New Fe I level energies and line identifications from stellar spectra

Abstract

The spectrum of the Fe I atom is critical to many areas of astrophysics and beyond. Measurements of the energies of its high-lying levels remain woefully incomplete, however, despite extensive analysis of ultraviolet laboratory iron absorption spectra, optical laboratory iron emission spectra, and the solar infrared spectrum. In this work we use as sources the high-resolution archival absorption-line ultraviolet and optical spectra of stars, whose warm temperatures favor moderate Fe I excitation. We derive the energy for a particular upper level in Kurucz's semiempirical calculations by adopting a trial value that yields the same wavelength for a given line predicted to be about as strong as that of a strong unidentified spectral line observed in the stellar spectra, then checking the new wavelengths of other strong predicted transitions that share the same upper level for coincidence with other strong observed unidentified lines. To date this analysis has provided the upper energies of 66 Fe I levels. Many new level energies are higher than those accessible to laboratory experiments; several exceed the Fe I ionization energy. These levels provide new identifications for over two thousand potentially detectable lines. Almost all of the new levels of odd parity include UV lines that were detected but unclassified in laboratory Fe I absorption spectra, providing an external check on the energy values. We motivate and present the procedure, provide the resulting new level energies and their uncertainties, list all the potentially detectable UV and optical new Fe I line identifications and their gf-values, point out new lines of astrophysical interest, and discuss the prospects for additional Fe I energy-level determinations in the near future.