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

2: %\documentclass{emulateapj}

3: \usepackage{natbib,amssymb}

4: \newcommand{\mic}{ $\mu$m }

5: \bibpunct{(}{)}{;}{a}{}{,}

6: \usepackage{graphics,graphicx,subfigure}

7: 

8: \begin{document}

9: 

10: \title{Spitzer IRS Observations of Class I/II Objects in Taurus: \\ Composition and Thermal History of the Circumstellar Ices}

11: \shorttitle{Characteristics of Ices in Taurus Class I/II Objects}

12: \shortauthors{G.~Zasowski et al.}

13: 

14: \author{G.~Zasowski\altaffilmark{1},

15: F.~Kemper\altaffilmark{2,1}, Dan~M.~Watson\altaffilmark{3},

16: E.~Furlan\altaffilmark{4,5}, C.J.~Bohac\altaffilmark{3},

17: C.~Hull\altaffilmark{1}, and

18: J.D.~Green\altaffilmark{3}} 

19: 

20: \altaffiltext{1}{Department of Astronomy, PO Box 400325, University of Virginia, Charlottesville, VA 22904-4325: gailis@virginia.edu, chat.hull@gmail.com}

21: \altaffiltext{2}{Jodrell Bank Centre for Astrophysics, University of Manchester, Manchester, M13 9PL, UK: Ciska.Kemper@manchester.ac.uk}

22: \altaffiltext{3}{Department of Physics and Astronomy, University of Rochester, Bausch \& Lomb Hall, PO Box 270171, 500 Wilson Blvd, Rochester, NY 14627-0171: dmw@pas.rochester.edu, joel@pas.rochester.edu, urgradcb@gmail.com}

23: \altaffiltext{4}{NASA Astrobiology Institute and Department of Physics and Astronomy, UCLA, 430 Portola Plaza, Los Angeles, CA 90095: furlan@astro.ucla.edu}

24: \altaffiltext{5}{Current address: Jet Propulsion Laboratory, Caltech, Mail Stop 264-767, 4800 Oak Grove Drive, Pasadena, CA 91109}

25: 

26: \begin{abstract}

27: We present observations of Taurus-Auriga Class I/II protostars obtained with the Spitzer InfraRed Spectrograph.

28: Detailed spectral fits

29: to the 6 and 15.2 micron ice features are made, using publicly-available laboratory data, to constrain the

30: molecular composition, abundances, and levels of thermal processing along the lines of sight.  

31: We provide an inventory of the molecular environments observed, which have an average composition dominated

32: by water ice with $\sim$12\% CO$_2$ (abundance relative to H$_2$O), $\gtrsim$2-9\% CH$_3$OH, $\sim$14\% NH$_3$, $\sim$3\% CH$_4$,

33: $\sim$2\% H$_2$CO, $\sim$0.6\% HCOOH, and $\sim$0.5\% SO$_2$.  We find CO$_2$/H$_2$O ratios nearly equivalent to those observed

34: in cold clouds and lines of sight toward the galactic center.  The unidentified 6.8 micron profiles

35: vary from source to source, and it is shown to be likely that even combinations of the most common candidates

36: (NH$_4^+$ and CH$_3$OH) are inadequate to explain the feature fully.  

37: We discuss correlations among SED spectral indices, abundance ratios, and thermally-processed ice fractions and their implications for CO$_2$ formation and evolution.

38: Comparison of our spectral fits

39: with cold molecular cloud sight-lines indicates abundant prestellar ice environments made even richer by

40: the radiative effects of protostars.  Our results add additional constraints and a finer level of detail to current

41: full-scale models of protostellar and protoplanetary systems.

42: \end{abstract}

43: 

44: \keywords{stars: pre-main sequence --- ISM: abundances --- ISM: molecules --- infrared: ISM --- astrochemistry}

45: 

46: %\notetoeditor{}

47: 

48: \section{INTRODUCTION}

49: 

50: The chemical and thermal evolution of protostellar environments affects the spectral appearance

51: of ices observed in the sequential stages of pre-main-sequence evolution

52: \citep[e.g.][]{WGH_82_protostars,GWE_99_molcloud,vD_06_PNASarticle}. The

53: presence of the ice absorption features is often easily recognized, with the most common ones

54: in the mid-infrared wavelength range found at

55: 6.0, 6.8, and 15.2 $\mu$m, but the exact feature shapes depend strongly on

56: the composition of the ice as determined by the initial condensation

57: and subsequent thermal processing and UV irradiation. 

58: Mixing ices can cause wavelength shifts in the resonances of the individual components within the ice matrix

59: \citep[e.g.][]{EKS_99_laboratory,PB_00_icymixtures,OFB_07_effectco2onh2o}, and

60: differences in lattice structure caused by either radiation damage or thermal processing are spectroscopically identifiable

61: \citep[e.g.][]{HTG_81_waterice,MH_92_waterice,MSR_98_waterice,KK_90_UVamorphization}.  

62: 

63: The initial reservoir of star formation material, the diffuse ISM, is extremely ice-poor \citep[e.g.][]{Gillett_75_DISMice,Whittet_97_DISMdust}.  

64: It is not dense enough to shield its dust grains, upon which ices condense \citep[e.g.][]{TielensAlla_87_ISMdust}, from incident destructive photons.  Only when cold

65: molecular clouds of high density ($n_H$ $\gtrsim$ 10$^4$ cm$^{-3}$) are formed can interstellar ices survive and be observed.  

66: In the Taurus molecular cloud, for example, 

67: a well-known seat of starbirth, ices are observed towards background stars even in quiescent regions unaffected by the newly-formed stars 

68: \citep[e.g.][]{BMG_05_co2fieldstars,KBP_05_backgroundstars,WSB_07_quiescentICM}.  These studies reveal

69: absorptions attributable to H$_2$O, CO, and CO$_2$, as well as some less-constrained absorptions (e.g. the still unidentified band at 6.8 $\mu$m) 

70: due to presumably more complex species.  

71: The ice bands observed towards these field stars are smooth, indicating amorphous and unprocessed compounds, and low upper limits can be placed on some species

72: created by UV radiation \citep[e.g. ``XCN'';][]{GWB_04_ice}.  These observations support 

73: the theory that molecular cloud ice chemistry is dominated by efficient cold grain surface reactions.  The presence of larger complex molecules \citep[such as HCOOH, 

74: which can be formed at low temperatures;][]{Keane_01_makehcooh} demonstrates the potential complexity of these reactions and 

75: indicates that protostars are born into a pre-enriched chemical environment.

76: 

77: Surveys of deeply embedded young stellar objects (YSOs) 

78: reveal a limited range of thermal processing, indicative of differences in YSO environment

79: and temperature \citep[e.g.][]{GWE_99_molcloud,GWB_04_ice}.

80: Studies of less-extinguished protostars \citep[Class I in the classification of][]{Lada_87_starformation} show evidence 

81: of enhanced complex molecule formation (H$_2$CO, XCN, OCS, PAHs [polycyclic aromatic hydrocarbons]), 

82: ice crystallization, further thermal annealing, and ice matrix segregation 

83: \citep[see e.g.][]{ACA_03_isocam,KTB_01_5-8umregion,EDD_98_segregation,WKC_04_taurus,GWB_04_ice,BPL_04_lowmass}.

84: 

85: Finally, the least embedded Class II objects have seemingly accreted or dissipated most of their 

86: cold natal envelopes and have heated the outer layers of their dust disks above the 

87: sublimation temperature of astrophysical ices.  These objects tend to show silicate dust \emph{emission}, PAH emissions, and weak or 

88: nonexistant ice signatures \citep[e.g.][]{Forrest_04_ttauri,Furlan_06_ttauristars}, 

89: though some Class II edge-on systems have ice absorptions arising from the cold disk midplane \citep[e.g.][]{Pontopp_05_diskices,Furlan_08_classImodel}.

90: 

91: Differences in chemical and physical structure exist between YSOs with masses either above or below a few M$_\sun$

92: \citep{vanDishoeck_03_highlowmassYSOs}.  Some of these

93: dissimilarities may be attributed to observational biases (low-mass YSOs are generally closer and can be better resolved), but others are likely to be

94: dependent on the intrinsic nature of the object (e.g. the role of shocks, core temperature, and infall timescales).  

95: For example, low-mass YSO studies claim higher abundances of CO$_2$, relative to H$_2$O \citep{BPL_04_lowmass,Pontop_08_co2inlowmass} 

96: and amorphous silicate dust \citep{WKC_04_taurus}, than surveys of more massive YSOs.

97: This paper focuses on the relatively underexplored categories of low-mass Class I protostars and Class I/II transition objects 

98: (henceforth both groups will be collectively referred to as Class I/II objects)

99: with ice absorptions by analyzing mid-infrared spectroscopy of 16 Taurus-Auriga YSOs \citep[$<$1-3 Myr old;][]{White_01_TMCage}, observed with

100: the InfraRed Spectrograph \citep[IRS;][]{HRV_04_IRS} on board the Spitzer Space Telescope \citep{WRL_04_Spitzer}.

101: Our sample overlaps slightly with that of the extensive recent work by the {\it c2d} team \citep[e.g.][]{Boogert_08_lowmassices,Pontop_08_co2inlowmass,Oberg_08_ch4}, 

102: and we indicate differences in our approach and findings throughout this paper.

103: This study helps complete our view of the ice evolutionary sequence from cold molecular clouds through

104: warmer environments with protoplanetary disks.  These low-mass, possibly early solar analogs are essential pieces to the puzzle of solar system formation.

105: 

106: Section~\ref{sec:obs_reduction} contains a short description of the sample, observations, and data reduction.  In Section~\ref{sec:methods} we describe 

107: how the ice features are identified, isolated, and analyzed, and in Section~\ref{sec:trends} we look at trends seen in the feature optical depths.

108: In Section~\ref{sec:features} we derive the composition of the carriers of the primary ice absorptions and evaluate 

109: candidates for some of the less constrained features.  In Section~\ref{sec:discussion} we discuss composition and spatial distribution of the ices and place

110: our observations in the context of pre-main-sequence evolution.

111: 

112: \section{OBSERVATIONS} \label{sec:obs_reduction}

113: 

114: \subsection{Target Sample Description}

115: 

116: The Taurus-Auriga dark cloud is a nearby \citep[d $\sim$ 140 pc;][]{Bertout_99_TMCdistance}, low density \citep[$n_*$ $\sim$ 1-10 pc$^{-3}$;][]{Gomez_93_TMCdensity} 

117: star forming region containing predominantly young, low-mass protostars.  

118: It provides an opportunity to study closely low-mass star formation unaffected by the shocks and strong UV radiation associated with more massive protostars.  

119: Based on IRAS 25 and 60\mic photometry, an

120: extensive sample of Class I/II objects was selected from the list by \citet{KenHart_95_YSOlist} and observed with Spitzer's IRS (Section~\ref{sec:obsprocedure}). 

121: From this sample we chose for analysis the subset of 12 Class I and 4 Class II objects \citep[as classified in][]{KenHart_95_YSOlist} 

122: with distinct ice features in regions of unsaturated, well-calibrated data.  

123: %This set contains 5 spectra (IRAS 04016+2610, IRAS 04108+2803B, IRAS 04181+2654B, IRAS 04239+2436, DG Tau B)

124: %previously published by \citet{WKC_04_taurus}.  

125: The IRAS spectral energy distribution (SED) points used in the original selection are insufficient to determine system inclination, 

126: but modeling by \citet{Furlan_08_classImodel} indicates that 

127: our smaller sample spans a wide range of disk inclination angles, from nearly edge-on to face-on systems.

128: 

129: For comparison, we include a secondary sample comprising observations of Taurus and Serpens cloud field stars---Elias 16, Elias 13, and CK 2---as well 

130: as of the Class 0 protostars CepE-MM and L1448 IRS 2 (AOR IDs 3623680, 3620352).

131: We also include data from the line of sight towards the W3(OH) star formation region, which probes part of the foreground W3 molecular cloud (AOR ID 3566336).

132: 

133: \subsection{Observational Procedure and Data Reduction} \label{sec:obsprocedure}

134: Our sample was observed with Spitzer's IRS as part of the \emph{IRS Disks} project (PI: J.~Houck \& D.~Watson; PID 2)

135: during guaranteed time awarded to the IRS instrument team (2004 February-March).  All of our targets were observed using the full 5--37 $\mu$m

136: wavelength coverage, combining either the Short Low and Long Low (SL, LL; 5.2-14 $\mu$m, 14-38 $\mu$m; $\lambda$/$\Delta\lambda\,\sim$ 90)

137: modules on the IRS, or the SL with the Short High and

138: Long High (SH, LH; 10-19 $\mu$m, 19-37 $\mu$m; $\lambda$/$\Delta\lambda\,\sim$ 600) modules, depending on the expected flux levels. 

139: 

140: Most (12) of the objects were observed using a small spectral map ($3\times

141: 2$) rather than a single pointing; this strategy avoids the need for pointing peak-ups and allows, in principle, for a better reconstruction of

142: the spectrum in cases of small mispointings.

143: The three-step sequence was carried out in the dispersion direction, with 

144: the steps separated by three quarters (for the SL module) or one half (for the other modules) of the slit width.

145: The two-step sequence corresponds to the normal IRS nod distance of one-third the slit length.

146: The other 4 objects were observed in the IRS staring mode, using only the two nods along the slit.

147: Observational details of all our Class I/II targets are listed in Table~\ref{tab:obs}.

148: 

149: We analyzed data from pipeline version S13, using

150: the {\sc smart} package \citep{HDH_04_smart}.  The standard {\it IRS Disks} reduction procedures, as well as any spectra requiring special treatment, are 

151: described in \citet{Furlan_06_ttauristars} and \citet{Furlan_08_classImodel}.

152: 

153: The uncertainty in each wavelength bin is taken to be half of the nod difference in flux level, except where this value is smaller than 1\% of the flux.  In those cases,

154: the uncertainties are taken to be 1\% of the flux in order to prevent underestimation.  We estimate an absolute spectrophotometric accuracy of 5-10\%.

155: 

156: For the field stars, we use the previously published data for 

157: Elias 16 \citep{BMG_05_co2fieldstars} and Elias 13 and CK~2 \citep{KBP_05_backgroundstars}, without re-reducing the data.

158: 

159: \subsection{IRS Spectra Description}

160: 

161: In Figure~\ref{fig:fullspectra}, we show the IRS spectra of the 16 Class I/II stars in our sample.

162: The YSO mid-infrared spectra are dominated by broad solid state resonances, particularly those at 6.0, 6.8, 9.7, and

163: 15.2 $\mu$m.  Additional, but less pronounced, features are visible in some spectra at 7.7\mic and as a broad band centered at 18 $\mu$m.  Silicates are responsible

164: for the broad 9.7 and 18\mic features, which can occur in either absorption (e.g. L1551 IRS 5), emission (e.g. IRAS 04108+2803A), or both.

165: 

166: The 6.0\mic feature is primarily due to the O-H bending mode of water ice, while the 6.8\mic feature remains largely

167: unidentified, with CH$_3$OH and NH$_4^+$ (among others) as suggested carriers.  The 7.7\mic absorption is generally attributed to solid CH$_4$, and

168: the 15.2\mic band is due to the O-C-O bending mode of solid CO$_2$.

169: A complete inventory of absorption features identified in our spectra is contained in Table~\ref{tab:features}.

170: 

171: Spectra for W3(OH) and the Class 0 objects (Figure~\ref{fig:otherobjects}) contain many of the same features.

172: 

173: \section{METHODS} \label{sec:methods}

174: 

175: \subsection{Isolating the IR Features and Determining Optical Depths} \label{sec:isolate_features}

176: 

177: \subsubsection{YSO Spectra} \label{sec:isolate_features_irs}

178: 

179: The optical depth, $\tau$, and shape of the individual absorption features

180: under consideration are extracted from the astronomical spectra

181: using a mathematical continuum fit.  

182: We make a cubic spline fit to the $\sim$5-20\mic continuum of each

183: spectrum using the following relatively feature-free regions: bluewards of the 6\mic H$_2$O feature; between the unidentified 6.8\mic feature

184: and the onset of the 9.7\mic silicate band; between the red wing of the silicate band ($\sim$14.5 $\mu$m) and the 15.2\mic CO$_2$ feature;

185: and redwards of the CO$_2$ feature---while ensuring that

186: none of the feature wings are included in the continuum.

187: We calculate the optical depth in the spectral features using $F_\lambda=F_{\lambda,cont}e^{-\tau_\lambda}$.  

188: 

189: Because the 9.7\mic silicate feature is highly complex, and

190: detailed modeling of the silicate mineralogy is beyond the scope of this paper, accurate extraction of ice features blended with the 9.7\mic silicate complex 

191: (e.g. those of CH$_3$OH, NH$_3$, and H$_2$O) is impractical.

192: However, the 11-13\mic H$_2$O libration band and

193: the 18\mic absorption feature in particularly silicate-rich spectra are much broader and shallower than the 15.2\mic ice absorption they overlap, so 

194: we can treat these shapes as pseudo-continua in order to isolate the CO$_2$ ice absorption.

195: 

196: Additional steps are needed to extract small features in the 7-8\mic range.  The wide variation in 6.8\mic and

197: silicate absorption profile wings requires that each spectrum be treated individually.

198: For objects not affected by the silicate feature bluewards of 8 $\mu$m, a straight-line continuum ($\sim$7.1-8 $\mu$m) is imposed to isolate any features.  

199: For objects affected by both the 6.8 and the 9.7\mic features, two straight-line continua are imposed: 

200: one from $\sim$7.1-7.5\mic and one from $\sim$7.5-8 $\mu$m (Figure~\ref{fig:continuum-ex}).

201: 

202: \subsubsection{Laboratory Spectra} \label{sec:isolate_features_lab}

203: 

204: We use laboratory data from the Leiden University ice analog databases\footnote{http://www.strw.leidenuniv.nl/$\sim$lab/databases/databases.htm}

205: \citep{GSG_95_labsim,Gerakines_96_puredb,EBG_96_coco2iso,EKS_99_laboratory,vBGF_06_coco2mixes} to identify the composition and apparent temperatures of

206: the ices seen in our targets. In order to provide a direct comparison between

207: the feature optical depths measured in the laboratory and those observed in the

208: astronomical spectra, we removed a straight-line continuum from the laboratory data.  For

209: the wavelength ranges of interest, optical depths calculated using a straight-line continuum differ, on

210: average, by less than half of a percent from those calculated using a more complicated cubic spline

211: one.  The wavelengths of the baseline ends are

212: generally equal to the continuum points used in the astronomical

213: spectra's spline fitting, with an occasional

214: slight shift to ensure feature wings are excluded from the continuum.  In Figure~\ref{fig:h2oco2_temp} we show

215: the range of abundances contained in the Leiden databases' non-irradiated H$_2$O:CH$_3$OH:CO$_2$

216: mixtures (T=10-185K, all circles); note that this plot does not represent the fraction of the

217: database comprising other molecules (e.g. CH$_4$ and H$_2$OH) or UV-irradiated mixtures.

218: 

219: The temperatures given in this paper, except where otherwise noted, refer to laboratory conditions.  Interstellar temperature barriers for crystallization, sublimation,

220: and other physical processes are lower than their laboratory counterparts, due to the lower pressure and longer interaction timescales in interstellar space

221: \citep[e.g.][]{BEG_00_13co2}.

222: Assumptions about particle shapes have not been applied to the laboratory spectra used in this broad analysis, and such corrections would in any case

223: be inappropriate for the annealed inhomogeneous mixtures used \citep{GWE_99_molcloud}.

224: 

225: \subsection{Spectral Fitting} \label{sec:fitting}

226: To confirm the validity of using laboratory spectra to model circumstellar material, we estimate the total extinction $A_V$ along the line of sight towards

227: our YSOs using $A_V$/$\tau_{Si}$ $\sim$ 17 \citep{Rieke_85_extinctionlaw}.  We find $A_V$ $\sim$ 3-34 for our total sample---and $A_V$ $\sim$ 6-34 for those objects

228: without clear signs of dust emission contaminating the absorption feature---compared with the average 

229: extinction of $A_V$ $\sim$ 1-5 toward Taurus association stars \citep{Myers_83_densecores}.  

230: In addition, recent work has shown that in dense regions, this relation with a given $\tau_{Si}$ substantially 

231: underestimates $A_V$ \citep{Whittet_88_extinctionlaw,Chiar_07_extinctionlaw}, so the total circumstellar extinction is likely to be even higher.

232: 

233: Thus, we assume that the foreground absorption towards the Taurus region is relatively low and that the spectral ice features result from material in the 

234: circumstellar environment.

235: In contrast with a least-components type of analysis \citep[e.g.][]{Boogert_08_lowmassices,Pontop_08_co2inlowmass}, 

236: we determine the ice composition and column densities of the circumstellar ice

237: by fitting linear combinations of laboratory data to the observed IRS spectra.

238: With the laboratory data described in Section~\ref{sec:isolate_features_lab}, detailed spectral fitting is performed on the 6.0 and 15.2\mic features using a standard 

239: $\chi^2$-minimization technique.  

240: The equation solved is

241: \begin{equation} \label{eqn:fit}

242: \tau(\lambda)=\sum{a_i\tau_{lab,i}(\lambda),}

243: \end{equation}

244: where the scale factors $a_i$ of each component are the free parameters in the fit.  To compare goodness-of-fit among various combinations, we use $\chi^2_\nu$, the 

245: reduced $\chi^2$ parameter.  This is defined as $\chi^2_\nu$=${\chi^2}$/(${n-m}$), where $n$ is the number of spectral data points 

246: and $m$ is the number of free parameters in the fit.

247: Traditional statistics require $\chi^2_\nu\sim1$ for an acceptable fit, but our values are often significantly larger or smaller.  Artificially small $\chi^2_\nu$

248: values can be caused by uncertainty overestimation or by poor signal-to-noise.  Large values may be a result of uncertainty underestimation, unknown 

249: uncertainties in the laboratory spectra, or the existence of components not included in the fits.  

250: Therefore $\chi^2_\nu$ comparisons may be made to determine

251: the best laboratory fit to a single spectrum; however, comparisons of goodness-of-fit among different YSO spectra are not necessarily valid.  

252: Quantities derived using a fitted spectral feature (e.g. processed fractions, column densities) are calculated using an average of all fits for that feature with a

253: $\chi^2_\nu$ within 20\% of the minimum.

254: 

255: \subsection{Column Densities} \label{sec:column}

256: To calculate column densities along the lines of sight to the YSOs, we use the relation \citep[e.g.][]{ACA_03_isocam}

257: \begin{equation} \label{eqn:column}

258: N(cm^{-2})=\frac{\sum{\tau_\lambda\Delta\lambda}}{\lambda_{peak}^2A},

259: \end{equation}

260: where $\Delta\lambda$ is the wavelength bin size, $\lambda_{peak}$ is the wavelength of the feature's peak optical depth, and

261: $A$ (cm molec$^{-1}$) is the band strength as determined in the laboratory.  

262: 

263: \subsection{SED Spectral Index} \label{sec:alpha}

264: To provide better resolution in evolutionary status than the simple \citet{Lada_87_starformation} classification, we use the

265: spectral index $\alpha$ \citep[e.g.][]{KenHart_95_YSOlist}:

266: \begin{equation} \label{eqn:alpha}

267: \alpha\equiv\frac{d\log{\lambda F_\lambda}}{d\log{\lambda}}\sim\frac{\log{\lambda_2F_{\lambda_2}}-\log{\lambda_1F_{\lambda_1}}}{\log{\lambda_2}-\log{\lambda_1}},

268: \end{equation}

269: where we choose $\lambda_1$ and $\lambda_2$ to lie within well-calibrated and nearly-featureless

270: parts of the spectrum but as much separated as possible (i.e. $\lambda_1 = 7$ $\mu$m and $\lambda_2 = 30$ $\mu$m; Table~\ref{tab:obj}).

271: The slope of this section of the spectrum depends upon a variety of factors, including disk radius, total envelope density, outflow cavity

272: opening angle, and inclination angle.  With the exception of inclination angle, all of these factors depend upon the total dust column density in such a

273: way that more dust produces a steeper SED slope, and models demonstrate that inclination angle effects are generally less significant than these other factors

274: \citep{Furlan_08_classImodel}.  Thus small values for the spectral index $\alpha$ indicate a relatively blue or exposed stellar object, but with increasing

275: $\alpha$ the spectrum becomes redder, indicating the YSO is more heavily embedded in dust.

276: 

277: \section{OPTICAL DEPTH TRENDS} \label{sec:trends}

278: 

279: In Figure~\ref{fig:objratios}, we show correlations in the optical depths of the three strongest absorptions

280: attributed to ices (at 6.0, 6.8, and 15.2 $\mu$m) in our YSO sample.

281: Clearly, the 6.0 and 6.8 $\mu$m features are 

282: very closely correlated, with linear Pearson correlation coefficient $r=0.96$ (Figure~\ref{fig:objratios}a).  

283: The peak ratio is $\tau_{6.8}$/$\tau_{6.0}$ $\sim$ 0.96$\pm$0.03,

284: and the straight-line fit passes within 0.01 of the origin.

285: Since the 6.0 $\mu$m band is

286: primarily attributed to the bending mode of H$_2$O, this correlation indicates

287: that the carrier(s) of the 6.8 $\mu$m band forms and evolves in a similar, polar environment.   

288: 

289: There is a looser correlation between the 6.0 $\mu$m and the 15.2\mic optical depths in our 

290: YSOs ($r=0.7$, Figure~\ref{fig:objratios}b), resulting in a wider ratio range $\tau_{6.0}$/$\tau_{15.2}$ $\sim$ 0.43-1.65.  

291: When the same ratios are calculated for the laboratory spectra, though the spread is even wider (0.01-3.6), only a small fraction

292: fall within the range defined by the YSOs (Figure~\ref{fig:h2oco2_temp}, filled circles); the 15.2\mic CO$_2$ feature is generally much stronger than the 

293: 6.0\mic water ice feature, independent of temperature. Mixtures dominated by H$_2$O provide most of the closest matches.

294: In addition, there are two lab mixtures of nearly 1:1:1 H$_2$O:CH$_3$OH:CO$_2$ abundances (center of Figure~\ref{fig:h2oco2_temp}) which have 

295: $\tau_{6.0}$/$\tau_{15.2}$ ratios comparable to the YSO ones but only after they have been heated to high temperatures (T $\gtrsim$ 125 K),

296: nearing the ice sublimation points.

297: The Leiden database has a scarcity of water-rich, CO$_2$-poor ($\lesssim$20\%) H$_2$O:CH$_3$OH:CO$_2$ mixtures in the temperature range

298: 10-180 K, though based on this comparison, these compositions appear to be astronomically relevant.  Because of the undersampling of compositions in this range, we

299: are not able to constrain the ice composition well from the $\tau_{6.0}$/$\tau_{15.2}$ ratio alone.

300: 

301: Despite being closely correlated with $\tau_{6.0}$, the 6.8\mic peak optical depth is even less correlated with the 15.2\mic depths 

302: ($r=0.56$, Figure~\ref{fig:objratios}c).  This indicates that much of the inherent scatter in the 

303: $\tau_{6.8}$/$\tau_{6.0}$ ratio is due to variations in the 6.8\mic feature among different lines of sight.

304: 

305: \section{SPECTRAL FEATURE ANALYSES} \label{sec:features}

306: 

307: In order to set more accurate constraints on the ice properties than feature-strength analysis provides, we perform spectral fitting and detailed analysis

308: of the most prominent features.

309: 

310: \subsection{The 6.0 $\mu$m Feature} \label{sec:6um_feature}

311: 

312: Each YSO spectrum is fitted over the wavelength range 5.4-6.3\mic using the method described in 

313: Section~\ref{sec:fitting} and the component set listed in Table~\ref{tab:6um_components}.  An example result is shown in Figure~\ref{fig:6umfit}, and the total

314: sample residuals, after subtraction of the 6\mic components, are shown in Figure~\ref{fig:6.8um_spectra}.  

315: Due to overlap with the unidentified 6.8\mic feature, the range beyond 6.3\mic was excluded in order to avoid imposing artificial structure on any fit residuals.  

316: These fits focus on identifying the constituents of the 6.0\mic feature---H$_2$O, H$_2$CO, HCOOH, and NH$_3$, as described in the following subsections.

317: This suite of species is similar to that employed by \citet{Boogert_08_lowmassices} to explain their extracted component set C1-5.

318: 

319: Either noise or saturation of the absorption bands in the background star/Class 0 sample prevents detailed spectral fitting of the 6\mic features, 

320: so we scaled a pure H$_2$O (10 K) spectrum to match the peak optical depth at 6.0 $\mu$m in order to calculate $N$(H$_2$O) (Table~\ref{tab:molcloud_abun}).

321: The shape of the 6\mic W3 spectrum is too irregular to be reasonably fit with a solid H$_2$O lab spectrum in this manner.  

322:  

323: \subsubsection{H$_2$O} \label{sec:h2o}

324: 

325: Water is the least volatile and most abundant of the ices typically found along protostellar lines of sight. The mid-infrared spectrum offers three strong features of 

326: water ice: narrow ones at 3.1 and 6.0 $\micron$, and a broad ($\lambda$ $\sim$ 11-20 $\mu$m) feature peaking at 13 $\micron$. 

327: To turn these profiles into water ice column 

328: density is somewhat problematic in each case. The 3.1 $\micron$ ice feature is generally regarded as the cleanest measure of H$_2$O ice column density 

329: \citep[e.g.][]{Tielens_84_5-8micspectra}. 

330: However, inconvenient gaps in the spectrum transmitted by Earth's atmosphere make it difficult to determine the continuum on the short wavelength side of this 

331: feature in ground-based observations, which would be the source of most of the comparisons we could make to our data. In addition, rather significant extinction and 

332: scattering near 3 $\micron$ are inevitably associated with protostars and can affect the profile of the water ice feature. 

333: Futhermore, this may lead to the bulk of the 3\mic background continuum taking a significantly different path through the protostellar envelope 

334: than that which gives rise to the longer-wavelength features. The latter 

335: effect would be rather more serious in systems viewed closer to edge-on, such as IRAS 04169+2702 and DG Tau B. 

336: 

337: Extinction at 6 $\micron$ is similar to that at 13 $\micron$, and both are significantly smaller than that at 3 $\micron$ \citep[extinction in magnitudes smaller by 

338: factors of $\sim$ 2-3;][]{Mathis_90_extlaw,Indeb_05_extlaw}. 

339: Thus the optical paths sampled at 6 and 13 $\micron$, and the ice column densities derived, should be similar and may be 

340: more characteristic of the line of sight to the embedded protostar than is the case at 3 $\micron$. 

341: %The disadvantages of the 6 and 13 $\micron$ water ice features are different from each other. 

342: In observed spectra, the 6 $\micron$ feature is the easiest of the three to analyze, as there are fewer problems in determination of the continuum. However, the 

343: depth and profile of the 6 $\micron$ feature are possibly influenced by absorption from species other than water ice, as shown, 

344: for example, by \citet{Boogert_08_lowmassices}. The 

345: 13 $\micron$ water ice libration feature is broad and overlaps many other absorptions, including the loosely-constrained 9.7\mic silicate feature, 

346: so its optical depth and feature shape are hard to measure precisely.

347: %{\bf In addition, pure H$_2$O ice laboratory spectra from the Leiden database show a correlation between temperature and the 13\mic feature strength, for nearly-constant

348: %$N$(H$_2$O); this indicates a strong 13\mic band-strength dependence on temperature (far exceeding that of the 6\mic band) that renders this feature even less suitable

349: %for column density calculations.}

350: 

351: Taking account of all of these difficulties, we choose to measure water ice column density with the 6 $\micron$ feature, and check against the possibility of substantial 

352: ``contamination'' from other species using the 13\mic feature.  

353: To calculate the H$_2$O column density, we use the 6.0 $\micron$ feature fit results along with 

354: Equation 2 (summed over $\lambda$=5.5-7 $\mu$m), with a band strength of $A_{H_2O}$=1.2$\times$10$^{17}$cm molec$^{-1}$ for the

355: T = 50 K water ice components and $A_{H_2O}$=1.3$\times$10$^{17}$cm molec$^{-1}$ for the T = 120 K one. These values pertain to pure ${\rm H_2 O}$ ice, though they are 

356: also suitable for apolar dilutant concentrations of $\lesssim$30\% \citep{GSG_95_labsim}. We note that the derived column densities $N({\rm H_2 O})$ (Table 5) may in 

357: some cases represent upper limits, as the overall composition of the 6 $\micron$ feature is still a source of discussion \citep[e.g.][]{GWB_04_ice,Boogert_08_lowmassices}. 

358: However, as we will see below, the consistency of these results with those from 13 $\micron$ features in our spectra indicates that the 

359: contribution of unaccounted-for species at 6 $\micron$ is rather small compared to that of water ice. 

360: In general, our values of water ice column density are somewhat larger than those derived for the same objects from 

361: the 3.1 $\micron$ feature (e.g. Boogert et al. 2008), and tend to depart more the closer the object is to an edge-on view, 

362: so the differences may in part be due to the expected differences between the light paths viewed at the two wavelengths.

363: 

364: To look for signs of additional ice components with features near 6 $\micron$ that are not resolved spectrally from the water-ice feature, we have measured the 11-20 

365: $\micron$ solid H$_2$O absorption in the nine targets for which the 10 $\micron$ silicate feature is most deeply in absorption. 

366: To do so, we followed a procedure similar to 

367: that by \citet{Boogert_08_lowmassices}. From a low-order polynomial fit to the continuum, constrained to follow the spectrum at 5.3-5.6 and 30-35 $\micron$, we determined 

368: the total optical depth in the 8-30 $\micron$ range. To this we ``fit'' the profile of interstellar silicate absorption, exemplified by the galactic center source GCS 3 

369: \citep{Kemper_04_silicates} and constrained to match the observed spectrum at 9.9-10.1 $\micron$ and 23-30 $\micron$. 

370: Since this interstellar silicate profile may not be fully applicable to circumstellar silicates, this step introduces additional uncertainty.

371: Subtraction of the resulting profile from the optical 

372: depth spectrum reveals the water ice libration absorption, peaking at $\sim $13 $\micron$, and interrupted at 15.2 $\micron$ by the $\rm{CO}_2$ ice feature. We measured 

373: from such results the peak optical depth in this water ice feature around 13 $\micron$. By small variation of the wavelengths at which the continuum and silicate profile 

374: were constrained to follow the spectrum, we estimated the uncertainty in the peak optical depth. Because the shape of this very broad water feature is more sensitive to 

375: both the continuum and silicate profile fits than is the peak optical depth, the corresponding water 

376: ice column density is considerably more uncertain.  In Figure~\ref{fig:libh2o}, 

377: the values for peak optical depth are plotted against the column density $N$(H$_2$O) derived from the 6 $

378: \micron$ feature as described above. Evident in Figure~\ref{fig:libh2o} is a good correlation between the 13 and 6 $\micron$ water signatures, with Pearson's $r~=~0.74$.

379: That the correlation is tight indicates that, if an additional ice component is significant at 6 $\micron$, 

380: its abundance relative to water ice must be essentially constant.

381: 

382: \subsubsection{H$_2$CO and HCOOH} \label{sec:h2co_hcooh}

383: Many of the objects in our sample have additional absorptions centered around $\sim$5.8\mic that can be well-fit by a combination of 

384: H$_2$CO (formaldehyde) and HCOOH (formic acid) ices.  Solid H$_2$CO is observed in many protostellar systems 

385: \citep[e.g.][]{KTB_01_5-8umregion,GWB_04_ice} and plays an important role in the production of complex organic molecules \citep{Schutte_93_h2co}.

386: H$_2$CO is difficult to detect independently in YSO Spitzer spectra; its

387: 3.47\mic C-H stretch feature is beyond the wavelength range of the IRS, and its weaker 6.68\mic C=O stretch feature is blended with the 6.8\mic absorption,

388: itself not well understood.  The H$_2$CO column densities (Table~\ref{tab:abundances}) are derived 

389: using the observed integrated optical depth of the 5.8\mic feature (as determined from

390: the spectral fits) and a band strength of $A_{H_2CO}$=9.6$\times$10$^{-18}$cm molec$^{-1}$ \citep{Schutte_93_h2co}.

391: 

392: Solid and gas-phase HCOOH has been detected in both low- and high-mass star-forming regions 

393: \citep[e.g.][]{Cazaux_03_hcoohdetect,Schutte_99_7.2/7.4mic,Bottinelli_04_hcoohdetect,Remijan_04_hcoohdetect,KTB_01_5-8umregion}.

394: Within the IRS wavelength range, there are solid-state HCOOH absorptions at 5.8 and 7.2 $\mu m$, and several in

395: the 8-11\mic range \citep{Bisschop_07_hcooh}.  The latter ones are blended with the 9.7\mic silicate complex, but the 7.2\mic band,

396: though intrinsically $\sim$4 times weaker than the 5.8\mic resonance, is detectable and indeed is 

397: tentatively identified in 6 of our 16 objects (Section~\ref{sec:7-8mic}).  We have derived HCOOH column densities (Table~\ref{tab:abundances}) by integrating the 

398: observed optical depth of the 5.8\mic feature (as determined from the spectral fits) and using a band strength of $A_{HCOOH}$=6.7$\times$10$^{-17}$cm molec$^{-1}$ 

399: \citep{Marechal_87_hcooh}.

400: 

401: \subsubsection{NH$_3$} \label{sec:nh3}

402: Solid NH$_3$ (ammonia) has mid-infrared features at 6.15\mic and 9.3 $\mu$m \citep[e.g.][]{SA_93_ch3ohnh3}, 

403: blended with the 6\mic water ice and 9.7\mic silicate features, respectively.  The effect on the 6\mic H$_2$O feature is negligible until the NH$_3$ concentration 

404: reaches $\sim$9\% \citep{KTB_01_5-8umregion}.  At higher concentrations, an excess appears around 6.15\mic on the red side of the 6\mic band in laboratory spectra,

405: which fits well the broad 6\mic feature observed in many of our YSO spectra.  

406: Analyses of the 9\mic inversion mode resonance by \citet{Lacy_98_frozennh3} and

407: \citet{GWS_00_W33A} indicate that circumprotostellar NH$_3$ exists primarily in an H$_2$O-dominated matrix.

408: Therefore, an H$_2$O:NH$_3$ (1:0.2) mixture, rather than pure NH$_3$,

409: is used in the spectral fits.  Because no satisfactory band strength measurements for NH$_3$ in such a mixture exist,  

410: the upper limit NH$_3$ column densities (Table~\ref{tab:abundances}) are estimated by using the band strength 

411: of pure water ice for the H$_2$O:NH$_3$ mixture and extrapolating

412: the column density of NH$_3$ based on its fractional abundance in the final fit.

413: 

414: \subsubsection{PAHs} \label{sec:remnants}

415: 

416: After contributions due to H$_2$O, H$_2$CO, HCOOH, and NH$_3$ are subtracted, small residuals in the 6.1-6.4\mic range, varying in strength, shape, and peak position,

417: are observed in 11 of the sample's 16 YSOs (Figure \ref{fig:6.8um_spectra}).  PAH absorption has been suggested as a candidate for the 6.2-6.3\mic 

418: feature observed in many massive protostars \citep{Schutte_96_6/6.8inngc7538}.  Using the average interstellar PAH spectrum

419: compiled by \citet{Hony_01_PAHspectrum} as a reference, we examined the peak position and widths of the YSO residuals.  Of the objects within

420: our sample with excess absorptions peaking between 6.2 and 6.3 $\mu$m, only three (IRAS 04361+2540, DG Tau B, and HL Tau) have an excess with a width similar to 

421: that seen in the PAH spectrum, and none of these three match the PAH feature in shape and peak position.  

422: Though the PAH relative band strengths often depend upon molecular environment and ionization level, the broad 7.5-9\mic complex and the 6.2\mic feature are 

423: consistently observed at comparable strength in astronomical environments \citep{Hony_01_PAHspectrum}, yet we find no correlation

424: in the YSO spectra between the 6.2 and 7.7\mic optical depths---the ratios are scattered between 0 and 7.5.

425: Moreover, the prominent but narrow PAH peak at 11.3\mic \citep[consistently $\sim$1.4$\times$ as strong as the 6.2\mic peak;][]{Chan_00_PAHratios,Hony_01_PAHspectrum} 

426: and the lesser ones at 12.7, 13.5, and 14.3\mic are not observed in the YSO data.  Thus in agreement with \citet{KTB_01_5-8umregion}, 

427: we find it unlikely that PAHs contribute significantly around 6.2-6.3 $\mu$m.

428: 

429: \subsection{The 6.8 $\mu$m Feature} \label{sec:6.8um_feature}

430: 

431: Though the 6.8\mic interstellar absorption band was discovered nearly 30 years ago \citep{Puetter_79_6.8umdiscovery}, 

432: its identification remains ambiguous and inconclusive.  

433: Species ranging from hydrocarbons to alcohols to carbon dust have been proposed to contribute to the feature in a variety of sources

434: \citep[see][]{KTB_01_5-8umregion}.

435: Here we focus on CH$_3$OH and NH$_4^+$---the candidate carriers most commonly accepted as explaining at

436: least part of the absorption \citep[e.g.][]{BE_04_ISI}.

437: 

438: \subsubsection{CH$_3$OH}

439: 

440: The position of the 6.8\mic feature coincides with the C-H deformation mode of alcohols, and of these, methanol is the most likely to be found in interstellar abundance.

441: Theoretical models of ice processes predict the creation of solid methanol through the sequential hydrogenation of CO ice which has been accreted 

442: from the gas phase onto dust grains \citep{Tielens_97_methproduction}.  

443: The presence of methanol ice in interstellar and protostellar environments is observationally supported by detection of

444: its own resonances---e.g. the C-H stretching mode at

445: 3.54\mic \citep{Grim_91_methdetect} and the C-O stretch at 9.8\mic \citep{GWS_00_W33A}---as well as by signatures of its interactions with other molecules

446: \citep[especially CO$_2$; e.g.][]{EDD_98_segregation}. 

447: 

448: Our sample spectra do not contain the NIR wavelengths of CH$_3$OH's multiple stretching modes, and the 9.8\mic resonance is blended with the silicate

449: absorption.  However, constraints on the CH$_3$OH abundance can be derived using the 15.2\mic CO$_2$ band, as the interaction of CH$_3$OH with CO$_2$ changes

450: the feature shape near 15.4\mic (Sections~\ref{sec:15um_feature} and \ref{sec:chemdiscuss}).  The relative strength of the 15.4\mic shoulder implies

451: that at most 50\% of the 6.8\mic feature is due to CH$_3$OH in a CO$_2$ matrix, although CH$_3$OH unmixed with CO$_2$ could be present

452: to provide additional absorption at 6.8 $\mu$m. 

453: Since CH$_3$OH is thought to form in a

454: hydrogenated environment, the assignment of at least part of the 6.8\mic absorption to this molecule is then supported by 

455: the close correlation of the 6.0 and 6.8\mic optical depths (Section~\ref{sec:trends}).

456: 

457: \subsubsection{NH$_4^+$}

458: The NH$_4^+$ ion was proposed as the carrier of the 6.8\mic band by \citet{Grim_89_nh4plus}, and since then, this proposal has achieved varying levels of success

459: \citep[e.g][]{Schutte_96_6/6.8inngc7538,KTB_01_5-8umregion,Schutte_03_nh4plus_6.8um}.  The ion can be produced in

460: the laboratory by the acid-base reaction HNCO + NH$_3$ $\rightarrow$ OCN$^-$+ NH$_4^+$ \citep{Keane_97_makenh4plus} or

461: in UV-irradiated H$_2$O:CO:NH$_3$ mixtures (as in Figure~\ref{fig:6.8um_spectra}).  

462: 

463: NH$_4^+$ has expected counter-ion features at 4.62\mic (OCN$^-$) and 6.3 and 7.41\mic (HCOO$^-$, produced by photolysis).  The 4.62\mic 

464: feature does not fall within the IRS wavelength coverage, and the 7.41\mic signature is too weak to be conclusively identified in our data.

465: However, some of our sample

466: objects do have a small excess absorption near 6.3 $\mu$m (Section~\ref{sec:remnants}), which is unlikely to be attributable to PAHs.  

467: It may instead be due to HCOO$^-$, indicating photolysis and 

468: subsequent warming of a polar ice containing NH$_3$.  We use the ratios of the laboratory 6.3\mic (HCOO$^-$) and 6.8\mic (NH$_4^+$) peaks 

469: ($\tau_{6.8}/\tau_{6.3}\sim1.7$ after the H$_2$O feature is removed), along with the equivalent widths of the YSO and 

470: laboratory features, to estimate the NH$_4^+$ contribution to the 6.8\mic feature.  Using this method, we find that, within our sample, 6-58\% 

471: of the 6.8\mic feature can be attributed to NH$_4^+$ (HL Tau is an outlier whose particularly large 6.3\mic excess indicates $\lesssim$91\% of its

472: 6.8\mic feature may be due to NH$_4^+$ using this method).

473: 

474: \subsubsection{Fits and Summary} \label{sec:6.8micfits}

475: 

476: Figure~\ref{fig:6.8um_spectra} shows a 

477: comparison of the 6.8\mic features, after subtraction of the 6\mic contributors, with a variety of 6.8\mic candidate laboratory spectra,

478: all rebinned to the resolution of the observations.  The bottom two lab spectra (Lab B and C) have CH$_3$OH/H$_2$O abundance ratios of 100 and 0.1,

479: and the gray vertical lines indicate the peak positions of these mixtures.  The relative amount of methanol present affects the peak substructure.

480: While in some cases the YSO peak positions and relative strengths closely resemble the laboratory data (e.g. IC 2087 IR and IRAS 04239+2436),

481: in others there is no correlation at all (e.g. IRAS 04169+2702).

482: The abundance of CH$_3$OH (relative to H$_2$O) has been found to vary by almost an order of magnitude among different protostellar lines

483: of sight \citep[e.g][]{Pontopp_03_methdetect,GWB_04_ice}, which may partly explain the observed variety of peak substructures.

484: 

485: The top laboratory spectrum in Figure~\ref{fig:6.8um_spectra} (Lab A) is a H$_2$O:CO:NH$_3$ deposit, 

486: irradiated to create NH$_4^+$ and subsequently warmed from 10 to 165 K.

487: The irradiated spectrum at 10 K peaks 0.1\mic bluewards of our sample's bluest 

488: feature; in order to align with any of the observed

489: profiles, the irradiated mixture must be heated to near the H$_2$O sublimation temperature,

490: as the sublimation appears to cause the redward peak shift \citep{Schutte_96_6/6.8inngc7538}.  Moreover, the width of

491: the NH$_4^+$ feature is too narrow ($\sim$0.25 $\mu$m) to account for the entire 6.8\mic feature widths observed towards each YSO (average $\sim$0.4 $\mu$m).  Even if

492: multiple NH$_4^+$ laboratory spectra at various temperatures are combined in order to reproduce the observed feature width, 

493: the resultant peak position remains too blue to match the YSO data.

494: 

495: Noting that many of the YSO 6.8\mic features share characteristics with the profiles of both CH$_3$OH (triple peaks, 7\mic shoulder) and NH$_4^+$ (asymmetric peak near

496: 6.75 $\mu$m), we attempted fitting them with a combination of the laboratory components.

497: While no remarkably good fits were achieved, there emerged a general trend that equal contributions from NH$_4^+$ and from CH$_3$OH in a water-dominated matrix, 

498: with varying amounts of nearly-pure CH$_3$OH intermixed, came closest to reproducing many of the 

499: YSO feature width-to-depth ratios (i.e. mixtures of irradiated NH$_3$, CH$_3$OH/H$_2$O$\sim$0.1, and CH$_3$OH/H$_2$O$\sim$100 

500: in the approximate ratios 1:1:0 to 1:1:1).  

501: However, the overall 6.8\mic astronomical feature shapes could not be be well-matched in terms of either peak alignment or subpeak relative strengths.

502: 

503: Based on this work, we conclude that we cannot rule out NH$_4^+$ and CH$_3$OH as contributors each of at most $\sim$50\%

504: to the 6.8\mic band, but we emphasize that even the 50/50 combinations do not reproduce the overall feature shapes.  

505: Studies of this feature performed with higher-resolution ISO-SWS observations 

506: \citep[e.g.][]{KTB_01_5-8umregion,Schutte_96_6/6.8inngc7538} find a similar variety of structure and peak positions, confirming that it may be difficult or physically

507: impossible to adequately explain the feature using the same set of components for all sight-lines.

508: 

509: \subsection{The 7-8\mic Range} \label{sec:7-8mic}

510: 

511: Because of the uncertainties due to the low resolution and low signal-to-noise of the 7-8\mic optical depth spectrum, 

512: as well as the inherent narrowness of the 7-8\mic features, 

513: we did not attempt to fit the shape of the spectral features with laboratory profiles but rather focused on detection of the 7.25\mic HCOOH, $\sim$7.6\mic SO$_2$, 

514: and 7.7\mic CH$_4$ features.

515: 

516: \subsubsection{HCOOH}

517: 

518: The 7.25\mic HCOOH band is intrinsically $\sim$4 times weaker than the 5.85\mic one \citep{Bisschop_07_hcooh}.  

519: Based on the amount of HCOOH in the fitted spectra in Section~\ref{sec:6um_feature}, the expected 7.25\mic optical depths in the YSO spectra

520: are $\lesssim$0.02.  Indeed, we do observe small features at 7.25\mic in 6 of our 16 objects.  It should be noted, however, that the strength of these do not

521: correlate with the HCOOH column densities derived from the spectral fits and the stronger 5.85\mic feature ($r=0.06$);

522: there are also other species with weak

523: signatures very close to this wavelength \citep[e.g. HCOO$^-$ and HCONH$_2$;][]{Schutte_99_7.2/7.4mic}.  As the optical depth profiles here are neither strong

524: nor resolved enough to conclusively distinguish among these species' signatures, these HCOOH detections must remain tentative.

525: 

526: \subsubsection{CH$_4$ and SO$_2$} \label{sec:ch4so2}

527: 

528: In 13 of our 16 objects, we observe a feature near 7.7 $\mu$m, where the ($\nu_4$) C-H deformation mode of CH$_4$ is expected.  Laboratory studies have shown that this

529: resonance's position and width are strongly dependent on the molecular environment, becoming blue-shifted (as far as 7.64 $\mu$m) and 

530: broadened when mixed with a variety of other ices, including H$_2$O, CO, CH$_3$OH, and CO$_2$ \citep{Boogert_97_labch4}.

531: In addition, there is the possibility of blending with the nearby SO$_2$ feature, whose 7.58\mic S-O stretching band 

532: has been shown to vary its peak position from 7.45-7.62 $\mu$m, 

533: depending on the temperature and the molecular environment \citep{Boogert_97_labch4}.  

534: 

535: In order to disentangle absorption due to SO$_2$ and CH$_4$, we fit each YSO's 7.45-8\mic wavelength range

536: with two Gaussian curves, allowing the peak of one (representing SO$_2$) to shift between 7.45-7.62\mic and constraining the other peak (CH$_4$) to 7.62\mic or larger.

537: A Gaussian approximation was chosen because both the SO$_2$ and CH$_4$ profiles are too narrow, compared with the IRS resolution, to

538: evaluate any substructure.

539: In many cases, a single Gaussain was sufficient, indicating a single, distinct feature of either SO$_2$ or CH$_4$, depending on peak position.

540: Of our 16 YSOs, 7 show only a 7.7\mic feature with little or no asymmetrical excess, 3 only have a distinct feature peaking 

541: within the SO$_2$ range, 4 show both distinct features, and 2 show a broad absorption with additional peaks

542: within the SO$_2$ range (i.e. fit by overlapping Gaussians).  Using the range of laboratory feature FWHMs

543: measured by \citet{Boogert_97_labch4}, we find that in 4 of the 9 objects with either distinct

544: features or a Gaussian fit in the 7.45-7.62\mic range, the feature width is too narrow ($\Delta\lambda\lesssim$0.06 $\mu$m) to be attributed to solid SO$_2$.  

545: These may be due to random noise creating an apparent peak or to CH$_4$ gas lines found between $\sim$7.4-8 $\mu$m.

546: SO$_2$ also has gas lines in this region, but these are very weak \citep{Boogert_97_labch4}.  For the 5 remaining features, we use 

547: Equation~\ref{sec:column} with a band strength $A_{SO_2}$=3.4$\times$10$^{-17}$cm molec$^{-1}$ \citep{Khanna_88_SO2bandstrength} to determine the SO$_2$ column

548: density; for the other objects, an upper limit to $N$(SO$_2$) is determined (Table~\ref{tab:abundances}).

549: 

550: Of the 13 objects with a distinct 7.7\mic CH$_4$ feature or Gaussian fit, 7 have FWHMs comparable to those observed in laboratory mixtures \citep{Boogert_97_labch4}, 

551: particularly in those mixtures where CH$_4$ and either H$_2$O, CH$_3$OH, or NH$_3$ exist in roughly equal abundance.  The remaining 6 objects have wider features,

552: which may be explained by variations in temperature or ice mantle composition along the line of sight.  To determine the column density of CH$_4$ towards each YSO, 

553: we use Equation~\ref{sec:column} with a band strength $A_{CH_4}$=7.3$\times$10$^{-18}$cm molec$^{-1}$ \citep[pure CH$_4$; discussion in][]{Boogert_97_labch4} and

554: integrate the optical depth between 7.6-7.9 $\mu$m.  Upper limit estimates for the 3 non-detections are provided by 

555: the noise in the spectrum (Table~\ref{tab:abundances}).

556: 

557: \subsection{The 15.2 $\mu$m Feature} \label{sec:15um_feature}

558: 

559: Double-peaked structure in the 15.2\mic feature, often observed in systems with an exposed source of heating and radiation 

560: (such as Class I/II objects), is associated with the dual peaks of the degenerate $\nu_2$ bending mode in pure CO$_2$ ice.  

561: This structure becomes less pronounced or disappears altogether when the apolar CO$_2$ molecules are embedded in an H$_2$O-rich polar matrix 

562: \citep[e.g.][]{EKS_99_laboratory}.  Thus the presence of a double-peaked feature indicates that thermal processing has occured, segregating the CO$_2$ from the polar

563: matrix.  The broad shoulder

564: on the red side of the absorption, centered around 15.4 $\mu$m, has been attributed to CO$_2$ interactions with other molecules, particularly CH$_3$OH,

565: and the strength and shape of the shoulder depend on both temperature and relative abundances \citep[e.g][]{DDd_99_co2ch3oh}.

566: Theoretical studies suggest that solid CO$_2$ is formed by oxidation of CO ice

567: \citep[e.g.][]{Tielens_82_compositionmodels,Ruffle_01_co2formation}, so the co-existence and interaction with solid CH$_3$OH 

568: \citep[also formed from CO;][]{Tielens_97_methproduction} is not unexpected.

569: 

570: In the fits, made for the high-resolution spectra\footnote{The low-resolution spectra are insufficient to resolve any structure which might be present.} 

571: between 14.7 and 16 $\mu$m, we include cold (10-30 K) deposits of H$_2$O:CO$_2$, pure CO$_2$, and an irradiated deposit of CO:H$_2$O, as well as

572: warmer ($\gtrsim$100 K) mixtures of H$_2$O:CH$_3$OH:CO$_2$ (Table~\ref{tab:15um_components}).  

573: The cold mixtures containing H$_2$O are used to estimate the contribution from 

574: nonprocessed material along the line of sight.  The warmer H$_2$O:CH$_3$OH:CO$_2$ spectra represent ices which have been warmed and have undergone 

575: the chemical and physical changes, including segregation of the CO$_2$, resulting from that processing.  

576: The pure CO$_2$ component, though measured at 10 K, does not greatly differ in shape until sublimation at $\sim$80 K 

577: and is here used as a measure of segregated, highly-processed ice.

578: 

579: Our fitting components differ from the recent work by \citet{Pontop_08_co2inlowmass}, 

580: which is focused almost exclusively on detailed methanol-free CO:CO$_2$:H$_2$O mixtures. Certainly these combinations can produce good fits to the CO$_2$ ice feature.

581: However, in the following we will adhere to 

582: the conventional approach involving a warm CH$_3$OH ice component, since we find that in most cases this produces better fits to our spectra than methanol-free 

583: mixtures, and it takes into account the strong evidence of solid methanol in many protostellar environments \citep[e.g.][]{Pontopp_03_methdetect,GWB_04_ice}.

584: It is worth noting that, since good fits are obtained by both classes of models, neither method should be thought to provide a unique solution to the ice-composition 

585: analysis. 

586: 

587: Using a single component fit, we find that a nearly equal-parts mixture of H$_2$O:CH$_3$OH:CO$_2$ at temperatures of $\gtrsim$100 K provides the lowest-$\chi_\nu^2$ 

588: fit to the observed YSO 15.2\mic absorption bands, in accordance with previous results for more massive protostars \citep[e.g.][]{EKS_99_laboratory,GWE_99_molcloud}.  

589: Where present, the dual peak positions align, and the approximate strength of the 15.4\mic shoulder is replicated.  

590: The fit improves, however, by an average of $\sim$50\% of $\chi_\nu^2$, when we allow multiple components (Figure~\ref{fig:15umfit}).  The combination of these

591: additional components reproduces the relative peak strengths, the inter-peak dip depth, and the shape of the 15.4\mic shoulder without requiring a

592: specific single temperature.

593: 

594: Our multicomponent fits indicate

595: that for our YSO sample, a smaller amount of CH$_3$OH ($\sim$14\%) at lower temperatures is sufficient to account for the observed 15.4\mic shoulder, 

596: and equal parts H$_2$O, CO$_2$, and CH$_3$OH are reflected at high temperatures \citep[higher than those identified by e.g][]{EKS_99_laboratory}, 

597: when the CO$_2$ has at least partly migrated out of the matrix.

598: Laboratory mixtures at these higher temperatures that also contain the smaller 

599: fractions of CH$_3$OH have an asymmetrical dual peak that is not observed in the YSO spectra; 

600: mixtures with H$_2$O, CH$_3$OH, and CO$_2$ in a 1:1:1 ratio at lower temperatures show a stronger shoulder at 15.4\mic (relative to the main peak) as well as a 

601: nearly-single peak at 15.2 $\mu$m.  

602: While the cold-surface formation processes and expected abundances of CH$_3$OH are not tightly-constrained 

603: \citep[e.g.][and references therein]{Tielens_96_starformingices,KTB_01_5-8umregion}, the generally-higher

604: abundances observed towards high-mass protostars are evidence that warmer environments may be conducive to additional CH$_3$OH production.

605: This is consistent with the stronger 15.4\mic shoulders observed towards massive protostars 

606: \citep[compared with low-mass ones; e.g.][]{EDD_98_segregation,BPL_04_lowmass},

607: and the equal-parts H$_2$O:CH$_3$OH:CO$_2$ mixtures required to fit them.

608: Most of our YSO fits are further improved by including the very cold ($\sim$30 K) CO:H$_2$O layer, representing absorption 

609: along the line of sight due to cold material, affected by the region's radiation field but too distant or too shielded from the central source

610: to have undergone thermal heating.

611: 

612: Methanol-bearing ice mixtures have distinctive absorptions elsewhere in the mid-infrared spectrum, notably at 6.8 and 9.8 $\micron$. 

613: The former feature is discussed above (Section~\ref{sec:6.8um_feature}). 

614: The latter unfortunately lies in the very bottom of the silicate absorption feature. 

615: In the mixtures that best fit the 15.2 $\micron$ ice feature, the 

616: 9.8 $\micron$ feature is either very weak or not particularly sharp. Thus it is difficult to separate possible 9.8 $\micron$ methanol ice absorption from small 

617: silicate-feature profile variation.

618: In general, a combination of absorption from the ice mixture that best fits the 15.2 $\micron$ feature, and from interstellar silicate grains 

619: \citep[toward GCS 3;][]{Kemper_04_silicates}, can be made to fit the spectrum, as illustrated in Figure~\ref{fig:silfit}. 

620: However, this result depends sensitively on the fit of 

621: the continuum and the silicate absorption profile, which as noted above is not very well-defined for these types of environments. We found no meaningful, 

622: quantitative constraints on a putative CH$_3$OH ice component from the 9.8 $\micron$ feature.

623: 

624: The 15.2\mic features in the molecular cloud sight-lines have previously been fitted with predominantly cold, polar H$_2$O:CO$_2$ mixtures

625: \citep{BMG_05_co2fieldstars,KBP_05_backgroundstars}.  We find that the $\chi^2_\nu$ of the fits can be matched or reduced

626: by including components containing methanol---specifically,

627: the cold counterparts of the H$_2$O:CH$_3$OH:CO$_2$ components used in our Class I/II fits described above (Figure~\ref{fig:bkgdfit}).  

628: The percentage of polar ice found ($\sim$75-80\%)

629: is comparable to that identified by \citet{BMG_05_co2fieldstars} and \citet{KBP_05_backgroundstars}, 

630: and the abundance ratios $N$(CO$_2$)/$N$(H$_2$O) from those studies are preserved.

631: 

632: We have calculated the total CO$_2$ column density (Table~\ref{tab:abundances}) observed along each embedded-object and molecular cloud

633: line of sight using Equation~\ref{eqn:column}. 

634: The exact band strengths of some of the laboratory mixtures used in the 15.2\mic fits

635: are undetermined.  The fitting results indicate that the vast majority of the CO$_2$ ice resides alongside H$_2$O ice,

636: so we use a band strength of 1.5$\times$10$^{-17}$cm molec$^{-1}$, 

637: which is appropriate for the 15.2\mic absorption from a H$_2$O:CO$_2$ 1.6:1 ice \citep{GSG_95_labsim}.  

638: The observed optical depth spectra were integrated between 14.7-16 $\mu$m.

639: We use the relative contributions of each fitting component to estimate the

640: column density of solid CO$_2$ in each stage of thermal processing (Table~\ref{tab:procseg}): radiation-only (10-30 K, ``cold''), some thermal heating (``warm''), 

641: and CO$_2$ segregation.  We have also estimated the column density of solid CH$_3$OH 

642: embedded in CO$_2$ using the the CH$_3$OH abundance and relative contribution of each laboratory fit component (Table~\ref{tab:abundances}).

643: 

644: \section{DISCUSSION} \label{sec:discussion}

645: 

646: \subsection{Ice Composition: Abundances With Respect to H$_2$O} \label{sec:chemdiscuss}

647: 

648: \subsubsection{CO$_2$ Column Density}

649: 

650: The column densities observed in our sample of Class I/II YSOs yield $N$(CO$_2$)/$N$(H$_2$O) $\sim$ 0.12($\pm$0.04) 

651: (Table~\ref{tab:abundances}, Figure~\ref{fig:waterco2_Nratio}), which is consistent with the values previously derived for many

652: different lines-of-sight, averaging $N$(CO$_2$)/$N$(H$_2$O) $\sim$ 0.17($\pm$0.03) 

653: \citep[][]{GWE_99_molcloud,Nummelin_01_co2inlowmass,GWB_04_ice,KBP_05_backgroundstars,ACA_03_isocam}, but significantly lower than the value of 

654: $N$(CO$_2$)/$N$(H$_2$O) $\sim$ 0.3, with large scatter, reported by \citet{BPL_04_lowmass} and 

655: \citet{Pontop_08_co2inlowmass}.  

656: Our study is unique in the sense that we use the band strength $A_{CO_2}$ of an H$_2$O:CO$_2$ mixture.  If we instead use the band strength for pure CO$_2$,

657: we find $N$(CO$_2$)/$N$(H$_2$O) $\sim$ 0.16($\pm$0.04), nearly equivalent to the results of the majority of previous studies, 

658: which include high- and low-mass YSOs, background field stars sampling molecular clouds, and galactic center sight-lines sampling 

659: diffuse interstellar material and molecular clouds.

660: 

661: The dashed line in Figure~\ref{fig:h2oco2_temp} identifies the locus of compositions with CO$_2$/H$_2$O = 0.12, which are clearly underrepresented in the heated samples

662: from the Leiden University database.  Good profile fits can only be achieved by combining ice mixtures, and this gap in composition-space may explain some of the

663: difficulty experienced in fitting 6\mic features.  Further laboratory exploration of the water-rich mixtures around the observed ratio with CO$_2$ may help

664: to improve constraints on the column density of H$_2$O ice based on the 6\mic feature.

665: 

666: \subsubsection{Evolutionary Trends in CO$_2$}

667: 

668: When considering the Class I/II sample in conjunction with the Class 0 objects, we find $N$(H$_2$O) to be well-correlated ($N$(CO$_2$) less so) with the

669: spectral index $\alpha$, where more embedded sources (higher $\alpha$) show higher column densities, as is expected (Figure~\ref{fig:alpha_abun}; top 2

670: panels).  The ratio $N$(CO$_2$)/$N$(H$_2$O) increases as the objects evolve towards lower $\alpha$-values (Figure~\ref{fig:alpha_abun}; bottom panel). 

671: 

672: The sublimation temperature of solid H$_2$O is $\sim$40 K higher than that of pure CO$_2$ ice \citep{GSG_95_labsim}.

673: Thus with increased exposure to thermal radiation, $N$(CO$_2$)/$N$(H$_2$O) is expected to go down, seemingly contradicting our observations. 

674: Although CO$_2$ molecules embedded in a polar matrix

675: may become ``trapped'' and able to survive to temperatures nearing the sublimation limit of H$_2$O, 

676: the thermal destruction of H$_2$O before CO$_2$ is unlikely.

677: 

678: An explanation of the high column density of CO$_2$ in more exposed environments may be found in the formation of additional CO$_2$ in

679: Class I/II objects.  

680: Photodissociation of H$_2$O leads to the production of radicals (e.g. OH) which can react to form CO$_2$ on the grain surface

681: \citep[e.g. CO+OH$\rightarrow$CO$_2$;][]{Alla_88_iceevolution}.

682: Some other reaction pathways to form CO$_2$ in molecular clouds and star-forming regions have activation barriers which become less obstructive in warmer,

683: more exposed environments \citep[e.g][]{Ruffle_01_co2formation,vanDishoeck_98_chemevolSFR}.  Indeed, in UV-irradiated H$_2$O:CH$_3$OH:CO$_2$ laboratory samples,

684: CO$_2$ eventually reaches a constant abundance level due to an equilibrium between destruction and formation \citep{EKS_99_laboratory},

685: while the abundance of H$_2$O further decreases due to photodissociation.

686: 

687: The values measured towards the background stars Elias 16, Elias 13, and CK~2 cannot be

688: extrapolated from this trend.  The $N$(CO$_2$)/$N$(H$_2$O) ratios for

689: these lines-of-sight (Table~\ref{tab:molcloud_abun}) are not lower than the ratios observed

690: towards the YSOs.  This discrepancy may be due to the fact that the

691: interstellar lines-of-sight probe a variety of environments and

692: compositional structure, while protostellar sight-lines are dominated

693: by the circumstellar material.  Possibly, the lower temperatures in

694: interstellar regions stifle CO$_2$ formation mechanisms with activation

695: barriers, while areas of increased density favor H$_2$O production.

696: 

697: \subsubsection{HCOOH}

698: 

699: Apart from one outlier at $N$(H$_2$O) $\sim$ 1.1$\times$10$^{19}$cm$^{-2}$ (L1551 IRS 5), the column densities of HCOOH and H$_2$O are strongly related ($r=0.61$,

700: Figure~\ref{fig:waterhcoohmeth_Nratio}).  This supports laboratory experiments demonstrating that the most likely formation environments

701: of solid HCOOH are those that also produce H$_2$O ice, through either radiative processing or cold surface reactions \citep{Bisschop_07_hcooh}.

702: We find $N$(HCOOH)/$N$(H$_2$O) $\sim$ 0-2\%, with an average of 0.6\%, 

703: well within the $\lesssim$5\% abundance ratio observed in lines-of-sight towards massive and low-mass protostars \citep{Bisschop_07_hcooh,Boogert_08_lowmassices}.

704: 

705: \subsubsection{CH$_3$OH}

706: The observed $N$(CH$_3$OH)/$N$(H$_2$O) ratios in our sample do not show a clear trend (Figure~\ref{fig:waterhcoohmeth_Nratio}).  

707: It is known that the $N$(CH$_3$OH)/$N$(H$_2$O) abundance varies greatly from source to source \citep[e.g][]{BE_04_ISI}, and it may be better

708: to analyze the CH$_3$OH abundance with respect to CO$_2$.

709: 

710: CH$_3$OH can be detected in the spectrum in three places. When it is intermixed with CO$_2$, a shoulder at 15.4 $\mu$m due to CH$_3$OH can

711: be observed in the 15.2 $\mu$m feature; and CH$_3$OH also shows 6.8 and 9.8 $\mu$m bands (the latter is blended in the 9.7\mic silicate complex).

712: In Figure~\ref{fig:6.8um15.2um_vs_15.4um15.2um} we show the ratio of the optical depth at

713: 15.4 $\mu$m (vertical axis) and 6.8 $\mu$m (horizontal axis), both with respect to the optical depth at 15.2 $\mu$m.  Filled circles

714: represent the astronomical measurements, while the plus signs indicate the ratios measured in the laboratory spectra.  The laboratory spectra

715: show a correlation between the two features, whose exact position depends on the temperature.  The 15.4/15.2 $\mu$m $\tau$ ratios

716: observed in the astronomical data are within the range measured in the laboratory, but the corresponding 6.8 $\mu$m optical depths are

717: significantly greater than the laboratory values.  This could be understood if only a small fraction of the CH$_3$OH along the line of

718: sight is embedded in a CO$_2$ matrix or if there are significant contributions to the 6.8 $\mu$m feature from other components.

719: 

720: Using the laboratory fits to the 15.2\mic complex in the high spectral resolution data, we estimate the column density of solid CH$_3$OH intermixed with CO$_2$

721: as 30-56\% with respect to CO$_2$, corresponding to $N$(CH$_3$OH)$_\mathrm{{CO_2}}$/$N$(H$_2$O) $\sim$ 2-9\% (Table~\ref{tab:abundances}).

722: Assuming that $\lesssim$50\% of the 6.8\mic feature is due to to CH$_3$OH (Section~\ref{sec:6.8micfits}), we estimate $N$(CH$_3$OH)$_\mathrm{{total}}$/$N$(H$_2$O) 

723: $\lesssim$ 5-17\%.  There are two outlying values, for IRAS 04108+2803A and IRAS 04181+2654B, at 32 and 28\%.  

724: Although we do not place high confidence in these as absolute values because of the 6.8\mic uncertainties, they are

725: consistent with the detection of solid CH$_3$OH with abundances (relative to H$_2$O) in this range up to $\sim$25\% towards low-mass YSOs \citep[as well as the overall

726: scatter in abundance;][]{Pontopp_03_methdetect,Boogert_08_lowmassices}.  Thus our data support the presence of a significant CO$_2$-poor H$_2$O:CH$_3$OH ice component.

727: 

728: \subsubsection{CH$_4$}

729: 

730: We observe a range of $N$(CH$_4$)/$N$(H$_2$O) ratios from $\sim$0.3 to 8\% in our sample (Figure~\ref{fig:watermethane_Nratio}), 

731: with an average of 2.6\% and a single outlier (IRAS 04154+2823) at 23\% .  

732: Our measurements largely overlap with the ratios of $\sim$0.5-4\% reported in the literature \citep[e.g.][]{Boogert_96_ch4inprotostars,GWB_04_ice},

733: though ratios up to 13\% are reported by \citet{Oberg_08_ch4}.

734: The higher values in our sample ($>$4\%, dashed line in Figure~\ref{fig:watermethane_Nratio}) may be 

735: partially ascribed to additional absorptions unresolvable at the SL-module resolution,

736: such as gas-phase CH$_4$ lines \citep{Boogert_97_labch4}, or to matrix-dependent variations in band strength.  

737: \citet{Hudgins_93_ch4bandstrength} found that the band strength for the 7.7\mic

738: CH$_4$ feature may vary by as much as a factor of 2.8 between H$_2$O-rich and CO-rich matrices.  

739: We use the value for pure CH$_4$, which lies approximately midway between the two extremes.

740: 

741: Laboratory studies of solid CH$_4$ indicate that it can be formed by a variety of processes---including UV photolysis of CH$_3$OH ices \citep{Alla_88_iceevolution}

742: and grain-surface hydrogenation of carbon \citep{Tielens_82_compositionmodels}---and that the resulting absorption profile

743: has characteristics unique to its formation pathway and embedding matrix \citep{Boogert_97_labch4}.  

744: Unfortunately, the low signal-to-noise of the CH$_4$ detections, and the low spectral resolution, prevent a detailed analysis of the spectral profiles;

745: however, the broad range of peak positions and widths implies that the CH$_4$ embedding matrices vary across the sample.

746: 

747: \subsection{Segregation of CO$_2$ Ice} \label{sec:co2seg}

748: 

749: In the colder parts ($\lesssim$20 K) of a star-forming region, well-removed from the hot cores and embedded protostars or buried in dense clumps, 

750: apolar CO$_2$ ice is formed embedded in a polar H$_2$O matrix \citep[e.g.][]{BE_04_ISI}.

751: The 15.2\mic CO$_2$ resonance will appear to be broad and single-peaked.  As these ices are heated,

752: as for instance happens in the circumstellar environments of YSOs,

753: the CO$_2$ can migrate out of the polar matrix and aggregate, 

754: either as a surface layer or in bulk apolar clumps within the matrix \citep[e.g.][]{SA_90_co2inanalogs,EKS_99_laboratory}.

755: The spectral appearance of the CO$_2$ resonance changes, and a double-peaked structure, with peaks at 15.1 and 15.25 $\mu$m, becomes apparent rather

756: rapidly (Figure \ref{fig:d2p_vs_temp}).

757: The migration process also destabilizes the CO$_2$:CH$_3$OH interactions

758: and reduces the 15.4\mic shoulder.  Due to the apparent spectral difference between segregated CO$_2$ ice

759: and CO$_2$ ice embedded in an H$_2$O:CH$_3$OH matrix, it is possible to determine the relative mass contained in both of these phases.

760: 

761: In Figure~\ref{fig:15um_hires_spectra}, we show the variation in the CO$_2$ 15.2 $\mu$m absorption feature in our sample, along with laboratory spectra of

762: CO$_2$ fully embedded in the matrix (top; 10 K) and fully segregated CO$_2$ (bottom; 130 K). The dashed lines indicate the positions of the

763: double-peaked segregated CO$_2$ and the shoulder at 15.4 $\mu$m due to embedded CO$_2$ interacting with CH$_3$OH. We show

764: only the Class I/II objects for which high resolution spectra are available, as the double-peaked structure is not resolved

765: in the low resolution spectra. 

766: Several of the spectra clearly show the double-peaked structure, while all deviate from the smooth, unprocesed Taurus cloud profiles

767: as published by \citet{WSB_07_quiescentICM}, \citet{BMG_05_co2fieldstars}, and \citet{KBP_05_backgroundstars}.

768: Indeed, from the multi-component fits described in Section~\ref{sec:15um_feature},

769: we find evidence in all of our low-mass YSO spectra for moderate-temperature thermal processing 

770: ($\lesssim$100 K; CO$_2$ warmed but not segregated),

771: in agreement with \citet{BPL_04_lowmass} but in contrast with \citet{Nummelin_01_co2inlowmass}.

772: However, unlike \citet{BPL_04_lowmass}, we also find

773: evidence for high-temperature processing---5 of the 9 objects (IRAS 04016+2610, IRAS 04239+2436, IRAS 04365+2535, DG Tau B, and IC 2087 IR)

774: have 15.2\mic profiles entirely dominated by warmed ices in which a significant fraction ($\sim$20-40\%)

775: of the CO$_2$ has segregated from the polar formation matrix.  

776: 

777: From the multi-component fits, we can derive the mass distribution of CO$_2$ ice over the {\it phases} (cold, warm, segregated).

778: The phase transitions correspond to laboratory temperatures, which in turn translate to (lower) astrophysical temperatures.

779: The onset of CO$_2$ segregation occurs at a temperature near 105 K in the laboratory (Figure \ref{fig:d2p_vs_temp}), which corresponds

780: to an astrophysical temperature of $\sim$65 K \citep[e.g.][]{BEG_00_13co2}.

781: The CO$_2$ evaporation temperature depends on the relative abundance of other 

782: species, particularly H$_2$O, but the mixtures with the most physically likely abundances 

783: (i.e. CO$_2$/H$_2$O $<$ 1) show an average sublimation temperature of $\sim$150 K, or 80 K in astrophysical environments.

784: Thus, the {\it segregated} column densities of Table~\ref{tab:procseg} represent CO$_2$ ice at circumstellar radii where the

785: ice temperature is between 65 and 80 K, and the {\it cold}+{\it warm} column densities indicate ice farther from the central source (i.e. where T $<$ 65 K).

786: The inner boundary of the segregated CO$_2$ ice zone is defined by the evaporation temperature at $\sim$80 K.

787: The low-mass protostellar envelope models produced by \citet{Furlan_08_classImodel} for nearly all the stars in our sample 

788: indicate these approximate temperature ranges in the densest inner regions

789: ($R$ $\lesssim$ 100 AU); however, detailed physical structure models that also include the appropriate ice opacities are needed to further constrain the location and

790: processing levels of the ices.

791: 

792: In Figure~\ref{fig:alpha_seg} we show the correlation of thermal processing levels with the SED spectral index (indicative of the line-of-sight dust column).  

793: While there is a large amount of scatter in these plots, the overall fraction of cold CO$_2$ ice increases 

794: (and the warm fraction decreases) for more deeply embedded objects, while the segregated fraction does not show a clear trend with total dust opacity.

795: %This suggests that once part of a protostellar environment has been warmed enough to segregate CO$_2$, 

796: %the process becomes dominated by the rapid rate with which segregation and subsequent sublimation occur.

797: 

798: \subsubsection{Segregation of CO$_2$ Ice in Molecular Clouds and W3(OH)}

799: All three interstellar cloud sight-lines show

800: single-peaked 15.2\mic profiles with very little substructure; CK 2 and Elias 16 are remarkably similar in shape (Figure~\ref{fig:bkgdfit}), 

801: despite indications that ice mantles in

802: Serpens may be fundamentally different from those in Taurus \citep[e.g][]{Eiroa_89_serpensice,WSB_07_quiescentICM}.  

803: These results give support for the cold grain origin of at

804: least some of the CH$_3$OH observed towards protostars; the presence of the unidentified 6.8\mic band in background star sight-lines

805: \citep[in particular CK 2 and Elias 16;][]{KBP_05_backgroundstars} is consistent with the assignment of CH$_3$OH as a carrier.

806: 

807: Interestingly, the spectrum of W3(OH), which has the reddest SED of our entire sample, shows a clearly double-peaked CO$_2$ feature, 

808: indicating pure CO$_2$ along the line of sight, which could have been formed by thermal segregation (Figure~\ref{fig:w3ohfit}).  

809: The segregation fraction is $\sim$26\%, comparable to the Class I/II average of $\sim$30\%.  Given the difference in segregation with 

810: the pure molecular cloud lines of sight, the most likely

811: explanation is that this line of sight samples not only the cold foreground W3 cloud but also the warmer, 

812: thermally-processed CO$_2$ ices around the massive protostars in

813: the W3(OH) region.  The strong similarity of the 15.2\mic absorption profile towards W3(OH) to those of the low-mass Taurus

814: objects is curious, as the line of sight toward W3(OH) is highly diverse, containing an ultra-compact H{\sc ii} region,

815: strong outflows, and H$_2$O/OH maser emissions \citep{Turner_84_w3ohtw,Alcolea_93_whohtwmaser,Argon_03_w3ohtwmaser}.  

816: Future spectroscopic observations with higher angular resolution will resolve the spatial structure and

817: provide useful information on the complexity of the early stages of CO$_2$ heating in embedded objects.

818: 

819: \section{CONCLUSIONS} \label{sec:conclusions}

820: 

821: We have analyzed the mid-infrared molecular ice features in 16 Class I/II low-mass YSOs using Spitzer IRS observations.  

822: Using detailed comparisons to laboratory ice analog spectra, we find the ice mantle compositions to be dominated by H$_2$O and CO$_2$ ice but with significant

823: abundances of additional species such as CH$_3$OH, H$_2$CO, HCOOH, NH$_3$, CH$_4$, and SO$_2$ ($\lesssim$10\% each).  

824: 

825: Detailed spectral fits to the 6\mic feature reveal that $\gtrsim$80\% of the band can be attributed to H$_2$O, with additional absorptions near 5.85\mic identified as

826: HCOOH and H$_2$CO.  Some of the observed spectra have a broader 6\mic peak that can be well-fit by H$_2$O:NH$_3$ ices, 

827: providing evidence for the presence of solid NH$_3$.  

828: On the long-wavelength wing of the 6\mic feature, small absorption remnants are observed, varying

829: in strength, shape, and peak position among the sources.  

830: We show that these are highly unlikely to be the result of PAH absorption, 

831: due to the absence of other expected PAH features at 7.5-9 and 11-14 $\mu$m.

832: 

833: We have examined the unidentified 6.8\mic feature and discussed its assignment to solid CH$_3$OH and NH$_4^+$, the two most common candidates.  While no high-quality

834: spectral fits could be achieved using these two components, we find that on average the observed YSO features could be approximately fit by equal contributions

835: from NH$_4^+$ and from a CH$_3$OH/H$_2$O$\sim$0.1 mixture, with varying amounts of nearly-pure CH$_3$OH intermixed.  However, we find a wide variety

836: of feature shapes in our sample, indicating that the interstellar component(s) producing this band vary greatly from object to object, and a single simple

837: carrier or set of carriers may be inadequate to explain the feature.

838: 

839: In the 7-8\mic range, we are able to isolate and identify features attributed to CH$_4$, SO$_2$, and

840: (tentatively) HCOOH ices.  The high CH$_4$ ice abundances derived from these, along with the variety of feature shapes and peak positions within the sample, 

841: suggest variations in line-of-sight molecular environment or possible absorption contamination by unresolvable gas-phase CH$_4$ lines.  

842: The solid SO$_2$ and CH$_4$ band profiles are

843: very sensitive to molecular environment and temperature, and additional high-resolution studies of this wavelength range will yield further information

844: on molecular ice interactions and abundance structures along the line of sight.

845: 

846: The 15.2\mic solid CO$_2$ bending mode feature is well-reproduced by multi-component laboratory fits comprising ice analogs (H$_2$O, CH$_3$OH, CO, CO$_2$)

847: in various stages of radiative and 

848: thermal processing.  Using the relative contribution of the components to each high-resolution feature profile, we find that all of the YSO ice environments

849: have undergone at least some thermal heating, and 5 of the 9 have profiles entirely dominated by processed ices in which a significant fraction ($\sim$20-40\%)

850: of the CO$_2$ has segregated from the polar formation matrix.  

851: 

852: The molecular ice abundance ratios, particularly $N$(CO$_2$)/$N$(H$_2$O), are consistent with the values observed in a wide variety of sight-lines, but within the sample,

853: we observe a negative correlation between this ratio and the mid-infrared SED spectral index.  

854: This evidence describes a more complicated polar CO$_2$ environment than indicated by previous studies.  

855: Comparison of the SED spectral index with the ice temperatures derived from the 15.2\mic fits

856: confirms the presence of cold ices in more deeply embedded objects and warmer, annealed ices in more evolved ones.

857: 

858: Comparison with molecular cloud sight-lines show that our fits are consistent with a scenario

859: in which H$_2$O-rich grain mantles with CO$_2$ and CH$_3$OH relative abundances of $\sim$20 and $\lesssim$10\%, respectively, as well as traces of other species, 

860: form in the cold molecular clouds and are

861: subsequently heated by young protostars.  This thermal processing sublimates the most volatile species while inducing molecular interactions and migrations

862: that become spectroscopically detectable.  The wide range of both molecular abundances and thermal processing levels observed within our sample indicates that

863: young stars have a strong effect on their surrounding environments.  

864: 

865: \begin{acknowledgements}

866: We are grateful to Edwin Bergin and Claudia Knez for providing us with the field star spectra, to Melissa McClure for assistance with data reduction issues, and

867: to Adwin Boogert for stimulating discussion.

868: Extensive use was made of Leiden University's Sackler Laboratory ice analog databases and of the NASA ADS Abstract Service.

869: This work is based on observations made with the Spitzer Space

870: Telescope, which is operated by the Jet Propulsion Laboratory,

871: California Institute of Technology under NASA contract 1407. Support

872: for this work was provided by NASA through Contract Number 1257184

873: issued by JPL/Caltech.

874: E.F. acknowledges support from a NASA Postdoctoral Program Fellowship.

875: \end{acknowledgements}

876: 

877: \bibliographystyle{aa}

878: \bibliography{gaillib}

879: 

880: \begin{deluxetable}{l c c c} \tablewidth{0 pt} \tabletypesize{\small} 

881: \tablecaption{Overview of the Spitzer IRS observations of our sample of Class I/II objects.  The first column gives the object name, and the second to fourth columns

882: give the observations' AOR key, date, and IRS modules/mode used (SL: short-low, SH: short-high, LL: long-low, LH: long-high).} 

883: \tablehead{\colhead{Object} & \colhead{AOR} & \colhead{Obs. Date (2004)} & \colhead{IRS Modules/Mode}}

884: \startdata

885: IRAS 04016+2610 &3528960 & 2/07 &SL,SH,LH Map \\

886: IRAS 04108+2803A &3539712 & 2/06 &SL,LL Stare \\ 

887: IRAS 04108+2803B &3529472 & 2/06 &SL,SH,LH Map \\ 

888: IRAS 04154+2823 &3534336 & 3/03 &SL,LL Map \\

889: IRAS 04169+2702 &3534848 & 3/04 &SL,LL Map \\

890: IRAS 04181+2654A &3546112 & 2/29 &SL,LL Stare \\ 

891: IRAS 04181+2654B &3545856 & 2/07 &SL,LL Stare \\ 

892: IRAS 04239+2436 &3530752 & 2/08 &SL,SH,LH Map \\

893: IRAS 04295+2251 &3537408 & 2/27 &SL,LL Map \\

894: IRAS 04361+2547 &3533056 & 2/29 &SL,SH,LH Map \\ 

895: IRAS 04365+2535 &3533312 & 2/27 &SL,SH,LH Map \\

896: IRAS 04381+2540 &3538944 & 2/27 &SL,LL Map \\ 

897: DG Tau B &3540992 & 2/08 &SL,SH,LH Stare \\

898: HL Tau &3531776 & 3/04 &SL,SH,LH Map \\

899: IC 2087 IR &3533312 & 2/27 &SL,SH,LH Map \\

900: L1551 IRS 5 &3531776 & 3/04 &SL,SH,LH Map \\ 

901: \enddata

902: \label{tab:obs}

903: \end{deluxetable}

904: \clearpage

905: 

906: \begin{deluxetable}{c c c c c} \tabletypesize{\scriptsize} \tablewidth{0 pt}

907: \tablecaption{Absorption features identified in our YSO spectra.} 

908: \tablehead{\colhead{Wavelength ($\mu$m)} & \colhead{Carrier(s)} & \colhead{Mode} & \colhead{{\it A} (10$^{-17}$cm molec$^{-1}$)} & \colhead{Reference}}

909: \startdata

910: Ice Features & & & & \\

911: \hline

912: 5.81 & H$_2$CO & C=O stretching & 0.96 & 1 \\

913: 5.85 & HCOOH & C=O stretching & 6.7 & 2 \\

914: 6.0 & H$_2$O & H-O-H bending & 1.2-1.3\tablenotemark{a} & 3 \\

915: 6.8 & CH$_3$OH?, NH$_4^+$?, organics & O-H bending, C-H deforming & $\cdots$ & \\

916: 7.58 & SO$_2$ & S-O stretching & 3.4 & 4 \\

917: 7.7 & CH$_4$ & C-H deforming & 0.73 & 4 \\

918: 15.2 & CO$_2$ & O-C-O bending & 1.5\tablenotemark{b} & 3 \\

919: \hline

920: Other Features & & & & \\

921: \hline

922: 9.7 & Silicates & Si-O stretching & $\cdots$ & \\

923: 14.97 & CO$_2$ & gas-phase $\nu_2$ bend & $\cdots$ & 5 \\

924: 18-20 & Silicates & O-Si-O bending & $\cdots$ & \\

925: \enddata

926: \tablenotetext{a}{{\it A} depends on temperature; range represents 10-120K}

927: \tablenotetext{b}{{\it A} for H$_2$O:CO$_2$ mix, see Section~\ref{sec:column}}

928: %\tablecomments{tablecomment}

929: \tablerefs{(1) \citet{Schutte_93_h2co}, (2) \citet{Marechal_87_hcooh}, (3) \citet{GSG_95_labsim}, (4) \citet{Boogert_97_labch4}, (5) \citet{vDHG_96_co2gasphase}}

930:  \label{tab:features}

931: \end{deluxetable}

932: \clearpage

933: 

934: \begin{deluxetable}{l c c c c c c} \tablewidth{0 pt} \tabletypesize{\small} 

935: \tablecaption{Characteristics of our sample of Taurus Class I/II objects.  The first column gives the object name, the second and third columns

936: give the 6.0\mic and 15.2\mic peak optical depths, and the fourth column gives the SED spectral index $\alpha$ (7-30 $\mu m$; Equation~\ref{eqn:alpha}).} 

937: \tablehead{\colhead{Object} & \colhead{$\tau_{6.0\mu m}$} & \colhead{$\tau_{15.2\mu m}$} & \colhead{SED $\alpha$}}

938: \startdata

939: IRAS 04016+2610 & 0.38($\pm$0.01) & 0.73($\pm$0.01) & 0.43($\pm$0.02) \\

940: IRAS 04108+2803A & 0.19($\pm$0.01)& 0.19($\pm$0.01) & 0.29($\pm$0.10) \\

941: IRAS 04108+2803B & 0.26($\pm$0.01)& 0.43($\pm$0.01) & 0.21($\pm$0.07) \\

942: IRAS 04154+2823 & 0.08($\pm$0.01) & 0.16($\pm$0.01) & -0.35($\pm$0.04) \\

943: IRAS 04169+2702 & 0.25($\pm$0.01) & 0.31($\pm$0.01) & 0.98($\pm$0.02) \\

944: IRAS 04181+2654A & 0.2($\pm$0.01) & 0.3($\pm$0.01) & 0.08($\pm$0.02) \\

945: IRAS 04181+2654B & 0.32($\pm$0.03) & 0.44($\pm$0.01) & 0.51($\pm$0.03) \\

946: IRAS 04239+2436 & 0.25($\pm$0.01) & 0.44($\pm$0.01) & 0.41($\pm$0.05) \\

947: IRAS 04295+2251 & 0.2($\pm$0.02)& 0.18($\pm$0.01) & 0.47($\pm$0.02) \\

948: IRAS 04361+2547 & 0.45($\pm$0.01) & 0.27($\pm$0.03) & 2.22($\pm$0.02) \\

949: IRAS 04365+2535 & 0.33($\pm$0.01) & 0.63($\pm$0.01) & 0.54($\pm$0.04) \\

950: IRAS 04381+2540 & 0.44($\pm$0.01) & 0.62($\pm$0.01) & 0.61($\pm$0.03) \\

951: DG Tau B & 0.34($\pm$0.04) & 0.3($\pm$0.01) & 0.38($\pm$0.07) \\

952: HL Tau & 0.12($\pm$0.01) & 0.12($\pm$0.01) & 0.34($\pm$0.02) \\

953: IC 2087 IR & 0.14($\pm$0.01) & 0.33($\pm$0.01) & -0.89$\pm$0.01) \\

954: L1551 IRS 5 & 0.54($\pm$0.01) & 0.58($\pm$0.03) & 1.41($\pm$0.02) \\

955: \enddata

956: \label{tab:obj}

957: \end{deluxetable}

958: \clearpage

959: 

960: 

961: \begin{deluxetable}{l c c} \tabletypesize{\small} \tablewidth{0 pt}

962: \tablecaption{Laboratory spectra used in 6\mic fits.} 

963: \tablehead{\colhead{Composition} & \colhead{Temperature (K)} & \colhead{Radiation\tablenotemark{a}}}

964: \startdata 

965: Pure H$_2$O & 10 & $\cdots$ \\

966: Pure H$_2$O & 30 & $\cdots$ \\

967: Pure H$_2$O & 50 & $\cdots$ \\

968: Pure H$_2$O & 120 & $\cdots$ \\

969: Pure H$_2$CO & 10 & $\cdots$ \\

970: Pure HCOOH & 10 & $\cdots$ \\

971: H$_2$O:CH$_4$\tablenotemark{b} 1:0.33 & 50 & 1 hr \\

972: H$_2$O:NH$_3$ 1:0.2 & 10 & 1 hr \\

973: \enddata

974: \tablenotetext{a}{The equivalent of 10$^{15}$cm$^{-2}$s$^{-1}$ photons (E$_\gamma>$ 6 eV) for the noted length of time}

975: \tablenotetext{b}{CH$_4$ does not have a resonance at 6$\mu$m, but its presence in a water ice matrix affects the shape of the feature.}

976: %\tablecomments{tablecomment}

977: \tablerefs{All spectra from Leiden University's Sackler Laboratory ice analog databases.}

978: \label{tab:6um_components}

979: \end{deluxetable}

980: \clearpage

981: 

982: \begin{deluxetable}{l c c c c c c c c} \rotate

983: \tablewidth{0 pt}

984: \tabletypesize{\footnotesize}

985: \tablecaption{Derived solid-state molecular abundances in our YSO sample.  The first column gives the object name, and the second through eighth columns

986: give the column densities of H$_2$O, CO$_2$, H$_2$CO, HCOOH, NH$_3$, SO$_2$, CH$_4$, and CH$_3$OH.}

987: \tablehead{\colhead{Object} & \colhead{$N$(H$_2$O)\tablenotemark{a}} & \colhead{$N$(CO$_2$)\tablenotemark{b}} & \colhead{$N$(H$_2$CO)} & \colhead{$N$(HCOOH)\tablenotemark{c}} & \colhead{$N$(NH$_3$)} & \colhead{$N$(SO$_2$)} & \colhead{$N$(CH$_4$)} & \colhead{$N$(CH$_3$OH)\tablenotemark{d}} %\\ \colhead{} & \colhead{10$^{18}$cm$^{-2}$} & \colhead{10$^{17}$cm$^{-2}$} & \colhead{10$^{16}$cm$^{-2}$} & \colhead{10$^{16}$cm$^{-2}$} & \colhead{10$^{17}$cm$^{-2}$} & \colhead{10$^{16}$cm$^{-2}$} & \colhead{10$^{17}$cm$^{-2}$} & \colhead{10$^{17}$cm$^{-2}$}}

988: \\ \colhead{} & \colhead{10$^{18}$cm$^{-2}$} & \colhead{\% H$_2$O} & \colhead{\% H$_2$O} & \colhead{\% H$_2$O} & \colhead{\% H$_2$O} & \colhead{\% H$_2$O} & \colhead{\% H$_2$O} & \colhead{\% H$_2$O}}

989: \startdata

990: IRAS 04016+2610 & 5.91 $\pm$ 0.2 & 17.4 $\pm$ 1.4 & 1.9 & 1.8 & $<$12.1 & $<$0.1  &  0.4 $\pm$ 0.4 &  8.3 $\pm$ 1.4 \\

991: IRAS 04108+2803A & 2.90 $\pm$ 0.3 & 9.5 $\pm$ 2.4 & 0.3 & 0.6 & $<$9.0 &  1.7 $\pm$ 1.1  &  8.0 $\pm$ 2.6 & $\cdots$  \\

992: IRAS 04108+2803B & 4.44 $\pm$ 0.2 & 11.1 $\pm$ 1.5 & 2.2 & 0.8 & $<$9.0 & $<$0.7\tablenotemark{e} &  5.6 $\pm$ 1.2 &  6.3 $\pm$ 1.5 \\

993: IRAS 04154+2823 & 1.07 $\pm$ 0.2 & 19.3 $\pm$ 7.9 & 0.1 & 0.0 & $<$28.4 & $<$0.5  &  23.2 $\pm$ 9.1 & $\cdots$  \\

994: IRAS 04169+2702 & 4.33 $\pm$ 0.2 & 10.1 $\pm$ 1.5 & 2.4 & 1.0 & $<$9.0 & $<$0.5\tablenotemark{e} &  3.9 $\pm$ 1.2 & $\cdots$  \\

995: IRAS 04181+2654A & 3.63 $\pm$ 0.2 & 11.4 $\pm$ 1.8 & 3.6 & 0.2 & $<$9.0 & $<$0.1\tablenotemark{e} &  3.1 $\pm$ 1.4 & $\cdots$  \\

996: IRAS 04181+2654B & 4.18 $\pm$ 0.4 & 15.6 $\pm$ 2.3 & 7.1 & 0.7 & $<$25.6 & $<$0.3  &  4.5 $\pm$ 2.0 & $\cdots$  \\

997: IRAS 04239+2436 & 4.44 $\pm$ 0.2 & 12.6 $\pm$ 1.5 & 2.4 & 0.0 & $<$12.5 &  0.6 $\pm$ 0.2  & $<$0.5 &  3.8 $\pm$ 1.5 \\

998: IRAS 04295+2251 & 2.82 $\pm$ 0.3 & 9.4 $\pm$ 2.3 & 1.7 & 0.4 & $<$15.3 & $<$1.0\tablenotemark{e} &  2.4 $\pm$ 1.8 & $\cdots$  \\

999: IRAS 04361+2547 & 7.38 $\pm$ 0.3 & 4.9 $\pm$ 1.5 & 0.0 & 0.9 & $<$9.0 & $<$0.1  &  4.3 $\pm$ 0.7 &  1.7 $\pm$ 1.5 \\

1000: IRAS 04365+2535 & 5.31 $\pm$ 0.2 & 15.2 $\pm$ 1.6 & 1.8 & 0.5 & $<$13.0 & $<$0.1  &  0.8 $\pm$ 0.7 &  8.5 $\pm$ 1.6 \\

1001: IRAS 04381+2540 & 7.92 $\pm$ 0.2 & 11.8 $\pm$ 0.8 & 0.8 & 0.5 & $<$9.2 &  0.2 $\pm$ 0.1  &  0.3 $\pm$ 0.3 & $\cdots$  \\

1002: DGTauB & 5.65 $\pm$ 0.6 & 7.5 $\pm$ 1.5 & 0.0 & 0.5 & $<$17.7 &  1.2 $\pm$ 0.6  & $<$0.4 &  3.5 $\pm$ 1.5 \\

1003: HLTau & 1.91 $\pm$ 0.2 & 7.3 $\pm$ 2.8 & 5.9 & 0.7 & $<$11.0 &  1.0 $\pm$ 0.5  & $<$1.2 &  4.0 $\pm$ 2.8 \\

1004: IC2087IR & 2.13 $\pm$ 0.2 & 19.5 $\pm$ 4.5 & 0.8 & 0.1 & $<$26.7 & $<$0.2  &  1.8 $\pm$ 2.2 &  8.7 $\pm$ 4.5 \\

1005: L1551IRS5 & 10.90 $\pm$ 0.2 & 8.0 $\pm$ 1.3 & 0.0 & 0.0 & $<$9.0 & $<$0.0  &  1.7 $\pm$ 0.4 &  3.2 $\pm$ 1.3 \\

1006: \enddata

1007: \tablecomments{$N$(H$_2$CO) and $N$(HCOOH) have typical absolute uncertainties of $\sim$10\%.}

1008: \tablenotetext{a}{Uncertainties listed are statistical, and derived from the residual differences between the data and our models. There are potential systematic errors as well, as described in Section~\ref{sec:h2o}.}

1009: \tablenotetext{b}{Using band strength of H$_2$O:CO$_2$ mixture; see Section~\ref{sec:chemdiscuss}.}

1010: \tablenotetext{c}{Based on 5.8\mic feature alone.}

1011: \tablenotetext{d}{$N$(CH$_3$OH) intermixed with CO$_2$, from high-resolution 15.2\mic feature fits.}

1012: \tablenotetext{e}{A feature was detected in the SO$_2$ wavelength range but with an overly-narrow width; value given is an upper limit.}

1013: \label{tab:abundances}

1014: \end{deluxetable}

1015: \clearpage

1016: 

1017: \begin{deluxetable}{l c c c} \tabletypesize{\small} \tablewidth{0 pt}

1018: \tablecaption{Laboratory spectra used in 15\mic fits.} 

1019: \tablehead{\colhead{Composition} & \colhead{Temperature (K)} & \colhead{Radiation\tablenotemark{a}} & \colhead{Reference}}

1020: \startdata 

1021: H$_2$O:CH$_3$OH:CO$_2$ 0.9:0.3:1 & 98 & $\cdots$ & 1 \\

1022: H$_2$O:CH$_3$OH:CO$_2$ 1:1:1 & 130 & $\cdots$ & 1 \\

1023: CO:H$_2$O 10:1 & 30 & 1 hr & 2 \\

1024: H$_2$O:CO$_2$ 10:1 & 10 & $\cdots$ & 3 \\

1025: Pure CO$_2$ & 10 & $\cdots$ & 3 \\

1026: \enddata

1027: \tablenotetext{a}{The equivalent of 10$^{15}$cm$^{-2}$s$^{-1}$ photons (E$_\gamma>$ 6 eV) for the noted length of time}

1028: %\tablecomments{tablecomment}

1029: \tablerefs{(1) \citet{EKS_99_laboratory}, (2) \citet{GSG_95_labsim}, (3) \citet{EBG_96_coco2iso}}

1030: \label{tab:15um_components}

1031: \end{deluxetable}

1032: \clearpage

1033: 

1034: \begin{deluxetable}{l c c c c c c c c} \tablewidth{0 pt} \tabletypesize{\small}

1035: \tablewidth{0 pt}

1036: \tablecaption{Percentage and column density $N$ (10$^{17}$cm$^{-2}$) of CO$_2$ ice present in each stage of processing along the 

1037: line of sight: cold, warm, and segregated (due to rounding, percentages may not sum to exactly 100\%).} 

1038: %\tablehead{\colhead{Object} & \colhead{Cold \%} & \colhead{$N$(CO$_2$)$_{cold}$} & \colhead{Warm \%} & \colhead{$N$(CO$_2$)$_{warm}$} & 

1039: %\colhead{Segregated \%} & \colhead{$N$(CO$_2$)$_{seg}$}}

1040: \tablehead{ & \multicolumn{2}{c}{Cold} & & \multicolumn{2}{c}{Warm} & & \multicolumn{2}{c}{Segregated} \\ \cline{2-3} \cline{5-6} \cline{8-9}

1041: \colhead{Object} & \colhead{\% Total} & \colhead{$N$} & & \colhead{\% Total} & \colhead{$N$} & & \colhead{\% Total} & \colhead{$N$}}

1042: \tablecolumns{7}

1043: \startdata

1044: IRAS 04016+2610 & 3 & 0.28 & & 71 & 7.3 & & 26 & 2.6 \\

1045: IRAS 04108+2803B & 33 & 1.6 & & 30 & 1.5 & & 37 & 1.8 \\ 

1046: IRAS 04239+2436 & 10 & 0.57 & & 87 & 4.9 & & 2 & 0.14 \\ 

1047: IRAS 04361+2547 & 62 & 2.2 & & 12 & 0.45 & & 26 & 0.95 \\ 

1048: IRAS 04365+2535 & 2 & 0.17 & & 61 & 4.9 & & 37 & 2.9 \\ 

1049: DG Tau B & 1 & 0.05 & & 76 & 3.2 & & 23 & 1.0 \\ 

1050: HL Tau & 32 & 0.45 & & 32 & 0.44 & & 36 & 0.51 \\ 

1051: IC 2087 IR & 9 & 0.35 & & 68 & 2.8 & & 23 & 0.95 \\ 

1052: L1551 IRS 5 & 22 & 1.9 & & 54 & 4.6 & & 24 & 2.1 \\ 

1053: \enddata

1054: \label{tab:procseg}

1055: \end{deluxetable}

1056: \clearpage

1057: 

1058: \begin{deluxetable}{l c c c c c c c c} \tablewidth{0 pt} \tabletypesize{\footnotesize}

1059: \tablewidth{0 pt}

1060: \tablecaption{Molecular cloud and Class 0 abundances.}

1061: \tablehead{ & \colhead{$N$(H$_2$O) (10$^{18}$cm$^{-2}$)} & \colhead{$N$(H$_2$O) (10$^{18}$cm$^{-2}$)} & \colhead{$N$(CO$_2$)/$N$(H$_2$O)\tablenotemark{a}} & \colhead{$N$(CO$_2$)/$N$(H$_2$O)} \\

1062: \colhead{Source} & \colhead{[This work]} & \colhead{[Lit., (Refs)]} & \colhead{[This work]} & \colhead{[Lit., (Refs)]}}

1063: \startdata

1064: CK 2 [Serpens] & 2.9 & 3.5 (1) & 0.29 & 0.33 (2) \\

1065: Elias 13 [Taurus] & 0.95 & 1.0 (3) & 0.14 & 0.15,0.19 (2,5) \\

1066: Elias 16 [Taurus] & 2.2 & 2.5 (4) & 0.2 & 0.18,0.23 (3,5) \\

1067: \hline

1068: CepE-MM & $<$29.7\tablenotemark{b} & $\cdots$ & $>$0.09 & $\cdots$ \\

1069: L1448 IRS2 & $<$28.6\tablenotemark{b} & $\cdots$ & $>$0.1 & $\cdots$ \\

1070: W3(OH) + W3 & $\cdots$ & $\cdots$ & $\cdots$ & $\cdots$ \\

1071: \enddata

1072: \tablenotetext{a}{$N$(CO$_2$) calculated using the band strength $A$ of pure CO$_2$ in order to facilitate comparison with published values.}

1073: \tablenotetext{b}{Calculated from a scaled pure H$_2$O spectrum.}

1074: \tablerefs{(1) \citet{Eiroa_89_serpensice}, (2) \citet{KBP_05_backgroundstars}, (3) \citet{Whittet_88_extinctionlaw}, (4) \citet{CAK_95_solidcotaurus}, (5) 

1075: \citet{Nummelin_01_co2inlowmass}}

1076: \label{tab:molcloud_abun}

1077: \end{deluxetable}

1078: \clearpage

1079: 

1080: \begin{figure}

1081:  \begin{center}

1082: % \includegraphics[width=0.9\textwidth]{fig_full_spectra.eps}

1083:  \includegraphics[width=0.9\textwidth]{f1.eps}

1084:  \end{center}

1085:  \caption{Our sample IRS spectra.  The lower left object has been marked with the locations and primary carriers of the most prominent absorption features.  Note

1086: that the 9.7\mic silicate feature may also appear partly or wholly in emission.}

1087:  \label{fig:fullspectra}

1088: \end{figure}

1089: \clearpage

1090: 

1091: \begin{figure}

1092:  \begin{center}

1093:  \includegraphics[width=0.7\textwidth]{f2.eps}

1094:  \end{center}

1095:  \caption{IRS spectra of the Class 0 protostars CepE-MM (in Cepheus) and L1448 LRS 2 (in Perseus), and the line of sight to W3(OH) through the W3 molecular cloud.

1096: The 9.7\mic silicate absorption feature in CepE-MM is saturated.}

1097:  \label{fig:otherobjects}

1098: \end{figure}

1099: \clearpage

1100: 

1101: \begin{figure}

1102:  \begin{center}

1103: % \includegraphics[width=0.8\textwidth]{fig_continuum-example.eps}

1104:  \includegraphics[width=0.8\textwidth]{f3.eps}

1105:  \end{center}

1106:  \caption{Example of the applied continuum determination using the spectrum of L1551 IRS 5.  The inset shows the additional continuum imposed on the 7-8\mic range.}

1107:  \label{fig:continuum-ex}

1108: \end{figure} 

1109: \clearpage

1110: 

1111: \begin{figure}

1112: % \includegraphics[]{fig_h2oco2_temp_triangular.eps}

1113:  \includegraphics[]{f4.eps}

1114:  \caption{Range of non-irradiated H$_2$O:CH$_3$OH:CO$_2$ laboratory compositions for which there is infrared spectroscopy in the Leiden University database (all circles).

1115: The vertices represent the pure molecular ices, and distance from each 

1116: molecule's vertex indicates decreasing relative abundance of that molecule (solid lines at increments of 20\%).  

1117: The filled circles represent mixtures whose $\tau_{6.0}$/$\tau_{15.2}$ 

1118: ratio falls within the range spanned by the ratios from the YSO spectra (Section~\ref{sec:trends}),

1119: and the dashed line indicates the compositional locus where CO$_2$/H$_2$O=0.12 (Section~\ref{sec:chemdiscuss}).}

1120:  \label{fig:h2oco2_temp}

1121: \end{figure} 

1122: \clearpage

1123: 

1124: \begin{figure}

1125:  \begin{center}

1126:  \includegraphics[width=0.45\textwidth]{f5a.eps}

1127:  \includegraphics[width=0.45\textwidth]{f5b.eps} 

1128:  \includegraphics[width=0.45\textwidth]{f5c.eps} 

1129:  \end{center}

1130:  \caption{Relationships between feature peak optical depths for our low-mass YSOs.  Panel (a)

1131: shows $\tau_{6.8\mu m}$ vs $\tau_{6.0\mu m}$, panel (b) shows $\tau_{15.2\mu m}$ vs $\tau_{6.0\mu m}$, and panel (c) shows $\tau_{15.2\mu m}$ vs $\tau_{6.8\mu m}$.  The

1132: dashed lines indicate the best linear fit.}

1133:  \label{fig:objratios}

1134: \end{figure} 

1135: \clearpage

1136: 

1137: \begin{figure}

1138:  \begin{center}

1139: % \includegraphics[width=0.7\textwidth]{fig_6umfits_example.eps}

1140:  \includegraphics[width=0.8\textwidth]{f6.eps}

1141:  \end{center}

1142:  \caption{The 6\mic feature of IRAS 04016+2610, fit by a combination of pure H$_2$O at 10 and 50 K (dots and short dashes), pure H$_2$CO at 10 K (dot-dash), 

1143: pure HCOOH at 10 K (dash-triple dots), and H$_2$O:NH$_3$ 10:2 at 10 K (long dashes).  The solid line is the sum.  The 6.8\mic feature is deliberately excluded from the fit.}

1144:  \label{fig:6umfit}

1145: \end{figure} 

1146: \clearpage

1147: 

1148: \begin{figure}

1149:  \begin{center}

1150: % \includegraphics[width=0.7\textwidth]{fig_cp_libh2o.eps}

1151:  \includegraphics[width=0.7\textwidth]{f7.eps}

1152:  \end{center}

1153:  \caption{Peak optical depth in the water ice libration mode ($\lambda$ $\sim$ 13 $\micron$), plotted against the water ice column density $N$(H$_2$O) 

1154: derived from the 6 $\micron$ water ice feature as described in Section~\ref{sec:h2o}.  Points and uncertainties are results for the nine Taurus objects 

1155: in which the 9.7\mic silicate features are strongly in absorption.}

1156:  \label{fig:libh2o}

1157: \end{figure} 

1158: \clearpage

1159: 

1160: \begin{figure}

1161:  \begin{center}

1162: % \includegraphics[width=0.75\textwidth]{fig_6.8um_spectra.eps}

1163:  \includegraphics[width=0.75\textwidth]{f8.eps}

1164:  \end{center}

1165:  \caption{The sample's 6-7\mic profiles (after subtraction of the H$_2$O contributions (dashed) and all contributions (solid) at 6 $\mu$m) 

1166: normalized to 1 and shifted vertically, sorted by 6.8\mic peak position (the point near 5.6\mic indicates the approximate optical depth uncertainty).

1167: The 5.6-5.7\mic residuals are almost entirely byproducts of continuum mismatches.  

1168: Also shown for comparison are normalized laboratory spectra---{\it Lab A}: H$_2$O:CO:NH$_3$ 10:2.5:2 at 165 K (irradiated to create NH$_4^+$), 

1169: {\it Lab B}: H$_2$O:CH$_3$OH:CO$_2$ 9:1:2 at 10 K, {\it Lab C}: H$_2$O:CH$_3$OH 1:100 at 10 K.}

1170:  \label{fig:6.8um_spectra}

1171: \end{figure} 

1172: \clearpage

1173: 

1174: \begin{figure}

1175:  \begin{center}

1176: % \includegraphics[width=0.7\textwidth]{fig_15umfits_example.eps}

1177:  \includegraphics[width=0.8\textwidth]{f9.eps}

1178:  \end{center}

1179:  \caption{15.2\mic feature of IRAS 04365+2535, shown repeated with a vertical offset.  The spectrum on the bottom is overlaid with a fit of 

1180: H$_2$O:CH$_3$OH:CO$_2$ 0.9:0.3:1 at 98 K (dots), H$_2$O:CH$_3$OH:CO$_2$ 1:1:1 at 130 K (dashes), and CO:H$_2$O 10:1 at 30 K 

1181: (dash-dot).  The solid line is the sum.  The heavy dashed fit on the upper spectrum is a fit using only H$_2$O:CH$_3$OH:CO$_2$ 1:1:1 at 118 K. 

1182: The excess between 14.95 and 15\mic is due to gas-phase CO$_2$ ro-vibrational $\nu_2$ Q-branch lines, blended together at this resolution.}

1183:  \label{fig:15umfit}

1184: \end{figure} 

1185: \clearpage

1186: 

1187: \begin{figure}

1188:  \begin{center}

1189: % \includegraphics[width=0.8\textwidth]{fig_silfit.eps}

1190:  \includegraphics[width=0.8\textwidth]{f10.eps}

1191:  \end{center}

1192:  \caption{Model of the absorption around 10\mic in IRAS 04016+2610.  Shown are the observed data (points), the GCS 3 silicate profile

1193: \citep[long dashes;][]{Kemper_04_silicates}, pure water ice at 10 K (dash-triple dots), and the two methanol-bearing mixtures described in Section~\ref{sec:15um_feature}

1194: and listed in Table~\ref{tab:15um_components} (short dashes and dot-dashes), scaled per this object's 15.2\mic fit.  The remnant peak at 9\mic is likely NH$_3$ ice.

1195: The model and observation are consistent with one another in the relevant wavelength range.}

1196:  \label{fig:silfit}

1197: \end{figure} 

1198: \clearpage

1199: 

1200: \begin{figure}

1201:  \begin{center}

1202: % \includegraphics[width=0.7\textwidth]{fig_bkgdfit_ck2.eps}

1203: % \includegraphics[width=0.7\textwidth]{fig_bkgdfit_e16.eps}

1204:  \includegraphics[width=0.7\textwidth]{f11a.eps}

1205:  \includegraphics[width=0.7\textwidth]{f11b.eps}

1206:  \end{center}

1207:  \caption{Comparison of fits to the Serpens background star CK 2 (top) and Taurus background star Elias 16 (bottom).  

1208: The left panels are repeats of the fits by \citet{KBP_05_backgroundstars}, where the

1209: dotted line is a laboratory spectrum of H$_2$O:CO$_2$ 10:1 at 10 K, the dashed line is H$_2$O:CO$_2$ 1:1 at 10 K, and the dash-dot line is CO:N$_2$:CO$_2$ 10:5:2 at 30 K.

1210: The

1211: mismatch on the red wing is caused by the forced continuum at $\sim$16.5 $\mu$m, which ignores the extent of some of the ice feature wings as measured in the laboratory.

1212: In the right panels, the dash-dot line is the same, but the H$_2$O:CO$_2$ 10:1 spectrum has been replaced with H$_2$O:CH$_3$OH:CO$_2$ 9:2:1 at 30 K (long dashes) and the

1213: H$_2$O:CO$_2$ 1:1 with H$_2$O:CH$_3$OH:CO$_2$ 0.9:0.3:1 at 30 K (dash-triple dot).}

1214:  \label{fig:bkgdfit}

1215: \end{figure}

1216: \clearpage

1217: 

1218: \begin{figure}

1219:  \begin{center}

1220: % \includegraphics[width=0.8\textwidth]{fig_waterco2_Nratio.eps}

1221:  \includegraphics[width=0.8\textwidth]{f12.eps}

1222:  \end{center}

1223:  \caption{Correlation between solid $N$(H$_2$O) and $N$(CO$_2$) for many different lines of sight.  

1224: Our sample comprises low-mass protostars, \citet{GWE_99_molcloud} contains

1225: mostly massive protostars and some galactic center sources, \citet{Nummelin_01_co2inlowmass} comprises low- and high-mass YSOs as well as some molecular cloud sight-lines,

1226: and \citet{BPL_04_lowmass} contains two low-mass protostars.  The solid line is the ratio between column densities using only the current work's sample, 

1227: and the dashed line is the ratio using all objects except the upper outlier from \citet{BPL_04_lowmass}.  Note that this work uses a CO$_2$ band

1228: strength (at 15.2$\mu$m) $\sim$36\% higher than that of the previous studies, which may account for most of the difference in $N$(CO$_2$)/$N$(H$_2$O).}

1229:  \label{fig:waterco2_Nratio}

1230: \end{figure} 

1231: \clearpage

1232: 

1233: \begin{figure}

1234:  \begin{center}

1235: % \includegraphics[width=0.7\textwidth, angle=90]{fig_alpha_abun.eps}

1236: % \includegraphics[width=0.7\textwidth]{fig_alpha_co22h2o.eps}

1237:  \includegraphics[width=0.45\textwidth]{f13a.eps}

1238:  \includegraphics[width=0.45\textwidth]{f13b.eps}

1239:  \end{center}

1240:  \caption{Relationships between the SED spectral index ($\alpha$; 7-30 $\mu$m) and $N$(CO$_2$), $N$(H$_2$O), and $N$(CO$_2$)/$N$(H$_2$O).

1241: The filled circles represent our Class I/II data, and the open circles represent the Class 0 objects.}

1242:  \label{fig:alpha_abun}

1243: \end{figure}

1244: \clearpage

1245: 

1246: \begin{figure}

1247:  \begin{center}

1248: % \includegraphics[width=0.6\textwidth,angle=90]{fig_waterhcoohmeth_Nratio.eps}

1249:  \includegraphics[width=0.8\textwidth]{f14.eps}

1250:  \end{center}

1251:  \caption{Ratios between $N$(H$_2$O), and $N$(HCOOH) and $N$(CH$_3$OH), for the Class I/II sample (Section~\ref{sec:chemdiscuss}).  The bottom plot refers to 

1252: $N$(CH$_3$OH) embedded in a H$_2$O:CO$_2$ matrix.}

1253:  \label{fig:waterhcoohmeth_Nratio}

1254: \end{figure}

1255: \clearpage

1256: 

1257: \begin{figure}

1258:  \begin{center}

1259: % \includegraphics[]{fig_6.8um15.2um_vs_15.4um15.2um_2.eps}

1260:  \includegraphics[]{f15.eps}

1261:  \end{center}

1262:  \caption{The correlation between 6.8\mic and 15.4\mic optical depths in the laboratory ice spectra (plus signs) and the observed YSO spectra (solid circles).

1263: Both have been scaled by peak 15.2 $\mu$m optical depth.  The laboratory data cover temperatures from 10-185K and CH$_3$OH/CO$_2$ abundance ratios of 0-1.4, which are 

1264: roughly indicated by the solid lines: [$\lesssim$0.5] {\bf /} [0.5-1.4] {\bf /} [$\gtrsim$1.4].  The vertical spread depends largely on temperature, with warmer mixtures

1265: of a constant composition showing weaker 15.4\mic shoulders.}

1266:  \label{fig:6.8um15.2um_vs_15.4um15.2um}

1267: \end{figure} 

1268: \clearpage

1269: 

1270: \begin{figure}

1271:  \begin{center}

1272: % \includegraphics[width=0.8\textwidth]{fig_watermethane_Nratio.eps}

1273:  \includegraphics[width=0.8\textwidth]{f16.eps}

1274:  \end{center}

1275:  \caption{Comparison of solid $N$(H$_2$O) and $N$(CH$_4$) for the YSO sample.  

1276: The dashed line indicates the maximum of $\sim$4\% (CH$_4$ relative to H$_2$O) observed towards massive protostars \citep[e.g.][]{GWB_04_ice}, and the 

1277: dotted line is the newer 13\% maximum reported by \citet{Oberg_08_ch4} towards low-mass protostars.}

1278:  \label{fig:watermethane_Nratio}

1279: \end{figure}

1280: \clearpage

1281: 

1282: \begin{figure}

1283:  \begin{center}

1284: % \includegraphics[width=0.4\textwidth,angle=90]{fig_d2p_vs_temp.eps}

1285:  \includegraphics[width=0.4\textwidth,angle=90]{f17.eps}

1286:  \end{center}

1287:  \caption{The ``dip-to-peak'' ratio as a measure of the amount of CO$_2$ segregation.  {\it (a)}---Demonstrates how the values were measured for each laboratory ice

1288: spectrum, using pure CO$_2$ as an example.  

1289: {\it (b)}---The relative dip strength as a function of laboratory ice temperature, showing the rapid onset of the process near $\sim$105 K.

1290: Much of the scatter in (b) is due to the dependence of CO$_2$ segregation temperature on the exact composition of the matrix in which it is embedded.  

1291: The large asterisks in the upper left represent pure CO$_2$ ice.}

1292:  \label{fig:d2p_vs_temp}

1293: \end{figure} 

1294: \clearpage

1295: 

1296: \begin{figure}

1297:  \begin{center}

1298: % \includegraphics[width=0.7\textwidth]{fig_15um_hires_spectra.eps}

1299:  \includegraphics[width=0.7\textwidth]{f18.eps}

1300:  \end{center}

1301:  \caption{The high-resolution 15\mic Class I/II profiles visually ordered by ``peakiness,'' normalized to 1, and vertically offset for clarity. 

1302: The dotted lines at 15.1 and 15.25\mic mark the peak positions of pure CO$_2$, and the one

1303: at 15.4\mic indicates the CH$_3$OH:CO$_2$-complex shoulder.  The top and bottom spectra are laboratory mixtures as indicated.}

1304:  \label{fig:15um_hires_spectra}

1305: \end{figure} 

1306: \clearpage

1307: 

1308: \begin{figure}

1309:  \begin{center}

1310: %  \includegraphics[width=0.6\textwidth]{fig_alpha_process.eps}

1311:   \includegraphics[width=0.6\textwidth]{f19.eps}

1312:  \end{center}

1313:  \caption{Relationship between the SED spectral index ($\alpha$; 7-30 $\mu$m) and the CO$_2$ processing fractions in our Class I/II sample.}

1314:  \label{fig:alpha_seg}

1315: \end{figure}

1316: \clearpage

1317: 

1318: \begin{figure}

1319:  \begin{center}

1320: % \includegraphics[width=0.7\textwidth]{fig_w3ohfit.eps}

1321:  \includegraphics[width=0.7\textwidth]{f20.eps}

1322:  \end{center}

1323:  \caption{15.2\mic CO$_2$ feature probing W3(OH) and the foreground W3 cloud.  

1324: It has been fitted with a combination of H$_2$O:CH$_3$OH:CO$_2$ 0.9:0.3:0.1 at 98 K (dotted line),

1325: H$_2$O:CH$_3$OH:CO$_2$ 1:1:1 at 130 K (dashed line), CO:H$_2$O at 10:1 at 30 K (dot-dashed line), and H$_2$O:CO$_2$ 10:1 at 10 K (dash-triple dot).  

1326: The solid line is the sum.}

1327:  \label{fig:w3ohfit}

1328: \end{figure}

1329: \clearpage

1330: 

1331: \end{document}

1332: 

1333: