Tribology and Contact Mechanics of Diamond-related Materials

Yifei Mo, Izabela Szlufarska, Materials Science Program

Diamond is an excellent material for tribological applications, because of its superior properties, e.g., extreme stiffness, wear resistance, surface stability, and low friction.

The advent of chemical vapor deposition (CVD) makes the use this perfect material for thin films coating possible. While the conventionally grown microcrystalline diamond films are too rough and contain a large number of defects, the nanocrystalline (NCD) and ultrananocrystalline diamond (UNCD) films made by new techniques have very smooth surface, which decreases friction and wear and make it an attractive material for micro- and nanoelectromechanical systems (MEMS/NEMS).

The fundamental understanding of friction and wear at the atomic scale will enable designing of surfaces with superior properties and preventing premature failure of MEMS/NEMS. Currently, there are no viable commercial MEMS/NEMS devices with sliding interface. The challenge in understanding tribology at a single asperity level lies in the multiple energy dissipation mechanisms participating in the process.

Many AFM studies have shown that in the wearless regime friction force at the nanoscale contact is proportional to the real interfacial contact area, F_f = τ ·A, where A is the real contact area and τ is the interfacial shear strength. The contact area A  does not scale linearly with load L anymore, which is qualitatively different from the macroscale Amontons’ law. The single asperity real area of contact A is not measured directly in experiments, but it is inferred based on continuum contact mechanics models, such as, Hertz, Johnson-Kendall-Roberts (JKR), Derjaguin-Muller-Toporov (DMT), and Maugis-Dugdale models.

The continuum level models have been developed for macroscopic contacts and there is no a priori reason to believe that these models will still be capable to interpret or predict the tribological behavior at the nanoscale. The aforementioned assumptions need to be reevaluated at the nanoscale, in particular that a considerable number of cases have been reported where these assumptions break down in experiments and simulations.

MD studies have been used to reveal the break-down of continuum mechanics at the nanoscale. They all agree on certain aspects of the problem, e.g., that continuum models underestimate the contact area, and disagree of some other fundamental questions, e.g., dependence of friction on load. All these simulations suffer from some limitations, i.e., unrealistic system setup (rigid tips), and/or unphysical interatomic potential, and/or very small system size.

We are conducting multi-million to over a billion atom molecular dynamics (MD) simulations with a state-of-the-art reactive empirical bond-order potential (2nd generation REBO) integrated with long-range interactions to study friction of single crystal diamond and UNCD at the single asperity level. Our atomistic simulations are carried out for systems at length-scales comparable with experimental atomic force microscope (AFM) tips and therefore our model predictions can be directly compared to experiments. We are in a unique position to address all of these limitations due to our capability to perform massively parallel MD simulation with the REBO potential. We will use the developed tools to study the laws of contact mechanics at the nanoscale and we will compare our results to the widely available experimental data for diamond.