Astron. Astrophys. 363, 958-969 (2000)

1. Introduction

ROSAT observations of several open stellar clusters have been carried out in the last few years with the primary purpose of studying the X-ray properties of their solar-type and low-mass members and investigating in detail their evolution with age and the age-rotation-activity connection (e.g., Randich 1997, 2000 and references therein; Jeffries 1999 and references therein). As a very important by-product, these X-ray surveys have allowed the identification of optically unknown low-mass candidate members of numerous clusters (e.g., Randich et al. 1995; Prosser et al. 1996; Randich et al 1996a; Prosser & Randich 1998). As is well known, the level of X-ray activity decays, on average, with increasing stellar age and therefore low-mass stars in young clusters are expected to be vigorous X-ray emitters and they should be detected in sensitive enough X-ray surveys (Randich 1997; Jeffries 1999). As it is the case for X-ray selected T Tauri stars, optical photometric and spectroscopic follow-up observations are needed in order to investigate cluster membership and to infer at least the basic stellar properties (e.g., colours and spectral types) of the X-ray selected candidate members.

The open cluster in Coma Berenices (Cl Melotte 111) is known for its apparent deficit of low-mass stars. It has an estimated age Myr, it is located at a distance of 86 pc (Van Leeuwen 1999), and its brightest stars are of A0 spectral type. Starting with the first proper motion/photometric/radial velocity survey of Coma (Trumpler 1938; km s^-1), different investigations have addressed the problem of the stellar population in this cluster (e.g., Artyukhina 1955; Argue & Kenworthy 1969; De Luca & Weis 1981; Bounatiro 1993); all the studies agreed in pointing out that the cluster has relatively few members (no more than 50-60) and, particularly, that the number of objects with 10.5 (or , corresponding to K0V spectral type) in the cluster is extremely small. De Luca & Weiss (1981) carried out BVRI photometry for 88 faint/red stars in the cluster region and concluded that the number of stars with 10.5 is probably not larger than 10. In other words, Coma Berenices appears to be generally poor in stars and, particularly, it is poor in low-mass stars. According to both Trumpler (1938) and Argue & Kenworthy (1969), this result can be interpreted by assuming that the mass of the cluster and the average mass density are lower than what is required for stability, and therefore less massive cluster members could have escaped the cluster. Very recently Odenkirchen et al. (1998) analyzed astrometric and photometric data from the Hipparcos, Tycho, and ACT catalogs covering 1200 deg² around the center of Coma Berenices, and measured a radial velocity of km s^-1; their primary aim was evidencing the process of escape and cluster dissolution by identifying cluster members in the surroundings of the cluster. They found that the cluster consists of a core and a halo extending up to 5^o from the cluster center and of a moving group of extratidal stars, which witness the cluster dissolution. The luminosity function of stars in the core-halo region shows a steep decline beyond , while that of stars in the moving group continues to rise. They also found that faint stars are preferentially located at distances between 3^o and 4^o from the cluster center (but note that two of the faint stars from Artyukhina 1955 and Argue & Kenworthy 1969 are instead located very close to the cluster center).

Most of the optical searches for Coma Berenices members used proper motions as the prime criterion for selecting possible cluster candidates. However, there is one main difficulty in using proper motions for picking out possible Coma members: the cluster proper motion itself is very small (see section on proper motions). Odenkirchen et al. (1998) selected kinematic members on the basis of tangential space velocities, but as they point out, the sample suffers from incompleteness beyond . The question then arises whether the cluster is really poor in low-mass stars, or whether additional low-mass cluster members do exist, but have not been identified so far.

Randich et al. (1996b - hereafter RSP) carried out a ROSAT X-ray survey of Coma and found 12 X-ray sources that could not be identified with known Coma members, but that had an optical counterpart whose properties (namely, apparent visual magnitude, X-ray over optical flux, hardness ratio) were consistent with cluster membership. The sensitivity of the X-ray observations of Coma ranged between  erg s^-1 and  erg s^-1, and the median X-ray luminosities of the K- and M-type dwarfs in the well studied Hyades open cluster are and 28.21 erg s^-1, respectively (Pye et al. 1994). Under the assumption that the X-ray luminosity distribution function (XLDF) of Coma low-mass dwarfs does not significantly differ from the one of the Hyades, if a substantial number of K/M Coma members exists, the RSP X-ray survey should reveal at least part of them. If these 12 objects, or a fraction of them, turned out to be actual cluster members, there would be an indication that Coma Berenices' main-sequence does not virtually truncate around the K spectral-type and there would be an additional motivation to search for still unidentified faint, late-type Coma members. We carried out VRI photometry and low resolution spectroscopy in the visible for the X-ray cluster candidates in order to ascertain or reject membership for these objects and to infer their properties. The results of our optical follow-up are presented here.

© European Southern Observatory (ESO) 2000

Online publication: December 5, 2000
helpdesk.link@springer.de