Juliet C. Pickering

NLTE analysis of Co i /Co ii lines in spectra of cool stars with new laboratory hyperfine splitting constants★

The analysis of stellar abundances for odd-Z Fe-peak elements requires accurate non-local thermodynamic equilibrium (NLTE) modelling of spectral lines fully taking into account the hyperfine structure (HFS) splitting of lines. Here, we investigate the statistical equilibrium of Co in the atmospheres of cool stars and the influence of NLTE and HFS on the formation of Co lines and abundances. Significant departures from LTE level populations are found for Co I; number densities of excited states in Co II also differ from LTE at low metallicity. The NLTE level populations are used to determine the abundance of Co in solar photosphere, log = 4.95 ± 0.04 dex, which is in agreement with that in C I meteorites within the combined uncertainties. The spectral lines of Co I were calculated using the results of recent measurements of hyperfine interaction constants by UV Fourier transform spectrometry. For Co II, the first laboratory measurements of HFS A and B factors were performed. These highly accurate A factor measurements (errors of the order of 3-7 per cent) allow, for the first time, reliable modelling of Co II lines in the solar and stellar spectra and, thus, a test of the Co I/Co II ionization equilibrium in stellar atmospheres. A differential abundance analysis of Co is carried out for 18 stars in the metallicity range -3.12 < [Fe/H] < 0. The abundances are derived by the method of spectrum synthesis. At low [Fe/H], NLTE abundance corrections for Co I lines are as large as +0.6, ... , + 0.8 dex. Thus, LTE abundances of Co in metal-poor stars are severely underestimated. The stellar NLTE abundances determined from the single UV line of Co II are lower by ∼0.5-0.6 dex. The discrepancy might be attributed to possible blends that have not been accounted for in the solar Co II line and its erroneous oscillator strength. The increasing [Co/Fe] trend in metal-poor stars, as calculated from the Co I lines under NLTE, can be explained if Co is overproduced relative to Fe in massive stars. The models of Galactic chemical evolution are wholly inadequate to describe this trend suggesting that the problem is in supernova yields.

Fe i oscillator strengths for the Gaia-ESO survey

The Gaia-ESO Public Spectroscopic Survey (GES) is conducting a large-scale study of multielement chemical abundances of some 100 000 stars in the Milky Way with the ultimate aim of quantifying the formation history and evolution of young, mature and ancient Galactic populations. However, in preparing for the analysis of GES spectra, it has been noted that atomic oscillator strengths of important Fe I lines required to correctly model stellar line intensitiesaremissingfromtheatomicdatabase.Here,wepresentnewexperimentaloscillator strengths derived from branching fractions and level lifetimes, for 142 transitions of Fe I between 3526 and 10 864 A, of which at least 38 are urgently needed by GES. We also assess the impact of these new data on solar spectral synthesis and demonstrate that for 36 lines that appear unblended in the Sun, Fe abundance measurements yield a small line-by-line scatter (0.08 dex) with a mean abundance of 7.44 dex in good agreement with recent publications.

Atomic and molecular data for optical stellar spectroscopy

High-precision spectroscopy of large stellar samples plays a crucial role for several topical issues in astrophysics. Examples include studying the chemical structure and evolution of the Milky Way galaxy, tracing the origin of chemical elements, and characterizing planetary host stars. Data are accumulating from instruments that obtain high-quality spectra of stars in the ultraviolet, optical and infrared wavelength regions on a routine basis. These instruments are located at ground-based 2- to 10-m class telescopes around the world, in addition to the spectrographs with unique capabilities available at the Hubble Space Telescope. The interpretation of these spectra requires high-quality transition data for numerous species, in particular neutral and singly ionized atoms, and di- or triatomic molecules. We rely heavily on the continuous efforts of laboratory astrophysics groups that produce and improve the relevant experimental and theoretical atomic and molecular data. The compilation of the best available data is facilitated by databases and electronic infrastructures such as the NIST Atomic Spectra Database, the VALD database, or the Virtual Atomic and Molecular Data Centre (VAMDC). We illustrate the current status of atomic data for optical stellar spectra with the example of the Gaia-ESO Public Spectroscopic Survey. Data sources for 35 chemical elements were reviewed in an effort to construct a line list for a homogeneous abundance analysis of up to 100000 stars.

The spectrum and term analysis of Co II

High-resolution spectra of singly ionized cobalt have been recorded by Fourier transform spectrometry in the region 1420-33333 A (70422-3000 cm-1) with cobalt-neon and cobalt-argon hollow cathode lamps as sources. Most of the Co II lines exhibit broadening due to hyperfine structure, but the wavenumber uncertainty for the center of gravity of the strongest lines is less than 0.002 cm-1. Of the previously listed energy levels of Co II, 215 have been revised and six have been discarded. In addition, 171 new levels have been found of which 125 are reported here for the first time. The number of classified lines has doubled with wavenumbers and classifications being presented for 2373 Co II lines, of which 1242 involve the new levels.

Measurements of the Hyperfine Structure of Atomic Energy Levels in Co I

Most of the Co I lines recorded by high-resolution Fourier Transform Spectrometry in the region 3000–45000 cm−1 (30000–2222 A) show significant hyperfine Structure. Computer fits to 1020 of these line profiles have yielded values of the magnetic dipole hyperfine interaction constant A for 297 energy levels with an uncertainty of between 0.2% and 3%. The A Factors range from −0.0236 to +0.0598 cm−1 for the even levels, and from −0.03387 to +0.1028 cm−1 for the odd levels. For 208 of these A factors no previous measurements are known. Approximate values of the electric quadrupole hyperfine interaction constant B were found for 156 levels. These measurements of A and B factors allow, for the first time, correct allowance to be made for the effects of hyperfine structure in Co I both for abundance analysis and in the resolution of blended lines.

High‐resolution photoabsorption cross‐section measurements of SO2 at 160 K between 199 and 220 nm

Photoabsorption cross sections of sulfur dioxide over a range of temperatures are required to interpret observations and to support models of the atmospheres of Io and Venus. We report high-resolution (lambda/Delta lambda approximate to 450,000) photoabsorption cross-section measurements by Fourier transform spectrometry of SO2 at 160 K in the wavelength region 199 to 220 nm, which encompasses the strongest features in the prominent (C) over tilde (1) B-2 -(X) over tilde (1)A(1) system. Our results are compared with literature values obtained at lower resolutions and with 295 K cross sections recorded earlier with the same instrument.

Hyperfine Structure Measurements of Neutral Manganese with Fourier Transform Spectroscopy

We report experimental hyperfine structure constants of levels in the neutral manganese atom, measured using Fourier transform spectroscopy of hollow cathode discharges. In total, 208 spectral lines of astrophysical interest have been analyzed to obtain hyperfine structure constants for 106 levels in Mn i, of which 67 have no previous hyperfine structure measurements. The uncertainties in the magnetic dipole constants, A ,a re between 1; 10 � 4 and 5 ;10 � 4 cm � 1 . Hyperfine structure constants for an additional 18 levels compiled from the literature are also given. Subject headingg atomic data — line: profiles — methods: laboratory

Fe I oscillator strengths for transitions from high-lying even-parity levels

New radiative lifetimes, measured to ±5% accuracy, are reported for 31 even-parity levels of Fe I ranging from 45061 cm–1 to 56842 cm–1. These lifetimes have been measured using single-step and two-step time-resolved laser-induced fluorescence on a slow atomic beam of iron atoms. Branching fractions have been attempted for all of these levels, and completed for 20 levels. This set of levels represents an extension of the collaborative work reported in Ruffoni et al. The radiative lifetimes combined with the branching fractions yields new oscillator strengths for 203 lines of Fe I. Utilizing a 1D-LTE model of the solar photosphere, spectral syntheses for a subset of these lines which are unblended in the solar spectrum yields a mean iron abundance of log[e(Fe)] = 7.45 ± 0.06.

The FERRUM project: Branching ratios and atomic transition probabilities of Fe II transitions from the 3d 6 (a 3 F)4p subconguration in the visible to VUV spectral region ?

We report measurements of the relative intensities of 81 emission lines of Fe II between 160 nm and 350 nm (62 168 cm 1 to 28 564 cm 1 ) from 4 levels by high resolution Fourier transform spectrometry, using a Penning discharge lamp as light source. These relative intensities have been used to determine the line branching fractions, which have then been combined with accurate experimental radiative lifetime measurements reported recently to give absolute transition probabilities and oscillator strengths for 81 lines. The accuracy of these f-values is compared with other previous experimental measurements, and with theoretical values. The new transition probabilities will allow accurate determinations of Fe II abundances in a wide variety of astrophysical objects.

Goddard High-Resolution Spectrograph Observations of the B III Resonance Doublet in Early B Stars: Abundances and Isotope Ratios

Boron abundances and isotope ratios in two early B stars have been measured by using the Hubble Space Telescopes Goddard High-Resolution Spectrograph (GHRS) in order to observe the B III resonance doublet near 2066 A. We also report new experimental and theoretical results for the wavelengths, isotope shift, and hyperfine structure of these boron lines. In HD 886 (γ Pegasi; B2 IV), and HD 35299 (B1 V) the abundance of boron is about a factor of 2 lower than that inferred from solar system meteorites. The 11B/10B isotope ratio in HD 886 is determined to be 4.7, and for HD 35299 a value of 3.7 is found. These values are consistent with the solar system isotope ratio. In a third star, HD 3360 (ζ Cassiopeiae; B2 IV), our GHRS observations show that the boron abundance is a factor of 40 below the solar system abundance. Our results for HD 886 and HD 35299 appear to be consistent with little dispersion in the initial boron-to-oxygen ratio among Population I stars and a universal 11B/10B isotope ratio. However, the abundance measured for HD 3360 shows that either some stars suffer significant boron depletion while still in the core-hydrogen burning phase of their evolution, or there is indeed a very wide variation in initial Population I boron abundances. A much-improved understanding of boron astration in early B stars will be required before this question can be resolved.