Ann Johnson, former associate professor, University of South Carolina (deceased)
Cooled to a few degrees above absolute zero, a superconducting quantum interference device, or SQUID, can do something amazing: detect a magnetic field only a millionth as strong as the human brain’s, or less than 5 quintillionths of a tesla.
Measuring such minute magnetic fields turns out to be useful for many things, including geophysical and archeological surveys, detection of the cosmic microwave background, nondestructive testing of materials and devices, and imaging the brain, heart, and other body parts. Invented some 50 years ago, the SQUID now comes in dozens of varieties, with different materials and circuitries, and operating temperatures both high (at the liquid-nitrogen range of around 77 kelvins) and low (less than 10 K, in the realm of liquid helium).
Exquisite though they are, SQUIDs seem to have been invented almost accidentally. During the 1950s and 1960s, industrial and academic labs pursued superconductors with nearly the same zeal they were devoting to semiconductors. Bell Telephone Laboratories, General Electric, IBM, RCA, and Westinghouse all had programs in superconductivity.
And yet, the SQUID did not come from any of these august groups. It was invented instead at the Ford Motor Co., whose scientific lab was then a relative upstart on the corporate research scene. Created in 1951 in Dearborn, Mich., the lab operated according to a philosophy that was established by AT&T’s Bell Labs and IBM’s Thomas J. Watson Research Center and that has now been essentially abandoned. These great institutions pursued research topics not because they were likely to contribute to the parent company’s bottom line anytime soon but because the corporation believed that research for research’s sake was something a real company did.
“We had the freedom to do what we were interested in,” says Arnold Silver, who worked in the Ford lab in its heyday. “We could follow our noses—and particularly, we could follow the data.” In 1963, Silver was a member of the talented team of scientists and engineers who noticed a curious phenomenon in a sample of supercooled phosphorous-doped silicon and then followed it to its logical conclusion.
Like the discovery of the cosmic microwave background at Bell Labs in 1964 and the exposition of fractal geometry at IBM in 1982, the invention of the SQUID at Ford offered no obvious benefit to the company’s principal business. And yet all of these breakthroughs eventually revolutionized major categories of science and brought the company enormous prestige. The SQUID story, like those of the other breakthroughs, also speaks to the question of where, in a radically more restrained corporate research environment, these serendipitous and fundamental discoveries will come from. The question is all the more important when it concerns a device—like the SQUID—that takes decades to find widespread use.
The seeds for the SQUID were planted well before the work in Dearborn. Back in 1911, the Dutch physicist Heike Kamerlingh Onnes [PDF] first observed superconductivity when he succeeded in cooling mercury to a few degrees above absolute zero. At this temperature, atomic vibrations in the material are reduced to the point where they no longer produce any resistance to the flow of electrons. So a current can be sustained in the supercooled material indefinitely and without needing a voltage to push it—that is, it becomes superconducting.
Research in superconductivity continued on the margins of physics until the 1950s. One high point came in 1957 when John Bardeen (a co-inventor of the transistor), Leon Cooper, and Robert Schrieffer published an atomic explanation of the phenomenon, which became known as the BCS theory. The three later shared the Nobel Prize in physics for their work.
Even as Bardeen, Cooper, and Schrieffer were refining their theoretical explanation, work elsewhere focused on making useful superconducting devices. Indeed, much like nanotechnology at the turn of this century, superconductivity in the 1950s and ’60s was considered to be the next big thing. Superconductors, it was thought, would revolutionize computers and computation and transform the power grid by eliminating resistive losses in transmission and distribution. Even more fantastic were the proposed uses of superconductors in space colonization and levitating cars. In an influential 1968 article, “Economic Aspects of Superconductivity,” physicists Roland W. Schmitt and W. Adair Morrison suggested that superconductors might succeed the bipolar transistor and eventually duplicate the market success of semiconductors. No surprise, then, that every sizable tech-oriented company in the world had a superconductivity group.
A giant step toward commercializing superconductivity came in 1962 when Brian Josephson, a graduate student at the University of Cambridge, theorized that a superconducting current, or supercurrent, could tunnel through an insulating barrier between two superconductors without any resistance, thereby completing a superconducting circuit. Once the supercurrent exceeded a certain critical current, however, an AC voltage would develop across the junction, with a frequency approaching 500 gigahertz. The following year, Philip W. Anderson and John M. Rowell at Bell Labs built the first such circuit, subsequently named the Josephson junction. A commercially viable Josephson junction then became the goal of many industry researchers. An electronic circuit fashioned from Josephson junctions would be capable of switching at very high speeds—particularly desirable for computer logic chips.
Ford’s Dearborn lab wasn’t caught up in the fervor surrounding superconductivity, but much like the more established corporate labs, it had a research program in cryogenics and nuclear magnetic resonance. Lab managers had been very successful at recruiting top scientific talent, who were well funded and relatively unfettered in their investigations. When later asked what magnetic resonance research had to do with cars, Terry Cole, a member of the Ford magnetics group, replied, “It really didn’t matter.” Indeed, none of the Ford researchers responsible for the SQUID—a group that included Robert Jaklevic, John Lambe, James Mercereau, Arnold Silver, and James Zimmerman—was drawn to the work because of its possible commercial applications.
Lambe made the first crucial observation, which came about like many a noteworthy discovery: by accident. In 1963, he was studying nuclear double resonance—the interaction between the nuclear and electron spins of an atom—in silicon-29. After placing the silicon sample in his magnetic resonance instrument, he cooled it to 4 K and then watched the signals on an oscilloscope. Normally, magnetic resonance is helpful in studying the physical and chemical properties of a material in the presence of a powerful magnet. A given material will emit a distinctive pattern of signals, or spectra. In this case, though, the sample began emitting signals immediately, while the magnet was still off.
Cole later gave this account in an oral history interview: “Suddenly, before they even turned the magnet on, they began to see what looked like magnetic resonance spectra coming out of their apparatus. They noticed that even with the power off, if they rotated the magnet, these lines seemed to move around, move back and forth in the spectral domain. They turned the magnet on. They were seeing not dozens of lines, which was not unusual, but thousands, tens of thousands of lines. A mystery! They had looked at the spectrum in a nearly identical sample just before. How could something change that much?”
Puzzled, Lambe consulted with some colleagues. The silicon sample had indium solder contacts, and so the team “decided it had something to do with superconductivity, because they began to appear temperature-wise at about the transition temperature of indium,” Silver later recalled.
They then experimented with different thin-film samples. When the samples had no cracks or other flaws, no lines appeared. But when they were damaged even slightly, the effect returned. “Eventually we patterned the films…we cut little notches in them,” Silver said. The oscilloscope readings from these intentionally notched films were even stronger than those from the original sample.
Mercereau, a Caltech Ph.D. who had worked on diffraction waves in liquid helium for his thesis, had a possible explanation for this puzzling phenomenon. He had just come back from a low-temperature physics conference, where he had met Brian Josephson. He suggested that they had in fact created a Josephson junction.
Realizing that their creation was sensitive to tiny magnetic fields, the researchers set about fashioning a Josephson-junction device that could actually measure the intensity of those fields—an interferometer, in other words.
To build the device, Jaklevic started by depositing a film of tin onto a glass microscope slide; he patterned the film by passing it through a stainless steel mask, which he had cut with a razor blade. Next he painted on a mask of Formvar (a type of plastic) and deposited a second tin film to cover the first film. The plastic mask was now the insulator between the two films, and the junction was created through an opening in the plastic between the bottom film, which oxidized in air, and the top film. The device was placed in liquid helium, a coil was inserted to apply magnetic fields to the device, and the resulting signals were viewed on an oscilloscope. The result was the world’s first functioning SQUID, as described by the team in an article in Physical Review Letters in early 1964. This type of SQUID would subsequently be referred to as a DC SQUID because its current stays constant. Interestingly, the Ford researchers were hesitant about using the term “SQUID” in official publications, even though, according to Silver, it soon came into common use around the lab; he credits Zimmerman with coining the term.
Thin-film junctions proved very time consuming and difficult to reproduce. “Months would go by when we could not make junctions,” Silver later wrote. So Zimmerman focused instead on making the circuits from niobium wire, which he happened to have on hand. Niobium thereafter became the superconducting element of choice in subsequent SQUIDs at Ford and elsewhere.
Zimmerman, working with Silver, continued to simplify the devices, which were still hard to reproduce, in no small part because niobium is an especially tricky metal to machine. Finally in 1965 the pair succeeded in making a SQUID consisting of a superconducting ring with just a single Josephson junction interrupting it. Silver used a 27-megahertz radio-frequency detection system from his magnetic resonance lab to measure the SQUID’s signals. When an external oscillating flux was applied to the ring, this low-noise detector captured the change in the internal flux. The researchers dubbed this device an RF SQUID. It was cheaper and easier to produce and became the basis for commercializing the SQUID.
However, as mentioned earlier, the SQUID had nothing to do with cars, and the Ford team made no serious attempts to profit from it. Zimmerman would later recount how team members watched the oscilloscope signal change as steel chairs were moved around the apparatus, minutely altering the surrounding magnetic fields. The lab’s SQUID was “an extremely sensitive detector of lab chairs,” he quipped. The fruits of their intellectual labors were 25 academic papers and nine U.S. patents related to SQUIDs and superconductivity. Even before the team broke up in 1967 or 1968, some members had already moved on to other topics.
What finally put a stop to Ford’s SQUID work, though, was an internal dispute over who had actually developed the device. Mercereau had spent much of 1965 and 1966 touting the team’s work at science conferences—so much so that Phil Anderson at Bell Labs began referring to superconducting quantum interference as the “Mercereau effect.” But within the Ford team, he was not considered the sole or even the main contributor to the SQUID’s invention. To ease tensions, Jacob E. “Jack” Goldman, director of the Ford labs, transferred Mercereau to Aeroneutronic, a lab in Newport Beach, Calif., that had been purchased by Ford.
The story of the SQUID then takes a new direction, with Zimmerman emerging as the protagonist. Within a year of Mercereau’s move in 1967, Zimmerman also joined Aeroneutronic, though not to work with his former colleague. In his new position, Zimmerman led the lab’s cryogenics division and worked on advancing the RF SQUID toward a marketable product.
He was temperamentally suited to the task. His varied background included a stint in Australia during and shortly after World War II, working on radar and on photometry. Later, after earning his Ph.D. from the Carnegie Institute of Technology (now Carnegie Mellon University) in 1951, he took a position with the Smithsonian Institution at the Table Mountain Observatory in California. There, he worked on measuring the electromagnetic radiation emitted by the sun, otherwise known as the solar constant, and he continued that research at the Montezuma Observatory in Chile’s Atacama Desert. The flexibility of mind that allowed Zimmerman to shift easily from radar to cryogenics to astronomy and back to cryogenics would serve him well in his quest to commercialize the SQUID.
In 1969, while still working for Aeroneutronic, Zimmerman cofounded—with physicist John Wheatley and several others—a company in San Diego called Superconducting Helium Electronics. SHE’s main focus was on manufacturing RF SQUIDs and the helium-refrigeration units needed to operate the devices. Initially, the company sold niobium devices like the ones that Zimmerman had designed in Dearborn. Eventually, that design evolved into a niobium-aluminum oxide-niobium device, which was much easier to make, although it had an inferior signal-to-noise ratio. SHE’s devices were the world’s first commercially successful SQUIDs.
In late 1969, Zimmerman was contacted by physicist David Cohen, who had become intrigued by biomagnetism and was experimenting with measuring electrical fields of the human heart and brain using copper coils. To do this, Cohen built a large magnetically shielded room and took hundreds of passes of each subject, using signal averaging to eke out a decent signal-to-noise ratio. “I badly needed a better detector and had heard about Zimmerman and his SQUID,” Cohen recently recalled.
In late December 1969, Zimmerman visited the Francis Bitter National Magnet Laboratory at MIT to meet with Cohen. Zimmerman had with him a portable SQUID demonstrator that he’d built; it consisted of an RF SQUID in a small helium-filled Dewar vessel and an RF amplifier in a small aluminum box. On New Year’s Eve, they tested it. Cohen later offered this account: “Finally we were ready to look at the easiest biomagnetic signal: the signal from the human heart, because it was large and regular. Jim stripped down to his shorts, and it was his heart that we first looked at. The resulting MCG [magnetocardiogram] signal exceeded my best expectations. It was as clear as a conventional ECG [electrocardiogram], and several orders of magnitude better than the MCG from a coil detector. Although I didn’t realize it, a new era had arrived in biomagnetism.”
Cohen likened his own effort with the copper coils to “trying to explore a new continent…using a rowboat.” The SQUID, he said, gave him “a powerboat.”
Cohen and Zimmerman’s experiment was the first use of a SQUID on a living subject. Though the apparatus wasn’t ready for clinical use, the results turned both men into true believers about the medical applications of the SQUID. Cohen next began using SQUIDs for measuring the tiny magnetic fields in the brain—even trickier to detect than those from the heart. Meanwhile, in San Diego, SHE shifted its focus to designing SQUIDs specifically for medical use; eventually, the company changed its name to Biomagnetic Technologies and sold SQUIDs for performing magnetoencephalograms, or MEGs.
Around the time of his visit with Cohen, Zimmerman was offered a job at the National Bureau of Standards (now the National Institute of Standards and Technology) at its facility in Boulder, Colo. He resigned from SHE due to a possible conflict of interest, but he never let up on making the SQUID more portable and responsive. He also helped create the first arrays of SQUIDs, for taking measurements over a wider area. With Martin Reite at the Colorado Health Sciences Center, for instance, he designed a SQUID array for MEGs; Reite used the apparatus to study the brain’s auditory response.
If Zimmerman’s work on SQUIDs was groundbreaking, his keen interest in refrigeration was equally important. The lack of good cooling technology, he realized, was one of the main things holding back the adoption of the SQUID in industry, medicine, and elsewhere. Available commercial refrigerators were too big and, worse, contained magnetic components, which could overwhelm the faint signals the SQUID was supposed to detect. Zimmerman spent more than a decade designing portable cryogenic systems, including an ingenious Stirling-cycle refrigerator no bigger than a bicycle pump. Made largely of plastic, it could still cool a niobium SQUID to 8.5 K. Later, after he retired in 1985, he and his grandson devised a liquid-helium cryostat that could maintain temperatures of 1 millikelvin. Now known as the Z cryostat (“Z” for Zimmerman), it is a standard piece of equipment in many high-temperature superconductivity labs.
Fifty years after its invention, the SQUID is only now coming into its own. Two companies currently manufacture the devices, and dozens of research groups around the world are investigating new applications for them. The Swedish company Elekta, for instance, makes neurological “stations” that each incorporate 306 SQUID circuits for verifying the abnormal magnetic activity associated with epilepsy and other conditions. Aided by high-speed computer processing, the apparatus yields a three-dimensional magnetic field map of a patient’s brain that can be used to guide surgical treatment of the condition.
The SQUID is also a principal building block for numerous electronics applications, including analog-to-digital converters and both traditional and quantum computing. In nondestructive testing, the devices have been used to detect aluminum corrosion in aircraft. Such anomalies would otherwise be nearly impossible to detect without dismantling or otherwise damaging the components, and the magnetic fields produced by such corrosion are exceedingly weak compared with those from other components on the plane, such as the steel fasteners.
Researchers are also considering the SQUID as a tool for measuring the effectiveness of magnetically activated drug delivery; this technique involves dispersing drugs through the blood using magnetic nanoparticles. SQUID arrays offer a way to noninvasively detect where the nanoparticles have dispersed and where the drug has been delivered. There have even been reports of nanoSQUIDs, which their inventors claim can measure the magnetic field of a single atom.
If the SQUID continues on its way to a glorious future, it will be a testament to Zimmerman’s energy and persistence. Building on those fragile lab curiosities devised in Dearborn, he doggedly refined and improved them over the course of several decades. Along the way, he spread the SQUID gospel, collaborating eagerly with researchers outside his own discipline and enthusiastically tackling the hard engineering problems to ensure its success. What the story of the SQUID elegantly shows is that the moment of invention—however surprising, revelatory, and exciting—is but the first small step in the long road to reality. It is also a poignant reminder of what’s been lost. As today’s corporations move away from unfettered basic research, we should not forget the crucial role of the 20th century’s industrial labs and the ingenious ideas and inventions that emerged from them.
This article originally appeared in print as “How Ford Invented the SQUID.”
About the Author
Ann Johnson is an associate professor of history at the University of South Carolina. She first learned about the Ford Motor Co.’s unlikely involvement in basic superconductivity research while writing a paper on the origins of nanotechnology. She’s no stranger to the automotive industry, however. Her Ph.D. dissertation was on the invention of car antilock braking systems; it later became the book Hitting the Brakes: Engineering Design and the Production of Knowledge (Duke University Press, 2009). The research for this article was supported by a grant from the U.S. National Science Foundation.