What it is
Microsoft's MatterGen generates stable, diverse inorganic crystals across the periodic table and can be fine-tuned to hit target chemistry, symmetry, and mechanical, electronic, or magnetic properties. Its structures are more than twice as likely to be new and stable as prior generative models and more than ten times closer to the local energy minimum. As a proof of concept, the team synthesized one generated material and measured a property within 20% of the target.
Why it matters
Materials discovery is a bottleneck for energy storage, catalysis, and carbon capture. A generative model that proposes synthesizable, property-targeted crystals shifts the search from screening to direct design.
Underlined numbers link to their source. Every metric and quoted figure is listed under Sources and data below.
Filed undermaterials discovery, generative model, inorganic crystals, MatterGen, ai for science