2007A&A...470..523D

We present the observations of a weak MgII absorption system detected at z∼0.452 in the UVES high-resolution spectrum of the QSO HE0001-2340. The weaker of the two MgII components forming the system shows associated absorptions due to SII, CaI, and FeI observed for the first time in a QSO spectrum. We investigate the nature of this absorber by comparing its properties with those of different classes of absorbers and reproducing its ionization conditions with photoionization models. We measured column densities of MgI, MgII, SII, CaI, CaII, MnII, FeI, and FeII with Voigt profile fitting. Although most of the observed velocity profiles are not resolved in the UVES spectrum, a curve of growth analysis excluded a significant underestimation of the column densities, in particular for the FeII and MgII multiplets. We compared our measurements with a sample of 28 weak MgII systems detected in the interval 0.4<z<1.4 in the 18 UVES spectra of the ESO Large Programme, plus 24 weak MgII systems in the same redshift range taken from the literature. Then, we performed a comparison with 11 damped systems for which ionic column densities were measured component by component and with 44 Galactic lines of sight with measured CaI, CaII and FeI. We also ran a grid of photoionization models with Cloudy to reproduce the observed MgI/MgII, CaI/CaII, and FeI/FeII column density ratios for the studied system. The observed absorber belongs to the class of weak MgII systems on the basis of its equivalent width, however the relative strength of commonly observed transitions deviates significantly from those of the above mentioned absorbers. A rough estimate of the probability of crossing such a system with a QSO line of sight is P∼0.03. The presence of rare neutral transitions suggests that the cloud is shielded by a large amount of neutral hydrogen. A detailed comparison of the observed column densities with the average properties of damped Lyman-α systems and local interstellar cold clouds shows, in particular, deficient MgII/MgI and CaII/CaI ratios in our cloud. The results of photoionization models indicate that the cloud could be ionized by the UV background. However, a simple model of a single cloud with uniform density cannot reproduce the observed ionic abundance ratios, suggesting a more complex density structure for the absorber. Supposing that ionization corrections are negligible, the most puzzling result is the underabundance of magnesium with respect to iron which is hard to explain both with nucleosynthesis and with differential dust depletion.