The Impact of Black Hole Formation on Population-averaged Supernova Yields

Abstract

Abstract The landscape of black hole (BH) formation—in which massive stars explode as core-collapse supernovae (CCSN) and which implode into BHs—profoundly affects the initial-mass-function (IMF)-averaged nucleosynthetic yields of a stellar population. Building on the work of Sukhbold et al., we compute IMF-averaged yields at solar metallicity for a wide range of assumptions, including neutrino-driven engine models with extensive BH formation, models with a simple mass threshold for BH formation, and a model in which all stars from 8 to 120 M ⊙ explode. For plausible choices, the overall yields of α-elements span a factor of 3, but changes in relative yields are more subtle, typically 0.05–0.2 dex. To constrain the overall level of BH formation, ratios of C and N to O or Mg are promising diagnostics. To distinguish complex, theoretically motivated landscapes from simple mass thresholds, abundance ratios involving Mn or Ni are promising because of their sensitivity to the core structure of the CCSN progenitors. We confirm previous findings of a substantial (factor 2.5–4) discrepancy between predicted O/Mg yield ratios and observationally inferred values, implying that models either overproduce O or underproduce Mg. No landscape choice achieves across-the-board agreement with observed abundance ratios; the discrepancies offer empirical clues to aspects of massive star evolution or explosion physics still missing from the models. We find qualitatively similar results using the massive star yields of Limongi & Chieffi. We provide tables of IMF-integrated yields for several landscape scenarios, and more flexible user-designed models can be implemented through the publicly available Versatile Integrator for Chemical Evolution (VICE; https://pypi.org/project/vice/).