Massive black holes regulated by luminous blue variable mass loss and magnetic fields - implications for the progenitor of LB-1

Jose H. Groh, Eoin Farrell, Georges Meynet, Nathan Smith, Laura Murphy, Andrew Allan

Published 2019-12-02Version 1

We investigate the effects of mass loss during the main-sequence (MS) and post-MS phases on the final black hole (BH) masses of massive stars. We compute solar metallicity Geneva stellar evolution models of an 85 $M_\odot$ star with mass-loss rate ($M_\odot$) prescriptions for MS and post-MS phases. Such models could lead to massive BHs such as the recently detected 70 $M_\odot$ BH in the LB-1 system. Based on the observational constraints for $M_\odot$ of luminous stars, we discuss two possible scenarios that could produce such a massive BH at high metallicity. First, if the progenitor of LB-1 evolved from the observed population of WNh stars, we show that its average mass loss rate during its post-MS evolution was less than $1\,\times10^{-5}\,M_\odot/$yr. However, this is lower than the typical observed mass-loss rates of LBVs. A second possibility is that the progenitor evolved from a yet undetected population of 80-85~\msun\ stars with strong surface magnetic fields, which could quench mass loss during the MS evolution. In this case, the average mass-loss rate during the post-MS luminous blue variable (LBV) phase has to be less than $5\,\times10^{-5}\,M_\odot/$yr. This value is still low, considering that LBVs such as AG Carinae typically have average mass-loss rates from quiescent stellar winds close or above this level. LBVs directly collapsing to massive BHs are apparently at odds with the evidence that LBVs are the direct progenitors of some supernovae (SNe). To reconcile this, we suggest that LBVs from single stars (or mergers that fully mixed) have large cores and form BHs, while binary LBVs may have smaller cores and larger envelopes (either mergers or mass gainers that do not fully mix) and could produce a SN.