Scientist Recounts Superconductivity Discovery

Ashburn's path also included a love of science, especially mathematics.

All of those factors were in play 25 years ago when Ashburn, then a first year physics graduate student at The University of Alabama in Huntsville, created the formula for the world's first “high temperature” superconductor.

The material he formulated from yttrium, barium, copper and oxygen was the first to become a superconductor at a temperature warmer than liquid nitrogen, a scientific milestone akin to running the four-minute mile or breaking the sound barrier. (A superconductor lets electricity flow through without any resistance.)

His role in creating the “123” superconductor earned him a seat (in the audience) at the circus-like announcement at a scientific meeting of the UAHuntsville team's success, known now as the “Woodstock of physics,” and a place on one of the most-cited scientific papers in the physics literature. His superconductor was also the subject of Ashburn's doctoral dissertation. It did not, however, earn him fortune or lasting fame.

“Not too many people have recognized my name in recent years,” Ashburn said recently. “It has all been such a bizarre experience. Personally, it was a good experience for me in terms of growing up fast.

“My dad likes to tell the story that I did it (created the formula) on Super Bowl Sunday, but that's not quite right.”

Instead, it happened a week earlier, when football played its unlikely part in Ashburn's story.

A scientific 'sprint'
The first superconductor was pure mercury. In 1910, scientists found that mercury becomes superconductive at 4 Kelvin (four degrees above absolute zero and about 452 degrees below zero Fahrenheit).

For the next 76 years, scientists worked to find superconductors at higher and higher temperatures with slow and intermittent success. Then, in mid-December 1986, researchers at IBM's laboratories in Zurich, Switzerland, discovered a material (a combination of lanthanum, strontium, copper and oxygen) that is a superconductor at about 30 Kelvin — about 405° below zero Fahrenheit.

That discovery sparked renewed interest in laboratories around the world, including Dr. M.K. Wu's physics lab at UAHuntsville. Wu, Ashburn and graduate student C.J. Torng were working on a NASA research project, studying possible metal alloys to be processed in space. After NASA agreed to let them change their research to the new superconductors, things moved quickly: In the scientific equivalent of a sprint, UAH's superconductor research took only six weeks from starting in December 1986 to the discovery in late January 1987.

The UAHuntsville team understood that the competition was fierce. Key players included research groups at IBM and AT&T's Bell Labs. (“I was scared to say anything over the phone,” Ashburn recalls. “I was being really paranoid.”)

“We knew it was a race to find something, but we didn't know what,” he said. “I was eating one meal a day, supper, about 8 p.m. We were working 12 hours a day, seven days a week. The time stamps on some of our experiments were at 1 a.m. and 2 a.m.

“On Dec. 24 we were still testing. We did four tests that day. There were no tests on Christmas, but on the 26th we were back in the lab.”

The difficult thing about that kind of intensive activity, however, is that it can't be sustained. By Saturday, Jan. 17, Wu was ready to get back to NASA's alloy research, telling his team to stop testing new samples. But Ashburn had some ideas of his own and he had some outside coaching, starting with a buddy from high school.

The thermite team
Daniel Schultz was Ashburn's classmate at Huntsville High School: “He was one of my pyromaniac friends. He introduced me to playing in the lab. We made gunpowder.

Our senior year in high school we were required to do a project and we convinced the teacher to let us make thermite.”

Thermite is a powerful incendiary. The burning compound was so hot it broke the ceramic crucible intended to contain it and sent a cloud of smoke billowing out into the lab before they could close the fume hood.

“I was thinking to myself, 'We're going to get an F for that,'” he recalled. (They got an A, and Ashburn was his class valedictorian.)

After high school, while Ashburn was earning an undergraduate degree in physics at UAH, Schultz was at Georgia Tech studying ceramic engineering. The superconductors that Ashburn, Wu and Torng were studying are ceramics. By the fall of 1985, Schultz was at UAH working on a graduate degree, when he wasn't counseling his friend on the intricacies of ceramics.

“Jim and I had a lot of discussions about crystal structures,” Schultz wrote in an account of his involvement written for Ashburn in November, 2011. “Jim was really thinking about the crystalline structures and potential 'formulas' for these new type of compounds.”

“One day Jim came into his lab and he showed me a hand-drawn table of element sizes, volumetric proportions and the resultant transition temperatures of those compounds,” Schultz wrote. “He said that over that night he came to the conclusion that yttrium … was the next element in the series to make a higher transition temperature compound. I thought his argument made a lot of sense and I was really impressed that he had reduced it down to a table based on his own formula for the superconducting compound structure.”

“He gave me copy of The Ceramics Bible and introduced me to yttrium,” Ashburn said. “He helped me build a ball mill that we used to grind the powders of the compounds we were trying to mix.”

More efficient than a mortar and pestle, the ball mill helped Ashburn and Torng get the extremely fine, uniform particles they needed for the compounds they were mixing.

Another coach on Ashburn's unofficial team was Jones Hamilton, also a physics graduate student and “yttrium's biggest cheerleader.”

“He deserves a lot of the credit,” Ashburn said. “Yttrium is an element that doesn't get a lot of respect and he was yttrium's chief proponent. If he hadn't been pushing, I might not have tried it.”

1986 technology
Earlier Ashburn automated Wu's lab, connecting the temperature probes to a new Hyundai Blue Chip personal computer (remember, it was 1986) that sported an impressive 512,000 characters of memory. Characters.

The new setup meant that instead of having to sit in the lab taking data for hours while a sample cooled in liquid helium, “I could actually go to the library while the measurements ran unattended, allowing me to make the most of my time,” he said. “I'm way too impatient and competitive to have been content with reading 40 points by hand when I could collect hundreds of points, each averaging 40 or so individual measurements.”

A day for football
With Wu calling a halt (at least temporarily) to superconductor work on Jan. 17, Ashburn decided it was time to take a day off. So the following day he went across town to visit his parents. He wanted to take some time off and, having missed the entire college bowl season while working in the lab, watch some football.

It was a perfectly normal thing for a young man raised in Alabama to do. There was only one problem: On Jan. 18, 1987, there was no football on TV. None. The NFL was taking a week off before the Giants took on the Broncos in the Super Bowl. College football was long since done. Even the East-West Shrine Game was the week before.

So much for watching football.

Disappointed, Ashburn decided he might as well get some work done. He had his own ideas about why the lanthanum strontium copper oxide (LSCO) material was a superconductor. (The exact mechanism that causes superconducting is still not known, although there are two leading competing theories.) He thought the size of the ceramic crystals might play a role, so he was interested in pursuing what he called “volume matching.”

“I was learning that you have to put in things that were the right size with the right charge,” he explained. “It was basic crystal chemistry.”

Wu wanted to try one-to-one substitutions of chemicals in the LSCO compound using elements from the same families and groups in the periodic table. They had tried replacing copper with aluminum, silver and zinc, all without success. (It turns out that lanthanum strontium zinc oxide is an excellent insulator.)

“He wanted me to make straight substitutions,” Ashburn said. With straight substitution, one atom of lanthanum in the LSCO compound, for instance, would be replaced in the formula by going one row up or down in the periodic table and using one atom of scandium or one of yttrium. “He wanted me to be more systematic. But I thought, 'If that was a superconductor, it would already have been discovered.'

“I was not going to test the obvious things. I wasn't going to waste my time on that.”

Working at his parent's home, without the resources of either the lab or the university’s library, Ashburn set to work.

That's when serendipity took a hand.

Working at home, the only reference book available was an “Introduction to Solid State Physics” textbook he was using in class that semester. On page 76 (“I still remember the page.”), it shows the atomic radii for lanthanum as 1.15 Angstroms and strontium at 1.13.

“I was concerned about yttrium being too small,” he said. “I knew the crystals we needed couldn't exist with a lot of yttrium in them. So I went back to barium (which has nice, big atoms) to match the volume, and I put in a lot more four times as much to compensate.”

He found later that “only with all three of those changes do you get close enough to get a superconductor.”

He took that formula and three others concocted that day with him to the lab on Monday, where Wu's team continued working on superconductors. He and Ashburn both had new formulae to test, including one of Ashburn's that coupled copper oxide with mercury oxide.

In the meantime, Wu had secured a two-kilogram can of yttrium oxide from a ceramics lab at NASA's Marshall Space Flight Center. He took it to his UAHuntsville lab on Jan. 23, but it sat unopened on a shelf.

“It said it was 99.999 percent pure,” Ashburn said. “I was afraid to open it.”

Instead, they set to work preparing some of the other new compounds, including Ashburn's mercury mix. They tested it on Jan. 26, predicting that it might be a 40 Kelvin superconductor.

The mercury compound worked, only the fourth “high temperature” copper oxide superconductor found to that point.

“We got a partial transition (a reduced resistance) at 40 Kelvin,” Ashburn said, although he discovered later that the important mercury milestone happened for the wrong reason: The mercury had evaporated before the compound was tested.

“It was the fact that there were vacancies in the crystal instead of mercury that made that work,” he said. “But it was the ideal motivation when we needed some encouragement. It was the first new thing that had shown superconducting properties for us.

“Running the mercury sample and getting that first success, that's when I got the courage to get the yttrium off the shelf and open the can.”

It took a couple of days to process the chemicals and prepare the compounds for firing. Fortunately, the oven Wu's group had been using to bake their superconductors was not available when the yttrium barium copper oxide (YBCO) compound was ready to cook. The only oven available heated only to 1,000 degrees Fahrenheit, instead of the 1,500 degrees that were normal for preparing ceramics of that kind.

That was fortunate, because the YBCO compound doesn't do well at extreme temperatures, as Ashburn would learn later. “These are not friendly materials,” he said. “If you overheat it, that's a bad thing. When it gets too hot it melts and separates into two things, so when it solidifies it turns into two different compounds.”

Two compounds went into the oven, Ashburn's and one of Wu's, and “when they came out Wu's looked better,” Ashburn said. “His was black and shiny, while mine was dull and green. Things that are green and dull are not typically going to be a good conductor.

“It was green, but there were flecks of black in it.”

On Jan. 29, the samples were ready to test.

“We tested in (liquid) helium that day,” Ashburn said. Electrical resistance through his YBCO compound dropped to zero at 95 Kelvin, about 288 degrees below zero Fahrenheit. That was a 50 Kelvin (90 degrees Fahrenheit) improvement over what had been the best superconductor at the time, such a major leap that they had trouble accepting it.

“Our first reaction was that something was wrong with the temperature … As soon as we saw that transition in helium we ran down the hallway to a chemistry lab.”

Ashburn vividly remembers filling a container with liquid nitrogen, taking it back to the lab and running the test again. Successfully.

“That's when we knew,” he said. “I didn't have any doubt that we had a superconductor. It looked too right.”

The new superconductor was the first “high temperature” superconductor, the first to be a superconductor in liquid nitrogen. All previous superconductors functioned only in liquid helium or liquid hydrogen, both of which are dangerous and difficult to handle.

It was the wrong oven. The wrong temperature. The wrong atomic sizes, leading to the wrong formula for the wrong (but much better) arrangement of the atoms. But if any single one of those things goes “right,” there's no superconductor.

The formula for the compound is Y1Ba2Cu3O7, ergo the “123” superconductor.

“The neat thing about it was, it wasn't the crystal structure I was trying to make,” Ashburn said. “But it was close enough to get it in the right class of structures. When I missed, I missed in the direction that was what we needed. And it has the atomic weight of 666.”

Devil in the details
Things did not go smoothly after that. Wu insisted on taking the compound to his former graduate advisor in Texas for independent testing to confirm its superconductivity.

The problem was that days before the discovery Wu's former advisor filed an application for a patent on every possible superconducting combination of about one-third of the periodic table. That included the entire extended chemical families of the elements in superconductors known at that time.

With all the combinations, it adds up to more than half a billion possible materials. Start mixing them in different proportions (like H2SO4) and the possible combinations quickly soar into the billions, and every superconducting one of them (known, unknown and not yet imagined) is apparently covered by that patent.

It was not one of the U.S. Patent Office's prouder moments.

The contentious dispute over who created the formula, how, where and when, dragged on for years, frequently with Ashburn at the center of the maelstrom. Wu left UAH for Columbia. With a new advisor, Ashburn would spend the next three years studying the YBCO compound as the subject for his doctoral dissertation.

Super potential
Touted as a miracle material that would change the world, superconductors have had a long, difficult journey toward realizing their potential. The materials are not easy to manufacture or process into useful products.

The potential benefits include revolutionizing solid-state electronics and high-speed computers, optical detectors like those used on satellites, or large-scale applications such as levitating trains or superconducting ships' engines. Electric generators made with superconducting wires would be more efficient that conventional generators.

The first obvious potential use for superconductors is moving electricity around without the loss of power due to resistance: It is estimated that between 6 and 7 percent of the electricity generated in the U.S. is lost to resistance in the wires. Unfortunately, installing thousands of miles of superconducting wires and the support facilities needed to keep them at operating temperatures is prohibitively expensive.

In 2008, the world's first superconducting transmission power cable was commissioned at a substation on Long Island, NY. It carries the power for 300,000 homes when it is operating at full capacity.

Moving on
M.K. Wu is director of the Superconductor Laboratory at the Institute of Physics in Taiwan. C.J. Torng is vice president of engineering for Headway Technologies, a division of TDK, in California.

While the invention's fallout was stressful for him, Ashburn has taken up documenting and preserving data and materials related to the invention as a kind of hobby. He has several paper grocery bag-sized boxes of material in the hall of his comfortable Huntsville home, where he lives with his wife Greta, their five children and Thor, a black poodle who is afraid of thunder. Books, newspaper clippings, magazine articles and court depositions peak out at the edges, around bottles of failed test samples and 5.25-inch floppy disks.

(By 1999 he could find no one with a computer that would read the outdated disks. In 2003 he had to find help to recovering most of the data from them. “A lot of it was just bits and pieces,” he said. “We made two or three copies of everything and I was able to reassemble about 90 percent of it, including all of the original data from Jan. 29.”)

“People just send me stuff and I don't have the heart to throw it away,” he said. “I have all sorts of stuff from that time period. I have some mementos. I have several samples, mostly the stuff that didn't work. For some reason it seemed important to keep it all.

“I have all of the old data. I can pin the tests down to the minute and I can show the place it happened within a few feet, despite the fact that the building has been remodeled and the old lab is now a classroom.”

After receiving his doctoral degree, Ashburn left superconductors behind. Today, he is a staff scientist for Lynx Support Specialist, Inc. (LSSI), in Huntsville, Ala., supporting the Army.

He still enjoys football, but with his own analytical twist: His interests in football and mathematics are regularly exercised on the Atomic Football website (http://www.knology.net/jashburn/football/). His algorithm (or formula) for ranking football teams — and predicting winners — is an official part of the system the NCAA uses to select teams for the FCS national championship tournament.

“My main interest is designing algorithms,” he said. “I like to model things with math. That's what I did then. It's my job and it's my happy place.”