Atomistic insights into the tribology of lubricant additives and black phosphorus: from ab initio to machine learning potentials

Tribology is the science that studies what occurs between two surfaces in relative motion, including friction, wear, and lubrication. Its results have an impact on the efficiency, durability, and sustainability of mechanical systems across all scales, from large industrial machinery to nanoscale devices. Understanding the interfacial mechanisms that govern energy dissipation and material degradation remains a central challenge in modern engineering. In this thesis, first-principles simulations based on Density Functional Theory (DFT) and Molecular Dynamics (MD) based on Machine Learning Potentials (MLPs) are integrated with experiments to investigate atomistic processes that rule the functionality of lubricant materials. The research examines two main classes of systems. The first focuses on molecular lubricant additives, particularly zinc dialkyldithiophosphates (ZDDPs) and other well established phosphorus-based compounds. Simulations reveal how isomerization, oxidation, and bond cleavage at iron surfaces affect tribofilm formation. The use of machine-learning interatomic potentials enables the exploration of realistic time and length scales, providing new insight into interfacial chemistry of ZDDP. The second class involves solid lubricants and, specifically, black phosphorus, a two-dimensional material with unique mechanical and electronic properties. Its interaction with oxygen and water, as well as the effects of chemical modification (i.e., fluorination) and adhesion to substrates, are investigated to understand how surface chemistry influences stability and frictional behavior. By bridging atomistic mechanisms with macroscopic performance, this thesis deepens the understanding of interfacial phenomena and aims at providing a computational framework for the rational design of next-generation lubricants and coatings that support more efficient and sustainable technologies.