Annu. Rev. Astron. Astrophys. 2000. 38: 761-814 Copyright © 2000 by Annual Reviews. All rights reserved

2.2. Dark Matter in the Halos and Outer Parts of Disks

Evidence has been mounting for several decades that the outer parts of disk galaxies are dominated by dark matter in the form of a spheroidal halo distribution (e.g. Trimble 1987, Carr 1994). Of the various possibilities for the nature of this halo, dark baryonic matter in the form of stellar remnants, low-mass stars, and substellar objects seems the most conservative solution, and is still consistent with current Big Bang nucleosynthesis limits (Carr 1994). It is of substantial interest to complement the existing deep optical and near-infrared observations with a search for possible IR halos in edge-on galaxies. Such halos could signal the presence of cool, low-mass stars or of substellar ( 0.08 M) brown dwarfs.

With ISOCAM, Gilmore & Unavane (1998), Beichman et al (1999) have searched for IR halos in five edge-on galaxies. No halos are detected in any of them. Gilmore & Unavane (1998) concluded that main-sequence, hydrogen-burning stars of all masses and metallicities down to the hydrogen-burning limit are excluded to dominate the halos. A mixture of low-mass stars with a mass function similar to that in the solar neighborhood can also be strongly ruled out in one galaxy (NGC 2915). Young ( 1 Gyr) brown dwarfs are weakly ruled out in UGC 1459. In contrast the EROS/MACHO microlensing data toward the Large Magellanic Cloud favor a mean mass of the deflectors in the halo of our own Galaxy of about 0.5 M (Alcock et al 1998). A possible way out may be that the lensing objects are not in the Galactic halo but in the LMC (Sahu 1994), or in a dwarf galaxy debris between the Galaxy and the LMC (Zhao 1998). The combined ISO, HST, and MACHO/EROS results, with the additional assumption that the composition of halos is similar in the different galaxies, suggest that, with the exception of brown dwarfs between 0.01 and 0.08 M (still permitted by EROS to make up to 35% of the halo mass, (Afonso et al 1999), the entire range of possible stellar and substellar entities is excluded or is highly unlikely (Gilmore 1998).

Another possible form of baryonic dark matter in the outer disks of spirals may be cold molecular gas (e.g. Pfenniger et al 1994). There is not much direct evidence for such a component from CO mm spectroscopy, but in the outer low-metallicity regions CO may have low abundance. Evidence for cold dust more extensive than the stellar distributions comes from 200 µm ISOPHOT and 450/850 S(0) and S(1) H₂ emission throughout the disk of the edge-on galaxy NGC 891. Remarkably, H₂ is detected out to 11 kpc galactocentric radius, implying the presence of warm (150-230 K) H₂ clouds that may represent a significant fraction of the total molecular content of the galaxy. Based on the line shapes of the two H₂ lines at the outermost positions, Valentijn & van der Werf argue that there must be a second cooler H₂ component (~ 80 K) that would then have to contain 4 to 10 times the mass estimated from CO and HI observations. If confirmed, such a molecular component (although much warmer than originally proposed by Pfenniger et al 1994) could make a significant contribution to the dynamical mass at that radius.