Linear Scaling Quantum Transport Methodologies

In recent years, the role of predictive computational modeling has become a cornerstone for the study of fundamental electronic, optical, and thermal properties in complex forms of condensed matter, including Dirac and topological materials. The simulation of quantum transport in realistic materials calls for the development of linear scaling, or order-$N$, numerical methods, which then become enabling tools for guiding experimental research and for supporting the interpretation of measurements. In this review, we describe and compare different order-$N$ computational methods that have been developed during the past twenty years, and which have been intensively used to explore quantum transport phenomena. We place particular focus on the electrical conductivities derived within the Kubo-Greenwood and Kubo-Streda formalisms, and illustrate the capabilities of these methods to tackle the quasi-ballistic, diffusive, and localization regimes of quantum transport. The fundamental issue of computational cost versus accuracy of various proposed numerical schemes is also addressed. We then extend the review to the study of spin dynamics and topological transport, for which efficient approaches of inspecting charge, spin, and valley Hall conductivities are outlined. The supremacy of time propagation methods is demonstrated for the calculation of the dissipative conductivity, while implementations fully based on polynomial expansions are found to perform better in the presence of topological gaps. The usefulness of these methods is illustrated by various examples of transport in disordered Dirac-based materials, such as polycrystalline and defected graphene models, carbon nanotubes as well as organic materials.