Dust in high redshift galaxies

Measurements of Zn and Cr abundances in 18 damped Ly? systems (DLAs) at absorption redshifts zabs = 0.692-3.390 (but mostly between zabs 2 and 3) show that metals and dust are much less abundant in high-redshift galaxies than in the Milky Way today. Typically, [Zn/H] -1.2; as Zn tracks Fe closely in Galactic stars of all metallicities and is only lightly depleted onto interstellar grains, we conclude that the overall degree of metal enrichment of damped Ly? galaxies ? 13.5 Gyr ago (H0 = 50 km s-1 Mpc-1, q0 = 0.05) was ~1/15 solar. Values of [Cr/Zn] span the range from 0 to - 0.65 which we interpret as evidence for selective depletion of Cr onto dust in some DLAs. On average Cr and other refractory elements are depleted by only a factor of ? 2, significantly less than in local interstellar clouds. We propose that this reflects an overall lower abundance of dust?which may be related to the lower metallicities, likely higher temperature of the ISM and higher supernova rates in these young galaxies?rather than an exotic composition of dust grains. Combining a metallicity ZDLA 1/15 Z? with a dust-to-metals ratio ? ? of that in local interstellar clouds, we deduce that the typical dust-to-gas ratio in damped Ly? galaxies is ? 1/30 of the Milky Way value. This amount of dust will introduce an extinction at 1500 ? of only A1500 ? 0.1 in the spectra of background QSOs. Similarly, we expect little reddening of the broad spectral energy distribution of the high-z field galaxies now being found routinely by deep imaging surveys. Even such trace amounts of dust, however, can explain the weakness of Ly? emission from star-forming regions. We stress the approximate nature of such general statements; in reality, the range of metallicities and dust depletions encountered indicates that some sight lines through high-redshift galaxies may be essentially dust-free, while others could suffer detectable extinction. Finally, we show that, despite claims to the contrary, these conclusions are not inconsistent with recent high-resolution observations of DLAs with the Keck telescope. We point out that the star formation histories of high-z galaxies are not necessarily the same as that of the Milky Way and that, if depletions of some elements onto dust are not taken into account correctly, it is possible to misinterpret the clues to early nucleosynthesis provided by nonsolar element ratios.