1: %%% This file is now modified for ApJ use (instead of A&A before)
2: %%
3: %% Beginning of file 'sample.tex'
4: %%
5: %% Modified 03 Jan 01
6: %%
7: %% This is a sample manuscript marked up using the
8: %% AASTeX v5.x LaTeX 2e macros.
9:
10: %% The first piece of markup in an AASTeX v5.x document
11: %% is the \documentclass command. LaTeX will ignore
12: %% any data that comes before this command.
13:
14: %% The command below calls the preprint style
15: %% which will produce a one-column, single-spaced document.
16: %% Examples of commands for other substyles follow. Use
17: %% whichever is most appropriate for your purposes.
18:
19: \documentclass[12pt,preprint]{aastex}
20:
21: %% manuscript produces a one-column, double-spaced document:
22:
23: % \documentclass[manuscript]{aastex}
24:
25: %% preprint2 produces a double-column, single-spaced document:
26:
27: % \documentclass[preprint2]{aastex}
28:
29: %\usepackage{emulateapj5} % emulates original format
30: %\topmargin1cm % to accomodate linus-sun differences
31: \usepackage{natbib}
32: \usepackage{longtable}
33: \slugcomment{Draft version from \today}
34:
35: \shorttitle{Spectral line data toward IRDCs}
36: \shortauthors{Beuther \& Sridharan}
37:
38: %% This is the end of the preamble. Indicate the beginning of the
39: %% paper itself with \begin{document}.
40:
41: \begin{document}
42:
43: \title{Outflow and dense gas emission from massive Infrared Dark Clouds}
44:
45: %% Use \author, \affil, and the \and command to format
46: %% author and affiliation information.
47: %% Note that \email has replaced the old \authoremail command
48: %% from AASTeX v4.0. You can use \email to mark an email address
49: %% anywhere in the paper, not just in the front matter.
50: %% As in the title, you can use \\ to force line breaks.
51:
52: \author{H.~Beuther$^1$ \& T.K. Sridharan$^2$}
53: \altaffiltext{1}{Max-Planck-Institute for Astronomy, K\"onigstuhl 17, 69117 Heidelberg, Germany}
54: \altaffiltext{2}{Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138, USA}
55: \email{beuther@mpia.de, tksridha@cfa.harvard.edu}
56:
57: \begin{abstract}
58: Infrared Dark Clouds are expected to harbor sources in different,
59: very young evolutionary stages. To better characterize these
60: differences, we observed a sample of 43 massive Infrared Dark
61: Clouds, originally selected as candidate high-mass starless cores,
62: with the IRAM\,30\,m telescope covering spectral line tracers of
63: low-density gas, high-density gas, molecular outflows/jets and
64: temperature effects. The SiO(2--1) observations reveal detections
65: toward 18 sources. Assuming that SiO is exclusively produced by
66: sputtering from dust grains, this implies that at least in 40\% of
67: this sample star formation is on-going. A broad range of SiO
68: line-widths is observed (between 2.2 and 65\,km\,s$^{-1}$), and we
69: discuss potential origins for this velocity spread. While the
70: low-density tracers $^{12}$CO(2--1) and $^{13}$CO(1--0) are detected
71: in several velocity components, the high-density tracer
72: H$^{13}$CO$^+$(1--0) generally shows only a single velocity
73: component and is hence well suited for kinematic distance estimates
74: of IRDCs. Furthermore, the H$^{13}$CO$^+$ line-width is on average
75: 1.5 times larger than that of previously observed NH$_3$(1,1). This
76: is indicative of more motion at the denser core centers, either due
77: to turbulence or beginning star formation activity. In addition, we
78: detect CH$_3$CN toward only six sources whereas CH$_3$OH is observed
79: toward approximately 40\% of the sample. Estimates of the CH$_3$CN
80: and CH$_3$OH abundances are low with average values of $1.2\times
81: 10^{-10}$ and $4.3\times 10^{-10}$, respectively. These results are
82: consistent with chemical models at the earliest evolutionary stages
83: of high-mass star formation. Furthermore, the CH$_3$OH abundances
84: compare well to recently reported values for low-mass starless
85: cores.
86: \end{abstract}
87:
88: \keywords{ stars: formation -- stars: early-type -- ISM: jets and
89: outflows -- ISM: abundances -- ISM: clouds -- ISM: kinematics and
90: dynamics}
91: %________________________________________________________________
92:
93:
94: \section{Introduction}
95: \label{intro}
96:
97: While observational research in high-mas star formation over the last
98: few decades has focused on relatively evolved regions like
99: Ultracompact H{\sc ii} regions (UCH{\sc ii}s) and High-Mass
100: Protostellar Objects (HMPOs) (e.g., see reviews by
101: \citealt{kurtz2000,churchwell2002,beuther2006b}), the earliest stages
102: of massive star formation have become regularly accessible only
103: through the advent of large-scale mid-infrared surveys of the Galactic
104: plane by satellite missions like the Infrared Space Observatory (ISO),
105: the Midcourse Space Experiment (MSX) and most recently Spitzer. The
106: basic observational characteristics of the earliest stages of
107: (massive) star formation, prior to the formation of any embedded
108: heating source, should be that they are strong cold dust emitters at
109: (sub)mm wavelengths, and weak or non-detections in the mid-infrared
110: because they have not yet heated a warm dust cocoon. In the low-mass
111: regime, so-called Infrared Dark Clouds (IRDCs) have been studied
112: regularly since a few years (e.g.,
113: \citealt{egan1998,carey2000,bacmann2000}), in high-mass star formation
114: the focus has shifted to these objects only more recently (e.g.,
115: \citealt{sridharan2005,hill2005,klein2005,simon2006,beltran2006,rathborne2006,pillai2006}).
116:
117: Massive IRDCs (in short IRDCs from now) are not a well defined class,
118: but they likely harbor various evolutionary stages, from genuine
119: High-Mass Starless Cores (HMSCs) via High-Mass Cores harboring
120: accreting Low/Intermediate-Mass Protostars to the youngest HMPOs
121: \citep{beuther2006b}. While the first stage provides good targets to
122: study the initial conditions of massive star formation prior to cloud
123: collapse, the other stages are important to understand the early
124: evolution of massive star-forming clumps. Here, we are targeting a
125: sample of massive IRDCs first identified as candidate HMSCs by
126: \citet{sridharan2005} to differentiate between the different
127: evolutionary stages existing within this sample. To facilitate this,
128: we observed the target sample with the IRAM 30\,m telescope in
129: different spectral lines sensitive to various physical processes like
130: molecular outflows (e.g., SiO, CO), dense gas (e.g., H$^{13}$CO$^+$,
131: CH$_3$CN, CH$_3$OH) or hot core conditions (CH$_3$CN, CH$_3$OH). Table
132: \ref{observing} lists the observed spectral lines, the upper level
133: excitation energies $E_{\rm{upper}}$ and the critical densities
134: $n_{\rm{crit}}$.
135:
136: \section{Observations and Data Reduction}
137: \label{obs}
138:
139: We observed a sample of 43 IRDCs selected from \citet{sridharan2005}
140: in August 2005 at the IRAM 30\,m telescope on Pico Veleta with four
141: receivers simultaneously. The flexible multi-receiver setup of the
142: 30\,m telescope allowed us to observe usually a few lines within the
143: same spectral setup of each receiver ((1) H$^{13}$CO$^+$(1--0) \&
144: SiO(2--1), (2) $^{13}$CO(1--0) and CH$_3$CN, (3) $^{12}$CO(2--1), (4)
145: CH$_3$OH). Table \ref{observing} lists the corresponding frequencies,
146: the spectral resolution and the system temperatures for each line. The
147: observations were conducted in a position-switching mode with OFF
148: positions carefully chosen from the Massachusetts-Stony Brook Galactic
149: plane CO survey \citep{sanders1986}, usually approximately of the
150: order $1000''$ offset from the source positions.
151:
152: \begin{table*}[htb]
153: \caption{Observing parameters}
154: \begin{center}
155: \begin{tabular}{lrrrrrr}
156: \hline \hline
157: Line & Freq. & $\Delta v$ & $T_{\rm{sys}}^a$ & $1\sigma$\,rms$^a$ & $E_{\rm{upper}}$ & $n_{\rm{crit}}^b$ \\
158: & GHz & km\,s$^{-1}$ & K & K & K & cm$^{-3}$\\
159: \hline
160: H$^{13}$CO$^+$(1--0) & 86.754 & 1.1 & 110 & 0.02 & 4.2 & 1.7e5\\
161: SiO(2--1) & 86.847 & 1.1 & 110 & 0.02 & 6.3 & 7.3e5\\
162: $^{13}$CO(1--0) & 110.201 & 0.9 & 220 & 0.04 & 5.3 & 2.2e3\\
163: CH$_3$CN$(6_0-5_0)$ & 110.384 & 0.9 & 220 & 0.03 & 18.5 & 1.5e6$^c$\\
164: CH$_3$CN$(6_1-5_1)$ & 110.381 & 0.9 & 220 & 0.03 & 25.7 & 1.5e6$^c$\\
165: CH$_3$CN$(6_2-5_2)$ & 110.375 & 0.9 & 220 & 0.03 & 47.1 & 1.6e6$^c$\\
166: $^{12}$CO(2--1) & 230.538 & 1.6 & 660 & 0.30 & 16.6 & 1.0e4\\
167: CH$_3$OH$(5_1-4_1)$E & 241.879 & 1.2 & 760 & 0.13 & 55.9 & 1.1e6\\
168: CH$_3$OH$(5_0-4_0)$A+& 241.791 & 1.2 & 760 & 0.13 & 34.8 & $^d$\\
169: CH$_3$OH$(5_{-1}-4_{-1})$E&241.767&1.2&760 & 0.13 & 40.4 & 9.5e5\\
170: CH$_3$OH$(5_0-4_0)$E & 241.700 & 1.2 & 760 & 0.13 & 47.9 & 1.0e6\\
171: \hline \hline
172: \end{tabular}
173: \end{center}
174: $^a$ Average system temperature and $1\sigma$\,rms values.\\
175: $^b$ Calculated from $A=0.3\lambda_{100}^{-3}\mu ^2$ ($\lambda_{100}$ in units of
176: 100\,$\mu$m and $\mu=3.9$\,debye) and $\gamma(\rm{CH_3CN}(6-5))_{20K}$ from \citet{pei1995a,pei1995b}.\\
177: $^c$ $n_{\rm{crit}} = A/\gamma$ with the Einstein coefficient $A$ and
178: the collisional rate $\gamma$. Except for CH$_3$CN, $A$ and $\gamma$
179: were taken from LAMBDA \citep{schoeier2005} at $T=20$\,K.\\
180: $^d$ No data for this transition in LAMBDA.
181: \label{observing}
182: \end{table*}
183:
184: \section{Results and discussion}
185:
186: \subsection{$^{12}$CO(2--1), $^{13}$CO(1--0) and H$^{13}$CO$^+$(1--0)}
187:
188: Figures \ref{sample1}, \ref{sample2} and \ref{sample3} show the
189: $^{12}$CO(2--1), $^{13}$CO(1--0) and H$^{13}$CO$^+$(1--0) spectra for
190: the whole source sample. Except of IRDC\,18431-0312-4, where the
191: H$^{13}$CO$^+$(1--0) line remained undetected, toward all other
192: sources the three lines were detected. The most obvious difference
193: between $^{12}$CO/$^{13}$CO on the one side and H$^{13}$CO$^+$ on the
194: other side is that the CO isotopologues are detected at many different
195: velocities whereas H$^{13}$CO$^+$ usually only shows a single spectral
196: peak (except of a few double-peaked structures listed in Table
197: \ref{sourceparameters}). This difference can be explained
198: straightforwardly by the location of the sources within the Galactic
199: plane because CO traces more or less all molecular clouds along the
200: line of sight, hence contributions from various Galactic spiral arms,
201: whereas H$^{13}$CO$^+$ has significantly higher critical densities
202: (Table \ref{observing}) and thus traces only the dense cores we are
203: interested in. Therefore, if one wants to know the velocity of a
204: specific region CO observations are mostly not useful and one has to
205: refer to high density tracers. The peak velocities $v$ of
206: H$^{13}$CO$^+$ (Table \ref{sourceparameters}) correspond well to the
207: previously derived velocities from NH$_3$ observations of the same
208: sample \citep{sridharan2005}. Identifying the correct associated
209: velocity component to derive accurate velocities is an important step
210: to get distance estimates from the Galactic rotation curve as reported
211: in \citet{brand1993} and \citet{sridharan2005}.
212:
213: Table \ref{sourceparameters} also lists the Full Width Half Maximum
214: (FWHM) line-widths $\Delta v(\rm{H^{13}CO^+})$ of the observed
215: H$^{13}$CO$^+$ lines from Gaussian fits, the average H$^{13}$CO$^+$
216: line-width is 2.6\,km\,s$^{-1}$. Figure \ref{dv} shows a comparison
217: of the H$^{13}$CO$^+$ line-width with the previously derived
218: line-width from the NH$_3$(1,1) observations from
219: \citep{sridharan2005}. We find that the H$^{13}$CO$^+$ line-width
220: nearly always exceeds the line-width from the NH$_3$(1,1) with an
221: average line-width ratio $\frac{\Delta v(\rm{H^{13}CO^+(1-0)})}{\Delta
222: v(\rm{NH_3(1,1)})}=1.5$. This is interesting since the critical
223: densities for both molecules differ by about 2 orders of magnitude
224: ($n_{\rm{crit}}(\rm{NH_3}(1,1))\sim 2\times 10^3$\,cm$^{-3}$, e.g.
225: \citealt{zhou1989}, $n_{\rm{crit}}(\rm{H^{13}CO^+(1-0)})\sim 1.7\times
226: 10^5$\,cm$^{-3}$, Table \ref{observing}), whereas the thermal
227: line-widths ($\Delta
228: v_{\rm{th}}=\sqrt{8{\rm{ln}}2\frac{kT}{m_{\rm{mol}}}}$ with the
229: molecular mass $m_{\rm{mol}}$) are similar ($\Delta
230: v_{\rm{th}}(\rm{NH_3}(1,1))_{15\rm{K}}\sim 0.2$\,km\,s$^{-1}$ and
231: $\Delta v_{\rm{th}}(\rm{H^{13}CO^+(1-0)})_{15\rm{K}}\sim
232: 0.15$\,km\,s$^{-1}$), $\Delta v_{\rm{th}}(\rm{H^{13}CO^+(1-0)})$ being
233: even smaller than $\Delta v_{\rm{th}}(\rm{NH_3}(1,1))$ because of its
234: larger molecular weight. Therefore, differences in the non-thermal
235: motions have to be responsible for the line-width variations
236: between both species. Similar results were obtained for low-mass
237: cores in a comparison of CS with NH$_3$ line-widths \citep{zhou1989}.
238:
239: There exist several potential ways to explain such a line-width
240: difference. For example, there is the well-known line-width--size
241: relation $\Delta v \propto R^{0.5}$ \citep{larson1981}. This relation
242: would imply that our target regions would be on average 2.3 times
243: larger in H$^{13}$CO$^+$ than they are in NH$_3$. Although we have no
244: maps of the regions, this appears very unlikely. A different
245: possibility was discussed by \citet{zhou1989} and \citet{myers1983}:
246: If the stability of a core is determined by the entire
247: line-broadening, for a gravitationally stable core the critical
248: density should relate to the line-width and size of a molecular core
249: as
250: $$n_{\rm{crit}}\propto (\Delta v/R)^2 \Longleftrightarrow \Delta v \propto R\sqrt{n_{\rm{crit}}}.$$
251: To fulfill this criterion the H$^{13}$CO$^+$ source size would need
252: to be $\sim$7 times smaller than that of NH$_3$ which
253: again appears to be very unlikely. In this scenario, one could then
254: conclude that molecular clumps are not gravitationally stable but
255: maybe collapsing already. However, overall collapse is not sure at all
256: from such an analysis, and the clumps could also be fragmenting or
257: only partially collapsing (see discussion in \citealt{zhou1989}).
258:
259: As discussed below in \S\ref{sio}, a significant fraction of sources
260: exhibits broad line wings in the jet/outflow tracer SiO, which is
261: indicative of early outflow and hence early star formation activity.
262: Since H$^{13}$CO$^+$ has a larger critical density it traces denser
263: regions of the gas clumps. With ongoing star formation activity, one
264: expects additional line broadening effects from various physical
265: processes like outflows, infall and potential rotation of embedded
266: accretion disks, all spatially unresolved by the current observations.
267: Most of these effects are likely contributing more strongly in the
268: dense gas components and hence could explain the excess in
269: H$^{13}$CO$^+$ line-width.
270:
271: \subsection{SiO(2--1)}
272: \label{sio}
273:
274: One of the aims of this study was to identify how many of the sources
275: would show clear SiO emission caused by molecular outflow activity.
276: SiO is believed to be produced in the gas-phase after sputtering
277: Si-bearing species from the dust grain surfaces when the outflow/jet
278: impinges on the ambient gas and dust \citep{schilke1997a}. SiO is usually
279: barely detectable in quiescent clouds. Figure \ref{siospectra}
280: presents the 18 SiO(2--1) detections toward our source sample of 43
281: regions. If this source sample is representative for typical Infrared
282: Dark Clouds, this implies that in at least 40\% of the IRDCs star
283: formation has already started (absence of SiO emission does not
284: necessarily imply absence of star formation activity). However, it is
285: not obvious whether this sample is representative for all IRDCs
286: because it was selected to be close to sites of already ongoing
287: massive star formation.
288:
289: Table \ref{sourceparameters} also lists the width-to-zero-velocity
290: $\Delta v_0$ of the observed SiO(2--1) lines. If one divides the
291: observed SiO spectra by their width-to-zero-velocity $\Delta v_0$ into
292: 3 regimes (a) low velocity $\Delta v_0 < 10$\,km\,s$^{-1}$, (b)
293: intermediate velocity 10\,km\,s$^{-1}<\Delta v_0<20$\,km\,s$^{-1}$ and
294: (c) high velocity $\Delta v_0 > 20$\,km\,s$^{-1}$, one finds six
295: sources in the low-velocity regime, five at intermediate velocities
296: and seven in the high-velocity regime. While the sources at
297: intermediate to high velocities obviously show gas at relatively high
298: velocities with respect to the ambient gas, this is less clear for the
299: sources in the low-velocity regime: those sources with $\Delta v_0 >
300: 5$\,km\,s$^{-1}$ probably also have outflow contributions, but the
301: SiO(2--1) spectra toward two sources with $\Delta v_0 <
302: 2.5$\,km\,s$^{-1}$ (IRDC\,18223-6 \& IRDC\,18530-2) hardly look like
303: typical outflow spectra. A similar spread in SiO velocities has also
304: previously been observed by, .e.g., \citealt{lefloch1998,codella1999}.
305: What are possible reasons for the different observed SiO $\Delta v_0$?
306:
307: The observed velocity spread could partly be produced by the angles
308: between the outflows and the plane of the sky since we only observe
309: line-of-sight velocities. However, assuming a random distribution of
310: outflow inclination angles $\theta$ ($\theta = 0$ is an outflow along
311: the line of sight) and similar unprojected outflow velocities, one
312: would expect fewer outflows at high velocities than at low velocities
313: ($\int_0^{45} sin(\theta)d\theta\sim 0.29$ vs. $\int_{45}^{90}
314: sin(\theta)d\theta\sim 0.71$). This is in contrast to our relatively
315: even distribution of $\Delta v_0$ from low to high values. Hence,
316: inclination effects are unlikley to explain the observations.
317:
318: Outflows/jets from massive protostars have intrinsically higher
319: velocities than those of their low-mass counterparts (e.g.,
320: \citealt{richer2000}). In addition, more luminous sources usually
321: exhibit stronger SiO emission (e.g., \citealt{codella1999}).
322: Therefore, the sources with smaller line-width could harbor less
323: massive protostars that may or may not form a massive star at the end
324: of their evolution. Furthermore, all objects with low SiO velocity
325: spread are also the weakest or non-detections in CH$_3$CN and CH$_3$OH
326: (see \S\ref{ch3cn_ch3oh}) which is also expected in the framework of
327: embeedded protostars with different masses. Thus, different
328: masses/luminosities of the embedded outflow-driving sources can
329: potentially cause the observed SiO line-width difference.
330:
331: Since IRDCs are believed to represent the earliest evolutionary stages
332: of massive (and low-mass) star formation, we expect that in some of
333: them low- to intermediate-mass protostars may be embedded that are
334: likely destined to grow, accrete and become high-mass stars at the end
335: of their evolution (e.g., IRDC\,18223-3,
336: \citealt{beuther2005d,beuther2007a}). However, a priori we cannot
337: differentiate whether an object will become massive at the end of its
338: evolution or whether it will remain a low-mass protostar. In this
339: scenario, one can also speculate about an evolutionary sequence where
340: the growing protostars accelerates the gas to higher velocities in the
341: course of its early evolution, i.e., the smaller line-widths may
342: correspond to the still less massive and younger objects whereas the
343: broad line-width sources may have already accreted more mass and could
344: hence be a little bit more evolved. This picture is consistent with
345: the above assertion that the line-width distribution can depend on the
346: mass and luminosity of the central object, just with the additional
347: assumption that the lower-mass objects may still be able to become
348: massive stars during their ongoing evolution. This scenario may appear
349: counter-intuitive compared to typical low-mass outflows where the
350: strongest emission is observed from class 0 sources and where the SiO
351: emission gets weaker formore evolved class I and II sources (e.g.,
352: \citealt{gibb2004}). However, even class 0 sources have typically
353: already ages of the order $10^4$\,yrs, and IRDCs like those presented
354: here are likely still significantly younger (e.g.,
355: \citealt{beuther2007a,krumholz2006b}). Furthermore, this low-mass
356: classification scheme is unlikely to be well applicable to high-mass
357: star formation because, among other reasons, massive stars reach the
358: main sequence while accretion processes can still be ongoing (e.g.,
359: \citealt{schaller1992,mckee2003}). Therefore, our picture applies only
360: to the earliest evolutionary stages of strongly accreting protostars
361: destined to become massive stars at the end of their evolution. In
362: principle, the non-detection of SiO line-wings could also be caused by
363: poorer signal-to-noise ratios or larger distances to the respective
364: region. However, both scenarios are unlikely the case here because we
365: have approximately the same rms for all sources, and the line-width
366: differences are found also toward sources at the same distances (e.g.,
367: IRDC\,18223-1243-3 \& IRDC\,18223-1243-4 versus IRDC\,18223-1243-6).
368:
369: Do the $\Delta v_0 < 2.5$\,km\,s$^{-1}$ fit into any of these
370: pictures? Since more than 50\% of the target sources exhibit no SiO
371: emission at all down to our sensitivity limit, it is interesting that
372: two objects show these extremely narrow line-widths.
373: \citet{lefloch1998} and \citet{codella1999} try to explain similar
374: observations by outflow/dense clump interactions and in an outflow
375: evolutionary model. In the latter scenario, the highest velocity
376: components are found at the earliest evolutionary class 0 stages and
377: the smaller line-width components could be remnants from earlier
378: interactions of molecular outflows with the ambient gas
379: \citep{codella1999}. They estimated the corresponding time-scale to
380: decrease the SiO emission and the velocities to be of the order
381: $10^4$\,yrs. We consider this scenario as unlikely for our sample
382: because massive star formation proceeds on relatively short
383: time-scales (of the order $10^5$ years, \citealt{mckee2002}), and
384: after $10^4$\,yrs the regions are expected to be strong enough to be
385: detetable at near- and mid-infrared wavelengths (e.g., the samples
386: presented by \citealt{molinari1996} or \citealt{sridha}) which is in
387: contrast to our sample selection as Infrared Dark Clouds. Another
388: possibility is to speculate whether these objects could represent the
389: earliest observed stage of star formation within our source sample: In
390: this picture, the Si-bearing species would just be released from the
391: grains by shock interaction and SiO would only have begun to form in
392: the gas phase, but the young jet/outflow would not have had the time
393: yet to accelerate large amounts of gas to detectable higher
394: velocities. To confirm or discard this picture additional indicators
395: of star formation activity would need to be observed, e.g., infall of
396: gas via asymmetric line profiles (e.g., \citealt{myers1996}) or
397: shocked H$_2$ emission either in the K-band (e.g.,
398: \citealt{davis2004}) or in the IRAC 4.5\,$\mu$m band (e.g.,
399: \citealt{rathborne2005}).
400:
401: \subsection{CH$_3$CN$(6_k-5_k)$ and CH$_3$OH(5--4)}
402: \label{ch3cn_ch3oh}
403:
404: In addition to the previously discussed molecules, the spectral setup
405: allowed to observe also the typical hot-core tracer CH$_3$CN as well
406: as a series of four CH$_3$OH lines. Although the critical densities
407: and excitation parameters of both molecules do not differ that much
408: (see Table \ref{observing}), CH$_3$OH typically has one to two orders
409: of magnitude larger gas column densities and is hence easier to
410: observe (e.g., \citealt{vandishoek1998,nomura2004}). Furthermore,
411: CH$_3$OH is known to be present in massive star formation right from
412: the beginning, hence it is a so-called parent molecule, whereas
413: CH$_3$CN is a daughter molecule which needs time to form through
414: chemical processes during the massive star formation process (e.g.,
415: \citealt{nomura2004}). Therefore toward a sample of young IRDCs, one
416: would expect CH$_3$CN to be either undetectable or only weakly
417: detected in a few sources, whereas CH$_3$OH should be detectable
418: toward a larger fraction of the sample.
419:
420: These expectations are largely confirmed by our observations (Figs.
421: \ref{ch3cn} \& \ref{ch3oh}), we detect CH$_3$CN only toward 6 sources
422: ($\sim$14\%) whereas CH$_3$OH is observed toward 17 regions
423: ($\sim$40\%). Furthermore, if detected, CH$_3$CN$(6_k-5_k)$ is
424: observed only in its two lowest $k$-components with excitation
425: temperatures $<30$\,K, the $k=2$ transition is detected in no source
426: at all. Similarly, for CH$_3$OH, we detected toward the 17 regions the
427: two lower excitation lines, whereas the two higher excitation lines
428: are detected only toward a small sub-sample of 5 regions. This is
429: additional confirmation of the youth and low average temperatures of
430: the selected source sample. Table \ref{sourceparameters2} lists the
431: line-widths $\Delta\nu_{k=0}$ and $\Delta\nu_{k=1}$ of the CH$_3$CN
432: $k=0,1$ components from Gaussian fits to each line. While we cannot
433: fit one source because of multiple components and in one other source
434: $\Delta\nu_{k=0}$ and $\Delta\nu_{k=1}$ are of approximately equal
435: width, for the four other sources we find consistently larger
436: line-widths in the $k=1$ component compared to the $k=0$ component.
437: This is consistent with already ongoing star formation activity in
438: these regions, where the warmer more central gas is subject to more
439: non-thermal motion, either of turbulent nature or due to infall,
440: rotation and/or outflow processes.
441:
442: One can try to use the observed lines also to get more quantitative
443: temperature and column density estimates. Following \citet{loren1984},
444: assuming optically thin emission and that a single temperature
445: characterizes the populations within a $k$-ladder and between
446: different $k$-ladders, the ratios of the intensities of different $k$
447: lines can be used to estimate this temperature. Specifically, for the
448: $J=6-5$ line and $k=0$ and $k=1$ components, we have
449:
450: $$T = 7.2/ln(I_0/I_1)$$
451:
452: with $I_k$ the corresponding observed peak intensities. For
453: IRDC\,18102$-$1800-1 and IRDC\,20081+2720-2, the relative line
454: intensities are consistent with optically thin emission and we
455: estimate the temperatures to be $16\pm 16$ and $16\pm 13$\,K,
456: respectively. For the other sources, while they are unlikely to be
457: optically thick, the intensities are inconsistent with optically thin
458: emission, so we do not estimate temperatures. It is interesting that
459: the estimates, where possible, although not very accurate, turn out to
460: be low and similar to estimates from NH$_3$ observations
461: \citep{sridharan2005}. We cannot use the two mainly detected CH$_3$OH
462: lines for temperature estimates because they are of different symmetry
463: (A- and E-CH$_3$OH) and can be considered for radiative transport
464: purposes as distinct molecules.
465:
466: However, we can calculate the column densities in the detected
467: CH$_3$OH and CH$_3$CN lines, and then derive the total molecular
468: column densities following \citet{rohlfs2006}. As partition functions
469: we used the values listed in the JPL catalog at $T=18.75$\,K
470: \citep{poynter1985}. Furthermore, we are interested in CH$_3$OH and
471: CH$_3$CN molecular abundances. To estimate these, we extracted the
472: 1.2\,mm peak fluxes from the images presented in
473: \citet{beuther2002a}\footnote{\citet{beuther2002a} presented peak
474: fluxes from 2D Gaussian fits to the images whereas here we use the
475: exact fluxes toward the peak positions. The former are on average a
476: bit lower.}. Following \citet{hildebrand1983} and
477: \citet{beuther2002erratum}, we then calculated the H$_2$ column
478: densities at $T=19$\,K (corresponding to the partition functions for
479: the molecular emission) assuming optically thin emission and a dust
480: opacity index $\beta=2$. The spatial resolution of the 1.2\,mm and
481: CH$_3$OH data is approximately the same, for the comparison with the
482: lower-resolution CH$_3$CN data, we smoothed the 1.2\,mm images also to
483: the corresponding spatial resolution of $23''$. The 1.2\,mm peak
484: fluxes and H$_2$ column densities (both at $11''$ resolution) are
485: listed in Table \ref{sourceparameters}. The different H$_2$ column
486: densities were then used to estimate the CH$_3$OH and CH$_3$CN
487: abundances. The calculated molecular column densities are estimated to
488: be correct within a factor 5, the accuracy of the estimated abundances
489: is correspondingly worse. Table \ref{sourceparameters2} lists the
490: derived molecular CH$_3$OH and CH$_3$CN column densities and
491: abundances which are on average relatively low.
492:
493: The average CH$_3$CN column density and abundance are
494: $3.1\times10^{12}$\,cm$^{-2}$ and $1.2\times 10^{-10}$, respectively.
495: This CH$_3$CN abundance compares well to the CH$_3$CN abundance
496: observed for example in the relatively cold and quiescent Orion
497: extended ridge whereas it is two orders of magnitude below the values
498: obtained for the Orion hot core \citep{vandishoek1998}. Furthermore,
499: it fits well to the chemical modeling of young high-mass star-forming
500: regions by \citet{nomura2004}: they find very low CH$_3$CN column
501: densities of the order $10^{11}-10^{12}$\,cm$^{-2}$ after only a few
502: hundred years which then progressively increase to values
503: $>10^{15}$\,cm$^{-2}$ after several $10^5$ years. In this picture,
504: even the sources with detected CH$_3$CN emission are as expected
505: extremely young massive star-forming regions.
506:
507: The average CH$_3$OH column density and abundance are
508: $8.2\times10^{13}$\,cm$^{-2}$ and $4.3\times 10^{-10}$, respectively.
509: This abundance is orders of magnitude below the values obtained toward
510: the typical regions in Orion (e.g., $8\times10^{-9}$ for the extended
511: ridge and $2\times10^{-7}$ for the hot core,
512: \citealt{vandishoek1998}). However, it is very close to the recently
513: reported CH$_3$OH abundances of $6\times 10^{-10}$ in the low-mass
514: starless cores L1498 and L1517B \citep{tafalla2006}. Although the
515: statistics are still small, this similarity between CH$_3$OH
516: abundances in low-mass starless cores and high-mass IRDCs indicates
517: that the initial CH$_3$OH molecular abundances at early evolutionary
518: stages do not differ significantly between low- and high-mass
519: star-forming regions.
520:
521: \section{Conclusions and summary}
522:
523: These single-dish mm spectral multi-line observations toward a sample
524: of 43 massive IRDCs revealed a range of interesting results. Since IRDC are
525: expected to harbor sources in different evolutionary stages from
526: genuine HMSCs to High-Mass cores with embedded low- to
527: intermediate-mass protostars potentially destined to become massive stars,
528: it is not surprising that the observational signatures do vary
529: significantly over the sample. Because our spectral line selection
530: covered tracers sensitive to different processes (outflows, dense gas,
531: lower density gas, hot core chemistry) we are able to constrain various
532: characteristics. In summary they are:
533:
534: \begin{itemize}
535:
536: \item Toward 40\% of the sources SiO(2--1) as a typical outflow/jet
537: tracer was detected. Assuming that SiO is exclusively produced by
538: gas phase reactions after sputtering Si-bearing species from dust
539: grains, this implies that at least 40\% of the IRDCs presented here
540: have ongoing star formation activity (non-detection of SiO does not
541: imply no star formation activity, hence the lower limit). The
542: observed width-to-zero-velocity of the SiO lines varies between 2.2
543: and 65\,km\,s$^{-1}$. While inclination effects are unlikely to
544: cause this velocity spread, we discuss other potenatial origins,
545: e.g., embedded objects with different masses and/or possible
546: evolutionary effects.
547:
548: \item While we usually detect many velocity components in the
549: low-density tracers $^{12}$CO(2--1) and $^{13}$CO(1--0), the
550: high-density tracer H$^{13}$CO$^+$(1--0) is also present in all
551: except one source, but in general tracing only one velocity
552: component. Therefore, it is well suited to derive velocities with
553: respect to the standard of rest and hence kinematic distances for
554: IRDCs. The observed FWHM line-width of H$^{13}$CO$^+$(1--0) is on
555: average a factor 1.5 larger than that of the previously observed
556: NH$_3$(1,1) line. Several possibilities for that behavior are
557: discussed: While the typical line-width--size relation is unlikely
558: the cause for this behavior, the data are consistent with
559: gravitational instabilities of the cores. However, this
560: interpretation is not unambiguous. Since H$^{13}$CO$^+$ has higher
561: critical densities, it traces the denser central cores with likely
562: larger internal motions. This motions could be caused by turbulence
563: or by active star formation processes like outflows/infall,
564: rotation.
565:
566: \item As expected CH$_3$CN is only rarely detected (toward $\sim$14\%
567: of the sources), whereas CH$_3$OH is observed toward a larger
568: fraction of the sample ($\sim$40\%). Both molecules are only
569: detected in their lower-energy transitions. While temperature
570: estimates are hardly feasible with these data, we are able to derive
571: the CH$_3$CN and CH$_3$OH abundances with average values of
572: $1.2\times 10^{-10}$ and $4.3\times 10^{-10}$, respectively. The low
573: CH$_3$CN column densities are consistent with chemical models where
574: the CH$_3$CN abundances are very low at the earliest evolutionary
575: stages and progressively increase with time. The derived CH$_3$OH
576: abundances are close to recently reported values for low-mass
577: starless cores, indicating that the initial CH$_3$OH abundances at
578: the onset of star formation processes do not vary significantly
579: from low- to high-mass cores.
580:
581: \end{itemize}
582:
583: \acknowledgments{We like to thank Sven Thorwirth for continuous
584: spectroscopic help. H.B.~acknowledges financial support by the
585: Emmy-Noether-Program of the Deutsche Forschungsgemeinschaft (DFG,
586: grant BE2578).}
587:
588: %\bibliography{/home/beuther/tex/bibliography}
589: %\bibliography{/Users/henrikbeuther/paper/bibliography}
590: %\bibliographystyle{aa} % this does the style, aa.bst necessary
591: \input{ms.bbl}
592:
593: \begin{figure*}[h]
594: \includegraphics[angle=-90,width=5.4cm]{fig1a.eps}
595: \includegraphics[angle=-90,width=5.4cm]{fig1b.eps}
596: \includegraphics[angle=-90,width=5.4cm]{fig1c.eps}\\
597: \includegraphics[angle=-90,width=5.4cm]{fig1d.eps}
598: \includegraphics[angle=-90,width=5.4cm]{fig1e.eps}
599: \includegraphics[angle=-90,width=5.4cm]{fig1f.eps}\\
600: \includegraphics[angle=-90,width=5.4cm]{fig1g.eps}
601: \includegraphics[angle=-90,width=5.4cm]{fig1h.eps}
602: \includegraphics[angle=-90,width=5.4cm]{fig1i.eps}\\
603: \includegraphics[angle=-90,width=5.4cm]{fig1j.eps}
604: \includegraphics[angle=-90,width=5.4cm]{fig1k.eps}
605: \includegraphics[angle=-90,width=5.4cm]{fig1l.eps}\\
606: \includegraphics[angle=-90,width=5.4cm]{fig1m.eps}
607: \includegraphics[angle=-90,width=5.4cm]{fig1n.eps}
608: \includegraphics[angle=-90,width=5.4cm]{fig1o.eps}\\
609: \includegraphics[angle=-90,width=5.4cm]{fig1p.eps}
610: \includegraphics[angle=-90,width=5.4cm]{fig1q.eps}
611: \includegraphics[angle=-90,width=5.4cm]{fig1r.eps}
612: \caption{$^{12}$CO(2--1) (grey), $^{13}$CO(1--0) (black line) and
613: H$^{13}$CO$^+$(1--0) (red line) spectra toward the whole sample of
614: IRDCs. The H$^{13}$CO$^+$(1--0) are scaled up by the factor labeled
615: in each panel for presentation purposes.}
616: \label{sample1}
617: \end{figure*}
618:
619: \begin{figure*}[h]
620: \includegraphics[angle=-90,width=5.4cm]{fig1s.eps}
621: \includegraphics[angle=-90,width=5.4cm]{fig1t.eps}
622: \includegraphics[angle=-90,width=5.4cm]{fig1u.eps}\\
623: \includegraphics[angle=-90,width=5.4cm]{fig1v.eps}
624: \includegraphics[angle=-90,width=5.4cm]{fig1w.eps}
625: \includegraphics[angle=-90,width=5.4cm]{fig1x.eps}\\
626: \includegraphics[angle=-90,width=5.4cm]{fig1y.eps}
627: \includegraphics[angle=-90,width=5.4cm]{fig1z.eps}
628: \includegraphics[angle=-90,width=5.4cm]{fig1aa.eps}\\
629: \includegraphics[angle=-90,width=5.4cm]{fig1ab.eps}
630: \includegraphics[angle=-90,width=5.4cm]{fig1ac.eps}
631: \includegraphics[angle=-90,width=5.4cm]{fig1ad.eps}\\
632: \includegraphics[angle=-90,width=5.4cm]{fig1ae.eps}
633: \includegraphics[angle=-90,width=5.4cm]{fig1af.eps}
634: \includegraphics[angle=-90,width=5.4cm]{fig1ag.eps}\\
635: \includegraphics[angle=-90,width=5.4cm]{fig1ah.eps}
636: \includegraphics[angle=-90,width=5.4cm]{fig1ai.eps}
637: \includegraphics[angle=-90,width=5.4cm]{fig1aj.eps}
638: \caption{Figure \ref{sample1} continued.}
639: \label{sample2}
640: \end{figure*}
641:
642: \begin{figure*}[h]
643: \includegraphics[angle=-90,width=5.4cm]{fig1ak.eps}
644: \includegraphics[angle=-90,width=5.4cm]{fig1al.eps}
645: \includegraphics[angle=-90,width=5.4cm]{fig1am.eps}\\
646: \includegraphics[angle=-90,width=5.4cm]{fig1an.eps}
647: \includegraphics[angle=-90,width=5.4cm]{fig1ao.eps}
648: \includegraphics[angle=-90,width=5.4cm]{fig1ap.eps}\\
649: \includegraphics[angle=-90,width=5.4cm]{fig1aq.eps}
650: \caption{Figures \ref{sample1} \& \ref{sample2} continued.}
651: \label{sample3}
652: \end{figure*}
653:
654: \begin{figure}[h]
655: \includegraphics[angle=-90,width=8.6cm]{fig2.eps}
656: \caption{Comparison of line-widths between H$^{13}$CO$^+$(1--0) (this
657: work) and NH$_3$(1,1) \citep{sridharan2005}.}
658: \label{dv}
659: \end{figure}
660:
661:
662: \begin{figure*}[h]
663: \includegraphics[angle=-90,width=5.4cm]{fig3a.eps}
664: \includegraphics[angle=-90,width=5.4cm]{fig3b.eps}
665: \includegraphics[angle=-90,width=5.4cm]{fig3c.eps}\\
666: \includegraphics[angle=-90,width=5.4cm]{fig3d.eps}
667: \includegraphics[angle=-90,width=5.4cm]{fig3e.eps}
668: \includegraphics[angle=-90,width=5.4cm]{fig3f.eps}\\
669: \includegraphics[angle=-90,width=5.4cm]{fig3g.eps}
670: \includegraphics[angle=-90,width=5.4cm]{fig3h.eps}
671: \includegraphics[angle=-90,width=5.4cm]{fig3i.eps}\\
672: \includegraphics[angle=-90,width=5.4cm]{fig3j.eps}
673: \includegraphics[angle=-90,width=5.4cm]{fig3k.eps}
674: \includegraphics[angle=-90,width=5.4cm]{fig3l.eps}\\
675: \includegraphics[angle=-90,width=5.4cm]{fig3m.eps}
676: \includegraphics[angle=-90,width=5.4cm]{fig3n.eps}
677: \includegraphics[angle=-90,width=5.4cm]{fig3o.eps}\\
678: \includegraphics[angle=-90,width=5.4cm]{fig3p.eps}
679: \includegraphics[angle=-90,width=5.4cm]{fig3q.eps}
680: \includegraphics[angle=-90,width=5.4cm]{fig3r.eps}
681: \label{siospectra}
682: \caption{SiO(2--1) detections.}
683: \end{figure*}
684:
685: \begin{figure*}[h]
686: \includegraphics[angle=-90,width=5.4cm]{fig4a.eps}
687: \includegraphics[angle=-90,width=5.4cm]{fig4b.eps}
688: \includegraphics[angle=-90,width=5.4cm]{fig4c.eps}\\
689: \includegraphics[angle=-90,width=5.4cm]{fig4d.eps}
690: \includegraphics[angle=-90,width=5.4cm]{fig4e.eps}
691: \includegraphics[angle=-90,width=5.4cm]{fig4f.eps}
692: \caption{CH$_3$CN$(6_k-5_k)$ detections. The rest frequencies of the
693: $k=0, 1, 2$ components are at 110383.522, 110381.404 and
694: 110375.052\,MHz, respectively. Because of the small intrinsic
695: line-widths, the $k=0,1$ components are spectrally resolved which is
696: usually not the case for more evolved High-Mass Protostellar Objects
697: or Hot Molecular Cores.}
698: \label{ch3cn}
699: \end{figure*}
700:
701: \begin{figure*}[h]
702: \includegraphics[angle=-90,width=5.4cm]{fig5a.eps}
703: \includegraphics[angle=-90,width=5.4cm]{fig5b.eps}
704: \includegraphics[angle=-90,width=5.4cm]{fig5c.eps}\\
705: \includegraphics[angle=-90,width=5.4cm]{fig5d.eps}
706: \includegraphics[angle=-90,width=5.4cm]{fig5e.eps}
707: \includegraphics[angle=-90,width=5.4cm]{fig5f.eps}\\
708: \includegraphics[angle=-90,width=5.4cm]{fig5g.eps}
709: \includegraphics[angle=-90,width=5.4cm]{fig5h.eps}
710: \includegraphics[angle=-90,width=5.4cm]{fig5i.eps}\\
711: \includegraphics[angle=-90,width=5.4cm]{fig5j.eps}
712: \includegraphics[angle=-90,width=5.4cm]{fig5k.eps}
713: \includegraphics[angle=-90,width=5.4cm]{fig5l.eps}\\
714: \includegraphics[angle=-90,width=5.4cm]{fig5m.eps}
715: \includegraphics[angle=-90,width=5.4cm]{fig5n.eps}
716: \includegraphics[angle=-90,width=5.4cm]{fig5o.eps}\\
717: \includegraphics[angle=-90,width=5.4cm]{fig5p.eps}
718: \includegraphics[angle=-90,width=5.4cm]{fig5q.eps}
719: \caption{CH$_3$O(5--4) detections. The rest frequencies of the four
720: detected lines are: CH$_3$OH$(5_1-4_1)$E at 241879.073\,MHz,
721: CH$_3$OH$(5_0-4_0)$A+ at 241791.431\,MHz, CH$_3$OH$(5_{-1}-4_{-1})$E
722: at 241767.224\,MHz, CH$_3$OH$(5_0-4_0)$E at 241700.219\,MHz. The few
723: single-channel spikes are artifacts caused by the correlator.}
724: \label{ch3oh}
725: \end{figure*}
726:
727: \newpage
728: \input{table.tex}
729: \input{table2.tex}
730:
731:
732: \end{document}
733:
734: