A reliable analysis of photospheric parameters (, ) requires precise element abundance determinations, even of putative trace elements, in order to model the temperature structure correctly.
In the case of LB 3459 the Balmer-line problem appears in the analysis based on pure H+He spectra. With the consideration of carbon, nitrogen, oxygen, magnesium, silicon, iron, and nickel in addition, the agreement of the theoretical Balmer lines with the observation has been improved. However, it is not perfect and there might be additional absorbers present in the photosphere which have not been considered in our analysis.
Since the ionization balance is a very sensitive indicator for , the simultaneous fit of the He I1 and He II (Fig. 7), C III and C IV (Fig. 8), N III - N V (Fig. 9), and O IV and O V lines (Fig. 10) to the UV spectrum of LB 3459 verifies our determination (Sect. 3.2) within a small error range.
The UV spectrum of LB 3459 is full of lines of iron-group elements. Their consideration improves the fit of the synthetic spectrum to the observation. However, several features remain unidentified and may stem from "light metals" Li - Ca which are not considered in this analysis. However, from the consideration of the iron-group elements we can judge that this has only a neglectable influence on our analysis of the photospheric parameters.
The derived low helium and metal abundances (Fig. 19) indicate that diffusion is likely to be efficient in the primary of LB 3459 and responsible for the depletion. Due to the increasing number of lines with the atomic number, the radiation force and the relatively low surface gravity of LB 3459 allows the heavier elements to stay at the surface - iron and nickel appear even to be solar. Magnesium is enriched by a factor of six - no simple explanation can be given for this phenomenon. High-resolution and high-S/N UV spectra are required in order to determine the photospheric metal abundances of the primary of LB 3459 more precisely.
The spectroscopically determined rotational velocity ( of the primary of LB 3459 is marginally less than which can be calculated from its radius and orbital period (Sect. 1). Thus it cannot be excluded that the rotation of the primary is bound. However, new optical spectra with high S/N ratio and short exposure times are highly desirable in order to significantly improve the determination of and to minimize the influence of orbital motion on the line broadening.
Based on the assumption that the primary, subdwarf component in LB 3459 has a mass of , HHH found the cool secondary star entirely consistent with ZAMS models. However, the newly determined, low masses of the primary () and secondary () component of LB 3459 suggest that this is a "low mass case B" system (Iben & Livio 1993). Thus, it appears possible that the secondary has formerly been a planet () which has survived the CE phase () and even has gained mass.
The uncertainties in the evolutionary scenario of LB 3459 may be a reason for the disagreement with the results from light curve and velocity curve analyses. Although the character of the primary appears clear, almost nothing is known about the secondary. Moreover, the experienced CE phase questions the validity of the evolutionary models of Driebe et al. (1998) for LB 3459 - although these are the most elaborated for similar, non-CE stars.
© European Southern Observatory (ESO) 2000
Online publication: April 10, 2000