When we decided to feature the first patent for a superconducting magnet in this issue’s logbook, we had no idea how difficult it would be to find it. The earliest weak superconducting magnets weren’t patented, since no one saw any practical application for them. It was only as superconducting materials matured that companies began to seek patents for superconducting magnet configurations, in a competitive boom of invention in the early 1960s. Recognition of the first patent seems to have gotten lost in the scuffle.

The first proof of concept for a superconducting magnet came from Heike Kamerlingh Onnes. Kamerlingh Onnes discovered superconductivity in 1911 when he cooled mercury to near absolute zero and found that its electrical resistance disappeared. He immediately saw the potential of this finding for generating high magnetic fields using little electrical power. But when he tested the concept of a superconducting magnet in 1913 by winding superconducting lead wire into a coil and running current through it, he found that magnetic fields above 500 gauss quenched the superconductivity.

Onnes concluded that superconducting magnets weren’t practical, and the field languished for four decades. During this time, superconducting materials with higher transition temperatures were discovered, including cold-worked niobium with a transition temperature around 9 K and an ability to sustain high fields without quenching. In 1954 George Yntema at the University of Illinois used niobium to make the first successful superconducting magnet.

“I saw no reason why a magnet could not be made with superconducting windings, so I gave it a try,” Yntema recalled in a 1987 article. He achieved 7.1 kilogauss with cold-worked niobium wound around an iron core. He reported his findings in a tiny abstract buried at the back of the May 1955 Physical Review Letters (page 1197), part of the minutes of the annual meeting of the American Physical Society.

John Hulm at Westinghouse saw the paper, “and was inspired to improve on my design. He is the only person I know of who noticed that abstract,” Yntema wrote. Yntema credits Hulm with making the second superconducting magnet in 1955. A niobium solenoid without an iron core, it achieved 6 kilogauss.

The next published paper, however, came in April 1960 from Stanley Autler of Lincoln Laboratory at M.I.T. In 1959 he independently made iron-cored niobium magnets that achieved 14 kilogauss. Autler’s magnets were the first to have a practical use, providing the magnetic field for a solid state maser.

“Things began to look interesting at this point for superconducting magnets,” as Autler said in a 1964 talk.

By that time, other groups were making niobium magnets as well, mostly for demagnetization refrigerators-Stanford University, the Bureau of Standards laboratory in Boulder, and General Electric. But no patents were filed. There was a general tendency among scientists of the time to publish papers but not to patent, according to a 2006 study. Also, there was no market demand for superconducting materials.

Autler’s paper generated attention, however. In early 1960, John Eugene Kunzler at Bell Labs began working on superconducting magnets for maser applications. His team wound a solenoid of molybdenum-rhenium wire that achieved 15 kilogauss. Kunzler filed his patent “Superconducting Magnet Configuration” on September 19, 1960, 15 days before Autler filed his patent for a superconducting magnet.

In December 1960, Kunzler’s group discovered that a compound of niobium and tin could sustain magnetic fields up to 100 kilogauss without quenching. High-field superconducting magnets were now a possibility, and the gauss race was on.

Kunzler’s group concentrated on finding a way to wind the brittle niobium-tin compound into a wire. As incentive, Kunzler’s boss Morris Tanenbaum promised Kunzler a bottle of scotch for every 3 kilogauss he achieved above 25, Kunzler recalled in a 1987 article. With this motivation, Kunzler and colleague Bernd Matthias, following a brainstorm from technician Ernie Buehler, found a way to pack niobium and tin powders inside a hollow tube of niobium, draw the tube into a wire, wind the wire into a coil, and then bake the coil at 1000 °C to chemically react the niobium and tin into a compound. With this configuration, Kunzler’s team expected to achieve the first 70 kilogauss superconducting magnet.

Meanwhile, Westinghouse, Atomics International, Lincoln Labs, and Oak Ridge Laboratories were also in the hunt for the first high-field superconducting magnet. The competition came to a head at a November 1961 conference at M.I.T. The sessions on superconducting magnets came near the end of the meeting, but the speaker’s abstracts were posted on a bulletin board when the meeting started, most reporting achievements of around 50 kilogauss. Many of the speakers made frequent phone calls back to their laboratories, revising their estimates upward as the meeting progressed, Kunzler wrote, an event which became known as “gaussian poker.”

Kunzler was not enjoying the race, he wrote, as his niobium-tin magnet was achieving less than 50 kilogauss and out of the running. The night before he presented, Buehler and Frank Hsu back at Bell Labs discovered a short in the magnet and repaired it. The magnet produced 68 kilogauss, and Kunzler’s group edged out Hulm from Westinghouse and Richard Hake and Ted Berlincourt from Atomics International, who each reported 60 kilogauss with niobium-zirconium windings.

Kunzler’s group later achieved over 100 kilogauss and earned two cases of scotch, before Bell Labs abandoned the magnet program for lack of Bell Labs applications. For his part, John Hulm went on to discover the most important superconducting alloy, niobium-titanium. In the 1970s, the building of the Tevatron accelerator by Fermilab provided an application for superconducting wire and magnets. A mass-production industry sprang up that would supply the wire for the new medical technology of magnetic resonance imaging.

And the first patent? Although Bell Labs doesn’t claim the honor, the U.S. Patent Office database reveals it is Kunzler’s patent, issued on April 14, 1964, nearly four years after submission. Autler’s first patent was issued more than a year later, in August 1965. Oak Ridge Labs and Hake and Berlincourt received superconducting magnet patents in spring 1965, and Hulm in 1966. In the game of gaussian poker, Kunzler’s group had won.

by Madolyn Bowman Rogers