1: \documentclass[12pt,preprint]{aastex}
2: \newcommand{\beq}{\begin{equation}}
3: \newcommand{\eeq}{\end{equation}}
4: \newcommand{\e}{$^{-1}$}
5: \newcommand{\calm}{{\cal M}}
6: \newcommand{\ee}{$^{-2}$}
7: \newcommand{\eee}{$^{-3}$}
8: \newcommand{\ellp}{\ell_{P}}
9: \newcommand{\ellpc}{\ell_{\rm pc}}
10: \newcommand{\K}{{\rm K}}
11: \newcommand{\pmax}{P_{\rm max}}
12: \newcommand{\pmin}{P_{\rm min}}
13: \newcommand{\pwnm}{P_{\rm WNM'}}
14: \newcommand{\pth}{P_{\rm th}}
15: \newcommand{\pthave}{P_{\rm th,\,ave}}
16: \newcommand{\sth}{\sigma_{\rm th}}
17: \newcommand{\ag}{\mbox{ \raisebox{-.4ex}{$\stackrel{\textstyle >}{\sim}$} }}
18: \newcommand{\al}{\mbox{ \raisebox{-.4ex}{$\stackrel{\textstyle <}{\sim}$} }}
19: \newcommand{\tmin}{T_{\rm (min)}}
20: \newcommand{\nmin}{n_{\rm min}}
21: \newcommand{\gs}{G_{0}^\prime}
22: \newcommand{\zd}{Z_{\rm d}^\prime}
23: \newcommand{\zg}{Z_{\rm g}^\prime}
24: \newcommand{\zio}{\zeta_{t}^\prime}
25: \newcommand{\phied}{\phi_{{\dot\epsilon}}}
26: \newcommand{\gturb}{\Gamma_{\rm turb}}
27: \newcommand{\muh}{\mu_{\rm H}}
28: %\newcommand{\cicol}{N\mbox{(\ion{C}{1})}}
29: %\newcommand{\ciicol}{N\mbox{(\ion{C}{2})}}
30: %\newcommand{\hicol}{N\mbox{(\ion{H}{1})}}
31: \newcommand{\cicol}{N_{\rm C\,I}}
32: \newcommand{\ciicol}{N_{\rm C\,II}}
33: \newcommand{\hicol}{N_{\rm H\,I}}
34:
35: \slugcomment{Submitted for Publication in the Astrophysical Journal}
36:
37: \shortauthors{Wolfire et al.}
38: \shorttitle{}
39: %\received{}
40: \begin{document}
41:
42: \title{Chemical Rates on Small Grains and PAHs: ${\rm C^+}$ Recombination
43: and ${\rm H_2}$ Formation}
44:
45: \author{
46: Mark\ G. Wolfire}
47: \affil{Department of Astronomy, University of Maryland,
48: College Park, MD 20742-2421}
49: \email{mwolfire@astro.umd.edu}
50:
51: \author{
52: A. G. G. M. Tielens}
53: \affil{Space Science and Astrobiology Division, NASA Ames Research
54: Center,
55: MS 245-3, Moffett Field,
56: CA 94035}
57: \email{atielens@mail.arc.nasa.gov}
58:
59: \author{
60: David Hollenbach}
61: \affil{
62: Space Science and Astrobiology Division, NASA Ames Research
63: Center,
64: MS 245-3, Moffett Field,
65: CA 94035}
66: \email{hollenbach@ism.arc.nasa.gov}
67:
68: \and
69: \author{
70: M. J. Kaufman}
71: \affil{Department of Physics, San Jose State University,
72: One Washington Square, San Jose, CA 95192-0106
73: }
74: \email{mkaufman@email.sjsu.edu}
75:
76: %\newpage
77:
78: \begin{abstract}
79:
80: We use observations of the \ion{C}{1}, \ion{C}{2}, \ion{H}{1}, and
81: ${\rm H_2}$ column densities
82: along lines of sight in the Galactic plane to determine the formation rate
83: of ${\rm H_2}$ on grains and to determine chemical reaction rates
84: with Polycyclic
85: Aromatic Hydrocarbons. Photodissociation region models
86: are used to find the best
87: fit parameters to the observed columns.
88: We find the ${\rm H_2}$ formation rate on grains
89: has a low rate ($R \sim 1\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$)
90: along lines
91: of sight with low column density ($A_{\rm V} \al 0.25$) and low molecular
92: fraction
93: ($f_{\rm H_2} \al 10^{-4}$). At higher column densities
94: ($0.25 \le A_{\rm V} \le 2.13$),
95: we find a rate
96: of $R\sim 3.5 \times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$. The
97: lower rate at low
98: column densities could be the result of grain processing by
99: interstellar shocks which may deplete the grain surface area or
100: process the sites of ${\rm H +H}$ formation, thereby inhibiting
101: ${\rm H_2}$ production.
102: Alternatively, the formation rate may be normal, and the
103: low molecular fraction may be the result of lines of
104: sight which
105: graze larger clouds. Such lines of sight would have a reduced
106: ${\rm H_2}$ self-shielding compared to the line-of-sight column.
107: We find the reaction
108: ${\rm C^+} + {\rm PAH^-} \rightarrow {\rm C} + {\rm PAH^0}$ is best fit
109: with a rate $2.4\times 10^{-7}\Phi_{\rm PAH} T_2^{-0.5}$ ${\rm cm^{3}}$
110: ${\rm s^{-1}}$ with $T_2= T/100$ K and
111: the reaction
112: ${\rm C^+} + {\rm PAH^0} \rightarrow {\rm C} + {\rm PAH^+}$ is best fit
113: with a rate $8.8\times 10^{-9}\Phi_{\rm PAH}$ ${\rm cm^{3}}$
114: ${\rm s^{-1}}$. In high column density gas we find $\Phi_{\rm PAH} \sim 0.4$.
115: In low column density gas, $\Phi_{\rm PAH}$ is less well constrained with
116: $\Phi_{\rm PAH} \sim 0.2 - 0.4$.
117:
118: \end{abstract}
119:
120: \keywords{astrochemistry--ISM: clouds--ISM: general---ISM: molecules}
121:
122: \section{INTRODUCTION}
123: \label{sec:introduction}
124:
125: Much of the chemistry in the interstellar medium (ISM) proceeds via
126: two body gas phase reactions; however,
127: based on reaction time arguments it is well established that
128: the formation of ${\rm H_2}$ must proceed predominantly
129: by grain surface reactions \citep{hollenbach1971} and thus
130: the equilibrium abundance of ${\rm H_2}$ in diffuse gas depends on
131: the balance between dissociation from ultraviolet radiation
132: and the formation on grains.
133: In addition, \cite{lepp1988} suggested that
134: in diffuse clouds,
135: reactions with small grains or large molecules can affect the charge
136: balance and abundance of metal ions. More recent theoretical
137: \citep[e.g.,][]{bakes1998,weingartner2001,wolfire2003,lips2003, chang2006} and
138: laboratory, \citep{cazaux2004} investigations have
139: explored the rates of ${\rm H_2}$ formation and
140: chemical reactions on surfaces of small grains
141: and their dependence on environmental factors.
142: The chemical rates have been difficult to pin down
143: in part because the exact abundance, type, surface area, and physical
144: properties
145: of the grains/large molecules are not well determined.
146:
147:
148: Studies with a more observational approach
149: have also estimated ${\rm H_2}$ formation
150: rates and chemical rates with grains
151: \citep[e.g.,][]{jura1975,browning2003,welty2003}.
152: \cite{jura1975} estimated a formation
153: rate of about $R~\sim 3\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$ in
154: diffuse gas of low column density; however,
155: one could see in the {\em Copernicus} survey of \cite{savage1977}
156: that a single ${\rm H_2}$ formation
157: rate can not explain all of the observed column densities
158: even accounting for the
159: expected variation in ultraviolet radiation field and gas density.
160: A more recent analysis by \cite{gry2002} using data from the {\em Far
161: Ultraviolet Spectroscopic Explorer} (FUSE) find a formation
162: rate of $R\sim 4\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$ in
163: diffuse gas and \cite{browning2003} estimates the rate at 2.5 - 10
164: times lower in the low metallicity environments of the Large and Small
165: Magellanic Clouds.
166: \cite{habart2004} suggests that the ${\rm H_2}$
167: formation rate varies in molecular clouds exposed to intense
168: radiation by a factor of about five, perhaps due to the temperature
169: dependence of the formation process on grain surfaces \citep{cazaux2004}.
170: In this paper we take
171: a semi-empirical approach
172: by allowing the ensemble of recent observations to guide us in the proper
173: ``astrophysical'' rates rather than use rates based on
174: laboratory measurements with an assumed grain type and distribution.
175:
176: \cite{bakes1998} investigated the effects of Polycyclic Aromatic
177: Hydrocarbons (PAHs) on
178: the chemistry of dense regions exposed to intense ultraviolet
179: radiation. They found that charge exchange reactions with PAHs are
180: important in setting the column density of neutral metals, in
181: particular \ion{C}{1}. \cite{welty2003} found
182: patterns of neutral column densities in diffuse clouds higher
183: than suggested by
184: electron recombination alone and suggested that
185: recombination on grains might
186: be important. \cite{weingartner2001} carried out a detailed analysis
187: of the ``grain assisted'' rates required to explain the observed
188: column densities along the line of sight to
189: 23 Ori. \cite{wolfire1995,wolfire2003} found that reactions with PAHs
190: can affect the ionization balance of metals and also atomic hydrogen,
191: in diffuse gas, in particular in the warm neutral medium. This
192: has an important effect on the grain charge and thus the heating
193: produced by the grain photoelectric effect; the dominant heating
194: process in the diffuse ISM.
195:
196: In \cite{wolfire2003} we adopted rates with PAHs based
197: on the \cite{draine1987}
198: formalism for interactions with small grains, but modified
199: the rates to match a preliminary assessment of the
200: observed $\cicol/\ciicol$ ratio in diffuse gas. This ratio is a
201: particularly good
202: diagnostic because carbon is abundant, the rates of
203: ${\rm C^+}$ recombination on PAH anions effectively compete with
204: electron recombination, and carbon exists
205: in the diffuse ISM outside of \ion{H}{2} regions
206: as either \ion{C}{1} or \ion{C}{2}
207: but not higher ionization stages.
208:
209:
210: For several years, the {\em Copernicus} column densities of \ion{H}{1}
211: \citep{bohlin1978},
212: ${\rm H_2}$ \citep{savage1977}, \ion{C}{1} \citep{jenkins1979,jenkins1983},
213: and \ion{C}{2} \citep{hobbs1982}
214: were the only data available.
215: More recently the combined observations from
216: the {\em Hubble Space
217: Telescope} Goddard High Resolution Spectrograph (GHRS) and
218: Space Telescope Imaging Spectrograph (STIS) instruments along with
219: the {\em Far Ultraviolet Spectroscopic Explorer} (FUSE) have produced
220: many more sight lines to study in particular in ${\rm H_2}$
221: \citep[e.g.,][]{rachford2002, cartledge2004}, \ion{C}{1}
222: \citep[e.g.,][]{zsargo1997,zsargo2003}, and \ion{C}{2}
223: \citep[e.g.,][]{sofia2004}.
224: In \S~2 we present the observations
225: used in this study and in \S~3 we discuss our cloud models and our
226: fitting procedure. In \S~4 we present our results
227: for the best fit rates for grain reactions and for ${\rm H_2}$ formation.
228: We also discuss some of the interesting sight lines which seem to
229: have outlying points. We summarize our main conclusions in \S~5.
230:
231: \section{OBSERVATIONS}
232: \label{sec:observations}
233:
234: Observations of the \ion{H}{1}, ${\rm H_2}$, \ion{C}{1}, and
235: \ion{C}{2} columns are listed in Table~\ref{tbl:observations}
236: and are taken from the literature as noted.
237: We have included only those lines of sight which have measured values of
238: $\log \hicol$, $\log N_{\rm H_2}$, and $\log \cicol$, and
239: have not included lines of sight with upper or lower limits in these
240: quantities. The \ion{C}{2} column
241: is measured directly only towards thirteen sources \citep{sofia2004}.
242: Six of these have definite \ion{C}{1} columns and are included in our sample.
243: For all other sources we use the conversion
244: $\ciicol ={\cal A}_{\rm C}\times N$ where
245: $N\equiv \hicol + 2N_{\rm H_2}$, and
246: ${\cal A}_{\rm C}=1.6\times 10^{-4}$ is the mean gas phase abundance of carbon
247: per hydrogen nucleus
248: found by \cite{sofia2004} from all measured values of the \ion{C}{2}
249: carbon abundance in the diffuse ISM.
250:
251: Most of the \ion{H}{1} and ${\rm H_2}$ columns and uncertainties
252: are from the {\em Copernicus} ultraviolet absorption
253: spectroscopy observations of \cite{bohlin1978}
254: and \cite{savage1977}. More recently,
255: \cite{cartledge2004} updated
256: several of these lines of sight with FUSE observations.
257: The columns are in good
258: agreement with the previous results, but with smaller quoted
259: uncertainties. In cases of overlap we use the \cite{cartledge2004}
260: results. Additional ${\rm H_2}$ columns are provided by
261: FUSE surveys \citep[e.g.,][]{rachford2002}.
262: In several low column density lines of sight \cite{savage1977} did
263: not quote an uncertainty for the ${\rm H_2}$ columns.
264: %The uncertainty
265: %in the \ion{C}{1} columns are not quoted in a few sources as well.
266: In these cases we use the uncertainty $\sigma(\log N_{\rm H_2}) = 0.20$
267: as suggested
268: by D.\ Welty\footnote{http://astro.uchicago.edu/home/web/welty/coldens.html}
269: and is comparable to the maximum quoted value for all observations.
270:
271: Many of the $\cicol$ columns are from the {\em Copernicus} ultraviolet
272: absorption observations
273: of \cite{jenkins1979} and \cite{jenkins1983}. Additional \ion{C}{1} columns
274: are provided for example, by GHRS observations and FUSE
275: \citep{zsargo1997,zsargo2003,jt2001}.
276: The method used by \cite{jt2001} to determine \ion{C}{1} columns
277: does not yield a direct measure of the random errors in the observed
278: velocity integrated column density. We adopt an uncertainty comparable to
279: similar STIS and {\em FUSE}
280: observations at similar column densities \citep[e.g.,][]{sonn2002}.
281:
282: \section{MODELS}
283: \label{sec:models}
284:
285: \subsection{Reaction Rates with PAHs and Carbon Chemistry}
286: \label{subsec:PAH rates}
287:
288: The PAH rates in \cite{wolfire2003} were calculated using the equations
289: given by \cite{draine1987} for a disk shaped PAH with the number of Carbon
290: atoms $N_{\rm C} = 35$ and a PAH abundance of $6\times 10^{-7}$ per
291: hydrogen nucleus. \cite{wolfire2003} found it necessary to modify the rates
292: by a factor $\Phi_{\rm PAH}$ to match a preliminary assessment of the
293: observed $\cicol/\ciicol$ ratio in the diffuse ISM. The factor
294: $\Phi_{\rm PAH}$ is
295: to include a wide range of unknowns in the rates including
296: the PAH size, geometry, and abundance. Although the full \cite{draine1987}
297: formalism is used in our code, over typical temperatures and densities
298: found in the diffuse ISM the rates can be closely approximated by
299: simple formulae. For ${\rm C^+}$ recombination on PAHs we have
300: \beq
301: {\rm C^+ + PAH^-} \rightarrow {\rm C + PAH^{0}}, \,\,\,\,
302: k_1 = 2.4\times 10^{-7} \Phi_{\rm PAH} T_2^{-0.5}\,\,\,
303: {\rm cm^3\,\, s^{-1}}\, ,
304: \label{eq:cprecompahm}
305: \eeq
306: where $T_2 \equiv T/(100\,\,{\rm K})$,
307: and for charge exchange
308: \beq
309: {\rm C^+ + PAH^0} \rightarrow {\rm C + PAH^{+}}, \,\,\,\,
310: k_2 = 8.8\times 10^{-9} \Phi_{\rm PAH}\,\,\,
311: {\rm cm^3\,\, s^{-1}}\, .
312: \label{eq:cprecompahn}
313: \eeq
314: Several rates were given
315: in \cite{wolfire2003} Appendix C2, mainly related to the charge balance for
316: hydrogen and the electron abundance. We provide a summary of our
317: rate set in Appendix \ref{appen:pahrates}.
318:
319: In addition to reactions with PAHs, ${\rm C^+}$ can also recombine with
320: electrons in the gas phase via radiative or dielectronic recombination.
321: It has recently come to our attention that dielectronic recombination
322: of ${\rm C^+}$ can be significant in diffuse gas
323: (G. Ferland, private communication).
324: The total (dielectronic plus radiative) rates by \cite{altun2004} and
325: those posted on-line for lower temperatures
326: by Badnell ( http://amdpp.phys.strath.ac.uk/tamoc/DATA/) are higher
327: by a factor of $\sim 2$ at 100 K than rates by \cite{nahar1997}.
328: For the majority of this paper we use the Badnell rates, but discuss
329: the implications if they are lower as suggested by the previous data.
330: A fit to the total rate is given by
331: \beq
332: {\rm C^+}\,\, + e \rightarrow {\rm C^0} \,\,\,\, k_3 = 1.8\times 10^{-11}
333: T_2^{-0.83} \,\,\, {\rm cm^3\,\, s^{-1}}\, ,
334: \label{eq:cprecome}
335: \eeq
336: where the fit is good to within 8\% for 20 K $ < T < 300$ K.
337:
338: The ${\rm C^+}$ is formed mainly by the photoionization of C by the
339: interstellar radiation field
340: \beq
341: {\rm C\,\, + }\, h\nu \rightarrow {\rm C^+\,\, +}\,\, e, \,\,\,\,
342: k_4 = 2.1\times 10^{-10} G_0 \exp (-2.6A_{\rm V})\,\,\, {\rm s^{-1}}
343: \label{eq:c1ionization}
344: \eeq
345: where $G_0$
346: is the FUV (6 eV $\le h\nu \le 13.6$ eV) interstellar radiation
347: field measured in units of the \cite{habing1968} field
348: ($G_0 = 1$ is $1.3\times 10^{-4}$ erg ${\rm cm^{-2}}$ ${\rm s^{-1}}$
349: ${\rm sr^{-1}}$).
350: In \cite{wolfire2003}, we adopted $G_0=1.7$, a field strength
351: comparable to the
352: \cite{draine1978} interstellar field. Balancing formation and
353: destruction processes, in clouds optically thin to the FUV radiation,
354: the $\cicol/\ciicol$ ratio is mainly a function of $n\Phi_{\rm PAH}/G_0$.
355:
356: \subsection{${\rm H_2}$ Formation}
357: \label{subsec:H2formation}
358:
359: Molecular hydrogen formation in diffuse gas proceeds via reactions with
360: atomic hydrogen on grain surfaces. We adopt a formation rate per unit
361: volume which
362: goes as $n n_{\rm H\,I} R$ where $n$ is the hydrogen nucleus density,
363: $n_{\rm H\, I}$ is the density of atomic hydrogen, and $R$ is
364: a constant of order $3\times 10^{-17}$ ${\rm cm^{3}}$
365: ${\rm s^{-1}}$. The constant includes the surface area
366: of grains per hydrogen nucleus. (If the metallicity were to vary
367: as $Z$, and the grain surface area per H proportional to $Z$, then the rate
368: $R$ would go as $Z$.) Previous investigators have included both
369: a gas and grain temperature
370: dependence in this rate equation intended to include the collision
371: rate of atomic hydrogen with grains,
372: the sticking coefficient
373: of hydrogen on grains (which is a function of both the hydrogen
374: temperature and grain temperature), and the thermal diffusion
375: of atoms across grain surfaces and thermal evaporation from these
376: surfaces (which is a function of the grain temperature).
377: The collision rate increases with
378: temperature while the sticking coefficient and formation efficiency
379: decrease with temperature \citep{hollenbach1971,burke1983,pirronello1999,
380: cazaux2004}. In light
381: of the uncertainties of the details of the various processes,
382: \cite{kaufman1999} adopted an average rate which was independent of
383: temperature.
384:
385: The dissociation rate of ${\rm H_2}$ per unit volume goes as
386: \beq
387: G_0 I n_{\rm H_2} \beta (N_{\rm H_2})\exp(-2.5A_{\rm V})
388: \label{eq:h2dissociation}
389: \eeq
390: where $I$ is the unshielded
391: photodissociation rate in the local interstellar field ($G_0=1$),
392: and $\beta (N_{\rm H_2})$ is the ${\rm H_2}$ self-shielding
393: factor. We take $I = 4.7\times 10^{-11}$ ${\rm s^{-1}}$
394: from \cite{abgrall1992}.
395: The unshielded rate could change depending on the
396: detailed population of the molecular hydrogen rotational and
397: vibrational levels resulting from different
398: incident FUV fields,
399: %gas column densities (FUV backscattering),
400: gas densities, and temperatures.
401: We have compared our constant unshielded dissociation
402: rate $I$, with results produced with
403: the code available on-line from the
404: Meudon Group\footnote{See http://aristote.obspm.fr/MIS}.
405: As discussed in \cite{lepetit2006} they include a more detailed
406: treatment of ${\rm H_2}$ dissociation and shielding.
407: We ran their
408: code at $n=30$ and 100 ${\rm cm^{-3}}$, $G_0 = 1.7$,
409: and $A_{\rm V} = 0.1$, 0.3, and 1, and found good agreement
410: with less than a 10\% difference in values of $I$.
411: We use the self-shielding factor from
412: \cite{draine1996} (their equation 37) which is a fit to their detailed
413: ${\rm H_2}$ destruction calculation. The equilibrium
414: molecular fraction $2n_{\rm H_2}/(n_{\rm H\,I} + 2n_{\rm H_2})$,
415: for unshielded ${\rm H_2}$ ($\beta = 1$) equals
416: $2nR/G_0I$. For constant $I$, the molecular fraction is proportional
417: to $2nR/G_0$. Thus, to first order and ignoring shielding by dust
418: or ${\rm H_2}$, the
419: $\cicol/\ciicol$ ratio depends on $\Phi_{\rm PAH}n/G_0$ while the
420: $N_{\rm H_2}/\hicol$ ratio depends on $Rn/G_0$.
421:
422: \subsection{2-Sided Models}
423: \label{subsec:2sidedmodels}
424:
425: We see from sections \ref{subsec:PAH rates} and
426: \ref{subsec:H2formation} that the $\cicol/\ciicol$ ratio
427: and the ${\rm H_2}$ column $N_{\rm H_2}$ depends on
428: $\Phi_{\rm PAH}$, $R$, and the ratio of $G_0/n$.
429: The total column density $N= \hicol + 2N_{\rm H_2}$ also enters
430: due to the extinction of the incident FUV field
431: in equations (\ref{eq:c1ionization}) and (\ref{eq:h2dissociation})
432: and also in the self-shielding
433: factor $\beta$ (which depends on the ${\rm H_2}$ column density).
434: The total column density is known for each source. Thus we
435: can construct models for each source by holding the total column
436: density fixed, and varying $G_0/n$, $\Phi_{\rm PAH}$, and $R$, until
437: good fits are obtained to the observed columns densities,
438: $N_{\rm C\,I}/N_{\rm C\,II}$ and $2N_{\rm H_2}/N$
439:
440: Since many of the observations have cloud columns $A_{\rm V}\al 1$ the effects
441: of radiation incident on the near and far side of the cloud need to
442: be included. This radiation affects both the photoionization of C and
443: the dissociation of ${\rm H_2}$. We have modified the 1-sided
444: Photodissociation Region (PDR) code
445: of \cite{kaufman1999} to include the 2-sided incident radiation.
446: The code calculates the equilibrium chemical abundances and gas
447: temperature in a gas layer exposed to X-rays and FUV radiation.
448: In the one-sided case, calculation of the chemistry, cooling, and line
449: transfer proceeds from the surface to the cloud center in a single pass
450: since these parameters depend only on the cloud properties closer to the
451: surface.
452: %MGW I have lifted some of these lines from the Neufeld paper which
453: %Michael wrote to explain the 2-sided model.
454: In the two-sided case, at a given point in the cloud,
455: the optical depth in the cooling lines and the ${\rm H_2}$ column
456: is required towards both surfaces, thus an iterative procedure is required.
457: We first calculate the structure of a cloud illuminated from one side
458: to a depth of one-half the visual extinction of the desired two-sided model.
459: This provides us with initial estimates of line optical depths as well
460: as shielding columns of ${\rm H_2}$ and CO. We then iterate the
461: chemical abundance profiles, optical depths, and shielding factors.
462:
463: We have further modified the physics and chemistry of the
464: \cite{kaufman1999} code according to the discussion in
465: \cite{wolfire2003} who assumed a higher abundance of PAHs and
466: a slower rate of interaction of ions and PAHs.
467: These changes largely offset each other, with a resulting minor effect
468: on the grain heating rates and ion chemistry. In addition, we have adopted
469: the results of \cite{Pequignot1990} for the rate coefficient for collisional
470: excitation of O I by H, and those of \cite{mccall2003} for the ${\rm H_3^+}$
471: dissociative recombination rate coefficient.
472: We have also included the illumination of the cloud by the interstellar
473: soft X-ray
474: radiation field as discussed in \cite{wolfire2003}.
475: Results from this 2-sided code have been previously presented in
476: \cite{neufeld2005}, and \cite{snow2006}.
477:
478:
479: We have compared our calculated ${\rm H_2}$ column densities with
480: those produced by the Meudon code. For the same model parameters in
481: section \ref{subsec:H2formation}, and running the code in 2-sided mode,
482: we find that our ${\rm H_2}$ columns
483: agree to within a factor of 2 (with our columns lower)
484: and with the largest difference at low column densities
485: ($A_{\rm V}\sim 0.1$). The agreement is within the typical
486: observational error.
487:
488: \subsection{Fitting $\Phi_{\rm PAH}$ and $R$}
489:
490: We determine the best fit $\Phi_{\rm PAH}$ and $R$ by computing
491: a grid of models consisting of 12 values of $\Phi_{\rm PAH}$
492: (0.01, 0.1, 0.2 ,0.3,
493: 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.1)
494: and 11 values of $R$ (0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9,
495: 10) $\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$. At each grid point,
496: (i.e., for fixed $\Phi_{\rm PAH}$ and $R$), we
497: find the best fit $G_0/n$ values for each source by minimizing the
498: $\chi^2_s$
499: \beq
500: \chi^2_s(\Phi_{\rm PAH}, R) =
501: \left \{ \frac{\log f_{\rm H_2}^{\rm model} - \log f_{\rm H_2}}
502: { \sigma[\log f_{\rm H_2} ]} \right \}^2 +
503: \left \{ \frac{\log f_{\rm C\,I}^{\rm model} - \log f_{\rm C\,I}}
504: {\sigma [ \log f_{\rm C\,I} ]} \right \}^2 ,
505: \eeq
506: where $\chi^2_s$ is the $\chi^2$ value for source $s$,
507: $f_{\rm H_2}$ and $f_{\rm C\,I}$ are the
508: observed values for the column density ratios
509: $f_{\rm H_2} = 2N_{\rm H_2}/[\hicol + 2N_{\rm H_2}]$,
510: $f_{\rm C\,I} = {\cicol/\ciicol}$,
511: $f_{\rm H_2}^{\rm model}$, and $f_{\rm C\,I}^{\rm model}$ are the
512: calculated values of $f_{\rm H_2}$ and $f_{\rm C\,I}$ (dependent
513: upon $N$ -- which is measured, and on the unknowns
514: $\Phi_{\rm PAH}$, $R$, and $G_0/n$).
515:
516:
517: We will calculate the $\chi^2$ value over a subset of the sources
518: within a range of $A_{\rm V}$ values. Thus
519: the reduced $\chi^2$ value at grid point ($\Phi_{\rm PAH}, R)$ is the sum
520: over sources within the restricted $A_{\rm V}$ bin:
521: \beq
522: \chi^2(\Phi_{\rm PAH}, R) = \sum_s^{\rm N_s} \chi^2_s/(N_s-2) .
523: \eeq
524: where $N_s$ is the total number of sources in the bin.
525:
526:
527: We find the uncertainty in the measurements
528: $\sigma (\log f_{\rm H_2})$ and $\sigma (\log f_{\rm C\,I})$
529: using the quoted uncertainties
530: $\sigma(\log \hicol)$, $\sigma(\log N_{\rm H_2})$, $\sigma(\log \cicol)$,
531: for the \ion{H}{1}, ${\rm H_2}$, and \ion{C}{1} columns, along with
532: the standard propagation of error formula \citep[e.g.,][]{taylor1997}.
533: Thus
534: \beq
535: \sigma({\log f_{\rm H_2}}) = \left [ \sigma( \log N_{\rm H_2})^2 +
536: \sigma( \log N)^2 \right ]^{1/2},
537: \label{eq:sigmafh2}
538: \eeq
539: where $\sigma( \log N)$ is the uncertainty in the total column density,
540: \beq
541: \sigma(\log N) = \left [ \sigma( \log N_{\rm H_2})^2 f_{\rm H_2}^2+
542: \sigma( \log \hicol)^2 f_{{\rm H\,I}}^2 \right ]^{1/2},
543: \label{eq:sigman}
544: \eeq
545: and $f_{\rm H\,I}$ is the fraction of atomic hydrogen
546: $f_{\rm H\,I}=\hicol /[\hicol + 2N_{\rm H_2}]$. For observed \ion{C}{2}
547: column densities,
548: the uncertainty in $\log f_{\rm C\,I}$ is given by
549: \beq
550: \sigma( \log f_{\rm C\,I} ) = \left [\sigma (\log \cicol)^2 +
551: \sigma (\log \ciicol)^2 \right ]^{1/2},
552: \label{eq:sigmacicii1}
553: \eeq
554: while for estimated \ion{C}{2} column densities the uncertainty in
555: $\log f_{\rm C\,I}$ is given by
556: \beq
557: \sigma( \log f_{\rm C\,I} ) = \left [ \sigma (\log \cicol)^2 +
558: \sigma (\log N)^2 \right ]^{1/2}.
559: \label{eq:sigmacicii2}
560: \eeq
561:
562: In practice, for the $\chi^2$ test, we set $G_0 = 1.7$ and vary $n$.
563: We estimate the initial range of $n$ values
564: from the analytic expressions in Appendix
565: \ref{appen:estimatinggn}. We run models
566: stepping in values of $n$ until we cross the minimum in
567: $\chi^2$. Then we apply a parabolic interpolation to find
568: the minimum $\chi^2$ and associated values of density and
569: column densities.
570:
571: \section{RESULTS and DISCUSSION}
572: \label{sec:resultsanddiscussion}
573:
574:
575: \subsection{Fixed Column Density Models}
576: \label{subsec:fixedrates}
577:
578: We first carry out model runs for four fixed cloud column
579: densities $A_{\rm V} = 0.1$, 0.5, 1.0, and 1.5 with
580: variable $\Phi_{\rm PAH}$, $R$, and $G_0/n$.
581: The data and model results are displayed in plots
582: of $\cicol/\ciicol$ versus $f_{\rm H_2}$ with source column
583: densities grouped in four $A_{\rm V}$
584: bins. These plots are intended to show the range of
585: observed values, the effects of varying $\Phi_{\rm PAH}$
586: and $R$,
587: and an ``eye ball'' assessment of reasonable values.
588: The $A_{\rm V}$ bins
589: span the range of observed $A_{\rm V}$ values with the
590: smallest being $A_{\rm V} = 0.03$ and the
591: largest $A_{\rm V} = 2.13$. The $A_{\rm V}$ bins have been chosen
592: to display natural groupings of the data which we further test
593: using the $\chi^2$ results.
594:
595: We present in Figure \ref{fig:ratioav025}
596: the observed and calculated
597: ratio of $\cicol/\ciicol$ versus $f_{\rm H_2}$ for observations
598: in the range
599: $0.03 \al A_{\rm V} \al 0.25$ and $N_{\rm H_2} \al 10^{17}$
600: cm$^{-2}$.
601: Figure \ref{fig:ratioav025}
602: focuses on the low $f_{\rm H_2}$ ($\al 10^{-4}$)
603: and low ${A_{\rm V}}$ directions.
604: The curves are for a model cloud column density of
605: $A_{\rm V} = 0.1$,
606: with $G_0 = 1.7$ and 5.1, and density points at
607: $n = 10$, 20, 30, and 40 ${\rm cm^{-3}}$ (with the
608: exception of two curves with three density points at
609: $n = 5$, 10, and 20 ${\rm cm^{-3}}$).
610: Diamonds ($\Diamond$) indicate models with
611: $G_0 = 1.7$ and triangles ($\triangle$) indicate models
612: with $G_0 = 5.1$.
613: We have labeled the model curves with their corresponding
614: values of $\Phi_{\rm PAH}/R_{-17}$ where $R_{-17}=R/1\times 10^{-17}$.
615: The curves with
616: $\Phi_{\rm PAH}/R_{-17} \le 1/3$ use $R = 3\times 10^{-17}$ ${\rm cm^{3}}$
617: ${\rm s^{-1}}$ with values of $\Phi_{\rm PAH}=1$
618: ($\Phi_{\rm PAH}/R_{-17} = 1/3$),
619: $\Phi_{\rm PAH}=0.5$ ($\Phi_{\rm PAH}/R_{-17} = 1/6$), and
620: $\Phi_{\rm PAH}=0.01$ ($\Phi_{\rm PAH}/R_{-17} = 1/300$). We have
621: added three additional curves with lower values of $R$ namely
622: $R = 2\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$, $\Phi_{\rm PAH} = 1$
623: ($\Phi_{\rm PAH}/R_{-17} = 1/2$);
624: $R = 1\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$, $\Phi_{\rm PAH} = 0.5$
625: ($\Phi_{\rm PAH}/R_{-17} = 1/2$); and
626: $R = 1\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$, $\Phi_{\rm PAH} = 1.0$
627: ($\Phi_{\rm PAH}/R_{-17} = 1$)
628: and one additional curve with higher values
629: of $\Phi_{\rm PAH}$ namely $R = 2\times 10^{-17}$ ${\rm cm^{3}}$
630: ${\rm s^{-1}}$, $\Phi_{\rm PAH} = 2.0$ ($\Phi_{\rm PAH}/R_{-17} = 1$).
631:
632: The curves
633: with the same value of $\Phi_{\rm PAH}/R_{-17}$, although not exactly the same,
634: differ by less than a factor of 1.5.
635: We see that values of the ratio $\Phi_{\rm PAH}/R_{-17} \approx 1/2-1/3$
636: are required to match the bulk of the observations.
637: For $\Phi_{\rm PAH} \le 1$ this
638: requires $R\le 2 \times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$ and
639: an ${\rm H_2}$ formation rate smaller than the standard
640: $3 \times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$ is necessary.
641:
642: %In the limit of no FUV dust extinction and no ${\rm H_2}$
643: %self-shielding, the ratio $\cicol/\ciicol \propto n\Phi_{\rm PAH}/G_0$,
644: %and $f_{\rm H_2}\propto 2nR/G_0$, and thus
645: %$\cicol/\ciicol \propto (\Phi_{\rm PAH}/R) f_{\rm H_2}$. %
646:
647:
648: The subsequent Figures (\ref{fig:ratioav075}, \ref{fig:ratioav125},
649: and \ref{fig:ratioav213}) focus on the high molecular
650: abundance lines of sight ($f_{\rm H_2} > 10^{-3}$ and
651: $N_{\rm H_2} > 10^{18}$ cm$^{-2}$).
652: The observations are limited to the range
653: $0.25 \al A_{\rm V} \al 0.75$ (Fig.\ [\ref{fig:ratioav075}]),
654: $0.75 \al A_{\rm V} \al 1.25$ (Fig.\ [\ref{fig:ratioav125}]), and
655: $1.25 \al A_{\rm V} \al 2.13$ (Fig.\ [\ref{fig:ratioav213}])
656: and model curves are for $A_{\rm V} = 0.5$ (Fig.\ [\ref{fig:ratioav075}]),
657: $A_{\rm V} = 1.0$ (Fig.\ [\ref{fig:ratioav125}]), and
658: $A_{\rm V} = 1.5$ (Fig.\ [\ref{fig:ratioav213}]).
659: The models all
660: use $R=3\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
661: with values of ${\Phi_{\rm PAH}} = 0.01$, 0.5, and 1.0,
662: and density, $n = 10$, 30, 100, and
663: 300 ${\rm cm^{-3}}$.
664: Comparing Figures \ref{fig:ratioav025} and \ref{fig:ratioav075}
665: we see a large jump in the molecular fraction with
666: no observations at intermediate values. This jump is
667: probably the result of the abrupt turn on of
668: molecular hydrogen due to the effects of self-shielding.
669: Self-shielding of ${\rm H_2}$ at $N \ge 5\times 10^{20}$
670: ${\rm cm^{-2}}$ ($A_{\rm V} \ge 0.25$)
671: was seen in the \cite{savage1977} data
672: and also in the more recent FUSE surveys \citep{gillmon2006}.
673: In Figure \ref{fig:ratioav075},
674: two obvious outliers are 23 Ori ($A_{\rm V} = 0.28$)
675: at high $\cicol/\ciicol$ ratio ($= 1.1\times 10^{-2}$)
676: and $\pi$ Sco ($A_{\rm V} = 0.28$) at low
677: $\cicol/\ciicol$ ratio ($=1.2\times 10^{-3}$).
678: These are excluded from the $\chi^2$ tests and
679: discussed further in subsection~\ref{subsec:highav}.
680:
681: %The ${\rm H_2}$ fraction in unshielded gas is given by:
682: %\beq
683: % f_{\rm H_2} = \frac{2nR}{G_0 I}\,\, .
684: %\label{eq:h2fracmin}
685: %\eeq
686: %For typical conditions in diffuse gas ($n \sim 30$ ${\rm cm^{-3}}$,
687: %$G_0 = 1.7$, $R=3\times 10^{-17}$ ${\rm s^{-1}}$,
688: %$I = 4.7\times 10^{-11}$
689: %${\rm s^{-1}}$), $f_{\rm H_2} = 2.3\times 10^{-5}$ for
690: %$R=3\times 10^{-17}$ ${\rm s^{-1}}$ and
691: %$f_{\rm H_2} = 7.5\times 10^{-6}$ for
692: %$R=1\times 10^{-17}$ ${\rm s^{-1}}$.
693: %We see from Figure \ref{fig:ratioav025} that several low $f_{\rm H_2}$
694: %observations can not be fit with $R=3\times 10^{-17}$ ${\rm s^{-1}}$, and
695: %lower $R$ values are required. I think we are dropping this paragraph.
696:
697: The variation of the curves
698: shown in Figures \ref{fig:ratioav025}
699: through \ref{fig:ratioav213} as functions of $G_0/n$, $\phi_{\rm PAH}$,
700: and $R$
701: is discussed in Appendix \ref{appen:estimatinggn}.
702: Points with the same $G_0/(nR)$ and fixed column, or $A_{\rm V}$,
703: have the same molecular fraction $f_{\rm H_2}$ independent of
704: $\Phi_{\rm PAH}$. At the highest values of $G_0/n$ the molecular
705: fraction is lowest due to rapid photodissociation of molecular hydrogen.
706: In the limit of no ${\rm H_2}$ self-shielding and no dust extinction,
707: the limiting molecular
708: fraction is given by $f_{\rm H_2} = {2nR}/({G_0 I})$. As $G_0/n$ decreases,
709: the molecular fraction increases with
710: a limiting value of $f_{\rm H_2} = 1$ for fully molecular gas.
711: As the cloud column density (and thus $A_{\rm V}$) increases the local
712: FUV field drops due to dust extinction and the molecular fraction
713: rises due to a lower ${\rm H_2}$ photodissociation rate. In addition,
714: the self-shielding by ${\rm H_2}$ raises the molecular fraction as the
715: column density increases.
716:
717: The variation in the $\cicol/\ciicol$ ratio can be understood
718: in four limiting regimes: high/low $\Phi_{\rm PAH}$, and high/low $G_0/n$.
719: First consider the high $\Phi_{\rm PAH}$ ($0.5$ and $1.0$) curves.
720: For low $G_0/n$ the destruction of ${\rm C^+}$ is dominated
721: by recombination on ${\rm PAH^-}$ and
722: $\cicol/\ciicol \propto \Phi_{\rm PAH}^2 (n/G_0)^2$
723: (See Appendix \ref{appen:variation}, eq.~[\ref{eq:Clowgn}]).
724: For high
725: $G_0/n$, the destruction of ${\rm C^+}$ is dominated by charge
726: exchange with ${\rm PAH^0}$ and electron recombination and thus
727: $\cicol/\ciicol \propto \Phi_{\rm PAH} n/G_0$ (eq.~[\ref{eq:Chighgn}]).
728: In the case of low $\Phi_{\rm PAH}=0.01$ the
729: ${\rm C^+}$ destruction is dominated by recombination with
730: free electrons $n_e$. For low $G_0/n$ the electrons are provided by
731: photoionization of
732: C, and $\cicol/\ciicol \propto n/G_0$ (eq.~[\ref{eq:Clowgnlowphi}]).
733: For high $G_0/n$, i.e., low values of $n$, the electrons are provided
734: by the soft-X ray ionization of H, and
735: $\cicol/\ciicol \propto n^{1/2}/G_0$ (eq.~[\ref{eq:Chighgnlowphi}]).
736: In the molecular hydrogen case, the molecular fraction increases with
737: column density due to the drop in FUV field and larger ${\rm H_2}$
738: self-shielding.
739: However, in the case of
740: the $\cicol/\ciicol$ ratio, it is just the dust extinction that leads
741: to the increase in the ratio with column.
742:
743: %At high $n$ $e-$ come from C+ rather than H+. Increasing $\Phi_{\rm PAH}$
744: %therefore increases CI. At low $n$, $e-$ come from H+, increasing
745: %$\phi_{\rm PAH}$ increases the rate of electron recombination on H+
746: %and thus decreasing the electron abundance and thus decreasing CI.
747:
748: \subsection{$\chi^2$ Results}
749: \label{subsec:chisquareresults}
750:
751: We next present the $\chi^2$ results obtained by binning the
752: sources in four $A_{\rm V}$ bins. But note that for each source
753: model, we fix $ A_{\rm V}$ to the observed column.
754: % shown previously in Figures \ref{fig:ratioav025}
755: %through \ref{fig:ratioav213}.
756: Figure \ref{fig:chiav025} shows the results for
757: sources in the range $0.03 \al A_{\rm V} \al 0.25$ and
758: $N_{\rm H_2} < 10^{17}$ cm$^{-2}$.
759: %(including ten sources and excluding
760: %$\epsilon$ Per
761: %and 59 Cyg).
762: There appears to be two local minima: one at
763: low $R$ values with
764: $R = 1\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
765: and $0.2 \al \Phi_{\rm PAH} \al 0.4$ and one at
766: $R\approx 2\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
767: and $\Phi_{\rm PAH}\ag 1$. The $\cicol/\ciicol$ ratios
768: in this $A_{\rm V}$ range are $\al 10^{-3}$ and both
769: equations (\ref{eq:Agnpahm}) and (\ref{eq:Agnpahn}) apply.
770: At low $\Phi_{\rm PAH}$ the ${\rm C^+}$ destruction by
771: electron recombinations dominate (right hand side of eq.\
772: \ref{eq:Agnpahn})
773: and the dependence on
774: $\Phi_{\rm PAH}$ drops out. Moving to the upper right in
775: the $\chi^2$ plot corresponds to lower densities and higher
776: temperatures. There, electron recombination is less effective
777: in producing \ion{C}{1}, the \ion{C}{1} abundance drops,
778: and the fits become poorer. At sufficiently
779: high $\Phi_{\rm PAH}$, the PAH reactions dominate, the
780: \ion{C}{1} abundance rises, and the fits improve. We shall demonstrate
781: that the low $R$ and $\Phi_{\rm PAH}$ minimum corresponds to
782: gas densities that are closer to those expected for the diffuse
783: ISM and are therefore favored.
784: The minimum $\chi^2_{\rm min} = 1.6$ and lies within a closed set of contours.
785: %is indicated with an asterisk ($\ast$)
786: %and lies at a value of $R = 1\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
787: %and $\Phi_{\rm PAH} = 0.5$ within a closed set of contours.
788:
789: We next show the results for
790: sources in the remaining $A_{\rm V}$ ranges
791: $0.25 \al A_{\rm V} \al 0.75$
792: (Fig.\ [\ref{fig:chiav075}]: 10 sources),
793: $0.75 \al A_{\rm V} \al 1.25$
794: (Fig.\ [\ref{fig:chiav125}]: 12 sources), and
795: $0.25 \al A_{\rm V} \al 2.13$
796: (Fig.\ [\ref{fig:chiav025213noeps}]: 28 sources).
797: %(including $\epsilon$ Per
798: %and 59 Cyg for a total of 30 sources. and excluding
799: %23 Ori, $\pi$ Sco).
800: %The minimum $\chi^2_{\rm min} = 2.7$.
801: %$1.25 < A_{\rm V} < 2.13$
802: %(Fig.\ [\ref{fig:chiav213}]: 5 sources).
803: In these $\chi^2$
804: plots we do not consider the exact position of
805: the minimum to be as
806: significant as in Figure \ref{fig:chiav025} since it
807: generally lies in a trough of low values within
808: open contours.
809: %Nevertheless, we have indicated the
810: %position of the minimum in the figures. For the wide range
811: %of column densities
812: %$0.25 < A_{\rm V} < 2.13$
813: %(Fig.\ [\ref{fig:chiav025213noeps}]) the minimum
814: %$\chi^2$ is 1.5 at $R=6\times 10^{-17}$ ${\rm cm^3}$
815: %${\rm s^{-1}}$ and $\Phi_{\rm PAH} = 0.9$.
816: An approximate fit to the minimum trough is given by
817: \beq
818: \Phi_{\rm PAH} = 0.22 \left (\frac{R}{10^{-17}\,
819: {\rm cm^{3}}\,\, {\rm s^{-1}}} \right ) -0.40~~~R\ag
820: 3\times 10^{-17} ~{\rm cm^{3}}\,\, {\rm s^{-1}}~~~.
821: \label{eq:phitrough}
822: \eeq
823: and shown in Figure \ref{fig:chiav025213noeps}.
824:
825: %We also show the case when the points for $\epsilon$ Per
826: %and 59 Cyg are excluded (Fig.\ [\ref{fig:chiav025213noeps}])
827: %for a total of 28 points.
828: %An approximate fit to the trough is given by
829: %\beq
830: % \Phi_{\rm PAH} = 0.19 \left (\frac{R}{10^{-17}\,
831: % {\rm cm^{3}}\, {\rm s^{-1}}} \right ) -0.18~~~.
832: %\label{eq:phitroughnoeps}
833: %\eeq
834: %Removing these
835: %points reduces the minimum $\chi^2_{\rm min}$ from 2.7
836: %5to $1.5$.
837: %Such a sharp reduction in $\chi^2$ may indicate that these
838: %lines of sight are indeed anomalous. However, we will still
839: %include them in our analysis.
840:
841: %In addition to the obvious outliers (23 Ori, $\pi$ Sco,
842: %$\epsilon$ Per,
843: %and 59 Cyg) we have dropped two additional sources which have a large
844: %effect on the $\chi^2$ values.
845: %Including HD 102065 ($A_{\rm V} = 0.49$)
846: %raises the minimum $\chi^2$ value from 1.1 to 1.4
847: %for sources with $0.25 < A_{\rm V} < 0.75$ and
848: %including
849: %HD 210839 ($A_{\rm V} = 1.4)$ raises the minimum $\chi^2$ from
850: %1.6 to 3.2 for sources with $1.25 < A_{\rm V} < 2.14$.
851: %We note that there is no absorption measurement for the
852: % \ion{H}{1} column density towards HD 102065 and the
853: %total column is inferred from E(B-V). If the true
854: %column were slightly larger, $f_{\rm H_2}$ would
855: %decrease, bringing this point into better
856: %agreement with its neighbors.
857:
858: The $\chi^2$ contours
859: shows that $R$ and $\Phi_{\rm PAH}$ are correlated.
860: The minimum
861: lies at roughly a 45 degree angle with $\Phi_{\rm PAH}$
862: proportional to $R$.
863: For a given observation,
864: that is a fixed $f_{\rm H_2}$, and $A_{\rm V}$,
865: from equation
866: (\ref{eq:Afh2highcol}), the
867: ratio $Rn/G_0$ must be constant along the line of
868: minimum $\chi^2$. In addition, for a fixed $\cicol/\ciicol$,
869: from equations (\ref{eq:Agnpahm}) and (\ref{eq:Agnpahn})
870: the ratio $G_0/(n\Phi_{\rm PAH})$ must be constant along
871: the line of minimum $\chi^2$. Together these imply that
872: $R/\Phi_{\rm PAH}$ must be constant or $R\propto \Phi_{\rm PAH}$.
873:
874: There is a suggestion from our figures that the
875: best fit $\Phi_{\rm PAH}$ and $R$ increases
876: with column density. For
877: $0.25 \al A_{\rm V} \al 0.75$
878: (Fig.\ [\ref{fig:chiav075}]) the minimum $\chi^2$ lies at
879: $R= 3\times 10^{-17}$ ${\rm cm^3}$ ${\rm s^{-1}}$,
880: and $\Phi_{\rm PAH}=0.2$,
881: while for
882: $0.75 \al A_{\rm V} \al 1.25$
883: (Fig.\ [\ref{fig:chiav125}]), there is a minimum trough that extends
884: from $R= 4\times 10^{-17}$ ${\rm cm^3}$ ${\rm s^{-1}}$,
885: and $\Phi_{\rm PAH}=0.4$ to
886: $R= 6\times 10^{-17}$ ${\rm cm^3}$ ${\rm s^{-1}}$,
887: and $\Phi_{\rm PAH}=1.0$.
888: If this effect is real then it is not entirely clear what
889: is the physical reason behind it. Perhaps shock processing
890: along low $A_{\rm V}$ lines of sight tend to reduce grain surface area
891: and produce smaller $R$ (and therefore smaller $\Phi_{\rm PAH}$ since
892: $\Phi_{\rm PAH} \propto R$). The statistics however, are rather poor
893: for separate $A_{\rm V}$ bins and additional observations are required
894: to confirm the trend.
895:
896: %Plots over an extended range of $ A_{\rm V}$ values
897: %are shown in Figures \ref{fig:chiav025125}
898: %for $0.25 < A_{\rm V} < 1.25$. and
899: %\ref{fig:chiav025213}
900: %for $0.25 < A_{\rm V} < 2.13$.
901: %Including HD 102065 increases $\Phi_{\rm PAH}$
902: %from 0.93 to 1.3 for sources with
903: %$0.25 < A_{\rm V} < 1.25$.
904: %For $0.25 < A_{\rm V} < 2.13$ including HD 102065 raises
905: %the $\chi^2$ from 1.2 to 1.4, including HD 210939
906: %raises the $\chi^2$ to 1.6, and including both
907: %raises the $\chi^2$ to 1.8.
908:
909: Note that as the ${\rm H_2}$ formation rate $R$ increases
910: the density $n$ must decrease (for fixed $G_0$) in
911: order to maintain the same molecular fraction.
912: At some large value of $R$ the
913: densities become much
914: lower than typically observed in the diffuse ISM.
915: In Figures
916: \ref{fig:chiav025} and
917: \ref{fig:chiav025213noeps}
918: each $R$ and $\Phi_{\rm PAH}$ grid point
919: has several sources contributing to the $\chi^2$ value
920: and there is a best fit density for each source.
921: We show in Figures \ref{fig:denav025cp} and
922: \ref{fig:denav025213noeps}
923: the average density $\langle n \rangle$,
924: obtained by averaging the best fit density
925: over the sources contributing
926: to each $R$ and $\Phi_{\rm PAH}$ grid point.
927: Recall that in computing the $\chi^2$ grid we are
928: finding the best fit ratio $G_0/n$ and have fixed
929: $G_0$ at 1.7. Thus the density could be higher
930: for higher $G_0$.
931: %We restrict the sources to be in the range
932: %$0.25 < A_{\rm V} < 2.13$.
933: Also shown in Figure \ref{fig:denav025213noeps}
934: is the
935: line along the $\chi^2$ minimum trough
936: (eq.[\ref{eq:phitrough}]).
937: The density plot (Fig.\ \ref{fig:denav025213noeps}) combined with
938: the $\chi^2$ plot (Fig.\ \ref{fig:chiav025213noeps}) shows that the
939: typical diffuse cloud ($n \sim 30$ ${\rm cm^{-3}}$, $G_0\sim 1.7$)
940: lies at $R \approx 3.5\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
941: and $\Phi_{\rm PAH} = 0.4$. For the low column densities, the contours
942: centered on $R
943: \approx 1\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$, and
944: $\Phi_{\rm PAH} = 0.2-0.4$
945: correspond to gas densities closer to that expected in diffuse
946: gas and we therefore favor these over the larger $R$ and
947: $\Phi_{\rm PAH}$ values.
948:
949:
950: %We also show the mean density
951: %$\langle n \rangle$ when
952: %$\epsilon$ Per and 59 Cyg are excluded (Fig.\ \ref{fig:denav025213noeps})
953: %with results
954: %$R = 3.0\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
955: %and $\Phi_{\rm PAH} = 0.4$.
956:
957: \subsection{Effects of ${\rm C^+}$ Recombination Rate}
958: \label{subsec:dielectric}
959: The recombination rate of ${\rm C^+}$ with electrons
960: is most important at low $\phi_{\rm PAH}$ or high
961: values of $G_0/n$ (see Appendix~\ref{appen:estimatinggn}).
962: The total rates (including dielectronic recombination)
963: of \cite{nahar1997} are higher by a factor
964: of $\sim 2$ at 100 K compared to
965: \cite{altun2004}
966: and further diverge at lower temperatures.
967: The $\cicol/\ciicol$ ratio can thus vary by a factor of
968: $\sim 2$ when ${\rm C^+}$ recombination
969: with electrons dominates the production of \ion{C}{1}.
970: In previous sections of this paper we have used the
971: higher rates, but since the rates are uncertain, we have also
972: investigated the effects of the lower rates. Figure
973: \ref{fig:chiav025213noepsnocp} shows the resulting
974: $\chi^2$ plot for the range of column densities
975: $0.25 \la A_{\rm V} \la 2.13$ and $N_{\rm H_2} > 10^{18}$ cm$^{-2}$.
976: The effects of the \cite{nahar1997}
977: rates more narrowly constrain the best values of $\phi_{\rm PAH}$
978: and $R$ by excluding the parameter space of low $\phi_{\rm PAH}$ and
979: large $R$. This is because of the decreased importance of the
980: electron recombination rates and thus the observed $\cicol/\ciicol$ ratio can
981: only be achieved through higher PAH reaction rates. The best fit
982: minimum trough, however, is not changed significantly compared to
983: Figure \ref{fig:chiav025213noeps},
984: with only a slight
985: extension to lower $R \sim 2.0\times 10^{-17}$ $\rm cm^{3}$ ${\rm s^{-1}}$
986: and $\phi_{\rm PAH}\sim 0.2$.
987:
988: \subsection{$A_{\rm V} < 0.25$ Lines of Sight}
989: \label{subsec:lowav}
990:
991: For our low column density lines of sight, the combination
992: of a low molecular hydrogen fraction, $f_{\rm H_2}$ and a relatively
993: high $\cicol/\ciicol$ ratio indicate that $R$ must be low or at
994: least a low ratio of $R/\Phi_{\rm PAH}$ (ref.\ Figs.\ [\ref{fig:ratioav025}]
995: and [\ref{fig:chiav025}]).
996: Are there unusual physical conditions, or additional
997: %formation or
998: destruction processes that
999: are acting in these regions which could account for
1000: the low ${\rm H_2}$ columns?
1001: We first consider the effects of
1002: warm neutral medium (WNM) along the line of sight. The WNM consists
1003: of low density ($n \sim 0.3 $ ${\rm cm^{-3}}$) warm ($T\sim 8000$ K) gas.
1004: Such gas will likely contain
1005: \ion{C}{2} and \ion{H}{1}, but very little \ion{C}{1} and ${\rm H_2}$.
1006: This is because the formation rates of \ion{C}{1} and ${\rm H_2}$ are
1007: proportional to $n^2$, while the destruction rates are proportional
1008: to $n$ and thus the abundance drops in lower density gas. In addition, at
1009: high temperatures the ${\rm C^+}$ recombination rates fall. The ${\rm H_2}$ formation rate
1010: may also be lower in the WNM because at high temperatures the \ion{H}{1} atoms
1011: colliding with grains will more likely bounce than stick \citep{burke1983}.
1012: % (The unshielded fraction in cold diffuse gas is
1013: %$f_{\rm H_2} = 2.3\times 10^{-5}$ for $R=3\times 10^{-17}$ ${\rm cm^{3}}$
1014: %${\rm s^{-1}}$, but at higher WNM temperatures \ion{H}{1}
1015: %colliding with grains will more likely bounce than stick).
1016: We tested results for an $A_{\rm V} = 0.1$ cloud
1017: with $G_0 = 5.1$ and $n = 30$ ${\rm cm^{-3}}$ in which
1018: half the column is WNM gas. Here we have included the density
1019: and temperature effects in the rates
1020: but not reduced $R$ due to
1021: a smaller sticking coefficient
1022: (see Fig.\ \ref{fig:ratioav025var} arrow `W'). The model
1023: point which includes WNM moves towards lower $\cicol/\ciicol$ ratio
1024: (by a factor of $\sim 2$) and lower
1025: $f_{\rm H_2}$ (by a factor of $\sim 10$). Part of the
1026: drop in $f_{\rm H_2}$ is due to lower ${\rm H_2}$
1027: self-shielding in the $n=30$ ${\rm cm^{-3}}$ cloud.
1028: The resulting vector moves
1029: away from the low $A_{\rm V}$ observations, and along
1030: a path which is nearly parallel to the standard models.
1031: Because the addition of WNM along the line-of-sight only
1032: moves points parallel to the standard models, it is unlikely
1033: that the presence of a WNM component can explain the ${f_{\rm H_2}}$
1034: and $\cicol/\ciicol$ results at low column.
1035:
1036: We next consider the effects of a higher cosmic ray ionization rate along
1037: low column density lines of sight.
1038: The effects of increasing the primary ionization rate by a factor of 10
1039: has little affect on the ${\rm H_2}$ columns since
1040: photodissociation dominates the destruction of ${\rm H_2}$.
1041: The $\cicol/\ciicol$ ratio increases slightly (by $\sim 30$\%)
1042: due to the greater electron abundance from cosmic ray ionizations
1043: (Fig.\ \ref{fig:ratioav025var} arrow `C').
1044: We conclude that neither WNM nor enhanced cosmic ray rates
1045: can explain the low $A_{\rm V}$ observations.
1046:
1047: Another possibility is that the low $A_{\rm V}$ lines of sight
1048: are strongly affected by grain processing in interstellar shocks.
1049: \cite{jones1996} demonstrated that sputtering and
1050: grain-grain collisions can vaporize grains, and reduce the grain
1051: surface area in sufficiently fast
1052: $v_{\rm s} \al 150$ km ${\rm s^{-1}}$ shocks.
1053: In the interstellar
1054: medium, shock speeds tend to scale with $n^{-1/2}$, so that lower
1055: density gas will experience higher velocity shocks. Low $A_{\rm V}$
1056: gas may have higher fractions of low density WNM gas. We note however,
1057: that shocks of speeds $v_{\rm s} \approx 100$ km ${\rm s^{-1}}$ tend
1058: to shatter the larger grains and produce a large population of small
1059: grains which increases the surface area. We speculate that
1060: another type of grain processing
1061: occurring in shocks is to sputter the surfaces ``clean'' and thereby
1062: modify the characteristics of the atomic hydrogen adsorption
1063: sites (e.g., binding energy).
1064:
1065: A final possibility is that the low $A_{\rm V}$ columns are lines
1066: of sight which graze the edges of larger clouds. In this case, the
1067: typical shielding column density (which is weighted towards
1068: the smallest column direction, i.e., perpendicular to the line-of-sight)
1069: is much less than the line-of-sight column that we measure.
1070: We examine this
1071: possibility with a simple test. We take a cloud with a total
1072: $A_{\rm V}=1$ through the center and consider a ray which passes
1073: through the cloud with a column of $A_{\rm V} = 0.1$. At the midpoint,
1074: such a ray
1075: passes through a minimum depth of $A_{\rm V} = 2.5\times 10^{-3}$ measured
1076: from the cloud surface. We take the $n_{\rm H_2}$, $n_{\rm H\, I}$,
1077: $n_{\rm C\, I}$, and $n_{\rm C\, II}$ abundances as a function of
1078: cloud radius from model runs for $A_{\rm V} = 1$, $G_0 = 5.1$,
1079: and $n = 30$ ${\rm cm^{-3}}$. We then integrate the
1080: $\hicol $, $N_{\rm H_2}$, $\cicol$, and $\ciicol$ column densities
1081: along the grazing line of sight. Results are shown in Figure
1082: \ref{fig:ratioav025var}
1083: as an arrow labeled `G'. We see that compared to an $A_{\rm V} = 0.1$ cloud,
1084: the $f_{\rm H_2}$ decreases due
1085: to the diminished ${\rm H_2}$ self-shielding at the cloud edge.
1086: Only a small molecular column exists between the FUV illuminated
1087: surface and points along the ray and thus the self-shielding is
1088: ineffective and the ${\rm H_2}$ is dissociated.
1089: We also see that the $\cicol/\ciicol$ ratio
1090: rises. This is because the FUV field is lower by a factor of two compared
1091: to the $A_{\rm V} = 0.1$ cloud since the radiation can not penetrate
1092: from the backside of the larger cloud. The lower field reduces the
1093: photoionization rate
1094: of C to ${\rm C^+}$ and increases the abundance of ${\rm PAH^-}$
1095: which converts ${\rm C^+}$ to C. (The lower field also reduces the
1096: ${\rm H_2}$ dissociation rate but this reduction is offset by the
1097: lack of ${\rm H_2}$ self-shielding). The results significantly
1098: move the model points towards the observations. We consider
1099: grazing incident lines of sight
1100: a potential explanation for the low column density observations
1101: without resorting to unusual grain properties.
1102:
1103: For $A_{\rm V} \al 0.25$ we find two lines of sight to
1104: be anomalous. These are
1105: $\epsilon$ Per ($A_{\rm V} = 0.16$, $f_{\rm H_2} = 0.20$)
1106: and 59 Cyg ($A_{\rm V} = 0.11$, $f_{\rm H_2} = 0.19$).
1107: These lines of sight have molecular fractions which
1108: are $\ag 10^3$ times greater than others with similar (low)
1109: column densities and require different rates or environments
1110: than the low $f_{\rm H_2}$ cases.
1111: Including $\epsilon$ Per
1112: and 59 Cyg in the $\chi^2$ tests for the
1113: $0.03 \al A_{\rm V} \al 0.25$, $N_{\rm H_2} < 10^{17}$ sources
1114: raises the minimum $\chi^2$ from 1.6 to 9.2 and thus these sources
1115: can well be considered to be part of a separate population than the
1116: rest of the low $A_{\rm V}$ and low $f_{\rm H_2}$
1117: lines of sight.
1118: %two lines of sight
1119: %$\epsilon$ Per ($A_{\rm V} = 0.16$, $f_{\rm H_2} = 0.20$)
1120: %and 59 Cyg ($A_{\rm V} = 0.11$, $f_{\rm H_2} = 0.19$) to Figure
1121: %\ref{fig:ratioav075}.
1122: %are discussed further in
1123: %section \ref{subsec:lowav}.
1124: The $\kappa$ Ori line of sight at slightly higher
1125: column density ($A_{\rm V} = 0.17$)
1126: is part of the low $f_{\rm H_2}$ group as are
1127: $\epsilon$ Ori
1128: $(A_{\rm V} = 0.14)$ and 15 Mon $(A_{\rm V} = 0.13)$.
1129: One explanation for these peculiar sight lines could be
1130: that shocks of speeds $v_s\approx 100$ km ${\rm s^{-1}}$
1131: have significantly increased the population of small
1132: grains. The resulting increase in grain surface area
1133: will increase the FUV extinction (and decrease the
1134: ${\rm H_2}$ photodissociation rate) and also increase the
1135: ${\rm H_2}$ formation rate. We find that with the FUV extinction and
1136: ${\rm H_2}$ formation rate a factor of three higher than normal
1137: we can reproduce the $\cicol/\ciicol$ ratio and molecular
1138: fraction for the $\epsilon$ Per line of sight.
1139: Unfortunately, there are
1140: no low resolution IUE data for $\epsilon$ Per to examine
1141: the FUV rise in the extinction curve. We also note that
1142: 59 Cyg is a variable Be star and column densities for this
1143: line of sight are highly uncertain (D.\ Massa, private
1144: communication).
1145:
1146: %Decreasing the upper bound to $A_{\rm V} = 0.10$
1147: %would remove $\epsilon$ Per and 59 Cyg, but also
1148: %remove four additional sources on
1149: %the plot. We therefore consider $\epsilon$ Per
1150: %and 59 Cyg to be anomalous. We did not include them
1151: %in the $\chi^2$ tests and consider several possible
1152: %explanations for their high molecular abundance
1153: %in subsection~later.
1154:
1155:
1156: %2 odd points at low $A_{\rm V}$
1157:
1158: \subsection{$A_{\rm V} > 0.25$ Lines of Sight}
1159: \label{subsec:highav}
1160:
1161: At columns densities $A_{\rm V} > 0.25$ we found two points
1162: to be anomalous: 23 Ori and $\pi$ Sco. \cite{weingartner2001}
1163: modeled the line of sight towards 23 Ori
1164: and found that they
1165: had some difficulty matching the high $\cicol$ column even when
1166: recombination on grains was included. They proposed several
1167: additional processes
1168: which might be working to enhance the C abundance including the
1169: dissociative recombination of ${\rm CH^+}$. The ${\rm CH^+}$ abundance
1170: can be enhanced compared to normal equilibrium chemistry due to
1171: the effects of non-thermal chemistry
1172: \citep{flower1998,joulain1998,zsargo2003}. If
1173: turbulent velocities are sufficiently large ($v\ag 3-4$ km ${\rm s^{-1}}$)
1174: and turbulent dissipation can drive reactions with temperature barriers, then
1175: a large column of ${\rm CH^+}$ can be produced \citep{joulain1998}.
1176:
1177: Another process which might account for the anomalous columns is
1178: the time dependence of the chemistry. The line of sight towards
1179: 23 Ori shows a large $\cicol/\ciicol$ ratio at small molecular
1180: fractions $f_{\rm H_2}$ compared to the equilibrium abundances.
1181: Consider a parcel of gas which has been recently shocked (i.e.,
1182: ${\rm H_2}$ dissociated and \ion{C}{1} ionized). The
1183: molecular hydrogen is produced at a rate of $\sim 3\times 10^{-17}n^2$
1184: ${\rm s^{-1}}$. The time to double the current fractional abundance
1185: $n_{\rm H_2}/n \sim 5\times 10^{-3}$ is about
1186: $t\sim 5\times 10^{-3}/(3\times 10^{-17}n)$ s or $t\sim 5.3\times 10^{4}$
1187: yr for $n=100$ ${\rm cm^{-3}}$. On the other hand,
1188: the time to double the current ratio $\cicol/\ciicol\sim 0.01$
1189: is about $t\sim 0.01/(\Phi_{\rm PAH}2.4\times 10^{-7}n_{\rm PAH^-})$ s
1190: or $t\sim 62$ yr for $n=100$ ${\rm cm^{-3}}$, where we have used
1191: ${\rm C^+}$ recombination with ${\rm PAH^-}$,
1192: $\Phi_{\rm PAH} = 0.5$, and
1193: $n_{\rm PAH^-}$ from equation (\ref{eq:Anpahm}).
1194: The time to reach the equilibrium $\cicol/\ciicol$ ratio is much shorter than
1195: the time to reach the equilibrium $f_{\rm H_2}$ fraction. Thus, for
1196: a recent ($t\al 5\times 10^{4}$ yr) passage of a shock, the
1197: $\cicol/\ciicol$ ratio has had time to reach equilibrium while
1198: the ${\rm H_2}$ fraction is dissociated compared to equilibrium.
1199:
1200: For $\pi$ Sco consider a parcel of molecular gas which has been recently
1201: illuminated by a source of FUV radiation. The time to dissociate the current
1202: $f_{\rm H_2}\sim 0.01$ at cloud center is
1203: $t\sim (G_0 I \beta[N_{\rm H_2}] \exp[-2.5 A_{\rm V}])^{-1}$
1204: or $t \sim 1.1\times 10^{6}$ yr, where we have
1205: used $G_0=5.1$, $N_{\rm H_2} = 1\times 10^{19}$
1206: ${\rm cm^{-2}}$, $\beta[1\times 10^{19}\,{\rm cm^{-2}}]=1.7\times 10^{-4}$
1207: (Eq.[\ref{eq:dbself}]), and $A_{\rm V} = 0.13$.
1208: The time to ionize the current $\cicol/\ciicol \sim10^{-4}$ ratio is
1209: about $t\sim (2.1\times 10^{-10}G_0 \exp[-2.6 A_{\rm V}])^{-1}$ s
1210: or $t\sim 41$ yr. Thus for a molecular clump illuminated
1211: within the past $\sim 10^6$ yr, the $\cicol/\ciicol$ ratio is in equilibrium
1212: while the ${\rm H_2}$ fraction is much higher than its equilibrium
1213: value.
1214:
1215: One complicating factor which we have not yet considered is the
1216: effects of multiple clump components along the line of sight.
1217: In general, taking the observed column density and splitting
1218: it up into multiple clouds exposes more surfaces to the
1219: interstellar radiation field. In the case of optically
1220: thick clumps,
1221: this will decrease the
1222: $\cicol/\ciicol$ ratio and decrease the molecular fraction
1223: $f_{\rm H_2}$.
1224: For the case of optically thin clumps, the
1225: $\cicol/\ciicol$ ratio will not change but
1226: $f_{\rm H_2}$ will drop due to reduced self-shielding.
1227: However, the effects of multiple clumps is important only for
1228: components of approximately equal column density. If one component
1229: dominates then the remaining smaller components do not contribute
1230: much to the total column. Since in general, the UV absorption
1231: spectroscopy observations are dominated
1232: by a single component, our neglect of multiple clouds
1233: should not significantly alter our results.
1234:
1235: %${\rm H_2}$ columns from Lacour 2005, ApJ, 627, 251
1236:
1237: %\subsection{Individual Sources}
1238: %\label{subsec:individual}
1239:
1240: \section{Summary}
1241: \label{sec:summary}
1242:
1243: We have carried out an analysis of observations
1244: of the \ion{C}{1}, \ion{C}{2}, \ion{H}{1}, and
1245: ${\rm H_2}$ column densities in the diffuse ISM
1246: towards 42 lines of sight spanning a range of
1247: column densities $0.03 \le A_{\rm V} \le 2.13$.
1248: We have fitted these columns using a 2-sided
1249: PDR model which simultaneously solves for the
1250: thermal and chemical balance throughout the layer.
1251: The relative columns of \ion{H}{1} and ${\rm H_2}$ are dependent
1252: on the rate coefficient $R$ for ${\rm H_2}$ formation on grain surfaces,
1253: on the ratio $G_0/n$,
1254: and on the total column $N$ (or $A_{\rm V}$) of the illuminated cloud. The
1255: relative columns of \ion{C}{1} and \ion{C}{2} depend on the electron
1256: and PAH abundances, the rate coefficients for neutralization of C$^+$ with
1257: PAHs (or $\Phi _{\rm PAH}$), the total column in the illuminated cloud, and $G_0/n$.
1258: Observations set the total columns of each line of sight plus the
1259: fractional abundances of H, ${\rm H_2}$, C, and ${\rm C^+}$. Our PDR models self
1260: consistently determine the electron, PAH$^+$, PAH, and PAH$^-$ abundances
1261: for each model. We have performed a $\chi^2$ test using $R$,
1262: $\Phi _{\rm PAH}$, and $G_0/n$ as our free parameters to find the best
1263: fits to the fractional abundances along each line of sight.
1264:
1265: The results are shown in Figures \ref{fig:chiav025} through \ref{fig:denav025213noeps},
1266: where we show the $\chi ^2$
1267: contours in the $R$, $\Phi _{\rm PAH}$ plane, given the best fitting $G_0/n$
1268: for each individual line of sight. Note that Figures
1269: \ref{fig:chiav025} through \ref{fig:denav025213noeps}
1270: are
1271: the $\chi ^2$ for these values of $R$, $\Phi _{\rm PAH}$ for all 42 sources.
1272: Our main conclusions are as follows:
1273:
1274: \noindent
1275: 1. For low column ($A_{\rm V} < 0.25$) lines of sight, and for the low
1276: molecular columns and fractions that accompany them
1277: ($N_{\rm H_2}< 10^{17}$ cm$^{-2}$, $f_{H_2} < 10^{-4}$), we
1278: find $R \simeq 1 \times 10^{-17}$ cm$^3$ s$^{-1}$ and $\Phi_{\rm PAH}
1279: \sim 0.2 - 0.4$.
1280:
1281: \noindent
1282: 2. At higher column densities ($0.25 < A_{\rm V} < 2.13$),
1283: we find $R \simeq 3.5 \times 10^{-17}$ cm$^3$ s$^{-1}$ and
1284: $\Phi _{\rm PAH} \sim 0.4$.
1285:
1286: \noindent
1287: 3. Our $\chi ^2$ fits show a correlated range of $\Phi _{\rm PAH}$ and $R$
1288: values that provide as good fits as our ``best fits" quoted above. For
1289: $A_V > 0.25$, we find good fits for $\Phi _{PAH} = 0.22 (R/10^{-17} {\rm
1290: cm}^3\ {\rm s}^{-1}) - 0.40$ as long as $R> 3 \times 10^{-17}$
1291: cm$^3$ s$^{-1}$. Very high values of $R$ (or therefore $\Phi_{\rm PAH}$),
1292: however, can only be fit with very low average densities in
1293: the clouds along the 42 sight lines. Figure \ref{fig:denav025213noeps}
1294: shows that for
1295: densities (equivalently thermal pressures) thought to be typical
1296: of local diffuse interstellar clouds, a value of $R \sim 3.5 \times 10^{-17}$
1297: cm$^3$ s$^{-1}$ (and therefore $\Phi _{\rm PAH} \sim 0.4$)
1298: is preferred. This value is consistent with previous determinations
1299: of $R$ made by comparing observations of ${\rm H_2}$ with model
1300: calculations \citep[e.g.,][]{jura1975}.
1301:
1302: \noindent
1303: 4. The low values of $R$ for low column density sight lines may indicate
1304: shock processing that decreases grain surface area or modifies
1305: grain surfaces along these sightlines, thereby reducing the ${\rm H_2}$
1306: formation rates, or a line of sight which grazes a larger cloud.
1307:
1308: %(What about WNM where R may be low because of H atoms bouncing from grains?)
1309:
1310: \noindent
1311: 5. Our results for $\Phi _{\rm PAH}$ indicate that the crude estimates of these
1312: ratios adopted by \cite{wolfire2003} were roughly correct.
1313: %but perhaps a factor
1314: %of $\sim 2$ too high.
1315: They confirm the importance of PAHs in determining
1316: the ionization level of atomic ions (in this case C vs C$^+$) in
1317: interstellar clouds.
1318:
1319: \acknowledgments
1320:
1321: M.G.W. was supported in part by a NASA Long Term Space Astrophysics (LTSA)
1322: grant NNG05GD64G. We thank an anonymous referee for a careful reading of
1323: our paper and for comments that improved the presentation.
1324:
1325:
1326: \appendix
1327:
1328: \section{Chemical Rates}
1329: \label{appen:pahrates}
1330:
1331: The chemical rates with PAHs were presented in \cite{wolfire2003}
1332: Appendix C2 for reactions involving the ionization balance
1333: of hydrogen. Here we reproduce the rates along with the Carbon rates
1334: used in this paper. Our starting
1335: point are the rates from \cite{draine1987} with $N_{\rm C} = 35$
1336: carbon atoms, and disk PAHs with a radius $a=(N_{\rm C}/1.222)^{0.5}$.
1337: We use a PAH abundance of $n_{\rm PAH} = 6\times 10^{-7}n$.
1338: \cite{wolfire2003} introduced the factor $\Phi_{\rm PAH}$ to
1339: account for a wide range of unknowns in the PAH distribution
1340: and rate coefficients. Simple fits to the rates appropriate
1341: to diffuse ISM conditions
1342: are presented in Table \ref{tbl:reactions}.
1343:
1344: \section{Estimating Radiation Field $G_0$ and Density $n$}
1345: \label{appen:estimatinggn}
1346:
1347: In this appendix we derive simple expressions to estimate the
1348: FUV radiation field $G_0$ incident on the cloud and the cloud
1349: density $n$ from the observed column densities of $N({\rm H\, I})$,
1350: $N_{\rm H_2}$,
1351: $\cicol$, and $\ciicol$.
1352: In the limit of no dust extinction, the equilibrium abundance of
1353: ${\rm H_2}$ is given by
1354: \beq
1355: G_0 I n_{\rm H_2} \beta [N_{\rm H_2}] = R\,n \, n_{{\rm H\, I}}~~~,
1356: \label{eq:Ah2eq}
1357: \eeq
1358: where $G_0$ is the incident FUV radiation field, $I$ the unshielded
1359: photodissociation rate, $\beta [N_{\rm H_2}]$ is the ${\rm H_2}$
1360: self-shielding factor, and $R$ is the formation rate on grains.
1361: For ${\rm H_2}$ columns $\al 10^{14}$ ${\rm cm^{-3}}$,
1362: $\beta [N_{\rm H_2}]=1$ and the molecular fraction is given by
1363: \beq
1364: f_{\rm H_2} = \frac{2nR}{G_0 I}\,\,\, .
1365: \label{eq:Afh2lowcol}
1366: \eeq
1367:
1368: To estimate $G_0$ and $n$ we use a
1369: simple expression for the ${\rm H_2}$
1370: self-shielding formula from
1371: \cite{draine1996}
1372: \beq
1373: \beta [ N_{\rm H_2} ] = \left [ \frac{N_0}{N_{\rm H_2}+N_0}
1374: \right ]^{3/4}~~~,
1375: \label{eq:dbself}
1376: \eeq
1377: where $N_0 = 1\times 10^{14}$ ${\rm cm^{-2}}$ is the ${\rm H_2}$ column
1378: where line self-shielding starts to become significant. We can integrate
1379: equation (\ref{eq:Ah2eq}) through the cloud
1380: \beq
1381: G_0 I N_0^{3/4} \int \frac{d N_{\rm H_2}^{'}}{[N_0 + N_{\rm H_2}^{'}]^{3/4}}
1382: = Rn\int d N_{\rm H\, I}^{'}~~~.
1383: \eeq
1384: Letting $x = N_0 + N_{\rm H_2}$ the integral becomes
1385: \beq
1386: G_0 I N_0^{3/4} \int_{N_0}^{N_0 + N_{\rm H_2}}
1387: \frac{d x^{'}}{x^{'3/4}}
1388: = R\, n\, N_{\rm {H\, I}}~~~.
1389: \eeq
1390: Carrying out the integral and solving for $G_0/n$ we have
1391: \beq
1392: \frac{G_0}{n} = \frac{0.25 N_{\rm H\, I}}
1393: {\left \{ [N_0 + N_{\rm H_2}]^{1/4} - N_0^{1/4} \right \}
1394: N_0^{3/4}} \frac{R}{I}^{}~~~,
1395: \eeq
1396: or
1397: \beq
1398: \frac{G_0}{n} = {\cal X}_{\rm H} \frac{R}{I}^{}~~~,
1399: \label{eq:Agn}
1400: \eeq
1401: with
1402: \beq
1403: {\cal X}_{\rm H} = \frac{0.25 N_{\rm H\, I}}
1404: {\left \{ [N_0 + N_{\rm H_2}]^{1/4} - N_0^{1/4} \right \}
1405: N_0^{3/4}}~~~.
1406: \label{eq:Axh}
1407: \eeq
1408:
1409: For $N_{\rm H_2} > N_0$ from equations (\ref{eq:Agn}), (\ref{eq:Axh}) and
1410: \cite{hollenbach1999} the molecular fraction is given by
1411: \beq
1412: f_{\rm H_2} = 2 \left ( \frac{Rn}{4G_0I} \right )^4 \left ( \frac{N}{N_0}
1413: \right )^3\,\,\, .
1414: \label{eq:Afh2highcol}
1415: \eeq
1416: For clouds where the optical depth of the dust becomes significant the
1417: local FUV field within the cloud is less than the incident field. We
1418: assume the flux on each side of a slab is $1/2 G_0$ and the mean field,
1419: $\langle G_0 \rangle$
1420: within the slab is approximately at a depth of 1/4 the total column density
1421: or at a depth of $1/4 A_{\rm V}$. Thus the mean field is given by
1422: \beq
1423: \langle G_0 \rangle = \frac {1}{2}G_0 e^{ (-\frac{3}{4} \alpha A_{\rm V})}
1424: + \frac {1}{2}G_0 e^{ (-\frac{1}{4} \alpha A_{\rm V})}~~~,
1425: \eeq
1426: where $\alpha = 2.5$ converts the FUV extinction at visual wavelengths
1427: to optical depth at the FUV dissociation energies of ${\rm H_2}$. Solving
1428: for $G_0$ we have
1429: \beq
1430: G_0 = \frac{2 \langle G_0 \rangle e^{\frac{2.5}{4}A_{\rm V}}}
1431: {1 + e^{-\frac{2.5}{2}A_{\rm V}}}~~~.
1432: \eeq
1433: Setting $G_0$ in equation (\ref{eq:Agn}) to the mean field
1434: $\langle G_0 \rangle$
1435: we have
1436: \beq
1437: \frac{G_0}{n} = \frac{2 e^{\frac{2.5}{4}A_{\rm V}}}
1438: {1 + e^{-\frac{2.5}{2}A_{\rm V}}} {\cal X}_{\rm H} \frac{R}{I}~~~.
1439: \label{eq:Agmeann}
1440: \eeq
1441:
1442: The analytic solution is approximate since $1/4A_{\rm V}$ is an approximation
1443: to the depth where the local field equals the mean field.
1444: Comparing equation (\ref{eq:Agmeann}) with our numerical results
1445: we find we can achieve better agreement with an additional factor
1446: of $1/2$ on the right hand side.
1447:
1448: A second estimate of $G_0/n$ can be obtained from the $\cicol/\ciicol$ ratio.
1449: There are several
1450: destruction processes for ${\rm C^+}$ to consider depending on the
1451: ${\rm PAH^-}$
1452: abundance. If destruction occurs primarily through recombination on
1453: ${\rm PAH^-}$
1454: then the equilibrium abundance of ${\rm C^0}$ is found by balancing
1455: recombination
1456: with photoionization of C
1457: \beq
1458: G_0 I_{\rm C} n_{\rm C^0}= \Phi_{\rm PAH} \kappa_1 n_{\rm C^+}
1459: n_{\rm PAH^-}~~~,
1460: \label{eq:Arecombination}
1461: \eeq
1462: where $G_0I_{\rm C} = G_0 2.1\times 10^{-10}$ ${\rm s^{-1}}$ is the photoionization rate of
1463: ${\rm C^0}$, and $\Phi_{\rm PAH} \kappa_1 = \Phi_{\rm PAH} 2.4\times 10^{-7} T_2^{-0.5}$
1464: ${\rm cm^3}$ ${\rm s^{-1}}$ is the ${\rm C^+}$ recombination rate with ${\rm PAH^-}$.
1465: From \cite{wolfire2003}, the ${\rm PAH^-}$ abundance in the regime
1466: $n_{\rm PAH^-}/n_{\rm PAH^0} < 1$ is given by
1467: \beq
1468: n_{\rm PAH^-} = 4.8\times 10^{-5} n_e n
1469: \Phi_{\rm PAH}G_0^{-1}~~~{\rm cm^{-3}},
1470: \label{eq:Anpahm}
1471: \eeq
1472: with the electron density $n_e$ and hydrogen nucleus density $n$ in units of
1473: ${\rm cm^{-3}}$. Substituting this expression for $n_{\rm PAH^-}$ into equation
1474: (\ref{eq:Arecombination}), and solving for $G_0/n$ we have
1475: \beq
1476: \frac{G_0}{n} = 4.1 \times 10^{-3} \Phi_{\rm PAH} T_2^{-0.25}
1477: \left [ \frac{\cicol}{\ciicol} \right ]^{-1/2}
1478: \label{eq:Agnpahm}
1479: \eeq
1480: where we have used an electron abundance of
1481: $n_e \approx 3\times 10^{-4}n$ ${\rm cm^{-3}}$ (including metals, especially
1482: ionization of C, and ionization
1483: of H by soft X-rays), and substituted the observed ratio
1484: $\cicol/\ciicol$ for $n_{\rm C^0}/n_{\rm C^+}$.
1485:
1486: If the abundance of ${\rm PAH^-}$ is low enough, then the destruction of
1487: ${\rm C^+}$
1488: occurs through charge exchange with ${\rm PAH^0}$ or through
1489: electron recombination in the
1490: gas phase. The equilibrium abundance is then given by
1491: \beq
1492: G_0 I_C n_{\rm C^0}= n_{\rm C^+} (k_2 n_{\rm PAH^0} + k_3 n_e)
1493: \label{eq:Achargeexchange}
1494: \eeq
1495: with $k_2 = 8.8\times 10^{-9} \Phi_{\rm PAH}$ ${\rm cm^3\,\, s^{-1}}$,
1496: and $k_3 = 1.8\times 10^{-11}T_2^{-0.83}$ ${\rm cm^3}\,\, s^{-1}$.
1497: Using the ${\rm PAH^0}$ abundance of $\sim 0.7$ times
1498: the total PAH abundance
1499: ($n_{\rm PAH^0}=0.7 \times 6\times 10^{-7} n$;
1500: Wolfire et al.\ 2003),
1501: and an electron abundance of $n_e \approx 3\times 10^{-4}n$ ${\rm cm^{-3}}$,
1502: we can solve for $G_0/n$
1503: \beq
1504: \frac{G_0}{n} = (1.8\times 10^{-5} \Phi_{\rm PAH} + 2.6\times 10^{-5}
1505: T_2^{-0.83} ) \left [\frac{\cicol}{\ciicol} \right ]^{-1}
1506: \label{eq:Agnpahn}
1507: \eeq
1508: %The destruction of ${\rm C^+}$ by ${\rm PAH^0}$ should dominate when
1509: %\beq
1510: % n_{\rm PAH^0}k_2 > n_{\rm PAH^-} k_1
1511: %\eeq
1512: %Using $n_{\rm PAH^0} = 0.7 \times 6\times 10^{-7} n$ and
1513: %$n_{\rm PAH^-}$ from equation (\ref{eq:Anpahm}) we obtain a limit on
1514: %$G_0/n$ for destruction of ${\rm C^+}$ by neutral PAHs
1515: %\beq
1516: % \frac{G_0}{n} > 0.9 T_2^{-1/2}\Phi_{\rm PAH}
1517: %\eeq
1518: We find that the destruction of ${\rm C^+}$ by ${\rm PAH^0}$
1519: dominates at a limit of approximately
1520: %Over the full range of model space,
1521: %due to the variable abundances of the PAH ionization fractions,
1522: %the electron abundances, the gas temperature, and the FUV field,
1523: %we find a some what smaller cross over better matches the numerical results.
1524: %We adopt the limit of
1525: \beq
1526: \frac{G_0}{n} \ag \frac{T_2^{-1/2}\Phi_{\rm PAH}^2}
1527: {1.2\Phi_{\rm PAH} + 1.6 T_2^{-0.83}}
1528: \eeq
1529: From equation (\ref{eq:Agnpahm}), and setting $T_2=1$, and
1530: $\Phi_{\rm PAH} = 0.5$,
1531: this limit
1532: corresponds to an observed
1533: column density ratio ${\cicol}/{\ciicol} \al 3\times 10^{-4}$. Thus for
1534: ${\cicol}/{\ciicol} \al 3\times 10^{-4}$ we use equation (\ref{eq:Agnpahn})
1535: while for ${\cicol}/{\ciicol} >> 3\times 10^{-4}$ we use equation
1536: (\ref{eq:Agnpahm}).
1537:
1538: Similar to the $G_0/n$ estimate based on ${\rm H_2}$ columns we also
1539: correct the incident FUV field for dust extinction within the cloud
1540: by setting the $G_0$ in equations (\ref{eq:Agnpahm}) and (\ref{eq:Agnpahn})
1541: to the mean FUV field $\langle G_0 \rangle$ and solve for the extinction
1542: corrected $G_0/n$. In solving for the minimum $\chi^2$, we use the
1543: estimates for $G_0/n$ presented here along with a fixed value of
1544: $G_0 = 1.7$ to provide a first guess range of $n$. If the minimum
1545: in $\chi^2$ is
1546: not crossed within the first range of $n$ we take a second range given by
1547: $n_{\rm max} \times (n_{\rm max}/n_{\rm min})$
1548: if higher densities are required
1549: or $n_{\rm min} \times (n_{\rm max}/n_{\rm min})^{-1}$ if lower densities
1550: are required
1551: (where
1552: $n_{\rm max}$ and $n_{\rm min}$ refer to the maximum and
1553: minimum densities in the range).
1554: The minimum in $\chi^2$ is generally found within the first or second
1555: range. We then do a quadratic fit of the $\chi^2$ values
1556: as a function of density to find the density which provides the minimum
1557: $\chi^2$ value.
1558:
1559:
1560: \section{Variation of $\cicol/\ciicol$ as a function of
1561: $G_0/n$ and $\Phi_{\rm PAH}$}
1562: \label{appen:variation}
1563:
1564: In Figures
1565: \ref{fig:ratioav025} through \ref{fig:ratioav213} we show the variation
1566: in $\cicol/\ciicol$ versus $f_{\rm H_2}$ as a function of $G_0/n$ and
1567: $\Phi_{\rm PAH}$. In the limit of low $G_0/n$ the destruction of ${\rm C^+}$
1568: is dominated by recombination on ${\rm PAH^-}$ and rate equation
1569: (\ref{eq:Arecombination}) applies. From equation (\ref{eq:Agnpahm})
1570: we have
1571: \beq
1572: \frac{\cicol}{\ciicol} \propto \Phi_{\rm PAH}^2
1573: \left [ \frac{n}{G_0} \right ]^2\,\,\, .
1574: \label{eq:Clowgn}
1575: \eeq
1576: We see in Figures \ref{fig:ratioav025} through \ref{fig:ratioav213}, along
1577: curves of constant $G_0$, the $\cicol/\ciicol$ ratio rises steeply
1578: as $n$ increases.
1579:
1580: In the limit of high $G_0/n$ the destruction of ${\rm C^+}$ is dominated
1581: by charge exchange with ${\rm PAH^0}$ and rate equation
1582: (\ref{eq:Achargeexchange}) applies. From equation (\ref{eq:Agnpahn})
1583: the $\cicol/\ciicol$ ratio is given by
1584: \beq
1585: \frac{\cicol}{\ciicol} \propto \Phi_{\rm PAH}
1586: \left [ \frac{n}{G_0} \right ] \,\,\, ,
1587: \label{eq:Chighgn}
1588: \eeq
1589: and the curves start to flatten as $n$ decreases.
1590:
1591: At low values of $\Phi_{\rm PAH}$ the destruction of ${\rm C^+}$
1592: is dominated by gas phase recombination with electrons, $n_e$.
1593: In the limit of low values of $G_0/n$, the electron abundance
1594: is determined mainly by the photoionization and recombination of
1595: Carbon.
1596: From equation
1597: (\ref{eq:Agnpahn}) the $\cicol/\ciicol$ ratio is given by
1598: \beq
1599: \frac{\cicol}{\ciicol} \propto
1600: \left [ \frac{n}{G_0} \right ] \,\,\, .
1601: \label{eq:Clowgnlowphi}
1602: \eeq
1603: In the limit of high $G_0/n$, the electron abundance is mainly
1604: determined by the soft X-ray photoionization of H. In this case
1605: $n_e \propto n^{1/2}$ and
1606: \beq
1607: \frac{\cicol}{\ciicol} \propto
1608: \frac{n^{1/2}}{G_0} \,\,\, ,
1609: \label{eq:Chighgnlowphi}
1610: \eeq
1611: and we see that the curves flatten considerably at low values of $n$.
1612:
1613: %\newpage
1614: \begin{thebibliography}{}
1615: %\bibitem[Abel et al.(2007)]{abel2007} Abel, N., Federman, S., \&
1616: %Sheffer, Y.\ 2007, American Astronomical Society Meeting Abstracts, 210,
1617: %\#19.01
1618:
1619: \bibitem[Abgrall et al.(1992)]{abgrall1992} Abgrall, H., Le
1620: Bourlot, J., Pineau Des Forets, G., Roueff, E., Flower, D.~R., \& Heck, L.\
1621: 1992, \aap, 253, 525
1622:
1623: \bibitem[Altun et al.(2004)]{altun2004} Altun, Z., Yumak, A.,
1624: Badnell, N.~R., Colgan, J., \& Pindzola, M.~S.\ 2004, \aap, 420, 775
1625:
1626: \bibitem[Bakes \& Tielens(1998)]{bakes1998} Bakes, E.~L.~O.,
1627: \& Tielens, A.~G.~G.~M.\ 1998, \apj, 499, 258
1628:
1629: \bibitem[Boiss{\'e} et al.(2005)]{boisse2005} Boiss{\'e}, P.,
1630: Le Petit, F., Rollinde, E., Roueff, E., Pineau des For{\^e}ts, G.,
1631: Andersson, B.-G., Gry, C., \& Felenbok, P.\ 2005, \aap, 429, 509
1632:
1633: \bibitem[Bohlin et al.(1978)]{bohlin1978} Bohlin, R.~C., Savage,
1634: B.~D., \& Drake, J.~F.\ 1978, \apj, 224, 132
1635:
1636: \bibitem[Browning et al.(2003)]{browning2003} Browning, M.~K.,
1637: Tumlinson, J., \& Shull, J.~M.\ 2003, \apj, 582, 810
1638:
1639: \bibitem[Burke \& Hollenbach(1983)]{burke1983} Burke, J.~R., \&
1640: Hollenbach, D.~J.\ 1983, \apj, 265, 223
1641:
1642: \bibitem[Cartledge et al.(2004)]{cartledge2004} Cartledge, S.~I.~B.,
1643: Lauroesch, J.~T., Meyer, D.~M., \& Sofia, U.~J.\ 2004, \apj, 613, 1037
1644:
1645: \bibitem[Cazaux \& Tielens(2004)]{cazaux2004} Cazaux, S., \&
1646: Tielens, A.~G.~G.~M.\ 2004, \apj, 604, 222
1647:
1648: \bibitem[Chang et
1649: al.(2006)]{chang2006} Chang, Q., Cuppen, H.~M., \& Herbst, E.\ 2006,
1650: \aap, 458, 497
1651:
1652: \bibitem[Diplas \& Savage(1994)]{diplas1994} Diplas, A., \&
1653: Savage, B.~D.\ 1994, \apjs, 93, 211
1654:
1655: \bibitem[Draine(1978)]{draine1978} Draine, B.~T.\ 1978, \apjs, 36,
1656: 595
1657:
1658: \bibitem[Draine \& Bertoldi(1996)]{draine1996} Draine, B.~T., \&
1659: Bertoldi, F.\ 1996, \apj, 468, 269
1660:
1661: \bibitem[Draine \& Sutin(1987)]{draine1987} Draine, B.~T., \&
1662: Sutin, B.\ 1987, \apj, 320, 803
1663:
1664: \bibitem[Fitzpatrick \& Massa(1990)]{fitz1990} Fitzpatrick,
1665: E.~L., \& Massa, D.\ 1990, \apjs, 72, 163
1666:
1667: \bibitem[Flower
1668: \& Pineau des Forets(1998)]{flower1998} Flower, D.~R.,
1669: \& Pineau des Forets, G.\ 1998, \mnras, 297, 1182
1670:
1671: \bibitem[Gillmon \& Shull(2006)]{gillmon2006} Gillmon, K., \&
1672: Shull, J.~M.\ 2006, \apj, 636, 908
1673:
1674: \bibitem[Gry et al.(2002)]{gry2002} Gry, C., Boulanger,
1675: F., Nehm{\'e}, C., Pineau des For{\^e}ts, G., Habart, E., \& Falgarone,
1676: E.\ 2002, \aap, 391, 675
1677:
1678: \bibitem[Habart et al.(2004)]{habart2004} Habart, E.,
1679: Boulanger, F., Verstraete, L., Walmsley, C.~M.,
1680: \& Pineau des For{\^e}ts, G.\ 2004, \aap, 414, 53
1681:
1682: \bibitem[Habing(1968)]{habing1968} Habing, H.~J.\ 1968, \bain, 19,
1683: 421
1684:
1685: \bibitem[Hobbs et al.(1982)]{hobbs1982} Hobbs, L.~M., York,
1686: D.~G., \& Oegerle, W.\ 1982, \apjl, 252, L21
1687:
1688: \bibitem[Hollenbach \& Tielens(1999)]{hollenbach1999} Hollenbach,
1689: D.~J., \& Tielens, A.~G.~G.~M.\ 1999, Reviews of Modern Physics, 71, 173
1690:
1691: \bibitem[Hollenbach \& Salpeter(1971)]{hollenbach1971} Hollenbach, D.,
1692: \& Salpeter, E.~E.\ 1971, \apj, 163, 155
1693:
1694: \bibitem[Howk et al.(1999)]{howk1999} Howk, J.~C., Savage,
1695: B.~D., \& Fabian, D.\ 1999, \apj, 525, 253
1696:
1697: \bibitem[Jenkins et al.(1983)]{jenkins1983} Jenkins, E.~B., Jura,
1698: M., \& Loewenstein, M.\ 1983, \apj, 270, 88
1699:
1700: \bibitem[Jenkins \& Shaya(1979)]{jenkins1979} Jenkins, E.~B.,
1701: \& Shaya, E.~J.\ 1979, \apj, 231, 55
1702:
1703: \bibitem[Jenkins \& Tripp(2001)]{jt2001} Jenkins, E.~B., \&
1704: Tripp, T.~M.\ 2001, \apjs, 137, 297
1705:
1706: \bibitem[Jenkins \& Peimbert(1997)]{jenkins1997} Jenkins,
1707: E.~B., \& Peimbert, A.\ 1997, \apj, 477, 265
1708:
1709: \bibitem[Jenkins et al.(1999)]{jenkins1999} Jenkins, E.~B., Tripp,
1710: T.~M., Wo{\'z}niak, P.~R., Sofia, U.~J., \& Sonneborn, G.\ 1999, \apj, 520,
1711: 182
1712:
1713: \bibitem[Jenkins et al.(2000)]{jenkins2000} Jenkins, E.~B.,
1714: Wo{\'z}niak, P.~R., Sofia, U.~J., Sonneborn, G., \& Tripp, T.~M.\ 2000,
1715: \apj, 538, 275
1716:
1717: %\bibitem[Jura(1974)]{jura1974} Jura, M.\ 1974, \apj, 191,
1718: % 375
1719:
1720: \bibitem[Jones et al.(1996)]{jones1996} Jones, A.~P., Tielens,
1721: A.~G.~G.~M., \& Hollenbach, D.~J.\ 1996, \apj, 469, 740
1722:
1723: \bibitem[Joulain et
1724: al.(1998)]{joulain1998} Joulain, K., Falgarone, E., Des Forets, G.~P.,
1725: \& Flower, D.\ 1998, \aap, 340, 241
1726:
1727: \bibitem[Jura(1975)]{jura1975} Jura, M.\ 1975, \apj, 197,
1728: 575
1729:
1730: \bibitem[Kaufman et al.(1999)]{kaufman1999} Kaufman, M.~J.,
1731: Wolfire, M.~G., Hollenbach, D.~J., \& Luhman, M.~L.\ 1999, \apj, 527, 795
1732:
1733: \bibitem[Le Petit et al.(2006)]{lepetit2006} Le Petit, F.,
1734: Nehm{\'e}, C., Le Bourlot, J., \& Roueff, E.\ 2006, \apjs, 164, 506
1735:
1736: \bibitem[Lepp et al.(1988)]{lepp1988} Lepp, S., Dalgarno, A.,
1737: van Dishoeck, E.~F., \& Black, J.~H.\ 1988, \apj, 329, 418
1738:
1739: \bibitem[Lipshtat
1740: \& Biham(2003)]{lips2003} Lipshtat, A., \& Biham, O.\ 2003, \aap, 400, 585
1741:
1742: \bibitem[McCall et al.(2003)]{mccall2003} McCall, B.~J., et al.\
1743: 2003, \nat, 422, 500
1744:
1745: %\bibitem[McKee(1989)]{mckee1989} McKee, C.\ 1989, IAU Symp.~135:
1746: %Interstellar Dust, 135, 431
1747:
1748: \bibitem[Nahar \& Pradhan(1997)]{nahar1997} Nahar, S.~N., \&
1749: Pradhan, A.~K.\ 1997, \apjs, 111, 339
1750:
1751: \bibitem[Neufeld et al.(2005)]{neufeld2005} Neufeld, D.~A.,
1752: Wolfire, M.~G., \& Schilke, P.\ 2005, \apj, 628, 260
1753:
1754: \bibitem[P\'equignot(1990)]{Pequignot1990} P\'equignot, D.\ 1990, \aap,
1755: 231, 499 (erratum 313, 1026 [1996])
1756:
1757: \bibitem[Pirronello et al.(1999)]{pirronello1999} Pirronello, V., Liu,
1758: C., Roser, J.~E., \& Vidali, G.\ 1999, \aap, 344, 681
1759:
1760: \bibitem[Rachford et al.(2002)]{rachford2002} Rachford, B.~L.,
1761: et al.\ 2002, \apj, 577, 221
1762:
1763: \bibitem[Savage et al.(1977)]{savage1977} Savage, B.~D., Bohlin,
1764: R.~C., Drake, J.~F., \& Budich, W.\ 1977, \apj, 216, 291
1765:
1766: \bibitem[Shull \& van Steenberg(1985)]{shull1985} Shull,
1767: J.~M., \& van Steenberg, M.~E.\ 1985, \apj, 294, 599
1768:
1769: \bibitem[Snow(1977)]{snow1977} Snow, T.~P., Jr.\ 1977, \apj,
1770: 216, 724
1771:
1772: \bibitem[Snow et al.(1996)]{snow1996} Snow, T.~P., Black,
1773: J.~H.,
1774: van Dishoeck, E.~F., Burks, G., Crutcher, R.~M., Lutz, B.~L., Hanson,
1775: M.~M., \& Shuping, R.~Y.\ 1996, \apj, 465, 245
1776:
1777: \bibitem[Snow \& McCall(2006)]{snow2006} Snow, T.~P., \& McCall,
1778: B.~J.\ 2006, \araa, 44, 367
1779:
1780: \bibitem[Sofia et al.(2004)]{sofia2004} Sofia, U.~J., Lauroesch,
1781: J.~T., Meyer, D.~M., \& Cartledge, S.~I.~B.\ 2004, \apj, 605, 272
1782:
1783: \bibitem[Sonnentrucker et al.(2002)]{sonn2002}
1784: Sonnentrucker, P., Friedman, S.~D., Welty, D.~E., York, D.~G.,
1785: \& Snow, T.~P.\ 2002, \apj, 576, 241
1786:
1787: \bibitem[Sonnentrucker et al.(2003)]{sonn2003} Sonnentrucker,
1788: P., Friedman, S.~D., Welty, D.~E., York, D.~G., \& Snow, T.~P.\ 2003, \apj,
1789: 596, 350
1790:
1791: \bibitem[Snow et al.(1998)]{snow1998} Snow, T.~P., Hanson,
1792: M.~M., Black, J.~H., van Dishoeck, E.~F., Crutcher, R.~M., \& Lutz,
1793: B.~L.\ 1998, \apjl, 496, L113
1794:
1795: \bibitem[Taylor(1997)]{taylor1997}Taylor, J. R. 1997,
1796: The Study of Uncertainties in Physical Measurements,
1797: (Mill Valley: University Science Books)
1798:
1799: \bibitem[Wannier et al.(1999)]{wannier1999} Wannier, P., Andersson,
1800: B.-G., Penprase, B.~E., \& Federman, S.~R.\ 1999, \apj, 510, 291
1801:
1802: \bibitem[Weingartner \& Draine(2001)]{weingartner2001}
1803: Weingartner, J.~C., \& Draine, B.~T.\ 2001, \apj, 563, 842
1804:
1805: \bibitem[Welty et al.(2003)]{welty2003} Welty, D.~E., Hobbs,
1806: L.~M., \& Morton, D.~C.\ 2003, \apjs, 147, 61
1807:
1808: \bibitem[Wolfire et al.(1995)]{wolfire1995} Wolfire, M.~G.,
1809: Hollenbach, D., McKee, C.~F., Tielens, A.~G.~G.~M., \& Bakes,
1810: E.~L.~O.\ 1995, \apj, 443, 152
1811:
1812: \bibitem[Wolfire et al.(2003)]{wolfire2003} Wolfire, M.~G.,
1813: McKee, C.~F., Hollenbach, D., \& Tielens, A.~G.~G.~M.\ 2003,
1814: \apj, 587, 278
1815:
1816: \bibitem[Zsarg{\'o} \& Federman(2003)]{zsargo2003} Zsarg{\'o}, J.,
1817: \& Federman, S.~R.\ 2003, \apj, 589, 319
1818:
1819: \bibitem[Zsargo et al.(1997)]{zsargo1997} Zsargo, J.,
1820: Federman, S.~R., \& Cardelli, J.~A.\ 1997, \apj, 484, 820
1821: \end{thebibliography}
1822:
1823: \clearpage
1824: \begin{deluxetable}{llcccccccl}
1825: \rotate
1826: \tablewidth{0pt}
1827: \tablecaption{Observations\label{tbl:observations}}
1828: \tablehead{
1829: \colhead{ } &
1830: \colhead{ } &
1831: \colhead{$\log N_{\rm H\,I}$} &
1832: \colhead{$\log N_{\rm H_2}$} &
1833: \colhead{$\log N_{\rm C\,I}$} &
1834: \colhead{$\log N_{\rm C\,II}$} &
1835: \colhead{ } &
1836: \colhead{ } &
1837: \colhead{ } &
1838: \colhead{ } \\
1839: \multicolumn{2}{c}{Star} &
1840: \colhead{$({\rm cm^{-2}})$} &
1841: \colhead{$({\rm cm^{-2}})$} &
1842: \colhead{$({\rm cm^{-2}})$} &
1843: \colhead{$({\rm cm^{-2}})$} &
1844: \colhead{$A_{\rm V}$} &
1845: \colhead{$\log f_{\rm H_2}$\tablenotemark{a}} &
1846: %\colhead{$\log \frac{N_{C\,I}}{N_{\,CII}}$} &
1847: \colhead{$\log f_{\rm C\,I}$\tablenotemark{b}} &
1848: \colhead{References}
1849: }
1850: \startdata
1851: $\kappa$ Cas & HD 2905 & $21.21\pm 0.11$ & $20.27\pm 0.18$
1852: & $15.55\pm 0.20$ & $17.50\pm 0.22$ &
1853: 0.98 & -0.73 & -1.95 &1, 2, 7\\
1854: 40 Per & HD 22951 & $21.05\pm 0.11 $ & $20.46\pm 0.18 $
1855: & $15.25\pm 0.06 $ & $17.43\pm 0.11 $\tablenotemark{b} &
1856: 0.85 & -0.47 & -2.18
1857: &1, 2, 3 \\
1858: $o$ Per & HD 23180 & $20.91\pm 0.11 $ & $20.61\pm 0.15 $
1859: & $15.68\pm 0.06 $ & $17.41\pm 0.11 $ &
1860: 0.79 & -0.30 & -1.73
1861: &1, 2, 3 \\
1862: $\zeta$ Per & HD 24398 & $20.81\pm 0.04 $ & $20.67\pm 0.10 $
1863: & $15.52\pm 0.05 $ & $17.35\pm 0.08 $\tablenotemark{c}&
1864: 0.79 & -0.23 & -1.83
1865: &3, 4, 5, 6 \\
1866: $\chi$ Per & HD 24534 & $20.72\pm 0.06$ & $20.92\pm 0.04$
1867: & $16.30\pm 0.20$ & $17.53\pm 0.14$\tablenotemark{c}
1868: & 1.10 & -0.12 & -1.24 & 6, 17, 22, 24\\
1869: $\xi$ Per & HD 24912 & $21.07\pm 0.06 $ & $20.53\pm 0.08 $
1870: & $15.12\pm 0.14 $ & $17.51\pm 0.13 $\tablenotemark{c}&
1871: 0.98 & -0.44 & -2.39
1872: &3, 4, 6, 7 \\
1873: $\epsilon$ Per & HD 24760 & $20.42\pm 0.06 $ & $19.53\pm 0.15 $
1874: & $13.71\pm 0.05 $ & $16.72\pm 0.07 $ &
1875: 0.16 & -0.69 & -3.01
1876: &4, 8 \\
1877: & HD 34078 & $21.43\pm 0,10$ & $20.81\pm 0.03$
1878: & $16.24\pm 0.44$ & $17.80\pm 0.45$
1879: & 2.00 & -0.49 & -1.56 & 27\\
1880: 23 Ori & HD 35149 & $20.74\pm 0.08 $ & $18.30\pm 0.11 $
1881: & $14.99\pm 0.05 $ & $16.94\pm 0.09 $ &
1882: 0.28 & -2.14 & -1.95
1883: &4, 9, 10 \\
1884: $\delta$ Ori & HD 36486 & $20.19\pm 0.03 $ & $14.74\pm 0.05 $
1885: & $13.02\pm 0.09 $ & $16.39\pm 0.09 $ &
1886: 0.085 & -5.15 & -3.37
1887: &7, 11, 12 \\
1888: $\lambda$ Ori & HD 36861 & $20.79\pm 0.08 $ & $19.11\pm 0.11 $
1889: & $14.32\pm 0.04 $ & $17.00\pm 0.21 $\tablenotemark{c} &
1890: 0.32 & -1.40 & -2.68
1891: &4, 6, 8 \\
1892: $\iota$ Ori & HD 37043 & $20.16\pm 0.05 $ & $14.69\pm 0.20 $\tablenotemark{d}
1893: & $12.98\pm 0.14 $ & $16.36\pm 0.15 $ &
1894: 0.071 & -5.17 & -3.38
1895: &2, 4, 7 \\
1896: $\epsilon$ Ori & HD 37128 & $20.46\pm 0.07$ & $16.28\pm 0.20 $
1897: & $13.59\pm 0.10$ & $16.66\pm 0.12$ &
1898: 0.14 & -3.88 & -3.07 & 4, 7, 12\\
1899: $\zeta$ Ori & HD 37742 & $20.41\pm 0.08 $ & $15.82\pm 0.20$
1900: & $13.88\pm 0.20$ & $16.61\pm 0.22$ &
1901: 0.13 & -4.29 & -2.73 & 1, 9, 10, 13\\
1902: $\mu$ Col & HD 38666 & $19.86\pm 0.08$ & $15.51\pm 0.20$\tablenotemark{d}
1903: & $12.96\pm 0.03$ & $16.06\pm 0.08$ &
1904: 0.036 & -4.05 & -3.10 & 2, 23\\
1905: $\kappa$ Ori & HD 38771 & $20.53\pm 0.04$ & $15.68\pm 0.20$\tablenotemark{d}
1906: & $13.68\pm 0.08$ & $16.73\pm 0.09$ &
1907: 0.17 & -4.55 & -3.05 & 2, 4, 7\\
1908: 139 $\tau$ & HD 40111& $20.90\pm 0.08$ & $19.74\pm 0.20$
1909: & $14.23\pm 0.18$ & $17.16\pm 0.19$ &
1910: 0.46 & -0.92 & -2.93 & 1, 2, 7 \\
1911: 15 Mon & HD 47839 & $20.36\pm 0.06$ & $15.55\pm 0.20$\tablenotemark{d}
1912: & $13.40\pm 0.15$ & $16.56\pm 0.16$ &
1913: 0.13 & -4.51 & -3.16 & 4, 2, 14\\
1914: $\zeta$ Pup & HD 66811 & $19.99\pm 0.02$ & $14.45\pm 0.20$\tablenotemark{d}
1915: & $13.26\pm 0.09$ & $16.19\pm 0.09$ &
1916: 0.049 & -5.24 & -2.93 & 1, 2, 14\\
1917: $\gamma^2$ Vel & HD 68273 & $19.78\pm 0.04$ & $14.23\pm 0.20$\tablenotemark{d}
1918: & $13.12\pm 0.17$ & $15.98\pm 0.17$ &
1919: 0.031 & -5.25 & -2.86 & 1, 2, 14\\
1920: $\rho$ Leo & HD 91316 & $20.26\pm 0.08$ & $15.61\pm 0.20$\tablenotemark{d}
1921: & $13.33\pm 0.22$ & $16.46\pm 0.23$ &
1922: 0.091 & -4.35 & -3.13 & 1, 2, 14\\
1923: % & HD 102065 & $20.54\pm 0.30$\tablenotemark{e} &
1924: % $20.50\pm 0.06$
1925: % & $14.60\pm 0.10$ & $17.19\pm 0.15$ &
1926: % 0.49 & -0.19 & -2.59 & 17, 19\\
1927: & HD 112244 & $21.08\pm 0.08$ & $20.14\pm 0.11$
1928: & $14.69\pm 0.05$ & $17.37\pm 0.08$
1929: & 0.74 & -0.73 & -2.68 & 1, 2, 8\\
1930: 1 Sco & HD 141637 & $21.19\pm 0.08$ & $19.23\pm 0.18$
1931: & $14.00\pm 0.05$ & $17.40\pm 0.09$ &
1932: 0.80 & -1.67 & -3.40 & 1, 2, 8\\
1933: $\pi$ Sco & HD 143018 & $20.72\pm 0.04$ & $19.32\pm 0.20$
1934: & $13.03\pm 0.05$ & $16.95\pm 0.06$ &
1935: 0.28 & -1.13& -3.92 & 1, 2, 8\\
1936: $\delta$ Sco & HD 143275 & $21.14\pm 0.08$ & $19.41\pm 0.20$
1937: & $14.25\pm 0.05$ & $17.36\pm 0.09$ &
1938: 0.74 & -1.45 & -3.11 & 1, 2, 8\\
1939: $\beta^1$ Sco & HD 144217 & $21.10\pm 0.04$ & $19.83\pm 0.06$
1940: & $14.42\pm 0.06$ & $17.37\pm 0.10$\tablenotemark{c}&
1941: 0.70 & -1.01 & -2.92 & 1, 2, 6, 15\\
1942: $\omega^1$ Sco & HD 144470 & $21.18\pm 0.08$ & $20.05\pm 0.11$
1943: & $14.35\pm 0.05$ & $17.44\pm 0.09$ &
1944: 0.88 & -0.89 & -3.09 & 1, 2, 8 \\
1945: $\nu$ Sco & HD 145502 & $21.14\pm 0.15$ & $19.89\pm 0.15$
1946: & $14.56\pm 0.27$ & $17.39\pm 0.30$ &
1947: 0.78 & -1.00 & -2.83& 1, 2, 7\\
1948: $\sigma$ Sco & HD 147165 & $21.34\pm 0.15$ & $19.79\pm 0.15$
1949: & $14.38\pm 0.05$ & $17.56\pm 0.15$ &
1950: 1.15 & -1.27 & -3.18 & 1, 2, 8\\
1951: $\rho$ Oph & HD 147933A & $21.55\pm 0.10$ & $20.57\pm 0.15$
1952: & $15.52\pm 0.02$ & $17.83\pm 0.09$ &
1953: 2.13 &-0.76 & -2.31 & 2, 15, 16\\
1954: $\chi$ Oph & HD 148184 & $21.14\pm 0.08$ & $20.63\pm 0.18$
1955: & $15.33\pm 0.01$ & $17.55\pm 0.09$ &
1956: 1.15 & -0.42 & -2.22 & 1, 2, 15\\
1957: 22 Sco & HD 148605 & $20.95\pm 0.08$ & $18.74\pm 0.18$
1958: & $13.95\pm 0.50$ & $17.16\pm 0.51$ &
1959: 0.46 & -1.92 & -3.21 & 1, 2, 7\\
1960: $\zeta$ Oph & HD 149757 & $20.72\pm 0.02$ & $20.65\pm 0.05$
1961: & $15.51\pm 0.07$ & $17.36\pm 0.11$\tablenotemark{c} &
1962: 0.71 & -0.20 & -1.84 & 2, 4, 6\\
1963: & HD 154368 & $21.00\pm 0.05$ & $21.16\pm 0.07$
1964: & $16.22\pm 0.20$ & $17.79\pm 0.21$ &
1965: 1.96 & -0.13 & -1.57 & 17, 18\\
1966: $\gamma$ Ara & HD 157246 & $20.71\pm 0.06$ & $19.24\pm 0.13$
1967: & $13.87\pm 0.12$ & $16.94\pm 0.13$ &
1968: 0.26 & -1.20 & -3.07 & 4, 7\\
1969: $\kappa$ Aql & HD 184915 & $20.90\pm 0.11$ & $20.31\pm 0.15$
1970: & $14.37\pm 0.16$ & $17.28\pm 0.18$ &
1971: 0.61 & -0.47 & -2.91 & 1, 2, 7\\
1972: & HD 185418 & $21.11\pm 0.15$ & $20.76\pm 0.05$
1973: & $15.57\pm 0.09$ & $17.59\pm 0.12$ &
1974: 1.24 & -0.33 & -2.02 &17, 25, 26 \\
1975: & HD 192639 & $21.29\pm 0.09$ & $20.69\pm 0.05$
1976: & $15.37\pm 0.08$ & $17.67\pm 0.10$ &
1977: 1.55 & -0.48 & -2.30 & 4, 17, 20\\
1978: 59 Cyg & HD 200120 & $20.25\pm 0.08$ & $19.32\pm 0.18$
1979: & $13.94\pm 0.15$ & $16.54\pm 0.17$ &
1980: 0.11 & -0.72 & -2.60 & 1, 2, 7\\
1981: & HD 206267 & $21.31\pm 0.15$ & $20.86\pm 0.04$
1982: & $15.30\pm 0.08$\tablenotemark{e} & $17.74\pm 0.12$ &
1983: 1.74 & -0.38 & -2.44 & 17, 21\\
1984: & HD 210839 & $21.16\pm 0.10$ & $20.84\pm 0.04$
1985: & $14.98\pm 0.08$ & $17.65\pm 0.10$ &
1986: 1.42 & -0.31 & -2.67 & 17, 21, 22\\
1987: $\sigma$ Cas & HD 224572 & $20.88\pm 0.08$ & $20.23\pm 0.18$
1988: & $14.74\pm 0.19$ & $17.24\pm 0.21$ &
1989: 0.55 & -0.51 & -2.50 & 1, 2, 7\\
1990:
1991: \enddata
1992: \tablerefs{
1993: (1)~\cite{bohlin1978}; (2)~\cite{savage1977}; (3)~\cite{wannier1999};
1994: (4)~\cite{cartledge2004};
1995: (5)~\cite{snow1977}; (6)~\cite{sofia2004}; (7)~\cite{jenkins1983};
1996: (8)~\cite{zsargo2003}; (9)~\cite{welty2003}; (10)~Compilation of column densities by
1997: D. Welty http://astro.uchicago.edu/home/web/welty/coldens.html; (11)~\cite{jenkins1999};
1998: (12)~\cite{jenkins2000}; (13)~\cite{jenkins1997};
1999: (14)~\cite{jenkins1979}; (15)~\cite{zsargo1997};
2000: (16)~\cite{shull1985}; (17)~\cite{rachford2002}; (18)~\cite{snow1996};
2001: (19)~\cite{gry2002}; (20)~\cite{sonn2002}; (21)~\cite{jt2001};
2002: (22)~\cite{diplas1994}; (23)~\cite{howk1999}; (24)~\cite{snow1998};
2003: (25)~\cite{fitz1990}; (26)~\cite{sonn2003}; (27)~\cite{boisse2005}
2004: }
2005: \tablenotetext{a}{$f_{\rm H_2} = 2N_{\rm H_2}/[\hicol + 2N_{\rm H_2}]$.}
2006: \tablenotetext{b}{$f_{\rm C\,I} = \cicol/\ciicol$.}
2007: \tablenotetext{b}{$\ciicol$ is taken as $1.6\times 10^{-4}\times [ \hicol + 2N_{\rm H_2} ]$
2008: unless otherwise noted.}
2009: \tablenotetext{c}{Observed $\ciicol$.}
2010: \tablenotetext{d}{Uncertainties for low $N_{\rm H_2}$ columns were
2011: not given in \cite{savage1977}.
2012: We use the uncertainty suggested by Welty [Reference (10)].}
2013: %\tablenotetext{e}{\ion{H}{1} column from \cite{rachford2002} estimated
2014: % from E(B-V)}
2015: \tablenotetext{f}{\ion{C}{1} column uncertainty assumed to be similar
2016: to \cite{sonn2002}.}
2017:
2018: %}
2019:
2020: \end{deluxetable}
2021:
2022: \begin{deluxetable}{ll}
2023: %\rotate
2024: %\tablewidth{0pt}
2025: \tablecaption{Reaction Rates\label{tbl:reactions}}
2026: \tablehead{
2027: \colhead{ } &
2028: \colhead{ Rate } \\
2029: \colhead{Reaction} &
2030: \colhead{(${\rm cm^{3}}$ ${\rm s^{-1}}$)}
2031: }
2032: \startdata
2033: ${\rm PAH^-} + {\rm H^+} \rightarrow {\rm PAH^0} + {\rm H}$
2034: & $8.3\times 10^{-7} \Phi_{\rm PAH} T_2^{-0.5}$ \tablenotemark{a}\\
2035: ${\rm PAH^0} + {\rm H^+} \rightarrow {\rm PAH^+} + {\rm H}$
2036: & $3.1\times 10^{-8} \Phi_{\rm PAH}$ \\
2037: ${\rm PAH^+} + e \rightarrow {\rm PAH^0}$
2038: & $3.5\times 10^{-5} \Phi_{\rm PAH} T_2^{-0.5}$\\
2039: ${\rm PAH^0} + e \rightarrow {\rm PAH^-} $
2040: & $1.3\times 10^{-6} \Phi_{\rm PAH}$ \\
2041: ${\rm PAH^-} +{\rm C^+} \rightarrow {\rm PAH^0} + {\rm C}$
2042: &$2.4\times 10^{-7}\Phi_{\rm PAH} T_2^{-0.5}$ \tablenotemark{b} \\
2043: ${\rm PAH^0} + {\rm C^+}\rightarrow {\rm PAH^+} + {\rm C}$
2044: & $8.8\times 10^{-9}\Phi_{\rm PAH}$ \tablenotemark{b} \\
2045: ${\rm PAH^0} + h\nu \rightarrow {\rm PAH^+} + e$ &
2046: $4.6\times 10^{-9}G_0\exp (-1.8 A_{\rm V})$ \tablenotemark{c}\\
2047: ${\rm PAH^-} + h\nu \rightarrow {\rm PAH^0} + e$ &
2048: $1.2\times 10^{-8}G_0\exp (-1.8 A_{\rm V})$ \tablenotemark{c}\\
2049: ${\rm C} + h\nu \rightarrow {\rm C^+} + e$ &
2050: $2.1\times 10^{-10}G_0\exp (-2.6 A_{\rm V})$ \tablenotemark{c}\\
2051: ${\rm C^+} + e\rightarrow {\rm C} + h\nu$ &
2052: $1.8\times 10^{-11}T_2^{-0.83}$ \\
2053: \enddata
2054: %\tablerefs{}
2055: \tablenotetext{a}{$T_2 = T/100$ K.}
2056: \tablenotetext{b}{Additional collisonal rates scale as $(m)^{-0.5}$
2057: where $m$ is the mass of the collision partner.}
2058: \tablenotetext{c}{Photo rates have units ${\rm s^{-1}}$.
2059: $G_0$ is the FUV field measured in units of the \cite{habing1968}
2060: field ($=1.6\times 10^{-4}$ erg cm$^{-2}$ s$^{-1}$ sr$^{-1}$). The
2061: \cite{draine1978} field is a factor of 1.7 larger. }
2062: \end{deluxetable}
2063:
2064: \clearpage
2065: \begin{figure}
2066: %\plotone{ratioav025v2cphc.eps}
2067: \plotone{f1.eps}
2068: \caption{
2069: Column density ratio $\cicol/\ciicol$ versus molecular
2070: hydrogen fraction $f_{\rm H_2} = 2N_{\rm H_2}/[\hicol + 2N_{\rm
2071: H_2}]$. Observations are shown for cloud column densities
2072: in the range $0.03 \la A_{\rm V} \la 0.25$, $N_{\rm H_2} < 10^{17}$
2073: cm$^{-2}$, and molecular fraction
2074: $f_{\rm H_2} \la 10^{-4}$.
2075: Clouds with $A_{\rm V} < 0.1$ are indicated with an
2076: ``$\times$''.
2077: Curves show model results for cloud column $A_{\rm V} =
2078: 0.1$, and various values of
2079: $\Phi_{\rm PAH}/R_{-17}$ where $R_{-17} = R/10^{-17}$.
2080: $\Phi_{\rm PAH} = 2.0$, $R = 2\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
2081: ($\Phi_{\rm PAH}/R = 1$; {\em dot}),
2082: $\Phi_{\rm PAH} = 1$, $R = 1\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
2083: ($\Phi_{\rm PAH}/R = 1$; {\em dash-dot-dot}),
2084: $\Phi_{\rm PAH} = 0.5$, $R = 1\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
2085: ($\Phi_{\rm PAH}/R = 1/2$; {\em long dash}),
2086: $\Phi_{\rm PAH} = 1$, $R = 2\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
2087: ($\Phi_{\rm PAH}/R = 1/2$; {\em dash-dot}),
2088: $\Phi_{\rm PAH}=1$, $R = 3\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
2089: ($\Phi_{\rm PAH}/R = 1/3$;
2090: {\em short dash})), $\Phi_{\rm PAH}=0.5$,
2091: $R = 3\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
2092: ($\Phi_{\rm PAH}/R = 1/6$; {\em solid}), and
2093: $\Phi_{\rm PAH}=0.01$, $R = 3\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
2094: ($\Phi_{\rm PAH}/R = 1/300$; {\em dot}; lower curves).
2095: The ratio $G_0/n$ varies along each model curve
2096: with higher
2097: $G_0/n$ yielding smaller values of $f_{\rm H_2}$.
2098: Individual models are shown with $n = 10$, 20, 30, and 40
2099: ${\rm cm^{-3}}$
2100: (with the exception of the {\em dash-dot} and {\em dash-dot-dot} curves where
2101: $n=5$, 10, and 20 ${\rm cm^{-3}}$)
2102: and $G_0 = 1.7$ ($\Diamond$; {\em Blue}) and
2103: $G_0 = 5.1$ ($\triangle$; {\em Red}) .
2104: }
2105: \label{fig:ratioav025}
2106: \end{figure}
2107:
2108: \begin{figure}
2109: %\plotone{ratioav075v2cphc.eps}
2110: \plotone{f2.eps}
2111: \caption{Column density ratio $\cicol/\ciicol$ versus molecular
2112: hydrogen fraction $f_{\rm H_2} = 2N_{\rm H_2}/[\hicol + 2N_{\rm
2113: H_2}]$. Observations are shown for cloud column densities
2114: in the range $0.25 \la A_{\rm V} \la 0.75$, and $N_{\rm H_2} > 10^{18}$
2115: cm$^{-2}$.
2116: %We have also added the
2117: %$\epsilon$ Per ($A_{\rm V} = 0.16$, $f_{\rm H_2} = 0.20$)
2118: %and 59 Cyg ($A_{\rm V} = 0.11$, $f_{\rm H_2} = 0.19$) lines
2119: %of sight to this figure.
2120: Clouds with $A_{\rm V} < 0.5$ are indicated with an
2121: ``$\times$''.
2122: Curves show model results for cloud column $A_{\rm V} =
2123: 0.5$, $R=3\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$,
2124: and $\Phi_{\rm PAH} = 1.0$ ({\em dash}),
2125: $\Phi_{\rm PAH} = 0.5$ ({\em solid}), and
2126: $\Phi_{\rm PAH} = 0.01$ ({\em dot}). The ratio
2127: $G_0/n$ varies along each model curve
2128: with higher
2129: $G_0/n$ yielding smaller values of $f_{\rm H_2}$.
2130: Individual models are shown with $n = 10$, 30, 100,
2131: and 300 ${\rm cm^{-3}}$ and $G_0 = 1.7$ ($\Diamond$; {\em Blue}) and
2132: $G_0 = 5.1$ ($\triangle$; {\em Red}). For $f_{\rm H_2} = 0.02$,
2133: $G_0/n \sim 1.7/10$ and for $f_{\rm H_2} = 0.2$,
2134: $G_0/n \sim 1.7/30$.
2135: }
2136: \label{fig:ratioav075}
2137: \end{figure}
2138:
2139: \begin{figure}
2140: %\plotone{ratioav125v2cphp.eps}
2141: \plotone{f3.eps}
2142: \caption{Column density ratio $\cicol/\ciicol$ versus molecular
2143: hydrogen fraction $f_{\rm H_2} = 2N_{\rm H_2}/[\hicol + 2N_{\rm
2144: H_2}]$. Observations are shown for cloud column densities
2145: in the range $0.75 \la A_{\rm V} \la 1.25$ and $N_{\rm H_2} > 10^{18}$
2146: cm$^{-2}$.
2147: Clouds with $A_{\rm V} < 1.0$ are indicated with an
2148: ``$\times$''.
2149: Curves show model results for cloud column $A_{\rm V} =
2150: 1.0$, $R=3\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$,
2151: and $\Phi_{\rm PAH} = 1.0$ ({\em dash}),
2152: $\Phi_{\rm PAH} = 0.5$ ({\em solid}), and
2153: $\Phi_{\rm PAH} = 0.01$ ({\em dot}). The ratio
2154: $G_0/n$ varies along each model curve
2155: with higher
2156: $G_0/n$ yielding smaller values of $f_{\rm H_2}$.
2157: Individual models are shown with $n = 10$, 30, 100,
2158: and 300 ${\rm cm^{-3}}$ and $G_0 = 1.7$ ($\Diamond$; {\em Blue}) and
2159: $G_0 = 5.1$ ($\triangle$; {\em Red}) .
2160: }
2161: \label{fig:ratioav125}
2162: \end{figure}
2163:
2164: \begin{figure}
2165: %\plotone{ratioav213v2cphc.eps}
2166: \plotone{f4.eps}
2167: \caption{Column density ratio $\cicol/\ciicol$ versus molecular
2168: hydrogen fraction
2169: $f_{\rm H_2} = 2N_{\rm H_2}/[\hicol + 2N_{\rm H_2}]$.
2170: Observations are shown for cloud column densities
2171: in the range $1.25 \la A_{\rm V} \la 2.13$ and $N_{\rm H_2} > 10^{18}$
2172: cm$^{-2}$.
2173: Clouds with $A_{\rm V} < 1.5$ are indicated with an
2174: ``$\times$''.
2175: Curves show model results for cloud column $A_{\rm V} =
2176: 1.5$, $R=3\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$,
2177: and $\Phi_{\rm PAH} = 1.0$ ({\em dash}),
2178: $\Phi_{\rm PAH} = 0.5$ ({\em solid}), and
2179: $\Phi_{\rm PAH} = 0.01$ ({\em dot}). The ratio
2180: $G_0/n$ varies along each model curve
2181: with higher
2182: $G_0/n$ yielding smaller values of $f_{\rm H_2}$.
2183: Individual models are shown with $n = 10$, 30, 100,
2184: and 300 ${\rm cm^{-3}}$ and $G_0 = 1.7$ ($\Diamond$; {\em Blue}) and
2185: $G_0 = 5.1$ ($\triangle$; {\em Red}) .
2186: }
2187: \label{fig:ratioav213}
2188: \end{figure}
2189:
2190: \begin{figure}
2191: %\plotone{chiav025v2cphp.eps}
2192: \plotone{f5.eps}
2193: \caption{$\chi^2$ plot of $\Phi_{\rm PAH}$ versus ${\rm H_2}$
2194: formation rate $R$. Contour levels are
2195: $\chi^2 = 2$, 3, 4, 5, 6, 8, and 10.
2196: Observations are restricted to cloud column densities
2197: in the range $0.03 \la A_{\rm V} \la 0.25$ and
2198: $N_{\rm H_2} < 10^{17}$ cm$^{-2}$.
2199: The minimum value of $\chi^2_{\rm min} = 1.6$ with
2200: 10 sources included.
2201: }
2202: \label{fig:chiav025}
2203: \end{figure}
2204:
2205: %GOOD BYE FIGURES! Hello again!
2206: \begin{figure}
2207: %\plotone{chiav025075cphc.eps}
2208: \plotone{f6.eps}
2209: \caption{$\chi^2$ plot of $\Phi_{\rm PAH}$ versus ${\rm H_2}$
2210: formation rate $R$. Contour levels are
2211: $\chi^2 = 1.5$, 2, 3, 4, 5, 6, 8, and 10.
2212: Observations are restricted to cloud column densities
2213: in the range $0.25 \la A_{\rm V} \la 0.75$ and
2214: $N_{\rm H_2} > 10^{18}$ cm$^{-2}$.
2215: %An asterisk
2216: %($\ast$) marks the minimum value of $\chi^2_{\rm min} = 1.1$ with
2217: %10 sources included.
2218: }
2219: \label{fig:chiav075}
2220: \end{figure}
2221:
2222: %SO LONG, HAVE A GOOD TRIP Back again!
2223: \begin{figure}
2224: %\plotone{chiav075125v2cphc.eps}
2225: \plotone{f7.eps}
2226: \caption{$\chi^2$ plot of $\Phi_{\rm PAH}$ versus ${\rm H_2}$
2227: formation rate $R$. Contour levels are
2228: $\chi^2 = 1$, 1.5, 2, 3, 4, 5, 6, 8, and 10.
2229: Observations are restricted to cloud column densities
2230: in the range $0.75 \la A_{\rm V} \la 1.25$ and
2231: $N_{\rm H_2} > 10^{18}$ cm$^{-2}$.
2232: %An asterisk
2233: %($\ast$) marks the minimum value of $\chi^2_{\rm min} = 0.76$ with
2234: %12 sources included.
2235: }
2236: \label{fig:chiav125}
2237: \end{figure}
2238:
2239: %BYE DONT FORGET TO WRITE
2240: %\begin{figure}
2241: %\plotone{chiav213.eps}
2242: %\caption{$\chi^2$ plot of $\Phi_{\rm PAH}$ versus ${\rm H_2}$
2243: %formation rate $R$. Contour levels are
2244: %$\chi^2 = 2$,, 3, 4, 5, 6, 8, and 10.
2245: %Observations are restricted to cloud column densities
2246: %in the range $1.25 < A_{\rm V} < 2.13$. An asterisk
2247: %($\ast$) marks the minimum value of $\chi^2_{\rm min} = 1.6$ with
2248: %5 sources included.
2249: %}
2250: %\label{fig:chiav213}
2251: %\end{figure}
2252:
2253: %\begin{figure}
2254: %\plotone{chiav025125.eps}
2255: %\caption{$\chi^2$ plot of $\Phi_{\rm PAH}$ versus ${\rm H_2}$
2256: %formation rate $R$. Contour levels are
2257: %$\chi^2 = 1$, 1.5, 2, 3, 4, 5, 6, 8, and 10.
2258: %Observations are restricted to cloud column densities
2259: %in the range $0.25 < A_{\rm V} < 1.25$. An asterisk
2260: %($\ast$) marks the minimum value of $\chi^2_{\rm min} = 0.93$ with
2261: %22 sources included.
2262: %}
2263: %\label{fig:chiav025125}
2264: %\end{figure}
2265:
2266:
2267: %\begin{figure}
2268: %\plotone{chiav025213v2.eps}
2269: %\caption{$\chi^2$ plot of $\Phi_{\rm PAH}$ versus ${\rm H_2}$
2270: %formation rate $R$. Contour levels are
2271: %$\chi^2 = 2.8$, 3, 4, 5, 6, 8, and 10.
2272: %Observations are restricted to cloud column densities
2273: %in the range $0.25 < A_{\rm V} < 2.13$.
2274: %An asterisk
2275: %($\ast$) marks the minimum value of $\chi^2_{\rm min} = 1.5$ with
2276: %28 sources included.
2277: %The minimum trough is given by
2278: %$\Phi_{\rm PAH} = 0.18 (R/10^{-17}\,\,
2279: %{\rm cm^{3}}\,\, {\rm s^{-1}}) -0.22$
2280: %}
2281: %\label{fig:chiav025213}
2282: %\end{figure}
2283:
2284: \begin{figure}
2285: %\plotone{chiav025213v2cphc.eps}
2286: \plotone{f8.eps}
2287: \caption{$\chi^2$ plot of $\Phi_{\rm PAH}$ versus ${\rm H_2}$
2288: formation rate $R$. Contour levels are
2289: $\chi^2 = 1.6$, 2, 3, 4, 5, 6, 8, and 10.
2290: Observations are restricted to cloud column densities
2291: in the range $0.25 \la A_{\rm V} \la 2.13$ and $N_{\rm H_2} > 10^{18}$
2292: cm$^{-2}$.
2293: %An asterisk
2294: %($\ast$) marks the minimum value of $\chi^2_{\rm min} = 1.5$ with
2295: %28 sources included.
2296: The minimum trough is given by
2297: $\Phi_{\rm PAH} = 0.22 (R/10^{-17}\,\,
2298: {\rm cm^{3}}\,\, {\rm s^{-1}}) -0.40$
2299: }
2300: \label{fig:chiav025213noeps}
2301: \end{figure}
2302:
2303: %\begin{figure}
2304: %\plotone{denav025213v2.eps}
2305: %\caption{Mean density plot of $\Phi_{\rm PAH}$ versus ${\rm H_2}$
2306: %formation rate $R$. Contour levels are
2307: %$< n > = 15$, 20, 30, 40, 50, 60, 70, 80, 90, and 100
2308: %${\rm cm^{-3}}$.
2309: %Observations are restricted to cloud column densities
2310: %in the range $0.25 < A_{\rm V} < 2.13$
2311: %plus
2312: %$\epsilon$ Per ($A_{\rm V} = 0.16$)
2313: %and 59 Cyg ($A_{\rm V} = 0.11$).
2314: %The minimum trough from
2315: %Fig.\ (\ref{fig:chiav025213})
2316: %is given by
2317: %$\Phi_{\rm PAH} = 0.18 (R/10^{-17}\,\,
2318: %%{\rm cm^{3}}\,\, {\rm s^{-1}}) -0.22$
2319: %}
2320: %\label{fig:denav025213}
2321: %\end{figure}
2322:
2323: \begin{figure}
2324: %\plotone{denav025cp.eps}
2325: \plotone{f9.eps}
2326: \caption{Mean density plot of $\Phi_{\rm PAH}$ versus ${\rm H_2}$
2327: formation rate $R$. Contour levels are
2328: $< n > = 5$, 10, 15, 20, and 30
2329: ${\rm cm^{-3}}$.
2330: Observations are restricted to cloud column densities
2331: in the range $0.03 \la A_{\rm V} \la 0.25$ and $N_{\rm H_2} < 10^{17}$
2332: cm$^{-2}$.
2333: }
2334: \label{fig:denav025cp}
2335: \end{figure}
2336:
2337: \begin{figure}
2338: %\plotone{denav025213v3cphc.eps}
2339: \plotone{f10.eps}
2340: \caption{Mean density plot of $\Phi_{\rm PAH}$ versus ${\rm H_2}$
2341: formation rate $R$. Contour levels are
2342: $< n > = 15$, 20, 30, 40, 50, 60, 70, 80, 90, and 100
2343: ${\rm cm^{-3}}$.
2344: Observations are restricted to cloud column densities
2345: in the range $0.25 \la A_{\rm V} \la 2.13$ and $N_{\rm H_2}>10^{18}$.
2346: The minimum trough from
2347: Fig.\ (\ref{fig:chiav025213noeps})
2348: is given by
2349: $\Phi_{\rm PAH} = 0.22 (R/10^{-17}\,\,
2350: {\rm cm^{3}}\,\, {\rm s^{-1}}) -0.40$
2351: }
2352: \label{fig:denav025213noeps}
2353: \end{figure}
2354:
2355: \begin{figure}
2356: %\plotone{chiav025213v2hc.eps}
2357: \plotone{f11.eps}
2358: \caption{$\chi^2$ plot of $\Phi_{\rm PAH}$ versus ${\rm H_2}$
2359: formation rate $R$ using ${\rm C^+}$ recombination
2360: rate from \cite{nahar1997}.
2361: Contour levels are
2362: $\chi^2 = 1.7$, 2, 3, 4, 5, 6, 8, and 10.
2363: Observations are restricted to cloud column densities
2364: in the range $0.25 \la A_{\rm V} \la 2.13$ and $N_{\rm H_2} > 10^{18}$
2365: cm$^{-2}$.
2366: %An asterisk
2367: %($\ast$) marks the minimum value of $\chi^2_{\rm min} = 1.6$ with
2368: %28 sources included.
2369: }
2370: \label{fig:chiav025213noepsnocp}
2371: \end{figure}
2372:
2373: \begin{figure}
2374: %\plotone{ratioav025varv3cphp.eps}
2375: \plotone{f12.eps}
2376: \caption{
2377: Column density ratio $\cicol/\ciicol$ versus molecular
2378: hydrogen fraction $f_{\rm H_2} = 2N_{\rm H_2}/[\hicol + 2N_{\rm
2379: H_2}]$. Observations are shown for cloud column densities
2380: in the range $0.03 \la A_{\rm V} \la 0.25$ and $N_{\rm H_2} < 10^{17}$
2381: cm$^{-2}$.
2382: Clouds with $A_{\rm V} < 0.1$ are indicated with an
2383: ``$\times$''.
2384: Curves show model results for cloud column $A_{\rm V} =
2385: 0.1$, and
2386: %$\Phi_{\rm PAH} = 0.5$, $R = 1\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
2387: %($\Phi_{\rm PAH}/R = 1/2$; {\em long dash}),
2388: %$\Phi_{\rm PAH} = 1$, $R = 2\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
2389: %($\Phi_{\rm PAH}/R = 1/2$; {\em dash-dot}),
2390: %$\Phi_{\rm PAH}=1$, $R = 3\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
2391: %($\Phi_{\rm PAH}/R = 1/3$;
2392: %{\em short dash})),
2393: $\Phi_{\rm PAH}=0.5$,
2394: $R = 3\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
2395: ($\Phi_{\rm PAH}/R = 1/6$; {\em solid}).
2396: % and
2397: %$\Phi_{\rm PAH}=0.01$, $R = 3\times 10^{-17}$ ${\rm cm^{3}}$ ${\rm s^{-1}}$
2398: %($\Phi_{\rm PAH}/R = 1/300$; {\em dot}).
2399: The ratio $G_0/n$ varies along each model curve
2400: with higher
2401: $G_0/n$ yielding smaller values of $f_{\rm H_2}$.
2402: Individual models are shown with $n = 10$, 20, 30, and 40
2403: ${\rm cm^{-3}}$
2404: %(with the exception of the {\em dash dot} curve with
2405: %$n=5$, 10, and 20 ${\rm cm^{-3}}$)
2406: %and $G_0 = 1.7$ ($\Diamond$) and
2407: and $G_0 = 5.1$ ($\triangle$) .
2408: Arrows
2409: show the affects of
2410: lines of sight which graze a larger ($A_{\rm V}=1$) cloud ('G'),
2411: have cosmic rays enhanced by a factor of 10 ('C'), and have half
2412: the column density in the WNM phase ('W').
2413: }
2414: \label{fig:ratioav025var}
2415: \end{figure}
2416: \end{document}
2417:
2418:
2419:
2420:
2421:
2422: