A brief history of atomic simulation as it was used in chemistry, physics, and materials science is presented starting with seminal work by Eyring in the 1930s through to current work and future challenges. This article provides the background and perspective needed to understand the ways in which reactive many-body potentials developed over the last three decades and have impacted materials research. It also explains the way in which this substantial impact on the field has been facilitated by increases in computational resources and traces the development of reactive potentials, which have steadily increased in complexity and sophistication over time. Together with the other contributions in this issue of MRS Bulletin, this article will help guide and inspire the next generation of computational materials scientists and engineers as they build on current capabilities to expand atomic simulation into new and exciting areas of materials research.

This article discusses some of the many-body potentials used for simulations of processes and energies in materials at the atomic scale, emphasizing their motivation and underlying physical concepts, particularly where these are not entirely empirical. The perspective is somewhat historical and describes the importance of developments of the theory of electrons in solids for the derivation of many-body (or many-atom) potential models. The models include density-dependent pairwise potentials, effective medium and embedded-atom models, and polarizable ion models. As a recent radical departure from approaches derived from the physics of electrons, the development of models based on information theory is also described.

Many-body potentials were introduced in the early 1980s and have become a workhorse for the simulation of materials, especially metallic systems. The physical motivations for the main classes of the various many-body potentials are summarized, and the advantages of this approach are discussed. Some current examples related to grain growth, stress generation in thin films, shock loading, and nanowire deformation are presented to illustrate the continuing value of these approaches. Finally, some of the approaches that have been introduced in subsequent years are briefly described.

Coulson’s bond order is a chemically intuitive quantity that measures the difference in the occupation of bonding and anti-bonding orbitals. Both empirical and rigorously derived bond order expressions have evolved in the course of time and proven very useful for atomistic modeling of materials. The latest generation of empirical formulations has recently been augmented by screening-function approaches. Using friction and wear of diamond and diamond-like carbon as examples, we demonstrate that such a screened bond order scheme allows for a faithful description of dynamical bond-breaking processes in materials far from equilibrium. The rigorous bond order expansions are obtained by systematic coarse-graining of the tight binding approximation and form a bridge between the electronic structure and the atomistic modeling hierarchies. They have enabled bottom-up derivations of bond order potentials for covalently bonded semiconductors, transition metals, and multicomponent intermetallics. The recently developed magnetic bond order potential gives a correct description of both directional covalent bonds and magnetic interactions in iron and is able to correctly predict the stability of bulk Fe polymorphs as well as the intricate properties of dislocation cores. The bond order schemes hence represent a family of reliable and powerful models that can be applied in large-scale simulations of complex processes involving fracture, wear, and plasticity.

Recent developments in reactive potentials for the simulation of complex bonding and complex chemistry are reviewed. In particular, the reactive force field and charged optimized many-body methods are two paradigms that enable atoms to autonomously determine their charge state and the nature of their local bonding environments. The capabilities of these methods are illustrated by examples involving ionic-covalent systems, a metal-covalent system, a high-k dielectric gate stack, and the interaction of water with an oxide. Prospects for future development and applications are also discussed.

We discuss the relative complexity and computational cost of several popular many-body empirical potentials, developed by the materials science community over the past 30 years. The inclusion of more detailed many-body effects has come at a computational cost, but the cost still scales linearly with the number of atoms modeled. This is enabling very large molecular dynamics simulations with unprecedented atomic-scale fidelity to physical and chemical phenomena. The cost and scalability of the potentials, run in serial and parallel, are benchmarked in the LAMMPS molecular dynamics code. Several recent large calculations performed with these potentials are highlighted to illustrate what is now possible on current supercomputers. We conclude with a brief mention of high-performance computing architecture trends and the research issues they raise for continued potential development and use.

Insects constitute the most diverse and populated subclass of animals, with two million species identified. They also display a vast diversity of morphological and functional adaptations that allow them to thrive in various environments, which enables them to fly, swim, or walk nearly anywhere. Insects can be regarded as highly efficient and robust bio-machines, a precious source of material and information for bioinspired miniature technological devices. Yet, to date, little study of the functionality of insects has been undertaken with modern nanotechnology tools. Atomic force microscopy (AFM) is a technique generally used to study surface properties of materials at the nanoscale. Recently it has been shown that the AFM method can be extended to study complex living organisms, cells, and even entire animals, such as insects. AFM has demonstrated the feasibility of recording surface oscillations with sub-Angstrom spatial and sub-millisecond temporal resolutions while positioning the AFM probe at different parts of an insect with nanometer precision. In effect, it enables the AFM to function as a nanostethoscope. This article describes how such a nanostethoscope can be used to study the material properties, physiological reactions, and sensing mechanisms of insects.