The Gemini Deep Deep Survey.: II.: Metals in star-forming galaxies at redshift 1.3<z<2

The goal of the Gemini Deep Deep Survey (GDDS) is to study an unbiased sample of K<20.6 galaxies in the redshift range 0.8<z<2.0. Here we determine the statistical properties of the heavy-element enrichment in the interstellar medium (ISM) of a subsample of 13 galaxies with 1.34<z<1.97 and UV absolute magnitude M-2000<-19.65. The sample contains 38% of the total number of identified galaxies in the first two fields of the survey with z>1.3. The selected objects have colors typical of irregular and Sbc galaxies. Strong [O II] emission indicates high star formation activity in the H II regions (star formation rate similar to13-106 M-circle dot yr(-1)). The high signal-to-noise ratio composite spectrum shows strong ISM Mg II and Fe II absorption, together with weak Mn II and Mg I lines. The Fe II column density, derived using the curve of growth analysis, is log N-Fe II=15.54(-0.13)(+0.23). This is considerably larger than typical values found in damped Lyalpha systems (DLAs) along QSO sight lines, where only 10 of 87 (similar to11%) have log N(Fe II)greater than or equal to15.2. High Fe II column densities are observed in the z=2.72 Lyman break galaxy cB58 (log N-Fe II similar or equal to15.25) and in gamma-ray burst host galaxies (log N(Fe II)similar to14.8-15.9). Given our measured Fe II column density and assuming a moderate iron dust depletion (delta(Fe)similar to1 dex), we derive an optical dust extinction of A(V)similar to0.6. If the H I column density is log N-HI<21.7 (as in 98% of DLAs), then the mean metallicity is Z/Z(circle dot)>0.2. The high completeness of the GDDS sample implies that these results are typical of star-forming galaxies in the 1<z<2 redshift range, an epoch that has heretofore been particularly challenging for observational programs.