Wolf-Rayet stars and cosmic gamma-ray bursts

Abstract

The observational properties of cosmic gamma-ray bursts and of Wolf-Rayet (WR) stars and their CO cores at the end of their evolution are analyzed. WR stars do not have hydrogen envelopes, facilitating the transformation of the energy of collapse into observable gamma rays. Of the ≈90 well-localized gamma-ray bursts, 21 have optical identifications, of which 16 have measured redshifts (z=0.4–4.5). The distribution of gamma-ray bursts in energy N(ΔE) has a large scatter, from 3×1051 to 2×1054 erg. There is some evidence that the distribution N(ΔE) is bimodal if we include the gamma-ray burst GRB 980425, which is associated with the peculiar type Ic supernova SN 1998bw in the nearby elliptical galaxy ESO 184-G82, for which ΔE γ≈1048erg. These characteristics of gamma-ray bursts are reminiscent of the distribution of final masses for the CO cores of WR stars, which uniformly covers a broad range: M CO=(1–2)M ⊙−(20–44)M ⊙. The possible bimodality of the gamma-ray burst energy distribution (E 1=1048 erg; ΔE 2=3×1051−2×1054erg) could be associated with the bimodal mass distribution for stellar relativistic objects (M NS=(1.35±0.15)M ⊙; M BH=4–15M ⊙). The fact that SN 1998bw is a “peculiar” type Ic supernova, not typical for the collapses of WR stars (which usually give rise to type Ib/c supernovae), could be related to the rotation of the collapsing CO core. This “drags out/rd the time for the collapse, leading to the formation of a neutron star, a decrease in the gamma-ray burst energy, and an increase in the fraction of kinetic energy transferred to the supernova envelope. The expected rate of collapse of the CO cores of WR stars in the Galaxy is ≈10−3/yr. This is at least three orders of magnitude higher than the mean frequency of gamma-ray bursts per galaxy (≈10−6–10−7/yr). Two models for gamma-ray bursts with WR stars as progenitors are considered: the hypernova model of Paczynski (1998) and the pulsation instability CO-core collapse model proposed by Gershte $$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{l} $$ n (2000). In both models, the rate of CO-core collapses can be brought into agreement with the observed rate of gamma-ray bursts by taking into account the anisotropy of the gamma radiation, associated with either a relativistic jet or the random character of the initial CO-core collapse due to instabilities. It is concluded that WR stars could be the progenitors of gamma-ray bursts. This hypothesis predicts the existence of two types of gamma-ray bursts, corresponding to the bimodal mass distribution for stellar relativistic objects, and of three types of optical afterglow, associated with collapses of the CO cores of WR stars that are single, in WR+O binaries, and in hypothetical WR+(A-M) systems. The paper also briefly examines a model of gamma-ray bursts as transient phenomena in the early stages of the evolution of galaxies (z>1), when very massive stars (M>100M ⊙) weak in heavy elements could form. Such massive stars should also lose their hydrogen envelopes and be transformed into massive WR stars, whose collapses could be accompanied by gamma-ray bursts. It is suggested that WR galaxies are the most probable candidates for the host galaxies of gamma-ray bursts.