Hard-material Adhesion: Which Scales of Roughness Matter?

Abstract

Surface topography strongly modifies adhesion of hard-material contacts, yet roughness of real surfaces typically exists over many length scales. This investigation aims to determine which scales of topography have the strongest effect on macroscopic adhesion. Adhesion measurements were performed on technology-relevant diamond coatings of varying roughness using spherical ruby probes that are large enough ($0.5$-mm-diameter) to sample all length scales of topography. For each material, more than $2000$ measurements of pull-off force were performed in order to investigate the magnitude and statistical distribution of adhesion. Using sphere-contact models, the roughness-dependent effective values of work of adhesion were measured, ranging from $0.08$ to $7.15$ mJ/m$^2$ across the four surfaces. The data was more accurately fit using numerical analysis, where an interaction potential was integrated over the AFM-measured topography of all contacting surfaces. These calculations revealed that consideration of nanometer-scale plasticity in the materials was crucial for a good quantitative fit of the measurements, and the presence of such plasticity was confirmed with AFM measurements of the probe after testing. This analysis enabled the extraction of geometry-independent material parameters; the intrinsic work of adhesion between ruby and diamond was determined to be $46.3$ mJ/m$^2$. The range of adhesion was $5.6$ nm, which is longer than is typically assumed for atomic interactions, but is in agreement with other recent investigations. Finally, the numerical analysis was repeated for the same surfaces but this time with different length-scales of roughness included or filtered out. The results demonstrate a critical band of length-scales between $43$ nm and $1.8$ nm in lateral size that has the strongest effect on the total adhesive force for these hard, rough contacts.