Evidence for the presence of dust in elliptical galaxies was given by the optical observations of obscured regions in some systems (see for instance Bertola et al. 1985, Véron & Véron 1988) and was finally confirmed by the FIR observations of the IRAS satellite (Knapp et al. 1989, Roberts et al. 1991). The low resolution IRAS data ( 1 arcmin) have to be processed by adequate techniques to provide detailed ( arcmin at 100 µm) information about the dust spatial distribution in nearby elliptical galaxies. In general, only the integrated flux of the detected source is available in the IRAS bands.
Despite the intrinsic limits due to the low resolution, Knapp et al. (1989) showed that a significant fraction (48) of the nearby E and S0 galaxies from the Revised Shapley-Ames Catalogue (Sandage & Tammann 1981, hereafter RSA) have been detected by IRAS at 60 and 100 µm at the limiting sensitivity (about 3 times lower than in the IRAS Point Source Catalog).
It is not surprising that ellipticals contain dust, since the presence of dust is directly related to the stellar formation. But the coexistence of solid particles with the dominant gas component, which in these galaxies is heated to K and radiates at X-ray wavelengths, is a matter of discussion. In fact, dust grains should be quickly destroyed by sputtering (Draine & Salpeter 1979) when in direct contact with the hot gas. In such an environment the dust has a lifetime of yr. The question thus follows: where does the dust come from?
At FIR wavelengths (60 and 100 µm) the thermal emission is mainly due to large grains (radius µm) which may have different heating sources, e.g. the general interstellar radiation field or OB stars. In order to discriminate between the two different contributions and to estimate the weight of each of them, several efforts were undertaken (see for instance Calzetti et al. 1995). A reasonable approach is to consider the 60 µm flux entirely due to the warm dust (40 K), while the 100 µm flux should be considered the result of two contributions: the warm and the cold (10 K) dust. While a two-component dust model is used to explain the spectral trend for different types of galaxies, it is rarely adopted for elliptical galaxies, which are often characterized by weak FIR emission and which are not always detected in all IRAS bands. Therefore, a single color temperature (from the 60 and 100 µm data) is usually taken as the dust temperature. It follows that no information about the dust temperature distribution and the dust spatial distribution is available for elliptical galaxies. The availability of the ISO data will provide spectra in a wider IR wavelength range (2.5-240 µm). Several attempts to understand the dust nature and origin suggest interesting interpretations by comparing optical and FIR data (Goudfrooij & de Jong 1995, hereafter GJ95, Tsai & Mathews 1995, 1996), by studying the stellar content, or by using a severe and critical approach to the data (Bregman et al. 1998). Finally, few elliptical galaxies have been observed at sub-millimeter wavelengths by Fich & Hodge (1991, 1993).
GJ95 found that the dust masses determined from the IRAS flux densities are roughly an order of magnitude higher than those determined from optical extinction values of dust lanes and patches, in contrast with what happens for spiral galaxies. The authors suggest that this "mass discrepancy" may be explained by the existence of a diffuse component (within 2 Kpc from the center), which is not detectable at optical wavelengths. On the other hand, Tsai & Mathews (1996) suggested that, while the distributed dust component is associated with dust recently ejected from evolving stars, another "extra" component of dust is present in ellipticals both in dust lanes and rings and/or in other galactic regions. In particular, they postulated that a substantial mass of cold gas remains "observationally elusive without forming completely into stars". If the extra dust is optically thin in the visible it should be located far from the galactic core region, where the intensity of the starlight and therefore the grain heating is reduced.
The dust spatial distribution suggested by GJ95 and by Tsai & Mathews (1996) support the two most popular scenarios respectively: the evaporation flow picture and the cooling flow picture. In the evaporation flow scenario the clouds of dust and gas currently observed in ellipticals have mainly an external origin, being associated to events of galaxy interaction and/or mergers and being heated by thermal conduction in the hot gas (de Jong et al. 1990, Sparks et al. 1989). On the contrary (cooling flow ), the internal origin of gas and dust may be explained with both red giant winds (Knapp et al. 1992) and by the cooling flow mechanism, in which mass loss from stars within the galaxy is heated by supernova explosions and by collisions between expanding stellar envelopes during the galaxy formation stage, and then cools and condenses (Fabian et al.1991).
Therefore, in ellipticals, the presence of dust and the dust "mass discrepancy" are related with the dust spatial distribution, which depends on the nature and the evolution of these systems. Hence, in order to investigate the dust content, dust mass evaluations as accurate as possible are required to estimate the amount of the "mass discrepancy". Since both scenarios suggest the presence of different dust components and the very nature of the galactic environments involves the existence of dust grains at different temperatures, it is necessary to take into account a dust temperature distribution for the dust mass evaluations.
Kwan & Xie (1992) suggest a theoretical approach to take into account the effects of the dust temperature distribution in the dust mass evaluation. I present here an application of their method which is discussed and implemented in Sect. 2. The results obtained for the sample of ellipticals, introduced in Sect. 3, are presented and discussed in Sect. 4 by comparing the FIR with the visual dust mass evaluation and by discussing the correlation between blue and FIR luminosities.
© European Southern Observatory (ESO) 1998
Online publication: September 17, 1998