Optimal model of semi-infinite graphene for ab initio calculations of reactions at graphene edges by the example of zigzag edge reconstruction

TL;DR

This study refines a semi-infinite graphene model for ab initio edge reaction calculations, demonstrating optimal parameters for convergence and revealing effects of defects on edge magnetization and spin ordering.

Contribution

It provides specific model parameters for accurate ab initio calculations of graphene edge reactions and explores defect impacts on magnetic properties.

Findings

Optimal nanoribbon width of 6 zigzag rows for convergence

Reaction energy of -0.15 eV and activation barrier of 1.61 eV for edge reconstruction

Defects locally reduce edge magnetization and induce spin order switching

Abstract

We investigate how parameters of the model of semi-infinite graphene based on a graphene nanoribbon under periodic boundary conditions affect the accuracy of ab initio calculations of reactions at graphene edges by the example of the first stage of reconstruction of zigzag graphene edges, formation of a pentagon-heptagon pair. It is shown that to converge properly the results, the nanoribbon should consist of at least 6 zigzag rows and periodic images of the pair along the nanoribbon axis should be separated by at least 6 hexagons. The converged reaction energy and activation barrier for formation of an isolated pentagon-heptagon pair are found to be -0.15 eV and 1.61 eV, respectively. It is also revealed that such defects reduce the graphene edge magnetization only locally but ordering of spins at opposite nanoribbon edges switches from the antiparallel (antiferromagnetic) to parallel one (ferromagnetic) upon increasing the defect density.