Abundance analysis of normal and mercury-manganase type late-B stars from optical spectra.

摘要

The aim of this project was to perform an abundance analysis of a selection of mercury-manganese and normal A and B-type stars. Observations of thirteen stars were taken using the Hamilton Echelle Spectrograph and coude auxilliary telescope (CAT) situated at Lick Observatory. The high signal-to-noise optical spectra were reduced with the aid of the software packages FIGARO and VISTA. A technique for measuring the parasitic light parallel to the dispersion of the spectrum was devised; none was found (mean P = -1.0+/-0.8%), which implies that measured equivalent widths should be free of this important systematic error. Measurements were made of the wavelengths and equivalent widths of unblended spectral lines. The spectra were modelled using ATLAS-6 model atmospheres and the synthesis package UCLSYN. Effective temperatures and surface gravities for the programme stars were adopted from Smith and Dworetsky (1993). Abundances were generally determined by comparing synthetic equivalent widths to measured equivalent widths. Where line blends made this impossible, the observed spectra were compared to the synthetic spectra. The instrumental profile was determined from arc calibration lines and weis found to be asymmetric; the profile was measured and was convolved with the synthetic spectra to allow a comparison with observations to be made. Microturbulence parameters for these stars were determined according to the method of Magain (1984); stars with effective temperatures below 11000K showed a positive value, but stars above this temperature yielded microturbulences consistent with zero. The abundances generally compare well with the literature and are in many cases an improvement on existing work. Phosphorus showed a positive correlation with effective temperature in HgMn stars and manganese also exhibits a strong correlation, in agreement with previous observations. The strong manganese line at lambda4206.367 was found to exhibit a systematic discrepancy with respect to other manganese lines at high abundances; this is probably due to hyperfine structure and it is recommended that this line not be used for abundance analysis in the future unless its structure can be modelled explicitly.