3 Discussion

We have derived some model-independent limits on the size and mass of the material responsible for iron line features, assuming that they can be detected during the X-ray afterglow. This work has been stimulated by the recent claim of detection of such lines, but the validity of our conclusions are quite general. The most important implication of these lines, if confirmed, regard the progenitor of the burst. In fact it is inescapable to assume that about a solar mass of (probably iron rich) material is located in the close vicinity of the event. Among the proposed models, the Supranova of Vietri & Stella (1998) is the only one which naturally accounts for this. This in turn implies that at least a class of $\gamma$-ray burst have to be associated with supernovae exploded about a month earlier (assuming that the remnants have a velocity of about 10⁴ km s^-1).

The possible association of GRB with supernovae has been investigated recently in detail by Bloom et al. (1998), Kippen et al. (1998) and Wang & Wheeler (1998), following the explosion of GRB 980425, likely associated with the type Ic SN 1998bw. Among these works, only Wang & Wheeler (1998) find evidence for a connection while the other two limit to a few percent the bursts possibly associated with supernovae. In the Supranova scenario, however, the association of supernovae with bursts suffer for a time delay which would smear the time correlation between the two phenomena.

Should the iron lines possibly detected in GRB 970508 and GRB 970828 be real and confirmed by other cases, then we have a strong case for the connection between supernovae and $\gamma$-ray bursts. The next generation of experiments and satellites, such as XMM, AXAF and ASTRO-E, will provide us with the necessary information to draw more accurate conclusion on the puzzling problem of the $\gamma$-ray burst progenitor.