Uspex

Background and Founding

Uspex emerged from the theoretical and computational work of Artem R. Oganov, a prominent crystallographer and materials scientist. Oganov developed the evolutionary algorithm methodology for crystal structure prediction during his tenure at ETH Zurich and later at Stony Brook University, before establishing the lab at Skoltech around 2011. The name Uspex is an acronym derived from the Russian words for "Universal Structure Predictor: Evolutionary Xtallography."

Software and Methodology

The core product is the Uspex code, which uses an evolutionary algorithm coupled with density functional theory (DFT) calculations to explore the energy landscape of crystal structures. It can predict not only ground-state structures but also metastable phases, which are crucial for understanding pressure-induced phase transitions, novel materials synthesis, and mineral physics. The software is available for academic use under a license and is continuously updated with new features such as variable-composition prediction, surface structure prediction, and inclusion of external fields.

Impact and Applications

Uspex has been instrumental in predicting hundreds of new crystal structures that were later confirmed experimentally. Notable successes include the prediction of post-perovskite phases, superconducting hydrides under high pressure (e.g., LaH10), and novel carbon allotropes. The software is cited in thousands of scientific papers and is regarded as one of the most reliable tools for computational materials discovery. Its user base includes researchers from leading universities, national laboratories, and materials science departments worldwide.

Community and Licensing

Unlike a commercial product, Uspex is primarily disseminated through academic channels. The source code is not fully open, but it is provided free of charge to academic researchers upon request. The lab also conducts tutorials and workshops, and Oganov’s group publishes extensively on methodology improvements. There is no formal pricing structure or sales team; instead, the project relies on grant funding and institutional support from Skoltech.

Recent Developments

In recent years, the Uspex lab has expanded into machine-learning-accelerated predictions and high-throughput screening. The group has also released auxiliary tools like the "XtalOpt" package and maintains a dedicated website (uspex-team.org) with documentation, user forums, and preprints. The lab continues to collaborate with experimentalists to validate predictions, maintaining a strong presence in high-pressure physics and solid-state chemistry.