The objective is to apply a multiscale approach, coupling the atomistic to the macroscopic scale, for the study of the planned projects. Various computational techniques will be used ranging from atomistic Monte Carlo and Molecular Dynamics simulations, at the empirical, tight-binding and ab initio level,  to continuum micromechanical modeling. The investigation of a problem will start at the atomistic level using empirical potentials, in order to extract the overall trends. This will be followed with more accurate atomistic simulations using either the tight-binding method or first-principle methods to extract exact results for the properties of interest. The results from the atomistic simulations will then be used as a guide to continuum modeling to extract the macroscopic picture.

A variety of complementary state-of-the-art computational methods will be used, able to provide a global picture ranging from an accurate atomistic description to the macroscopic scale. For great statistical accuracy, continuous-space Monte Carlo (MC) simulations will be used at the HO, based on algorithms developed over the years by Kelires and collaborators. The method can trace the stability of the structures as a function of temperature, pressure, and chemical potential. Phenomena which can be modeled include segregation, phase separation, phase transformations, clustering and phase coarsening, uniformity of phase dispersion, control of compositions and size of nanoinclusions, etc. All these affect the mechanical and tribological properties. The method is ideal for the study of the multimaterial nanocomposite coatings. In addition, within this MC approach, we shall use a technique previously developed to analyze the energetics, the stress fields and the elastic moduli into atomic-level contributions, in order to study variations and gradients across discontinuities and inhomogeneities, and identify potential sources of mechanical wear and fatigue. The interactions will be described by well-tested empirical potentials. Specifically, for C-C and Si-C interactions the Tersoff potentials will be used and for C(ceramic)-metal interactions appropriate combinations will be developed. The use of empirical potentials allows for the use of large computational cells which realistically simulate experimental work.

For the study of nanostructured coatings at the atomistic level, we will also carry out at HO Tight-Binding Molecular Dynamics (TBMD) simulations. The TBMD method, based on a quantum-mechanical description of the interactions, provides greater accuracy in the energies than the empirical-potential-based MC method, at the expense of smaller statistical accuracy. For this reason, the TBMD will complement the MC simulations extracting accurate conclusions about the stability and structure, as a function of grain size, as well as about the optoelectronic properties of the nanomaterials under study. In particular, this method will provide us with the dielectric functions and all the pertinent optical parameters and quantities extracted from it, such as the absorption coefficients and the optical band gaps, which will be compared with the experimentally deduced in the consortium quantities. The TB methods of Wang & Ho and Mehl & Papaconstantopoulos will be used.

For even higher accuracy, ab initio quantum mechanical calculations and MD simulations based on Density Functional Theory (DFT) will be carried out at PA3, which is a world leader in this activity, and at HO. Emphasis will be given to obtaining accurate information about structural local configurations and their energetics, especially about conformations on the surfaces and interfaces of the nanostructured materials. This, for example, will be crucial for the optimization of mechanical strength properties and adsorption properties. Especially, for light absorption problems (e.g., solar coating absorption) and the associated structural changes, a new approach that employs DFT and time-dependent DFT (TDDFT), developed by the group at Harvard, will be utilized. It uses a local atomic basis-set representation and real time propagation of single-particle wavefunctions. Time evolution of the electronic states is coupled with the classical Newtonian motion of ions using MD.

The coupling of the atomistic description to larger length scales will be implemented through advanced continuum micromechanical modeling developed within the theory of linear homogenization, to be carried out at the HO. Appreciating that the nanoinclusion/interphase combination essentially results in a composite (with the interphase layer acting as the “matrix”) within the global nanocomposite coating, the objective of the homogenization methodology is to replace the nanoinlusion/interphase composite with an “effective” or “homogenized” particle of intermediate radius and homogeneous properties. The data pertinent to the interphase layer that are needed for the homogenization study will be obtained from the atomistic simulations (see previous paragraphs) and the experimental data obtained via nanoindentation techniques. The homogenization model will thus serve as the coupling of both the experimental work and the atomistic description to the larger length scales. There is significant experience at the HO in the use of homogenization and other analytical/numerical techniques for modeling of composites and smart composites of varying macroscopic geometries and microscopic morphology. The results of the continuum modeling will serve as a benchmark to compare with the observations of our experimental characterization techniques.