Josephson junction

The Josephson junction emerged not from a laboratory but from a graduate seminar. In 1962, a 22-year-old Cambridge student named Brian Josephson sat in a course taught by Philip Anderson, a Bell Labs physicist on sabbatical. Anderson had been explaining "broken symmetry" in superconductors—the idea that below a critical temperature, electrons pair up and march in quantum lockstep. Josephson became obsessed with what would happen if two superconductors were separated by an almost impossibly thin barrier.

The adjacent possible had aligned. BCS theory, published in 1957, finally explained why superconductors worked—Cooper pairs of electrons sharing a quantum phase. Ivar Giaever at General Electric had shown in 1960 that individual electrons could tunnel through nanometer-thin oxide barriers between superconductors. But Josephson saw further: if the barrier was thin enough—just ten atoms or so—entire Cooper pairs might tunnel while preserving their quantum phase coherence. Current would flow without voltage.

Josephson published his prediction in July 1962. John Bardeen, the only person to win two Nobel Prizes in Physics, publicly doubted him. But at Bell Labs, Anderson remembered those Cambridge discussions. He told colleague John Rowell about Josephson's theory. On January 21, 1963, Rowell saw the effect in a tin-lead junction—supercurrent flowing through an insulating barrier exactly as predicted. The graduate student had been right; the two-time Nobel laureate had been wrong.

The cascade was immediate. By 1964, researchers at Ford's laboratory had built the first SQUID magnetometer, capable of detecting magnetic fields a trillion times weaker than Earth's. These devices now map the brain's magnetic whispers in magnetoencephalography, finding epileptic foci before surgery. By 1990, every volt measured anywhere on Earth traced back to Josephson junctions—arrays of thousands of them replaced electrochemical cells as the international voltage standard, achieving precision of one part in ten billion.

IBM poured resources into Josephson computing from the 1960s through 1983, betting superconducting circuits would outpace semiconductors. They lost that bet to Moore's Law, but the technology persists. HYPRES, founded from IBM's ashes, builds circuits switching at 120 GHz. And in 1999, the junction found its most profound application: the superconducting qubit. Every major quantum computer today—from Google's Sycamore to IBM's Eagle—runs on Josephson junctions, the same quantum effect a graduate student predicted while his professor lectured about broken symmetry.