The Sensible Superconductor

Cheap to make, simple to cool, easy to shape into wires, magnesium diboride could throw the field of superconducting applications wide open. "It could replace niobium-titanium, a conventional low-temperature superconductor, in future MRI magnets," says visionary Paul Grant, science fellow at the Electric Power Research Institute (Palo Alto, Calif.). "It has to be given serious attention."

Besides magnetic resonance imaging (MRI), practical uses of the zero-resistance material in transformers and fault current limiters seem not too much to hope for. Much further off, perhaps, could be networks of superconducting power lines cooled by liquid hydrogen, as proposed by Grant.

Until the magnesium diboride revelation, engineers trying to apply superconductivity to the real world had to grapple with distinctly un-ideal materials. Low-temperature (metallic) superconductors, while not too costly and with excellent mechanical properties, require cooling down nearly to zero—4.2 K—a pricey proposition. High-temperature (ceramic) superconductors can be cooled at far less expense to a less chilly 77 K, but their manufacturing process requires a great deal of silver, a costly material.

Magnesium diboride falls between the two types on the temperature scale. With a superconducting transition temperature of 39 K, it can be conveniently cooled with commercial cryocoolers or liquid hydrogen (boiling point: 20.2 K).

A powder that can be found in any well-stocked chemistry laboratory, it had never been tested for superconductivity until very recently. The five Japanese researchers who did so announced their discovery in January 2001 at a small conference in the Japanese city of Sendai. The news spread among superconductor cognoscenti like wild fire. Lights burned late into the night in laboratories worldwide as scientists worked overtime studying the inexpensive metallic compound's properties with a view to harnessing its surprising potential.

Within two weeks, Paul C. Canfield of Iowa State University (Ames) and his physics/astronomy department collaborators had made 3-cm-long wires out of the compound by passing magnesium vapors over boron fibers. "This material is just full of good news," he said.

Two months after the initial announcement, the American Physical Society held its annual meeting in Seattle, Wash., and arranged a special evening session of magnesium diboride papers. The large hotel banquet hall chosen for the event was jammed with physicists; the session, which began after dinner, included dozens of oral presentations and went on until 1 a.m.

The attendees were fascinated by the wealth of data indicating the compound was a conventional (low-temperature metallic) superconductor. This augured well for the future. Such a material can be explained by the Bardeen-Cooper-Schrieffer (BCS) theory of superconductivity, which says that the interactions between the electrons in a superconductor are governed by the thermal vibrations of atoms in the crystal structure of the material. So the behavior of magnesium diboride would be easy to understand. (The physical mechanism responsible for superconductors of the so-called high-temperature type is not yet clearly understood.)

"It's rare that we get so excited," said Robert J. Cava of Princeton University, an expert on superconductivity and an attendee at the conference.

But initial data also showed researchers that they had some work to do. In order for the material to be truly useful, its current-carrying capability and magnetic field tolerance would have to be improved and a commercially viable and economical method of manufacturing wires would have to be developed. Today, little more than a year later, the news is very encouraging.

Economical ingredients

The materials that go into making magnesium diboride, magnesium and boron, are both dirt-cheap. EPRI's Grant believes that magnesium diboride cable may eventually be comparable in price to copper cable. When first discovered to be a superconductor, a kilogram of high-purity magnesium diboride cost about US $750.

But with larger demand, the cost could drop dramatically. Grant estimates that it could cost about $10/kg to chemically reduce raw boron pentahydrate to metallic boron. To react boron with magnesium to produce magnesium diboride for wires would probably cost another $10, he says. That adds up to $20/kg for the material.

These numbers are "wet-fingers-in-the-wind estimates and could wind up substantially in error," says the man from EPRI. "But let's say they represent a lower limit. As an upper limit, commercially prepared magnesium diboride should drop to $300 per kilogram with volume demand."

In the power transmission industry, the cost-to-performance ratio is measured in cost per kiloampere-meter. The cost of making magnesium diboride wires will be around $1/kA-m, Grant says. In contrast, high-temperature superconducting ceramic wires sell for about $200/kA-m today. (The price is expected to fall to about $50/kA-m in two or three years). Made of bismuth, strontium, calcium, and copper oxide, often called BSCCO, they are sheathed expensively in silver for structural, metallurgical, and electrical reasons.

Length no problem

Two companies are working to commercialize magnesium diboride wires for applications like coils and cable: Hyper Tech Research in the United States (Troy, Ohio) and Diboride Conductors in the United Kingdom (Cambridge). Hyper Tech has made wires as long as 100 meters, but so far has electrically tested only shorter segments.

The Hyper Tech method of wire-pulling is a modified powder-in-tube approach, explains Mike Tomsic, the company's president. In fact, it is the process used to make most superconducting wires that start out as powders, except that in the Hyper Tech approach, the iron tube is being formed continually. A U-shaped iron tube is filled with the powder and then closed down to form a round wire about 6 mm across, which in turn is drawn down to about a quarter of its starting thickness. To form multifilament wires, several filaments are stacked into a tube, which is drawn down to a width of about 1.4 mm.

Progress in making magnesium diboride wire and tape has been rapid. As early as May 2001, Sungho Jin and colleagues at Agere Systems (Allentown, Pa.), a spinoff of Lucent Technologies, had made magnesium diboride tape in lengths of almost 1 meter with very little fuss. They took iron tubes with outside diameters of about 6 mm, filled them with magnesium diboride powder, then flattened and stretched the tubes until they became ribbons about 5 mm wide and 0.5 mm thick. Lastly, they baked the ribbons at a temperature of 900 °C, which fused the powder into a solid.

"In principle, using this process, you could make the wires as long as you like," said Jin. He pointed out, too, that inexpensive iron could be used for the magnesium diboride sheath as opposed to the silver used with high-temperature superconducting ceramics.

Cool comfort

A distinct advantage of the high-temperature ceramic superconductors is that they can be cooled by liquid nitrogen, which is very cheap in itself and, with its 77-K transition temperature, also cheap to keep cool. Conversely, wires made of conventional superconductors like niobium-titanium need to be cooled, at great expense, to liquid helium temperatures. Magnesium diboride, which falls in between, needs to be chilled to 20-30 K for it to be useful. While this is colder than liquid nitrogen, it is within the range of a standard commercial cryocooler, and the cost is not that high. "There's an entire world of economic difference between liquid helium temperatures and the temperature at which this material operates," says Grant.

For applications such as transformers for electric utility grids, magnesium diboride seems a good choice, he continues. In fact, they may be the entry application that the material needs to gain a foothold in the marketplace. According to his estimates, magnesium diboride transformers will be more economically attractive not only than those based on high-temperature superconductors but even than those using copper wire.

According to Grant, the ownership costs associated with a magnesium diboride transformer, operating at 25 K, would be about $59/kA-m, most of it due to cooling costs. For a copper transformer, operating at 300 K (room temperature), this figure is around $65/kA-m, mostly due to resistive losses. For a high-temperature ceramic superconductor, operating at 77 K, this figure is between $80 and $180/kA-m, most of it coming from the cost of the wire.

The wires may also be used to improve the performance of electric motors and generators, because superconducting coils can carry much more current than copper. Thin superconducting films could serve as electronic components and sensitive magnetic sensors. Other potential roles would be in superconducting magnets and wireless base stations.

On 15-16 November 2001, a conference on the practical applications of superconductors was organized in Boston by the Knowledge Foundation (Brookline, Mass.). To attendee Deborah Van Vechten, magnesium diboride looked "very positive" for wireless receivers with ultrahigh signal sensitivity and transmitters with ultralow phase noise. Van Vechten is the program officer for superconductivity for the U.S. Office of Naval Research.

Also present was Jun Akimitsu, the leader of the Japanese team that announced the superconducting behavior of magnesium diboride. He told IEEE Spectrum that the Japanese government has started funding a big effort to find industrial applications for the material. "This material has many features—it's very light, cheap, and easily made," he said. "I believe it has a big potential for applications."

Few disagree with Akimitsu.

"If magnesium diboride had been discovered in the 1960s and 1970s," said Cava, the Princeton expert who wrote a commentary on the importance of the discovery, then "the whole culture of superconductivity research would have been different."

Nothing's perfect

With the announcement of magnesium diboride's unsuspected talent, superconductivity has, almost overnight, shot into prominence again. Yet, compared with other superconductors, magnesium diboride did have a few shortcomings to overcome.

The material's inability to carry much current was one that surfaced early. In May 2001, Jin and his colleagues at Agere reported that their wires could carry 35 000 A/cm². (Real-life superconductor applications require a larger value, at least 80 000 A/cm².) However, that figure has moved sharply upward, and is currently at around 200 000 A/cm2 for a magnetic field of 1 tesla, typical of transformers and motors.

More worrisome is the material's inability to stand up to very strong magnetic fields. Early data showed that its superconductivity vanished in fields greater than 2 T, which produces magnetic vortices inside the alloy. These vortices move under the Lorenz force created by the current, and their motion dissipates energy, which shows up as electrical resistance.

The solution turns out to lie in making the material less structurally perfect. A team led by David Caplin at London's Imperial College found that structural defects induced in the alloy pin the vortices and keep them from moving. The team bombarded samples of magnesium diboride with protons and found the resulting defects greatly enhanced the material's ability to shepherd current through a magnetic field. While proton bombardment is impractical for large-scale manufacture of wire, the research suggests that chemical doping may work just as well.

"We've shown that modest disorder, at the level of about 1 percent, can generate a material whose performance is technologically attractive," and attainable in an economically viable way, either by chemical doping or mechanical processing, Caplin explained.

As it happens, chemical doping is now yielding the predicted improvement. A technique that adds oxygen to thin films of the superconductor is being pioneered at the University of Wisconsin at Madison by Chang-Beom Eom, professor of materials science and engineering, and colleagues.

"We believe that oxygen goes into the lattice and replaces some of the boron atoms," Eom told Spectrum. He says it is an alloying process that improves both the current-carrying capacity and magnetic field tolerance. More than one mechanism is at work, it appears. Adding oxygen also creates minute particles of magnesium oxide that aggravate the disorder.

But can the techniques Eom uses to improve the superconducting properties of magnesium diboride films also be used for wires? Eom is unsure. And Hyper Tech's Tomsic cautions that although there has been a proof of principle that there are ways to improve vortex pinning, doing it in a wire uniformly is a big challenge.

If these magnetic issues are resolved without harm to the economics of production, magnesium diboride wires could prove of use in medical diagnostic machines that use powerful magnets for magnetic resonance imaging.

A challenger in the wings

Given the heady excitement of the past year, some physicists who have worked in superconductivity dream of the day when magnesium diboride wires are in common use. But even if the material proves its mettle against BSCCO, it may be on a collision course with a so-called second-generation high-temperature superconductor, a compound of yttrium, barium, copper, and oxygen, which is in the early stages of development.

Like BSCCO, YBCO is superconducting above liquid nitrogen temperature. But it is cheaper to manufacture and has better magnetic field characteristics. To make matters more interesting, magnesium diboride and YBCO are on the same development time line. Hyper Tech's Tomsic expects to be selling magnesium diboride wire commercially by 2004 or 2005 at $1-2/kA-m. On the YBCO side, Philip J. Pellegrino, president of IGC-SuperPower (Schenectady, N.Y.) told Spectrum that he plans to sell the second-generation superconductor at about $10/kA-m in the second half of this decade.

Linda Geppert, Editor