Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Computational materials design

The role of computers in scientific research has notably increased over the last several decades. This has been driven by the increasing computational power of these machines as well as more efficient computational approaches, which has enabled the analysis of the increasingly complex materials that are investigated for widely differing needs such as energy, information, or applications at extreme conditions. Better materials need to be designed, but in order to achieve improved materials there needs to be comprehensive insight into behaviour, from the ångström to the micrometre, and from the femtosecond to the millisecond. Computation can certainly help in this endeavour.

In this Insight we aim to capture the role of computation for the understanding, prediction, and design of materials. Different computational methodologies are used to study phenomena at different lengthscales and timescales. Electronic structure methods such as density functional theory can be used to predict the fundamental properties of materials from their smallest constituents: atoms, chemical bonds between atoms, and unit cells, obtaining ground-state properties such as elastic constants as well as excited-state properties such as optical absorption. The study of larger-scale phenomena, for example the bottom-up assembly of nanoparticles into aggregates or crystals, requires less granular but more efficient techniques such as molecular dynamics or Monte Carlo approaches. Multiscale modelling can allow computation to model even greater lengthscales, allowing the determination of properties at sizes directly relevant for engineering, but without sacrificing important information at finer scales. And the rise of machine learning over the last few years has certainly made an impact, allowing the analysis of big data with the extraction of hard-to-see correlations, or the development of improved atomistic approaches that could aid materials design through enhanced accuracy and speedup, and resulting in more reliable descriptions of complex systems.

Many of these approaches are still under active development and refinement. Moreover, the digital infrastructure required to support these simulations, and allow data sharing and validation, is still emerging. The role of computation in providing insight into materials, and enabling materials design, will play a key role in materials innovation. We hope that you enjoy the glimpse into the future of computational materials design that this collection aims to provide. The fast pace of new developments in these research fields will certainly continue to dramatically advance our way of life.

Stephen Shevlin, Senior Editor

Bruno Castro, Senior Editor

Xin Li, Senior Editor

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Stephen Shevlin.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Shevlin, S., Castro, B. & Li, X. Computational materials design. Nat. Mater. 20, 727 (2021). https://doi.org/10.1038/s41563-021-01038-8

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41563-021-01038-8

Further reading

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing