An accidental observation in a German tokamak revealed a new regime of plasma confinement — doubling performance overnight and reshaping the design of every major fusion device since.
By the early 1980s, tokamak physicists had made enormous progress in heating plasmas to fusion-relevant temperatures, but a stubborn problem remained: energy leaked out of the plasma faster than theoretical models predicted. This “anomalous transport” meant that achieving ignition would require ever-larger, more expensive machines. The standard operating regime, later dubbed L-mode (low confinement mode), seemed to impose a hard ceiling on performance. Researchers at laboratories worldwide searched for ways to reduce transport, but no one expected the solution to arrive by accident.1
On February 4, 1982, Friedrich Wagner and his colleagues at the Max Planck Institute for Plasma Physics in Garching, Germany, were running experiments on the ASDEX tokamak (Axially Symmetric Divertor Experiment). During a neutral beam heating shot, the plasma abruptly transitioned into a new state. The hydrogen-alpha emission — a measure of particle losses at the plasma edge — dropped sharply, and energy confinement roughly doubled. Wagner recognized that something fundamental had changed: the plasma had spontaneously formed a transport barrier at its edge, creating a steep pressure gradient that insulated the hot core from the cooler boundary.2
The new regime was designated H-mode (high confinement mode). Over the following years, experiments at ASDEX and then at tokamaks around the world — DIII-D in San Diego, JET in England, JT-60 in Japan — confirmed and extended the discovery. The L-H transition was found to require a minimum heating power that depended on plasma density, magnetic field strength, and machine size. The edge transport barrier was associated with the suppression of turbulence by sheared plasma flows, though a complete first-principles theory remains elusive to this day.3
H-mode came with a complication. The steep edge pressure gradient periodically collapsed in bursts called edge-localized modes, or ELMs — violent instabilities that expelled energy and particles onto the tokamak walls. In large machines, unmitigated ELMs could damage plasma-facing components. Managing ELMs became one of the central challenges of fusion engineering, driving the development of techniques such as resonant magnetic perturbations, pellet pacing, and the search for naturally ELM-free regimes like QH-mode and I-mode.4
Friedrich Wagner received the Hannes Alfvén Prize in 2007 for the H-mode discovery. His serendipitous observation in 1982 remains one of the most consequential experimental findings in the history of fusion research, a reminder that breakthrough physics sometimes announces itself without warning on an ordinary Tuesday in a German laboratory.5