth and crystallization) in protoplanetary 

191: disks was analyzed to great extend in young, intermediate mass Herbig Ae/Be stars (HAeBes) \citep[e.g.,][]{bouwman2001,meeus2001,acke2004,vanboekel2005}, but also in the young, less massive TTauri 

192: stars \citep[e.g.,][]{forrest2004,kessler2006,sargent2006}. In addition, the ice and dust 

193: features of younger and more deeply embedded objects were studied \citep[e.g.,][]{watson2004} and 

194: the ice inventory of molecular clouds was investigated \citep[e.g.,][]{knez2005,bergin2005}. 

195: %Thanks to the large wavelengths coverage, {\sc Spitzer} spectra are also an ideal

196: %input to the spectral energy distributions of protoplanetary disks and help to model disk structures and

197: %geometries \citep[e.g.][]{furlan2006}. 

198: 

199: One special sub-group of young low-mass objects are FU Orionis objects (FUors), named after the prototype

200: FU Ori. For most objects of this group evidence was found for a tremendous outburst in optical or NIR light

201: over short timescales (months to years) followed by a decline in luminosity typically over several decades. 

202: Other objects were included to this group as they shared peculiar spectral features, e.g, CO bandhead absorption 

203: in the NIR and a changing spectral type with wavelength. Most observational data can be well explained with 

204: highly active accretion disks surrounding these objects, possibly fed with fresh material from

205: a remnant envelope \citep[for a review, see,][]{hartmann1996}. Observations in the MIR are thus ideally suited 

206: to probe the dusty component of the circumstellar material of these objects, either contained

207: in the accretion disks or, additionally, in the envelopes. Recent findings from NIR and MIR 

208: interferometers support the presence of accretion disks several tens of AU in size \citep{quanz2006,malbet2005,millan-gabet2006}. 

209: However, while the SED and the NIR and MIR visibilities of FU Ori itself can be sufficiently explained 

210: by a simple accretion disk model \citep{quanz2006}, \citet{millan-gabet2006} found that accretion disks alone can not reproduce the SED and observed low K-band visibilities of V1057 Cyg, V1515 Cyg and Z CMa-SE simultaneously. 

211: They concluded that additional uncorrelated flux is required possibly arising due to scattering by large dusty envelopes.  

212: 

213: While \citet{lorenzetti2000} presented far-infrared spectroscopy data for 6 FUors observed with  

214: ISO-LWS, a dedicated MIR study of a larger sample of FUors is still missing.

215: \citet{larsson2000} showed the ISO-SWS spectra for 6 FUors for comparison, 

216: but no analysis was carried out. The ISO data for Z CMa were presented in publications related 

217: to Herbig Ae/Be stars by \citet{acke2004} and \citet{elia2004},

218: as Z CMa presumably is a binary system consisting of a Herbig star and a FUor. \citet{white2000} used the ISO observations of L1551 IRS 5 as input for a radiative transfer model of this source. 

219: \citet{hanner1998} discussed ground-based 8-13$\mu$m spectra for four FUors (FU Ori, V1515 Cyg, V1057 Cyg, V1735 Cyg) and fitted a simple dust model to the data to check whether silicate particles from the interstellar medium (ISM) can reproduce the observed features. 

220: \citet{schuetz2005} published additional ground-based data for another four objects in the same wavelength regime.  As three objects (Z CMa, V346 Nor, V883 Ori) showed the 10$\mu$m feature in absorption they inferred the optical depth from fitting an ISM dust model to the spectra. The fourth object (Bran 76, alias BBW 76) was not 

221: analyzed in greater detail. \citet{polomski2005} presented data on RNO 1B, Z CMa, and Parsamian 21 and derived

222: dust temperatures and optical depths. 

223: 

224: To our knowledge, the first MIR spectra of FUors observed with {\sc Spitzer} were presented in 

225: \citet{green2006}. The main focus of this publication was an accurate SED modeling of FU Ori, V1515 Cyg, and V1057 Cyg. Spectral solid state features were not analyzed in greater detail. 

226: The spectrum of V346 Nor was presented for comparison.   

227: In \citet{quanz2006} the first detailed dust composition modeling for the FU Ori spectrum was presented

228: and evidence for grain growth in the accretion disk was found. The {\sc Spitzer} 

229: spectra of RNO 1B/1C were shown in \citet{quanz2007rno}. However, the dust composition was not yet analyzed in detail.

230:  

231: In this paper we compile MIR spectra for 14 FU Orionis objects 

232: observed with ISO and/or {\sc Spitzer}. 

233: As up to now only $\sim$\,20 FUors or FUor candidates are known, 

234: this is the largest sample of these objects analyzed so far in a single MIR study.

235: Part of the data have not been published before. 

236: For objects where the spectra show a sufficient signal-to-noise ratio, 

237: the dust and ice composition of the circumstellar material is investigated. 

238: %______________________________________________________________

239: 

240: %% In a manner similar to \objectname authors can provide links to dataset

241: %% hosted at participating data centers via the \dataset{} command.  The

242: %% second curly bracket argument is printed in the text while the first

243: %% parentheses argument serves as the valid data set identifier.  Large

244: %% lists of data set are best provided in a table (see Table 3 for an example).

245: %% Valid data set identifiers should be obtained from the data center that

246: %% is currently hosting the data.

247: %%

248: %% Note that AASTeX interprets everything between the curly braces in the 

249: %% macro as regular text, so any special characters, e.g. "#" or "_," must be 

250: %%\left[ \left\lbrace %% preceded by a backslash. Otherwise, you will get a LaTeX error when you 

251: %% compile your manuscript.  Special characters do not 

252: %% need to be escaped in the optional, square-bracket argument.

253: 

254: \section{Observations and Data Reduction}

255: The mid-infrared spectra we present in this paper are compiled from the archives of the ISO\footnote{The ISO archive can be accessed via http://www.iso.vilspa.esa.es/ida/} and the 

256: {\sc Spitzer}\footnote{http://ssc.spitzer.caltech.edu/archanaly/archive.html} satellite. 

257: The Short Wavelength Spectrograph (SWS) onboard {\sc ISO} consisted of two nearly independent grating

258: spectrometers with a spectral resolution between 1000 and 2500 (depending on the band and order) and covering a wavelength range from 2.4\,-\,12.0\,$\mu$m and from 12.0\,-\,45.0\,$\mu$m, respectively. 

259: While the field-of-view (FOV) for the shorter wavelength regime was 14$''\times 20''$, the FOV for the 

260: longer wavelength range was 14$''\times 27''$, with the exception of the wavelength range between 28.9\,-\,45.0\,$\mu$m which had a FOV of 20$''\times 33''$.

261: With {\sc ISO/SWS} seven objects classified as FUors were observed between April 1996 and October 1997. For one object (OO Ser) data were taken at five different epochs documenting a decay in luminosity over a few months. Table~\ref{isojournal} summarizes the {\sc ISO}-observations with object names, coordinates, {\sc ISO-SWS} observing mode and scan speed, integration time on target, possible pointing offsets (see below), and the date of the observation. 

262: For the data reduction Highly Processed Data Products (HPDP) or SWS Auto Analysis Results (AAR) were downloaded from the {\sc ISO} archive for speed 1 and 2 or speed 3 and 4 observations, respectively. With the OSIA software package (version 4.0)\footnote{http://sws.ster.kuleuven.ac.be/osia/} the following reduction steps were carried out:  

263: For each object the spectra from the {\sc ISO-SWS} up- and down-scan were flat-fielded and rebinned. 

264: After sigma clipping, the speed 3 and speed 4 spectra were de-fringed. This procedure was not required for the speed 1 and speed 2 data as the HPDP are already de-fringed. Finally, the spectra from the up- and down-scan were combined and rebinned to a spectral resolution of 100. In case the resulting spectrum showed signs of a pointing offset (e.g., aperture jumps) a correction based on the measured beam profiles along the different axes was applied to the raw data and the data reduction was repeated. The applied offsets are listed in Table~\ref{isojournal}.

265: 

266: The {\sc Spitzer} observations are summarized in Table~\ref{spitzerjournal}. 

267: The IRS onboard {\sc Spitzer} offers a short wavelength, low resolution module (SL) covering the wavelength

268: range between 5.2\,-\,14.5\,$\mu$m, and a short wavelength, high resolution module (SH) going from 9.9\,-\,19.6\,$\mu$m. The corresponding slit sizes are $\approx$\,3.6$''\times 136.0''$ 

269: (including both SL orders and an additional bonus segment connecting both orders) and 4.7$''\times 11.3''$, respectively. 

270: For the longer wavelength part a long wavelength, low resolution module (LL) ranging from 

271: 14.0\,-\,38.0\,$\mu$m and a long wavelength, high resolution model (LH) covering the regime from 18.7\,-\,37.2\,$\mu$m are available. The slits sizes are $\approx$\,10.5$''\times 360.0''$ 

272: (including both LL orders and an additional bonus segment connecting both orders) and 11.1$''\times 22.3''$, respectively. Both low resolution modules (SL and LL) offer a spectral resolution between 64\,-\,128 (depending on the wavelength) while the high resolution modules both have $\sim$\,600.

273: All objects listed in Table~\ref{spitzerjournal} were observed with the full wavelength coverage, either with 

274: a combination of SL+LL or with SL+SH+LH.

275: Two objects (HL Tau and XZ Tau) are not classified as FUors but were part of a small {\sc Spitzer}/IRS map including L1551 IRS 5, and data for all three objects could be downloaded simultaneously. Interestingly, XZ Tau is a binary system that recently was found to show EXor-type variations \citep{coffey2004}, i.e., another type of short term eruptions of young stars. Thus, a comparison to the FUor data presented here is reasonable. In addition, the data of HL Tau enable us to compare the FUor spectra to that of a well-studied young star with a remnant envelope and a highly inclined accretion disk seen almost edge-on \citep{close1997}. Part of the {\sc Spitzer} spectrum of HL Tau was already published in \citet{bergin2005}.

276: The object V1647 Ori was observed three times

277: within a period of roughly 5 months to monitor its brightness as it underwent an eruption beginning of 2004.

278: However, to our knowledge, thus far no spectrum was published.

279: 

280: The data reduction process of the {\sc Spitzer} data is described in detail 

281: in \citet{quanz2007rno}. However, we should mention that the error bar for each individual spectral point

282: represents the formal standard deviation from the mean flux of at least two independent measurements 

283: (two telescope nod position and possibly several observation cycles). Also taken into account is the formal

284: error of the spectral response function. For details on the method we refer the reader also to \citet{bouwman2006}.

285: We estimate a relative flux calibration error across a spectral order of $\approx 5$\% and an absolute calibration error between orders or modules of $\approx 10$\%.

286: In particular, for the objects RNO 1B and RNO 1C apparent flux density offsets between the SL and the SH part of the spectra, as well as between the short and long wavelengths part of the high-resolution spectra, are 

287: already discussed in \citet{quanz2007rno}. 

288: For the other objects presented here, the discrepancies in the flux densities between the SL and 

289: the SH part of the spectrum were $<$\,10\% and we matched the longer wavelength part to the shorter regime by multiplying a scalar factor. Only Bran 76 (also known as BBW 76) showed a larger offset 

290: of $\approx$\,15\% as already mentioned by \cite{green2006}. 

291: For Parsamian 21 (HBC 687) we do not show the SH spectrum between 

292: 14 and 20\,$\mu$m as the slit of the spectrograph was not centered on the source and significant flux loss

293: occurred for which we could not correct. An additional correction between the SH and the LH module of the spectrum was

294: required for XZ Tau, HL Tau and L1551 IRS 5 where the flux densities of the LH part had to be scaled down 

295: by 10\,-\,15\%. This offset can be explained by the larger aperture of this module which possibly probed 

296: additional large scale emission from the surroundings of these objects. 

297: 

298: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

299: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

300: 

301: \section{Results}

302: \subsection{General overview}

303: Figures~\ref{iso_short1}-\ref{spitzer_long2} show the complete sample of spectra. For 

304: Bran 76 and the third observation of V1647 Ori only low-resolution {\sc Spitzer} data were available. 

305: To increase the signal to noise, all {\sc Spitzer} LH data and the SH data of V1735 Cyg were smoothed by a factor 

306: of three. Most of the yet remaining spikes in these spectra are not real but rather flux jumps between the 

307: different orders of the spectrographs.

308: As mentioned in the introduction of this paper, parts of the data shown here were already published: \citet{green2006} presented the {\sc Spitzer} data for 

309: FU Ori, V1515 Cyg, V1057 Cyg, Bran 76, and V346 Nor and used disk-envelope models to 

310: explain the SEDs. The spectrum of HL Tau was shown by \citet{bergin2005}. 

311: \citet{schuetz2005} used the ISO data for Z CMa and V346 Nor to compare with their ground-based data.

312: Finally, \citet{larsson2000} showed the ISO/SWS SEDs of the outbursting object OO Ser and used the ISO/SWS data

313: for RNO1B, Z CMa, V1057 Cyg, and V1735 Cyg for comparison. 

314: 

315: Unfortunately, for most objects the quality of the ISO/SWS data is significantly worse than that of the {\sc Spitzer} observations. Even after the data have been rebinned to a spectral resolution of 100, artefacts, i.e. 

316: potential emission features that were only detected either during the up- or the down-scan, remain  

317: in the spectra. 

318: In Figure~\ref{compare_spectra} we show data for four objects that were observed with both {\sc Spitzer} and ISO.  In particular, for the objects with lower flux levels the noise in the SWS data is significant. 

319: The reason for this is the short integration time for most objects which is reflected in the {\tt speed} parameter in Table~\ref{isojournal}. In consequence, the ISO data are mainly used for qualitative statements rather then 

320: for quantitative analyses throughout the rest of the paper. Only the ISO data for Z CMa (with a high flux level) and Reipurth 50 (with long integration time) will be examined in more detail in one of the consecutive sections. 

321: The data on Reipurth 50 is published for the first time. 

322: 

323: 

324: From Figures~\ref{iso_short1}-\ref{spitzer_long2} it becomes clear that for all objects the flux densities

325: increase toward longer wavelengths indicative of warm dusty material surrounding all objects. However, Figures~\ref{iso_short1},~\ref{iso_short2},~\ref{spitzer_short1}, and~\ref{spitzer_short2} show that there are 

326: striking differences within the group of FUors: While some objects show a silicate emission feature in

327: the 10\,$\mu$m region, other objects show deep absorption profiles. Since other spectral features 

328: do further support such a differentiation we will in the following distinguish the objects via the behavior 

329: of their 10\,$\mu$m feature and discuss the two categories separately in the following subsections.

330: A complete overview of the most prominent spectral features between 3 and 16\,$\mu$m is given in Table~\ref{features}. If we count in V883 Ori, which was observed by \citet{schuetz2005}, and disregard 

331: XZ Tau and HL Tau, then 9 FUors show the silicate feature in absorption while 6 FUors show silicate emission.

332: 

333: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

334: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

335: 

336: 

337: \subsection{Objects with 10\,$\mu$m emission}\label{10muem}

338: 

339: \subsubsection{Qualitative analysis of the 10\,$\mu$m region}

340: In Figure~\ref{emission} we compiled all objects showing signs of silicate emission in the 10\,$\mu$m band.

341: To subtract the underlying continuum we fitted a polynomial of first order to the flux at 8 and 13\,$\mu$m\footnote{For Parsamian 21 we had to shift the left point of the fit to 6.6\,$\mu$m due to the special shape of the 

342: spectrum (see below).}. 

343: As we are thus far only interested in a qualitative comparison among the various objects,

344: the exact shape of the continuum is not important and higher oder polynomial fits did not alter the 

345: results relative to each other. For comparison we overplot the silicate emission feature of typical interstellar medium (ISM) dust grains, scaled to the observed spectra \citep[red, dashed lines in Figure~\ref{emission}; ][]{kemper2004}. We note that the spectrum of each object is shown twice: 

346: While the first plot shows the observed spectrum of the source, the second one shows

347: the dereddened spectrum. For the dereddening we compiled optical extinction values from the literature (see caption of Figure~\ref{emission}) and used the extinction law described in \citet{mathis1990} to derive corresponding extinction values for the MIR. It is assumed that all of the extinction is caused by material in the line

348: of sight toward the objects and that no self-shadowing effect (e.g., by an inclined accretion disk with a certain flaring angle) is present. The continuum was fitted for the dereddened spectra separately.

349: It shows that for most objects the extinction is not negligible and that

350: it can have substantial influence on the shape of the silicate feature (see, e.g., V1057 Cyg and V1647 Ori). For a fair comparison it thus seems reasonable to evaluate the dereddened and

351: not the observed spectra. 

352: 

353: At first glance, when comparing the spectra, it is noteworthy that all emission features differ from the typical 

354: shape of the typical ISM dust feature. \citet{green2006} stated that the emission peak for the 

355: first four objects in Figure~\ref{emission} was close to 9.5\,$\mu$m and that the dust features appeared to be 

356: pristine\footnote{It should be mentioned that \citet{green2006}

357: determined the underlying continuum from a fit to the 6\,-\,8\,$\mu$m region and not as we did from fitting a 

358: straight line between 8 and 13\,$\mu$m.}. We find that all features peak longward of 9.7\,$\mu$m \citep[the typical peak position for ISM type dust; ][]{kemper2004} and that they show additional flux excess compared to the ISM feature at even longer wavelengths. This indicates that dust grain processing has already set in. 

359: Furthermore, it should be noted that the spectrum of Parsamian 21 looks significantly different compared to the others. The most prominent characteristic

360: are strong emission bands around 8.2\,$\mu$m and probably also around 11.3\,$\mu$m from 

361: polycyclic aromatic hydrocarbons (PAHs). \citet{polomski2005} already suspected the existence of PAH emission in the spectrum of this source based on ground-based observations, but

362: a firm confirmation was thus far lacking. We discuss the spectrum of Parsamian 21 in more detail below. 

363: 

364: For a better comparison of the emission features, we plotted in Figure~\ref{normalizedemission} the normalized 

365: fluxes of the objects between 8 and 13\,$\mu$m. Following \citet{vanboekel2005}, the normalization was done via

366: \begin{equation}

367: F_{\rm norm}(\lambda)=1+\frac{F_{\rm obs}(\lambda)-F_{\rm cont}(\lambda)}{<F_{\rm cont}>}

368: \end{equation}

369: where $F_{\rm obs}(\lambda)$ is the observed flux, $F_{\rm cont}(\lambda)$ is the continuum flux and $<F_{\rm cont}>$

370: denotes the mean value of the underlying continuum in the considered wavelength regime. This normalization ensures 

371: that the shape of the emission feature is preserved. From Figure~\ref{normalizedemission} it becomes clear that the 

372: emission features of Bran 76, FU Ori, and V1515 Cyg are quite similar in terms of shape and strength.

373: The features of V1057 Cyg and XZ Tau are far less pronounced and much broader\footnote{In contrast to our analysis \citet{green2006} found the feature of V1057 Cyg to be comparable in shape (and thus in dust composition) to those of Bran 76, FU Ori, and V1515 Cyg.}. The emission profiles of V1647 Ori and Parsamian 21 are slightly stronger than those of the other objects and, as mentioned above, the latter object is the 

374: only one showing strong PAH emission bands. The three epochs of data for V1647 Ori allow us to study the

375: variability of this object in the 10\,$\mu$m region over a period of approximately five months. Figure~\ref{normalizedemission} shows that between October 2004 (epoch 1) and mid of March 2005 (epoch 2) the flux level decreased significantly. At the third epoch (end of March 2005) the flux appears to have 

376: slightly increased again (see also Figure~\ref{spitzer_short1}). The overall shape of the feature during the six months period did, however, not change.

377: 

378: To put these results in a broader context with other young objects and to get a first idea

379: on the dust grain properties, we plot in Figure~\ref{fluxratio} the flux ratio at 11.3 and 9.8\,$\mu$m against

380: the computed peak over the continuum in the normalized spectra, i.e., the maxima of equation (1). 

381: This figure also shows the region typically occupied by young TTauri stars and the slightly 

382: more massive Herbig Ae/Be stars 

383: \citep[see, e.g.,  ][]{przygodda2003,vanboekel2003,vanboekel2005,kessler2006}. We find that the objects presented here tend to have in general a relatively

384: weak peak over continuum emission and a flux ratio between 0.8 and 1.0. While the strength of the 

385: peak over the continuum is interpreted as a tracer for grain sizes (with higher peak values denoting smaller grains), the flux ratio is a more general tracer for grain processing, i.e., grain growth as well as crystallization. 

386: This is explained by the fact that not only the growing of grains leads to a broader and flatter silicate 

387: feature between 9.8 and 12.0\,$\mu$m \citep{bouwman2001}, but also the crystallization process introduces

388: distinct emission peaks of forsterite and enstatite longward of 10\,$\mu$m \citep[see, e.g., ][]{vanboekel2005}. 

389: Taking these considerations into account, we find that our sample shows clear evidence for grain growth and dust processing. For a more quantitative analysis of the dust composition we fitted a dust model to the data as 

390: explained in the following subsection.

391: 

392: \subsubsection{Dust composition}

393: In Figure~\ref{dust_model} we fitted an analytical dust model to the spectrum of FU Ori between 7 and 17\,$\mu$m.

394: To equally weight the low- and high-resolution part of the spectrum, we rebinned the data to a spectral resolution 

395: of 100, still sufficient to disentangle contributions from different dust species.

396: Instead of fitting a two component model to the data, with one temperature for the underlying continuum and

397: one temperature for the emission layer, as we did in \citet{quanz2006}, we use a new model 

398: described in Juh\'asz et al. 2007 (to be submitted). In this model the temperatures for 

399: the disk midplane, the disk surface layer and an inner disk rim are fitted by powerlaw distributions instead of  

400: single black body temperatures. In this respect, the model represents more realistically the temperature distribution

401: in a protoplanetary disk. The input parameters for the model were taken from the simple accretion 

402: disk model in \citet{quanz2006}. Table~\ref{temperature_table} summarizes the fitted parameters of the individual

403: temperature components.

404: 

405: The dust model was based on opacities for six dust species (amorphous olivine and pyroxene, crystalline forsterite 

406: and enstatite, silica, and amorphous carbon) in three grain sizes (0.1, 1.5, and 6.0\,$\mu$m). 

407: References for the optical properties of the grains are given in Juh\'asz et al. 2007 (to be submitted) and

408: \citet{quanz2006}. In contrast the analysis presented in \citet{quanz2006}, 

409: this time we included also carbon grains in the fit. However, as shown in Figure~\ref{dust_model} and summarized in Table~\ref{dust_table}, the best fit ($\chi^2$\,=\,45.3) does not require any carbon grains. The derived mass fractions agree well with the results based on the more simple two component dust model used in \citet{quanz2006}. 

410: While a large fraction of big, amorphous grains contribute to a high-temperature continuum, the smaller

411: grains are responsible for the observed emission feature. Still, we emphasize that, according to our analysis, 

412: also in the disk surface layer large grains are present: Comparing the derived composition to that of the interstellar medium \citep{kemper2004,vanboekel2005}, 

413: we find a significantly higher fraction of 1.5\,$\mu$m grains in the 

414: dust of FU Ori. Furthermore, the model does predict a negligible amount of crystalline grains ($<$\,0.01), no silica dust and, as mentioned above, no carbon grains. Certainly, the values for the derived mass fractions should not be taken literally, as, for instance, the temperature structure in the disk and the dust 

415: opacities are not perfectly known. However, the models presented here and in \citet{quanz2006} 

416: agree on two very important aspects: (1) there is 

417: clear evidence for dust coagulation in the spectra of FUors, and (2) there is hardly any

418: contribution from crystalline grains detectable in the data.

419: This lack of crystalline grains was already noticed and discussed in \citet{quanz2006}, but we will

420: reinvestigate this issue below. 

421: 

422: 

423: \subsubsection{The PAH features of Parsamian 21}\label{par21}

424: As its features are significantly different than those of the other FUors, we plot in Figure~\ref{parsamian_pah}

425: the dereddened spectrum of Parsamian 21 and analyze it in more detail. 

426: The vertical lines illustrate, where typically PAH emission bands are located. While prominent PAH emission is present at 6.3, 8.2 and (possibly) 11.3\,$\mu$m, no bands are seen at 7.7, 8.6 and 12.7\,$\mu$m. In addition, the spectrum bears signs of underlying silicate emission and it is difficult to 

427: disentangle in the 11.3\,$\mu$m emission band the possible contributions from crystalline forsterite and 

428: PAHs. 

429: 

430: PAHs are commonly seen in the spectra of HAeBe stars \citep[e.g.,][]{meeus2001,acke2004,sloan2005} and, more

431: recently, PAH emission was also detected in TTauri spectra \citep{geers2006}. However, the 

432: spectrum of Parsamian 21 differs significantly from the PAH spectra of other young objects: it shows 

433: a broad and strong feature at 8.2\,$\mu$m which is barely detected and the main characteristic of 

434: group C objects in \citet{peeters2002}. These objects show no PAH complex 

435: at 7.7\,$\mu$m and 8.6\,$\mu$m but the peculiar broad band at 8.2\,$\mu$m whose origin yet remains unknown \citep{peeters2002}. In section~\ref{par21_discussion} we

436: will critically review the findings presented here and discuss their implications with respect to the 

437: FUor status of Parsamian 21.

438: 

439: 

440: 

441: \subsubsection{Qualitative analyses of the 5\,-\,8\,$\mu$m region}

442: As already noted and explained 

443: by \citet{green2006} FU Ori, Bran 76, V1515 Cyg, and also V1057 Cyg show absorption bands of gaseous

444: H$_2$O at 5.8 and 6.8\,$\mu$m (Figure~\ref{spitzer_short1}) 

445: coming from a collection of rotation-vibration bands.\footnote{Interestingly, already \citet{sato1992} found evidence for water vapor absorption between 1 and 3\,$\mu$m in the spectra of FU Ori, V1515 Cyg, and V1057 Cyg, 

446: but also L1551 IRS 5, V1057 Cyg, and Z CMa.}

447: V1647 Ori, however, shows 

448: evidence for ice absorption bands at 6.0 and 6.85\,$\mu$m (see also Figure~\ref{absorption6}). 

449: We mention already here that this objects shows an additional ice feature at 15.2\,$\mu$m 

450: due to CO$_2$ (Figure~\ref{absorption15}).

451: V1647 Ori is the only object in the sample showing 

452: the 10\,$\mu$m silicate feature in emission accompanied by significant ice absorption bands. A more detailed 

453: analysis of these ice features is given in section~\ref{10muabs}. Finally, as we will discuss in the following section, Parsamian 21 shows clear evidence for a PAH emission feature at 6.2\,$\mu$m.

454: 

455: 

456: 

457: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

458: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

459: 

460: \subsection{Objects with 10\,$\mu$m absorption}\label{10muabs}

461: \subsubsection{Analyses of the 10\,$\mu$m region}

462: In Figure~\ref{absorption} we plot the optical depth in the 10\,$\mu$m region for the objects showing the silicate  band in absorption. To derive the optical depth, we fitted the continuum with a polynomial of first order to the observed spectra between 8 and 13\,$\mu$m, assuming that no absorption due to silicates is present at these wavelengths. Also here the exact shape of the continuum is difficult to estimate and any fit always carries

463: uncertainties. However, even fits with higher order polynomials did not alter the results significantly and the 

464: main conclusions remained valid. Based on the assumed continuum we then computed the optical depth $\tau$ using  

465: \begin{equation}

466: F_{\rm obs}(\lambda)=F_{\rm cont}(\lambda)\cdot e^{-\tau}

467: \end{equation}

468: where $F_{\rm obs}(\lambda)$ is the observed flux and $F_{\rm cont}(\lambda)$ the assumed continuum. 

469: To derive the wavelength position and value of the maximum optical depth, we did the following:

470: As some spectra are quite noisy (e.g., V1735 Cyg, RNO 1B, Z CMa, and Reipurth 50) we fitted the data 

471: between 8.8 and 10.2\,$\mu$m, i.e., the region with the greatest optical depth, with a polynomial of second order

472: (shown as red, dash-dotted lines in Figure~\ref{absorption}).

473: The wavelength position of the maximum optical depth (vertical, red, dash-dotted lines in Figure~\ref{absorption}) 

474: and the corresponding value are then derived from the fit. 

475: From the spectra with a high signal to noise ratio and strong absorption profiles (e.g., the {\sc Spitzer} spectra of L1551 IRS 5 and HL Tau) it becomes clear that the fit matches the shape of the absorption feature quite well and that 

476: this approach seems reasonable. The black, dotted horizontal line in Figure~\ref{absorption} shows the 

477: assumed continuum level. In Table~\ref{extinctiontable} we summarize the derived maximum optical depth for each object $\tau_{\rm silicate}$ and the corresponding

478: wavelength. We give also values for the extinction in the silicate band A$_{\rm silicate}$ and 

479: estimates for the corresponding extinction in the optical A$_{\rm V}$. A$_{\rm silicate}$ was derived via

480: \begin{equation}

481: A_{\rm silicate}=-2.5\,{\rm log_{10}}\,({\rm e}^{-\tau{_{\rm silicate}}})\quad .

482: \end{equation}

483: The errors for

484: $\tau_{\rm silicate}$ and A$_{\rm silicate}$ are based on the 1-$\sigma$ uncertainties in the spectra

485: and the resulting fluctuations of the polynomial fit. 

486: A$_{\rm V}$ and its error are derived from averaging over the extinction models of \citet{mathis1990} and \citet{weingartnerdraine2001} and assuming two different values of R$_{\rm V}$ for each model (R$_{\rm V}$\,=\,3.1 and 5.0 for \citet{mathis1990} and R$_{\rm V}$\,=\,3.0 and 5.5 for \citet{weingartnerdraine2001}). 

487: As already noted by \citet{cohenkuhi1979} one has to be careful when deriving a value for the optical extinction based on the silicate feature, as uncertainties in the underlying continuum and unknown dust compositions 

488: influence the results. Hence, for comparison, Table~\ref{extinctiontable} gives also values for A$_{\rm V}$ found in the literature. One has to keep in mind, though, that also here certain assumptions have been made and 

489: partly different observing techniques were applied potentially leading to discrepancies in the derived

490: values for A$_{\rm V}$.

491: 

492: For V346 Nor and RNO 1C the values for A$_{\rm V}$ agree quite well. Also for Z CMa, Reipurth 50 and L1551 IRS 5

493: the values are in general agreement. Due to the high extinction toward the latter two objects,

494: a good estimate of A$_{\rm V}$ is difficult. We would like to point out, however, that a value of

495: A$_{\rm V}$ $\gtrsim$\,150 mag for L1551 IRS 5 as found by \citet{campbell1988} and mentioned, for instance, by \citet{stocke1988}, \citet{white2000} and \citet{fridlund2002} 

496: appears to be an overestimate. In yet unpublished NIR data taken

497: with the {\sc Omega2000} camera at the Calar Alto Observatory (Spain) we do see a near-infrared (NIR)

498: counterpart to the L1551 IRS 5 system with K$_{\rm S}$\,$\approx$\,9.7 mag. Even if this was purely scattered

499: light from a disk the intrinsic K magnitude of the central object(s) (it might be a multiple system) can not

500: be lower. Thus, the observed K-magnitude is a lower limit for the intrinsic K-luminosity of the source.

501: The above mentioned value for A$_{\rm V}$ would translate into A$_{\rm K}\approx$\,15 mag, and thus 

502: the absolute K-band magnitude of L1551 IRS 5 would be $\approx -11$ mag for an assumed distance of 140 pc. 

503: For young low-mass objects this appears to be definitely too luminous even if accretion luminosities 

504: are taken into account as well. For V1735 Cyg and

505: HL Tau our derived values for A$_{\rm V}$ are lower than those found in the literature. However, for 

506: V1735 Cyg no errors are given for the value found in the literature. 

507: For HL Tau the optical extinction derived from the radiative transfer 

508: model by \citet{menshchikov1999} may suffer from insufficient information about the source geometry or dust opacities which can easily lead to large uncertainties in the computed figures. 

509: The comparison for these objects clearly shows that, in particular for embedded objects, it is difficult to derive 

510: consistent values for A$_{\rm V}$ if different techniques are applied.

511: 

512: Finally, RNO 1B also shows a discrepancy between our value

513: for A$_{\rm V}$ and that given by \citet{staudeneckel1991}. In this case, however, the 

514: silicate absorption feature bears evidence for additional superposed emission as will be described in the 

515: next paragraph. Thus, we attribute at least part of the missing optical depth to an underlying

516: silicate emission feature. 

517: 

518: To measure not only the depth of the absorption profile, but to get also an idea of the

519: dust composition responsible for the observed absorption, we analyzed the shape of the 

520: silicate feature. In Figure~\ref{abs_profiles} we show the optical depths computed or measured for different 

521: dust grain compositions. Depending on the composition the wavelengths of the maximum optical depth

522: changes. The most extreme cases are amorphous olivine grains with a size of 0.1\,$\mu$m peaking 

523: at a wavelength slightly longward of 9.7\,$\mu$m, and amorphous pyroxene grains with the same size 

524: peaking around 9.2\,$\mu$m wavelength. As mentioned above, in Table~\ref{extinctiontable} 

525: we already give the wavelengths

526: of the maximum optical depths observed toward our sources and it shows that there is indeed a broad 

527: range of values, indicating different dust compositions. 

528: 

529: In Figure~\ref{absorption2} we plot again the observed optical depth of our objects 

530: as shown in Figure~\ref{absorption}, but now we overplot one or two of the absorption profiles presented

531: in Figure~\ref{abs_profiles}, scaled to the maximum observed optical depth. Whether this scaling

532: is appropriate is difficult to determine but it shows that most observed features agree

533: reasonably well with one or two of the reference profiles. Three objects (L1551 IRS 5, V346 Nor and V1735 Cyg) 

534: show additional absorption longward of 11\,$\mu$m when compared to the dust features, but at least part

535: of this absorption can be attributed to H$_2$O libration bands \citep[e.g., ][]{pontoppidan2005}.

536: V1735 Cyg, RNO 1C and Z CMa seem to agree best with 

537: the dust model computed by \citet{draine2003} based on the particle size distribution from \citet{weingartnerdraine2001}. 

538: V346 Nor requires shortward of 10\,$\mu$m a mixture of 

539: the same model and the dust composition observed toward the galactic center \citep{kemper2004}.

540: Longward of 10\,$\mu$m the former model alone fits better. For Reipurth 50 a combination 

541: of both models explains the data well over the whole wavelength regime with the \citet{draine2003} model

542: fitting the short and the \citet{kemper2004} model fitting the long wavelength part.

543: L1551 IRS 5 is best fitted with a mixture of small, amorphous pyroxene grains and the model from \citet{draine2003}.

544: The profile of HL Tau agrees extremely well with the galactic center dust profile from \citet{kemper2004}.

545: To account for a small shift toward shorter wavelengths a little bit more amorphous pyroxene seems to be required 

546: or small uncertainties in the position of the peak absorption from the continuum fit can explain this shift. 

547: Finally, the absorption profile of RNO 1B is special in two ways: First, the 

548: short wavelength range is best fitted solely with small, amorphous 

549: pyroxene grains. And secondly, between 10 and 11.5\,$\mu$m

550: the profile shows a significantly lower optical depth than any of the reference profiles. In combination with 

551: our findings for the visual extinction (see above) this suggests that RNO 1B shows silicate emission 

552: superposed on the absorption feature. Interestingly, the apparent decrease in optical depth longward of 10\,$\mu$m 

553: indicates that at these wavelengths the contribution of the emission is relatively larger compared to the 

554: other wavelengths in the silicate band. This in turn means that the emission profile does not

555: have the typical shape of the ISM dust feature but a broader and 

556: more evolved profile like the other FUors 

557: emission profiles shown in Figure~\ref{emission}. 

558: 

559: 

560: \subsubsection{The 15.2\,$\mu$m CO$_2$ ice feature}

561: The shape of the bending mode profile of CO$_2$ ice around 15.2\,$\mu$m does not only allow 

562: a detailed analysis of the involved ice inventory, but it enables us to derive 

563: information about potential ice processing due to heating effects \citep[e.g.,][]{pontoppidan2005}. 

564: In Figure~\ref{absorption15_comparison} we present two reference cases which we will use in the following

565: as a benchmark for comparison with the FUor spectra. The first object in Figure~\ref{absorption15_comparison}

566: (CK2) is a highly extincted background star behind the Serpens dark cloud \citep{knez2005}, and the second object (HH46) is an embedded low-mass protostar \citep{boogert2004}. 

567: 

568: While the spectrum of CK2 probes mainly the ice composition of the 

569: intervening dark cloud, the spectrum of HH46 bears information about the immediate environment of the young 

570: star. Apart from being slightly broader, the spectrum of HH46 shows a double-peaked sub-structure which 

571: is caused by crystallization and effective segregation of the  CO$_2$ and H$_2$O ice involved in the absorption.

572: These processes take place when ice mixtures with concentrations of  CO$_2$/H$_2$O$\ge$1 are heated \citep{boogert2004}. In space, this phase transition from amorphous to crystalline ice is expected to occur between 50-90 K and thus at higher temperatures than typically found in cold, dark molecular clouds. Thus, the 

573: 15.2\,$\mu$m ice feature of HH46 shows that (at least) part of the ice must already have been heated by

574: the embedded protostar. In fact, \citet{boogert2004} fit the feature with a two-component ice model

575: based on laboratory spectra with one component being highly processed polar ice with a laboratory 

576: temperature of T$_{\rm lab}$\,=\,125 K, and the other component being an H$_2$O-rich, CH$_3$OH-deficient cold ice

577: with T$_{\rm lab}$\,=\,10 K.\footnote{The presence of CH$_3$OH in the ice feature can potentially be traced by 

578: the shape of the long-wavelength wing of the CO$_2$ profile showing additional absorption if CH$_3$OH 

579: is present in higher abundances.}

580: The spectrum of CK2 was fitted by \citet{knez2005} solely with cold ice

581: components. They used a polar mixture of H$_2$O:CO$_2$=1:1 and H$_2$O:CO$_2$=10:1 at 10 K with a ratio of 2:1 and 

582: an additional apolar component of CO:N$_2$:CO$_2$=100:50:20 at 30 K. The overall polar fraction was

583: assumed to be 78\,\%.

584: 

585: In Figure~\ref{absorption15} we present the 15.2\,$\mu$m features observed toward our FUor sample. 

586: To derive the optical depths we fitted the continuum with a straight line fixed around 14.65 and 16.3\,$\mu$m.  

587: For each object we overplot either the (scaled) spectrum of HH46 or CK2 depending on the shape of the profile. While

588: the profiles of V1647 Ori and V346 Nor agree better with the profile of CK2 (the spectrum representing

589: unprocessed ice), L1551 IRS 5, RNO 1B and 1C, and HL Tau show evidence for a double-peaked sub-structure 

590: and thus heating effects and processed ice. The comparatively bad quality of the V1735 Cyg spectrum does not 

591: allow a solid comparison to either reference spectrum. It is noteworthy that the spectrum of V1647 Ori 

592: is almost an exact copy of CK2 indicating that the ice composition is mostly identical. Based on the 

593: fitted reference spectrum we computed also the optical depth $\tau_{15.2\,\mu{\rm m}}$ 

594: for each object and summarized the results in 

595: Table~\ref{iceextinctiontable}. The errors are based on the 1-$\sigma$ uncertainties in the observed spectra.

596: 

597: 

598: \subsubsection{The 6.0 and 6.8\,$\mu$m ice features}

599: Although frequently observed toward high- and low-mass sources,

600: the two well-known ice features at 6.0 and 6.8\,$\mu$m are 

601: quite complex and difficult to interpret. Certainly, a large fraction of the 

602: optical depth of the 6.0\,$\mu$m band can be attributed to H$_2$O ice, but also other species might contribute 

603: to this absorption feature \citep{keane2001}. For instance, slightly shortward, at roughly 5.85\,$\mu$m, an additional absorption shoulder is sometimes superposed \citep[e.g.,][]{pontoppidan2005,keane2001} for which formaldehyde (H$_2$CO) and formic acid (HCOOH) are theoretical candidates. Without any additional information

604: (e.g., the 3.08\,$\mu$m band of H$_2$O or the 3.47\,$\mu$m band of H$_2$CO) it is thus difficult to 

605: determine the true water ice content in the 6.0\,$\mu$m band.

606: 

607: An additional absorption feature at 6.85\,$\mu$m is often observed toward protostars \citep[e.g.,][]{keane2001}, 

608: but also toward the extincted background star CK2 \citep{knez2005} and the edge-on disk CRBR 2422.8-3423 \citep{pontoppidan2005}. Although a final identification of this band has yet to be provided,

609: NH$_4^+$ seems to be one of the most promising candidates \citep{schutte2003,pontoppidan2005}. 

610: However, \cite{dishoeck2004} for example, mention also methanol (CH$_3$OH) as a potential carrier of this

611: absorption band. 

612: 

613: Given all these uncertainties we restrict ourselves in this section to the computation

614: of the optical depth of both of the above mentioned absorption bands. Like in the previous section

615: we assumed a straight line for the continuum anchored at 5.4 and 7.6\,$\mu$m.

616: To derive the optical depths we then fitted a polynomial

617: of fourth order to both absorption dips to eliminate the noise in the spectra. 

618: Figure~\ref{absorption6}

619: shows the observed spectra between 5.5 and 7.5\,$\mu$m on an optical depth scale (black lines) with the

620: resulting fits overplotted (red, dashed-dotted lines). The computed optical depth for each band and each 

621: object is given in Table~\ref{iceextinctiontable}. Like in the previous section the errors are based on the 1-$\sigma$ uncertainties in the observed spectra. 

622: 

623: 

624: \subsection{The fading of OO Ser and V1647 Ori}

625: For two objects (OO Ser and V1647 Ori) we have multi-epoch data and can derive some conclusions

626: on the variability of these objects.

627: As illustrated in Figure~\ref{iso_long2}, we fitted a straight line to the data of OO Ser between 15 and 30\,$\mu$m 

628: to estimate the decay in flux density observed over the 5 epochs (see, Table\ref{isojournal}).

629: The wavelength range was chosen as in this regime all spectra are still relatively clear of artefacts and spikes partly seen at longer wavelengths. In Table~\ref{ooserflux} we summarize the flux density level at 

630: 20 and 30\,$\mu$m at each epoch and give also the slope measured between 15 and 30\,$\mu$m. The 

631: errors are derived from the goodness of the fit to the data. Between the 

632: first observations on April 14, 1996, and the last observations, carried out September 22, 1997, the flux densities 

633: decreased to roughly 50\% of the initial values. These data demonstrate that OO Ser faded rapidly over relatively short timescales and that it might be an intermediate object between a typical FUor and an EXor having

634: fading timescales of several decades and months, respectively. Based on photometric 

635: monitoring at infrared wavelengths, \citet{kospal2007} came to a similar conclusion and predicted that 

636: OO Ser should return to its pre-outburst luminosity not before 2011. 

637: 

638: In Table~\ref{ooserflux} we also summarize the flux densities of V1647 Ori observed at 8, 20 and 30\,$\mu$m at three different epochs. The errors are taken directly 

639:  from the {\sc Spitzer} spectra. 

640: It shows that also this object faded significantly over  a period of less than 5 months between October 2004 and March 2005. In addition, short time variations 

641: in the flux levels seem to be present, as between the two epochs in March 2005 the object became slightly brighter again. In consequence, these data 

642: support the assumption that the outburst of V1647 Ori may also be intermediate between FUor- and EXor-type events  similar to OO Ser \citep{muzerolle2005,acosta-pulido2007}. 

643: 

644: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

645: \subsection{Additional emission lines}

646: For completeness we show in Figure~\ref{lines} absorption and emission lines identified in the high-resolution regime

647: of the {\sc Spitzer} spectra, part of which are difficult to identify in Figures~\ref{spitzer_long1} and~\ref{spitzer_long2}. Already \citet{green2006} noted the [S III] emission lines at 18.7 and 33.4\,$\mu$m 

648: in the spectrum of V1515 Cyg and argued that they originate from extended emission in the region and not

649: from the object itself. The spectra of RNO 1B and RNO 1C show evidence for H$_2$ quadrupole emission 

650: around 17\,$\mu$m and additional H$_2$ lines in the low-resolution part of the spectrum at shorter wavelengths. In \citet{quanz2007rno} we analyzed these emission lines in detail and concluded that they are related to shocks within a molecular outflow powered by the nearby embedded object IRAS 00338+6312. [Fe II] lines around 17.9\,$\mu$m are

651: present in the spectra of L1551 IRS 5 and RNO 1B, and L1551 IRS 5 shows also 

652: the [Fe II] line near 26.0\,$\mu$m. The [Fe II] lines of L1551 IRS 5 were already detected in the ISO/SWS 

653: spectrum of this source and attributed to hot and dense material located close to the root of the outflow \citep{white2000}. The line intensities did not fit the predictions from shock models. Concerning the [Fe II]

654: line in the spectrum of RNO 1B, it seems likely that it arises in the outflow shocks that also excite the 

655: H$_2$ emission lines discussed in \citet{quanz2007rno}. 

656: Finally, V1057 Cyg shows gaseous CO$_2$ absorption slightly shortward of 15.0\,$\mu$m.

657: 

658: 

659: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

660: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

661: 

662: 

663: \section{Discussion}

664: \subsection{Two categories of FUors}

665: The results presented here suggest that the sample of FUors can be  

666: divided into two categories based on the observational appearance of the 10\,$\mu$m silicate feature.

667: We decided to call the absorption 

668: feature objects "Category 1"-FUors (9 objects in Table~\ref{features}) 

669: and the emission feature objects "Category 2"-FUors (6 objects in Table~\ref{features}). 

670: In the following we discuss the properties of the two 

671: categories in more detail.

672: 

673: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

674: 

675: 

676: \subsubsection{Category 1 FUors: silicate and ice absorption features}

677: In general it is possible to observe the silicate feature in absorption if the 

678: circumstellar disk surrounding a young object 

679: is seen close to edge-on. \citet{menshchikovhenning1997} showed that typically for a disk with an opening angle of 20$^{\circ}$ between the upper and lower disk surface the silicate feature appears 

680: in absorption only if the disk inclination is $\lesssim 10^{\circ}$ from edge-on.

681: Given the amount of objects showing silicate absorption compared to 

682: the number of objects with silicate emission and assuming a random distribution of the orientation

683: of the accretion disks, it is unlikely that all Category 1 FUors are seen edge-on. 

684: Rather, 

685: these objects are still more deeply embedded in their molecular envelopes covering a larger 

686: solid angle than the edge-on disk alone. These envelopes cause the ice and dust absorption features. 

687: 

688: Figure~\ref{absorption2} illustrates that all but one 

689: silicate absorption bands are fitted best either with 

690: the silicate composition observed toward the Galactic center by \citet{kemper2004} 

691: or with the astronomical silicates from \citet{weingartnerdraine2001} and \citet{draine2003}\footnote{L1551 IRS 5 requires a fraction of additional amorphous pyroxene grains.}. This finding shows that all absorption features can be fitted with small, amorphous silicates and hence the extinction is caused by pristine and not processed dust. 

692: Only the spectrum of RNO 1B is difficult to fit with any ISM dust compostition, but, as already mentioned, there are strong hints that the absorption feature is altered by a superposed emission feature. 

693: 

694: In Figure~\ref{absorption_correlation} we show the observed optical depths of the ices and the silicate

695: feature in a scatter plot to search for any correlation. Although the range of optical depths we probe here

696: is limited, it seems that at least for the ices there seem to be correlations (left and right plot in 

697: Figure~\ref{absorption_correlation}). Fitting a straight line to the data we find:

698: \begin{equation}

699: \tau_{15.2\,\mu{\rm m}}=(0.349\pm0.037)\cdot\tau_{6.0\,\mu{\rm m}}+(0.040\pm0.009)

700: \end{equation}

701: \begin{equation}

702: \tau_{6.8\,\mu{\rm m}}=(0.858\pm0.067)\cdot\tau_{6.0\,\mu{\rm m}}+(0.007\pm0.016)

703: \end{equation}

704: This implies that the physical and chemical conditions within the envelopes and clouds 

705: causing the absorption are similar. A more detailed analysis, e.g., the determination of different 

706: ice abundances relative to water ice, is beyond the scope of this work. For such an analysis the

707: water ice feature around 3\,$\mu$m is required as it suffers less from additional contributions of other 

708: ice species compare to the 6\,$\mu$m feature. 

709: 

710: For a correlation between the optical depths of the silicates and ices the situation is slightly different because, as we have already pointed out, 

711: the silicate feature can be influenced by superposed emission. This shows nicely in the middle plot in  Figure~\ref{absorption_correlation}, where RNO 1B is shifted with respect to the other objects. 

712: 

713: Finally, it is interesting to note that the objects showing evidence for ice processing in 

714: Figure~\ref{absorption15} tend to show higher optical depths in the ice features than the other

715: sources. This might suggest that the extinction for the latter objects (V346 Nor, V1735 Cyg) 

716: might be caused by ices somewhere in the line of sight to the source, 

717: rather than by material related to the young star. This might also explain why V1647 Ori shows

718: weak silicate emission, but ice absorption: the extinction is caused by cold foreground material,

719: reflected also in the observed high value for A$_{\rm V}$ (see caption Figure~\ref{emission}).

720: Ground-based observations find the spectrum to be flat in the 10\,$\mu$m regime \citep{abraham2006}.

721: However, the sensitivity of {\sc Spitzer} allows the detection of a weak feature 

722: resulting from an extinguished but intrinsically strong silicate emission band.

723: 

724: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

725: 

726: \subsubsection{Category 2 FUors: the silicate emission feature and its dust composition}

727: %\subsubsection{No crystalline grains?}

728: As presented in Figure~\ref{emission} and described in the related section, the emission profiles bear

729: evidence for dust grain processing. Even after the correction for apparent interstellar extinction, the 

730: shape of the silicate profile differs from that of typical ISM dust. Like for TTauri stars and the 

731: slightly more massive HAeBes, the origin of the emission feature in the 

732: spectra of the FUors is the heated surface layer of the accretion disk. The apparent grain

733: processing is believed to be only possible in circumstellar disks and not in he less dense circumstellar envelopes.

734: While the emission layer for TTauri stars and HAeBes is mainly heated be the central stellar object, for 

735: FUors the hot inner parts of the accretion disk itself can act as the main illuminating source

736: \citep[e.g.,][]{lachaume2004}. Due to the high accretion rates of FUors those inner regions are extremely 

737: hot and account for a significant fraction of the total flux even at optical wavelengths \citep{quanz2006}.

738: Furthermore, accretion disk models with a flared geometry were not only able to explain the SEDs 

739: of FU Ori and Bran 76 \citep{green2006}, but they could also reproduce interferometric observations in the 

740: NIR \citep{malbet2005} and MIR \citep{quanz2006}. V1057 Cyg and V1515 Cyg show more emission 

741: at longer infrared wavelengths than the previous objects, which can be accounted for assuming a remnant infalling 

742: envelope in addition to the accretion disks. This is supported by the results from K-band interferometry 

743: \citet{millan-gabet2006}, where envelopes are required to explain the low NIR visibilities. 

744: 

745: Concerning the dust composition it is interesting to note that, given the low peak-over-continuum ratio

746: illustrated in Figure~\ref{fluxratio}, grain growth must 

747: have already set in. This is supported by the results of our dust model fit.

748: Furthermore, the spectra do not show  

749: evidence for crystalline dust particles. The mass fraction of crystalline particles in 

750: the dust model computed for the spectrum of FU Ori was negligible (see, Table~\ref{dust_table}). 

751: Since strength and shape of the silicate feature of FU Ori is

752: comparable to those of Bran 76 and V1515 Cyg (Figure\ref{normalizedemission}), the dust composition

753: in all objects is similar. The spectrum of V1057 Ori is even broader and less pronounced than the other 

754: spectra, indicating even larger grains in the disk surface layer. The intrinsic feature of V1647 Ori 

755: is stronger than those of the other FUors, but no prominent signs of crystalline

756: silicates are present either.\footnote{The spectrum of Parsamian 21 does show a

757: prominent feature at 11.3\,$\mu$m but we attribute most of the related flux to 

758: PAH emission and not to crystalline silicates (see section~\ref{par21}).} 

759: 

760: As already mentioned in \citet{quanz2006}, there are several reasons 

761: for which stronger crystalline features could have been expected to be detected. 

762: The high disk accretion rates should ease the detection of crystalline particles in two ways:

763: (1) high accretion rates lead to high disk temperatures \citep{bell1997} which 

764: in turn should increase the amount of crystalline particles produced by annealing processes at T$\ge$800 K.

765: (2) an increase in the accretion rate should also increase 

766: the radial and vertical mixing in the disk \citep{gail2001}

767: transporting the crystalline particles farther out and to the disk surface, where they can 

768: be detected by means of MIR spectroscopy. However, neither in the innermost disk regions probed with 

769: MIR interferometry \citep{quanz2006} nor in the {\sc Spitzer} spectra presented here in 

770: Figure~\ref{emission} or at longer wavelengths in Figure~\ref{spitzer_long1} we see any striking 

771: evidence for crystallinity. This means, that either those grains do not exist in large amount in these

772: disks or they are somehow hidden.

773: 

774: The disks of FUor objects are different from those of TTauri or Herbig star

775: disks in the sense that in the radially innermost regions of FUor disks the luminosity

776: is accretion dominated (TTauri and Herbig star disks are irradiation

777: dominated everywhere). At larger radii, also FUor disks are irradiation

778: dominated, albeit the main heating source may not be the central star itself, but

779: rather the hot inner disk regions close to the star where accretion dominates \citep[e.g.,][]{lachaume2004}. 

780: Disk regions that are irradiation dominated

781: have a surface layer that is warmer than the underlying disk interior; such an

782: "inverted" temperature profile causes dust features to appear in emission.  In

783: the accretion dominated regions, the main heating source is the release of

784: gravitational energy in the disk midplane. Here, the disk interior is at least

785: as hot (and likely hotter) than the disk surface, causing dust features to be

786: effectively hidden, or even to appear in absorption\footnote{The gas absorption features

787: between 5 and 8\,$\mu$m described in 3.2.4 likely originate in this hot disk region.}. 

788: Thus, it is conceivable

789: that there are significant amounts of crystalline silicates present in the hot

790: inner disk regions of our targets, even if they do not show up prominently in

791: the spectra.

792: 

793: But what about crystalline silicates further from the central object, in

794: irradiation dominated regions?  Given the high accretion luminosity in our

795: objects it seems unavoidable that large amounts of crystalline silicates are

796: produced in the innermost disk regions. In addition to this, crystalline silicates may be produced \emph{in

797:   situ} at large radii in shocks \citep{harker2002} or electric

798: % bibtex entries are at the end of this file!

799: discharges \citep{pilipp1998,desch2000}. If the disk is

800: well mixed in the vertical direction, the crystalline silicates should be

801: present in the surface layer of the disk at large ($\gtrsim$\,few~AU) radii,

802: and show up in emission, which they evidently do \emph{not} (at least, to a

803: much lesser extent than in many TTauri and Herbig Ae/Be star disks, which are

804: more evolved). This inevitably leads to the conclusion that

805: crystalline silicates are \emph{not} abundant in the surface layer of

806: FUor disks at radii of more than a few AU from the central object,

807: \emph{contrary to expectations}. 

808: 

809: At this point, we have no proper explanation for this observation. In order to

810: animate the discussion, we will post the following idea, but stress that at

811: this point this is mere speculation. The crystalline silicates are not present

812: in the surface layer at larger radii, but may be present in the disk interior,

813: where they are expected to be formed by virtue of the processes mentioned

814: before. Thus, we need to explain why the crystalline silicates do not get

815: mixed in the vertical direction all the way to the disk surface. One

816: possibility is that the crystalline particles are somehow less well coupled to

817: the gas, possibly because they are more compact (or less "fluffy") than the

818: unprocessed dust particles. This may cause the selective settling of

819: crystalline dust.

820: %(this, however, contrasts observations of interplanetary dust particles where

821: %crystalline particles are found as small inclusions of otherwise mostly

822: %amorphous, fluffy particles).

823: A second possibility is the following. The accretion rate in FUor objects is

824: high, suggesting that the disk is still being supplied with fresh material

825: from the maternal envelope. This material, which is thought to contain

826: exclusively small, amorphous dust particles, will in part "rain" onto the disk

827: surface at larger radii. If the rate at which this material falls onto the

828: disk is higher than the rate at which it is mixed through the disk in the

829: vertical direction, the disk may be covered by a "blanket" of pristine

830: material, effectively hiding the more processed material present in the disk

831: interior. It is unclear whether this idea can be harmonized with the

832: observation that the dust we see at the disk surface has undergone significant

833: grain growth. This issue clearly calls for a much more detailed and

834: quantitative investigation, which is beyond the scope of the present work.

835: 

836: 

837: 

838: 

839: \subsection{Unifying the two categories of FUors: An evolutionary sequence}

840: The spectral properties of the sample presented here draw a rather

841: inhomogeneous picture of the group of FU Orionis objects. However, as most of them 

842: are convincingly classified as FUors, they do share some common observational and physical

843: properties. In the following, we present an idea, how the apparent differences can be 

844: explained within a unified paradigm of FU Orionis objects. 

845: 

846: Already 20 years ago \citet{herbig1977} and \citet{hartmannkenyon1985} suggested that each 

847: FUor might undergo several outbursts. More recent theoretical models show that

848: gravitational instabilities in the accretion disk, driven by continuing 

849: infall from a remnant envelope, can account for intense bursts of high accretion rates which 

850: are intersecting more prolonged, quiescent periods of low accretion \citep{vorobyovbasu2006,boley2006}.

851: After several of these outbursts have occurred within a time span of several 10$^5$ years, 

852: the envelope, which is the trigger of the disk instabilities, vanishes, and the object 

853: enters finally a state of permanently low accretion. If one relates this final phase to the

854: classical TTauri phase of low-mass YSOs, then FUors are younger than classical TTauri stars 

855: as already pointed out by \citet{weintraub1991} and \citet{sandell2001}.

856: 

857: \citet{kessler2005} analyzed the 10\,$\mu$m silicate feature of 34 young stars and confirmed that 

858: the evolution of the feature of low-mass stars and the overall SEDs is similar to that of intermediate-mass stars:

859: embedded objects, showing a pristine silicate band in absorption, evolve into objects showing

860: a combination of silicate absorption and emission, and finally pure emission features appear, where 

861: dust processing leads to a broad range of shapes and strength. Since all these stages are 

862: represented in our sample, we believe that, indeed, the FUor-phase is the link between the 

863: embedded Class I objects and the more evolved Class II objects. In this context, the 

864: objects showing silicate absorption are younger and at the beginning of the period where 

865: subsequently numerous FUor outbursts will occur, while the objects showing pure emission features

866: are more evolved and possibly near the end of their FUor period. As illustrated in Figure~\ref{spectral-indices}

867: the spectral indices of the emission objects are indistinguishable from Class II objects in Taurus \citep{furlan2006}. In particular, the objects FU Ori and Bran 76 appear to be the most evolved

868: objects, because, as mentioned above, their SEDs do not require the presence of a large remnant envelope.

869: A nice example of an intermediate object is RNO 1B, where the dominating absorption feature is 

870: altered by the underlying silicate emission from the accretion disk. Figure~\ref{categories} illustrates with simple 

871: sketches the main features of the the two categories of FUors. It should be mentioned that 

872: theoretically a Class II FUor seen close to edge-on might be interpreted as a Class I FUor. As outlined

873: in 4.1.1 it is, however, statistically very unlikely that all of the Class I objects shown here are 

874: Class II FUors "in disguise".

875: 

876: One question arises, though. If FUors are younger than classical TTauri stars but do already show clear

877: evidence for dust coagulation, then why do some classical TTauri stars show silicate emission features 

878: consisting purely of pristine and unprocessed dust? One possible explanation is linked to the

879: high accretion rates of the FUors. It is believed that high disk accretion rates are related to a higher

880: degree of turbulence and thus mixing in the disk. As a result, larger particles are coupled to the turbulent

881: gas and mixed throughout the disk and to the disk surface, where they can be observed. Once the

882: accretion rate, and hence the amount of mixing, drops, larger particles tend to settle to the

883: disk midplane much faster than small particles, which remain in the disk surface layer. Thus, although 

884: large particles are present in the disk of a TTauri star, they might not reach the disk surface, and only

885: the small grains produce a silicate emission feature. Again, as for the issue concerning the

886: lack of crystalline grains, a more detailed theoretical modeling is certainly

887: required to see to what extend this qualitative explanation is valid. 

888: 

889: 

890: \subsection{Parsamian 21 - intermediate mass FUor or Post-AGB star?}\label{par21_discussion}

891: The strong PAH emission of Parsamian 21 is unique among the sample of FUors presented here, and the

892: analysis rise serious doubts whether Parsamian 21 is indeed a member of this group. The initial classification 

893: was made by \citet{staudeneckel1992} based on optical spectroscopy. These authors derived a spectral 

894: type of F5Iab, found a prominent P-Cygni profile in H$\alpha$ and observed shock-induced emission in [O I],

895: [N II] and [S II]. They also mention the detection of Li in their spectra, which, if present, does not

896: seem to be significantly above the noise level in the data. 

897: New observations with higher spectral resolution and higher signal-to-noise would certainly be 

898: eligible. Although some of these features are commonly observed in FUors, they are not unique to this group 

899: of objects, as also HAeBes show, for instance, P-Cygni profiles in H$\alpha$. 

900: In addition, two of the main properties of FUors, an outburst in optical light and CO bandhead 

901: absorption profiles in the NIR \citep[e.g.,][]{hartmann1996}, have not yet been observed for Parsamian 21. 

902: Looking at the immediate surroundings of Parsamian 21 in optical data from the Digitized Sky Survey\footnote{http://archive.eso.org/dss/dss} it is also

903: striking that no dark cloud complex is linked to Parsamian 21 and hence no connection to a larger

904: star-forming region is present.

905: 

906: To excite the PAH emission, sufficiently strong UV radiation is required. However, most 

907: FUors have a later spectral type than Parsamian 21 and lack, unlike TTauri stars, in general UV continuum 

908: excess emission \citep{hartmann1996}.

909: %Only HD 135344 shows also a peculiar PAH spectrum in the 7-9\,$\mu$m range, probably caused by

910: %a blend of two peaks at 8.0 and 8.2\,$\mu$m, respectively \citep{sloan2005}. The spectrum of Parsamian 21

911: %fits best to the objects defining group C$'$ in \citet{peeters2002}. However, only two objects belong to this

912: %group in that paper and both are Post-AGB stars (CRL 2688, IRAS 13416-6243). Like Parsamian 21

913: %they show no PAH complex at 7.7\,$\mu$m and no 8.6\,$\mu$m but the peculiar broad band at 8.2\,$\mu$m. However,

914: %in contrast to Parsamian 21, their feature at 11.2\,$\mu$m is extremely weak. 

915: %While most PAH emission bands are well characterized by their C-H or C-C modes, the identification of the

916: %8.2\,$\mu$m band yet remains unknown \citep{peeters2002}. In section~\ref{par21_discussion} we

917: %will critically review the findings presented here and discuss their implications with respect to the 

918: %FUor status of Parsamian 21.

919: As mentioned above, the positions of the PAH emission bands are untypical

920: for young stars and so far only 7 objects are known to show characteristics of PAH class C objects \citep{sloan2007} as defined by \citet{peeters2002}. Among these 7 objects only two presumably young stars (SU Aur and HD 135344) 

921: show the peculiar PAH band close to 8.2\,$\mu$m. All other PAH class C objects are

922: post-AGB stars or red giants \citep{sloan2007,peeters2002}. 

923: 

924: In addition, one of the Post-AGB stars from 

925: \citet{peeters2002} and also HD 56126, a Post-AGB star with PAH and silicate emission \citep{hony2003}, have

926: the same spectral type as Parsamian 21. However, we should mention 

927: that \citet{dibai1969} and \citet{the1994} give a 

928: spectral type of A5Ve$\alpha$ for Parsamian 21. This is too early for a typical 

929: FU Orionis object, but could also explain the PAH emission.

930: Finally, the bipolar emission knots seen in H$\alpha$ and [N II] in the optical spectrum of Parsamian 21 \citep{staudeneckel1992}, are also found in the bipolar outflows of evolved stars and "pre-planetary" nebulae

931: \citep[e.g., in the "Butterfly Nebula", ][]{solf2000}. They are thus not only seen in young objects.

932: 

933: Based on these findings, we believe that the FUor status of Parsamian 21 is at least very questionable.

934: Either this object represents an intermediate mass FUor object, suggesting that also stars in this 

935: mass regime undergo phases of enhanced accretion, or, and this appears even more likely, Parsamian 21

936: is not even a young object but an evolved star, sharing typical properties with Post-AGB stars.

937: 

938: \section{Conclusions and future prospects}

939: Our conclusions can be summarized as follows:

940: \begin{itemize}

941: \item{We presented the first coherent space based spectroscopic MIR study 

942: of 14 FUors observed with the {\sc Spitzer Space Telescope} or the

943: {\sc Infrared Space Observatory}. The sample includes roughly two thirds of the known FUors or FUor candidates.}

944: \item{Based on the appearance of the 10\,$\mu$m silicate feature, we divided the sample into two categories: Category 1 objects

945: show the silicate feature in absorption, and the spectra show additional absorption bands at 6.0, 6.8 and 15.2\,$\mu$m due to ices. 

946: Category 2 objects show silicate 

947: emission and (most of them also) indications for water vapor absorption at shorter wavelengths. Only one Category 2 object (V1647 Ori) 

948: shows ice absorption bands which we

949: explain by foreground extinction and an intrinsically strong 10\,$\mu$m emission feature.}

950: \item{The silicate absorption is best explained with dusty and icy envelopes surrounding the Category 1 objects. Statistical 

951: reasons argue against all objects having accretion disks seen edge-on. The silicate emission of the Category 2 FUors arises from 

952: the surface layer of the surrounding accretion disks.}

953: \item{The shape of the silicate band of the Category 1 objects is in agreement with typical dust compositions of the ISM. For one object

954: (RNO 1B)  the shape of the feature and the decrease in optical depth longward of 10\,$\mu$m can be explained with a superposed emission feature.}

955: \item{Optical depths for the observed silicate and ice absorption bands were derived. We find an apparent correlation among the optical depths of the 

956: ices indicating similar environmental conditions for the objects. For the silicate feature, no correlation is expected.}

957: \item{Using different extinction curves we computed the optical extinction 

958: toward the objects based on the depth of the 10\,$\mu$m silicate features. The results are in agreement with values found in the literature, given the uncertainties in the dust models and for the optical 

959: extinction values in the references.

960: For RNO 1B the derived value for A$_{\rm V}$ is smaller than expected, indicating that on-top silicate emission might influence the optical depth at 10\,$\mu$m.}

961: \item{The emission profiles of the Category 2 objects show clear evidence for grain growth. Fitting a dust model to the spectrum of FU Ori reveals that, indeed,

962: larger grains than typically observed in the ISM are required to explain the shape. As FUors are presumably younger than TTauri stars (see below), this 

963: indicates that grain growth sets in very early during disk evolution.}

964: \item{Despite the high accretion rates of the FUor accretion disks and the resulting higher disk temperatures and mixing rates, we find hardly any evidence

965: for crystalline grains. So far we lack a clear explanation for this observational results and leave it to further investigations.}

966: \item{The two categories of FUors can be explained within a single paradigm, where Category 1 objects are younger and similar to Class I objects, while

967: Category 2 FUors are more evolved and show already properties of Class II sources. This explanation is in agreement with theoretical models which 

968: expect FUors to undergo several outbursts before they enter the more quiescent classical TTauri phase. Thus, the FUor-phase might indeed be the link between

969: Class I and Class II objects and common to most young low-mass stars.}

970: \item{For OO Ser and V1647 Ori the multi-epoch data allows an analysis of the post-outburst fading of the objects. Both objects fainted significantly 

971: over timescales of a few months, suggesting the outbursts might be intermediate between the long-lived FUor-eruptions and the short-lived EXor-type events.}

972: \item{Only one object (Parsamian 21) shows PAH emission similar to that often observed in Post-AGB stars. 

973: We find that most other observational data for Parsamian 21

974: can also be explained with the object being an evolved star. In consequence, the FUor-status for this object is questioned.  }

975: \end{itemize}

976: 

977: Based on these findings, future investigations might include the following points:

978: \begin{itemize}

979: \item{A complete MIR spectroscopic census of all known FUors would complement the data presented here and might

980: help to derive conclusions concerning the duration of the FUor phase using statistical arguments.}

981: \item{Multidimensional radiative transfer models have not yet been applied to FUor accretion disk, 

982: but they are required to 

983: derive a coherent picture of the disk structure, including the emission layer.}

984: \item{Models for the dust evolution in accretion disks could try to explain quantitatively the observed large grain population and the

985: apparent lack of crystalline silicates.}

986: \item{Finally, if our conclusions are correct, then some of the known low-mass Class I objects might be FU Ori objects hidden in a 

987: quiescent phase between two consecutive outbursts. The observational properties of these objects might be revisited to search for any

988: indication of FUor properties.}

989: \end{itemize}

990: 

991: 

992: %% The \notetoeditor{TEXT} command allows the author to communicate

993: %% information to the copy editor.  This information will appear as a

994: %% footnote on the printed copy for the manuscript style file.  Nothing will

995: %% appear on the printed copy if the preprint or

996: %% preprint2 style files are used.

997: 

998: %% The eqnarray environment produces multi-line display math. The end of

999: %% each line is marked with a \\. Lines will be numbered unless the \\

1000: %% is preceded by a \nonumber command.

1001: %% Alignment points are marked by ampersands (&). There should be two

1002: %% ampersands (&) per line.

1003: 

1004: %______________________________________________________________

1005: 

1006: 

1007: 

1008: %______________________________________________________________

1009: 

1010: 

1011: %% If you wish to include an acknowledgments section in your paper,

1012: %% separate it off from the body of the text using the \acknowledgments

1013: %% command.

1014: 

1015: %% Included in this acknowledgments section are examples of the

1016: %% AASTeX hypertext markup commands. Use \url without the optional [HREF]

1017: %% argument when you want to print the url directly in the text. Otherwise,

1018: %% use either \url or \anchor, with the HREF as the first argument and the

1019: %% text to be printed in the second.

1020: 

1021: \acknowledgments

1022: S.~P.~Q.~kindly acknowledges support from the German

1023: \emph{Friedrich-Ebert-Stiftung}. We are grateful to Henrik Beuther and Kees Dullemond for 

1024: interesting and insightful discussions and thank the referee for a detailed report helping to

1025: improve the style and content of this paper.  

1026: The version of the ISO data presented in this paper correspond to Highly Processed Data Product (HPDP) sets  available for public use in the ISO Data Archive. OSIA is a joint development of the SWS consortium. Contributing institutes are SRON, MPE, KUL and the ESA Astrophysics Division. This research has made use of the SIMBAD database,

1027: operated at CDS, Strasbourg, France. 

1028: 

1029: %% To help institutions obtain information on the effectiveness of their

1030: %% telescopes, the AAS Journals has created a group of keywords for telescope

1031: %% facilities. A common set of keywords will make these types of searches

1032: %% significantly easier and more accurate. In addition, they will also be

1033: %% useful in linking papers together which utilize the same telescopes

1034: %% within the framework of the National Virtual Observatory.

1035: %% See the AASTeX Web site at http://www.journals.uchicago.edu/AAS/AASTeX

1036: %% for information on obtaining the facility keywords.

1037: 

1038: %% After the acknowledgments section, use the following syntax and the

1039: %% \facility{} macro to list the keywords of facilities used in the research

1040: %% for the paper.  Each keyword will be checked against the master list during

1041: %% copy editing.  Individual instruments or configurations can be provided 

1042: %% in parentheses, after the keyword, but they will not be verified.

1043: 

1044: {\it Facilities:} \facility{Spitzer},\facility{ISO}

1045: %% Appendix material should be preceded with a single \appendix command.

1046: %% There should be a \section command for each appendix. Mark appendix

1047: %% subsections with the same markup you use in the main body of the paper.

1048: 

1049: %% Each Appendix (indicated with \section) will be lettered A, B, C, etc.

1050: %% The equation counter will reset when it encounters the \appendix

1051: %% command and will number appendix equations (A1), (A2), etc.

1052: 

1053: 

1054: %% The reference list follows the main body and any appendices.

1055: %% Use LaTeX's thebibliography environment to mark up your reference list.

1056: %% Note \begin{thebibliography} is followed by an empty set of

1057: %% curly braces.  If you forget this, LaTeX will generate the error

1058: %% "Perhaps a missing \item?".

1059: %%

1060: %% thebibliography produces citations in the text using \bibitem-\citep

1061: %% cross-referencing. Each reference is preceded by a

1062: %% \bibitem command that defines in curly braces the KEY that corresponds

1063: %% to the KEY in the \citep commands (see the first section above).

1064: %% Make sure that you provide a unique KEY for every \bibitem or else the

1065: %% paper will not LaTeX. The square brackets should contain

1066: %% the citation text that LaTeX will insert in

1067: %% place of the \citep commands.

1068: 

1069: %% We have used macros to produce journal name abbreviations.

1070: %% AASTeX provides a number of these for the more frequently-citepd journals.

1071: %% See the Author Guide for a list of them.

1072: 

1073: %% Note that the style of the \bibitem labels (in []) is slightly

1074: %% different from previous examples.  The natbib system solves a host

1075: %% of citation expression problems, but it is necessary to clearly

1076: %% delimit the year from the author name used in the citation.

1077: %% See the natbib documentation for more details and options.

1078: 

1079: \input{ms.bbl}

1080: %\bibliographystyle{apj}

1081: %\bibliography{bibliography.bib}

1082: 

1083: \clearpage

1084: 

1085: \begin{deluxetable}{lcccccl}

1086: \rotate

1087: \tablecaption{Journal of {\sc ISO-SWS} observations. The coordinates denote the pointing

1088: position of the telescope as saved in the header of the data files. The speed of the observations, the time spent on the target and possible pointing offsets in the telescope's y and z axis are given as well. The offsets were

1089: corrected for during the data reduction applying the measured beam profiles along the different axes. 

1090: \label{isojournal}}           % title of Table

1091: \tablewidth{0pt}

1092:       % is used to refer this table in the text

1093: \tablehead{

1094: \colhead{Object} & \colhead{RA (J2000)} & \colhead{DEC (J2000)} & 

1095: \colhead{AOT / Speed} & \colhead{Time on} & \colhead{Offset [$''$]} & \colhead{Date} \\

1096: 	 &  &     &  &	\colhead{target [sec]}  &  \colhead{y / z}  & 

1097: }

1098: \startdata

1099: \object{OO Ser}  &  18h29m49.05s & +01d16$'$19.2$''$ & SWS01 / 1 &  1062 &  - / - &1996-04-14 \\  

1100:                &  18h29m49.08s   & +01d16$'$19.8$''$ & SWS01 / 1 &  1140 &  4 / 3 &1996-10-24 \\

1101:   	     &  18h29m49.05s & +01d16$'$19.2$''$ & SWS01 / 2 &  1140 &  - / - & 1997-03-08 \\

1102:   	    &  18h29m49.05s & +01d16$'$19.2$''$ & SWS01 / 3 &  3454 &  - / - & 1997-04-12 \\

1103:   	    &  18h29m49.09s & +01d16$'$19.8$''$ & SWS01 / 3 &  3454 &  - / - & 1997-09-22 \\

1104: RNO 1B  & 00h36m46.24s & +63d28$'$54.3$''$	& SWS01 / 2 & 	1912 &  - / - &  1996-08-27\\

1105: V346 Nor &  16h32m32.05s & -44d55$'$28.9$''$ & SWS01 / 2 & 1912 &  - / - &  1996-08-31\\

1106: \object{Z CMa} &	 07h03m43.17s &  -11d33$'$06.6$''$ & SWS01 / 2 &  3454 &   3 / 3 &  1997-11-07\\

1107: V1735 Cyg  & 21h47m20.60s & +47d32$'$04.9$''$ & SWS01 / 2 & 1912 &  4 / 4 & 1996-08-06\\

1108: Reipurth 50 & 05h40m17.89s &  -07d27$'$29.3$''$	& SWS01 / 3 &	3454 &  5 / 5 & 1997-10-13\\

1109: L1551 IRS 5 & 04h31m34.06s & +18d08$'$04.8$''$	& SWS01	/ 4 & 	6538 & - / -  & 1997-09-06\\

1110:  

1111: \enddata

1112: \end{deluxetable}

1113: 

1114: %\clearpage

1115: 

1116: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1117: 

1118: \begin{deluxetable}{lcclll}

1119: \rotate

1120: \tablecaption{Journal of {\sc Spitzer-IRS}  observations. The coordinates denote the average slit 

1121: position of the low resolution spectrograph as computed by the on board software. 

1122: The AOR of the observations, the time spent on the target for the different modules and the observation date are given as well. 

1123: \label{spitzerjournal}}           % title of Table

1124: \tablewidth{0pt}

1125:       % is used to refer this table in the text

1126: \tablehead{

1127: \colhead{Object} & \colhead{RA (J2000)} & \colhead{DEC (J2000)} & 

1128: \colhead{AOR} & \colhead{Integration time $[$sec$]$} & %\colhead{Offset $['']$} & 

1129: \colhead{Date} 

1130: }

1131: \startdata

1132: \object{Bran 76} (BBW 76) & 07h50m35.52s   & -33d06$'$24.12$''$ & 3571200 &  12 (SL, LL)   & 2004-04-14 \\  

1133: \object{FU Ori}  &  05h45m22.39s   & +09d04$'$12.5$''$ & 3569920 &  12 (SL, SH, LH)   & 2004-03-04 \\

1134: \object{L1551 IRS 5}  &  04h31m34.08s & +18d08$'$04.92$''$ & 3531776 &  6 (SL, SH, LH) &  2004-03-04 \\

1135: \object{Parsamian 21} (HBC 687) &  19h29m00.72s & +09d38$'$47.11$''$ &  5039872 &  36 (SL) &     2004-04-18 \\

1136:              &              &                   &          & 48 (SH, LH)   &        \\

1137: \object{RNO 1B}   &  00h36m46.34s &  +63d28$'$53.76$''$    & 6586624  &  36 (SL) &    2004-01-07\\

1138:              &              &                   &          &  60 (SH, LH)   &         \\

1139: \object{RNO 1C}   &  00h36m46.89s &  +63d28$'$58.44$''$    & 6586624  &  36 (SL) &     2004-01-07 \\

1140:              &              &                   &          &  60 (SH, LH)   &        \\

1141: \object{V1057 Cyg} & 20h58m53.76s & +44d15$'$28.44$''$     & 3570176  &  12 (SL, SH, LH)&   2003-12-15\\

1142: \object{V1515 Cyg} & 20h23m48.00s & +42d12$'$25.56$''$     & 3570432  &  12 (SL, SH, LH)&   2004-05-11 \\

1143: 

1144: \object{V1647 Ori}\tablenotemark{a} & 05h46m13.13s & -00d06$'$05.21$''$     & 12261120 &  12 (SL) &     2004-10-20 \\

1145:              &              &                   &          &  24 (SH, LH)   &          \\

1146:           & 05h46m13.15s & -00d06$'$04.41$''$      & 11569920 & 48 (SL) &     2005-03-11 \\

1147:              &              &                   &          &  484 (SH), 240 (LH)   &       \\

1148:           & 05h46m13.14s & -00d06$'$04.69$''$       & 12644096 &  12 (SL, LL) &     2005-03-24 \\

1149: \object{V1735 Cyg} & 21h47m20.6s &  +47d32$'$00.7$''$	& 3570944  &  12 (SL, SH, LH) &  2003-12-17 \\

1150: \object{V346 Nor}  & 16h32m32.1s &  -44d55$'$28.6$''$	& 3570688  &  12 (SL, SH, LH) & 2004-02-27 \\\hline

1151: \object{HL Tau} & 04h31m38.4s &  +18d13$'$57.9$''$	& 3531776  &  6 (SL, SH, LH) & 2004-03-04 \\

1152: \object{XZ Tau} AB& 04h31m40.1s &  +18d13$'$57.4$''$	& 3531776  &  6 (SL, SH, LH) & 2004-03-04 \\

1153: \enddata

1154: \tablenotetext{a}{The object was observed at three different epochs.}

1155: \end{deluxetable}

1156: 

1157: 

1158: %\clearpage

1159: \thispagestyle{empty}

1160: \begin{deluxetable}{llccccccc}

1161: \rotate

1162: \tablecaption{Overview of prominent spectroscopic features (mostly ices and dust) 

1163: seen in the spectra presented in Figures~\ref{iso_short1},~\ref{iso_short2},~\ref{spitzer_short1},~\ref{spitzer_short2},~\ref{spitzer_long1},~\ref{spitzer_long2}. "abs" denotes an absorption feature, "em" an emission feature. For objects where

1164: ISO and {\sc Spitzer} data are available, the {\sc Spitzer} data is of higher quality and thus more reliable.  

1165: \label{features}}           % title of Table

1166: \tablewidth{0pt}

1167:       % is used to refer this table in the text

1168: \tablehead{

1169: \colhead{Object} & 

1170: \colhead{Instrument} &

1171: \colhead{H$_2$O} (ice) & 

1172: \colhead{H$_2$CO(?)/HCOOH(?)/} &

1173: \colhead{H$_2$O (gas)} &

1174: \colhead{CH$_3$OH(?)/} &

1175: \colhead{silicates} &

1176: \colhead{CO$_2$ (ice)} &  

1177: \colhead{PAH\tablenotemark{a}} \\

1178:  & 

1179:  &

1180:  & 

1181: \colhead{H$_2$O (ice)} &

1182:  &

1183: \colhead{NH$_4^+$(?)} &

1184:  &

1185:  &  

1186:  \\

1187:  &   &

1188: \colhead{3.08$\mu$m} & 

1189: \colhead{5.85 / 6.0\,$\mu$m} & 

1190: \colhead{5.8 / 6.8$\mu$m}&

1191: \colhead{6.85$\mu$m} &

1192: \colhead{10.0$\mu$m} &

1193: \colhead{15.2$\mu$m} & 

1194: }

1195: \startdata

1196: OO Ser       &  ISO-SWS       &   - & -   & -  &  -   & abs & abs & - \\

1197: V346 Nor     &  ISO-SWS       & abs & abs & -  &  -   & abs & -   & - \\

1198:              &  {\sc Spitzer} IRS   &   - & abs & -  &  abs & abs & abs & - \\

1199: Z CMa         &  ISO-SWS       &   - & -   & -  &  -   & abs & -   & - \\

1200: Reipurth 50  & ISO-SWS        & abs & abs & -  &  abs & abs & -   & - \\

1201: L1551 IRS 5  & ISO-SWS        & abs & abs & -  &  abs & abs & abs & - \\

1202:              & {\sc Spitzer} IRS    &   - & abs & -  &  abs & abs & abs & - \\

1203: RNO 1B       &  ISO-SWS       & abs & -   & -  &  -   & flat & - & - \\

1204: RNO 1B       &  {\sc Spitzer} IRS   &   - & abs & -  &  abs & abs & abs & - \\

1205: RNO 1C       &  {\sc Spitzer} IRS   &   - & abs & -  &  abs & abs & abs & - \\

1206: V1735 Cyg    &  ISO-SWS       & abs & abs?& -  &  -   & flat & - & - \\

1207:              &  {\sc Spitzer} IRS   &   - & -   & -  &  -   & abs  & abs? & - \\

1208: Parsamian 21 & {\sc Spitzer} IRS    &   - & -   & -  &  -   & em  & -  & em \\  

1209: Bran 76 (BBW 76)       & {\sc Spitzer} IRS    &   - & -   & abs&  -   & em  & -  & -  \\

1210: V1057 Cyg    & {\sc Spitzer} IRS    &   - & -   & abs&  -   & em  & abs (gas) & - \\

1211: V1515 Cyg    & {\sc Spitzer} IRS    &   - & -   & abs&  -   & em  & -   & -  \\

1212: V1647 Ori    & {\sc Spitzer} IRS    &   - & abs & -  &  abs & em  & abs & - \\

1213: FU Ori       & {\sc Spitzer} IRS    &   - & -   & abs&  -   & em  & -   & -  \\\hline

1214: V883 Ori\tablenotemark{b}     & ESO Timmi2     &   - & -   & -  &   -  & abs & -   & - \\\hline

1215: XZ Tau AB    & {\sc Spitzer} IRS    &   - & -   & -  &  -   & em  & -   & - \\

1216: HL Tau       & {\sc Spitzer} IRS    &   - & abs & -  &  abs & abs   & abs & - \\

1217: \enddata 

1218: \tablenotetext{a}{Any significant emission at 6.2 or 7.7\,$\mu$m (C-H modes), 8.6, 11.3 or 12.7\,$\mu$m (C-C modes)

1219: or at 8.2\,$\mu$m (origin not clear yet).}

1220: \tablenotetext{b}{Observations from \citet{schuetz2005}.}

1221: \end{deluxetable}

1222: 

1223: %\clearpage

1224: 

1225: \begin{deluxetable}{lcccc}

1226: \tablecaption{Temperatures and power law indices for the different components of the 

1227: analytical dust model for FU Ori shown in Figure~\ref{dust_model}.  

1228: \label{temperature_table}}           % title of Table

1229: \tablewidth{0pt}

1230:       % is used to refer this table in the text

1231: \tablehead{

1232: \colhead{Component} & \colhead{T [K]}& \colhead{Exponent $p$}

1233: }

1234: \startdata

1235: Inner disk rim  & 2246 & -0.84 \\

1236: Disk surface & 1128 & -0.49  \\

1237: Disk midplane & 865 & -0.11 \\

1238: \enddata

1239: \end{deluxetable}

1240: 

1241: %\clearpage

1242: 

1243: \begin{deluxetable}{lcccc}

1244: \tablecaption{Mass fractions of dust species with different grain sizes 

1245: derived from the dust model fit for FU Ori shown in Figure~\ref{dust_model}.  

1246: \label{dust_table}}           % title of Table

1247: \tablewidth{0pt}

1248:       % is used to refer this table in the text

1249: \tablehead{

1250: \colhead{Species} & \colhead{0.1\,$\mu$m}&\colhead{1.5\,$\mu$m}& \colhead{6.0\,$\mu$m}& \colhead{Total}

1251: }

1252: \startdata

1253: Amorphous olivine & 0.16 & 0.20 & 0.21 &  0.57 \\

1254: Amorphous pyroxene & 0.00 & 0.00 & 0.43 & 0.43 \\

1255: Crystalline forsterite & $<$0.01 & 0.00 & 0.00 & $<$0.01 \\

1256: Crystalline enstatite & 0.00 & 0.00 & 0.00 & 0.00\\

1257: Silica & 0.00 & 0.00 & 0.00 & 0.00\\

1258: Amorphous carbon & 0.00 & 0.00 & 0.00 & 0.00\\\hline\\

1259: Total & 0.16 & 0.20 & 0.64 & \\

1260: \enddata

1261: \end{deluxetable}

1262: 

1263: %\clearpage

1264: 

1265: \begin{deluxetable}{lccccc}

1266: \tablecaption{Optical depths and extinction values as derived from the spectra shown in Figure~\ref{absorption}.

1267: The last column list reference values for $A_{\rm V}$ from the literature.

1268: \label{extinctiontable}}           % title of Table

1269: \tablewidth{0pt}

1270:       % is used to refer this table in the text

1271: \tablehead{

1272: \colhead{Object} & \colhead{$\tau_{\rm silicate}$} & \colhead{$\lambda_{\rm silicate}$ [$\mu$m]} & 

1273: \colhead{$A_{\rm silicate}$ [mag]} & \colhead{$A_{\rm V}$ [mag]} & \colhead{$A_{\rm V}^{literature}$ [mag]}  

1274: }

1275: \startdata

1276:      L1551 IRS 5 &  1.33$\pm$0.03  & 9.44 &  1.44$\pm$0.03 &   21.83$\pm$   4.70 &

1277:       19-20\tablenotemark{a}, 30\tablenotemark{b}, $\gtrsim$150\tablenotemark{c} \\

1278:       

1279:        V346 Nor  &  0.32$\pm$0.01  & 9.68 &  0.35$\pm$0.01 &    4.98$\pm$   0.86 &  

1280:        6.2\tablenotemark{d}\\

1281:       V1735 Cyg  &  0.29$\pm$0.03  & 9.52 &  0.31$\pm$0.03 &    4.60$\pm$   0.91 &  

1282:        10.0\tablenotemark{e,}\tablenotemark{f}\\

1283:          RNO 1B  &  0.24$\pm$0.01  & 9.30 &  0.27$\pm$0.01 &    4.15$\pm$   1.03 &

1284: 	 9.2\tablenotemark{g}  \\

1285:          RNO 1C  &  0.63$\pm$0.01  & 9.58 &  0.69$\pm$0.01 &   10.11$\pm$   1.87 &  

1286: 	 $\gtrsim$12\tablenotemark{g}  \\

1287:          HL Tau  &  0.47$\pm$0.01  & 9.61 &  0.51$\pm$0.01 &    7.38$\pm$   1.32 &  

1288: 	 $\approx$24.0\tablenotemark{h,}\tablenotemark{i}, $\approx$38\tablenotemark{j}\\

1289:           Z CMa  &  0.31$\pm$0.06  & 9.48 &  0.34$\pm$0.20 &    5.05$\pm$   1.04 &  

1290: 	 2.8$\pm$0.1\tablenotemark{k}  \\

1291:         Reipurth 50 &  1.68$\pm$0.34  & 9.68 &  1.82$\pm$0.20 &   26.18$\pm$   4.34 &  

1292:      $\approx$50\tablenotemark{l}  \\

1293: \enddata

1294: \tablenotetext{a}{\citet{snell1985}}

1295: \tablenotetext{b}{\citet{smith1987}}

1296: \tablenotetext{c}{\citet{campbell1988}}

1297: \tablenotetext{d}{\citet{grahamfrogel1985}, assuming E(B-V)=2.0}

1298: \tablenotetext{e}{\citet{sato1992}}

1299: \tablenotetext{f}{\citet{levreault1988}}

1300: \tablenotetext{g}{\citet{staudeneckel1991}, based on optical data and assuming spectral type.}

1301: \tablenotetext{h}{\citet{bergin2005}}

1302: \tablenotetext{i}{\citet{close1997}}

1303: \tablenotetext{j}{\citet{menshchikov1999}, derived from radiative transfer model.}

1304: \tablenotetext{k}{\citet{cohenkuhi1979}, based on optical data and assuming spectral type.}

1305: \tablenotetext{l}{\citet{casali1991}, based on NIR colors, assumed intrinsic color temperatures and

1306: strength of additional ice absorption features.}

1307: \end{deluxetable}

1308: 

1309: %\clearpage

1310: 

1311: \begin{deluxetable}{lccc}

1312: \tablecaption{Optical depths of the main ice features at 6.0, 6.85 and 15.2\,$\mu$m. 

1313: \label{iceextinctiontable}}           % title of Table

1314: \tablewidth{0pt}

1315:       % is used to refer this table in the text

1316: \tablehead{

1317: \colhead{Object} & \colhead{$\tau_{6.0\,\mu{\rm m}}$}&\colhead{$\tau_{6.85\,\mu{\rm m}}$}& \colhead{$\tau_{15.2\,\mu{\rm m}}$}

1318: }

1319: \startdata

1320:      L1551 IRS 5 &  0.51$\pm$0.01	&0.47$\pm$0.01	&	0.54$\pm$0.05	\\

1321:        V346 Nor  &  0.05$\pm$0.01	&0.06$\pm$0.01	&	0.12$\pm$0.02	\\

1322:       V1735 Cyg  &  0.05$\pm$0.02	&0.09$\pm$0.01	&	$\approx$0.3 	\\

1323:          RNO 1B  &  0.20$\pm$0.01	&0.18$\pm$0.01	&	0.32$\pm$0.04	\\

1324:          RNO 1C  &  0.35$\pm$0.01	&0.26$\pm$0.01	&	0.41$\pm$0.02	\\

1325:          HL Tau  &  0.10$\pm$0.01	&0.07$\pm$0.01	&	0.13$\pm$0.06	\\ 

1326:       V1647 Ori  &  0.07$\pm$0.01	&0.06$\pm$0.01	&	0.20$\pm$0.02	\\

1327:       V1647 Ori 2&  0.10$\pm$0.01	&0.08$\pm$0.01	&	0.19$\pm$0.02 	\\

1328:       V1647 Ori 3&  0.09$\pm$0.01	&0.07$\pm$0.01	&	-\tablenotemark{a}  \\

1329: \enddata

1330: \tablenotetext{a}{only low-resolution data}

1331: \end{deluxetable}

1332: 

1333: %\clearpage

1334: 

1335: \begin{deluxetable}{lcccc}

1336: \tablecaption{\emph{ Upper half:} Flux densities at 20 and 30\,$\mu$m and the slope of the SED between 15 and 30\,$\mu$m 

1337: for OO Ser observed at 5 different epochs. \emph{ Lower half:} Flux densities at 8, 20 and 30\,$\mu$m for V1647 Ori at three 

1338: different epochs. 

1339: \label{ooserflux}}           % title of Table

1340: \tablewidth{0pt}

1341:       % is used to refer this table in the text

1342: \tablehead{

1343: \colhead{Date} &  \colhead{Flux density} & \colhead{Flux density} &\colhead{Flux density} & \colhead{Slope [Jy/$\lambda$]}\\

1344: \colhead{} &  \colhead{at 8\,$\mu$m [Jy]} &\colhead{at 20\,$\mu$m [Jy]} &\colhead{at 30\,$\mu$m [Jy]} & \colhead{}

1345: }

1346: \startdata

1347: \multicolumn{5}{c}{OO Ser}\\\hline\\

1348:      1996-04-14 &  &16.5$\pm$1.8		&43.0$\pm$2.3	&	2.65$\pm$0.04 \\

1349:      1996-10-24 &  &15.2$\pm$2.1		&41.2$\pm$2.6	&	2.61$\pm$0.05 \\

1350:      1997-03-08 &  &12.8$\pm$1.2		&32.5$\pm$1.5	&	1.98$\pm$0.03 \\

1351:      1997-04-12 &  &11.9$\pm$1.5		&31.1$\pm$1.8	&	1.92$\pm$0.04 \\

1352:      1997-09-22 &  &8.8$\pm$1.2		        &23.9$\pm$1.4	&	1.52$\pm$0.03 \\\hline\\

1353: \multicolumn{5}{c}{V1647 Ori}\\\hline\\

1354:      2004-10-20 & 3.95$\pm$0.02 & 11.56$\pm$0.16 & 16.52$\pm$0.51 & \\

1355:      2005-03-11 & 2.49$\pm$0.01 & 7.21$\pm$0.20 & 10.35$\pm$0.24 & \\

1356:      2005-03-24 & 3.04$\pm$0.01 & 7.82$\pm$0.17\tablenotemark{a} & 10.62$\pm$0.05\tablenotemark{a} & \\

1357: \enddata

1358: \tablenotetext{a}{\phantom{}low-resolution data}

1359: \end{deluxetable}

1360: 

1361: 

1362: %\clearpage

1363: 

1364: %% Use the figure environment and \plotone or \plottwo to include

1365: %% figures and captions in your electronic submission.

1366: %% To embed the sample graphics in

1367: %% the file, uncomment the \plotone, \plottwo, and

1368: %% \includegraphics commands

1369: %%

1370: %% If you need a layout that cannot be achieved with \plotone or

1371: %% \plottwo, you can invoke the graphicx package directly with the

1372: %% \includegraphics command or use \plotfiddle. For more information,

1373: %% please see the tutorial on "Using Electronic Art with AASTeX" in the

1374: %% documentation section at the AASTeX Web site,

1375: %% http://www.journals.uchicago.edu/AAS/AASTeX.

1376: %%

1377: %% The examples below also include sample markup for submission of

1378: %% supplemental electronic materials. As always, be sure to check

1379: %% the instructions to authors for the journal you are submitting to

1380: %% for specific submissions guidelines as they vary from

1381: %% journal to journal.

1382: 

1383: %% This example uses \plotone to include an EPS file scaled to

1384: %% 80% of its natural size with \epsscale. Its caption

1385: %% has been written to indicate that additional figure parts will be

1386: %% available in the electronic journal.

1387: 

1388: \begin{figure}

1389: \centering

1390: \epsscale{1.}

1391:     \plotone{f1.eps}

1392:    \caption{ISO-SWS spectra in the 2-15\,$\mu$m range for 6 of our targets. The two 

1393:    plots in the top panel (V1735 Cyg and RNO 1B) suffer from a poor signal-to-noise 

1394: ratio resulting from a combination of relatively low flux levels and a short integration times (Table~\ref{isojournal}). All sub-structures in the spectra are not believed to be real. }

1395:  \label{iso_short1}

1396: \end{figure}

1397: 

1398: 

1399: 

1400: \begin{figure}

1401: \centering

1402: \epsscale{1.}

1403:     \plotone{f2.eps}

1404:    \caption{Same as Figure~\ref{iso_short1} but for the object OO Ser which was observed at 5 

1405:    different epochs. As shown in Table~\ref{isojournal} the integration time of the 

1406:    observations were increased during the 5 epochs. However, as the object faded significantly over this period,

1407:    the signal-to-noise ratio could not be improved and it is difficult to identify spectroscopic features in the data

1408:    apart from the silicate absorption profile seen in all data sets.}

1409:  \label{iso_short2}

1410: \end{figure}

1411: 

1412: 

1413: 

1414: \begin{figure}

1415: \centering

1416: \epsscale{1.}

1417:     \plotone{f3.eps}

1418:    \caption{Same as Figure~\ref{iso_short1} but between 15 and 40\,$\mu$m. All spikes seen in the data are artefacts 

1419:    and not believed to be real, as they showed up either in the up- or in the down-scan but not in both.}

1420:  \label{iso_long1}

1421: \end{figure}

1422: 

1423: 

1424: 

1425: \begin{figure}

1426: \centering

1427: \epsscale{1.}

1428:     \plotone{f4_color.eps}

1429:    \caption{Same as Figure~\ref{iso_short2} but between 15 and 40\,$\mu$m. The red, dashed line illustrates 

1430:    a fit to the data between 15 and 30\,$\mu$ to estimate the change of the slope and flux density levels during

1431:    the decay of OO Ser (see text section 3.4.). All spikes seen in the data are artefacts 

1432:    and not believed to be real, as they showed up either in the up- or in the down-scan but not in both. }

1433:  \label{iso_long2}

1434: \end{figure}

1435: 

1436: 

1437: 

1438: \begin{figure}

1439: \centering

1440: \epsscale{1.}

1441:     \plotone{f5.eps}

1442:    \caption{{\sc Spitzer} low-resolution spectra between 5 and 14\,$\mu$m.}

1443:  \label{spitzer_short1}

1444: \end{figure}

1445: 

1446: 

1447: 

1448: \begin{figure}

1449: \centering

1450: \epsscale{1.}

1451:     \plotone{f6.eps}

1452:    \caption{Same as Figure~\ref{spitzer_short1}.}

1453:  \label{spitzer_short2}

1454: \end{figure}

1455: 

1456: 

1457: 

1458: \begin{figure}

1459: \centering

1460: \epsscale{1.}

1461:     \plotone{f7.eps}

1462:    \caption{Same objects as in Figure~\ref{spitzer_short1}, but now showing the high-resolution 

1463:    part of the spectrum between 14 and 34\,$\mu$m. The spectrum of V1515 Cyg shows a prominent emission line

1464:    longward of 33\,$\mu$m which is discussed in section 3.5. Between 31 and 32\,$\mu$m all objects 

1465:    where high-resolution data is present show artefacts from the data redcution process.}

1466:  \label{spitzer_long1}

1467: \end{figure}

1468: 

1469: 

1470: 

1471: \begin{figure}

1472: \centering

1473: \epsscale{1.}

1474:     \plotone{f8.eps}

1475:    \caption{Same objects as in Figure~\ref{spitzer_short2}, but now showing the high-resolution 

1476:    part of the spectrum between 14 and 34\,$\mu$m. Between 31 and 32\,$\mu$m all objects 

1477:    where high-resolution data is present show artefacts from the data redcution process.}

1478:  \label{spitzer_long2}

1479: \end{figure}

1480: 

1481: 

1482: 

1483: \begin{figure}

1484: \centering

1485: \epsscale{1.}

1486:     \plotone{f9_color.eps}

1487:    \caption{Comparison between ISO/SWS and {\sc Spitzer}/IRS data for objects observed with both satellites.

1488:    The ISO data (red, solid lines) are significantly more noisy. In our sample 

1489:    only data for objects with high flux levels or long integration times are suitable for quantitative 

1490:    analyses.  

1491:     }

1492:  \label{compare_spectra}

1493: \end{figure}

1494: 

1495: 

1496: 

1497: %\clearpage

1498: 

1499: \begin{figure}

1500: \centering

1501: \epsscale{.85}

1502:     \plotone{f10_color.eps}

1503:    \caption{Continuum subtracted silicate emission profiles (black, solid lines). 

1504:    Each spectrum is shown twice, i.e., with and without being corrected for interstellar extinction as explained 

1505:    in the text. The assumed values for A$_{\rm V}$ are: 2.2 mag \citep[Bran 76; ][]{reipurth2002},

1506:    2.4 mag \citep[FU Ori; ][]{skinner2006}, 3.2 mag \citep[V1515 Cyg; ][]{herbig1977}, 

1507:    3.2 mag \citep[V1057 Cyg; ][]{herbig1977},

1508:    11.0 mag \citep[V1647 Ori; ][]{muzerolle2005}, 1.4 mag \citep[XZ Tau; ][]{white2001}, 

1509:    4.0 mag \citep[Parsamian 21; ][]{staudeneckel1992}. 

1510:    We used the extinction law from \citet{mathis1990} with R$_{\rm V}$=3.1.

1511:    For comparison the red, dashed line shows the typical ISM silicate feature observed 

1512:    toward the galactic center from \citet{kemper2004}. 

1513:    As mentioned in the text V1647 Ori was observed at three different epochs. 

1514:    }

1515:  \label{emission}

1516: \end{figure}

1517: 

1518: \clearpage

1519: 

1520: \begin{figure}

1521: \centering

1522: \epsscale{1.}

1523:     \plotone{f11_color.eps}

1524:    \caption{Dereddened and normalized 10\,$\mu$m dust emission features of the objects shown in Figure~\ref{emission}. For a better comparison the dereddened spectra are shown.}

1525:  \label{normalizedemission}

1526: \end{figure}

1527: 

1528: %\clearpage

1529: 

1530: 

1531: \begin{figure}

1532: \centering

1533: \epsscale{1.}

1534:     \plotone{f12.eps}

1535:    \caption{The flux ratio between 11.3 and 9.8\,$\mu$m plotted against 

1536:    the normalized peak over the continuum.

1537:    The dashed line indicates the region where typically TTauri stars \citep{przygodda2003} and Herbig Ae/Be stars 

1538:    \citep{vanboekel2003,vanboekel2005} are located.}

1539:  \label{fluxratio}

1540: \end{figure}

1541: 

1542: %\clearpage

1543: 

1544: \begin{figure}

1545: \centering

1546: \epsscale{1.}

1547:     \plotone{f13_color.eps}

1548:    \caption{Results from a dust model fit to the 7-17\,$\mu$m region of the spectrum of FU Ori 

1549:    using the model described in the text. The observed spectrum 

1550:    is shown in the black, solid line, the 

1551:    blue, dotted line illustrates the contribution from the fitted continuum, and the green, dashed line

1552:    denotes the computed emission feature. The sum of the fitted 

1553:    components is shown in the red, dash-dotted line, matching the observed spectrum.

1554:    The mass fractions of the fitted dust species are given in Table~\ref{dust_table}. 

1555:    The spectrum is fitted solely with amorphous grains, part of which have grown significantly.}

1556:  \label{dust_model}

1557: \end{figure}

1558: 

1559: 

1560: %\clearpage

1561: 

1562: 

1563: \begin{figure}

1564: \centering

1565: \epsscale{1.}

1566:     \plotone{f14_color.eps}

1567:    \caption{Dereddened {\sc Spitzer} MIR spectrum of Parsamian 21 (assuming A$_{\rm V}$\,=\,4.0 mag)

1568:    illustrating the strong PAH emission features not seen

1569:    in any other FUor. Note also the steeply rising continuum and the underlying silicate dust emission.

1570:    The vertical lines indicate the positions of typical PAH emission bands: strong emission is detected

1571:    at 6.3, 8.2 and 11.3\,$\mu$m (red, dashed lines), no emission is seen at 7.7, 8.6 and 12.7\,$\mu$m.

1572:    }

1573:  \label{parsamian_pah}

1574: \end{figure}

1575: 

1576: %\clearpage

1577: 

1578: 

1579: \begin{figure}

1580: \centering

1581: \epsscale{1.}

1582:     \plotone{f15_color.eps}

1583:    \caption{Observed optical depths of the 10\,$\mu$m silicate absorption feature (black, solid line). 

1584:    The dotted line shows the

1585:    assumed continuum. The vertical 

1586:    red, dash-dotted lines indicate the position of the maximum optical depth of the second order polynomial fitted 

1587:    to the data between 8.8 and 10.2 (also plotted in red, dash-dotted lines).}

1588:  \label{absorption}

1589: \end{figure}

1590: 

1591: %\clearpage

1592: \begin{figure}

1593: \centering

1594: \epsscale{1.}

1595:     \plotone{f16_color.eps}

1596:    \caption{Optical depths for different dust grain populations on an arbitrary scale. 

1597:    The black, solid line shows the profile for 0.1\,$\mu$m sized amorphous 

1598:    olivine grains \citep{dorschner1995}, the red, dashed 

1599:    line shows the ISM feature toward the galactic plane from \citet{kemper2004}, the green, dash-dotted line is the 

1600:    profile based on the astronomical silicates from \citet{weingartnerdraine2001} and \citet{draine2003}, 

1601:    and the blue, dotted line is for 0.1\,$\mu$m sized amorphous pyroxene

1602:    grains \citep{dorschner1995}.}

1603:  \label{abs_profiles}

1604: \end{figure}

1605: 

1606: %\clearpage

1607: 

1608: 

1609: \begin{figure}

1610: \centering

1611: \epsscale{1.}

1612:     \plotone{f17_color.eps}

1613:    \caption{Same observed spectra as in Figure~\ref{absorption} (black, solid lines),  

1614:    but now overplotted with reference spectra with different dust compositions (same color and

1615:    line style code as in Figure~\ref{abs_profiles}). }

1616:  \label{absorption2}

1617: \end{figure}

1618: 

1619: %\clearpage

1620: 

1621: \begin{figure}

1622: \centering

1623: \epsscale{1.}

1624:     \plotone{f18_color.eps}

1625:    \caption{Comparison of the 15.2\,$\mu$m CO$_2$-ice feature between the heavily extincted background 

1626:    object CK2 \citep{knez2005} and the embedded young source HH46 \citep{boogert2004}. Due to

1627:    higher temperatures in the vicinity of the embedded protostar crystallization occurs and 

1628:    and a double-peaked sub-structure appears characteristic of the pure CO$_2$ matrix \citep{ehrenfreund1998}.}

1629:  \label{absorption15_comparison}

1630: \end{figure}

1631: 

1632: %\clearpage

1633: 

1634: \begin{figure}

1635: \centering

1636: \epsscale{1.}

1637:     \plotone{f19_color.eps}

1638:    \caption{The 15.2\,$\mu$m CO$_2$-ice feature observed toward some of our objects (black, solid lines).  

1639:     The spectra of HL Tau, RNO 1B and V1735 Cyg were smoothed by a factor of 20 to increase the signal to noise. 

1640:     Overplotted is either the scaled spectrum of HH46 or CK2 (red, dashed-dotted line; see

1641:     Figure~\ref{absorption15_comparison}) depending on which profile fits better to the FUor data.

1642:     Due to the remaining high noise level in the spectrum of V1735 Cyg no clear comparison to 

1643:     either reference object was possible. The absorption feature of V1647 Ori observed at the third 

1644:     epoch is not shown as these are only low-resolution data, while the data presented here (first and second epoch)

1645:     were taken with the high-resolution spectrograph.}

1646:  \label{absorption15}

1647: \end{figure}

1648: 

1649: %\clearpage

1650: 

1651: \begin{figure}

1652: \centering

1653: \epsscale{1.}

1654:     \plotone{f20_color.eps}

1655:    \caption{The 6.0 and 6.85\,$\mu$m ice bands observed toward some of our objects (black lines). Overplotted

1656:    are the polynomial fits of fourth order (red, dashed-dotted line) to determine the optical depths of the

1657:    absorption bands. }

1658:  \label{absorption6}

1659: \end{figure}

1660: 

1661: %\clearpage

1662: 

1663: \begin{figure}

1664: \centering

1665: \epsscale{1.}

1666:     \plotone{f21.eps}

1667:    \caption{Zoom into emission lines and absorption lines detected in the high-resolution part 

1668:    of the spectra of some objects (see text).}

1669:  \label{lines}

1670: \end{figure}

1671: 

1672: %\clearpage

1673: 

1674: \begin{figure}

1675: \centering

1676: \epsscale{1.}

1677:     \plotone{f22_color.eps}

1678:    \caption{Correlation between the optical depths of ice features at 15.2 and 6.0\,$\mu$m (left), 

1679:    ices and silicates at 15.2 and 9.8\,$\mu$m (middle), and ice features at 6.8 and 6.0\,$\mu$m (right).

1680:    V1647 Ori is not shown in the middle plot as its silicate feature is seen in emission and not 

1681:    in absorption (see text). }

1682:  \label{absorption_correlation}

1683: \end{figure}

1684: 

1685: 

1686: %\clearpage

1687: 

1688: \begin{figure}

1689: \centering

1690: \epsscale{1.}

1691:     \plotone{f23.eps}

1692:    \caption{Spectral index $n$ of the silicate emission objects 

1693:    evaluated between 13 and 25\,$\mu$m vs. the spectral index between 6 and 13\,$\mu$m.

1694:    $n$ is computed as $n\equiv d\,{\rm log}(\lambda F_{\lambda})/d\,{\rm log}(\lambda)$. 

1695:    All objects populate the same area in the plot as the Class II objects in Taurus 

1696:    presented by \citet{furlan2006} (dashed area). Due to the steep rise in its SED, Parsamian 21 lies slightly 

1697:    off from the other objects.}

1698:  \label{spectral-indices}

1699: \end{figure}

1700: 

1701: 

1702: %\clearpage

1703: 

1704: \begin{figure}

1705: \centering

1706: \epsscale{1.}

1707:     \plottwo{f24a.eps}{f24b2.eps}

1708:    \caption{Simple graphical representation explaining the two 

1709:    categories of FUors (not to scale). 

1710:    Category 1 objects (left) show silicate absorption and are younger than than Category 2 

1711:    objects (right) which show silicate emission. While Category 1 objects are still more deeply embedded in 

1712:    their envelopes and appear to be Class I sources, Category 2 FUors are 

1713:    similar to Class II sources.

1714:    The spectrum of V1647 Ori can be explained assuming that extinction in the line of sight to this object

1715:    creates the ice absorption features and reduces also the strength of the observed silicate emission.  

1716:    }

1717:  \label{categories}

1718: \end{figure}

1719: 

1720: %\clearpage

1721: 

1722: 

1723: 

1724: 

1725: %% Tables should be submitted one per page, so put a \clearpage before

1726: %% each one.

1727: 

1728: %% Two options are available to the author for producing tables:  the

1729: %% deluxetable environment provided by the AASTeX package or the LaTeX

1730: %% table environment.  Use of deluxetable is preferred.

1731: %%

1732: 

1733: %% Three table samples follow, two marked up in the deluxetable environment,

1734: %% one marked up as a LaTeX table.

1735: 

1736: %% In this first example, note that the \tabletypesize{}

1737: %% command has been used to reduce the font size of the table.

1738: %% We also use the \rotate command to rotate the table to

1739: %% landscape orientation since it is very wide even at the

1740: %% reduced font size.

1741: %%

1742: %% Note also that the \label command needs to be placed

1743: %% inside the \tablecaption.

1744: 

1745: %% This table also includes a table comment indicating that the full

1746: %% version will be available in machine-readable format in the electronic

1747: %% edition.

1748: \end{document}

1749: 

1750: %%

1751: %% End of file `sample.tex'.

1752: