On the composition of GRBs' Collapsar jets

The duration distribution of long gamma-ray bursts (GRBs) reveals a plateau at durations shorter than similar to 20 s (in the observer frame) and a power-law decline at longer durations. Such a plateau arises naturally in the Collapsar model. In this model, the engine has to operate long enough to push the jet out of the stellar envelope and the observed duration of the burst is the difference between the engine's operation time and the jet breakout time. The jet breakout time inferred from the duration distribution (similar to 10 s in the burst's frame) is comparable to the breakout time of both analytic estimates and numerical simulations (both 2D and 3D) of a hydrodynamic jet (similar to 10 s for typical parameters). Recently, we have estimated analytically the breakout time of a Poynting-flux-dominated jet and have shown that it is consistent with 2D numerical simulations. We find that such a jet with the same overall energy breaks out much faster (less than or similar to 1 s). If this result holds for 3D simulations it implies that only hydrodynamic jets are compatible with the duration of the plateau in the GRB duration distribution and hence the jet should be hydrodynamic during most of the time that its head is within the envelope of the progenitor star and around the time when it emerges from the star. This would naturally arise if the jet forms as a hydrodynamic jet in the first place or if it forms Poynting flux dominated but dissipates most of its magnetic energy early on within the progenitor star and emerges as a hydrodynamic jet.