1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %% FCRAO IC5146 DIFFERENTIAL DEPLETION PAPER %%
3: %% %%
4: %% REVISIONS 7/10/00 BY CJ LADA %%
5: %% ADDED D. CIARDI's CORRECTIONS 8/1/00 EAB %%
6: %% REVISIONS 8/2/00 EAB %%
7: %% %%
8: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
9: \documentclass[preprint]{aastex}
10: %\documentclass[preprint2]{aastex}
11:
12:
13: \def\Hthreep{H$_3^+$} % H3+
14: \def\NtwoHp{N$_2$H$^+$} % N2H+
15: \def\CeiO{C$^{18}$O} % N2H+
16: \def\CsvO{C$^{17}$O} % N2H+
17: \def\CtfS{C$^{34}$S} % N2H+
18: \def\Ntwo{N$_2$} % N2
19: \def\Htwo{H$_2$} % N2
20: \def\av{$A_V$} % N2
21: \newcommand{\nhtwo}{\mbox{n$_{H_{2}}$}}
22: \newcommand{\cc}{\mbox{cm$^{-3}$}}
23:
24:
25:
26: \begin{document}
27:
28: %\title{Depletion of CO \& CS and the Production of
29: %N$_2$H+ in the Dark Cloud Associated with IC 5146}
30: \title{Molecular Excitation and Differential Gas-Phase Depletions in
31: the IC 5146 Dark Cloud}
32:
33: \author{Edwin A. Bergin}
34: \affil{Harvard-Smithsonian Center for Astrophysics, 60 Garden Street,
35: Cambridge, MA 02138} \email{ebergin@cfa.harvard.edu}
36:
37: \author{David R. Ciardi}
38: \affil{Department of Astronomy, University of Florida, Gainesville, FL
39: 32611} \email{ciardi@astro.ufl.edu}
40:
41: \author{Charles J. Lada}
42: \affil{Harvard-Smithsonian Center for Astrophysics, 60 Garden Street,
43: Cambridge, MA 02138} \email{clada@cfa.harvard.edu}
44:
45: \author{Joao Alves}
46: \affil{European Southern Observatory, Karl-Schwarzschild-Strasse 2, 85748
47: Garching, Germany}
48: \email{jalves@eso.org}
49:
50: \author{Elizabeth A. Lada}
51: \affil{Department of Astronomy, University of Florida, Gainesville, FL
52: 32611} \email{lada@astro.ufl.edu}
53:
54: \slugcomment{Accepted by The Astrophysical Journal}
55: \begin{abstract}
56: We present a combined near-infrared and molecular-line study of
57: 25$' \times 8'$ area in the Northern streamer of the IC 5146 cloud.
58: Using the technique pioneered
59: by \citet{lada94}, we construct a Gaussian smoothed map of the
60: infrared extinction with the same resolution as the molecular line observations
61: in order to examine correlations of integrated intensities
62: and molecular abundances with extinction for \CsvO ,
63: \CtfS, and \NtwoHp .
64: We find that over a visual extinction range of 0 to 40 magnitudes,
65: there is good evidence
66: for the presence of differential gas-phase depletions in
67: the densest portions of IC 5146. Both CO and CS exhibit a statistically
68: significant (factor of $\sim 3$) abundance reduction
69: near \av\ $\sim 12$ magnitudes while, in direct contrast,
70: at the highest extinctions, \av $> 10$ magnitudes,
71: \NtwoHp appears relatively undepleted.
72: Moreover, for \av\ $< 4$ magnitudes there exists little or no \NtwoHp .
73: This pattern of depletions is
74: consistent with the predictions of chemical theory.
75: Through the use of a time and depth dependent
76: chemical model we show that the near-uniform or rising \NtwoHp\ abundance with
77: extinction is a direct result of a reduction in
78: its destruction rate at high extinction due to the
79: predicted and observed depletion of CO molecules.
80: The observed abundance threshold for \NtwoHp , $A_V^{th} \sim 4$ mag, is examined
81: in the context of this same model and we demonstrate how this technique can
82: be used to test the predictions of depth-dependent chemical models.
83: Finally, we find that cloud density gradients can have a significant effect
84: on the excitation and detectability of high dipole moment molecules,
85: which are typically far from local thermodynamic equilibrium. Density
86: gradients also cause chemical changes as reaction rates and
87: depletion timescales are density dependent. Accounting
88: for such density/excitation gradients is crucial to a correct determination
89: and proper interpretation of molecular abundances.
90:
91: \end{abstract}
92:
93:
94: \keywords{dust, extinction --- ISM: abundances --- ISM: clouds --- ISM:
95: individual (IC 5146) --- ISM: molecules --- stars: formation}
96:
97: \section{Introduction}
98:
99: The determination of masses, densities, energetics, and chemistry for cold
100: molecular cores is of fundamental importance for our understanding of how
101: and why stars form. Unfortunately, the dominant molecule in interstellar
102: molecular clouds, H$_2$, is generally unobservable at the low temperatures
103: associated with star forming cores. Consequently, the most fundamental
104: properties of molecular clouds (and the embedded star-forming cores) have
105: been determined almost exclusively from observations of rarer trace
106: molecules (and their isotopes) such as CO, CS, and NH$_3$. However, these
107: secondary molecular tracers have been calibrated only in a handful of
108: sources, and because of variations in chemical abundances, chemical
109: evolution, and excitation conditions, application to other sources is
110: inherently uncertain, making accurate determination of important physical
111: properties (e.g., sizes, masses, and energetics) of molecular clouds not
112: always easy or possible.
113:
114: Because of the apparent constancy of the gas-to-dust ratio in molecular
115: clouds \citep{js74, bsd78}, the most direct method to trace the hydrogen
116: content of a molecular cloud is to determine the distribution of dust
117: throughout the cloud. In principle, the distribution of the dust can be
118: derived from far-infrared and millimeter wavelength dust emission, but the
119: grain opacity at high densities is poorly constrained, limiting the
120: accuracy of the dust (and hence, H$_2$) column density determination
121: \citep{kram98}. Recently, \citet{lada94} developed a powerful technique for
122: mapping the large scale distribution of dust using multi--wavelength
123: near-infrared imaging. By measuring the near-infrared color excess of
124: stars behind a cloud the line of sight dust extinction (and hence, total
125: column density) can be directly determined.
126: The technique enables measurements of the dust
127: distribution over a significant of range of angular scales and extinction
128: \citep{crd98, lal99}. Because the near-infrared extinction law does not
129: vary significantly with grain growth in cold cores \citep{math90}, high
130: extinction observed towards cloud cores can be anchored to the
131: well-calibrated H$_2$-to-dust ratio at lower extinction.
132:
133: When the near-infrared excess data is combined with molecular gas
134: emission-line observations, direct determination of the molecular
135: abundances relative to H$_2$ can be obtained; thus, allowing a more
136: accurate determination of the physical and chemical properties of
137: molecular clouds and their embedded cores \citep{all99}. Indeed,
138: $^{12}$C$^{18}$O, a molecule often assumed to trace accurately the distribution
139: of material in cold cores, has been found to show evidence for a reduction
140: in abundance at high extinction which has been attributed to the freezing
141: of gas-phase CO onto the surfaces of cold dust grains (i.e., depletion)
142: in the cores of IC 5146 \citep{kram99} and L 977
143: \citep{all99}.
144:
145: \citet{kram99} found compelling evidence for depletion of CO on to dust grains in a
146: cold core located in the IC 5146 dark cloud. The finding was based
147: primarily upon the relationship between the integrated intensity of
148: C$^{18}$O and visual extinction derived from near-infrared color excess
149: measurements. The C$^{18}$O integrated intensity was found to be
150: well-correlated with $A_V$ for $A_V \lesssim 10$ mag. For $A_V \gtrsim
151: 10$ mag, the relationship flattens out, and the C$^{18}$O integrated
152: intensity is nearly constant out to a visual extinction of $A_V \sim 30$
153: mag. C$^{17}$O was used to check for opacity effects, and the C$^{18}$O
154: was found to be optically thin throughout the region. \citet{kram99}
155: concluded that the most likely explanation was the depletion of CO on to
156: the surface of dust grains. Thus, masses, and densities of cold dark cores
157: in IC5146 derived from $^{12}$C$^{18}$O may underestimate the true mass and
158: densities by as much as 30\%.
159:
160: However, the study by \citet{kram99} covered a very small region ($3\farcm
161: 3 \times 3\farcm 3$) and was limited to a single core in the IC 5146 dark cloud.
162: In this paper, we present a more detailed study of the molecular gas
163: abundances over a much larger region in the IC 5146 dark cloud ($25\arcmin
164: \times 8\arcmin$) at moderate resolution (50\arcsec). Our primary goal is
165: to perform, for a variety of molecular species, a direct comparison of the
166: molecular emission and the line of sight dust column density in the dark
167: cloud associated with the young cluster IC 5146. By using the line of
168: sight dust extinction to trace the H$_2$ column density and comparing that
169: to the measured molecular emission, a better understanding of the chemical
170: abundances and evolution within dark molecular cores can be achieved. As
171: rarer trace-molecules are routinely used to discern the distribution of
172: molecular material in star forming clouds, understanding the abundance and
173: chemistry of these molecules is vital to our understanding of star
174: formation.
175:
176: We have obtained new radio observations of the rotational transitions of
177: $^{12}$C$^{18}$O, $^{12}$C$^{17}$O, $^{12}$C$^{32}$S, $^{12}$C$^{34}$S,
178: $^{13}$C$^{32}$S, and N$_2$H$^+$ towards the Northern Streamer in the dark
179: cloud (B168) associated with IC 5146. Using the near-infrared extinction data of
180: the same region published by \citet{lal99}, we present a detailed analysis
181: of the correlation between the molecular gas and the dust column density
182: in the IC 5146 dark cloud. In \S 2 of this paper,
183: we describe the near-infrared data utilized in this study.
184: In \S 3, we describe the
185: acquisition and reduction of the radio molecular line observations.
186: In \S 4 we examine the excitation and present molecular abundance profiles against
187: extinction. In \S 5 we discuss the implications of these observations and
188: in \S 6, we detail our conclusions.
189:
190: \section{Observations and Data Reduction}
191:
192: \subsection{Near-Infrared Extinction}
193:
194: The near-infrared data used in this paper are taken from \citet{lal99},
195: where a detailed description of the data acquisition, reduction, and
196: source extractions may be found. The area surveyed is approximately
197: $27\arcmin \times 8\arcmin$, and the photometry is complete (at the
198: 10$\sigma$ level) to a depth of $H \approx 17$ mag and $K \approx 16$ mag.
199: The infrared color excess for each star was determined via $E(H-K) =
200: (H-K)_{obs} - (H-K)_{intr}$, where $(H-K)_{intr} \equiv
201: \langle(H-K)\rangle_{control}=0.13\pm0.01$ magnitudes, the mean color
202: of field stars observed in a nearby, unextincted control field
203: \citep{lada94, lal99}. The
204: color excess for each star was converted to an extinction, using the
205: reddening law of \citet{rl85}: $A_K = 1.78E(H-K)$ \& $A_V = 15.9E(H-K)$.
206:
207: The extinction derived in this manner is directly proportional to the {\em
208: true} dust column density along that line-of-sight, under the assumption
209: that the H$_2$-to-dust ratio remains constant. The conversion of the
210: measured near-infrared color excess to the conventional visual extinction
211: may not truly represent the actual visual reddening of a background star,
212: because grain growth in cold clouds alters the extinction law at
213: $\lambda~\ll~1$~\micron. However, at near-infrared wavelengths, the reddening law
214: is not known to vary significantly with grain growth
215: \citep[e.g.,][]{math90}, and the derived near-infrared extinctions remain
216: proportional to the line-of-sight column density.
217: Moreover, the conversion of the measured infrared extinction
218: to the conventional visual extinction maintains the proportionality. Use
219: of $A_K$ instead of $A_V$ would avoid this confusion, but to facilitate
220: comparison to other work \citep[e.g.,][]{lada94, crd98, alves98, kram99,
221: lal99}, we report visual extinction throughout this paper.
222:
223: To enable a direct comparison between the near-infrared observations and
224: the radio molecular-line observations (obtained with the FCRAO
225: telescope and described in the next section), the mean near-infrared
226: extinction along each line-of-sight associated with a molecular-line observation
227: is calculated by convolving the individual infrared excess
228: measurements with a two-dimensional Gaussian filter. The gaussian filter
229: has a FWHM of 50\arcsec\ to match the approximate beamsize of the FCRAO
230: telescope, and is truncated at $r=3\sigma$. The uncertainty on the
231: mean extinction is estimated from the Gaussian weighted rms dispersion of
232: extinction measurements falling within each ``beam.''
233: The gaussian convolved map of visual extinction is
234: shown in Figure~\ref{fig_avgrid}a. This map is quite similar to the
235: larger extinction map of IC 5146 presented by \citet{lal99}.
236:
237: \subsection{Molecular-Line Observations}
238:
239: Over two separate observing periods (1999 March 23--25 and 2000 Jan 31 --
240: Feb 01), the 14~m telescope of the Five College Radio Astronomy
241: Observatory (FCRAO) in New Salem, MA was used to observe a $25\arcmin
242: \times 8\arcmin$ region covering the Northern Streamer of the IC 5146
243: dark cloud. The rotational transitions of the following molecules were
244: observed: $^{12}$C$^{18}$O (109.782182 GHz, $J =1\rightarrow$ 0; hereafter,
245: C$^{18}$O), the hyperfine triplet of $^{12}$C$^{17}$O (112.359277 GHz, $J
246: =1\rightarrow 0$; hereafter, C$^{17}$O -- note that the two reddest
247: components are blended and unresolved at our spectral resolution of 78
248: kHz; see below), $^{12}$C$^{32}$S (97.981011 GHz, $J =2\rightarrow$1;
249: hereafter, CS), $^{12}$C$^{34}$S (96.412962 GHz, $J = 2\rightarrow$1;
250: hereafter, C$^{34}$S), $^{13}$C$^{32}$S (92.4979 GHz $J = 2\rightarrow$1;
251: hereafter, $^{13}$CS), and N$_2$H$^+$ (93.173777 GHz, $J =1\rightarrow 0$). The
252: N$_2$H$^+$ transition consists of seven hyperfine components spanning 4.7
253: MHz; however, at our spectral resolution (78 kHz, see below) only the
254: three main groups of lines centered at rest frequencies of 93.1719 GHz (a
255: blend of three lines), 93.1738 GHz (a blend of three lines), and 93.1763
256: (a single line) are resolved \citep[e.g.,][]{cmt95}.
257:
258: The observations reported in this paper were obtained with the
259: newly commissioned SEQUOIA 16-element focal plane array receiver.
260: The sixteen elements of the SEQUOIA array are aligned in a square
261: $4 \times 4$ pattern; each element is separated from its neighbor by
262: 88\arcsec\ in both the X and Y directions. A full beam sampled map
263: requires 4 pointings of the array on the sky and is referred to as a
264: ``footprint''. All maps were centered with respect to $\alpha(J2000) =$
265: 21$^h$47$^m$27$^s$ and $\delta(J2000) =$ 47$^{\circ }31'00''$.
266:
267: The FAAS autocorrelator spectrometer was utilized in the 512-channel, 40
268: MHz bandwidth mode, yielding a channel spacing of 78 kHz ($\Delta v \sim
269: 0.21$ km s$^{-1}$ at 112 GHz and $\Delta v \sim 0.25$ km s$^{-1}$ at 93
270: GHz) The 1999 data were taken in position switching mode, while the
271: 2000 data were taken in frequency switching mode. For the frequency switched data,
272: the signal frequency was shifted by 4 MHz. Typical system temperatures
273: ranged from $140-300$ K. The integrations times for each molecule varied
274: (see Table 1), but typical $T_A^*$ rms noise temperatures per channel were
275: $\lesssim 0.02$ K. Internal calibration was done via a chopper wheel which
276: allows switching between the sky and an ambient temperature load. Focusing
277: and pointing were checked by observing periodically the SiO masers T Ceph
278: (1999) and R Cas (2000). Typical, rms pointing uncertainties were
279: 5\arcsec.
280:
281: The Northern Streamer was mapped at full beam sampling in C$^{18}$O, CS, and
282: N$_2$H$^+$. The $25\arcmin \times 8\arcmin$ map was rotated by 14$^{\circ}$
283: North of East with respect to equatorial coordinates to follow the long
284: axis of the cloud and the near-infrared survey data
285: \citep[e.g.,][]{lada94, lal99}. The full map required four ``footprints''
286: for a total of 256 grid points. The streamer contains two dense cores
287: which are readily detectable in a 50$''$ beam and were mapped at full beam
288: sampling in C$^{34}$S. The positions of the
289: grid points in these two individual ``footprints'' are no different than the
290: positions found in the the full map. A single additional ``footprint''
291: (64 grid points) was observed towards the eastern core in C$^{18}$O. This
292: ``footprint'' was offset from the regular map grid by $-0\farcm 369$
293: ($\sim 0.5$ beamsize) in both right ascension and declination. The
294: purpose of this ``footprint'' was to increase the sampling of the
295: C$^{18}$O in the eastern core. Additional single array placements (16
296: grid points) on the eastern core were observed in C$^{18}$O, C$^{17}$O,
297: and $^{13}$CS. These single array observations were used to investigate
298: the C$^{18}$O and CS optical depth of the core (see \S 3.1 and \S 3.2 below).
299:
300: A summary of the map grid points is shown in Figure~\ref{fig_avgrid}, and details of the
301: observing parameters for each molecular observation can be found in Table
302: 1. Several points on the southeastern edge of
303: the radio grid fall outside of the infrared survey region
304: and these data points are not included in this study.
305:
306: For the remainder of the paper, the ``Eastern Core'' refers to the
307: eastern-most ``footprint'' (i.e., the eastern-most $8 \times 8$ grid),
308: the ``Western Core'' refers to
309: the western-most ``footprint'' (i.e., the western-most $8 \times 8$ grid),
310: and the ``Mid-Streamer'' refers to the two ``footprints'' ($16 \times 8$
311: grid) lying between the ``Eastern'' and ``Western'' Cores (see Fig. 1).
312:
313:
314: All data were reduced using SPA and CLASS spectral line reduction packages
315: in conjunction with custom-made analysis routines. A second order
316: polynomial baseline was subtracted from the position-switch spectra; the
317: frequency-switched spectra
318: were folded and a second-order polynomial baseline was subtracted. Main
319: beam efficiencies of $\eta_{MB} \approx 0.50-0.55$ were adopted from the
320: SEQUOIA documentation. Final $T_R$ rms noise temperatures were
321: generally $\lesssim 0.04$ K.
322: All data presented in the paper are corrected using the main beam efficiency and are
323: on the $T_R$ scale.
324: %The spectral lines (including the individual
325: %components of the C$^{17}$O and N$_2$H+ hyperfine transitions) were all
326: %fit with single component Gaussian profiles. Integrated intensity maps
327: %for C$^{18}$O, CS, C$^{34}$S, and N$_2$H+ are presented in Figures 2a-d.
328:
329:
330: \section{Results: Comparison of the Dust and Gas}
331:
332: Figures 2a and b present maps of the integrated
333: emission of \CeiO\ and \NtwoHp\ in
334: IC 5146. The map of \CeiO\ J $= 1 \rightarrow 0$ integrated intensity appears
335: broadly similar in distribution to the visual extinction map shown in
336: Figure~\ref{fig_avgrid}. Both extinction and \CeiO\ trace column density
337: enhancements, or cores, at the eastern and western edges of the mapped region,
338: although the \CeiO\ emission peaks are slightly offset from those of \av .
339: The integrated emission of \NtwoHp\ shows strong emission
340: toward the eastern core, and only weak and somewhat clumpy distribution throughout
341: the rest of the streamer. Curiously, the \NtwoHp\ intensity appears to correlate
342: reasonably well with the regions of highest extinction. This is more clearly demonstrated
343: in Figure~\ref{fig_wc} where the integrated intensity distribution of \NtwoHp\
344: and \CeiO\ are compared with visual extinction in the western core.
345: Here the strongest \NtwoHp\
346: emission appears directly on the peak of extinction, while the \CeiO\ emission
347: maxima are offset by nearly 1$'$. This is unlikely to be the result of differences
348: in excitation requirements between these two transitions. The upper state energy of
349: each transition is $\sim 5$ K and, while \NtwoHp\ has a large critical density
350: ($\sim 10^5$ \cc ) and preferentially samples only the highest densities, the
351: \CeiO\ emission should also trace the same density regimes. Instead these
352: morphological differences are likely due to differences in the chemistry as
353: will be discussed later in this paper.
354:
355: Figure~\ref{fig_csdist} presents the total integrated emission for the
356: J $= 2 \rightarrow 1$ transitions of CS and \CtfS .
357: The CS emission morphology is similar to that of \CeiO\ as
358: CS strongly peaks on both cores and shows weak, but detectable, emission from
359: the gas between the cores. Because of its weaker intensity, the J $= 2
360: \rightarrow 1$ transition of \CtfS\ was only observed in the two cores.
361: Within these smaller regions the
362: \CtfS\ emission is roughly similar to that of its more abundant isotope, CS.
363:
364: In Figure~\ref{fig_spec} selected spectra at two different positions are presented.
365: The two chosen positions are ($\Delta \alpha = +0\farcm 2$, $\Delta \delta = +0\farcm 7$) and
366: ($\Delta \alpha = -0\farcm 9$, $\Delta \delta = +2\farcm 5$) are both associated with the Eastern
367: core and sample extinctions that differ by a factor of two. Here we see that the
368: intensity of C$^{17}$O is significantly lower at the highest extinction (for \av\ = 36 mag the
369: detection is 16$\sigma$ on the integrated intensity, while at \av\ = 19 mag the
370: detection is 4$\sigma$). In contrast the emission of C$^{34}$S and \NtwoHp\ appears to
371: be stronger at the position with higher extinction.
372: These differences, shown at only two positions, will be examined in
373: greater detail with higher sampling in the following sections.
374:
375: \subsection{C$^{18}$O}
376:
377: The direct comparison between the C$^{18}$O integrated intensity and
378: visual extinction at each point in the mapping grid
379: is presented in Figure~\ref{fig_c18oint}. The behavior of these two
380: tracers is very similar to what was found previously for the IC 5146 region \citep{lada94,
381: kram99}. The integrated intensity and the extinction appear
382: well-correlated for $A_V
383: \lesssim 10-15$ mag. For $A_V \gtrsim 10-15$ mag, the relationship
384: shows a high degree of scatter, and appears to flatten out.
385:
386: A bivariate linear fit was performed for the data located at $A_V < 10$ mag, with the
387: following result:
388: \begin{equation}
389: I(C^{18}O) = \left(-0.4 \pm 0.1\ K\ km\ s^{-1}\right) + \left(0.20 \pm
390: 0.01\ \frac{K\ km\ s^{-1}}{mag}\right)\cdot [A_V].
391: \end{equation}
392:
393: \noindent The above fit is in excellent agreement with the fit presented by
394: \citet{lal99}, for a similar region but at lower angular resolution
395: (102\arcsec) than the observations presented here.
396: The fit is also in
397: agreement with the C$^{18}$O-$A_V$ relationship found for the smaller
398: region (and finer beamsize -- 30\arcsec) studied by \citep{kram99}.
399: However, if the entire extinction range is used in the fit, the resulting slope is
400: 15\% shallower, consistent with earlier findings for both IC5146
401: \citep{lada94} and also in L 977 \citep{all99}.
402:
403: The two most likely effects which could explain the change in slope
404: at high extinction
405: are: (1) high C$^{18}$O optical depth for $A_V \gtrsim 10$ mag, and (2)
406: depletion of CO onto the surface of dust grains. A third possibility is
407: a decrease of the CO excitation temperature to $T_{ex} < 5$ K for $A_V
408: \gtrsim 10$ mag. However, \citet{all99}, for a similar result in L~977,
409: make a convincing argument that this is not a likely possibility. They argue that
410: heating mechanisms (cosmic ray rates, ambipolar diffusion, and natural
411: radioactivity) are sufficient to prevent the gas in cloud cores to cool
412: below 5 K. Although we cannot completely rule out this possibility, we
413: regard the cooling of the gas to such extreme temperature as unlikely, and
414: proceed with a discussion of the optical depth and depletion.
415:
416: The paired C$^{18}$O--C$^{17}$O single pointing observations provide
417: information regarding the optical depth of the C$^{18}$O in the densest
418: portions of the eastern core (see Table 1). C$^{18}$O and
419: C$^{17}$O should be collisionally excited under the same physical
420: conditions deep inside cold cores where the molecules are well-shielded
421: from external ultraviolet radiation \citep{lfd98, vdb88}. Additionally,
422: the ratio [$^{18}$O]/[$^{17}$O]$=3.65 \pm 0.15$ appears to be relatively
423: constant throughout the interstellar medium \citep{penz81}. Thus, the
424: only difference in the emission between C$^{18}$O and C$^{17}$O should
425: arise from the oxygen relative abundance; i.e., completely optically thin
426: emission of C$^{18}$O and C$^{17}$O should have an integrated intensity
427: ratio of 3.65.
428:
429: A comparison of the integrated intensities of C$^{17}$O vs C$^{18}$O is
430: shown in Figure~\ref{fig_c17o} (top).
431: The solid line represents the expected ratio of
432: 3.65 for optically thin emission and does {\em not} represent a best fit
433: to the data. The dashed lines represent the $1\sigma = \pm0.15$
434: uncertainty of the [$^{18}$O]/[$^{17}$O] ratio. The excellent agreement
435: of the data distribution with the predicted ratio indicates that the
436: C$^{18}$O emission is indeed optically thin.
437:
438: To test directly for optical depth effects as a function of total column
439: density, the C$^{18}$O/C$^{17}$O ratio has been plotted as a function of
440: extinction (bottom, Fig.~\ref{fig_c17o}). The extinction provides an independent
441: assessment of the total column density along each line of sight. If the
442: C$^{18}$O was optically thick at high column density (high extinction), the
443: C$^{18}$O/C$^{17}$O ratio should systematically decrease as a function of
444: increasing extinction. The ratio is remarkably constant over a range of
445: 40 magnitudes of visual extinction. Figure~\ref{fig_c17o} implies that (1) the
446: [$^{18}$O]/[$^{17}$O] ratio in IC 5146 is very near the previously
447: determined interstellar value of $3.65 \pm 0.15$, and (2) the C$^{18}$O
448: emission is {\em not} highly optically thick. However, due to the errors on the
449: ratios we cannot rule out the C$^{18}$O emission being thin, but with moderate
450: optical depth ($\tau \lesssim$ 0.5 ).
451:
452: The ratio of the three hyperfine components of C$^{17}$O can be used to
453: determine whether or not C$^{17}$O is thin.
454: The relative intensities of the
455: three components (central:blue:red) are 4:3:2 for optically thin emission
456: \citep{lfd98}. At the
457: spectral resolution of our data (78 kHz), the central and red components
458: are blended; thus, the measured relative intensity of the two components
459: (central-red blend:blue) should be 2:1, for optically thin emission
460: and indeed the measured median
461: relative intensity is $2 \pm 0.5$. If this is the case then the difference
462: in the two C$^{17}$O spectra presented in Figure~\ref{fig_spec}, with the highest extinction
463: having lower emission, is suggestive of a drop in abundance.
464:
465:
466: \subsection{CS}
467:
468: The depletion of sulfur bearing molecules (e.g., CS and SO) is predicted to be
469: a sharp function of the density, and these species should be robust tracers
470: of gas phase molecular depletion on to the surface of dust grains
471: in cloud cores \citep{bl97}. As there
472: was already evidence to suggest that CO may be depleted in the cores of IC
473: 5146 (\S 3.1), CS should also show evidence for gas-phase
474: depletion.
475:
476: To explore this hypothesis, the C$^{34}$S optical depth was computed for
477: both the eastern and western cores at positions where we have observations
478: of both CS and \CtfS , using the following relationship:
479: \begin{equation}
480: \frac{\rm T(C^{34}S)}{\rm T(CS)} = \frac{1-\exp{[-\tau]}} {1-\exp{[-a\tau]}},
481: \end{equation}
482: where $\tau$ is the C$^{34}$S optical depth and $a$ is the abundance ratio
483: of $[^{32}\rm{S}]/[^{34}\rm{S}] = 14$ \citep{pratap97}. Using the ratio of the
484: integrated intensities, we iteratively solved Equation 2 for $\tau$, until
485: the calculated intensity ratio matched the measured intensity ratio to
486: within $\sim$5\%. In Figure~\ref{fig_cstau}, the C$^{34}$S optical depth is shown
487: as a function of visual extinction $A_V$.
488: The uncertainties shown in Fig.~\ref{fig_cstau} are a result of the
489: uncertainties in the integrated
490: intensity measurements; the uncertainties associated with the
491: convergence of the numerical calculations are significantly smaller.
492: The derived opacities clearly indicate that the emission from C$^{34}$S is
493: optically thin in both cores for all A$_V$. However, the emission from CS ($\tau_{\rm CS}
494: = \tau_{\rm C^{34}S} * 14$) must be optically thick throughout the cloud.
495: Surprisingly, despite being thin, the C$^{34}$S optical depth
496: is nearly constant for $A_V \gtrsim 10$ mag.
497:
498: As a check of the derived optical depths, the optical depth of $^{13}$CS
499: was calculated in a similar manner to the C$^{34}$S optical depth. Only
500: the position ($-0\farcm 54, 0\farcm 89$) was sufficiently strong to permit
501: such a calculation ($\int T(^{13}{\rm CS}) dv = 0.089\pm0.02$ K km
502: s$^{-1}$). An optical depth of $\tau = 0.033 \pm 0.008$ was found,
503: implying that the C$^{34}$S optical depth (assuming an abundance ratio of
504: $[\rm{C}^{34}S]/[^{13}\rm{CS}] = 4.29$) is
505: $\tau = 0.14 \pm 0.03$ (see Fig. 5). The C$^{34}$S optical depth
506: measured using the T(\CtfS )/T(CS) ratio is $\tau = 0.14 \pm 0.04$,
507: at the same position, confirming the small opacities of C$^{34}$S
508: in this cloud.
509:
510: Examining the dependence of \CtfS\ opacity with extinction
511: in Figure~\ref{fig_cstau} we find that while the extinction rises
512: by a factor of 2.5 ($15-40$ mag), the C$^{34}$S
513: optical depth remains nearly constant at $\tau \approx 0.15-0.25$. Indeed, there
514: is even a small hint that the optical depth of the C$^{34}$S in the
515: eastern core is decreasing as a function of increasing extinction.
516: The eastern core also appears to have an opacity that is, on average,
517: slightly higher than the western core.
518: As optical depth is a direct tracer of column density in either the
519: upper or lower state, this difference suggests that the abundance
520: of CS could be different between the two cores.
521: Moreover, the constant opacity with extinction for both cores
522: would then be indicative of
523: a \CtfS (and CS) abundance decrease with increasing extinction.
524: However, CS and \CtfS\ have high dipole moments and their emission
525: is sensitive to gas density, as opposed to
526: \CeiO\ emission which is fairly insensitive to the density (provided
527: \nhtwo\ $> 1000$ \cc ). Thus, for CS and its isotopic
528: variants, changes in the excitation conditions can play an important,
529: and with regards to depletion analyses, potentially confusing role.
530: These effects are examined in \S 4.
531:
532: \subsection{N$_2$H+}
533:
534:
535: Unlike CS, or even CO, \NtwoHp\ is predicted by evolutionary
536: chemical models to have a low depletion rate, due to the
537: relatively low binding energy of its precursor molecule, \Ntwo\
538: \citep{bl97}.
539: Thus, we might expect that \NtwoHp\ would exhibit a different
540: behavior as a function of extinction when compared to CS or CO.
541:
542: In Figure~\ref{fig_n2hpint}, the
543: the N$_2$H+ integrated intensity, summed over all hyperfine transitions,
544: is compared to the visual extinction. This figure clearly shows
545: quite a different relationship between \NtwoHp\ integrated emission and
546: visual extinction when compared to
547: the other species included in our study.
548: For $A_V
549: \lesssim 15$ mag, the N$_2$H+ integrated intensity
550: vs $A_V$ is nearly flat and follows the
551: following relationship:
552: \begin{equation}
553: I(N_2H+) = \left(-0.03 \pm 0.04\ K\ km\ s^{-1}\right) + \left(0.017 \pm
554: 0.006\ \frac{K\ km\ s^{-1}}{mag}\right)\cdot [A_V]\ (A_V < 15\ {\rm mag}).
555: \end{equation}
556: Note that the slope of the linear fit is non-zero at the 3$\sigma$ level,
557: indicating that the N$_2$H+ intensity gradually increases with extinction.
558: At $A_V \gtrsim 15$ mag, the intensity displays a sharp increase and the
559: N$_2$H+$-A_V$ relationship steepens by nearly a factor of 15:
560: \begin{equation}
561: I(N_2H+) = \left(-4.3 \pm 0.6\ K\ km\ s^{-1}\right) + \left(0.26 \pm
562: 0.03\ \frac{K\ km\ s^{-1}}{mag}\right)\cdot [A_V]\ (A_V > 15\ {\rm mag}).
563: \end{equation}
564: Curiously, the increase in \NtwoHp\ intensity occurs at a similar
565: visual extinction to where the CO-\av\ relationship begins to show
566: evidence of potential depletion.
567: The derivation of \NtwoHp\ abundances and the physical cause for the
568: sharp increase at \av\ $\gtrsim 15$ will be discussed in the following
569: section.
570:
571: \section{Analysis: Molecular Excitation and Abundances}
572:
573: In the previous section we searched for correlations between integrated
574: emission and visual extinction for each of the surveyed molecules.
575: Here we examine whether any departures from the correlation
576: are the result of excitation and/or abundance gradients.
577: In this section we concentrate on trends in relative or normalized
578: abundances (i.e. abundances normalized using the abundance derived at the lowest extinction with
579: reliable data points). In this fashion,
580: in the determination of molecular abundance for a given species,
581: we eliminate common uncertainties (such as collision rates).
582:
583: To calculate total column densities from the molecular data we
584: account for the radiation transfer through the use of the
585: large-velocity gradient approximation. To reduce the effects of opacity
586: we use here only the C$^{17}$O, \CtfS , and \NtwoHp\ data.
587: Since the emission of C$^{17}$O and \CtfS\ (\S 3.1 and 3.2) is optically
588: thin, and, based upon ratios of hyperfine components, much
589: of the \NtwoHp\ emission is also thin, this
590: reduces the LVG approximation to simply solving the equations of statistical equilibrium.
591: The results are therefore less dependent on the details of the radiative transfer solution.
592: To improve the statistics of our results we average abundances in bins of
593: 5 magnitudes starting with \av\ = 5 mag. Below this value there are little or no significant
594: ($> 3\sigma$) data points (for \NtwoHp , C$^{17}$O, and \CtfS ).
595:
596: In this study we observe only a single transition of a given molecular species.
597: However, in each case we have information on the optical depth, using
598: CS for \CtfS , and the hyperfine ratios for \CsvO\ and \NtwoHp .
599: This provides an
600: additional and limiting constraint on the column density determination.
601: With the observed intensity and opacity from a single transition the total column density can
602: be derived with knowledge of the collision rates, density, temperature, and line width.
603: The collision rates are available from the literature
604: and the line width or velocity dispersion is an observed
605: quantity. However, to derive a map of the total column density from
606: the molecular line maps requires
607: a priori knowledge of the density and temperature structure of the
608: IC 5146 cloud.
609:
610: For the density structure we use the analysis in (Lada, Alves, \& Lada 1999; hereafter LAL99)
611: who show that the extinction (or column density) gradient in the same
612: region of IC 5146 is nicely
613: reproduced by a cylindrical geometry (for the entire streamer)
614: with $\rho \propto r^{-2}$. The procedure LAL99 use to determine the density structure
615: is identical to the commonly used method of deriving the volume density structure
616: from mm and sub-mm dust continuum emission (see Shirley et al 2000 and references
617: therein). To be consistent with the results in LAL99, which show
618: a large systematic increase in volume density from low to high extinction, we
619: adopt the following density profile (\nhtwo\ $=$ \nhtwo (r = r$_0$) $\times$ (r$_0$/r)$^{2}$
620: $= 10^5$(0.047 pc/r)$^2$ \cc ). The density profile is shown in Figure~\ref{fig_denpro} and is
621: in reasonable agreement with the radial profile of total column density (Figure 8 in LAL99).
622: In our calculations
623: we use average densities in bins of 5 mag (shown as horizontal hash marks in
624: Figure~\ref{fig_denpro}). For example, if a given data point has an extinction between 5 and 10
625: mag then we use the average density of \nhtwo\ $= 1.4 \times 10^{4}$ \cc\ within that bin.
626:
627: There exists some information on the temperature structure from
628: observations of CO and NH$_3$. CO observations generally show gas temperatures ranging from
629: 10 to 13 K \citep{dobashi92}. However, these estimates are likely reflective of the
630: temperature at the cloud surface and not the denser interior \citep{bgsu94}.
631: In the eastern core NH$_3$ observations provide some information on the dense gas
632: temperature with T = 13 K \citep{jma99}. We also have a single detection
633: of the J = 6 $\rightarrow$ 5 (K = 0 and 1) transition of CH$_3$C$_2$H, which can be used to
634: estimate the temperature \citep{bgsu94}. Since this detection is at a single position
635: ($\Delta \alpha = -0\farcm 5$, $\Delta \delta = +0\farcm 9$)
636: we have not shown the spectra; however, the K = 0 integrated intensity
637: is $\int T \delta v =$ 0.11$\pm 0.02$ K km s$^{-1}$ and K = 1 is
638: $\int T \delta v =$ 0.13$\pm 0.02$ K km s$^{-1}$.
639: Using the method described in \citet{pratap97} the ratio
640: of these two is consistent, within the errors, with a gas temperature $< 13$ K.
641: Given this information, it is unlikely that there are strong temperature gradients
642: throughout most of the IC5146 molecular streamer
643: with the probable range between (perhaps) 5 and the CO temperature
644: of 13 K. In the following we present abundances calculated with a constant temperature
645: of 10 K and discuss the results of additional analysis assuming constant temperatures
646: of 5 and 13 K (and combinations thereof).
647:
648: \subsection{C$^{17}$O Abundance}
649:
650: Because of concerns that moderate opacities in \CeiO\ emission could potentially
651: mask small abundance changes in order to search for relative abundance differences
652: we predominantly use the more limited \CsvO\ observations.
653: For the \CsvO\ column density derivation, we use the collision rates of CO with para-H$_2$
654: \citep{f88}. We adopt the density determined as a function of extinction
655: as described above and the temperature is assumed to be constant at 10 K for each position
656: (changes in the temperature will be discussed below).
657: The line width is determined via gaussian fits to
658: each spectrum.
659: Because the \CsvO\ J $= 1 \rightarrow 0$ transition has hyperfine structure we use
660: this additional information in our derivation of column density. We therefore
661: simultaneously fitted the two resolved hyperfine components (as discussed in \S 3.1 the
662: F $= 7/2 \rightarrow 5/2$ and F = $= 3/2 \rightarrow 5/2$ are blended), assuming that the
663: hyperfine levels are populated according to LTE. The two constraints on the
664: on the $\chi^2$ search for the best fit column density are: (1) the intensity of the lines
665: and (2) the relative intensities of the hyperfine ratios which limits the opacity of the
666: solution. In general the reduced $\chi^2$ of most solutions are $< 1$ and
667: in all cases the \CsvO\ emission is found to be optically thin.
668: For each position we derive column densities if the observed integrated intensity is $> 3\sigma$.
669: The abundance relative to H$_2$ is derived using N$_{H_2} = 10^{21}$\av .
670:
671: Table~3 presents the weighted average abundance of C$^{17}$O
672: (within 5 mag bins of extinction) and in Figure~\ref{fig_abun}
673: the relative abundance of \CsvO\ is given as a function
674: of visual extinction. In this figure we normalize the abundance to the lowest
675: extinction bin with a statistically significant abundance determination (between 5 and 10 mag).
676: Excluding the point at \av\ = 27.5 mag with large errors,
677: we see that the relative \CsvO\ abundance shows a steady and significant decrease, by
678: a factor of 3, from the lowest to highest visual extinction.
679: This result is found assuming a constant temperature of 10 K. To investigate
680: whether a systematic temperature gradient can change the result we repreated the same
681: procedure assuming constant temperatures of 5 and 13 K. To mimic temperature gradients
682: we examined combinations of these results. In the absence of a significant population
683: of luminous embedded stars (the case here) the most likely systematic temperature gradient
684: is the temperature decreasing with increasing extinction. In this case we find that
685: the \CsvO\ abundance at high \av\ (with T = 5 K) does increase due to higher
686: opacity in the J = 1 state, allowing for greater column density, but by only 15\%.
687: At the cloud edges (provided T = 13 K) the abundance also shows a slight increase,
688: primarily due to the lower population in the J = 1 state. If the temperature were even
689: greater then the abundance continues to rise. In all, this examination
690: still provides a factor of 3 decrease in the \CsvO\ abundance from \av\ = 5 to 40 mag.
691: It is worth emphasizing that even if CO and its isotopic variants are completely depleted
692: in the dense core center the molecular observations will detect emission from the relatively
693: undepleted gas residing in the low density regions along the line of sight through the core.
694: Thus, if CO shows evidence for depletions at 10 mag in a cloud with a total extinction (from
695: Fig~5 for example) of 40 mag, then the measured abundance will drop by at most a factor of four.
696:
697: A similar analysis was performed using the entire \CeiO\ data set and we find that the
698: relative abundance is essentially constant until the highest extinction bin, whereupon the
699: abundance decreases by a factor of 1.5. This confirms our suspicion that the \CeiO\
700: emission is likely to have moderate optical depth ($\tau \sim 0.5$) which masks
701: the larger abundance difference (factor of 3) seen in \CsvO .
702: We note that one core in our data set ($\Delta \alpha = -9\farcm 2$, $\Delta \delta = 2\farcm 9$)
703: has been examined with the IRAM
704: 30m antenna at higher ($\sim 30''$) resolution by \citet{kram99}.
705: They find a clear decrease in the CO abundance near 12 mag.
706: Our results are not as dramatic as found in the higher resolution study, which
707: allowed for more data points to be placed at higher extinctions.
708: For instance, the \citet{kram99} core
709: shows a peak extinction in a 30$''$ beam of $\sim 28$ mag,
710: at a resolution of 50$''$ this reduces to a peak extinction of 16 mag.
711: Thus, our sampling is lower in the exact regions where gas-phase
712: depletions are most likely.
713: %Moreover, there are several points near
714: %the centers of dense cores that, due to a lack of stars within the
715: %central portions of the gaussian NIR beam, we are able to derive only
716: %lower limits to their extinction.
717:
718: \subsection{\CtfS\ (and CS) Abundance}
719:
720: To derive the total column density of \CtfS\ we use a similar
721: method described for \CsvO\ in the previous section. That is, with
722: a density and temperature assumed (and constant) for each grid
723: position we perform a $\chi^2$ search for the best fit total column
724: density. For collision rates we use the rates given by
725: \citet{gc78}, and the line width is
726: determined via gaussian fits to the spectrum.
727: \CtfS\ does not have hyperfine structure, but we have additional information
728: in that the \CtfS\ opacity has been determined for all positions
729: with $> 3\sigma$ integrated intensities (\S 3.2). Thus,
730: for each data point we again
731: two constraints for the radiative transfer model: (1) the integrated intensity
732: and (2) the opacity of the J $= 2 \rightarrow 1$ transition.
733: The majority of column density solutions have reduced $\chi^2 < 1$.
734:
735: The \CtfS\ integrated intensity as a function of visual extinction is
736: presented in Figure~\ref{fig_c34sint}. This Figure is notable in that
737: the \CtfS\ integrated intensity appears to be reasonably well correlated
738: with \av . This result is somewhat surprising given the near constant
739: opacity as a function of extinction shown in Figure~\ref{fig_cstau}.
740: The near uniform opacity would suggest that the \CtfS\ (and therefore CS)
741: abundance is sharply decreasing with extinction. However the dependence
742: in Figure~\ref{fig_c34sint} is at odds with that conclusion,
743: as it requires a constant abundance.
744: This difference, an integrated intensity correlated with \av , combined with a
745: constant optical depth, can be reproduced,
746: provided that there was a change in excitation
747: from low to high extinctions.
748: Indeed we can expect such a increase in the \CtfS\ excitation temperature
749: proceeding from lower to higher \av , as a direct result of the known density gradient.
750: Even a small change in the excitation temperature could replicate the observed
751: dependence.
752:
753: The results of the \CtfS\ excitation analysis are presented in Table~3 in the form
754: of abundances and in Figure~\ref{fig_abun} as normalized abundance
755: plotted against visual extinction. Examining the data points we find
756: a steady statistically significant abundance ($> 3\sigma$)
757: decrease by an overall factor of 3 from low to high extinction.
758: Thus, by accounting for the rise in density as required by the NIR observations, and
759: the observed constant opacity, along with an integrated intensity that increases with
760: extinction, we find that the CS abundance decreases with \av .
761: If we assume that the most likely temperature gradient is with warm gas at core edges
762: (T = 13 K)
763: and cold gas (T = 5 K) in the center then the abundance still systematically declines
764: by a factor of 2. In all, these results suggest that CS, like CO, shows an abundance
765: decrease in the dense gas likely due to depletion onto grains.
766:
767: \subsection{\NtwoHp\ Abundance}
768:
769: To derive the abundance of \NtwoHp\ we use the total intensity of the
770: lines along with the hyperfine ratios as described earlier for C$^{17}$O.
771: We use the collisional
772: rates for HCO$^+$ excited by para-H$_2$ \citep{f99}, which
773: has comparable collisional rates to \NtwoHp\ \citep{m84}.
774: The results of the excitation calculations are given in Table~3 and shown
775: in Figure~\ref{fig_abun}.
776:
777: First, there are only 3 \NtwoHp\ detections
778: between 0 and 5 mag of extinction, and of these, two have
779: \av\ $>$ 4.5 mag (these data are not provided due to high errors).
780: Although \CtfS\ and \CsvO\ also have little emission
781: in this regime, the more abundant isotopes, \CeiO\ and CS, both are detected
782: at low extinction.
783: Thus, unlike for other species, there appears to be little \NtwoHp\ below
784: \av\ $\lesssim$ 4 mag.
785: Beyond the lowest extinctions, for T = 10 K the \NtwoHp\ normalized abundance
786: in Figure~\ref{fig_abun} shows a high abundance for low \av\ ($\sim 7.5$ mag)
787: then decreases by a factor of 3 for \av\ = 17.5 mag, whereupon the abundance
788: increases by a similar factor at high \av . In our solutions with different
789: temperatures (T = 5, 10, and 13 K) we find that the initial abundance decrease
790: for \av\ $< 17.5$ mag is lessened, provided that the cloud is warmer in these regions.
791: If the gas is cooler in the core center then we find an even greater increase
792: in relative abundance for \av\ $> 17.5$ mag.
793:
794: In summary we find, at least, two regimes for the \NtwoHp\
795: abundance: (1) there is a threshold of extinction, $A_V^{th} \lesssim 4$
796: mag below which we find little or no \NtwoHp\
797: in the gas-phase and (2) for higher extinctions the abundance shows a complicated
798: structure that initially declines in value until \av\ = 15 mag whereupon
799: it the relative abundance increases with extinction. Consideration of potential temperature
800: changes reduces the initial decline, but would increase the rise in relative abundance
801: towards to the core center. However, there is
802: no evidence for a systematic decrease in the \NtwoHp\ abundance at large \av .
803: In the following section we examine the physical and chemical
804: causes that give rise to this dependence.
805:
806: \section{Discussion}
807:
808: \subsection{Differential Molecular Depletions in IC 5146}
809:
810: In the preceding section we examined the excitation of the three
811: molecules included in our survey. The results of this analysis is
812: that we find good evidence for the presence of differential molecular
813: depletions in IC 5146. That is, \CeiO\ and CS (and \CtfS ) are
814: depleting from the gas phase at high extinctions and \NtwoHp\ is not (at least at a
815: resolution of 50$''$). Similar evidence of differential
816: depletions are found in the literature, for a summary see \citet{bergin00}.
817: The depletion of CCS relative to both ammonia and \NtwoHp\ (with the
818: latter two tracers typically coincident with the dust continuum emission
819: peak) found in isolated low-mass cores is quite similar to the
820: results found in our work \citep{klv96, olwh99}.
821: Recently, \citet{caselli99} found evidence of
822: CO depletion in the starless L1544 cloud core in a region traced by
823: both \NtwoHp\ and dust continuum.
824: The detection of gas phase molecular depletions should not be considered as unexpected,
825: given that molecular ice features have been observed
826: along numerous lines of sight in the ISM \citep{whittet93, ttgb91}.
827:
828: In general, previous searches for
829: chemical differences related to molecular depletion relied on
830: finding morphological dissimilarities between the emission of
831: two or more species with each other or with dust continuum
832: (similar to that shown in Figure~\ref{fig_wc}) and
833: then performing an excitation analysis to prove or disprove that
834: such differences are the result of chemical abundance variations.
835: Our study has found comparable results, but has placed the evidence
836: for molecular abundance changes, which are attributed to
837: depletion, on firmer statistical grounds and removes any ambiguity in terms
838: of the total hydrogen column density, such as those associated with
839: using dust continuum emission \citep{kram98}.\footnote{Of course, another
840: method to search for molecular depletions is to directly observe
841: molecules in the solid state. This method works quite well for
842: the dominant molecules on grain surfaces (such as H$_2$O or CO).
843: However, it fails for lesser abundant species such as CS because, even
844: if they depleted entirely from the gas-phase their abundances are too
845: low to produce observable absorption features.}
846:
847: \subsection{Comparison with Chemical Theory}
848:
849: This depletion pattern with sulfur bearing molecules
850: (e.g. CS, CCS), along with CO, showing depletions onto grains and
851: \NtwoHp\ remaining in the gas-phase is in good agreement with the
852: theoretical models presented by \citet{bl97}. In these models,
853: the gas-phase chemistry includes the effects of
854: molecules both depleting onto and evaporatoring from grain surfaces.
855: \NtwoHp\ remains in the gas-phase longer at higher
856: densities than other molecules because the high
857: volatility of its pre-cursor molecule,
858: \Ntwo , allows for various desorption mechanisms (such as thermal
859: evaporation or cosmic-ray spot heating of grain surfaces) to keep
860: a significant \Ntwo\ abundance, and therefore \NtwoHp ,
861: in the gas phase at high densities.
862: %By comparing
863: %our abundance structure as function of extinction (Figure~\ref{fig_abun})
864: %to these same chemical models
865: %we have found that CO depletion is also an important factor in allowing
866: %\NtwoHp\ to preferentially trace only the high density (and high \av ) gas.
867:
868: To compare our results with the gas-grain chemical model of \citet{bl97},
869: we use the density profile given in Figure~\ref{fig_denpro}. This profile
870: matches the one derived for IC 5146 from the NIR extinction measurements
871: (LAL99). We adopt the notation $\tau_V$ with regards to cloud depth
872: because the model is computed using cloud radius, thus \av\ $= 2\tau_V$.
873: With extinction and density (we also assume a constant dust and gas temperature
874: of 10 K and a UV enhancement factor of $G_{0} = 2.2$; Kramer et al 1999)
875: predetermined as a function of radius,
876: we use the chemical model of \citet{bl97} to predict abundances as a function of
877: both time and visual extinction and then directly compare these to the observational
878: results. This model is slightly different from \citet{bl97} in that we
879: do not allow for the density to evolve with time, but rather
880: hold the density constant for each radius (and therefore extinction).
881: However, the density does increase with increasing extinction, properly
882: matching the observed structure.
883:
884: Figure~\ref{fig_chemmod} presents two different chemical models.
885: The top two panels show a pure gas-phase
886: chemical model sampled at two different times ($10^5$ and 10$^7$ yr).
887: This model includes the effects of grain absorbing
888: UV radiation (thereby allowing for molecular formation) but does not
889: include any gas-grain interactions.
890: We note that to match the models to observations $\tau_V$ must be
891: multiplied by a factor of two to account for both sides of the cloud
892: (i.e. the model was done using radius and not diameter).
893: In Figure~\ref{fig_chemmod}, we see that the abundances of most
894: species rise as the UV radiation field is attenuated
895: with increasing extinction and then remain constant. This behavior
896: exists despite the nearly two order of magnitude rise in density from edge
897: to center. However, for \NtwoHp\ there is a slow steady decrease in abundance
898: towards greater extinction. This behavior is readily understood by examining
899: the primary formation and destruction processes for \NtwoHp .
900: Using the pathways outlined in Appendix A, we see
901: that the formation rate of \NtwoHp\ (via cosmic ray ionization) should
902: be essentially constant with density, with the assumption that cores are
903: threaded by cosmic rays. However,
904: due to the higher probability for collisions with higher density,
905: the destruction rate via dissociative recombination and reactions
906: with CO is raised. Thus, for pure gas-phase models the prediction
907: is that the abundance of \NtwoHp\ decreases with extinction.
908:
909: The second model shown in the bottom two panels of Figure~\ref{fig_chemmod}
910: is a gas-grain chemical model where molecules are allowed to collide
911: and stick to the surfaces of dust grains. The primary desorption mechanism
912: is cosmic-ray spot heating. For
913: the binding energy of molecules to the grain surfaces we assume that
914: the grains are coated by a layer of CO molecules. For more details
915: of the model see \citet{bl97}.
916: Comparing the pure gas-phase model with the gas-grain model at t = 10$^{5}$ yr
917: the abundance profiles are similar. At this
918: early time depletion has yet to play a role. For later times, we
919: see demonstrative effects due to grain depletion: (1)
920: for $\tau_V > 6$ mag (\av\ $> 12$ mag) the abundance of CS is
921: dramatically reduced (2) the abundance
922: of CO shows a small, factor of $\sim$3, decrease between $\tau_V$ $= 0$ to
923: 16 mag and (3) the abundance of \NtwoHp\ is constant with \av . For \NtwoHp\
924: this behavior is in direct contrast to that seen in
925: the pure gas-phase case. This indicates
926: that the \NtwoHp\ destruction rate must be progressively lower with
927: higher extinction and density in the
928: gas-grain model than for pure gas-phase chemistry. From Appendix A the major
929: destroyers of \NtwoHp\ are CO and electrons. In Figure~\ref{fig_chemmod}
930: there is increasing CO depletion with higher density and extinction
931: which results in a progressively lower \NtwoHp\ destruction rate.
932: Thus, the observed uniform abundance of \NtwoHp\ at high extinctions is
933: also a direct consequence of CO depletion and is not solely the result of
934: the high volatility of the \Ntwo\ molecule.
935: It is worth noting that if more CO molecules freeze onto grains, as would occur when
936: time progresses, then \NtwoHp\ abundance rises accordingly.
937:
938: Finally, we note that at low \av ,
939: for CS and CO (x(\CeiO ) $\times$ 500),
940: the models can be directly compared to the observed abundances and
941: predicted values are in reasonable agreement with observations
942: (compare Figures~\ref{fig_abun} and~\ref{fig_chemmod}).
943: For high \av , the models, as presented here as abundance profiles
944: with extinction, cannot be directly compared to the observations.
945: This is because the observations are averages over the entire line of
946: sight, including
947: both depletion zones and undepleted gas. For a direct comparison,
948: the chemical model would have to be
949: averaged over the line of sight, stepping through \av ,
950: in similar fashion to the observations.
951: However, this is not the case for undepleted species (e.g. \NtwoHp ), which
952: can be directly compared to observations.
953: If the relative abundance profile in Figure~\ref{fig_abun} is correct then
954: the observations would suggest a combination of undepleted models (to account
955: for the low \av\ decline in abundance) and the depleted models (accounting for
956: the rise in abundance at high extinction) would be required.
957: In each case the models are in reasonable agreement with the
958: overall value of abundance.
959:
960: \subsection{Testing Interstellar Photodissociation Rates}
961:
962: In \S 4.3 we found two regimes for the \NtwoHp\ abundance:
963: (1) there is a threshold extinction, $A_V^{th} \lesssim 4$ mag, below
964: which there is little or no \NtwoHp , and
965: (2) for higher extinctions the abundance is roughly constant. In
966: the previous section we discussed how the structure in
967: latter regime can be reproduced. However, in Figure~\ref{fig_chemmod} the
968: dependence found in the former regime is also addressed.
969: In the chemical model
970: the lack of \NtwoHp\ for \av\ $\lesssim 4$ mag
971: is the result of the photodissociation of its parent molecule (\Ntwo ),
972: combined with a higher electron abundance at cloud edges due to
973: photoionization.
974: In Figure~\ref{fig_chemmod}, \NtwoHp\ shows a sharp rise in abundance for
975: $\tau_V \sim 3$ mag, corresponding to \av\ $\sim 6$ mag. This is
976: quite close to the observed threshold. The
977: \Ntwo\ photorate is taken from \citet{vd88}, where the depth dependence of
978: the photodissociation rate
979: depends on its absorption cross section\footnote{For \Ntwo\ predissociation occurs
980: mainly between 912--1000 \AA\ \citep{carter72, rtt81}}, the assumed radiation
981: field, the continuum attenuation, and on the albedo and
982: scattering phase function of the grains \citep{vd88}. Despite the
983: complications involved in this computation, given the match
984: between the observed and predicted threshold, the depth dependence of the
985: rate and the ionization structure, appears to be reasonably taken into account.
986:
987: We note that the chemical model in Figure~\ref{fig_chemmod}
988: does not properly account
989: for the self-shielding of CO molecules, as \CeiO\ is emission is observed
990: for \av\ $< 4$ mag (Figure~\ref{fig_c18oint}). However, the effect of
991: increased CO shielding in the model would be to lower the electron abundance at the
992: cloud edge and increase the importance of \Ntwo\ photodestruction
993: as the method for lowering the \NtwoHp\ abundance. Our assertion
994: that \Ntwo\ photodissociation is contributing to the observed \NtwoHp\ abundance
995: threshold would then be strengthened.
996:
997: \citet{vd88} present a different photorate computed using a different
998: grain model than used in our model, one with the grains more forward
999: scattering (their grain model 3). If we use the depth dependent photorates from
1000: that model (for all steps in \Ntwo\ and \NtwoHp\ production) we find that
1001: the \Ntwo\ photodissociation rate is larger deeper into the cloud and
1002: the \NtwoHp\ threshold shifts to beyond \av\ = 10 mag. This is in
1003: disagreement with the observations and points to an important use of the
1004: technique outlined in this paper.
1005: In the future, by observing and modeling clouds with simple and
1006: well-characterized structure, the depth
1007: dependent photorates (along with the gas-grain interaction) can be tested.
1008: Indeed, most of the depth dependent photorates for many molecules are
1009: highly uncertain \citep{vd88}. In some cases this is due to a lack of viable
1010: cross-sections, but also because the detailed rate calculations could
1011: not be observationally verified. Although, there certainly are complications in
1012: application; this technique offers a viable opportunity to
1013: provide observational tests of these rates, which can help in constraining
1014: not only the depth dependence of the photorates but also the grain physics.
1015:
1016:
1017: \section{Summary}
1018:
1019: We have examined the correlation of the emission of three molecules:
1020: \CeiO , CS, and \NtwoHp\ with visual extinction in the northern streamer
1021: of the IC 5146 cloud. We use the molecular data, along with a model
1022: of the excitation, to determine molecular abundances and examine the
1023: abundance profile with \av . The principle results are:
1024:
1025: (1) We find good evidence for a reduction in the abundance of CO and
1026: CS (using \CtfS ) for \av\ $\gtrsim 10 - 15$ mag. This reduction is
1027: attributed to the depletion of these molecules onto the surfaces of cold
1028: dust grains in the densest portions of the IC 5146 cloud.
1029:
1030: (2) For \NtwoHp , we find two separate regimes for its abundances: (a)
1031: for low extinction there is a threshold extinction, $A_V^{th} \lesssim 4$ mag,
1032: below which little \NtwoHp\ is found and (b) for \av\ $\gtrsim 4$ mag the
1033: \NtwoHp\ abundance is either constant or declining until \av\ = 15 mag whereupon
1034: the abundance rises as a function of extinction in the visual.
1035:
1036: (3) We find that density gradients in cloud cores can have a
1037: profound effect on the derivations of the spatial distributions
1038: of total column densities and molecular abundances for species with
1039: high dipole moments. Chemical interactions, such as reaction rates
1040: and depletion timescales, are also strongly density dependent.
1041: Thus the interpretation and determination of accurate abundances in dense
1042: cloud cores requires both knowledge of the density and temperature
1043: structure in the cloudy material and a proper accounting for such
1044: structure in abundance determinations and chemical models.
1045:
1046: (4) The observed patterns of differential depletion shows CS and CO exhibiting
1047: molecular depletions while \NtwoHp\ remains in the gas phase. This is in good
1048: agreement with the predictions of chemical theory by \cite{bl97}.
1049:
1050: (5) By combining the observationally derived density profile for IC 5146 from
1051: \cite{lal99} with the gas-grain chemical model of \cite{bl97}, we show that the
1052: observed constant \NtwoHp\ abundance with extinction can be reproduced
1053: provided that CO, which is a major destroyer of \NtwoHp , is
1054: depleting in the dense cores. Since CO depletion is observed
1055: this result is an important clue to the
1056: physical and chemical reason that allows \NtwoHp\ molecules to trace
1057: the densest regions of molecular cores.
1058:
1059: (6) We also demonstrate that the observed \NtwoHp\ abundance threshold
1060: ($A_V^{th} \lesssim 4$ mag)
1061: is partially due to the photodissociation of \Ntwo . This highlights
1062: an important application of this technique, which is the ability to
1063: test not only the chemical models of dense regions, and thereby
1064: directly probe the gas-grain interaction, but also to provide a viable
1065: and completely new method for observationally testing models of
1066: depth-dependent photodissociation rates.
1067:
1068:
1069:
1070:
1071: \appendix
1072: \section{Formation of N$_2$H$^{+}$ Abundance}
1073:
1074: The primary route to form N$_2$H$^{+}$ is through the following reaction:
1075:
1076: \begin{equation}
1077: \eqnum{A1}
1078: \rm{H_3^+ + N_2 \rightarrow N_2H^+ + H_2}.
1079: \end{equation}
1080:
1081: \noindent \Hthreep\ is the produced via cosmic ray ionization of
1082: H$_2$. The main destruction pathways are
1083: dissociative electron recombination and a reaction with CO (reactions
1084: with O and C are of lesser import).
1085: In equilibrium, the following expression can be derived for
1086: the concentration of N$_2$H$^{+}$,
1087:
1088: \begin{equation}
1089: \eqnum{A2}
1090: \rm{
1091: n(N_2H^+) = \frac{n(H_3^+)n(N_2)k_{H_3^+,N_2}}{n(e^-)\alpha(N_2H^+)
1092: + n(CO)k_{CO,N_2H^+}}
1093: }.
1094: \end{equation}
1095:
1096: \noindent
1097: Where $\alpha(\rm{N_2H^+})$ is the dissociative recombination coefficient
1098: and k$_{\rm{x,N_2H^+}}$ are the various reaction rates given
1099: by \citet{mfw97}. See also \citet{wzw92}.
1100: %This expression can be then solved for the abundance of N$_2$, which is
1101: %
1102: %\begin{equation}
1103: %\eqnum{A3}
1104: %\rm{
1105: %n(N_2) = \frac{n(N_2H^+)}{n(H_3^+)k_{H_3^+,N_2}}
1106: %\{n(e^-)\alpha(N_2H^+) + n(CO)k_{CO,N_2H^+}}.
1107: %}
1108: %\end{equation}
1109:
1110:
1111:
1112:
1113: \begin{thebibliography}{}
1114:
1115: \bibitem[Alves, Lada, \& Lada (1999)]{all99} Alves, J., Lada, C. J., \& Lada,
1116: E. A. 1999, ApJ, 515, 265
1117:
1118: \bibitem[Alves et al. (1998)]{alves98} Alves, J., Lada, C. J., Lada, E. A., Kenyon,
1119: S. J., \& Phelps, R. 1998, ApJ, 506, 292
1120:
1121: \bibitem[Bergin (2000)]{bergin00} Bergin, E.A. 2000, in Astrochemistry: From
1122: Molecular Clouds to Planetary Systems, eds. Y.C. Minh \&
1123: E.F. van Dishoeck, IAU Symposium, 197, in press
1124:
1125: \bibitem[Bergin \& Langer (1997)]{bl97} Bergin, E. A. \& Langer, W. D.
1126: 1997, ApJ, 486, 316
1127:
1128: \bibitem[Bergin et al (1994)]{bgsu94}
1129: Bergin, E.\ A., Goldsmith, P.\ F., Snell, R.\ L.\ \& Ungerechts, H.\ 1994,
1130: \apj, 431, 674
1131:
1132: % \bibitem[Benson \& Myers (1989)]{bm89} Benson, P. J. \&
1133: % Myers, P. C. 1989, \apjs, 71, 89
1134:
1135: \bibitem[Bohlin, Savage, \& Drake (1978)]{bsd78} Bohlin, R. C., Savage, B. D., \&
1136: Drake, J. F. 1978, \apj, 224, 132
1137:
1138: \bibitem[Carter (1972)]{carter72} Carter, V. 1972, J. Chem. Phys., 56, 4195
1139:
1140: \bibitem[Caselli et al. (1999)]{caselli99} Caselli, P.,
1141: Walmsley, C. M., Tafalla, M., Dore, L. \& Myers, P. C. 1999, \apjl, 523,
1142: L165
1143:
1144: \bibitem[Caselli, Myers, \& Thaddeus (1995)]{cmt95} Caselli, P., Myers, P. C., \&
1145: Thaddeus, P. 1995, \apj, 455, L77
1146:
1147: \bibitem[Ciardi et al. (1998)]{crd98} Ciardi, D. R., Woodward, C. E.,
1148: Clemens, D. P., Harker, D. E., \& Rudy, R. J. 1998, \aj, 116, 349
1149:
1150: \bibitem[Dobashi et al. (1992)]{dobashi92}
1151: Dobashi, K., Yonekura, Y., Mizuno, A. \& Fukui, Y. 1992, \aj, 104, 1525
1152:
1153: \bibitem[Flower (1988)]{f88} Flower, D.R. 1988, Molecular Collisions in
1154: the Interstellar Medium, (Cambridge: Cambridge University Press)
1155:
1156: \bibitem[Flower (1999)]{f99} Flower, D.R. 1999, MNRAS, 305, 651
1157:
1158: %\bibitem[Frerking, Langer, \& Wilson (1982)]{flw82}
1159: %Frerking, M.A., Langer, W.D., \& Wilson, R.W. 1982, \apj, 262, 590
1160:
1161: \bibitem[Green \& Chapman (1978)]{gc78} Green, S., \& Chapman, S. 1978, \apjs, 37,
1162: 169
1163:
1164: \bibitem[Jenkins \& Savage (1974)]{js74} Jenkins, E. B. \& Savage, B. D.
1165: 1974, \apj, 187, 243
1166:
1167: \bibitem[Jijina, Myers \& Adams (1999)]{jma99} Jijina, J.,
1168: Myers, P.\ C.\ \& Adams, F.\ C.\ 1999, \apjs, 125, 161
1169:
1170: \bibitem[Kramer et al. (1998)]{kram98} Kramer, C. Alves, J., Lada, C.,
1171: Lada, E., Sievers, A., Ungerechts, H., \& Walmsley, M. 1998, \aap, 329,
1172: L33
1173:
1174: \bibitem[Kramer et al. (1999)]{kram99} Kramer, C. Alves, J., Lada, C.,
1175: Lada, E., Sievers, A., Ungerechts, H., \& Walmsley, M. 1999, \aap, 342,
1176: 257
1177:
1178: \bibitem[Kuiper, Langer, \& Velusamy (1996)]{klv96} Kuiper,
1179: T. B. H., Langer, W. D. \& Velusamy, T. 1996, \apj, 468, 761
1180:
1181: \bibitem[Lada et al. (1994)]{lada94} Lada, C. J., Lada, E. A., Clemens, D.
1182: P., \& Bally, J. 1994, \apj, 429, 694
1183:
1184: \bibitem[Lada, Alves, \& Lada (1999)]{lal99} Lada, C. J., Alves, J., \&
1185: Lada, E. A. 1999, \apj, 512, 250
1186:
1187: \bibitem[Ladd, Fuller, \& Deane (1998)]{lfd98} Ladd, E. F., Fuller, G. A.,
1188: \& Deane, J. R. 1998, ApJ, 495, 871
1189:
1190: \bibitem[Mathis (1990)]{math90} Mathis, J. S. 1990, \araa, 28, 37
1191:
1192: \bibitem[Millar, Farquhar, \& Willacy (1997)]{mfw97} Millar,
1193: T. J., Farquhar, P. R. A. \& Willacy, K. 1997, \aaps, 121, 139
1194:
1195: \bibitem[Monteiro (1984)]{m84} Monteiro, T. 1984, MNRAS, 210, 1
1196:
1197: \bibitem[Ohashi et al. (1999)]{olwh99}
1198: Ohashi, N., Lee, S. W., Wilner, D. J. \& Hayashi, M. 1999, \apjl, 518, L41
1199:
1200: \bibitem[Penzias (1981)]{penz81} Penzias, A. A. 1981, ApJ, 249, 518
1201:
1202: \bibitem[Pratap et al. (1997)]{pratap97} Pratap, P., Dickens,
1203: J. E., Snell, R. L., Miralles, M. P., Bergin, E. A., Irvine, W. M. \&
1204: Schloerb, F. P. 1997, \apj, 486, 862
1205:
1206: \bibitem[Reike \& Lebofsky (1985)]{rl85} Reike, G. H. \& Lebofsky, M. J.
1207: 1985, ApJ, 288, 618
1208:
1209: \bibitem[Richards, Torr, \& Torr (1981)]{rtt81} Richards, P.G., Torr, D.G., \&
1210: Torr, M.R. 1981, J. Geophys. Res., 86, 1495
1211:
1212: \bibitem[Shirley et al. (2000)]{serc00} Shirley, Y.L., Evans, N.J. II, Rawlings, J.M.C.,
1213: \& Gregerson, E.M. 2000, ApJ, in press
1214:
1215: \bibitem[Tielens et al. (1991)]{ttgb91} Tielens, A.G.G.M., Tokunaga, A.T.,
1216: Geballe, T.R., \& Baas, F. 1991, ApJ, 381, 181
1217:
1218: \bibitem[van Dishoeck \& Black (1988)]{vdb88} van Dishoeck, E. F. \& Black,
1219: J. H. 1988, ApJ, 334, 771
1220:
1221: \bibitem[van Dishoeck (1988)]{vd88}van Dishoeck, E. F.
1222: 1988, in Rate Coefficients in Astrochemistry, eds. T.J. Millar \& D.A. Williams,
1223: (Kluwer: Dordrecht), 49
1224:
1225: \bibitem[Whittet (1993)]{whittet93} Whittet, D. C. B. 1993, in Dust and Chemistry
1226: in Astronomy, eds. T. J. Millar \& D. A. Williams (Bristol:
1227: Institute of Physics Publishing), p9
1228:
1229: \bibitem[Womack, Ziurys, \& Wyckoff (1992)]{wzw92} Womack,
1230: M., Ziurys, L. M. \& Wyckoff, S. 1992, \apj, 387, 417
1231: \end{thebibliography}
1232: \newpage
1233:
1234: \begin{figure}
1235: \plotone{fig1.eps}
1236: \caption{
1237: (a) Gaussian convolved map of visual extinction toward the northern streamer
1238: of IC 5146. Contours start at 4 mag of \av\ and increase in steps of 4 mag.
1239: The stars show the positions of IRAS point sources.
1240: (b) Points delineating the radio grid overlayed on the region surveyed
1241: in the infrared (as shown by the map of visual extinction).
1242: The near infrared survey region is outlined in each.
1243: The central reference position is listed in Table~1.
1244: }
1245: \label{fig_avgrid}
1246: \end{figure}
1247:
1248: \begin{figure}
1249: \plotone{fig2.eps}
1250: \caption{
1251: (a) Map of C$^{18}$O J = 1 $\rightarrow$ 0 integrated intensity in IC 5146. Contour levels begin
1252: with 0.3 K-km/s and are spaced by 0.2 K-km/s. (b) Map of integrated
1253: N$_2$H$^+$ J = 2 $\rightarrow$ 1 emission. Contour levels begin at 0.6 K-km/s and increase
1254: in steps of 0.6 K-km/s.
1255: The stars show the positions of IRAS point sources.
1256: }
1257: \label{fig_c18o_n2hp}
1258: \end{figure}
1259:
1260: \begin{figure}
1261: \epsscale{0.8}
1262: \plotone{fig3.eps}
1263: \caption{
1264: Comparison of C$^{18}$O J = 1 $\rightarrow$ 0 integrated emission distribution
1265: (solid contours), N$_2$H$^+$ J = 1 $\rightarrow$ 0 (dashed contours), with
1266: visual extinction (grey scale) towards the western core of the IC 5146 northern
1267: streamer. Note how the N$_2$H$^+$ emission maximum corresponds with the peak
1268: of visual extinction. For C$^{18}$O the contour levels begin with
1269: and are spaced by 0.3 K-km/s, while for N$_2$H$^+$ contour levels begin with
1270: and are spaced by 0.6 K-km/s. }
1271: \label{fig_wc}
1272: \end{figure}
1273:
1274: \begin{figure}
1275: \plotone{fig4.eps}
1276: \caption{
1277: (a) Map of CS J = 1 $\rightarrow$ 0 integrated intensity in IC 5146. Contour levels begin
1278: with 0.5 K-km/s and are spaced by 0.5 K-km/s. (b) Smaller maps of the integrated
1279: C$^{34}$S J = 2 $\rightarrow$ 1 emission. Contour levels begin at 0.2 K-km/s and increase
1280: in steps of 0.1 K-km/s.
1281: The stars show the positions of IRAS point sources.
1282: }
1283: \label{fig_csdist}
1284: \end{figure}
1285:
1286: \begin{figure}
1287: \plotone{fig5.eps}
1288: \caption{ Plot showing selected spectra at two different positions in the Eastern core:
1289: ($\Delta \alpha = +0\farcm 2$, $\Delta \delta = +0\farcm 7$) with \av\ = 36 mag and
1290: ($\Delta \alpha = -0\farcm 9$, $\Delta \delta = +2\farcm 5$) with \av\ = 19 mag.
1291: }
1292: \label{fig_spec}
1293: \end{figure}
1294:
1295: \begin{figure}
1296: \epsscale{0.8}
1297: \plotone{fig6.eps}
1298: \caption{
1299: C$^{18}$O integrated intensity vs. visual extinction for the
1300: entire Northern Streamer. The best fit linear relationship for $A_V \leq
1301: 10$ mag is given as the solid line, and is extrapolated as the
1302: dash-dot line for $A_V > 10$ mag. The median uncertainties in the
1303: measurements are shown in the lower-right corner. The eastern core,
1304: western core, and mid-streamer are plotted as different symbols as
1305: indicated by the legend.}
1306: \label{fig_c18oint}
1307: \end{figure}
1308:
1309: \begin{figure}
1310: \epsscale{0.8}
1311: \plotone{fig7.eps}
1312: \caption{
1313: Comparison of the C$^{17}$O and C$^{18}$O integrated
1314: intensities in the eastern core. The solid line represents the expected
1315: trend ($\pm 1\sigma$ -- dashed lines) for optically thin emission. (b)
1316: C$^{18}$O--C$^{17}$O ratio of integrated intensities vs. visual
1317: extinction. The solid line represents the expected ratio for optically
1318: thin emission. The dashed lines represent the median $1\sigma$
1319: uncertainty of the measured ratios.}
1320: \label{fig_c17o}
1321: \end{figure}
1322:
1323: \begin{figure}
1324: \epsscale{0.8}
1325: \plotone{fig8.eps}
1326: \caption{
1327: C$^{34}$S optical depth vs. visual extinction for the eastern
1328: and western cores. The eastern core and western core are
1329: plotted as different symbols as indicated by the legend. The {\em circle}
1330: ($\circ$) at $A_V \approx 27$ mag represents the optical depth as derived
1331: from the $^{13}$CS data. The C$^{34}$S optical depth measured for that
1332: same line of sight overlaps the $^{13}$CS derived optical depth.}
1333: \label{fig_cstau}
1334: \end{figure}
1335:
1336: \begin{figure}
1337: \epsscale{0.8}
1338: \plotone{fig9.eps}
1339: \caption{
1340: N$_2$H+ integrated intensity vs. visual extinction for the
1341: entire Northern Streamer. The best fit linear relationship for $A_V < 15$
1342: mag is plotted as the dashed line, and the best fit linear relationship
1343: for $A_V > 15$ mag is plotted as the solid line. The median uncertainties
1344: in the measurements are shown in the lower-right corner. The eastern
1345: core, western core, and mid-streamer are plotted as different symbols as
1346: indicated by the legend.}
1347: \label{fig_n2hpint}
1348: \end{figure}
1349:
1350: \begin{figure}
1351: %\epsscale{0.8}
1352: \plotone{fig10.eps}
1353: \caption{
1354: Volume density of molecular hydrogen as a function of visual extinction and radius. This
1355: profile is in good agreement with the profile of extinction with cloud radius determined in
1356: \citet{lal99}. The horizontal hash marks found along the line representing the density
1357: profile are the average densities used in the molecular abundance analysis for respective
1358: bins of 5 mag starting with \av\ = 5 -- 10 mag.
1359: }
1360: \label{fig_denpro}
1361: \end{figure}
1362:
1363: \begin{figure}
1364: %\epsscale{0.8}
1365: \plotone{fig11.eps}
1366: \caption{
1367: Normalized molecular abundances derived from combined molecular column densities and NIR
1368: extinction measurements of C$^{17}$O (circles and solid line), \CtfS\ (squares and
1369: dashed line), and
1370: \NtwoHp\ (triangles and dotted line) relative to H$_2$ shown as a function of extinction in
1371: the visual.
1372: The data are weighted averages
1373: within bins of 5 mag, thus the errors are reflective of the total number of
1374: data points included in the average and of the percentage error of the
1375: individual data points. The abundances are normalized using the data in the first bin
1376: (A$_V$ = 5 -- 10 mag) in order to see the trend in relative abundances as a function
1377: of extinction.
1378: }
1379: \label{fig_abun}
1380: \end{figure}
1381:
1382: \begin{figure}
1383: \epsscale{0.8}
1384: \plotone{fig12.eps}
1385: \caption{
1386: C$^{34}$S integrated intensity vs. visual extinction for the
1387: eastern and western cores. The best fit linear relationship for $A_V \leq
1388: 15$ mag is overplotted as a solid line, and is extrapolated as the
1389: dash-dot line for $A_V > 15$ mag. The eastern core, western core, and
1390: mid-streamer are plotted as different symbols as indicated by the legend.
1391: The {\em circle} ($\circ$) at $A_V \approx 27$ mag represents the
1392: C$^{34}$S integrated intensity inferred from the measured $^{13}$CS
1393: integrated intensity. The C$^{34}$S integrated intensity measured for
1394: that same line of sight overlaps the $^{13}$CS derived integrated
1395: intensity.
1396: }
1397: \label{fig_c34sint}
1398: \end{figure}
1399:
1400: \begin{figure}
1401: %\epsscale{0.8}
1402: \plotone{fig13.eps}
1403: \caption{
1404: Profile of gas-phase abundances (abundances are relative to H$_2$) against
1405: optical depth for surveyed species and \Ntwo . The top panels are a
1406: pure gas-phase chemical model, while the lower panels are a gas-grain chemical
1407: model. Similarly the left panels are for time (t) $= 10^5$ yr and
1408: the right panels t $= 10^7$ yr. The physical conditions of the model
1409: are given in \S 5.2.
1410: }
1411: \label{fig_chemmod}
1412: \end{figure}
1413:
1414: %*********
1415: %Table 3
1416: %*********
1417: \begin{deluxetable}{lccccc}
1418: %\tabletypesize{\footnotesize}
1419: %\rotate
1420: \tablewidth{0pt}
1421: \tablecolumns{6}
1422: \tablecaption{Summary of Molecular Line Observations}
1423: \tablehead{
1424:
1425: & \colhead{} & \colhead{Observation} &
1426: \colhead{Mapping} & \colhead{Total $T_{int}$} & $T_{sys}$\\
1427:
1428: \colhead{Map Region} & \colhead{Molecule} & \colhead{Date}
1429: & \colhead{Mode} & \colhead{[sec]} & \colhead{[K]}
1430:
1431: }
1432: \startdata
1433: \underline{Full Map}\tablenotemark{a} & C$^{18}$O & 1999 Mar 23 & PS\tablenotemark{b} & 450 & 300\\
1434: & CS & 1999 Mar 23 & PS\tablenotemark{b} & 300 & 140\\
1435: & N$_2$H+ & 1999 Mar 23, 25 & PS\tablenotemark{b} & 900\tablenotemark{c} & 170\\
1436:
1437: \underline{East Core}\\
1438: Full-Beam Footprint & C$^{34}$S & 1999 Mar 25 & PS\tablenotemark{b} & 1800 & 145\\
1439: Single Array Pt.\tablenotemark{d} & $^{13}$CS & 1999 Mar 25 & PS\tablenotemark{b} & 2700 & 170\\
1440: 2 Single Array Pt.\tablenotemark{e}& C$^{18}$O & 2000 Jan 30 & FS & 600 & 180\\
1441: 2 Single Array Pt.\tablenotemark{e}& C$^{17}$O & 2000 Jan 30 & FS & 3600 & 300\\
1442: Full-Beam Footprint\tablenotemark{f}& C$^{18}$O & 2000 Feb 01 & FS & 1200 & 180\\
1443: \underline{West Core}\\
1444: Full-Beam Footprint & C$^{34}$S & 2000 Jan 30 & FS & 2400 & 160\\
1445:
1446: \enddata
1447: \tablenotetext{a}{(0,0) position of map: $\alpha$ = 21:47:27; $\delta$=47:31:00 (J2000).}
1448: \tablenotetext{b}{Reference position: $\alpha$ = 21:48:15; $\delta$=47:21:00 (J2000).}
1449: \tablenotetext{c}{Central two footprints of N$_2$H+ map were observed for
1450: a total integration time of 1500 sec.}
1451: \tablenotetext{d}{Offset from map (0,0) = (0,0).}
1452: \tablenotetext{e}{Offsets from map (0,0): Position 1 = (-2\farcm 69,1\farcm 4);
1453: Position 2 = (-2\farcm 51, 2\farcm 15).}
1454: \tablenotetext{f}{Offset from map (0,0) = (-0\farcm 369, -0\farcm 369).}
1455: \end{deluxetable}
1456: \clearpage
1457: \begin{deluxetable}{rrrrrrr}
1458: \tablenum{2}
1459: \tablecolumns{7}
1460: \tablewidth{5.0in}
1461: \tablecaption{Average Molecular Abundances Relative to H$_2$\tablenotemark{a}}
1462: \tablehead{
1463: \colhead{A$_V$} &
1464: \colhead{C$^{17}$O} &
1465: \colhead{n$_{\rm C^{17}O}$} &
1466: \colhead{C$^{34}$S} &
1467: \colhead{n$_{\rm C^{34}S}$} &
1468: \colhead{N$_2$H$^+$} &
1469: \colhead{n$_{\rm N_2H^+}$} \\
1470: \colhead{(mag)} &
1471: \colhead{($\times 10^{-7}$)} &
1472: \colhead{} &
1473: \colhead{($\times 10^{-11}$)} &
1474: \colhead{} &
1475: \colhead{($\times 10^{-11}$)} &
1476: \colhead{}
1477: }
1478: \startdata
1479: 7.5 & 0.29$\pm$0.08 & 3 & 17.0$\pm$1.7 & 12 & 12.8$\pm$ 1.8 & 8\nl
1480: 12.5 & 0.27$\pm$0.04 & 5 & 9.8$\pm$0.6 & 17 & 6.7$\pm$0.6 & 18 \nl
1481: 17.5 & 0.26$\pm$0.03 & 4 & 7.9$\pm$1.0 & 3 & 4.5$\pm$0.4 & 9 \nl
1482: 22.5 & 0.20$\pm$0.05 & 3 & 6.6$\pm$0.9 & 6 & 6.4$\pm$0.7 & 8 \nl
1483: 27.5 & 0.34$\pm$0.11 & 1 & 6.8$\pm$0.9 & 3 & 11.3$\pm$1.7 & 3 \nl
1484: 37.5 & 0.09$\pm$0.02 & 1 & 5.8$\pm$0.7 & 1 & 8.8$\pm$0.9 & 1 \nl
1485: \enddata
1486: \tablenotetext{a}{n$_{\rm C^{18}O}$, n$_{\rm C^{34}S}$, n$_{\rm N_2H^+}$ are the number
1487: of points included in binned average.}
1488: \end{deluxetable}
1489:
1490: \end{document}
1491:
1492:
1493:
1494:
1495: