The interstellar medium (ISM) encompasses only a modest fraction of the mass of a galaxy, typically a few percents of the stellar mass, less than 1% of the dark matter mass, of a large spiral galaxy such as the Milky Way. However, it is an essential component of spiral and irregular galaxies (see various books and reviews such as Spitzer 1968, 1998, Balian, Encrenaz & Lequeux 1975, Dyson & Williams 1980, Kaler 1997, Lequeux 2005, Tielens 2005). Most importantly, it is the material and the framework of star formation. Gravitational forces are constantly at work to squeeze and fragment its condensations; slowly starting at the scale of giant molecular clouds and ending in the final collapse of individual stars. Star formation is known as a highly complex process, yielding a number of intermediate stages and byproducts which correspond to various components of the ISM such as cold and hot condensations, accretion disks, bipolar flows and jets, energetic stellar winds and shocks, etc. These mechanisms introduce strong feedback actions upon the surrounding ISM, which in turn determine the global properties of star formation and account for its low efficiency.
The symbiosis between stars and the ISM, together with stellar evolution, governs the overall and chemical evolution of galaxies. The properties, chemical composition and global mass of the ISM depend on its interaction with stars not only through the engulfment of material into new stars and the violent action of massive young stars through winds and supernova blasts, but also through the milder winds of the less massive stars, especially during the final stages of their evolution when they disperse their matter in the AGB and planetary nebula phases.
The extremely low density of the interstellar gas, ~ 1-103
cm-3, outside of the final star-forming condensations, leads
to peculiar physical and chemical properties, as concerns time
constants, temperature, heating and cooling, chemical reactions, dust
and nanoparticles, and the associated radiative transfer which plays a
major role in the energy balance and radiative
multi-
emission
of galaxies (see e.g.
Tielens 2005
and references therein). The peculiarity and richness of the ISM is also
a consequence of the variety of the violent processes which permeate it,
especially in its lowest density phases: UV radiation, stellar winds,
supernova explosions, X-rays and
-rays,
cosmic rays, inducing
ionization, turbulence, magnetic fields, shocks, etc. Such violent
processes are fundamentally hostile to molecules. However, the latter
find efficient shelter within the densest parts of the ISM which are
much less affected.
Indeed, because of its dispersed nature, the ISM is a highly self-interacting and dissipating medium. It is thus at the origin of the formation of the disks of the spiral galaxies, and is subjected to strong interactions in galaxy collisions and merging, leading to major starbursts and feeding the central black hole. The ISM is also constantly fed by accretion of extragalactic gas and it plays a major role in the formation of galactic bars. It is strongly affected and compressed by spiral structures, leading to the spectacular visualisation of spiral arms through massive star formation. In major starbursts, it may convey huge outflows into the galactic halo, and even in the intergalactic medium, being probably at the origin of the enrichment of the extragalactic gas in heavy elements.
The basic features of the energetics - heating and cooling - and the dynamics of the interstellar medium have been understood for 30-50 years (see the general references above and references therein, and McKee & Ostriker 1977). The energy involved in interstellar gas processes is generally only a small fraction of the total energy generated in stars (and possibly the central AGN), while interstellar dust may channel into the far-infrared a substantial fraction, and even the quasi-totality, of the energy generated by star formation (and in some cases the AGN). Heating the gas is thus achieved through relatively marginal processes from UV stellar radiation - mainly through photo-electric ejection of electrons from dust grains (Watson 1972), and from atoms in the vicinity of UV stars -, from cosmic rays, stellar winds, interstellar shocks mainly from supernovae explosions, gravitational energy of molecular clouds, etc. Outside of adiabatic expansion, cooling is mainly achieved through spectral lines of atoms, ions and molecules; the main lines depending mostly on the gas temperature. The interplay between heating, cooling and the gas dynamics may generate special thermodynamical properties, such as major instabilities in some cases (Field 1965), the special physics of frequent shocks (McKee & Ostriker 1977, Shull & McKee 1979, McKee & Hollenbach 1980), peculiar turbulence (Williams et al. 2000, Falgarone et al. 2005a, Elmegreen & Scalo 2004), etc.
Mostly through dust grains (and atomic absorption in the far-UV and X-ray ranges), the interstellar medium completely reshapes the spectral energy distribution (SED) of galaxies. In most of them, a good part of the UV and visible radiation from stars is depleted and the corresponding energy comes out in the infrared. The main part of this energy is emitted in the far-infrared by relatively large dust grains, while nanoparticles (mostly Polycyclic Aromatic Hydrocarbons, PAHs, Léger & Puget 1984) emit spectacular aromatic bands in the mid-infrared.
The tenuous interstellar gas, theatre of very peculiar physical and chemical processes, is indeed an essential ingredient of galaxies. It forms galactic disks and stars, reprocesses a large fraction of their radiation into the infrared, constantly exchanging material with stars and often strongly interacting with them through violent processes.
1.2. Molecular gas is a major component of the ISM
As described e.g. in Tielens (2005), Lequeux (2005) and Dyson & Williams (1980), molecules constitute a significant fraction of the ISM, residing in the densest regions where star formation takes place. In the ISM, the overall abundance of molecules, i.e. in practice that of H2, is determined by the balance between the destruction processes, mostly by UV, and the formation, generally on dust grains for H2 (see Section 2.4). High gas densities favour molecules, both by accelerating their formation, and protecting them against the main destruction process, UV photodissociation, because of the exponential UV shielding by dust and self-shielding of H2. Therefore, at densities of about 103 cm-3, the interstellar gas is almost entirely molecular, except for about 25% in mass of He. Of course, the molecules are almost 100% H2 because of the overwhelming abundance of H nuclei (Table 1), and less than 1% being in other molecules, mostly CO and H2O (see Section 6). The ISM also contains ~ 1% in mass of dust grains intimately mixed with the gas, including most of the refractory elements, such as Fe and Si, and a significant fraction of O and C (see Section 8). The interstellar molecular gas is generally cold (10-50 K) because of inefficient heating, mostly by cosmic rays in the absence of UV radiation, and efficient cooling through molecular lines, mostly CO. However, there are also special cases with warm molecular gas, from ~ 100 K up to a few 103 K in strong starbursts, shocks, or in the vicinity of hot stars, AGB stars and AGN.
The molecular ISM is essential in star formation, from the initial condensation and fragmentation of giant molecular clouds, to the dense accretion disks with molecular flows and pre-planetary disks. The total mass of molecular gas is a determining factor for the global star formation in a galaxy. It is impossible to discuss the complex processes of star formation without dealing with the molecular physics of the interstellar gas. Molecules are particularly essential for cooling the molecular gas to the low temperatures which determine the properties of gravitational collapse and will eventually lead to star formation.
Molecules may also be minor components of the atomic interstellar gas. They have then little influence on the physical properties of the ISM and its evolution. However, they can be interesting probes of the ISM, e.g. for tracing the intensity of cosmic rays or UV radiation, and importantly contribute to the infrared emission through mid-IR PAH bands. There are even cases where tiny amount of molecules may have a dramatic influence. In particular, in the primordial gas, before the formation of the first stars, abundances of H2 as small as a few 10-4 may allow cooling proto-galactic condensations down to a few 102 K, allowing early collapse of relatively small masses, perhaps of globular cluster size, which could in turn give birth to the first stars (Section 10.4.2).
Molecular gas - mostly H2 and He, with ~ 1% of other molecules
and dust - is definitely the normal state of the dense ISM
(
103
cm-3). It is essential in the physics of the gravitational
collapse at the various scales which eventually lead to the formation
of stars and planetary systems.
The discovery of the first interstellar molecules occurred as early as 1936-1942, in the context of studies of atomic and ionic absorption lines in the sight-line of bright stars, through the detection of absorption bands of a few diatomic species with abundant atoms and strong optical bands: CH+, CH and CN. It is interesting to note that this list of optical detections has been little increased since this initial discovery. There is a relationship between the rotational excitation of the lower level of these transitions and the CMB 3K radiation temperature; however, this relationship was not properly understood until the CMB was discovered 30 years later.
With the development of interstellar dust studies in the 1940-1950's, various discussions took place about possible catalytic synthesis of molecules on the surface of dust grains, and their likely presence in abundance in the interstellar gas. However, because of the lack of adequate optical lines for most molecules, the exploration of the molecular ISM had to wait for the development of microwave, infrared and especially millimetre technics, and space astronomy for UV detections. After the revolution brought in interstellar studies by the radio astronomy with the first observation of the HI 21 cm line, the first organized searches for interstellar molecules were triggered by the success of laboratory microwave spectroscopy. They produced in the 1960's the discovery of the very important interstellar molecules OH, NH3, H2O and H2CO, with strong interstellar maser emission for OH and H2O. Then, because most simple molecules have rotation transitions in the millimetre range, the majority of the detections came with the advent of millimetre radio-astronomy. Indeed, the development of interstellar molecules studies has been always greatly dependent on technical advances: first with radio-astronomy technics in the stream of the enormous success of radio continuum and 21cm observations; and then mostly with the full development of millimetre telescopes and detectors, which was indeed a dedicated effort aiming at interstellar molecules studies, strongly motivated by early microwave detection of complex molecules, such as NH3 and H2CO.
The seventies were the golden age of the discoveries of interstellar molecules. After the early detection of CO, it was soon realized that its stability, abundance and widespread distribution in local interstellar clouds, in the entire MW Galactic disk, circumstellar AGB shells, external galaxies, etc., make it a very good tracer of H2 and thus of the whole molecular ISM. H2 itself is very difficult to detect in the normal cold molecular ISM, because of its lack of allowed electric dipole rotational and vibrational transitions, and the large opacity of molecular clouds to UV radiation (Section 10).
In addition to CO, early millimetre searches quickly discovered tens of other interstellar molecules, some of them which were expected, such as HCN, CH3OH, CS, SiO, C2H, etc., but others turned out to be unexpected such as HCO+, N2H+, HCnN, CnH, often unknown in the laboratory (see e.g. Watson 1976, and Section 2). This soon led to a very good understanding of the basic features of interstellar chemistry, in particular the need of dust grain synthesis of H2 and the essential role of ionic reactions driven by cosmic rays (see Section 2.4).
In parallel to millimetre observations, the first space UV telescopes allowed the exploration of the molecular content of translucent clouds through absorption of UV lines (Section 7). Such clouds are mostly atomic, but contain enough molecules, mostly H2, to be detectable.
The advent in the 1980-1990's of large single-dish telescopes and interferometers operating at millimetre wavelengths opened up a new area in our understanding of the interstellar medium through further discoveries of new interstellar molecules, detailed studies of star forming regions and their physics and chemistry in the Milky Way, and importantly initiated the first comprehensive molecular studies in external galaxies in the local Universe. However, it was realized only in the nineties that molecules could already be detected with the current equipment in strong starbursts in the most distant Universe. In parallel, ground-based and space infrared astronomy developed comprehensive studies of H2 lines and aromatic mid-IR bands.
1.4. Main achievements of molecular studies in external galaxies
As it is reviewed below, many of the main achievements of molecular studies in the last decades took place in observations of external galaxies. Such studies have become a major field for understanding galaxies, their evolution and especially star formation and starbursts at galactic scales.
First, numerous, comprehensive millimetre studies of CO and its isotope varieties have brought detailed information about the amount and distribution of the molecular gas in various kinds of local galaxies, and its relation with star formation and physical conditions. One has been able to infer the role of molecular gas in galaxy evolution, spiral arms, galactic bars and AGN feeding, galaxy collisions, etc., and associated physical processes. The detection of a number of other molecules in local galaxies have allowed comparative studies of molecular abundances in various galactic environments, including isotopic varieties, and inferences on physical and chemical modelling.
Mega-masers in the lines of OH and H2O, with enormous power, have been detected in local infrared luminous and ultra-luminous galaxies (LIRGs and ULIRGs), with discussion of their physics. Detection of H2O mega-masers in disks around AGN lead to a precise determination of the mass of their central super-massive black hole. Several other molecules have also been studied in the vicinity of the central torus around AGN, with continuously improving angular resolution at millimetre and infrared wavelengths.
The absorption of many molecular lines of many species at redshift ~ 0.5-1, in front of a few radio sources, have provided comparative molecular abundances and inferences on the evolution of physical conditions and atomic abundances in galaxies. Rotational emission of molecules is currently detected in ULIRGs at very high redshift up to z = 6.4, often around powerful quasars and radio galaxies. Unique information is deduced about the early formation and starbursts of the most massive galaxies.
Large aromatic compounds (PAHs) have shown to be ubiquitous in all kinds of galaxies. Infrared H2 emission lines begin to currently trace the warm molecular gas in the mid-IR from space, and at high angular resolution in the near-IR from ground-based adaptive optics.
Much more is expected with the prospects of order(s) of magnitude gains of sensitivity expected with worldwide new facilities: mostly with ALMA (Atacama Large Millimeter Array) in the millimetre range; and also from future large space infrared telescopes and SKA (Square Kilometer Array) in the radio. One will thus study fine details in local galaxies and make comprehensive global studies of standard galaxies at high redshift and detailed studies of IR starbursts galaxies at very high redshift. One may thus expect: i) deep progress in understanding the evolution, formation and merging of various kinds of galaxies at all early epochs when the galaxies and most of their present stars formed, and the connections between AGN and their host galaxies; and also ii) major advances advances in comparative interstellar chemistry in external galaxies.