We propose a common-envelope jets supernova (CEJSN) scenario for the fast-rising blue optical transient AT2018cow. In a CEJSN a neutron star (NS) spirals-in inside the extended envelope of a massive giant star and enters the core. The NS accretes mass from the core through an accretion disc and launches jets. These jets explode the core and the envelope. In the specific polar CEJSN scenario that we propose here the jets clear the polar regions of the giant star before the NS enters the core. The jets that the NS launches after it enters the core expand almost freely along the polar directions that contain a small amount of mass. This, we suggest, explains the fast rise to maximum and the fast ejecta observed at early times of the enigmatic transient AT2018cow. The slower later time ejecta is the more massive equatorial outflow. We roughly estimate the accretion phase onto the NS during the explosion phase to last for a time of ≈ 10 3 s, during which the average mass accretion rate is ≈ 10 −4 M s −1 . We outline the possible diversity of CEJSNe by listing five other scenarios in addition to the polar CEJSN scenario.

We calculate the outward energy transport time by convection and photon diffusion in an inflated common envelope and find this time to be shorter than the envelope expansion time. We conclude therefore that most of the hydrogen recombination energy ends in radiation rather than in kinetic energy of the outflowing envelope. We use the stellar evolution code MESA and inject energy inside the envelope of an asymptotic giant branch star to mimic energy deposition by a spiraling-in stellar companion. During 1.7 years the envelope expands by a factor of more than 2. Along the entire evolution the convection can carry the energy very efficiently outwards, to the radius where radiative transfer becomes more efficient. The total energy transport time stays within several months, shorter than the dynamical time of the envelope. Had we included rapid mass loss, as is expected in the common envelope evolution, the energy transport time would have been even shorter. It seems that calculations that assume that most of the recombination energy ends in the outflowing gas might be inaccurate.

We analyze the morphologies of three core collapse supernova remnants (CCSNRs) and the energy of jets in other CCSNRs and in Super Luminous Supernovae (SLSNe) of type Ib/Ic/IIb, and conclude that these properties are well explained by the last jets' episode as expected in the jet feedback explosion mechanism of core collapse supernovae (CCSNe). The presence of two opposite protrusions, termed ears, and our comparison of the CCSNR morphologies with morphologies of planetary nebulae strengthen the claim that jets play a major role in the explosion mechanism of CCSNe. We crudely estimate the energy that was required to inflate the ears in two CCSNRs, and assume that the ears were inflated by jets. We find that the energies of the jets that inflated ears in 11 CCSNRs span a range that is similar to that of jets in some energetic CCSNe (SLSNe), and that this energy, only of the last jets' episode, is much less than the explosion energy. This finding is compatible with the jet feedback explosion mechanism of CCSNe, where only the last jets, that carry a small fraction of the total energy carried by earlier jets, are expected to influence the outer parts of the ejecta. We reiterate our call for a paradigm shift from neutrino-driven to jet-driven explosion models of CCSNe.

We study r-process feasibility inside jets launched by a cold neutron star (NS) spiralling-in inside the core of a giant star, and find that such common envelope jets supernova events might be a significant source of heavy r-process elements in the early Universe. We run the stellar evolution code MESA to follow the evolution of low metalicity giant stars that swallow NSs during their late expansion phases and find that in some of the cases the NSs penetrate the core. The Bondi-Hoyle-Lyttleton (BHL) mass accretion rate onto a NS as it spirals-in inside the core is sufficiently high to obtain a neutron rich ejecta as required for the heavy r-process where the second and third r-process elements are synthesized. Due to the small radius of the NS the accretion is through an accretion disk and the outflow is in jets (or bipolar disk winds). The r-process nucleosynthesis takes place inside the jets. To account for the r-process abundances in the Galaxy we require that one in ten cases of a NS entering the envelope of a giant star ends as a CEJSN r-process event.

By using the stellar evolution code MESA we show that most of the hydrogen recombination energy that is released as the envelope expands during a regular common envelope evolution (CEE), namely, the initial dynamical phase or plunge-in phase, is radiated, and hence increases substantially the stellar luminosity. Only about ten per cent of the hydrogen recombination energy might be used to remove the envelope. We show that the key property of energy transport is that when convection becomes inefficient in the outer parts of the envelope, where the ionization degree of hydrogen falls below about 30 per cent, photon diffusion becomes very efficient and removes the recombination energy. The expanding envelope absorbs most of the gravitational energy that is released by the spiraling-in process of the secondary star inside the common envelope, and so it is the hydrogen recombination energy that is behind most of the luminosity increase of the system. The recombination energy of hydrogen adds only a small fraction of the energy required to remove the common envelope, and hence does not play a significant role in the ejection of the envelope.