1999ApJ...515..487T

We analyze the spectra of 12 high-redshift QSOs in the region of Mg II λ2798. Two six-spectrum composites, with <z≥4.47 and 3.35, are compared with the Francis et al. Large Bright Quasar Survey (LBQS) composite (<z≳0.8) and show remarkable similarity. The equivalent width of Mg II is somewhat larger in the LBQS composite than in the high-redshift composites; this may be identified as a weak Baldwin effect. The equivalent width of the Fe II blends to either side is the same in the <z≥3.35 composite as in the LBQS composite and marginally larger in the <z≥4.47 composite. We derive strong confidence limits on the absence of a significant increase in either the Fe II equivalent width or the Fe II/Mg II line ratio from the earliest epoch to the present. Hamann & Ferland predicted that the Fe II/Mg II ratio at high redshift should be a factor of ∼3 lower than for low-redshift QSOs if the age of the universe at the earlier epoch is much less than 1 Gyr. This is due to the delayed contribution of supernovae type Ia to the iron abundance. A 1 Gyr minimum age of the universe at z=4.47 gives interesting upper limits to q₀ for a given H₀ (e.g., q₀<0.20 for H₀=75 km.s^–1.Mpc^–1). The lack of change in Fe II/Mg II between the two high-redshift epochs indicates that 1 Gyr may be an underestimate of the universe age at z=4.47.

We also discuss alternate explanations. The abundance evolution in the central regions of galaxies that host QSOs may follow a significantly different path to that in the solar neighborhood, perhaps from the dominance of short-lived type Ia supernovae progenitors or through the lack of any significant type Ia contribution, in which case we are yet underestimating the efficiency of Fe⁺ ions in QSO broad-line clouds for production of the observed Fe II. Cosmologies with terms beyond zero-pressure matter are also briefly discussed. These considerations notwithstanding, the lack of observed Fe II evolution at high redshifts puts unique constraints on luminous QSO broad-line physics, nucleosynthesis in the QSO host galaxies, and/or cosmological parameters.