THE ORIGIN OF VARIABILITY OF THE INTERMEDIATE-MASS BLACK-HOLE ULX SYSTEM HLX-1 IN ESO 243-49

The ultra-luminous (L-X less than or similar to 10(42) erg s(-1)) intermediate-mass black-hole (IMBH) system HLX-1 in the ESO 243-49 galaxy exhibits variability with a possible recurrence time of a few hundred days. Finding the origin of this variability would constrain the still largely unknown properties of this extraordinary object. Since it exhibits a hardness-intensity behavior characteristic of black-hole X-ray transients, we have analyzed the variability of HLX-1 in the framework of the disk instability model that explains outbursts of such systems. We find that the long-term variability of HLX-1 is unlikely to be explained by a model in which outbursts are triggered by thermal-viscous instabilities in an accretion disk. Possible alternatives include the instability in a radiation-pressure-dominated disk but we argue that a more likely explanation is a modulated mass transfer due to tidal stripping of a star in an eccentric orbit around the IMBH. We consider an evolutionary scenario leading to the creation of such a system and estimate the probability of its observation. We conclude, using a simplified dynamical model of the post-collapse cluster, that no more than 1/100 to 1/10 of M-lozenge less than or similar to 10(4) M-circle dot IMBHs-formed by runaway stellar mergers in the dense collapsed cores of young clusters-could have a few x 1 M-circle dot main-sequence star evolve to an asymptotic giant branch on an orbit eccentric enough for mass transfer at periapse, while avoiding collisional destruction or being scattered into the IMBH by two-body encounters. The finite but low probability of this configuration is consistent with the uniqueness of HLX-1. We note, however, that the actual response of a standard accretion disk to bursts of mass transfer may be too slow to explain the observations unless the orbit is close to parabolic (and hence even rarer). Also, increased heating, presumably linked to the highly time-dependent gravitational potential, could shorten the relevant timescales.