A log N(HI) = 22.6 DLA in a dark gamma-ray burst: the environment of GRB 050401
The optical afterglow spectrum of GRB050401 (at z=2.8992+/-0.0004) shows the presence of a DLA, with log(nHI)=22.6+/-0.3. This is the highest column density ever observed in a DLA, and is about five times larger than the strongest DLA detected so far in any QSO spectrum. From the optical spectrum, we also find a very large Zn column density, allowing us to infer an abundance of [Zn/H]=-1.0+/-0.4. These large columns are supported by the X-ray spectrum from Swift-XRT which shows a column density (in excess of Galactic) of log(nH)=22.21^{+0.06}_{-0.08} assuming solar abundances (at z=2.9). The comparison of this X-ray column density, which is dominated by absorption due to alpha-chain elements, and the HI column density derived from the Ly-alpha absorption line, allows us to derive a metallicity for the absorbing matter of [alpha/H]=-0.4+/-0.3. The optical spectrum is reddened and can be well reproduced with a power-law with SMC extinction, where A_V=0.62+/-0.06. But the total optical extinction can also be constrained in a way which is independent of the shape of the extinction curve: from the optical-to-X-ray spectral energy distribution we find, 0.5<~A_V<~4.5. However, even this upper limit, independent of the shape of the extinction curve, is still well below the dust column that is inferred from the X-ray column density, i.e. A_V=9.1^{+1.4}_{-1.5}. This discrepancy might be explained by a small dust content with high metallicity (low dust-to-metals ratio). `Grey' extinction cannot explain the discrepancy since we are comparing the metallicity to a measurement of the total extinction (without reference to the reddening). Little dust with high metallicity may be produced by sublimation of dust grains or may naturally exist in systems younger than a few hundred Myr.