2017MNRAS.466.3133H

Ne VIII absorbers seen in QSO spectra are useful tracers of warm ionized gas, when collisional ionization is the dominant ionization process. While photoionization by the ultraviolet background (UVB) is a viable option, it tends to predict large line-of-sight thickness for the absorbing gas. Here, we study the implications of the recently updated UVB at low z to understand the ionization mechanisms of intervening Ne VIII absorbers. With the updated UVB, one typically needs higher density and metallicity to reproduce the observed ionic column densities under photoionization. Both reduce the inferred line-of-sight thicknesses of the absorbers. We find a critical density of >=5 x 10^–5 cm^–3 above which the observed N(Ne VIII)/ N(O VI) can be reproduced by pure collisional processes. If the gas is of near solar metallicity (as measured for the low ions) then the cooling time-scales will be small (<10⁸ yrs). Therefore, a continuous injection of heat is required in order to enhance the detectability of the collisionally ionized gas. Using photoionization models we find that in almost all Ne VIII systems the inferred low ion metallicity is near solar or supersolar. If we assume the Ne VIII phase to have similar metallicities then photoionization can reproduce the observed N(Ne VIII)/ N(O VI) without the line-of-sight thickness being unreasonably large and avoids cooling issues related to the collisional ionization at these metallicities. However, the indication of broad Lyα absorption in a couple of systems, if true, suggests that the Ne VIII phase is distinct from the low ion phase having much lower metallicity.