It has been suggested that damped Lyman absorption systems (DLAAS) - massive clouds of gas which produce saturated Lyman absorption features in the spectra of background quasars - are the high-redshift progenitors of current disk galaxies (Wolfe 1993). The properties of DLAAS are therefore interesting in terms of the formation process of galaxies such as our own (Lanzetta et al. 1991).
There is some evidence that the number or size of DLAAS evolves over the redshift range (White, Kinney & Becker 1993), which implies the conversion of gas into stars at ; indeed, some DLAAS at have metallicities of 10-20 per cent Solar (Pettini et al. 1994), and some contain sufficient dust to slightly redden their background quasars (Pei et al. 1991). Although their interstellar medium has clearly been enriched, direct evidence of star formation has yet to be demonstrated (Hu et al. 1993).
If DLAAS are protogalactic systems then they should have reservoirs of low-metallicity gas and high star-formation rates. In support of this, Frayer et al. (1994) reported CO(1-0) and (more tentatively) CO(3-2) emission from a DLAAS towards the quasar PC 1643 4631A (Schneider, Schmidt & Gunn 1991); their data were consistent with the gas being clumped, with dimensions similar to those of a galactic disk and a mass of around 1012 M (we assume , km s-1 Mpc-1 throughout this paper). The CO luminosity was estimated to be several orders of magnitude greater than that of the Milky Way.
Braine et al. (1996) attempted to confirm the Frayer et al. result using the IRAM interferometer, but concluded that the earlier CO detections were spurious. They pointed out that an interferometer is less prone to distorted baselines than the conventional single-dish approach, and noted that its ideal application is the detection of broad, weak lines from sources smaller than the primary beam (in this case, at 3 mm). It is worth mentioning, however, that weak sources spread over ( kpc) would be heavily resolved by the IRAM interferometer and hence very difficult to detect. The NRAO 140-ft and 12-m dishes used by Frayer et al. have HPBWs of and (115-135 kpc), so the CO detections of Frayer et al. can be understood (in the context of the Braine et al. data) only if the molecular gas proves to be extended on scales an order of magnitude larger than the Milky Way. (CO has also been reported towards PKS 0528-250 in a DLAAS at - Brown & Vanden Bout 1992 - and, again, the validity of the detection was disputed - Wiklind & Combes 1994).
In a galactic environment, skins of atomic gas are thought to cover each clump of molecular gas, with the thickness of the skin determined by the abundance of dust, which provides protection from the global UV field for the CO molecules within. In the atomic skin, incident UV photons left over from dissociating and ionizing are absorbed by dust grains, which cool via far-IR continuum emission; photoelectrons, ejected as a consequence of UV absorption by the grains, heat the atomic and molecular hydrogen. The C ions are then collisionally excited and emit the [C II ] ( ) fine-structure line (Hollenbach, Takahashi & Tielens 1991; Mochizuki et al. 1994). The [C II ] emission is an important cooling process in galaxies, accounting for up to 1 per cent of the far-IR luminosity (Stacey et al. 1991).
Our objective here was to search for [C II ] - an unmistakable signature of star-formation activity - in a DLAAS with a large column density of neutral hydrogen, cm-2 (White et al. 1993), and where CO had apparently been detected, and to thereby independently confirm that star formation is ongoing in that system and that enriched gas is present. Demonstrating the potential of [C II ] as a probe of metallicity, of DLAAS star-formation history, and of the evolution of galaxies at these redshifts, would represent a major advance in our studies of the early Universe.
For nearby galaxies, the rest frequency of the [C II ] transition (1.900537 THz) means that the line is inaccessible from the ground and, to date, all detections in the near Universe have been made by the balloon-borne experiments (Mochizuchi et al. 1994), the Kuiper Airbourne Observatory (Stacey et al. 1991) or the Cosmic Background Explorer (Bennett et al. 1994). However, at high redshifts the [C II ] line is shifted into windows observable from Mauna Kea with the 15-m JCMT; specifically, for the line appears in the B window, and for it appears in the C window. To date, there have been no detections of highly redshifted [C II ] (e.g. Isaak et al. 1994).
© European Southern Observatory (ESO) 1998
Online publication: January 16, 1998