Materials Project

Materials Project
Name	Materials Project
Type	Research initiative
Founded	2011
Founder	Gerbrand Ceder, Chris Wolverton, Jimmy Sage
Location	Berkeley, California
Focus	Computational materials science, materials discovery
Parent	Lawrence Berkeley National Laboratory

Materials Project

The Materials Project is a computational initiative that accelerates the discovery of inorganic materials using high-throughput density functional theory and open data sharing. Based at Lawrence Berkeley National Laboratory and associated with the Berkeley Lab, the project connects researchers from institutions such as Massachusetts Institute of Technology, University of California, Berkeley, and national facilities like Argonne National Laboratory to provide curated materials properties for theory-driven and experimental workflows. It serves as an infrastructural resource comparable in ambition to databases like PubChem, Protein Data Bank, and Cambridge Structural Database for the domain of crystalline solids.

Overview

The project compiles computed properties — formation energies, electronic structures, elastic tensors, optical spectra — for tens of thousands of inorganic compounds, enabling cross-cutting studies spanning Toyota Research Institute, Google Research, IBM Research, and university labs. Its data model integrates provenance and standardization practices from computational initiatives at Oak Ridge National Laboratory and benchmarking efforts similar to work at Joint Center for Energy Storage Research. Users from Stanford University, Harvard University, Caltech, and industrial partners access the resource through programmatic APIs, web interfaces, and data export pipelines used in studies ranging from battery materials at Toyota Motor Corporation to photovoltaics investigated by teams at National Renewable Energy Laboratory.

History and Development

Conceived in the early 2010s, the initiative emerged from efforts led by researchers affiliated with University of California, Berkeley and Massachusetts Institute of Technology, building on methodologies developed by groups at Harvard University and computational toolchains influenced by work at Argonne National Laboratory. Initial funding and infrastructure support came from agencies such as the U.S. Department of Energy and collaborative programs involving the National Science Foundation. Over time the project incorporated community feedback from conferences like Materials Research Society meetings and workshops at Sandia National Laboratories, expanding its scope to include battery-focused datasets influenced by consortia including Joint Center for Energy Storage Research.

Database and Methodology

The dataset is generated primarily via plane-wave density functional theory calculations following protocols comparable to those used in studies at Max Planck Institute for Solid State Research and benchmarking campaigns at Lawrence Livermore National Laboratory. Relaxations, total energies, and electronic density-of-states are computed with parameter sets aligned with conventions from VASP-based studies and verified against experimental summaries published in journals like Physical Review Letters and Nature Materials. Crystallographic inputs reference prototype structures cataloged in collections such as the Inorganic Crystal Structure Database and legacy compilations associated with Crystallography Open Database. The project implements thermodynamic analysis tools akin to those utilized by research groups at Columbia University and phase stability frameworks comparable to techniques described in publications from Northwestern University.

Software and Tools

To serve diverse user communities, the initiative provides software libraries and application programming interfaces inspired by toolkits developed at Lawrence Berkeley National Laboratory and community projects at Python Software Foundation ecosystems. Client libraries support integrations with environments common at Microsoft Research and interactive platforms used in labs at Yale University. Visualization modules parallel efforts found in projects such as ASE (Atomic Simulation Environment) and plotting conventions used in publications from Princeton University. Workflows for automated high-throughput calculations draw on scheduler integrations used at facilities like Oak Ridge National Laboratory and cluster environments at Sandia National Laboratories.

Research Applications

Researchers use the resource to accelerate design workflows in energy storage studied at Tesla, Inc. and Toyota Motor Corporation, catalysis programs at Argonne National Laboratory, and semiconductor investigations at Intel Corporation. Studies leveraging the dataset have enabled predictions of novel electrode materials investigated in collaborations with MIT Energy Initiative and light-absorbing compounds explored by teams at University of Cambridge. The data have supported machine learning models developed in partnerships with groups at Google DeepMind, Facebook AI Research, and academic labs at University of Chicago to screen materials for properties reported in journals such as Science and Advanced Materials.

Community and Collaborations

The project maintains collaborations with national laboratories including Lawrence Livermore National Laboratory and Oak Ridge National Laboratory, academic consortia at University of California, San Diego and University of Michigan, and industrial partners across energy and semiconductor sectors like Samsung Electronics and General Motors. It contributes to community standards alongside initiatives such as NOMAD Laboratory and engages with open-data movements represented by organizations like Open Data Institute. Educational outreach includes workshops at conferences such as American Physical Society meetings and tutorials delivered at summer schools affiliated with Materials Research Society.

Category:Materials science databases