A predictive model to link molecular structures to carbonate rock wettability: Adsorption of various organic molecules on flat and rugged calcite surfaces

Wettability is a key petrophysical property which impacts multiphase flow in porous media. It is generally believed that rock wettability in subsurface reservoirs has been established by adsorption of organic molecules during oil invasion and that the majority of carbonate reservoirs are mixed-wet or oil-wet as a consequence hereof. Thus, insight into which molecules actually bind to the carbonate rock, estimation of molecule-specific relative adsorption affinities and the importance of the crystalline calcite structure is pivotal to decipher the origin of wettability in carbonate reservoirs - and modification hereof in EOR processes. We applied advanced density functional theory (DFT) to calculate adsorption energies of 36 different organic molecules on top of calcite with two different crystalline structures, namely flat {104} and rugged calcite {318}. We made an in-depth statistical analysis of the data and the work stands out to our knowledge as the most comprehensive and systematic quantitative study of adsorption of organic molecules reported to date. We show that molecules with a carboxylic group adsorb to calcite with highest affinity and that all 36 molecules bind with significantly higher affinity to rugged calcite as compared to flat calcite. Strikingly, this adsorption enhancement was most significant for molecules with functional groups containing nitrogen and oxygen. This suggests that regions in carbonate rocks with rugged surfaces of calcite have an inherent thermodynamic preference to become more oil-wet as compared to flat calcite surfaces when exposed to organic molecules with polar functional groups. We applied COSMO (conductor-like screening model) theory in conjunction with our recently published Generalized Sigma Moment (GSM) model to derive a predictive correlation to estimate calcite type-specific molecular adsorption energies. We derived two models with significant prediction power (R²=0.95 and R²=0.9) and show from nonlinear regression and clustering analysis that the electrostatic polarity of the organic molecules is the single most important determinant of adsorption energy. The thorough analysis and the correlations we present here are valuable and powerful tools for quantitative prediction of adsorption energies of wettability-modifying organic molecules and for deciphering the origin of wettability of carbonate rocks.