The results of interstellar extinction measurements from the near IR to the far-UV are reviewed. The average interstellar extinction curve for the diffuse cloud medium exhibits a nearly linear rise (in 1/λ) from 1 μm−1 to the 2.25 μm−1 “knee” in extinction where the slope changes. In the UV there is a pronounced extinction bump near 4.6 μm−1 (2175 å) followed by a broad minimum and a steep rise in extinction to the shortest wavelengths for which measurements exist. For wavelengths shortward of about 5500 å, the interstellar extinction curve exhibits considerable variation in shape from one sight line to another. In addition to strength variations, the width (FWHM) of the 2175 å extinction bump has been observed to vary by more than a factor of two from 360 to 770 å with the average width being 480 å. In contrast, the central position of the feature only varies from 2110 å to 2195 å with the average central position being 2175 å. The most extreme variations in extinction are found at far-UV wavelengths where E(λ – V)/E(B – V) has been found to range from 3 to 12.5 at 1250 A. In the last few years significant progress has been made in determining various empirical relationships among extinction parameters in the different wavelength regimes and in determining how extinction curve shape changes are influenced by the interstellar environment in which the dust resides. Those relationships are discussed.

The abundances of free atoms and ions relative to hydrogen in the interstellar medium indicate how thoroughly different elements have condensed into solid form. Large contrasts in these depletions disclose differences in how tightly various elements are bound to the grains. While most grain nuclei probably form within astrophysical sites where densities are large, a substantial amount of the accretion of heavy elements must occur within interstellar clouds. Interstellar shocks created by supernova explosions or perhaps collisions of clouds destroy or significantly erode the grains. This viewpoint on the formation and destruction of grains is supported by the decrease in the severity of depletions in regions of lower than normal density or parcels of gas moving at high velocity. While depletions in gas away from the plane of the galaxy generally imitate the behavior of low density regions in the plane, mild anomalies for some elements may exist.

In this paper I review the existing measurements of the wavelength dependence of interstellar extinction arising from dust situated outside the Milky Way galaxy. This paper emphasizes studies of extinction in the Large and Small Magellanic Clouds. I discuss the UV extinction properties, as well as the optical and near-IR properties as determined from ground-based photometry.

The observed infrared properties of dust extinction along the line of site to the Galactic Centre are compared with those in the local interstellar medium. Further constraints on the nature of the dust grains are provided by polarimetric observations, which also lead to information on the stratification of the material along the line of sight. Finally some implications of the large polarized emission seen from the central regions are discussed.

A brief introduction is provided to the origin of interstellar polarization, grain model calculations, and alignment. The characteristics of the observed linear and circular continuum polarization are described, with an emphasis on various constraints that might be placed on the nature of the polarizing grains and what additional observations would be of value. Complementary information can be obtained from polarization structure at spectral features in the extinction curve. Interstellar polarization has been measured in some other galaxies, indicating an interesting frontier for further research.

The unidentified (since 1921) diffuse interstellar bands (DIBs) are discussed together with their relations to other interstellar absorptions sucn as: continuous extinction, polarization and atomic or molecular absorption lines. It is shown that DIBs do not form the absorption spectrum of one agent, but probably of several (3 or more). DIBs as well as other interstellar absorptions are usually formed in several clouds along a line-of-sight. Thus, they suffer Doppler splitting; the first high resolution profiles free of the latter effect are described. Since single interstellar clouds may differ not only in radial velocities but also in many physical (optical) parameters, the observed interstellar absorptions are ill-defined averages over all clouds situated along any line-of-sight. It is of basic importance to determine not only the single cloud profiles of diffuse bands, but also their relations to other interstellar absorptions in the same clouds. Intensity ratios of DIBs are shown to be sensitive to the shapes of extinction curves, depletion patterns of elements and molecular abundances in the considered clouds. The sensitivity of the DIBs to the variation in polarization is less documented but probably also present. Thus the diffuse lines are presented as the unidentified part of the absorption spectrum of interstellar matter. Their identification depends on the determination of their relations to other interstellar absorptions which must be determined precisely.

Studies of interstellar dust based on observations of scattered and dust-emitted radiation from grains are reviewed.

Observations of 1–25 μm. continuum emission and the interstellar infrared emission features in reflection nebulae are reviewed. These observations place important constraints on models of very small grains or large molecules such as PAHs, which these models must address in order to understand this fundamental component of interstellar dust.

Advances in infrared spectrometers and the theory that PAHs are responsible for the infrared emission bands have led to a wealth of new information in recent years. High quality data have shown many weak emission bands which are diagnostic of the material producing the bands. While correlations of the strengths of the narrow bands indicate that a single material can account for all of the narrow bands, independent spatial behavior of the narrow and broad components show that they have different origins. Predictions of the behavior of the spectra based on laboratory data have been confirmed observationally, strengthening the theory that PAH molecules are the origin of the infrared emission bands.

We discuss the fraction of the infrared cirrus emission radiated out of thermal equilibrium using the IRAS data. Recent spectrophotometric data (Giard et al., 1988a, 1988b) are presented which confirm that aromatic infrared bands account for at least a fraction of their near and mid-infrared cirrus emission as was suspected from indirect arguments. Large variations of the energy distribution among the IRAS bands are shown to be present in molecular clouds. Abundances of polycyclic aromatic hydrocarbons (PAH's) required to account for the emission are discussed.

The infrared evidence which supports the PAH hypothesis is briefly summarized. Rather than presenting a general discussion of these assignments, this paper focuses on the spectroscopic issues raised by recent observational and experimental developments. These issues include: the position and profile of the “1310” cm−1 (“7.7” μm) feature, the position and intensities of the bands in the 910-710 cm−1 (11-14 μm) region, the newly detected 1900 cm−1 (5.3 μm) band, and the spatial and spectral variations in the 3000 cm−1 (3 μm) region as well as in the 12 and 25 μm IRAS bands. It is concluded that the infrared evidence for interstellar PAHs and PAH-related species is compelling.

Various sources of non-equilibrium radiation from interstellar dust are discussed. It is shown that the existence of cirrus emission at 12 and 25 μm is consistent with the presence of amorphous carbon dust and arises from thermal spikes within ≃ 10å subvolumes of normal (0.01-0.1 μm radius) dust grains. The 3.28 μm unidentified infrared (UIR) feature also arises in this way, as the radiative relaxation of high energy vibrational modes accompanying a thermal spike in hydrogenated amorphous carbon. Extended red emission (ERE) and near-infrared (NIR) emission are also discussed and are postulated to originate as edge and defect luminescence from HAC solids with bandgaps Eg ≲ 2.5eV.

We discuss the optical properties of carbonaceous materials from the VUV to the FIR. Updated laboratory data on amorphous carbon (HAC) and some PAH mixtures are analysed and compared with relevant literature. We find that the 2175 å hump can be due to very small (< a > ≥ 10 å) HAC grains. The UIR bands seem better explained by mixtures of HAC particles and collections of PAHs.

A review is presented of some physical and chemical properties of polycyclic aromatic hydrocarbons (PAHs) that are relevant for interpreting various aspects of observations and speculations on PAHs in the interstellar medium. The subjects discussed are: the stability and reactivity of neutral and ionic PAHs; the spectroscopy and photophysics of neutral and monocationic PAHs; the photostabilities of PAH monocations and dications and their astrophysical implications.

The Polycyclic Aromatic Hydrocarbon (PAH) hypothesis states that a mixture of free PAH molecules is an ubiquitous and abundant component of the interstellar matter. It has been first formulated by Léger and Puget (1984) on the basis of molecular stability that made PAHs good candidates for the very small grains proposed by Sellgren (1984). Then, they obtained strong, although not final, support from spectroscopy.

The proposal that polycyclic aromatic hydrocarbons (PAHs) are the source of the unidentified infrared bands has several serious deficiencies that have not been discussed or satisfactorily treated: (1) no collection of neutral PAHs has been found which matches the observed wavelengths; (2) ion or dehydrogenated molecules have been predicted to be the dominant species in some regions, but no infrared spectra of either species have been obtained; (3) the restriction to small grains is based on grain temperatures ∼ 1000-1500 K which in turn followed from the infrared continuum color temperature – there is now a question whether that is thermal emission; if it is, the source cannot be molecules; (4) recent observations of the 12/100 μm flux ratio as a function of stellar temperature do not conform to predictions of the PAH hypotheses; (5) the photon excitation mechanism for infrared emission by neutral molecules should produce strong visible-ultraviolet fluorescence which is not observed – ions may not do this, but their infrared spectra are not known. There does not seem to be a ready explanation for these problems with the PAH hypotheses. Until definitive experimental study and analysis yielding unambiguous results have been carried out for PAHs, hydrogenated amorphous carbon or other forms of carbonaceous material, it is premature to assume any type of grain is the source of the infrared bands.

Quenched carbonaceous composite (QCC) was synthesized by quenching the plasma of methane gas. Chemical properties as well as optical and infrared spectra of the QCC and “oxidized” QCC were measured. Good agreement of the IR spectra of the QCCs to the unidentified infrared (UIR) emission bands was obtained. Correspondence of their features to molecular structures in the QCC was estimated. It is concluded that a “cross-conjugated ketone” structure (CCK) caused the 6.2, 7.7, and 8.6 μm features and “solo” H atoms on carbon are responsible for the 3.3 and 11.3 μm features.

The effects of size and energy on infrared fluorescence (IRF) and on chemical reaction rates are investigated, using polycyclic aromatic hydrocarbons (PAHs) as examples. The range of validity of the Thermal Approximation (TA) is examined. It is found that for properties that have a near-linear dependence on the internal energy, the TA provides an adequate description of the non-thermal, time-dependent processes associated with ultraviolet photon absorption. Since IRF at high energy is nearly linear, the TA is adequate for IRF at high excitation energies, but care must be taken, because the TA fails at low energies. The TA is never adequate for chemical reactions under these conditions.