On the current emissions trajectory, Earth will only avoid catastrophic global warming through large-scale negative-emissions technology, likely backed by storage of CO2 in underground aquifers. In addition, underground H2 storage looks to be an essential part of the global energy transition due to the large volumes required.  Once injected underground, these partially-soluble gases can flow through the porous rock in a connected plume, become disconnected and trapped as tiny ganglia surrounded by groundwater, and also dissolve and then diffuse through the aqueous phase ("Ostwald ripening").  While convective flow dominates how the gases migrate in the short-term, diffusive transport is important for long-term storage security.  

Recent theoretical work on diffusive transport neglected or excessively simplified multiple factors, leaving uncertainty over the direction of  CO2 transport: does it diffuse downwards to safety or upwards towards escape?   Due to the length- and time-scales involved, it isn't feasible to study this problem experimentally.   

In this talk I'll describe the long journey of how we resolved this question through first-principles thermodynamic analysis, and made the surprising finding that entropic and thermodiffusive effects — arising from geothermal gradients — dominate other effects and overturn previous results.