Physical conditions of five O vi absorption systems towards PG 1522+101

ABSTRACT We present the analysis of five O vi absorbers identified across a redshift path of z ∼ (0.6−1.3) towards the background quasar PG 1522+101 with information on five consecutive ionization stages of oxygen from O ii to O vi. The combined HST and Keck spectra cover UV, redshifted extreme-UV, and optical transitions from a multitude of ions spanning ionization energies in the range of ∼(13−300) eV. Low-ionization (C ii, O ii, Si ii, Mg ii) and very high-ionization species (Ne viii, Mg x) are non-detections in all the absorbers. Three of the absorbers have coverage of He i, in one of which it is a >3σ detection. The kinematic structures of these absorbers are extracted from C iv detected in HIRES spectra. The farthest absorber in our sample also contains the detections of Ne v and Ne vi. Assuming co-spatial absorbing components, the ionization models show the medium to be multiphased with small-scale density–temperature inhomogeneities that are sometimes kinematically unresolved. In two of the absorbers, there is an explicit indication of the presence of a warm gas phase (T ≳ 105 K) traced by O vi. In the remaining absorbers, the column densities of the ions are consistent with a non-uniform photoionized medium. The subsolar [C/O] relative abundances inferred for the absorbers point at enrichment from massive Type II supernovae. Despite metal enrichment, the inferred wide range for [O/H] ∼ [−2.1, +0.2] amongst the absorbers along with their anticorrelation with the observed H i suggest poor small-scale mixing of metals with hydrogen in the regions surrounding galaxies and the IGM.