|Annu. Rev. Astron. Astrophys. 1999. 37:
Copyright © 1999 by . All rights reserved
6.4. Binaries and Dynamical Effects
There has been much speculation about the possible origin of hot low-mass stars through dynamical interactions, especially in binary systems through Roche-lobe mass transfer or mergers. The various recognized mechanisms have been reviewed by Bailyn (1995), while their implications for the UVX problem were most extensively discussed by GR. The mildest form of interaction occurs when a star ascending the giant branch loses part of its expanding envelope to a companion, thereby appearing with lower MENV on the ZAHB (Mengel et al 1976) but evolving normally thereafter. Based on the frequency of binaries among the sdB stars in the Galactic field (Green 1999), this process may be fairly common there. Although "fine tuning" of binary mass ratios and separations would seem to be necessary to produce small MENV without suppressing the He flash altogether, in fact the hot-flash mechanism (D'Cruz et al 1996) would mitigate this problem here as it does for normal mass loss. One reason that UV star production in binaries might depend on Z is that stellar envelopes become more inflated at higher metallicities (GR). The fact that field and cluster hot horizontal branches have gravities and luminosities consistent with single-star models suggests that more drastic interactions (e.g. mergers) are considerably more rare.
Some support for dynamical effects is provided by the observation that the extent of horizontal branch "blue tails" in Galactic globular clusters appears to correlate with cluster concentration and density (e.g. Fusi Pecci et al 1993, Buonanno et al 1997). However, other expectations for dynamical mechanisms are not met. For instance, the hot stars are not necessarily concentrated to the centers of clusters, and the system with the largest EHB/post-EHB population ( Cen) is notably low density (e.g. Whitney et al 1994, Rich et al 1997). HST observations of 10 cluster cores (Sosin et al 1997) show that the EHB stars are not as centrally concentrated as the blue stragglers (which are almost universally agreed to be interaction products).
It is unclear how to translate the evidence for dynamical mechanisms in star clusters to the dynamical environment of galaxies, which is very different and less conducive to stellar interactions. Of course, the present field population in E galaxies may well have originated in concentrated cluster-like systems that have since disintegrated but which were responsible for establishing the binary frequency. Stellar rotation, which could depend on global dynamical characteristics of galaxies, may also influence the UVX through He mixing (Sweigart 1997). The only hint that the E galaxy UVX is related to dynamics is the connection between large UV upturns and boxy isophotes (Longo et al 1989 and Section 5.2). The core of M32 is the densest observable extragalactic system. No large radial gradients in blue light are apparent near its center at HST resolution (King et al 1992, 1995;, Bertola et al 1995;, Cole et al 1998;, Lauer et al 1998), though Brown et al (1998a) note that the density of resolved UV stars per unit total light increases slightly at smaller radii in both M31 and M32.
Mass transfer onto white dwarf binary companions can produce copious UV flux and has also been discussed in the context of the UVX by GR. One of the main problems is again the fine-tuning needed to ensure a sufficient but not excessive transfer. Because of the large parameter space involved, estimates are only tentative, but accreting white dwarfs seem unlikely to be major contributors to the UVX (GR). The UV spectrum of a population dominated by such objects is also expected to show fairly strong emission lines (e.g. Wu et al 1992), which are absent in E galaxy spectra.
Although dynamical interactions could certainly influence the UVX in galaxies, there is no strong evidence yet that they do so.