Friction is important in everyday life, but because it is everywhere we often take it completely for granted. Every time you stand still, walk, or hold an object, friction helps you to control movements. At the same time, the friction of moving parts in machines, and the associated wear, costs modern societies an enormous amount of energy and materials. Because friction is part of everything we do, human beings have been dealing with it since prehistoric times. We have phenomenological laws that describe how friction behaves, but they do not explain where it comes from.
Friction arises from microscopic interactions between atoms and molecules in the sliding interface at tiny contact asperities. In the past decades, we have finally become able to investigate and understand friction also at these small length scales, thanks to enormous advances in experimental techniques. Atomic Force Microscopy (AFM) and Friction Force Microscopy (FFM), measure the friction of a nanoscale tip moving over a surface at very low velocities, up to a maximum of around ten micrometer per second. While these experiments have given much insight, these speeds are far below the typical values in macroscopic applications. Techniques exist that allow a probe tip to move faster, but they lose the high resolution. The combination of small length scale and realistically high velocity has thus remained illusive.
In our recent paper in Nature Communications we present the first results of just such a technique: Intermodulation FFM (ImFFM). The ImFFM technique achieves high velocity by exciting a torsional resonance of the AFM cantilever that sweeps the tip back and forth over the substrate. We combine this with the high sensitivity of intermodulation AFM (ImAFM), which was previously developed at KTH. We can perform these experiments rapidly, and determine the velocity-dependence of friction with sub-nm resolution, as well as observe sub-nm features on the surface.
Nevertheless, we still pay a price for the high speeds and high sensitivity: while the information we are looking for is contained in the measured output data, interpreting it in a meaningful way is not easy. This is why we chose graphite as the surface for our experiments. It is relatively simple and well-characterised, so that we we knew what to expect. By simulation of a simple model, we were able to capture several characteristic signature features of the interaction, which helped us to better understand and explain the experiments.
This blog entry was written by Associate Professor Astrid de Wijn at the Department of Mechanical and Industrial Engineering.