Driven by an expected surge in demand for electric vehicles (EVs), wind turbines, and other applications requiring permanent magnets, consumption of many rare-earth (RE) elements is expected to outstrip the global supply within a decade. Coupled with an almost total US dependence on China for separated REs and the magnets made from them, the impending shortage has prompted the US government to subsidize and stimulate domestic RE mining, metal-making, and magnet manufacturing.
Rare-earth elements are those with atomic numbers between lanthanum (57) and lutetium (71). Scandium and yttrium are often considered REs because they share some properties. For a primer on rare earths’ electronic properties, see the Quick Study by Jianshi Zhou and Greg Fiete, Physics Today, January 2020, page 66.
Neodymium-iron-boron (NdFeB) magnets are the highest-performing, most compact, and most lightweight type of permanent magnet commercially available. Although neodymium is the principal RE used in magnets, NdFeB magnets often contain praseodymium because the two are difficult to separate.
The NdFeB magnets are manufactured either by sintering—melting and compressing powders—or by combining and pressing powders with a bonding material such as epoxy or a thermoplastic. Sintered magnets are capable of much stronger fields, whereas bonded ones can be injection molded, compression bonded, extruded, or 3D printed into more complex shapes. Sintered NdFeB magnets are the most commonly used by far: An estimated 160 000 tons were sold globally in 2020, according to industry consultant John Ormerod; bonded-magnet consumption last year was 12 000 tons. Peter Afiuny is executive vice president of the US’s sole manufacturer of sintered NdFeB, Texas-based Urban Mining. He says the US accounts for around 20% of the $20 billion global demand for those magnets.
An exploded view of a permanent-magnet electric-vehicle traction (propulsion) motor. The rare-earth-containing magnets are embedded in the rotor.
An exploded view of a permanent-magnet electric-vehicle traction (propulsion) motor. The rare-earth-containing magnets are embedded in the rotor.
Skyrocketing prices last year for the REs used for magnets reflected increasing tensions between China and the West. The price of Nd jumped 75% from January to December last year, and NdPr was up 50%. Terbium, used along with dysprosium in small amounts to prevent demagnetization of NdFeB magnets at elevated temperatures, more than doubled in price during the same period.
Demand for NdFeB magnets will double this decade, analysts say, faster than the rate of growth for REs in consumer electronics, catalysts, phosphors, and other uses. Ryan Castilloux, managing director of the research and consulting firm Adamas Intelligence, forecasts that demand for REs as a group will climb 10% annually over the next 10 years. Alex King, founding director of the Critical Materials Institute at Ames Laboratory, notes that annual demand growth for REs has returned to its long-term trend of 13% following a downturn caused by a price spike in 2011 (see the plot on page 21). There is every reason to expect that trend to continue, he says.
Global production of rare-earth oxides (REO), 1950–2019. The trend line (red) is fitted to the data (blue) through 2010 and extrapolated beyond that point. The drop of production below the trend line within the green band indicates the drop in demand following a price spike in 2011. (Adapted from A. King, Critical Materials, Elsevier, 2020, based on data from USGS Mineral Commodity Summaries.)
Global production of rare-earth oxides (REO), 1950–2019. The trend line (red) is fitted to the data (blue) through 2010 and extrapolated beyond that point. The drop of production below the trend line within the green band indicates the drop in demand following a price spike in 2011. (Adapted from A. King, Critical Materials, Elsevier, 2020, based on data from USGS Mineral Commodity Summaries.)
An “unfathomable” amount of new Nd, Pr, Tb, and Dy oxides—5000 to 6000 tons each year—will be needed by the second half of the decade to keep pace with NdFeB magnet demand, says Castilloux. That would require a new mine the size of Mountain Pass in southern California each year. Mountain Pass is the sole US mine and currently supplies 15% of the world’s REs. An annual shortage of 48 000 tons of NdFeB powders and alloys will develop by 2030, Castilloux predicts, roughly the amount needed to make propulsion motors for 25 million–30 million EVs.
Department of Defense acts
At 8%, Mountain Pass’s ore-bearing rock has one of the world’s highest RE ore concentrations. But all of its output is shipped to China for processing and separation into individual REs. With help from a $9.6 million grant last year from the Department of Defense, mine owner MP Materials plans to open a plant in 2022 to separate lighter-atomic-weight REs, primarily Nd and Pr, on site. It would be the first separation facility in the Western Hemisphere.
The Pentagon awarded a second grant to MP Materials to design a separation plant for heavy-atomic-weight REs, principally Dy and Tb. Mountain Pass ore has little heavy RE content; however, it could produce small quantities for defense purposes. The amount of the DOD grant wasn’t disclosed.
Both awards, and two others provided by DOD to other RE companies, were issued through the Defense Production Act of 1950 and were prompted by executive orders and directives issued by President Donald Trump beginning in 2017. In July federal agencies were instructed to take unspecified actions to lessen US dependence on foreign sources of REs and 34 other “critical minerals.” Without such assistance, Trump said, the US industry “cannot reasonably be expected to provide” RE oxides, metals, and alloys or NdFeB magnets. The orders stated that “purchases, purchase commitments, or other action taken under the Defense Production Act are the most cost-effective, expedient, and practical alternative method for meeting the need for this critical capability.”
Lynas, the only other major producer of REs outside China, last year received a DOD grant of an undisclosed amount to design a heavy RE separation plant that it plans to build in Texas. The company currently processes Australian-origin ores in Malaysia.
The Pentagon last year also gave $29 million to Urban Mining, which recycles REs from discarded NdFeB magnets to make new magnets. Afiuny says the grant will help the company expand production capacity at its plant, which he says is the only NdFeB sintered magnet manufacturing facility in the Western Hemisphere. There are three bonded-magnet manufacturers in the US, says Ormerod. Other companies in the US customize magnets for specific applications, but their material comes mostly from China.
The Department of Energy announced in November that it will consider applications from industry to receive loan guarantees to help finance projects that would increase domestic output of REs and other critical minerals ranging in scarcity from aluminum to platinum group metals. The agency has around $40 billion in unused loan guarantees designated for clean-energy development and electric-vehicle manufacturing.
Principal national; contributions to global rare-earth oxide (REO) production, 1950–2019. (Adapted from A. King, Critical Materials, Elsevier, 2020, based on data from USGS Mineral Commodity Summaries.)
Principal national; contributions to global rare-earth oxide (REO) production, 1950–2019. (Adapted from A. King, Critical Materials, Elsevier, 2020, based on data from USGS Mineral Commodity Summaries.)
Before November’s announcement, DOE had maintained that mining and separation operations were not eligible for loan guarantees because they are too far upstream in the manufacturing chain. In 2012 the department rejected a $280 million loan-guarantee application from Molycorp, the previous owner of Mountain Pass. That company filed for bankruptcy in 2015.
Tucked into the fiscal year 2021 National Defense Authorization Act is a requirement that within five years most military systems exclusively use REs mined and refined entirely outside China. It further directs the Pentagon to select the best options for improving the domestic availability of REs, such as providing funding or restricting foreign supplies. In 2018 Congress ordered the Pentagon to purchase only non-Chinese-made NdFeB magnets and magnet alloys but didn’t rule out raw materials from that nation. China’s share of the world’s RE mining has declined in recent years, but it still makes about 90% of NdFeB metal and magnets; Japan produces nearly all of the rest. Much, though not all, of Japan’s RE feedstock comes from Lynas. (See Physics Today, October 2018, page 22.)
Surging demand
The global market for EVs could grow a stunning 36% annually over the next decade if nations adhere to commitments made under the 2015 Paris Agreement, according to the International Energy Agency. The European Union, the UK, and California have announced bans on sales of new internal combustion vehicles, to take effect in 2035. President Biden has proposed adding 500 000 EV charging stations by 2030 and providing new incentives for buying EVs.
Most EV propulsion motors use permanent magnets; Tesla, which employed induction-motor technology invented by Nikola Tesla in its first-generation Roadster, switched to permanent magnets for its mass-market vehicles. “Electric motors using [NdFeB] magnets are the motor that will drive electrification,” Afiuny says. “There are no serious rivals.”
MP Materials forecasts 31% annual growth in demand for magnets used in EV propulsion motors through 2030, according to the company’s website. EVs alone could consume all the world’s mined Nd and Pr by 2035, compared with 5% of total demand today, the company says.
Rapid growth also is expected for magnets used in wind turbines, with estimates ranging up to 25% annually. The European Union plans a fivefold increase in offshore turbines by 2030, to a total of 60 GW. An increasing proportion of turbines are so-called direct drive and use permanent magnets to couple the slow-moving blades to the generator. They require less maintenance and have fewer moving parts to wear out compared with conventional turbines, which use a gearbox to couple the blades to the generator.
Direct-drive turbines require 1 ton of NdFeB per megawatt, says Afiuny. In 2019 the average land-based turbine in the US produced 2.5 MW, according to Lawrence Berkeley National Laboratory. Offshore, direct-drive turbines dominate and are larger; 10–12 MW systems are expected to become more prevalent in the coming years. GE just unveiled a direct-drive turbine that will produce 13 MW.
Small NdFeB magnets are found in dozens of automotive components, including power windows, speakers, switches, and actuators. Apart from cars, the magnets are used in consumer electronics, including hard disk drives, speakers, and cell phone vibrators, and for actuators used in aircraft and other applications.
China remains essentially the world’s only source of Dy and Tb, which are extracted from clays in the south of the country. Magnet manufacturers have achieved some success in reducing the amounts of the two elements without sacrificing performance. Meanwhile, engineers are figuring out ways to design motors to better dissipate heat, says King, which would lessen the need for Dy and Tb. Urban Mining developed an NdFeB magnet that uses 40% less Dy than conventional NdFeB and offers superior thermal performance, it says. Siemens, the world’s largest producer of wind turbines, has recently begun making permanent magnets that are free of heavy RE elements.
There is a huge, virtually untapped potential for recycling magnet metal. Less than 1% of the NdFeB contained in components that have reached the end of their useful lives is reused in the US, says Urban Mining’s Afiuny. Adamas forecasts that 90 000 tons of NdFeB magnets will be entering the waste stream globally each year by 2030. Still, demand growth will outpace the amount of material recovered.
As supply tightens, NdFeB magnets will likely become more limited to high-value applications such as EV propulsion, says King. That’s what happened during the 2011 price spike, when automakers substituted other types of magnets in many applications such as windshield wipers and power-window motors, where high performance and compactness aren’t so important.
Some studies have shown that cerium and lanthanum, both abundant, can supplant a portion of NdPr in some magnet applications. Meanwhile, in November 2020 Niron Magnetics of Minnesota announced a partnership with General Motors and Marquette University to develop an iron nitride permanent magnet for EV drivetrains. Niron claims the technology offers inherently higher magnetization and lower cost than NdFeB. DOE’s Vehicle Technologies Office supported that work with a $5 million grant.
Financing needed
MP Materials declined to comment for this article. Castilloux and King both say that Mountain Pass could meet North American demand for Nd and Pr through 2030. But since a typical RE mine contains only 15–25% Nd and Pr, globally 60 000 tons to 100 000 tons of RE mine production overall will be needed by the end of the decade just to meet magnet demand.
There are few new mining prospects in the US. Texas Mineral Resources is proposing to develop, in partnership with the privately owned company US Rare Earths, a property in West Texas, primarily for heavy REs. Production could begin in 2023. Another potential source is the Bear Lodge area, on US Forest Service land in Wyoming.
Canada has several RE mining prospects. They include Quebec’s Kwyjibo, owned by the provincial government, and the Ashram Deposit, owned by Commerce Resources. The Foxtrot mine in Labrador is owned by Search Minerals. In addition, Ucore Rare Metals of Nova Scotia plans to develop the Bokan-Dotson Ridge Project, a heavy RE deposit near Ketchikan, Alaska. That deposit is rich in heavy REs, and the plan has political support from the state’s congressional delegation. The outlook for its development brightened as a result of the domestic RE-sourcing requirement in the National Defense Authorization Act. The Army Research Laboratory in 2014—2016 awarded Innovation Metals Corp a total of $1.8 million to help fund development of its RE separation technology. Ucore acquired Innovation Metals last year.
Although those prospects are promising, all of them currently lack the hundreds of millions of dollars needed to bring them to production, says Castilloux. “Investors see prices going through the roof, but not many are to the point where they want to become a financer. They’re looking at which ones can we jump into and jump out of in six months and make a few million. There needs to be a longer-term interest in seeing these projects through.” If financing were located, gearing up to full production would take five years or more.
Even when Mountain Pass begins separating oxides, the US will remain dependent on China for metal making and magnet production.
Light rare-earth elements are generally considered to be those ranging from lanthanum to promethium. The remainder are known as heavy rare earths.
Light rare-earth elements are generally considered to be those ranging from lanthanum to promethium. The remainder are known as heavy rare earths.
“There are several chicken and egg situations for this industry,” says Castilloux. “Having oxides-processing capacity in the US will bolster the business case for establishing metal-making capacity and magnet capacity.” MP, for one, says it plans to manufacture NdFeB magnets by 2025.
Separating concentrated ores to individual RE oxides is the most capital-intensive part of the process, says King. China has heavily invested in plants that may not be the most efficient, he notes. “If we start today with all the research that has gone into separations and reduction, we should be able to build more efficient plants and compete with China on a free-market basis.”
Ucore, which declined an interview request, plans to build a separation plant at its Alaska prospect and hopes that sales of oxides will help pay to develop the mine. “With the ability to refine raw materials into high-purity RE oxides, the significant capital costs to build an RE mine can then be justified,” company chairman and interim CEO Pat Ryan said in an October 2020 statement. In the meantime, Ucore says its Alaska plant will separate REs from unspecified non-Chinese sources. But helping others process their material could undermine Ucore’s own mining project, says Castilloux, who views the Alaskan deposit as one of the more economically challenging.
An alternative source
Growing demand and prolonged high prices may stimulate the extraction of REs from monazite, an RE-rich component of mineral sands. Last August, the Saskatchewan provincial government provided Can$31 million ($24 million) to build an RE processing facility. It will have a capacity to produce 500 tons a year of individual RE oxides from monazite when it begins operation in late 2022.
Energy Fuels, the largest US uranium miner, announced in December an agreement to acquire monazite sands from the Chemours Co’s Georgia mine. Energy Fuels says it can recover enough mixed REs to meet 10% of US demand when it begins processing in the first quarter of this year. But while it considers whether to build its own separation facility, the company will ship its RE-rich concentrated ore to Europe or Asia for that purpose.
Medallion Resources in Vancouver, British Columbia, is completing a technical and economic feasibility study of a monazite-processing plant that would produce 3500 tons of REs annually, including 500 tons of NdPr.
Other nations with abundant monazite deposits include Australia, India, Indonesia, Madagascar, and the Philippines. Iluka Resources, an Australian mineral-sands mining company, has committed AU$35 million ($26.8 million) to upgrade its monazite-processing operation to produce RE enriched ores.
“Developing a US-based integrated RE magnet supply chain is a significant challenge requiring cooperative efforts from the mine to the magnet original equipment manufacturer,” Ormerod sums up. “A close private–public sector coordinated effort is essential if the capital, economic, and technical hurdles are to be overcome.”
Correction, 29 January 2021: This article originally stated Niron Magnetics received a grant from DOE’s Advanced Research Projects Agency–Energy instead of the Vehicle Technologies Office.
Correction, 11 March 2021: This article originally stated Ucore received a grant from the Army Research Laboratory. Ucore acquired the company, Innovation Metals, that has received the grant.