2. Sample selection and observations
Sample stars were selected according to two criteria. First we chose objects with a membership probability higher than 90 in the proper motion studies of Sanders (1977) and Girard et al. (1989). Second, we made an effort not to include known binaries in our sample: therefore, in all but one case (see Sect. 4.2) we chose stars that were not catalogued as binaries in the literature. We anticipate, however, that some of our objects turned out to be either binaries or radial velocity variables on the basis of later surveys. We will flag those stars in the following discussion.
The observations were carried out at ESO, La Silla, in February 1995 using the CASPEC Spectrograph at the 3.6m telescope (Randich and Pasquini 1996). CASPEC was used in combination with the red crossdisperser, the long camera and the 31.6 lines/mm echelle grating. Coupling the long camera with the Tektronix CCD ESO #37, which has 24 µm square pixels, a spectral coverage of 2400 Å was covered, with a dispersion of 0.080 Å/pixel in the lithium region.
Due to the cluster declination, the combined telescope plus instrument image quality was not very good, and a slit width of 2 arcsec had to be used, giving a resolving power of . This resolution corresponds to a FWHM for unblended calibration lines of 4 pixels. Stars with V=14 were observed for 1.5 hours, giving a typical signal to noise (S/N) ratio of about 50 per pixel.
In the reduction phase, one of the largest uncertainties is given by the subtraction of the interorder light. For this reason, the spectra were reduced (and normalized) in two independent ways, with two different background subtractions, finding that the differences in the measured Li equivalent widths were at most of 7 mÅ, in good agreement with what expected from the resolution and the S/N ratio.
In Table 1 the basic parameters of the observed stars are summarized, together with the measured Li equivalent widths. Note that at our resolution, the Li feature at 6707.81 Å cannot be separated from a nearby 6707.44 Fe line, whose contribution is thus included in the equivalent widths given in Table 1. This contribution will be subtracted before converting equivalent widths to Li abundances (see Sect. 3). Note that for star S1045 (the only double line spectroscopic binary in the sample) the measured equivalent widths of each component include, in addition to the 6707.44 blend, any contribution coming from the other component of the binary system. The abundances for this binary were derived using spectral synthesis (cfr. Sect. 4.2).
Table 1. Observed Stars. Column 1-8 give: 1) Sanders (1977) number, 2) apparent V magnitude, 3) B-V colour, 4) V-I colour, 5) measured equivalent width (in mÅ) for the Li+Fe blend, 6) effective temperature, 7) Li abundance, 8) binary flag.
The photometry from Montgomery et al. (1993) available for all but one object (S958) has been consistently adopted. In Table 1 stars are numbered according to Sanders (1977). Although, as mentioned in the previous section, we tried to avoid known binaries, some may still be present in the sample. After the observations were taken, Dr. Mathieu kindly provided us with his latest results from a radial velocity survey of the cluster (Mathieu and Latham 1996, in preparation); it turned out that some of our target stars are either binaries or show radial velocity variations. These objects have been flagged in the last column of Table 1.
© European Southern Observatory (ESO) 1997
Online publication: April 28, 1998