Astron. Astrophys. 319, 637-647 (1997)

4. The lithium doublet

The Li I doublet is formed by two Li⁷ components at 6707.761 Å  and 6707.912 Å . The very fragile Li⁶ isotope can also be observed in some stars. Its components lie at 6707.921 Å  and 6708.072 Å . If the hyperfine structure is also considered the number of the Li⁷ + Li⁶ components may amount up to 19 lines (Kurucz, 1995).

The terrestrial "representative isotopic ratio" is tabulated to be equal to 0.081 by Anders & Grevesse (1989), while the meteoritic value representative of the early solar system is 0.079 (Anders & Ebihara, 1982). We assumed the isotopic ratio from Anders & Grevesse (1989) as starting point for our computations. However, in the case of peculiar stars, especially for cold CP stars, the strong surface magnetic fields can make the Li problem very complicate, because they hamper the mixing of surface matter with the internal hotter one.

The lower plot of Fig. 4 shows that in the case of  CrB the feature at 6708 Å  can not be reproduced by computations with a Li abundance (i.e. 2.6 dex larger than the solar one) and an eventual Li⁶ /Li⁷ ratio equal to 0.081. In fact, the observed feature is redshifted by 0.2 Å with respect to the computed one. When the isotopic ratio is increased, the wavelength difference between the observed and computed absorptions slightly decreases, but it is never eliminated even if a highly improbable isotopic ratio equal 10 is assumed. All the 19 isotopic and hyperfine components listed in Kurucz (1995) were considered in the computations.

Fig. 4. Upper plot: The observed Li feature (thick line) compared with that computed for an isotopic ratio 0.081, , and . Lower plot: The observed Li feature (thick line) and the computed ones for and different isotopic ratios. From left to right: Li6 /Li7  = 0.081,  0.7,  2,  10.

Another explanation of the observed redshifted feature with respect to the Li I predicted wavelength is the blending of Li I with some other element. From the analysis of the known atomic lines in the 6707.0 - 6708.6 Å  region listed in Table 7 of Gerbaldi et al. (1995) and used by us for computing the spectrum of  CrB, we could infer that V I at 6708.094 Å  is the only possible component to the 6708 Å  feature observed in  CrB. Because it is not predicted in the Sun (Gerbaldi et al., 1995; Kurucz, 1995) we modified the from Kurucz (1993b) until we obtained a predicted feature consistent with the solar observations. In this way we fixed for V I 6708.094 Å . Then we modified the Li and V abundance of  CrB in order to fit the feature at 6708 Å  with the computations. The upper plot of Fig. 4 shows the comparison of the observed spectrum with a spectrum computed with and (i.e. 3.2 dex larger than the solar value). The agreement is good and furthermore the large vanadium abundance does not yield in the 6693 - 6721 Å  region strong vanadium lines predicted but not observed. Anyway, other wavelength ranges should be studied in order to check this high V abundance. In fact, previous works (see Table 4) have given V in excess by a factor lower than 10.

Table 5. List of all the lines contributing to the spectrum of  CrB in the lithium region

The last, most acceptable hypothesis, is that lithium is blended with some unknown element, possibly belonging to the heavy elements or to rare earth elements, which were found overabundant by factors ranging from 10 to 1000 (see for instance the abundances of Sr, Zr, La, Sm, Gd, Ce listed in Table 4). Actually some unidentified features are present in the studied region. The strongest ones are at 6697.3, 6700.4, 6705.0, 6707.0, 6709, 6710.3, 6713.6, 6716.3 Å  (Fig. 3). For instance, from a comparison of giants with different gravities, Lambert et al. (1993) concluded that in stars enriched in heavy elements, a Ce II line could blend the Li line. Since, in  CrB, Ce II is overabundant by 2 3 dex, a similar blending can be expected.

Finally, we have not investigated the effect of the magnetic field on the Li line. In fact, accurate determinations of the abundances in magnetic stars require to take into account the broadening by the magnetic field.

We have searched for the presence of the other Li I line at 6103.64 Å  in spectra of  CrB observed by us in this region. This Li I line is not observable, as expected, because it is fainter than the resonance doublet and is blended with a strong Fe II line at 6103.54 Å .

© European Southern Observatory (ESO) 1997

Online publication: July 3, 1998
helpdesk.link@springer.de