Effects of surface chemistry on the mechanochemical decomposition of tricresyl phosphate.

The growth of protective tribofilms from lubricant antiwear additives on rubbing surfaces is initiated by mechanochemically promoted dissociation reactions. These processes are not well understood at the molecular scale for many important additives, such as tricresyl phosphate (TCP). One aspect that needs further clarification is the extent to which the surface properties affect the mechanochemical decomposition. Here, we use nonequilibrium molecular dynamics (NEMD) simulations with a reactive force field (ReaxFF) to study the decomposition of TCP molecules confined and pressurised between sliding ferrous surfaces at a range of temperatures. We compare the decomposition of TCP on native iron, iron carbide, and iron oxide surfaces. We show that the decomposition rate of TCP molecules on all the surfaces increases exponentially with temperature and shear stress, implying that this is a stress-augmented thermally activated (SATA) process. The presence of base oil molecules in the NEMD simulations decreases the shear stress, which in turn reduces the rate constant for TCP decomposition. The decomposition is much faster on iron surfaces than iron carbide, and particularly iron oxide. The activation energy, activation volume, and pre-exponential factor from the Bell model are similar on iron and iron carbide surfaces, but significantly differ for iron oxide surfaces. These findings provide new insights into the mechanochemical decomposition of TCP and have important implications for the design of novel lubricant additives for use in high-temperature and high-pressure environments.

Simulating Surfactant-Iron Oxide Interfaces: From Density Functional Theory to Molecular Dynamics.

Understanding the behavior of surfactant molecules on iron oxide surfaces is important for many industrial applications. Molecular dynamics (MD) simulations of such systems have been limited by the absence of a force field (FF), which accurately describes the molecule-surface interactions. In this study, interaction energies from density functional theory (DFT) + U calculations with a van der Waals functional are used to parameterize a classical FF for MD simulations of amide surfactants on iron oxide surfaces. The original FF, which was derived using mixing rules and surface Lennard-Jones (LJ) parameters developed for nonpolar molecules, was shown to significantly underestimate the adsorption energy and overestimate the equilibrium adsorption distance compared to DFT. Conversely, the optimized FF showed excellent agreement with the interaction energies obtained from DFT calculations for a wide range of surface coverages and molecular conformations near to and adsorbed on α-Fe2O3(0001). This was facilitated through the use of a Morse potential for strong chemisorption interactions, modified LJ parameters for weaker physisorption interactions, and adjusted partial charges for the electrostatic interactions. The original FF and optimized FF were compared in classical nonequilibrium molecular dynamics simulations of amide molecules confined between iron oxide surfaces. When the optimized FF was employed, the amide molecules were pulled closer to the surface and the orientation of the headgroups was more similar to that observed in the DFT calculations. The optimized FF proposed here facilitates classical MD simulations of anhydrous amide-iron oxide interfaces in which the interactions are representative of accurate DFT calculations.