Constraints on the contributions to the observed binary black hole population from individual evolutionary pathways in isolated binary evolution

Gravitational waves from merging binary black holes can be used to shed light on poorly understood aspects of massive binary stellar evolution, such as the evolution of massive stars (including their mass-loss rates), the common envelope phase, and the rate at which massive stars form throughout the cosmic history of the Universe. In this paper we explore the \emph{correlated} impact of these phases on predictions for the merger rate and chirp mass distribution of merging binary black holes, aiming to identify possible degeneracies between model parameters. In many of our models, a large fraction (more than 70% of detectable binary black holes) arise from the chemically homogeneous evolution scenario; these models tend to over-predict the binary black hole merger rate and produce systems which are on average too massive. Our preferred models favour enhanced mass-loss rates for helium rich Wolf--Rayet stars, in tension with recent theoretical and observational developments. We identify correlations between the impact of the mass-loss rates of Wolf--Rayet stars and the metallicity evolution of the Universe on the rates and properties of merging binary black holes. Based on the observed mass distribution, we argue that the $\sim 10\%$ of binary black holes with chirp masses greater than $40$ M$_\odot$ (the maximum predicted by our models) are unlikely to have formed through isolated binary evolution, implying a significant contribution (> 10%) from other formation channels such as dense star clusters or active galactic nuclei. Our models will enable inference on the uncertain parameters governing binary evolution in the near future.