The radiative association rate coefficients and their temperature dependences have been estimated for several likely interstellar ion-molecule reactions from laboratory collisional association rate data. They include the CH3+ + H2 and CH3+ + H2O reactions, which we suggest lead to CH4 and CH3OH respectively, and the critical association reaction C+ + H2.

A possible mechanism for the formation of interstellar linear molecules is studied experimentally by means of synthesis apparatus. C2 and CN radicals are abundantly formed from plasmas containing C, N and H atoms. The C2 and CN radicals survive electron bombardment. They collide with each other and recombine to form linear molecules.

A chemical evolutionary model with a large number of species and a large chemical library is used to examine the principal chemical processes in interstellar clouds. Simple chemical equilibrium arguments show the potential for synthesis of very complex organic species by ion-molecule radiative association reactions.

Molecular abundances were calculated time-dependently using a chemical scheme which was carefully constructed to represent the molecular evolution in dense clouds. All ion-neutral reactions of species containing up to 2 heavy atoms (C, N, O, S and Si) were surveyed and a set of possible reactions was selected from numerous exothermic reactions. As molecules containing 3 to 4 heavy atoms are produced by condensation reactions, the scheme was extended. As a result, 234 species, their 2884 gas-phase reactions, and H recombination on grain surfaces were included in our chemical scheme.

A new method for estimating electron fractions in shielded molecular clouds is proposed on the basis of gas phase ion-molecule reactions which involves measuring the quantity . Applied to existing data, it yields upper limits to Xe in the range from 10−8 to 10−7 for a variety of clouds, warm as well as cool. An upper bound to the cosmic ray ionization rate is also obtained.

A discussion is given of the general processes and considerations that arise in attempting to understand molecular reactions in cool, dense interstellar clouds. Basic elements of the gas phase, “ion-molecule” scheme are given explicitly before surveying topics in which there is considerable current activity. These topics include: (i) refined comparisons of prediction and observation for species of “intermediate” complexity, (ii) numerical computations of cloud models which include numerous chemical reactions and molecular species, (iii) formation of the complex molecules, (iv) isotope fractionation in interstellar molecules and (v) possible contributions from chemistry in shocks.

Laboratory and theoretical studies have been made of the effects of ultraviolet photolysis of interstellar grain mantles. It has been shown that grain photolysis should be important even in dense clouds. A large number of molecules and radicals observed in the interstellar gas appear in the irradiated ices of CO, H2O, NH3 and CH4 which are deposited at 10 K. Energy released during warm-up is seen from visible and infrared luminescence and inferred from vapor pressure enhancement relative to unirradiated samples. Grains are pictured as a source as well as a sink (capture) of molecules. The photolysis of an individual grain provides the stored chemical energy which is sporadically released by relatively mild triggering events (such as low velocity grain-grain collisions in turbulent molecular clouds) to produce the impulsive heating needed to eject or evaporate a portion of the grain mantle. An extremely complex and non-volatile substance possessing the infrared signatures of amino and carboxylic acid groups and having a mass of 514 amu has been produced at a rate corresponding to a mass conversion rate of interstellar grains of between 2% and 20% in 107 years.

Condensation temperatures of various interstellar molecules in dark clouds are calculated. Chemical reactions in dust mantles are discussed.

Cold hydrogen atoms at T ≥ 7K were shown experimentally to react with graphite grains at the same temperature to produce CH4 and smaller amounts of C2H6, C2H4 and C2H2. At T < 20K the hydrocarbon mantle could polymerize to form carbonaceous substances, similar to those found in carbonaceous chondrites. Further encounters with H-atoms would result in their recombination on the hydrocarbon mantle around the grains. At higher grain temperatures, the hydrocarbons formed could be ejected into the gas phase.

Cold iron atoms at T < 5K were shown experimentally to react with molecular hydrogen in a T < 5K matrix. Mössbauer studies with 57Fe demonstrated the formation of an Fe-H2 bond. FeH2 and FeH molecules could be formed on grains by encounters of iron atoms with either H-atoms or H2 molecules.

The composition and physical properties of interstellar grain mantles continues to be an important problem in astrophysics. Part of this importance comes from the fact that grain mantle composition, photochemistry and photophysics are involved in interstellar chemistry (Greenberg et al. 1972, Greenberg, 1979, Greenberg et al. this volume). Because most molecules have a number of fundamental modes of vibration which possess activity between 2.5 and 25 μm (the middle infrared), spectroscopic measurement in this region can provide a direct probe of the molecules making up grain mantles. In addition to molecular composition, under favorable conditions, such measurements can yield molecular abundances, the solid/gas ratio for specific molecules and give an indication of such physical grain properties as temperature and thermal history.

Spectra of infrared sources associated with molecular clouds have shown absorption features at wavelengths of 6.0 and 6.8 μm. We suggest that the 6.0 μm feature can be identified with the stretching vibration of C=0 and the 6.8 μm feature with the bending vibrations of CH2 and CH3. The amount of carbon in the form of hydrocarbon molecules may be comparable to the amount in CO. This abundance of hydrocarbons is probably too large to be consistent with radio observations if the molecules are gaseous, but large abundances of hydrocarbons on the surfaces of grains may explain the infrared features, yet be unobservable in the radio.

The near-infrared spectrum of many sources associated with molecular clouds shows a broad absorption feature at 3.08 μm (e.g. Merrill et al., 1976; Harris et al., 1978). This feature has usually been attributed to absorption by H2O ice frozen on grains, but it has been impossible to satisfactorily reproduce the observed band shape (Merrill et al., 1976; Mukai et al., 1978). We have been able to obtain a complete fit of this absorption feature in the laboratory using very low temperature mixtures of H2O with other polar molecules. The preparation of these interstellar dust grain-mantle analogs has been described elsewhere (Greenberg, 1979; Hagen et al., 1979). They are prepared by allowing a gas mixture of simple molecules (e.g. CO, H2O, NH3, CH4 etc.) to condense on a low temperature (10 K) substrate. This frozen mixture can be heated and recooled. The samples are analyzed with an infrared spectrometer.

The λ2200 extinction feature has been measured from ANS observations of 30 stars. For each star, the depth of the λ2200 feature is compared to EB-V and the equivalent width of the diffuse band λ4430.

A good correlation appears out to EB-V = 1.13. The various mechanisms for producing the diffuse and λ2200 features are discussed, and a discriminatory test is put forward based on observations of the Magellanic Clouds.

Results are presented from a programme of optical spectroscopy and related studies which has the double purpose of investigating the structure of diffuse molecular clouds and of establishing tighter correlations for the unidentified interstellar diffuse bands.

While an examination of the available data reveals some seemingly contradictory results, a general framework having the following outlines can be put forward:

1. With the exception of the two galactic center sources SgrA and SgrB, the relative isotopic abundances exhibited by the giant molecular clouds in our Galaxy exhibit few, if any, significant variations from the values obtained by averaging the data from all these sources.

2. The 13C/12C and 14N/15N abundance ratios are ∼130% and ∼150%, respectively, of their terrestrial values throughout the galactic plane and somewhat higher, ∼300%, near the galactic center.

3. The 16O/18O and 17O/18O abundance ratios are ∼130% and ∼160%, respectively, of their terrestrial values throughout the Galaxy, although the former may be somewhat lower near the galactic center.

4. The S and Si isotopes have generally terrestrial abundances.

This report is a summary of the determination of the ratio of the column density of H212C16O (hereafter H2CO) or H212C18O (hereafter H2C18O) to that of H213C16O (hereafter H213CO). With one exception, all of the published ratios have been determined from measurements of the 110−111 lines. The exception is the Orion Molecular Cloud (see discussion of Kutner et al. 1976). The most complete surveys of H213CO are those of Wilson et al. (1976) and Gardner and Whiteoak (1979). The ratios obtained from 110−111 data are lower limits because of the effect of photon trapping in H2CO (see the discussion by Henkel et al. 1979b). From measurements of the 110−111 and 211−212 lines and model calculations, Henkel et al. estimate that the corrections for 10 clouds outside the Galactic center region is 1.6±0.7. This correction gives an average (H2CO/H213CO) ratio of 64±17. The models of Henkel et al. also predict the optical depth of the 312−313 line and the brightness temperature of the 212−1 11 line. Preliminary measurements of Wilson et al. (1979) show that the 312−313 results are consistent with the predictions; but more sensitive measurements are required to test the model fully. Unpublished 212− 1 11 emission-line spectra are weaker than predicted by the model, but this might be caused by beam dilution, since the observations were made so as to cover a 1.5′ uniformly weighted aperture. The lineshapes of the 110−111 and 2 11−212 absorption lines agree poorly with the 212−111 emission lines, indicating either that self-absorption in the millimeter lines of H2CO is important, or that there are H2CO clouds behind the continuum sources. Because of their position relative to the continuum sources these clouds would contribute to the millimeter emission lines but not to the centimeter absorption lines.

Formaldehyde is an excellent molecule for studying isotopic abundances. It is easily observed in several connected transitions, allowing an analysis of excitation differences among species. With formaldehyde, one can also measure the same ratio using millimeter emission lines and centimeter absorption lines, which can often be interpreted more directly. A comparison of the results from the millimeter and centimeter lines can give an idea of the uncertainties in the analysis of millimeter emission lines.

Since 13C is believed to be produced by non-equilibrium CNO processing in stellar evolution (Truran 1977), measurements of the carbon ratio Rc ≡ [12C] / [13C] in the interstellar medium may provide important information on nucleo-synthesis. Commonly, the ratio (N13/N18)LTE ≡ [13CO/C18O]LTE is measured and from this RLTE ≡ [12CO/13CO]LTE is deduced and these values are often identified with Rc. However, this line of reasoning can be misleading for two reasons (Dickman et al. 1979):

(1) The difficulty of determining accurate column densities, [13C16O] and [12C18O], because of the complexity of the radiative transfer problem;

(2) The possible role of fractionation, whereby RCO ≡ [12CO] / [13CO] does not necessarily reflect the initial atomic abundance ratio RC (Watson et al. 1976, Langer 1977, Liszt 1978).

In order to determine molecular fractional abundances, both the molecular density and hydrogen density must be known. In this study we have determined these parameters by fitting the observed intensities of J = 2 → 1 and J = 1 → 0 13CO and C18O transitions using a spherical cloud LVG radiative transfer model. The kinetic temperature is determined by observations of 12CO and is found to be between 9 K and 13 K for our sources. The fractional abundance of CO is expected to rise rapidly between regions of low extinction and those with Av ≥ 4 mag, due primarily to the decrease in photodestruction rates (Langer 1976). The C18O fractional abundance data plotted as function of Av support a nonlinear relationship between X(C18O) and Av for Av ≤ 4 mag, with an indication of an asymptotoic value X(C18O) = 2.2 × 10 (-7) in highly obscured regions. For 16O/18O ratios of 250 (suggested by our data although possible uncorrected saturation of 13CO makes this a lower limit) and 700 (A. Penzias, private communication) the fractions of carbon in CO in well-shielded regions are .08 and .23, respectively.

The ratio of ground-state densities Ro ≡ N (13CO) / N (C18O) has been used to infer physical and chemical conditions in giant molecular clouds (Wannier et al. 1976) and dark clouds (Mahoney et al. 1976; Langer et al. 1979). In dark clouds Ro is found to vary from values near the terrestrial ratio [13C] [16O] / [12C] [18O] ∼ 5 at positions of high extinction to values ∼20 at positions of low extinction. In this paper we present high-resolution J = 1 → 0 spectra of CO, 13CO, and C18O at positions of high extinction in TMC-2, L134, and L134N. The C18O lines have non-Gaussian wings and are ∼ half as wide as the 13CO lines. We find that RO must vary across the line, from a minimum of R0 ∼ 4 at the peak of the C18O line to a maximum of RO ∼ 10 in the wings, unless the 13CO line has peak opacity ∼5. The variation of RO with position and with velocity is consistent with models of clouds which have a dense core with low velocity-dispersion and low fractionation, and a rarefied envelope with high velocity-dispersion and high fractionation.