Atomic-scale friction, as accessed in tip-based experiments, is investigated theoretically in the full range of surface corrugations, temperatures, and velocities. Emphasis is given to the regime of thermal drift, when the regular stick-slip behavior is completely ruined by thermal effects. The possibility of nearly vanishing friction ("thermolubricity") is predicted even for strong (overcritical) surface corrugations, when traditional models would predict significant friction. The manifestation of this effect in recently published experimental data is demonstrated.

In this paper, we use a set of rate equations to describe the thermal activation of a tip moving along a one-dimensional lattice, including the possibility of multiple back and forth jumps between neighboring potential wells. This description of an atomic-scale friction experiment is used to investigate how temperature acts as a lubricant, an effect that we refer to as thermolubricity. We discuss the detailed theoretical aspects of the model, which explains many aspects of the variation in atomic friction over a wide range of temperatures, velocities, and surface corrugations. We conclude that friction at low velocities and low surface corrugations is much lower than the weak logarithmic velocity dependence predicted before. Another consequence of the model is the trivial result that friction is zero in the zero-velocity limit. We confront numerical results from our theoretical model with experiments, in which the surface corrugation was controlled by use of geometrical effects, to demonstrate the experimental existence of thermolubricity. Although the calculations produce excellent fits to our data, the values of the fitting parameters clearly indicate that the underlying single-spring model suffers from an intrinsic flaw, which we ascribe to either the absence of flexibility of the tip or the restriction to a one-dimensional sliding geometry

No abstract