Summary: USA Rare Earth, Pasqal and Riven Systems have announced a research partnership that will combine an autonomous laboratory, machine learning and quantum-computing experiments to look for improved rare-earth separation molecules. The work is promising, but it remains a development program rather than a proven commercial process.
Rare-earth separation is one of the most technically demanding links between a mixed mineral feedstock and the individual oxides needed for permanent-magnet production. A new partnership announced on September 17 aims to apply automated experiments and advanced computing to that chemistry.
According to the announcement, Riven Systems plans to run thousands of automated laboratory experiments and create data on how candidate extractants bind to different rare-earth elements. Pasqal would then compare quantum machine-learning models with classical computing models, while USA Rare Earth would assess promising candidates against its intended feedstocks and processing flowsheets.
Why separation chemistry matters to magnet supply
Mining a rare-earth-bearing deposit does not by itself create magnet-ready material. Mixed rare-earth products still have to be separated into individual oxides, converted into metals and alloys, and then processed into magnet material under tightly controlled specifications. Dysprosium and terbium are especially important for some high-temperature NdFeB grades, while neodymium and praseodymium are central to the main magnetic phase.
The partners say their discovery workflow could help identify extractants that reduce the number of processing stages, equipment requirements and chemical consumption. Those are objectives, not verified results. The announcement does not report a qualified extractant, commercial recovery rate, plant design, production schedule or customer approval.
What B2B magnet buyers should watch
For procurement and engineering teams, the relevant milestones will come after the initial research stage. Useful evidence would include repeatable selectivity data, recovery and purity results, performance across different feedstocks, pilot-scale validation, operating-cost estimates and a realistic route from laboratory testing to continuous production.
Until those results are available, the project should be viewed as an attempt to accelerate process development rather than as new supply entering the market. Buyers evaluating future rare-earth or magnet sources should continue to ask where separation, metalmaking, alloy production, magnet manufacturing and product qualification will take place, and which steps are already operating at commercial scale.