Quantum Computers Just Cracked a Key Fusion Fuel Puzzle

A team of scientists from Oak Ridge National Laboratory, Cleveland Clinic and IBM has used a quantum computer to model one of the trickiest chemistry problems standing between humanity and abundant fusion energy — the first-known calculation of its kind.

In a paper posted to arXiv this week, the researchers described how they mapped out nine molecular configurations of FLiBe, a molten salt made of fluorine, lithium and beryllium that is a leading candidate for producing tritium fuel inside future fusion reactors. Getting that chemistry right is central to the U.S. Department of Energy's Genesis Mission, a national push to unify high-performance computing, artificial intelligence and quantum computing to accelerate scientific discovery.

Fusion reactors run on a mix of deuterium and tritium — a heavier form of hydrogen that is extremely rare in nature. Ensuring a steady supply of tritium is one of the biggest engineering hurdles on the road to commercial fusion power. Most reactor designs plan to "breed" their own tritium inside a blanket of lithium-rich material that lines the reactor wall and absorbs neutrons flying off the fusion reaction. FLiBe is one of the strongest candidates for that blanket because it circulates as a liquid at high temperature, transports heat efficiently and contains lithium that can be converted into tritium.

There is a catch. FLiBe's composition shifts dynamically under intense neutron bombardment, extreme heat and powerful magnetic fields, and its behavior is governed by quantum-level interactions that classical computers struggle to simulate at scale. Traditional supercomputer methods can approximate the physics but often lack the accuracy needed to design better formulations. Physical experimentation is possible but slow and expensive.

That is why the team turned to quantum-centric supercomputing — pairing IBM's quantum processors with the classical exascale machines at Oak Ridge and Cleveland Clinic. The same techniques were recently used by the collaboration to model proteins spanning more than 12,000 atoms, another problem that had long resisted classical computing alone.

"In order to demonstrate the capabilities catalyzed by the Genesis Mission, we have built a team of leading experts across seven DOE national labs, four universities, three industry partners and Cleveland Clinic to pursue a multi-pronged discovery cycle aimed at optimizing tritium production in molten salt fusion blanket materials," said Tom Beck, section head for science engagement in the Computing and Computational Sciences Directorate at ORNL. "Quantum computers, such as those built by IBM and enhanced by AI and exascale computing, are key tools that accelerate the discovery and design cycles needed to produce sufficient tritium to fuel fusion reactors."

The nine molecular configurations the team calculated capture different arrangements of FLiBe clusters — combinations of ions that determine how the salt will behave inside a reactor. Understanding those configurations is a fundamental building block for figuring out how to optimize the recipe for tritium breeding, thermal management and long-term stability under real reactor conditions.

"This work builds on our advances in simulating complex biological systems at scale, including proteins spanning 12,635 atoms, and extends those techniques into materials science to explore fusion-relevant systems with greater accuracy and efficiency," said Kenneth Merz, a staff scientist at Cleveland Clinic and the paper's corresponding author. "By bringing these technologies together, researchers can provide solutions to challenging real-world problems with greater speed and precision."

Jerry Chow, IBM's chief technology officer for quantum-centric supercomputing, framed the milestone as further evidence that quantum computers are graduating from lab curiosities to genuine scientific instruments. "Bringing quantum, AI and classical computing together is essential to tackling our society's most fundamental scientific challenges — unlocking capabilities which none of these paradigms can access alone," he said. "These results add to mounting evidence that quantum-centric supercomputing is now a practical scientific tool for problems that have long challenged chemists, engineers and materials scientists."

The results are early. Nine configurations is a starting point, not a finished blueprint. But the calculation represents a specific, concrete example of quantum hardware being brought to bear on a problem with real-world consequences — in this case, taking a serious swing at one of the longest-running roadblocks to clean, abundant fusion energy.

As quantum systems continue to scale in qubit count and reliability, the collaboration expects to run larger and more accurate simulations of not only FLiBe but a wider family of candidate reactor materials. That, the team argues, is exactly the kind of ambitious, boundary-pushing science the Genesis Mission was set up to enable.