2003ApJ...590..296R

This paper presents an analysis of [Fe IV] emission based on new identifications and previous measurements of [Fe IV] lines in 30 Doradus, IC 4846, M42, SMC N88A, and SBS 0335-052. The Fe abundances obtained by adding the abundances of the relevant Fe ions (mainly Fe⁺⁺ and Fe³⁺) are found to be lower, by factors in the range 2.6-5.9, than the Fe abundances implied by [Fe III] emission and an ionization correction factor derived from ionization models. The most likely explanation of this discrepancy is that either the collision strengths for [Fe IV] or the Fe ionization fractions predicted by models are unreliable. The available data allow one neither to distinguish between these two possibilities nor to exclude another possible explanation: that the discrepancy implies the presence of a gradient in the Fe abundance within the ionized gas. Further measurements of [Fe IV] lines and checks on the Fe³⁺ atomic data and ionization models are needed to reach a definitive conclusion. The discrepancy introduces an uncertainty in the determination of Fe abundances in ionized nebulae. This uncertainty has implications for our understanding of both the evolution of dust in ionized nebulae and the chemical history of low-metallicity galaxies.