Insider Brief
- USA Rare Earth, Pasqal and Riven Systems have partnered to explore quantum machine learning for discovering molecules that could improve rare earth separation processes.
- Riven’s automated laboratory will generate experimental data on extractant selectivity, while Pasqal will benchmark quantum machine learning models against classical models using data from the lab.
- The project will focus on USA Rare Earth’s planned feedstocks, including material from the Round Top mine, mixed rare earth carbonates and recycled magnet-manufacturing swarf.
PRESS RELEASE — USA Rare Earth (Nasdaq: USAR), a rare earth, critical minerals and advanced materials company, Pasqal (Nasdaq: PSQL), a global leader in neutral-atom quantum computing and Riven Systems, a specialist in AI for industrial chemistry, today announced a strategic partnership to develop next-generation separation technology for the rare earth value chain. Using quantum machine learning, the project aims to identify new separation molecules that bind more effectively to rare earths than existing alternatives, enabling smaller, lower-cost, less energy-intensive processing facilities.
The collaboration brings together U.S. and French expertise in two strategically important sectors: critical minerals and quantum computing. Together, the companies aim to explore technologies that could help support a more secure, efficient and competitive Western rare earth supply chain.
The partnership brings together Riven’s self-driving minerals separation laboratory and Pasqal’s quantum computing power with USA Rare Earth’s rare earth processing expertise. Under the planned project, Riven would conduct thousands of automated experiments and generate the training data needed to build machine learning models of extractant selectivity for rare earth elements. Pasqal’s Neutral Atom QPU would then benchmark quantum machine learning models against classical computing-based models (each trained on data from the self-driving lab) to support USA Rare Earth in optimizing extractant selection.
“The key challenge the rare earth industry outside Asia faces is to separate the Mixed Rare Earth Carbonate (MREC) produced in upstream operations into individual, separated oxides. The Chinese dominate this capability, especially for vital heavy rare earths such as dysprosium, terbium and yttrium. Our goal is to accelerate the path to efficient, competitive and sustainable solutions by changing how the rare earth industry discovers the chemistry it runs on, replacing years of trial-and-error,” remarked Alex Moyes, Senior Vice President of Upstream (USA), USA Rare Earth. “This strategic partnership will institute an innovative discovery pipeline where new extractant molecules are identified and tested virtually, powered by a quantum machine learning model trained on real chemical data from an autonomous lab. Rapid discovery of an extractant built and optimized for USA Rare Earth’s flowsheets could improve our capital and operating efficiency by reducing stage count, equipment needs and raw material consumption, while lowering the carbon footprint and environmental impact of our facilities. Breakthroughs like these are what could help make a Western rare earth value chain possible and accelerate our journey to being competitive on the world stage.”
Pasqal and Riven will tailor their work to USA Rare Earth’s feedstocks, expected to include the Round Top mine in Sierra Blanca, Texas, as well as third-party mixed rare earth carbonates (MREC) and recycled cuttings (known as swarf) from the magnet manufacturing process. Targeting these key feedstocks means the experimental work can potentially impact all aspects of USA Rare Earth’s materials processing work.
“This partnership brings together three companies working at the forefront of technologies that are increasingly important to economic growth, industrial competitiveness and national resilience,” said Wasiq Bokhari, Chief Executive Officer of Pasqal. “Rare earth materials are essential and improving how they are processed has implications far beyond a single industry. By combining Pasqal‘s quantum computing capabilities with USA Rare Earth’s materials expertise and Riven’s advanced automation platform, we have an opportunity to accelerate innovation while demonstrating how French and American companies can work together to strengthen critical supply chains.”
USA Rare Earth, Pasqal, and Riven believe this initial machine learning project could create opportunities for a deeper, extended collaboration. Over time, the companies envision a first-of-a-kind, end-to-end extractant discovery process: a quantum machine learning model identifies high-potential extractants, those extractants are tested in Riven’s self-driving lab, and top candidates are validated in USA Rare Earth’s R&D facility in Wheat Ridge, Colorado before being integrated in USA Rare Earth’s flowsheets.
“AI and autonomous labs are the next frontier in critical mineral processing,” said Dr. Orion Archer Cohen, CTO and co-founder of Riven Systems. “We applaud USA Rare Earth’s proactive adoption of these technologies, which forge a path to significant improvements over legacy chemistry. This is exactly how the West can leverage its lead in advanced compute technologies to reindustrialize faster.”
Facts Only
* USA Rare Earth, Pasqal, and Riven Systems partnered on exploring quantum machine learning for molecule discovery.
* Riven’s automated laboratory will generate experimental data on extractant selectivity.
* Pasqal will benchmark quantum machine learning models against classical models using the lab data.
* The project targets USA Rare Earth feedstocks, including material from the Round Top mine, MREC, and recycled magnet-manufacturing swarf.
* The objective is to identify new separation molecules that bind more effectively to rare earths than existing alternatives.
* Riven will conduct thousands of automated experiments and generate training data for machine learning models of extractant selectivity.
* Pasqal’s Neutral Atom QPU will benchmark quantum machine learning models against classical computing-based models.
* The partnership aims to explore technologies supporting a more secure, efficient, and competitive Western rare earth supply chain.
* Planned future integration involves a process where ML models identify extractants, the lab tests them, and USA Rare Earth validates candidates before flowsheet integration.
Executive Summary
USA Rare Earth, Pasqal, and Riven Systems formed a partnership to explore quantum machine learning for discovering molecules that improve rare earth separation processes. The project involves using the collaboration to identify new separation molecules that bind more effectively to rare earths than current alternatives, aiming to enable smaller, lower-cost, and less energy-intensive processing facilities. The collaboration merges U.S. expertise in critical minerals, French expertise in quantum computing, and specialized AI for industrial chemistry.
Riven will conduct automated experiments in a self-driving laboratory to generate experimental data on extractant selectivity. Pasqal will use its Neutral Atom QPU to benchmark quantum machine learning models against classical models using the lab data. The project will focus on USA Rare Earth feedstocks, including material from the Round Top mine, mixed rare earth carbonates (MREC), and recycled magnet-manufacturing swarf.
The overall goal is to accelerate the path to more efficient and sustainable solutions for the rare earth value chain by discovering optimized extractants through a virtual testing pipeline. The collaborators envision a future end-to-end process where quantum machine learning models identify candidates, Riven tests them in an automated lab, and USA Rare Earth validates the top candidates before integration into processing flowsheets.
Full Take
The narrative frames technological convergence—quantum computing, autonomous labs, and industrial chemistry AI—as a direct pathway to geopolitical and economic security within the critical minerals sector. The motivation explicitly targets a historical inefficiency: the reliance on trial-and-error chemistry for separating mixed rare earth carbonates (MREC). This shift implies that current methods are inherently suboptimal, positioning the partnership not just as an optimization exercise but as a necessary disruptive intervention.
The structure presents a clear progression from data generation (Riven) to theoretical exploration (Pasqal/QML) to practical application and integration (USA Rare Earth R&D). This layered approach suggests a potential point of leverage: leveraging advanced computation to fundamentally redefine chemical discovery, rather than simply optimizing existing industrial processes using classical ML. The focus on Western supply chain resilience acts as a powerful external driver, lending urgency to the technical exploration.
The implied pattern here is an appeal to technological determinism where quantum/AI solutions are presented as the inevitable means to solve geopolitical resource competition. The underlying assumption is that solving the chemical separation problem inherently solves the security problem. The critical tension lies in whether the complexity of implementing this novel pipeline—integrating autonomous physical experimentation with abstract quantum modeling—is sufficiently addressed, or if the focus remains primarily on showcasing the potential rather than mapping the practical engineering and operational friction points required for true end-to-end industrial deployment. What specific bottlenecks in integrating real-world chemical noise into quantum learning models are being addressed by this framework? What metrics will validate that a "virtually discovered" molecule translates reliably into tangible, lower-cost physical separation at scale?