AI Instead of Rare Earths: Munich Start-up Seeks Magnets from Common Elements
Munich, July 10, 2026 – alqem, a Munich-based materials discovery company, is pioneering the use of artificial intelligence (AI) to identify and bring to market high-performance magnets that do not rely on rare-earth elements. This strategic move aims to address critical supply chain vulnerabilities for materials essential to the energy transition.
The company’s research operations are primarily located in Portugal, while its headquarters remain in Munich. The founding team brings together a diverse range of expertise, combining strategic and supply chain knowledge with computational materials science and experimental physics.
Addressing a Critical Supply Chain Challenge
The global energy transition heavily depends on permanent magnets, which are integral components in electric vehicles, wind turbines, robotics, and industrial drives. Currently, the strongest magnets require rare-earth elements, with dysprosium and terbium being crucial for high-performance applications at operational temperatures. However, the supply chain for these elements is highly concentrated, leading to geopolitical and economic vulnerabilities, as evidenced by recent export restrictions.
alqem’s core mission is to use AI to discover magnetic compounds made from abundant and easily accessible elements. The goal is to achieve or surpass the performance of existing rare-earth magnets and, crucially, to predict their synthesis methods. This approach is also being applied to thermoelectric materials, which convert waste heat into usable energy.
“The big vision is clear: a material base for the energy transition that Europe can actually control itself – discovered and commercialized in years instead of decades,” stated Hanh Nguyen, Co-Founder & CEO of alqem.
The Genesis of an Innovative Idea
The concept for alqem originated from an academic collaboration between Dr. Tiago Cerqueira, alqem’s CTO, and Prof. Milan Allan, CSO. Their work in SuperC, an international research consortium focused on finding a room-temperature superconductor, highlighted the power of integrating theory, simulation, synthesis, and characterization in material discovery. They recognized that physically informed AI could be a potent tool for exploring uncharted material territories and that this methodology could be applied to various material classes beyond superconductors.
Hanh Nguyen’s fifteen years of experience in energy and chemistry underscored the limitations imposed by existing materials. “If you experience that often enough, you begin to see the opposite: how much technological progress is quietly held back by material boundaries and how much could be unleashed if those boundaries fell away,” she explained. Rare-earth-free magnets with superior performance represent such a potential breakthrough.
Development Process and Achievements
alqem’s platform development involved three stages: mapping candidate structures, predicting relevant properties (such as magnetization and anisotropy for magnets), and forecasting synthesis methods. A significant challenge was the scarcity of real-world data for training the AI model. To overcome this, alqem built ab-initio training datasets and utilized its synthesis capabilities to validate training data across a broad spectrum of chemical systems.
A notable success has been the accuracy achieved from prediction to experiment, with predicted values falling within 15% of experimental results. Another challenge involves synthesizing predicted materials in the lab, particularly for novel systems. To address this, alqem is developing an LLM-scientist to assist in formulating informed hypotheses and learning from experimental outcomes.
The company also boasts significant achievements, including attracting top-tier advisors, establishing robust validation capabilities, identifying promising candidate materials that could outperform current NdFeB and SmCo magnets, and securing strong interest from industrial partners facing immediate supply chain pressures.
Market Reception and Future Outlook
The market and industry response has been highly positive. The need for secure supply chains has become a boardroom priority in automotive and industrial manufacturing, making alqem’s solution particularly timely. There is strong interest from end-users to test new materials once they are validated in alqem’s lab, with plans for collaborative development with selected partners.
On the investor side, alqem has attracted deep-tech specialists who understand the unique development trajectory of research-intensive companies compared to software ventures.
The alqem team emphasizes three key takeaways from their journey: AI functions most effectively as an amplifier of deep domain expertise, not a replacement; choosing problems where combined backgrounds offer an advantage is crucial; and honesty in storytelling resonates with investors.
alqem’s rapid progress and innovative approach position it as a key player in developing sustainable material solutions for the energy transition, aiming to create a material base that Europe can control independently.
Source: https://www.chemie.de/news/1189188/ki-statt-seltene-erden-muenchner-start-up-sucht-magnete-aus-gewoehnlichen-elementen.html
