Chemical transformation of CO in evolving protoplanetary discs across stellar masses: A route to C-rich inner regions

Context. Protoplanetary discs around very low mass stars (VLMSs) show hydrocarbon-rich MIR spectra indicative of C/O>1 in their inner discs. This is in contrast to such discs around higher-mass hosts, which are typically richer in O-bearing species. Aims. The two scenarios proposed to elevate C/O around the inner discs of VLMSs are the release of C by eroding carbonaceous grains or the advection of O-depleted gas from the outer disc. However, if CO gas remains abundant, sufficiently O-depleted material cannot be produced. We tested whether the chemical transformation of CO into other species allows the transport scenario to produce C/O significantly in excess of 1. Methods. We tracked the inner disc C/H and O/H over time using a 1D disc evolution code that models the transport of gas and ice phase molecules and includes the conversion of some species into others to represent key reaction pathways operating in the midplane. We explored the influence of disc mass, size, ionisation rate, and the presence of a dust trap. Results. The inner disc C/O increases over time due to sequential delivery where O-rich species (e.g. H2O) give way to C-rich species (e.g. CH4). To reach C/O>1, separating C and O is key, and hence the gas phase destruction of CO by He+, which liberates C, is critical. Ionisation drives the midplane chemistry and must have rates ≳10−17 s−1 (at least for VLMSs) for significant chemical evolution within the disc lifetime. However, the rates must be ≲10−17 s−1 for T Tauri stars to ensure their C/O remains less than 1 for the first few megayears. Initially more compact discs lose O-rich ices faster and reach a higher C/O. A warm dust trap between the CH3OH and CH4 snowlines traps CH3 OH ice (formed via hydrogenation of CO ice) for long enough to be photodissociated, providing an alternative way to liberate the C that started in CO in the form of CH4 gas that keeps the inner disc significantly C rich. Conclusions. The destruction of gaseous CO combined with gas advection and radial drift can deplete O enough and produce sufficient hydrocarbons to explain the typical C/O>1 of VLMSs. While their C/O is typically higher than for T Tauri stars due to the faster sequential delivery, achieving values significantly in excess of 1 likely also requires higher ionisation rates and more compact discs than for T Tauri stars. Observations of older discs may distinguish whether a higher ionisation rate is indeed required or if the faster physical evolution timescales alone are sufficient.