Computational materials science has begun to accelerate the discovery of advanced materials, and many of the most notable of theory-inspired discoveries have been in energy-related fields like catalysis, electrochemical energy storage, and solar photovoltaics. Since its inception in 2010, The Materials Project has sought to use best-practices from various information technology fields to make the process of computational simulation of solid materials more efficient, well-documented, and shareable among the materials science community. In this work, we highlight the most recent developments from The Materials Project geared towards furthering the goal of open high-throughput computational materials science.

Among these is an expansion of various computational materials science methods into high-throughput capable workflows in new code repositories geared towards allowing users to execute workflows in a way that enables efficient replication and documentation of computational procedures. In addition, our workflow infrastructure has been adapted to supercomputing architectures not traditionally used in computational materials science by parallelizing individual DFT calculations across nodes. We also have new workflows geared towards simulating various spectroscopies, including XAS and Raman, as well as calculating adsorption energies on solid surfaces for the purpose of screening electrocatalysts and predicting condition-specific nanoparticle shapes.

Furthermore, updates to our online infrastructure allow users to request advanced properties like the elastic tensor and to monitor the resultant procedure that occurs in our simulation infrastructure. During the period from September to December of 2016, Over 600 elastic property requests, which involve over 15000 individual DFT calculations of deformed materials, have been fulfilled using a voting system by which users may indicate materials on materialsproject.org which should be prioritized in our queue of calculations. In addition, we have worked with the Office of Science and Technological Information (OSTI) at the Department of Energy to provide digital object identifiers (DOIs) for individual materials, enabling researchers to cite our online infrastructure more accurately and appropriately.

These milestones evidence how valuable high-throughput materials science via efforts like the Materials Project can be, but a number of challenges still remain towards leveraging the computational materials science community to its full potential. These include more effective educational outreach, which we have begun by designing workshop and classroom curricula. Our first workshop was held in July 2016, and we hope to continue making our approach and our results more accessible to the general community via efforts like these. We also have begun deploying a platform, MPContribs, designed to enable our users to contribute both theoretical and experimental results such that they may be accessible on The Materials Project. We hope to use platforms like these to increase user engagement and more closely compare ours and others' theoretical approaches to community-sourced experimental data.

In summary, efforts like The Materials Project are accelerating the discovery of new materials for energy applications. In this work, we outline recent strategies for furthering the rate of materials discovery by more effectively leveraging our own resources and by promoting a more robust computational materials science community.