In 1965, a small team at Oak Ridge National Laboratory in Tennessee switched on a reactor that broke every rule of nuclear engineering. The Molten Salt Reactor Experiment (MSRE) used no solid fuel rods, no pressurized water, and no uranium pellets. Instead, it dissolved its fuel — uranium-233 bred from thorium — directly into a molten fluoride salt mixture heated to 650°C. It ran successfully for four years. Then the government killed it.

The MSRE was the brainchild of Alvin Weinberg, the director of Oak Ridge and one of the inventors of the pressurized water reactor (PWR) — the very design that dominates nuclear power today. Weinberg considered the PWR his greatest mistake. He knew its weaknesses intimately: the need for massive pressure vessels, the risk of hydrogen explosions, the meltdown potential that would later materialize at Three Mile Island, Chernobyl, and Fukushima. The molten salt reactor was his answer.

The advantages were remarkable even by modern standards:

• Walk-away safety. If power failed, a frozen salt plug at the bottom of the reactor melted, and the fuel drained by gravity into passively cooled tanks. No operator action required. No backup generators. No pumps.
• Atmospheric pressure operation. No massive steel pressure vessels. No risk of steam explosions.
• Thorium fuel cycle. Thorium is three to four times more abundant than uranium in Earth's crust. India alone sits on an estimated 963,000 tonnes of it.
• Waste reduction. The reactor could burn actinides — the long-lived components of nuclear waste — reducing storage timelines from hundreds of thousands of years to roughly 300.
• Online reprocessing. Gaseous fission products bubbled out continuously. The liquid fuel could be cleaned without shutting down.

So why did it die? Politics and institutional momentum. In 1972, the Atomic Energy Commission was backing Milton Shaw's liquid-metal fast breeder reactor (LMFBR) program, which promised to breed plutonium — useful for weapons stockpiles as well as energy. Weinberg openly questioned the safety of the breeder program and advocated for molten salt. He was fired in 1973. The MSRE was defunded. Oak Ridge's molten salt team was disbanded. The LMFBR program itself was later canceled in 1983, having consumed billions with no commercial reactor to show for it.

The technical challenges were real but not insurmountable. The fluoride salts corroded the Hastelloy-N alloy used in the reactor's piping, particularly via tellurium embrittlement at grain boundaries. Tritium management was difficult. Online fuel reprocessing at scale remained unproven. In 1969, these were hard problems.

In 2026, they are solvable problems. Modern nickel superalloys and silicon carbide composites offer dramatically better corrosion resistance than 1960s-era Hastelloy-N. Advanced computational fluid dynamics can model salt behavior with precision Weinberg's team could only dream of. Companies like Kairos Power (whose Hermes reactor in Oak Ridge received NRC construction approval in 2023), Terrestrial Energy, and Copenhagen Atomics are actively building on MSRE's legacy. China's SINAP completed a 2-MW experimental molten salt reactor test in Wuwei, Gansu Province in 2023, the first since MSRE.

The most bitter irony: Weinberg's reactor already solved problems we've spent sixty years failing to solve with conventional designs. Passive safety without engineered backup systems. A fuel cycle resistant to weapons proliferation. A path to burning existing nuclear waste as fuel. We had the proof of concept running for 1,530 days in the 1960s, then walked away because the wrong reactor had the right political backers.