What stress components drive mechanochemistry? A study of ZDDP tribofilm formation

Abstract

Zinc dialkyldithiophosphate (ZDDP), the most widely used antiwear additive in engine oils, has been extensively studied over the last few decades to help understand the origin of its effectiveness. Glassy phosphate-based tribofilms, approximately 100 nm thick, are often formed on surfaces sliding in ZDDP-containing oils, which help to prevent or reduce wear. Recent studies reveal that a combination of applied shear and compressive stresses drive mechanochemical reactions that promote tribofilm growth, and that growth is further accelerated by increased temperature. While recent work has shown that compressive stress alone is insufficient to form tribofilms, the individual effects of the shear stress and compressive stress are not fully understood. Here, shear and compressive stresses are studied separately by using different ratios of high-viscosity, high-traction fluids for testing. This allows the areal mean compressive and shear stresses in the fluid when confined at a loaded sliding interface, to be independently controlled while driving tribofilm growth, which is a system we refer to as a stress-controlled mechanochemical reactor. Tribofilms derived from a secondary ZDDP were generated using a tungsten carbide/tungsten carbide ball-on-disk contact in the full elastohydrodynamic lubrication (EHL) regime using a mini-traction machine (MTM), meaning that solid–solid contact is avoided. The MTM was equipped with a spacer layer imaging (SLIM) capability, permitting in situ measurement of the tribofilm thickness during its growth. The well-separated sliding surfaces generated by the high-viscosity fluids confirm that solid–solid contact is not required for tribofilm formation. Under these full fluid film EHL conditions, shear stress and temperature promote tribofilm growth in accordance with stress-augmented thermal activation. In contrast, under constant shear stress and temperature, compressive stress has the opposite effect, inhibiting tribofilm growth. Using the extended Eyring model for shear- and hydrostatic pressure-affected reaction kinetics, an activation energy of 0.54 ± 0.04 eV is found, consistent with prior studies of ZDDPs. The activation volume for shear stress is found to be 0.18 ± 0.06 nm³, while that for the compressive stress component is much smaller, at 0.010 ± 0.004 nm³. This not only confirms prior work supporting that shear stress drives tribofilm growth, but demonstrates and quantifies how compressive stress inhibits growth, consistent with the rate-limiting step in tribofilm growth involving a bond-breaking reaction. Implications of these findings are discussed.