Formation and evolution of compact binaries

Abstract

In this thesis we investigate the formation and evolution of compact binaries. Chapters 2 through 4 deal with the formation of luminous, ultra-compact X-ray binaries in globular clusters. We show that the proposed scenario of magnetic capture produces too few ultra-compact X-ray binaries to explain the observations if we assume strong magnetic braking. Magnetic capture produces no ultra-compact systems at all if we assume a more realistic, weaker magnetic-braking law or a more modern law that takes into account saturation at short rotation periods. This means that the observed systems in globular clusters are probably formed by another mechanism such as the collision of a (sub)giant star with a neutron star, and that the negative period derivative of the binary with an orbital period of 11 minutes should be apparent and probably caused by acceleration in the gravitational potential of the cluster. In chapter 5 we investigate the X-ray binary 2S 0918-549 in the Galaxy that is believed to be ultra-compact. An observation of a long X-ray burst is presented, which can only be explained if hydrogen is present on the neutron star. This is clearly in contradiction with the presumed ultra-compact nature of the binary. We suggest that slow accretion of helium can cause the neutron star to accumulate a thick helium layer and that this can explain the length of the burst. The helium nature of the donor is further supported by the observation that the neon-to-oxygen ratio in the system is 2.4 times the solar value. We show that this high ratio can be naturally explained if the donor star is a helium white dwarf, whereas a carbon-oxygen white dwarf would yield a neon-to-oxygen ratio that is lower than solar. We conclude that the donor is probably (the inner part of) a helium white dwarf that was produced by a star of 2.25Mo or less. Chapter 6 discusses the formation of double white dwarfs. For ten of these systems both components have been observed so that both masses are known. The low masses of the components and the short orbital periods of these ten binaries suggest that the white dwarfs were formed after two mass-transfer episodes and that the orbit of the binary must have shrunk appreciably during the last mass transfer. The strong orbital shrinkage can be explained by a so-called spiral-in based on energy balance. During the spiral-in of the orbit, the envelope of the donor is ejected from the system. We show that stable, conservative mass transfer combined with a canonical spiral-in cannot explain all observed systems. Instead we need two phases of a spiral-in, of which the first needs balance of angular momentum rather than energy. The second spiral-in may have either energy or angular-momentum balance. We can explain the observed masses and orbital periods well with angular-momentum conservation, although it turns out to be more difficult to explain the observed age differences between the components in addition.