THE DISCOVERY earlier this year of a new class of high-temperature superconductors sparked a wave of intense research that quickly produced—for the first time in more than 20 years—a whole family of new superconductors. The recent developments have rekindled discussions of advanced superconductor applications and boosted excitement in the field to levels rarely seen in more than two decades.
Superconductors are an exotic group of materials endowed with unique properties, most notably a knack for conducting electricity without losing energy along the way. In ordinary conductors, such as copper and an enormous number of other metals and alloys, as electrons flow through the conductors' lattice, they ricochet off impurities and vibrating atoms. The scattering process leads to loss of energy as heat, akin to an electronic form of friction. In contrast, electrons glide through superconductors in a frictionless manner.
With a talent for mediating smooth electronic sailing, superconductors would seem the ideal material from which to build efficient electric motors and power storage and distribution systems. Compared with ordinary equipment, superconductor-based gear would operate at extreme savings—energetically and financially.
So it would seem in principle. In practice, however, it doesn't work that way. "Superconductivity is a delicate phenomenon arising from a subtle balance of forces," says Princeton University chemistry professor Robert J. Cava. Relatively few materials exhibit superconductivity, Cava points out, and many of those materials become superconductors only when they are chilled below an impractically low critical transition temperature (Tc)—often within a few degrees of absolute zero.
Mercury, for example, turns into a superconductor when chilled with liquid helium to roughly 4 K (−452 °F). That observation, which launched the field of superconductivity, was first made in 1911 by physicist Heike Kamerlingh Onnes of Leiden University, in the Netherlands. Onnes was honored for his work with the Nobel Prize in Physics in 1913.
Like mercury, other metals, such as lead, niobium, tin, and various alloys, also become superconductors, but only at transition temperatures near absolute zero. Various materials of this type, collectively referred to as low-temperature superconductors, have been known for decades, but their extreme cooling needs have largely ruled out opportunities for large-scale applications, such as municipal power systems. Some of these materials, however, for example niobium alloys, are used to build the powerful electromagnets that lie at the heart of modern magnetic resonance (NMR and MRI) instruments.
Seventy-five years after Onnes' pioneering discovery, so-called high-temperature superconductors made their debut. In 1986, Karl Alexander M??ller and J. Georg Bednorz, staff scientists at IBM's Zurich Research Laboratory, showed that a ceramic compound, a copper oxide containing barium and lanthanum, supports superconductivity at a relatively balmy Tc of 35 K. Although the transition temperature isn't exactly toasty, it pegged the thermometer a full 12 K above the Tc of the record-holding superconductor at the time.
Not only was the high transition temperature unprecedented, it was inexplicable. The prevailing theory predicted that superconductivity could not persist above 30 K.
With the discovery of superconductivity in cuprates, "the entire universe of condensed matter physics and materials science was turned on its head," Cava asserts. "That was the event of a lifetime," he says, adding that "there was simply no premonition that such an exotic state could exist in an oxide."
The shocking discovery of high-temperature superconductors set off a worldwide race to find other cuprates with even higher transition temperatures, as well as to find new theories to explain the high-temperature behavior. It also netted its discoverers, Müller and Bednorz, the 1987 Nobel Prize in Physics.
Within months of the seminal announcement about the La-Ba-Cu-O superconductor, researchers prepared a La-Sr-Cu-O compound and reported a Tc of roughly 40 K. Shortly thereafter, the record-setting value jumped to a whopping 90 K with the synthesis of yttrium barium copper oxides. That temperature is noteworthy then and now, experts stress, because it can be reached by cooling a wire or device made from that kind of superconductor with liquid nitrogen (boiling point 77 K).
Compared with liquid helium, liquid nitrogen is inexpensive and easy to handle. Reducing the stringency of the cooling demands—or largely eliminating them by coming up with a superconductor with a room-temperature Tc—would greatly increase the practicality and range of superconductor applications. So the search for higher Tc materials continued.
Not long after reaching a Tc of 90 K in the late 1980s, the value topped 100 K. The Tc of Bi2Sr2Ca2Cu3O10 was measured at 107 K, and a Tl-Ba-Ca copper oxide checked in at 125 K. Eventually, a superconductor with a Tc of 138 K was prepared and tested. First reported in 1995, (Hg0.8Tl0.2)Ba2Ca2Cu3O8.33 is generally considered the present record holder for a Tc measured at ambient pressure. (Under extreme pressure, the value climbs even higher.)
Cuprates with a Tc near room temperature have thus far eluded scientists. Nonetheless, during the past several years, some of the high-temperature superconductors have been put to use. Specialized cables, motors, and components of power transmission equipment, for example, have been demonstrated in a few countries. High-Tc materials are also used to make advanced research magnets and electronic filters for cellular communications equipment.
In 2001, after scientists spent years searching for superconductors with complex compositions, researchers in Japan surprisingly found that MgB2, a simple inexpensive metal boride, is a superconductor with a Tc of 39 K. Although that value is far lower than the Tc values of nearly all cuprates, it is 16 K higher than that of any other simple metallic superconductor. Since 2001, researchers have failed to find other superconductors that are related chemically to MgB2.
Fast-forward to 2008, and history just might be getting set to repeat itself. In February, researchers at Tokyo Institute of Technology reported the discovery of a new superconductor based on a rare-earth iron arsenide compound. The team measured a Tc of 26 K in samples of LaOFeAs that were doped with roughly 10 atom % of fluoride ions (J. Am. Chem. Soc. 2008, 130, 3296).
"It was beyond imagination that an iron compound should be a superconductor," says materials science professor Hideo Hosono, who led the study. Iron is magnetic, and generally, magnetism and superconductivity don't mix, he explains.
IF THAT'S SO, why was his group screening iron compounds for superconductivity? "We weren't, at least not initially," Hosono says. His team was trying to prepare magnetic semiconductors—not superconductors. And while evaluating the materials' low-temperature electronic properties, the group found that the iron arsenides were indeed superconductors. "It was quite a surprise," Hosono admits.
Word of the new Fe-As superconductors spread quickly, and researchers began looking for related compounds with higher Tc values. As with the cuprates 20 years ago, researchers didn't take long at all to begin finding other members of this brand-new family of superconductors. Just weeks after Hosono's paper was published, various research groups quickly started broadcasting their results, some of which were posted on the preprint server arxiv.org.
For example, Xianhui Chen of the University of Science & Technology of China, in Hefei, reported that SmO1–xFxFeAs superconducts at temperatures up to 43 K. Hosono's group together with colleagues at Nihon University, in Tokyo, also reached a Tc of 43 K by using high-pressure methods to synthesize the F-doped LaOFeAs compound reported in their JACS paper (Nature 2008, 453, 376).
Meanwhile, Zhongxian Zhao of the Institute of Physics at the Chinese Academy of Sciences, in Beijing, found that the praseodymium analog of the Fe-As material has a Tc of 52 K. Zhao and coworkers then followed up with another paper reporting that the samarium iron arsenide compound synthesized under pressure has a Tc of 55 K (Chin. Phys. Lett. 2008, 25, 2215).
Just a few weeks ago, the Tc inched up further when Guanghan Cao and Zhuan Xu of Zhejiang University, in Hangzhou, China, reported results of an alternative doping strategy. Rather than replacing some of the oxygen ions with fluoride ions, as other researchers have done, the Zhejiang team partially substituted thorium ions for gadolinium to create Gd0.8Th0.2FeAsO. The group measured a Tc of 56 K, which, as of C&EN press time, stands as the record value for a noncuprate superconductor (Europhys. Lett. 2008, 83, 67006).
"The race is on, just as for the cuprates, to discover materials with increasingly higher Tc," says Michael R. Norman, head of the Condensed Matter Theory Group at Argonne National Laboratory. Research on this new family of superconductors, which are known as iron pnictides (compounds of the nitrogen group elements), is proceeding at a hectic pace, and already, the initial Tc value has been raised from 26 to 56 K. "If a material could be discovered with a Tc above 77 K, then things would become really interesting," Norman says. Scientists aren't there yet, but they're working on that challenge and related ones.
One of those challenges is coming up with a satisfactory theoretical explanation for the basis of superconductivity in these materials. Unfortunately, no widely accepted theory exists to rationalize the cuprates' behavior, let alone that of the brand-new iron pnictides.
The properties of low-temperature or conventional superconductors, such as mercury, are understood according to the BCS theory developed by and named for the American physicists John Bardeen, Leon N. Cooper, and John R. Schrieffer. A key element of that theory, for which the researchers were honored with the 1972 Nobel Prize in Physics, is a mechanism that coaxes electrons, which would ordinarily repel one another, to form pairs. These "Cooper pairs" glide in an unimpeded manner through the material's lattice, thereby allowing electric current to flow without losing energy. How such pairs form in high-temperature superconductors is not well understood even after two decades of research. Scientists hope that the new high-Tc materials will shed light on that question.
"For 22 years, it appeared that this high-temperature behavior was unique to copper oxides," says Douglas Scalapino, a theoretical physicist at the University of California, Santa Barbara. With the discovery of the pnictides, it's clear that high-Tc superconductivity isn't just for an exclusive club of substances. Now that two families of compounds can be compared and contrasted, researchers may soon be able to uncover properties that guide them toward explanations of superconductivity in both sets of materials. That's a big deal, in Scalapino's view. "It's like getting the opportunity for a 'do-over' in a ball game."
In terms of structure and some other properties, the two classes of compounds look alike. Similar to the copper oxides, the iron arsenides are built up from layers of alternating chemical composition. In the newer materials, Fe-As planes are sandwiched between layers composed of oxygen and lanthanum or other rare-earth elements.
THE NEED for doping is another similarity shared by both classes of compounds. Inserting dopants into the lattices alters the number of electrons in the parent materials and leads to superconductivity. In the case of LaOFeAs, for example, researchers explain that introducing fluoride ions in place of some of the oxygen ions injects electrons into the Fe-As layers, through which they are conducted across the crystal.
Another common feature stems from the materials' magnetic properties. Both families of undoped compounds contain rows of ions in which the pattern of magnetic spins alternates from row to row. And in both types of materials, that pattern fades as the material is doped and turns into a superconductor.
Magnetism is connected to superconductivity in a number of ways. For example, a standard method for characterizing superconductors is subjecting the material to a magnetic field and measuring the material's ability to conduct current without resistance—that is, its tendency to behave like a superconductor. Superconductors tend to rebuff and exclude magnetic fields in their vicinity in a display of a well-known phenomenon known as the Meissner effect. The effect is powerful enough to levitate magnets (or superconductors) as shown in popular superconductor demonstrations. When the applied magnetic field exceeds some characteristic threshold value, however, superconductivity shuts down.
Measuring that value, known as the upper critical field, can reveal key pieces of information about superconductors, according to Frank L. Hunte, a postdoc at the National High Magnetic Field Laboratory at Florida State University. For example, it indicates whether a new superconductor can sufficiently support high magnetic fields to be a good candidate material from which to build research magnets, such as the type used in particle accelerators. In addition, measuring the variation of the upper critical field with temperature can shed light on the material's superconductivity mechanism.
Hunte and coworkers recently conducted such experiments on LaO0.89F0.11FeAs and were surprised to find that the material remained a superconductor even when exposed to the lab's 45-tesla magnet, the world's most powerful continuous-field magnet (Nature 2008, 453, 903). In terms of nailing down the mechanism (or mechanisms) that govern superconductivity in pnictides and cuprates, Hunte says without hesitation that "the jury is still out."
However, on the basis of the group's study, he concludes, "It's likely that we're dealing with different mechanisms." At this early stage in iron arsenide superconductor history, it's no surprise that some researchers have reached the same conclusion while others propose that pnictides and cuprates superconduct via similar mechanisms.
While some teams of scientists continue to address the mechanism question or strive to raise iron arsenides' transition temperatures, other groups are probing different fundamental issues. At the University of Edinburgh, in Scotland, chemistry professor J. Paul Attfield and coworkers have examined the usefulness of incorporating high-pressure and high-temperature methods into synthetic routes for preparing "late" rare-earth superconductors, referring to elements of higher atomic number. That strategy is known to stabilize many late rare-earth analogs of various early rare-earth compounds, the team says.
The Edinburgh group finds that that approach also works well for the pnictide superconductors. After reports of RO1–xFxFeAs (R represents a rare-earth element) superconductors made from La, Ce, Pr, Nd, Sm, and Gd, Attfield's team synthesized Tb and Dy analogs. The procedure required preparing TbAs and DyAs under vacuum at high temperature and then reacting each reagent with iron compounds in an inert atmosphere at over 1,100 °C while maintaining pressures of 10???12 gigapascal (Chem. Commun. 2008, 3634).
IN ADDITION to the general conclusion that these extreme synthesis methods are effective techniques for preparing such compounds, the team reports that the properties of superconductors remain fairly constant as the elements from Ce (element 58) through Dy (element 66) are incorporated into the materials.
Working at even higher pressures, Haozhe Liu of Harbin Institute of Technology, in China, and coworkers squeezed a neodymium arsenide superconductor at more than 30 GPa while measuring the effect of pressure on the lattice structure. As Liu explains, extreme pressure can compress the lattice structure and shrink the spacing between the Fe-As and Nd-O layers. Forcing the layers closer together may increase the material's Tc by improving charge transfer between the layers. But it doesn't always work that way, he cautions.
On the basis of synchrotron measurements conducted at Brookhaven National Laboratory, the Harbin team found that as pressure increases from 0 to 15 GPa, the interlayer spacing in F-doped NdOFeAs decreases by about 1 Å and Tc increases marginally. As the pressure increases further, however, one of the other lattice parameters abruptly increases and the Tc drops (J. Am. Chem. Soc., DOI: 10.1021/ja804229k).
Within a few months of the discovery of a new iron- and arsenic-containing superconductor, researchers in many labs responded with a whole slew of discoveries on related materials. It's anyone's guess at this point whether the new family of iron pnictides will include a future member with a Tc above the boiling of liquid nitrogen—let alone room temperature—and whether the new compounds will eventually lead to everyday applications for high-temperature superconductors.
For now, it's clear that the discovery has reinvigorated the field of superconductivity. The finding isn't the "event of a lifetime," aficionados say, but it is a powerful shot in the arm for an area of basic science that's closing in on its first century of life.