SIMBAD references

Stars of a wide range of metallicity have been observed and analyzed in the spectral region of the Be II resonant doublet (∼3130A), from the most metal-poor ones, which give us information on the primary mechanism(s) of Be production during the early phases of galactic evolution, to solar-type stars, which become fundamental tools to investigate the structure of the outermost layers of their atmospheres and to constrain which mixing and depletion mechanisms are more effective.

We tested the different theoretical scenarios proposed so far to explain the formation of beryllium, namely if the classical cosmic-ray theory of spallation of interstellar CNO atoms by colliding protons and alpha-particles, proposed 25 years ago by Reeves et al. (1970, Nature, 226, 727), remains a valid explanation, or if the most recent suggestions of a reversed mechanism, involving spallation between H and He interstellar atoms hit by accelerated CNO nuclei in the vicinity of supernovae (Duncan et al. 1992, ApJ, 401, 584; Casse et al. 1995, Nature, 373, 318), may better account for what we observe today. A reduced chi^2 analysis has shown a correlation of [Be/H] with [Fe/H] characterized by a slope of 0.82, that suggests a behaviour more akin to that of a primary element rather than a secondary one. This strongly favors the most recent view of creation of Be (and B) through spallation by CNO nuclei accelerated during a supernova explosion which gives the best reproduction of our set of abundances, although it needs to be tested with larger samples of data for both beryllium and boron abudances.

Our findings of a linear trend between [Be/H] and [Fe/H], of slope 0.82 ± 0.24 (1-sigma), and of a real dispersion among different objects at almost the same metallicity can be considered the most important result of this analysis. The observed scatter might suggest the onset of some different mechanism involved in the galactic production of beryllium, additional destruction mechanisms, or, in the most recent scenario of Be formation, it might also reflect the presence of inhomogeneities in the interstellar medium surrounding active star forming regions.

There is no suggestion of a break at the metallicity of [M/H]= -1.00 that usually marks the transition between halo and disk stars, although a larger homogeneous sample of observations could test this better. No other change of slope seems necessary to fit satisfactorily our observed trend in the logarithmic plane [Be/H] vs. [Fe/H], as previously suggested by other investigations. There is no plateau at the lowest metallicities which might indicate a primordial contribution to Be production.

Two of the most metal-poor stars we analyzed are subgiants, with effective temperatures Three stars have no detectable Be absorption, which seems to indicate that their Be content has been depleted, thus becoming useful data points in testing predictions of depth of mixing in stellar structure models. They increase the small number of objects already found with strong Be depletion, and along with the few halo stars which are Li-poor, suggest that their anomalies should be further tested by analysing their B abundances.

Analysis of HST (Hubble Space Telescope) spectra of both components A and B of the Alpha Centauri system yields a solar Be abundance in alfa Cen A and [Be/H]= -0.50 dex in alfa Cen B. The lack of Be destruction in component A and moderate destruction in B has allowed a rather accurate determination of the depth of mixing, to a temperature intermediate between 2.5 and 3.5 X 10^6 K in the case of alfa Cen A, and to very close to 3 X 10^6 K in the case of alfa Cen B (Primas et al. 1996, in preparation).