Galaxies and the first stars in the universe formed billions of years ago as a result of the cooperative effects of gravitational collapse and nonequilibrium chemistry. Gravity drew the primordial gas together into lumps; the formation of the first molecules in the universe, simple diatomic molecules like H2, H2+, HD, HeH+, LiH, and LiH+, may then have ensured that the heat generated by gravitational collapse and shock waves was radiated away rapidly enough to allow the gravitational collapse and fragmentation of these gaseous lumps to proceed to the point of forming stars and galaxies. We briefly mention a few of the latest studies of this primordial chemistry, including that in the evolving intergalactic medium (IGM) in a Cold Dark Matter (CDM) model cosmology and that in radiative shocks in the early universe.

We study the formation of Population III objects (and stars) taking into account the expansion of the Universe and a series of physical processes relevant to the primordial plasma. In particular, a very important way to cool the first condensations in the Universe is with molecular hydrogen. The minimum mass of the Population III objects formed is ~ 104M⊙. We show also that stars of mass ≳ 50M⊙ can be formed from the fragmentation of Population III objects of mass ~ 106M⊙.

We review the observational status in evaluating the interstellar deuterium abundance and show that the most reasonable value is of the order of 10−5. Although in general agreement with the standard Big Bang nucleosynthesis, the situation is still unclear and deserves much more observations.

This paper summarizes our understanding of the dominant D-reservoirs in molecular clouds and suggests possible direct determinations of the D-abundance. It is concluded that rotational HD lines from shock regions provide the best way to determine the gas phase D-abundance. In cold dense cores, the dominant gas phase D-reservoir is likely to be atomic D, because of expected inefficient HD formation on grain surfaces. The gas phase D-abundance derived from observations of dense cores is ≍5 × 10-6/n4 (with n4 the total density in units of 104 cm−3). A large fraction of the D (≍50%) may be locked up in deuterated molecules in grain mantles. A small fraction (≍2%) may be locked up in the photolyzed residues of such grain mantles. PAHs will also be deuterated (PADs), containing ≍1% of the D. Finally, it is likely that all of these processes have contributed to the D-enrichment observed in solar system materials.

Values of D/H measured in the methane on the giant planets and Titan indicate the presence of two distinct reservoirs of deuterium in the outer solar system. The dominant reservoir is in hydrogen gas, the second, multi-component reservoir is found in the hydrogen that is bound in condensed compounds. Both reservoirs appear to have originated in the interstellar medium. In contrast, the values of D/H in water vapor on Mars and Venus (especially) exhibit a large enrichment from the “condensed matter” starting value. Interpretation of this enrichment may illuminate the history of water on these two planets.

2 μm spectroscopic observations by many authors have revealed significant rotation-vibrational H2 emission is widespread from starburst to bare nucleus galaxies. Near-IR H2 emission lines can arise from various excitation sources: UV radiation by hot stars, shock excitation by supernova remnants or AGN driven winds, and UV/X-ray radiation by an AGN. In this review recent data will be compared with such H2 excitation models.

Data of our on-going program on the interstellar polarization in the Small Magellanic Cloud (SMC) seem to show a correlation between the UV extinction and the parameters of the Serkowski law which describe the polarization. Fits of popular interstellar dust models to the polarization data are not always satisfactory, being better for stars with small wavelength of maximum polarization, λmax.

Chemical properties of extragalactic molecular clouds are summarized. Column densities, relative abundances, and chemical implications are briefly discussed and references for more detailed studies are given. The need for model calculations is emphasized.

The first observations of the [CII] line toward the nuclei of gas-rich external galaxies, showed that the far-infrared line emission contributes up to 1% of the total luminosity and most likely originates from dense photon-dominated regions (PDRs) associated with the surfaces of molecular clouds exposed to FUV from external or embedded OB stars (Crawford et al. 1985, Lugten et al. 1986, Stacey et al. 1991). We have mapped the [CII] emission toward NGC 6946 over an 8' × 6' (23 × 17 kpc) (Madden et al. 1991) using the Max-Planck Instutute/U.C.Berkeley Far-Infrared Imaging Fabry-Perot Interferometer (FIFI) on the Kuiper Airborne Observatory (KAO).

We discuss the parameters needed to model chemistry in extragalactic clouds. While density and temperature can be constrained by multiline observations, molecular abundances may be severely affected by the adopted elemental abundances. While the observations of the Magelllanic Clouds can be reasonably interpreted in terms of dark cloud models, molecular gas in starburst galaxies could well be dominated by photoeffects. The detection of deuterium in extragalactic molecules would provide a valuable diagnostic.

Observations of interstellar CN, CH, and CH+ in diffuse and translucent clouds obtained with a spectral resolution of more the 500,000 are reported. First results from a survey of CH and CH+ are discussed. The CH+ isotopic ratio has been determined to be 68±5 toward ζ Oph and a component of 12CH+ has been found with a velocity width of 15 km s−1. A review of the rotational excitation temperature of interstellar CN and its implications is given.

The spectra of objects suffering extinction by diffuse interstellar dust contain a broad feature centered at ~3300 cm−1 (~3.0 μm), attributed to O-H stretching vibrations, and/or a feature near 2950 cm−1 (3.4 μm) attributed to C-H stretching vibrations. The 2950 cm−1 feature can be attributed to C-H stretching vibrations in the -CH2- and -CH3 groups of a fairly complex carbonaceous material containing aliphatic functional groups.

Polycyclic Aromatic Hydrocarbon (PAH) molecules have been proposed as candidates to explain the Diffuse Interstellar Bands (DIBs). We have performed laboratory measurements of coronene, using rare gas matrix isolation techniques and UV photolysis. Our aim was to search for a possible identification of the 4430 Å DIB, but also to provide data almost free from environmental band shifts and broadening, which can be used for astronomical identification of the species.

We modelled the IR emission of interstellar PAHs. Substantial differences between the IR properties of interstellar and laboratory PAHs are found, possibly resulting from ionization. The various IR features being dominated by distinctly different size PAHs, their observed relative intensities are sensitive indicators of the size distribution. A number of applications of our model related to future ISO and SIRTF IR data are pointed out.

Theoretically determined vertical excitation energies of C5 and C7 molecules have been shown to agree with the strongest diffuse interstellar band at 4430Å. Several other weak diffuse bands can be identified with vibrational transitions, if 4430Å band is taken as 0-3 transition and vibrational constant have a value of 2190cm−1. The vertical excitation energy of a second electronic transition of C7 molecule agrees with diffuse band at 6177Å. This electronic band system may account for diffuse bands at 5778Å, 6660Å and other bands near them.

The relative importance of gas-phase, shock and grain-surface chemistry is discussed with reference to recent observational data. The relation between diffuse, translucent and high-latitude clouds is illustrated with models in which carbon is transformed from atomic (C+, C) to molecular (CO) form with increasing extinction.

The large abundance of CH+ in diffuse clouds has been a mystery for more than 50 years (Dalgarno 1976). Many different explanations have been proposed, but only (shock) models with a substantial column of warm gas appear capable of approaching the observed column densities (Elitzur and Watson 1978). In these models, CH+ is formed in the warm postshock gas through the reaction C+ + H2 → CH+ + H, which is endoergic by 4650 K. Although the most sophisticated MHD shock models are consistent with various observational aspects of CH+, they require substantial “fine-tuning” of the parameters (Draine and Katz 1986; Hartquist et al. 1990). In addition, they predict a shift in velocity between CH+ and other molecules found in the cold quiescent gas, which is not observed in recent data (e.g. Crawford 1989). In order to test further the shock models, we have searched for CH+ in a number of translucent clouds (AV ≍ 1-5 mag), which have H2 column densities that are up to an order of magnitude larger than the clouds studied so far (cf. Souza 1979). Observations of the chemically related molecules CH, C2, CN and CO have been obtained as well.

We consider the propagation of low-amplitude MHD waves in partially-ionised plasmas. Ion-neutral drift (ambipolar diffusion) can lead to significant variations in the deuterium fractionation ratios of several molecules (e.g. HCO+ and N2H+) on spatial scales of between a few hundredths and a few tenths of a parsec, depending upon the fractional ionisation of the plasma. It is possible that interstellar Alfvén waves could be detected by molecular spectroscopy and that these waves may produce other small-scale abundance gradients in molecular clouds.

The observed H2 level population distribution of the rotational levels J = 5, 6 and 7 in the ground electronic and vibrational state X(v = 0) is analysed using available Copernicus data (Spitzer and Morton, 1976) for the diffuse clouds toward the stars ζ Oph, δ Per and ξ Per. The abundances of H2 in these high purely rotationally excited energy levels are due to the influence of one or more of three possible excitation mechanisms: collisions with abundant particles (e.g. H atoms and H2 molecules), UV-photoexcitation and H2 formation.

Several tracers of the cold interstellar medium in its quiescent phase reveal common unexpected properties: the fractal nature of the emission contour levels, the self-similar brightness distribution, the existence of small unresolved fragments of dense gas in various environments, and a highly turbulent velocity field. In spite of the difficulties met in interpreting these data, they are important bearings in our understanding of the structure, the physics and the chemistry of this medium.