First Comparison of Ionization and Metallicity in Two Lines of Sight toward HE 1104−1805 AB atz=1.66

Using new Hubble Space Telescope Faint Object Spectrograph, the New Technology Telescope ESO Multi Mode Instrument, and Keck HIRES spectra of the gravitationally lensed double QSO HE 1104-1805 AB ( z em =2.31) and assuming UV photoionization by a metagalactic radiation field, we derive physical conditions (ionization levels, metal abundances, and cloud sizes along the lines of sight) in five C IV+Mg II absorption systems clustered around z =1.66 along the two lines of sight. Three of these systems are associated with a damped Lyα (DLA) system with log N (H I)=20.85, which is observed in the UV spectra of the bright QSO image, A. The other two systems are associated with a Lyman-limit system with log N (H I)=17.57 seen in the fainter image, B. The C IV and Mg II line profiles in A resemble those in the B spectra and span Δ v ≈360 km s -1 . The angular separation, θ=3.195'', between A and B corresponds to a transverse proper separation of S ⊥ =8.3 h −1 50 kpc for q 0 =0.5 and a lens at z =1. Assuming that the relative metal abundances in these absorption systems are the same as those observed in the DLA system, we find that the observed N (C IV)/ N (Mg II) ratios imply ionization parameters of logΓ=-2.95 to -2.35. Consequently, these clouds should be small (0.5-1.6 kpc with a hydrogen density of n H ≲0.01 cm -3 ) and relatively highly ionized. The absorption systems to B are found to have a metallicity 0.63 times lower than the metallicity of the gas giving rise to the DLA system, Z DLA ≃1/10 Z ☉ . We detect O VI at z =1.66253 in both QSO spectra but no associated N V. Our model calculations lead us to conclude that the C IV clouds should be surrounded by large (~100 kpc), highly ionized low-density clouds ( n H ~10 −4 cm -3 ) in which O VI, but only weak C IV, absorption occurs. In this state, logΓ≥-1.2 reproduces the observed ratio of N (O VI)/ N (N V)>60. These results are discussed in view of the disk/halo and hierarchical structure formation models.