NEW PHOTOIONIZATION MODELS OF INTERGALACTIC CLOUDS

We present new photoionization models of optically thin, low-density intergalactic gas at constant pressure, photoionized by QSOs. To improve accuracy of ionization calculations, our code uses Opacity Project photoionization cross sections for the ground states of C, N, O, and Si and makes approximate corrections for the radiative recombination rates using detailed balance. We model all ion stages of H, He, C, N, O, Si, and Fe, plus H-2 and predict the column density ratios of clouds at specified values of ionization parameter U = n-gamma/n(H) and cloud metallicity. For gas photoionized by quasars, a two-phase (cloud-intercloud) medium cannot exist at T = (1-3) x 10(4) K, so that pressure confinement of the Ly-alpha clouds is problematic. We derive new metallicity limits for absorption systems at z = 3.321 (S5 0014 + 81) and at z = 2.9676 (PKS 2126-158). In addition, we find a stable molecular phase in Ly-alpha clouds, with U almost-equal-to 10(-7) to 10(-5) and n(H-2)/n(H I) almost-equal-to 10(-2), in which molecular cooling produces T < 10(4) K, as suggested by recent line width data. Photoelectron escape from low-column H I clouds can reduce the heating and explain the observed narrow line widths. If the ionizing radiation flux J0 = 4-pi-I0 almost-equal-to 10(-21) ergs cm-2 s-1 Hz-1, as inferred for several absorption systems in QSO 2206-199N, these 10(4) K clouds require high densities (n(H) almost-equal-to 0.1 cm-3) and have significant neutral fractions. The inferred cloud thicknesses are 10(5-9) times smaller than the lateral extents estimated by double-QSO observations. Such aspect ratios are difficult to understand, and the large pressures would make the sheet like clouds difficult to confine by means other than ram pressure. The value of U required to explain the H I line widths is also much smaller than that needed to explain the observed ratios N(C IV)/N(C II) greater-than-or-equal-to 1 in sharp metal-line systems.