Researchers from Oak Ridge National Laboratory (ORNL), Cleveland Clinic, and IBM have used quantum computers to calculate nine molecular configurations of FLiBe, a molten salt material widely considered a leading candidate for tritium breeding in future fusion reactors. The work, published on arXiv, marks the first time these specific molecular calculations have been performed using quantum computing.
The research tackles one of fusion energy’s most significant technical challenges: producing and recovering enough tritium to sustain reactor operations. Tritium is a critical fuel component for most proposed fusion reactor designs but is extremely scarce in nature. Improving tritium breeding efficiency is a central objective of the U.S. Department of Energy’s Genesis Mission.
FLiBe, a mixture of lithium fluoride and beryllium fluoride, serves multiple roles in many fusion reactor concepts, including tritium breeding and heat transfer. Predicting its behavior is computationally demanding because its chemical composition evolves under neutron irradiation, high temperatures, and strong magnetic fields. Accurately modeling these interactions at the quantum level is beyond the practical reach of many conventional computational approaches and often requires costly experimental validation.
To address this challenge, the research team combined quantum processors with traditional high-performance computing (HPC) resources in a quantum-centric supercomputing workflow. Rather than relying solely on classical approximations, quantum processors were used for portions of the electronic-structure calculations, while CPUs and GPUs handled the remaining computational tasks. This hybrid approach enabled the team to calculate the energetics of multiple FLiBe molecular configurations, with and without tritium, providing deeper insight into atomic interactions and tritium-binding mechanisms.
The project also builds on computational methods previously developed for large-scale biological simulations. The same quantum-enabled workflow has already been applied to protein systems spanning 12,635 atoms, and researchers have now adapted those techniques to materials science applications relevant to fusion energy.
The research team assembled for this effort under the Genesis Mission spans seven DOE national laboratories, four universities, three industry partners, and Cleveland Clinic. The broader mission aims to unify high-performance computing, AI, and quantum computing across the DOE’s 17 national laboratories to accelerate scientific discovery, and the team says integrating these technologies will speed the discovery and optimization of fusion blanket materials that can produce sufficient tritium for commercial reactors.
IBM’s contribution centers on advancing quantum-centric supercomputing, which unifies CPUs, GPUs, and quantum processing units (QPUs) within a single computational workflow. The company views hybrid quantum-classical computing as an emerging platform for tackling scientific problems in chemistry, materials science, and engineering that remain difficult for classical systems alone.
Looking ahead, the collaborators plan to reduce communication overhead between quantum and classical computing resources while increasing the size and complexity of molecular systems that can be simulated. The long-term goal is to make quantum-centric simulation a practical tool for fusion materials design and validation workflows.
The FLiBe research joins several scientific milestones demonstrated on IBM quantum systems during 2026, including simulations of magnetic materials, molecular chemistry studies, and large-scale biological modeling, further expanding the role of quantum computing as a tool for scientific research.
Amazon