Ann Merchant Boesgaard

During Big Bang nucleosynthesis (BBN) in the first 15 minutes of the Universe, some ⁷Li was created along with isotopes of H and He. The determination of that primordial value of Li can help constrain the conditions at that time. The oldest stars with known ages can be found in globular clusters which have well-determined ages through stellar evolution models. High-resolution spectra of Li have been obtained with the Keck I Telescope and HIRES in several unevolved stars in the clusters M13 and M71 with V magnitudes of 17.6–17.9. Abundances of Li have been determined with spectrum synthesis techniques and show a range of a factor of 4. We attribute that spread to differences in initial angular momentum resulting in different amounts of spin-down, related mixing, and destruction of Li. Our results are compared with similar results for main-sequence and turnoff stars in other globular clusters. The range in age of these clusters is 11.2–14.2 Gyr for an age span of 3 Gyr. These clusters range in [Fe/H] from −0.75 to −2.24 corresponding to a factor of 30 in metallicity. The maximum in the Li abundance for these unevolved stars in all eight clusters is the same corresponding to Li/H = 3.16 × 10⁻¹⁰, while the predicted Li abundance, based on the deuterium abundance and the BBN predictions, is 5.24 × 10⁻¹⁰.

The light element lithium is formed by nucleosynthesis during the Big Bang. Its abundance can help to define the parameters of the early Universe. To find this primordial value, it is necessary to determine Li abundances in the oldest stars because it is readily destroyed by nuclear reactions in stellar interiors. We have made high-resolution (∼45,000) spectroscopic observations of five identical unevolved turnoff stars in the 13 Gyr old globular cluster M5. In our analysis we find a range in Li abundance of a factor of 2; the spread is 5 times the individual error. The comparison of these results with those for turnoff stars from five other globular clusters reveals a similarly large range in Li. Lithium in M5 and the other clusters all have stars above the field star Li plateau, but none are as high as the predictions for primordial Li. The maximum values for Li are the same in all six clusters. Multiple generations of stars are found in many globular clusters; those later generations are expected to have formed from Li-depleted gas. Such second- and later-generation stars would have no Li. However, only one of the six clusters has a few unevolved stars with upper limits on the Li abundance.

We have derived new detailed abundances of Mg, Ca, and the Fe-group elements Sc through Zn (Z = 21−30) for 37 main-sequence turnoff very metal-poor stars ([Fe/H] ≲−2.1). We analyzed Keck HIRES optical and near-UV high signal-to-noise spectra originally gathered for a Be abundance survey. Using typically ∼400 Fe-group lines with accurate laboratory transition probabilities for each star, we have determined accurate LTE metallicities and abundance ratios for neutral and ionized species of the 10 Fe-group elements as well as α elements Mg and Ca. We find good neutral/ion abundance agreement for the six elements that have detectable transitions of both species in our stars in the 3100–5800 Å range. Earlier reports of correlated Sc−Ti−V relative overabundances are confirmed, and appear to slowly increase with decreasing metallicity. To this element trio we add Zn; it also appears to be increasingly overabundant in the lowest-metallicity regimes. Co appears to mimic the behavior of Zn, but issues surrounding its abundance reliability cloud its interpretation.

The light element lithium is formed by nucleosynthesis during the Big Bang. Its abundance can help to define the parameters of the early universe. To find this primordial value, it is necessary to determine Li abundances in the oldest stars because it is readily destroyed by nuclear reactions in stellar interiors. We have made high-resolution (~45,000) spectroscopic observations of five identical unevolved turn-off stars in the 13 Gyr old globular cluster M5. In our analysis we find a range in Li abundance of a factor of two; the spread is five times the individual error. The comparison of these results with those for turn-off stars from five other globular clusters reveals similarly large range in Li. Lithium in M5 and the other clusters all have stars above the field star Li plateau, but none are as high as the predictions for primordial Li. The maximum values for Li are the same in all six clusters. Multiple generations of stars are found in many globular clusters; those later generations are expected to have formed from Li-depleted gas. Such second- and later-generation stars would have no Li. However, only one of the six clusters has a few unevolved stars with upper limits on the Li abundance.

We have derived new detailed abundances of Mg, Ca, and the Fe-group elements Sc through Zn (Z = 21-30) for 37 main sequence turnoff very metal-poor stars ([Fe/H] <= -2.1). We analyzed Keck HIRES optical and near-UV high signal-to-noise spectra originally gathered for a beryllium abundance survey. Using typically about 400 Fe-group lines with accurate laboratory transition probabilities for each star, we have determined accurate LTE metallicities and abundance ratios for neutral and ionized species of the 10 Fe-group elements as well as alpha elements Mg and Ca. We find good neutral/ion abundance agreement for the 6 elements that have detectable transitions of both species in our stars in the 3100-5800 Angstrom range. Earlier reports of correlated Sc-Ti-V relative overabundances are confirmed, and appear to slowly increase with decreasing metallicity. To this element trio we add Zn; it also appears to be increasingly overabundant in the lowest metallicity regimes. Co appears to mimic the behavior of Zn, but issues surrounding its abundance reliability cloud its interpretation.

This is a report on some highlights of research on the rare light elements, lithium (Li), beryllium (Be), and boron (B), that were presented at my Henry Norris Russell Lecture in 2020 January. It is not a comprehensive review of work on these light elements, but contains sections on Big Bang nucleosynthesis of Li and the rarity of these light elements. It includes information on how they are observed, both historically and currently, and the difficulties entailed in determining their abundances. The production of Li, Be, and B is ongoing, so the youngest stars contain the most Li in their atmospheres and they have had less time to destroy it. All three elements are readily destroyed in stellar interiors, but have differing degrees of susceptibility to the particular nuclear fusion reactions which deplete their surface content. This feature makes them remarkably good probes into the otherwise unobservable interiors of stars and provides insights into internal-mixing processes. It also enhances the use of two or more of the three in sorting out the various processes at work in the insides of stars.

The surface content of lithium (Li) and beryllium (Be) in stars can reveal important information about the temperature structure and physical processes in their interior regions. This study focuses on solar-type stars with a sample that is more precisely defined than done previously. Our selection of stars studied for Be is constrained by five parameters: mass, temperature, surface gravity, metallicity, and age to be similar to the Sun and is focused on stars within ±0.02 of 1 M _⊙. We have used the Keck I telescope with HIRES to obtain spectra of the Be ii spectral region of 52 such stars at a high spectral resolution (∼45,000) and high signal-to-noise ratios. While the spread in Li in these stars is greater than a factor of 400, the spread in Be is only 2.7 times. Two stars were without any Be, perhaps due to a merger or a mass transfer with a companion. We find a steep trend of Li with temperature but little for Be. While there is a downward trend in Li with [Fe/H] from −0.4 to +0.4 due to stellar depletion, there is a small increase in Be with Fe from Galactic Be enrichment. While there is a broad decline in Li with age, there may be a small increase in Be with age, though age is less well determined. In the subset of stars closest to the Sun in temperature and other parameters, we find that the ratio of the abundances of Be to Li is much lower than predicted by models; there may be other mixing mechanisms causing additional Li depletion.

This is a report on some highlights of some research on the rare light elements, lithium (Li), beryllium (Be), and boron (B), that I presented in my Henry Norris Russell Lecture in January, 2020. It is not a comprehensive review of work on these light elements, but contains sections on Big Bang nucleosynthesis of Li and the rarity of these light elements. It includes information on how they are observed, both historically and currently, and the difficulties entailed in determining their abundances. %It traces the evolution of these elements through the history of the Galaxy. The production of Li, Be, and B is ongoing so the youngest stars contain the most Li, and they have had less time to destroy it. All three elements are readily destroyed in stellar interiors, but have differing degrees of susceptibility to the particular nuclear fusion reactions which deplete their surface content. This feature makes them remarkably good probes into the otherwise unobservable interiors of stars. It also enhances the use of two or more of the three in sorting out the various processes at work in the insides of stars.

The surface abundances of the light elements lithium (Li) and beryllium (Be) reveal information about the physical processes taking place in stellar interiors. The investigation of the amount of these two elements in stars in open clusters shows the effect of age on those mechanisms. We have obtained spectra of both Li and Be in main-sequence stars in NGC 752 at high spectral resolution and high signal-to-noise ratios with HIRES on the Keck I telescope. In order to make meaningful comparisons with other clusters, we have determined the stellar parameters on a common scale. We have found abundances of Li and Be by spectral synthesis techniques. NGC 752 is twice the age of the well-studied Hyades. We find that (1) the Li dip centered near 6500 K is wider in NGC 752, having expanded toward cooler temperatures; (2) the Be dip is deeper in the older NGC 752; (3) the Li “peak” near 6200 K is lower by about 0.3 dex; (4) although there is little Be depletion in the cooler stars, it is possible that Be may be lower in NGC 752 than in the Hyades; and (5) the Li content in both clusters declines with decreasing temperature, but there is less Li in NGC 752 at a given temperature by ∼0.4 dex. These differences are consistent with the transport of the light-element nuclei below the surface convection zone as predicted by theory. That connection to rotational spin-down is indicated by the pattern of rotation with temperature in the two clusters.

We have obtained new detailed abundances of the Fe-group elements Sc through Zn (Z = 21–30) in three very metal-poor ([Fe/H] ≈ −3) stars: BD+03^o740, BD−13^o3442, and CD−33^o1173. High-resolution ultraviolet Hubble Space Telescope/Space Telescope Imaging Spectrograph spectra in the wavelength range 2300–3050 Å were gathered, and complemented by an assortment of optical echelle spectra. The analysis featured recent laboratory atomic data for a number of neutral and ionized species for all Fe-group elements except Cu and Zn. A detailed examination of scandium, titanium, and vanadium abundances in large-sample spectroscopic surveys indicates that they are positively correlated in stars with [Fe/H] ≤ −2. The abundances of these elements in BD+03^o740, BD−13^o3442, CD−33^o1173, and HD 84937 (studied in a previous paper of this series) are in accord with these trends and lie at the high end of the correlations. Six elements have detectable neutral and ionized features, and generally their abundances are in reasonable agreement. For Cr we find only minimal abundance disagreement between the neutral (mean of [Cr i/Fe] = +0.01) and ionized species (mean of [Cr ii/Fe] = +0.08), unlike most studies in the past. The prominent exception is Co, for which the neutral species indicates a significant overabundance (mean of [Co i/H] = −2.53), while no such enhancement is seen for the ionized species (mean of [Co ii/H] = −2.93). These new stellar abundances, especially the correlations among Sc, Ti, and V, suggest that models of element production in early high-mass metal-poor stars should be revisited.