A massive wind turbine—capable of turning the breeze into two million watts of power—has 40-meter-long blades made from fiberglass, towers 90 meters above the ground, weighs hundreds of metric tons, and fundamentally relies on roughly 300 kilograms of a soft, silvery metal known as neodymium—a so-called rare earth.
This element forms the basis for the magnets used in the turbines. "Large permanent magnets make the generators feasible," explains materials scientist Alex King, director of the U.S. Department of Energy's (DoE) Ames Laboratory in Iowa, which started making rare earth magnets in the 1940s as part of the Manhattan Project. The stronger the magnets are, the more powerful the generator—and rare earth elements such as neodymium form the basis for the most powerful permanent magnets around.
In the modern world rare earths go far beyond magnets. Spanning 17 elements—from lanthanum to lutetium, plus scandium and yttrium—they find use in computers, screens, superconductors, oil refineries, hybrid or electric vehicles, catalytic converters, compact fluorescent lightbulbs, light-emitting diodes, lasers, audio speakers and microphones, cell phones, MRI machines, telecommunications, battery electrodes, advanced weapons systems, polished glass, and even the electric motors that run automobile windows. "There is no single military system in use by the Pentagon that does not contain rare earths," King notes, ranging from Abrams tanks to radar systems.
But, in large part, magnets drive the growth in demand for rare earths such as neodymium—swelling by 15 percent per year, according to an analysis published in Science—and dysprosium. Magnets are the key to generating electricity, of course, and electricity is the key to the use of cleaner sources of energy—whether wind turbines or electric vehicles.
At the heart of those devices sits the most powerful magnet available today—a mix of neodymium, iron and boron, which can produce an energy product of as much as 60 megagauss–oersteds (a unit of magnetic strength). For comparison, a typical iron magnet has an energy product of only four megagauss–oersteds and a refrigerator magnet is typically a mere 0.5 megagauss–oersted. "The stronger the magnet, the smaller the magnet can be," explains Luana Iorio, a manager at GE's High Temperature Alloys and Processing Laboratory.
Fortunately, the elements "are neither rare nor earths," as in soil (although they can be found there), King adds. The name comes from their seeming scarcity when first discovered in the late 18th century in an ore found near Ytterby, Sweden. Unfortunately, "finding reasonable concentrations of them that are economically extractable is quite rarer."
As it stands, 97 percent of the 124,000 metric tons of neodymium, dysprosium—the name means "hard to get"—and other important rare earth elements produced each year come from one place: China. "When demand for neodymium started to rise, driven in part by wind turbines and hybrid autos, the Chinese were the only providers left," King says. And demand for the rare earths continues to outpace supply, particularly as China has implemented export quotas.
So the hunt is on for both better ways to use rare earths as well as better ways to mine and, perhaps more important, cleanly separate the rare earths for use.
A critical pass
Vast waste ponds scar the landscape on the banks of the Yellow River, 190 kilometers from the city of Baotou in China. Visible from space, the Bayan–Obo iron mine in Inner Mongolia is the world's largest source of rare earths, and the Chinese companies supplying them employ acid to dissolve them out of ore rock that often also contains radioactive elements like thorium, radium or even uranium. Intensive boiling with strong acids—repeated thousands of times because the elements are so chemically similar—finally separates out the neodymium, dysprosium or cerium.
Such a difficult production process is one reason why the U.S. no longer mines them. Until the 1990s, the U.S. was the major supplier of rare earths, largely out of one mine in Mountain Pass, Calif., owned by oil company Unocal, now part of Chevron. Unocal shut down that mine—and processing facility—in 2002 because they could not compete with China's purer product that began to flood the market in the 1990s. Unocal "didn't think there would be a need for high-purity products," explains chemist John Burba, chief technology officer for Molycorp, the company hoping to reopen rare earth production at the Mountain Pass Mine.
Plus, the operations were not exactly environmentally friendly. "Back in the 1990s the plant was sending 850 gallons of wastewater per minute down a pipeline into evaporation ponds," Burba notes. "It was a devil's mixture because of the chemistry they employed." In addition, the rare earths at Mountain Pass are mixed with radioactive thorium, requiring special care in handling and disposal.
The whole slew of rare earth elements are a challenge to separate because of their chemical similarity—and they are never found alone. "The challenge with rare earths is they always occur together," GE's Iorio explains. "There are processing costs to separate them out from each other."
Regardless, the Chinese National Offshore Oil Corp. (CNOOC) wanted to purchase Unocal—and its California rare earth asset—in 2005, a move the U.S. military blocked, according to Burba. Now Molycorp wants to restart operations by 2012 using a new process, which will require Molycorp to essentially rebuild the entire operation at a cost of $500 million. The process employs a strong acid and a base to separate the rare earths—the so-called chlor–alkali solvent extraction method—but it still will not produce pure rare earths; rather it will yield oxides of cerium, lanthanum, praseodymium and neodymium.
In essence, Mountain Pass will become a chemical plant, sucking up electricity and steam from an on-site natural gas–fired boiler. In addition, the wastewater of the process will be recycled back to produce the strong acid and base necessary to start the process all over again—hydrochloric acid and sodium hydroxide. "Mining is a very small part of our operation," Burba says, noting that mining the ore containing the rare earths is only 10 percent of his company's cost. "The vast majority of what we do is advanced chemistry."
Of course, there will still be by-products—such as the residual ore, or tailings, from the mining and separation as well as calcium carbonate, magnesium carbonate and magnesium hydroxide from the chemical process, along with that pesky thorium. But the primary salt from the chlor–alkali process is sodium chloride (otherwise known as table salt), which will be recycled back into the process using some of the steam generated on site and use to make new acid and base using a chlor-alkali unit. "It's a big saltwater loop," Burba explains. "Our water consumption is 10 percent or less of what had been done historically at this site."
By next year, the site hopes to produce 2.7 million kilograms of rare earth oxides a year—separating the elements from the ore using a liquid ion-exchange process. By 2015, they hope to be at full production, producing 18 million kilograms of various rare earth oxides a year. "We have greater than 30 years of mining capacity at 40 million pounds per year," Burba says.
But that only represents 6 percent of the present global market, which is still growing. The U.S. Government Accountability Office estimates it will take seven to 15 years to find new rare earth deposits, build the infrastructure to process them, and make them available to manufacturers. "This is not like going out and panning for gold," Burba notes. "This processing requires a huge amount of chemical processing. You have to have good infrastructure."
Geologists have found deposits in Australia, Canada, Mongolia, Vietnam and even Greenland, and efforts are also underway to begin mining the deposits in the legions of discarded electronics available today. "There is a lot of rare earth material out there in used products; extracting from that urban mine will be viable," argues King of Ames Lab, which also has scientists working on better extraction methods for, say, neodymium from ores or old generators. "Recycling is definitely going to be a big part of the solution to this problem."
But for the foreseeable future China will continue to dominate rare earth production—and it holds the world's largest reserves, nearly twice as much as its neighbors to the north and west in the Commonwealth of Independent States (an organization of former Soviet republics formed after the dissolution of the U.S.S.R.) and three times as much as U.S. reserves. And China is the only producer of dysprosium—vital for the heat-resistant magnets favored by the U.S. military and hybrid car–makers.
The perils of that dominance became evident to the world this fall when China reportedly shut off rare earth supplies to car manufacturers and other users in Japan as a result of a diplomatic imbroglio. After all, by 2005, all U.S. manufacturers of the neodymium iron boron magnets—invented by General Motors researchers in the early 1980s—had shut down. But even before China flexed its market-dominating power a slew of scientific researchers had been investigating how to use less rare earths—or even none at all—by fabricating better magnetic materials.
Magnetism arises from the electrons orbiting the atomic nuclei of some elements. When atoms align in a certain fashion a strong magnetic field results. Magnetic elements like iron or neodymium typically arrange themselves this way, thus generating a permanent magnetic field.
But by tinkering with that alignment—spacing it out with other materials or embedding it in a lattice composite at the nanoscale (a nanometer is one billionth of a meter)—scientists can potentially exponentially increase the strength of these magnets: The stronger the magnet, the fewer of them you need. "There have been theories around for 15 years that if you had a very controlled nanostructured magnet, you could as much as double" its strength, GE's Iorio says. "If you can double the strength you can use a much smaller magnet and get the same performance—or have the same size and get much more performance." That also means GE could spend less on the metal materials; the company currently spends some $4 billion annually on the purchase of metals and alloys.
GE was awarded $2.2 million by DoE's ARPA–E program to develop bulk quantities of such nanocomposite magnets in a bid to cut by 80 percent the rare earth elements used. The challenge is to make such a tiny microstructure stable—and reproducible on a much larger, bulk scale. But, within two years, Iorio says, "we'll have a magnet big enough to sit in the palm of your hand. Something that's useful [to stick on] your refrigerator."
ARPA–E also gave $4.4 million in October 2009 to a group led by physicist George Hadjipanayis of the University of Delaware to create a nanostructured version of the neodymium iron boron magnet that eliminates the need for as much neodymium. The secret: mixing it with softer magnetic materials that remain magnetic only when exposed to a magnetic field. "Neodymium iron boron nanoparticles are difficult to make because they are very reactive," Hadjipanayis says, in addition to being hard to align within a given nanostructure. "It's very difficult to obtain an aligned magnetic material," he adds. "My challenge is to make larger amounts of these nanoparticles, make sure I protect them, and find a way to assemble them."
At the same time, Hadjipanayis is researching whether more widely available rare earths could serve as substitutes or whether powerful magnets could be made that don't employ any rare earth elements, such as iron cobalt alloys. "Anything that comes out, we'll take," he says of the hunt for an alternative to neodymium, although the only known alloy as strong as rare earths is composed of iron and platinum, which is too expensive to compete commercially. "By the time you find something, it takes five to 10 years to make it commercially. I don't see anything for the next 15 to 20 years that is rare earth–free."
And that means the material challenge posed by rare earths won't be solved anytime soon, particularly as more wind farms cover the land, compact fluorescent lightbulbs and light-emitting diodes proliferate, and greater numbers of hybrid or electric vehicles hit the road. "There are some materials crises, like worldwide demand for chromium, that are easily solved—we just stopped putting chromium on cars," King notes. "This one is not so easy to solve."