V. Makaganiuk, O. Kochukhov, N. Piskunov, S. V. Jeffers, C. M. Johns-Krull, C. U. Keller, M. Rodenhuis, F. Snik, H. C. Stempels, J. A. Valenti
Mercury-manganese (HgMn) stars have been considered as non-magnetic and
non-variable chemically peculiar (CP) stars for a long time. However, recent
discoveries of the variability in spectral line profiles have suggested an
inhomogeneous surface distribution of chemical elements in some HgMn stars.
From the studies of other CP stars it is known that magnetic field plays a key
role in the formation of surface spots. All attempts to find magnetic fields in
HgMn stars have yielded negative results. In this study, we investigate the
possible presence of a magnetic field in phi Phe (HD 11753) and reconstruct
surface distribution of chemical elements that show variability in spectral
lines.We also test a hypothesis that a magnetic field is concentrated in
chemical spots and look into the possibility that some chemical elements are
stratified with depth in the stellar atmosphere. Our analysis is based on
high-quality spectropolarimetric time-series observations, covering a full
rotational period of the star. Spectra were obtained with the HARPSpol at the
ESO 3.6-m telescope. Combining information from all suitable spectral lines, we
set an upper limit of 4 G on the mean longitudinal magnetic field. For chemical
spots, an upper limit on the longitudinal field varies between 8 and 15 G. We
confirmed the variability of Y, Sr, and Ti and detected variability in Cr
lines. Stratification analysis showed that Y and Ti are not concentrated in the
uppermost atmospheric layers. Our spectropolarimetric observations rule out the
presence of a strong, globally-organised magnetic field in phi Phe. This
implies an alternative mechanism of spot formation, which could be related to a
non-equilibrium atomic diffusion. However, the typical time scales of the
variation in stratification predicted by the recent time-dependent diffusion
models exceed significantly the spot evolution time-scale reported for phi Phe.
View original:
http://arxiv.org/abs/1111.6065
No comments:
Post a Comment