2013A&A...559A..94W

It has recently been proposed that one subclass of Type Ia supernovae (SNe Ia) is sufficiently distinct and sufficiently common to be classified separately from the bulk of SNe Ia; the suggested class name is Type Iax supernovae (SNe Iax), after SN 2002cx. However, their progenitors are still uncertain. We study whether the population properties of this class might be understood if the events originate from a subset of sub-Chandrasekhar mass explosions. In this potential progenitor population, a carbon-oxygen white dwarf (CO WD) accumulates a helium layer from a non-degenerate helium star; ignition of that helium layer then leads to ignition of the CO WD. We incorporated detailed binary evolution calculations for the progenitor systems into a binary population synthesis model to obtain rates and delay times for these events. The predicted Galactic event rate of these explosions is ∼1.5x10^–3yr^–1 according to our standard model, in good agreement with the measured rates of SNe Iax. In addition, predicted delay times are ∼70Myr-800Myr, consistent with the fact that most of SNe Iax have been discovered in late-type galaxies. If the explosions are assumed to be double detonations - following current model expectations - then, based on the CO WD masses at explosion, we also estimate the distribution of resulting SN brightness (-13>M_bol>-19mag), which can reproduce the empirical diversity of SNe Iax. We speculate on why binaries with non-degenerate donor stars might lead to SNe Iax if similar systems with degenerate donors do not. We suggest that the high mass of the helium layer necessary for ignition at the lower accretion rates typically delivered from non-degenerate donors might be necessary to produce SN 2002cx-like characteristics, perhaps even by changing the nature of the CO ignition.