Atomistic simulation of a graphene-nanoribbon–metal interconnect

Abstract

We report a molecular statics simulation of the physical processes responsible for binding and lattice distortions in a nanoscale electrical interconnect with realistic boundary conditions. The interconnect consists of a graphene ribbon interfaced with the (111) crystallographic surfaces (over 11,000 atoms overall) of two nickel electrodes. We quantify the graphene lattice distortions by mapping strains, as well as out-of-plane atomic displacements on a grid, throughout the simulated interconnect. The results suggest strongly localized graphene lattice distortions at the edges and strains that do not exceed 0.5% elsewhere. Such strains are not expected to affect the electrical properties of the graphene nanoribbon interconnect. A stand-alone graphene nanoribbon is simulated in order to identify the effect of electrodes partially supporting the graphene nanoribbon. Our results indicate that the electrodes reduce the in-plane strains induced by the nanoribbon edges, while causing rippling of the graphene lattice. The average graphene-nickel intersurface separation and the cohesive energy for the top-fcc configuration are calculated at ∼2.13 Å and 68.22 meV Å(-2). In order to describe the interatomic interactions in the simulation, we utilize a set of accurate atomistic potentials for graphene on a nickel surface. The approach is based on the modified embedded atom method (MEAM) for the C-C and Ni-Ni interactions, and a Morse-type potential, which takes the surface configuration into account, for the Ni-C interactions. Our focus is on the Ni-(111) crystallographic surface interfaced with graphene in top-fcc, top-hcp, and hcp-fcc initial configurations. The interactions were validated by calculating the equilibrium binding energy and intersurface distance. The resulting binding energies and equilibrium intersurface separations obtained are in very good agreement with previous experimental and ab initio data obtained by use of density functional theory (DFT).