Atomistic simulations represent very powerful tools to monitor the physical and chemical processes at buried interfaces. I will present recent studies that aim at advancing i) materials to reduce friction by controlling tribochemistry processes ii) solid interfaces by high throughput screening of adhesion and shear strength.

i) Hypericin, commonly used as medical treatment, possesses amazing tribological properties. A hypericin concentration as low as 0.3% in glycerol can reduce the friction coefficient below 0.01 (the so-called superlubricity regime) under boundary lubrication. First-principles calculations elucidate the thermodynamic driving force for graphene formation from hypericin. The process, promoted by the tribological conditions, was monitored in real-time by ab initio and classical molecular dynamics simulations. 

ii) We developed a modular workflow, TribChem, connected to publicly available databases, and performed the high-throughput ab initio screening of the adhesion energy of around a hundred metallic heterostructures, ranging from transition to noble metals. The dataset allowed us to identify general trends and by using a machine learning approach, we obtained a simple analytical expression for predicting the adhesion energy from the surface energy other intrinsic properties of the two heterostructure constituents alone.