Accelerating New Materials Development: Material Projects, Google DeepMind, and A-Lab Collaboration
Accelerating New Materials Development: Material Projects, Google DeepMind, and A-Lab Collaboration: Material Projects, an open-access database from Berkeley Lab, provides researchers with crucial information on various materials. Google DeepMind, through its GNoME tool, has contributed 380,000 crystal structures to Material Projects. Some GNoME calculations were utilized with Materials Project data to test A-Lab, a futuristic laboratory at Berkeley Lab that combines robotics and AI to expedite the discovery of new materials.
Founded in 1931, the Lawrence Berkeley National Laboratory (Berkeley Lab) is a multiprogram national laboratory managed by the University of California for the U.S. Department of Energy's Office of Science. The A-Lab, situated within Berkeley Lab, commenced operations last February and operates 24/7.
Material Projects: A Database for Material Discovery Initiated in 2011 by Berkeley Lab, the Materials Project is the world's most widely used open-access repository of information on inorganic materials, boasting 400,000 users worldwide. By calculating properties of known or predicted materials, it enables researchers to focus on promising materials for future technologies—such as lighter alloys for fuel efficiency, more efficient solar cells for renewable energy, or faster transistors for the next generation of computers.
Google DeepMind utilized information from the Materials Project to develop the GNoME tool, based on deep learning, leading to the discovery of 2.2 million new crystals. Among these, 380,000, considered stable and promising for future technological applications, were added to the Material Projects database.
The dataset includes information on how atoms are arranged in a material (crystal structure) and its stability (formation energy).
Kristin Persson, founder and director of the Materials Project at Berkeley Lab and professor at UC Berkeley, explains:
"We must create new materials if we want to address global environmental and climate challenges. Through materials innovation, we can potentially develop recyclable plastics, harness waste energy, manufacture better batteries, and build cheaper, longer-lasting solar panels, among other things."
A-Lab: AI-Guided Automation for New Material Synthesis Led by scientist Yan Zeng, A-Lab employs AI-guided robots to perform the complex steps of the material synthesis process. This automation allows processing 50 to 100 times more samples per day than a human researcher, significantly accelerating the discovery pace.
In just 17 days, A-Lab successfully created 41 new compounds predicted by the Materials Project in a trial of 58 attempts—a rate of over two new materials per day, compared to months for a human researcher. GNoME data served as additional verification of the stability of these predicted materials.
Gerd Ceder, principal investigator at A-Lab and scientist at Berkeley Lab and UC Berkeley, concludes:
"We achieved this staggering success rate of 71%, and we already have ways to improve it further. We've shown that combining theory and data with automation yields incredible results. We can manufacture and test materials faster than ever, and adding additional data points to the materials project means we can make even smarter choices."