The Chemical Enrichment History of Damped Lyman-alpha Galaxies Limin Lu, Wallace L. W. Sargent, & Thomas A. Barlow Caltech, 105-24, Pasadena, CA 91125 (email: [email protected]) 6 9 9 Abstract. Studies of damped Lyα absorption systems in quasar spec- 1 tra are yielding very interesting results regarding the chemical evolution n of these galaxies. We describe some preliminary results from such a pro- a gram. J 9 1 1 1. Introduction v 8 Damped Lyα absorption systems in quasar spectra are generally believed to 9 trace theabsorption frominterstellar gas inhigh-redshiftgalaxies, possiblyfrom 0 the disks or proto-disks of spirals (Wolfe 1988). They can be studied in the 1 0 redshift range 0 < z < 5 by combining UV and optical observations. The 6 damped Lyα galaxies are particularly suited for probing the chemical evolution 9 of galaxies over a large fraction of the Hubble time for several reasons: (1) they / h are relatively common and easy to identify in quasar spectra, so building up a p large sample is possible; (2) given their large neutral hydrogen column densities o- (N(HI)∼ 1020 − 1022 cm−2), most of the absorbing gas should be neutral so r ionization corrections should be minimal (cf. Viegas 1995); (3) the damped t s Lyα galaxies should be relatively representative of galaxies at high redshifts a since they are selected simply because they happen to lie in front of background : v quasars. i X The first systematic investigation of the chemical evolution of damped Lyα galaxies was conducted by Pettini and collaborators (Pettini et al. 1994; 1995), r a who studied the Zn and Cr abundances in ∼ 20 damped Lyα galaxies. The advent of the 10-m Keck telescope allows us to carry out similar investigations in a much more detailed fashion. In this short contribution, we present some preliminary results from such a program. Detailed analysis and discussion may be found in Lu, Sargent, & Barlow (1996; hereafter LSB96). 2. Results Figure 1 shows the abundance results so far obtained from our Keck program, with the addition of selected measurements from published papers where we believe the effect of line saturation has been treated properly. We also correct (when applicable) the abundance measurements from previous papers for the set of new oscillator strengths compiled by Tripp, Lu, & Savage (1995) so that all the measurements will be on the same footing. References to the data used in constructing figure 1 may be found in LSB96. 1 Figure 1. (a) Age-metallicity relation for our sample of damped Lyα galaxies. The conversion from redshift to age is calculated for q0 = 0.5 and H0 = 50. (b)-(f) Abundance ratios of selected elements found in damped Lyα galaxies. The notion [Fe/H] has the meaning [Fe/H]=log(Fe/H)damp−log(Fe/H)⊙, and similarly for others. Typical measurement errors of the abundances are 0.1 dex. 2.1. Age-Metallicity Relation Figure 1(a) shows the age-metallicity relation for our sample of damped Lyα galaxies (filled circles). The solid curve roughly indicates the age-metallicity relation for disk stars in the Galactic solar neighborhood determined by Ed- vardsson et al. (1993). We note the following: (1) The damped Lyα galaxies have Fe-metallicities ([Fe/H]) in the range of 1/10 to 1/300 solar, thus representing a population of very young galaxies at least in terms of the degree of chemical enrichment. (2) The mean metallicity appears to increase with age, providing direct evidence for the buildup of heavy elements in galaxies. It may be significant that all the four galaxies with z > 3 have [Fe/H]< −1.7, while at 2 < z < 3 at least some galaxies have achieved much higher metallicities. This may signal 2 an epoch of rapid star formation in galaxies. We also note that the intrinsic trend of increasing metallicity with age would be stronger if Fe is somewhat depleted by dust in these galaxies because the depletion should be the least for the highest redshift galaxies (see section 2.2, however). (3)Clearly thedampedLyαgalaxies have muchlower metallicities thanthe MilkyWay diskatanygiven timeinthepast. Thismaybearsignificantly onthe nature of the damped Lyα galaxies. It was suggested initially (cf. Wolfe 1988) that the damped Lyα absorbers may trace disks or proto-disks of high-redshift spirals. But the low metallicities of damped Lyα galaxies cast some doubts on this interpretation. Timmes, Lauroesch, & Truran (1995; also see Timmes 1995, this volume) suggested that the abundance measurements of damped Lyα galaxies are consistent with the chemical enrichment history of the Milky Way disk if the enrichment process in damped Lyα galaxies is delayed by ∼3 Gyrs for some reason; this seems to place the Milky Way at a privileged position. On the other hand, the metallicities found for our sample of damped Lyα galaxies are very similar to those found for Galactic halo stars and globular clusters, suggestingthepossibilitythatdampedLyαabsorbersmayrepresentaspheroidal component of high-redshift galaxies. This possibility has in fact already been suggested by Lanzetta, Wolfe, & Turnshek(1995) basedon considerations of gas consumptions in these galaxies. 2.2. Abundance Ratios and Nucleosynthesis Panels (b)-(f) of figure 1 show the abundance ratios of various elements in damped Lyα galaxies relative to their corresponding solar ratios. Elemen- tal abundance ratios, in principle, allow one to gain insight of what kind of nucleosynthetic processes may be responsible for the enrichment of the inter- stellar medium. For example, the well-documented overabundance of even-Z (Z=atomicnumber)α-groupelements relative totheFe-peak elements inGalac- tic halo stars is believed to reflect the nucleosynthetic products of massive stars through SN II explosions (cf. Wheeler, Sneden, & Truran 1989). It is interest- ing that the observed abundance patterns of N/O, Si/Fe, Cr/Fe and Mn/Fe in dampedLyαgalaxies areallconsistentwithmeasurementsinGalactic halostars (cf. Wheeler et al. 1989). In particular, we note that the observed N/O ratios are not easily explained with dust depletions because N and O are largely unaf- fected by dust in the Galactic ISM. The observed Mn/Fe ratios are also difficult toexplainwithdustdepletionsbecauseintheGalacticISMdustdepletionscause the gas-phase Mn/Fe ratio to be higher than the solar ratio, opposite to what is observed in damped Lyα galaxies. On the other hand, these ratios are easily understood in terms of the odd-even effect (ie, the odd-Z elements generally show underabundances relative to the even-Z elements of same nucleosynthetic origin at low Fe metallicities) and the different nucleosynthetic origins of these elements (cf. Wheeler et al. 1989). These results strongly indicate that we have observed these galaxies during the epoch when SN II are largely responsible for theenrichment of the interstellar mediumin thesegalaxies, whilelow mass stars have not had enough time to evolve and to dumptheir nucleosynthetic products into the interstellar medium through mass loss and SN Ia. Thus the chemical enrichmentprocessinthesegalaxies shouldnothaveproceededmorethan1Gyr when they were observed. 3 However, the observed Zn/Fe ratio in damped Lyα galaxies is inconsistent with the above nucleosynthesis interpretation. In Galactic stars, Zn/Fe is found to be solar at all metallicities (cf. Wheeler et al. 1989 and references therein). Thisdifferencemaysuggestthat,whiletheobservedrelativeabundancepatterns in damped Lyα galaxies are dominated by the effects of nucleosynthesis, there is somedustdepletioneffectontopofthat. Thepresenceofasmallamountofdust in dampedLyα galaxies has been claimed from the reddeningof the background quasars (cf. Pei, Fall, & Bechtold 1991). On the other hand, recent theoretical studies indicate that Zn can be produced in large quantities in the neutrino driven winds during SN II explosions (Hoffman et al. 1995; see also Woosley 1995, this volume). Since SN II makes little Fe, a Zn/Fe overabundance may be possible in the ejecta of SN II. The puzzle is then why Zn is observed to track Fe abundance in Galactic stars. If indeed the observed Zn/Fe overabundance in damped Lyα galaxies is caused by depletion of Fe onto dust grains, the [Fe/H] measurementsinfigure1(a)willunderestimatethetrueFe-metallicities by∼ 0.5 dex (on average). 3. Concluding Remarks Damped Lyα galaxies provide the unprecedented opportunity to directly probe the chemical enrichment history of galaxies over a large fraction of the Hubble time. Some intriguing results have already emerged from the current study. However, many questions remain, eg, why do the damped Lyα galaxies have so low metallicities compared to the past history of the Milky Way disk, and what are the implications? What is the significance of the large scatter in the measured[Fe/H]atanygiveredshift? Howbigaroledoesdustplayinmodifying the observed abundances and their interpretations? Some of these issues will be addressed in more details in LSB96. References Edvardsson, B., et al. 1993, A&A, 275, 101 Hoffman, R.D. et al. 1995, ApJ, submitted Lanzetta, K.M., Wolfe, A.M., & Turnshek, D.A. 1995, ApJ, 440, 435 Lu, L., Sargent, W.L.W., & Barlow, T.A. 1996, in preparation (LSB96) Pei, Y., Fall, S.M., & Bechtold, J. 1991, ApJ, 378, 6 Pettini, M. et al. 1994, ApJ, 426, 79 Pettini, M. et al. 1995, in QSO Absorption Lines, ed. G. Meylan (Springer- Verlag), 71 Timmes, F.X., Lauroesch, J.R., & Truran, J.W. 1995, ApJ, 451, 468 Tripp, T.M., Lu, L., & Savage, B.D. 1995, ApJS, in press Viegas, S. M. 1995, MNRAS, 276, 268 Wheeler, J.C., Sneden, C., & Truran, J.W. Jr. 1989, ARA&A, 27, 279 Wolfe, A.M. 1988, in QSO Absorption Lines: Probing the Universe, eds. J.C. Blades, D.A. Turnshek, & C.A. Norman (Cambridge Univ Press), 297 4