Rare earths recycling and recovery: the two sides of the industry

By Antonio Torrisi
Published: Wednesday, 29 October 2014

The rare earths supply risk and consequent panic buying that emerged in 2011 paved the way for sluggish demand, oversupply and low prices. In light of this, recycling rare earths from downstream products might appear an unnecessary economic burden. Antonio Torrisi, Reporter, investigates how this technology can integrate with new mining projects to promote future stability and sustainability in the rare earths market.

The stability the rare earths industry experienced between 2000 and 2009 was broken in 2011-2012 by surging panic in international markets, generated by China’s ban of rare earths exports to Japan and tightening export quotas, following the territorial dispute over the Senkaku (or Diaoyu) Islands.

As a result, rare earths prices spiked over the course of a year, as markets feared a potential supply shortage. Between June 2010 and August 2011, the average price of cerium oxide (Ce2O3) surged by 2,400%, from $6/kg to $149.5/kg, while prices for neodymium oxide (Nd2O3) and dysprosium (Dy2O3) leapt by 991% and 700%, from $33/kg to $360/kg and from $300/kg to $2,400/kg, respectively.

Three years later, the rare earths market looks very different. Exports from China have remained below the country’s quotas for the last two years, on the back of depressed demand. In addition, the extremely fragmented Chinese rare earths industry, characterised by hundreds of small producers and a persisting illegal black market, meant that a huge oversupply of rare earths has contributed to lower prices, even below values seen in 2009.

Average prices for Ce2O3, Nd2O3 and Dy2O3 amounted to $5.25/kg, $77.5/kg and $350/kg as of October 2014, according to the IM Prices Database.

Outside China, in 2012, US-based Molycorp Inc. restarted rare earths production at Mountain Pass, California, and in February 2013, Lynas Corp. started processing ore mined in Mount Weld, Australia, at its Lynas Advanced Materials Plant (LAMP) in Malaysia. Both companies are targeting an annual production capacity of 20,000-23,000 tpa rare earth oxides (REOs).

In the last year, persisting low prices and sluggish demand hit Chinese rare earths producers, with the world’s largest producer, Inner Mongolia Baotou Steel Rare Earth Hi-Tech Co., seeing its profits fall by over 70% in H1 2014 and China Minmetals Rare Earth Co. posting a 94% fall in its profits in the first quarter of this year.

The depressed market conditions also hit Molycorp and Lynas, which have been struggling to turn their rare earth businesses to profit since they entered production, and have seen losses widen in the first half of this year.

Japan, India and the US have stockpiled rare earths, while Sumitomo Corp. and Kazakhstan’s National Atomic Co. (Kazatomprom), have recently started test production at a plant in Kazakhstan, targeting a production capacity of 1,500 tpa REOs, which is expected to expand to 3,000 tpa in 2015 and to 5,000-6,000 tpa in 2017.

Meanwhile, junior companies started to explore for rare earths resources outside China. According to Gareth Hatch at Technology Metals Research, 429 new rare earths projects are under development in the world, involving 261 companies and 37 countries outside China and India.

In March, the World Trade Organization (WTO) ruled against China’s rare earths and tungsten export quotas, following a formal complaint filed by the US, Europe and Japan in 2012.

In an analysis of the current supply chain of rare earths published in February, the US department of Defence (DoD) concluded that there is not a supply risk and that “market supply and demand conditions overall have significantly improved in the US and internationally since 2010 and 2011 turmoil in the rare earths market”.

The report acknowledges that there has been an 11.8% decrease in demand from 2010 to 2013 and that the initial forecast growth has been revised downwards.

The DoD also said it saw no shortfalls in the supply of rare earths ores and concentrates for its own uses, adding that stronger demand is likely to occur through 2015 and 2018.

While acknowledging the present absence of a potential risk of shortages in rare earths supply, the US DoD is sponsoring research in the reclamation of rare earths from downstream products at the Critical Material Institute (CMI), based at the Ames Laboratory in Iowa. The CMI recently received $120m of funding from the US Department of Energy (DoE) to develop projects to recycle rare earths from permanent magnets.

Within the present rare earths market conditions, it might be surprising that several industries and countries are embarking on projects to study the reclamation of rare earths from end-products and e-waste, including lamp phosphorous, nickel-metal hydride (NiMH) batteries and permanent magnets.

With the exception of fluorescent light phosphors and nickel-metal hydride (NiMH) batteries, the current state of the technology does not allow for an economically feasible reclamation of rare earths, the DoD stated.

However, there are other important reasons that make recycling rare earths from downstream industries an interesting perspective, which could ultimately help the rare earths market to become more sustainable.



NdFeB magnet rising demand

Neodymium (Nd) and dysprosium (Dy) rare earth elements are used in neodymium-iron-boron (NdFeB) permanent magnets. Dy substitutes Nd in the magnet composition to increase the temperature stability of the magnet against demagnetisation, but the content varies according to the specification of the magnet. Typically, about 1% of Nd is substituted with Dy in permanent magnet formulations.

Despite the present oversupply, demand for rare earths is expected to increase in the near and long-term future, due to their application in several high-tech industries, including computers, mobile phones, light emitting diodes (LED) and catalysts.

In particular, demand for NdFeB magnets is expected to see a surge worldwide, due to their application in green technologies including hybrid electric vehicles (EVs) and wind turbines, which are expected to experience rapid growth in China and the US in the medium to long-term future.

According to consulting company, Walter Benecki LLC, the NdFeB market will continue to grow, with global production projected to expand from 63,000 tonnes in 2012 to 78,000 tonnes in 2015, and from 50,000 tonnes to 65,000 tonnes in 2015 in China.

The US DoE estimates it will need 17,000 additional tonnes of didymium (Nd+Pr) oxide by 2015, which translates into an additional 28,000 tonnes NdFeB magnets, representing a 30-40% increase compared with the total production in 2012.

China started new policies this year in order to regulate its domestic rare earths market, tackling illegal mining through thorough investigations and promoting the formation of corporations in the rare earths market.

Beijing is resolute about preserving China’s rare earth natural resources and mitigating the environmental pollution caused by the rare earths industry. The Ministry of Industry and Information Technology (MIIT) has pinpointed the development of downstream products and industries as the strategic route to improve the domestic market.

According to recently published analysis by Dudley Kingsnorth, executive director of Industrial Minerals Company of Australia (IMCOA), China could become a net importer of REOs in the next five years, while focusing its efforts on the production of permanent magnets, light phosphors and other value-added products.

New mining projects come to play a major role in assuring the supply of rare earths, but not all of them are at a stage where they can enter production in the next two years. Also, many have high costs due to the energy-intensive processing required to extract the products.

For this reason, the electronics industry has some concerns about a possible shortage of critical rare earth elements, such as neodymium and dysprosium, in the near future.



Recovery and recycling

Despite the large number of projects under development worldwide, none of the junior companies have entered the new production stage yet (although some are processing saleable material from tailings). This is because the recovery of rare earths, whether from new mining projects or from tailings and mining by-products, is a complex process with high energy and financial costs.

The production of rare earth concentrates from ore requires flotation and pre-leaching processes with large consumption of acids. In addition, rare earths separation from the concentrate with the traditional solvent extraction technique is a complex process, which requires the use of several acid and organic solvents, including kerosene, and multiple steps.

Colorado, US-based mineral engineering consultants Lyntek Inc. have frequently discussed this issue with IM (see July 2012: “Does it have to cost the earth?” and Processing 2013 “Costing the Earth”), stressing the need to liberate REO from the gangue, or ‘host’ mineral, before attempting to separate the REOs using magnetic, gravity, electrostatic or flotation methods.

The type of processing method is usually dictated by the composition of the ore, but also by cost considerations and environmental regulations.

TSX-V listed Geomega Resources Inc. has developed a new method to separate rare earths from a REO mineral concentrate, which is still generated through a conventional flotation and pre-leaching process of the ore from its Montviel mine in Quebec, Canada.

“During the process, charged particles and ions migrate in the separation channel perpendicular to the flow, under the effect of the electric field,” Kiril Mugerman, head of corporate development at Geomega Resources, told IM.

“The speed of migration depends on the electrophoretic mobility of the particles and ions, which varies based on charge and size,” he added.

The size of the ion makes a big difference, Mugerman told IM. Impurities in the concentrate can be separated very well from the rare earths in the free flow electrophoresis (FFE) process, owing to a great difference in size. They separate very fast, completely independently, without interfering with the rare earths separation.

“Some channels will get only one rare earth element with 100% purity, with no other ions, some other channels will get a mixture. Those can be just re-run through the process in order to further separate the ions,” he said.

According to Mugerman, the technique will have the advantage of simplifying the recovery of neodymium and praseodymium at high purity, as it is very difficult to separate neodymium and praseodymium from a didymium (Nd+Pr) mixture with solvent extraction separation.

The company explained that electricity becomes its only input in the separation stage of rare earths processing, which makes no use of multiple chemical solvents.

The lack of organic solvents has a very positive impact on the mitigation of environmental risks in addition to reducing operating costs, according to Geomega.

Mugerman said Geomega sees potential to attain 100% purity and complete recovery with no special adjustments required based on rare earth element distribution of concentrate.

The method could be applicable to many different types of concentrates, including those produced from mine tailings, by-products and end products waste - such as hard disk drives - containing rare earths, according to the company.

Recovery of rare earths from by-products and mine tailings might become an economically feasible approach to fast-tracking entrance into production.



Other projects looking at recycling

Canadian junior Orbite Aluminae Inc. is developing a process to recycle rare earths as a by-product of its high-purity alumina (HPA) produced from its aluminous clays in Quebec, Canada. The company is planning to extract the minerals from the liquid solution left behind in the process to produce HPA.

Japan-based Nippon Light Metal launched a pilot plant in February 2013 to study the possibility of extracting rare earths from red mud waste from bauxite deposits in Jamaica.

The plant, which was temporarily halted in August this year due to the presently depressed rare earths market, will treat 30 tonnes of dry red mud with acid in order to test the recovery of rare earths from bauxite residue.

Canadian junior Medallion Resources is planning to recover rare earths from monazite mineral sands in the Gulf States. The company signed a memorandum of understanding (MoU) with Takamul Investment Co., a subsidiary of Oman Oil Co., and it is planning to build a rare earths separation plant, expecting to bring it into production within three years from receiving permits and approvals.

In the US, tailings at Molycorp’s Mountain Pass rare earths deposit in California contain approximately 3-5% REOs, according to the DoD.

Monazite and xenotime reserves at the Pea Ridge iron mine, in Missouri, host an average concentration of 20.3% REOs, with estimated reserves of over 200,000 tonnes REOs, according to the Missouri Department of Natural Resources.



Recycling from NdFeB magnets

Meanwhile, the US Geological Survey (USGS) estimated that of the 90,400 tonnes of rare earths produced in 2011, 65% went to landfill, 23% to construction aggregate and 9% to downgrade use, while less than 1% was recycled.

Recycling rare earths is an energy-intensive process that requires complex steps and large amounts of chemicals, which need to be disposed of. For some products such as cerium and lanthanum the process may not be economically feasible due to their present low prices.

According to the DoE, the market has concerns on the supply of five critical rare earths, namely neodymium (Nd), dysprosium (Dy), europium (Eu), terbium (Tb) and yttrium (Y), and about one third of the rare earth tonnage is of high enough value to be recycled, according to the International Electronics Manufacturing Initiative (iNEMI).

Waste products from the electronics industry can be a valuable source of neodymium and dysprosium, which are contained in permanent magnets. According to Arnold Magnetic Technologies, hard disk drive (HDD), compact disks (CD) and DVD systems accounted for 14% of the global Nd demand in 2010, which is expected to rise to 16% in 2015.

According to Volker Zepf from the University of Ausburg, Germany, there is a theoretical recycling potential for the magnets in HDDs, by recovering Nd and Dy either from the scrap in the manufacture of the magnets, during which up to 30% of rare earths can be lost, or from end-of-life (EoL) products.

A simplified recycling flow sheet for NdFeB magnets considers recycling from the scrap - also called swarf - originated during the manufacturing process, from small magnets in EoL products and large magnets in hybrid and EVs and wind turbines (see figure 1).

Direct recycling of the magnets is only useful for large magnets, while for small magnets contained in loudspeakers, mobile phones and HDDs, pre-processing is needed in order to dismantle the magnets - the resin bonded magnet in the spindle motor and the sintered magnet in the voice coil motor - from the end-products.

As nearly all waste electronics and electrical equipment (WEEE) is presently shredded at its EoL, the magnets are powdered in the process and end up sticking to the ferrous waste contained in the shredder, which makes separation difficult. De-magnetisation of the magnets by heating above 3000C, or applying opposite magnetic fields is needed to allow separation, an operation which increases the energy costs of the process.

Specialised automatic techniques seem to be more economical than manual separation. Japanese NdFeB manufacturer Hitachi announced in 2012 that it was developing a recycling process to recover Nd from HDD and air conditioners.

The University of Birmingham conducted studies in 2012 to separate sintered rare earth magnets by hydrogen gas, with separation efficiencies of around 95% on small scale trials.

Following the shredding, recovery of Nd from the magnets is achieved by applying typical hydrometallurgical routes, which require large amounts of leaching agents, mainly hydrochloric (HCl), nitric (HNO3), sulphuric (H2SO4) acids and caustic soda (NaOH), that need to be disposed of after use.

Rhone-Poulenc, now Rhodia-Solvay Group, developed a method to recover samarium and cobalt from SmCo magnets and swarf using a hydrometallurgical process in 1994. The US Bureau of Mines developed a sulphuric acid-based aqueous process to recover Nd from NdFeB magnets in the same year.

Alternative routes to recover neodymium in permanent magnets include pyrometallurgical processing, electroslag refining, liquid metal extraction, glass slag method and gas-phase extraction.

In 2000, the Ames Laboratory developed a process based on a liquid-solid reaction system to recycle rare earths from Nd2Fe2B magnets and other alloys, using magnesium as an extraction reagent.

This year, the Anhui University of Technology in China and the University of Delft in the Netherlands developed a process to recycle Nd, Dy and Pr from NdFeB magnets scrap using molten magnesium chloride and potassium chloride. The researchers achieved recovery rates of 86.6% Nd, 89.2% Pr and 79.7% Dy running the process for three hours at 1,0000C.

The use of a specific process to recover rare earths in permanent magnets often depends on the size and composition of the magnets.

Recycling Nd from NdFeB magnets contained in HDDs could account for 10,000-14,000 tonnes Nd globally, according to Zepf, with 25% contribution from North America and Europe and 50% contribution from Asia-Pacific.

Despite steady growth in mobile phone demand, a 100% recovery of rare earths in magnets contained in these products would account for a total of 3,000 tonnes rare earths, according to Zepf, who forecast that about 5,200 tonnes Nd could be recovered from wind turbine generator permanent magnets by 2020.

Estimates of the use of permanent magnets containing Nd and Dy in hybrid and EVs indicate a total amount of 1,250 tonnes Nd and 450 tonnes Dy will be absorbed by the automotive markets in Germany and the US, and 4,200 tonnes Nd and 1,450 tonnes Dy in the global automotive market, by 2020.

Recycling from NiMH batteries

Spent NiMH batteries contain about 36-42% nickel and 8-10% mischmetal, a mixture of La, Ce, Nd and Pr, which have high hydrogen storage capacity.

In 2011, Umicore and Rhodia developed a solvent extraction based process to recycle rare earths from NiMH batteries. An industrial pilot plant has been operation near Antwerp, Belgium, since September 2011, with an annual capacity of 7,000 tpa rare earths. The company can process more than 350,000 tonnes of electronic waste, including photovoltaic cells, computer circuits and mobile phones.

Honda Motor Co. Ltd and Japan Metals & Chemicals announced in 2012 they had established a recycling plant to extract rare earths in NiMH from Honda (H)EV vehicles using molten salt electrolysis, which can remove about 80% of the original material.

Recycling from lamp phosphors

Lamp phosphors are a rich source of heavy rare earths such as europium, terbium and yttrium, and have the advantage of a more straightforward recycling process compared with permanent magnets, according to Koen Binnemans, research scientist at the University of Leuven, Belgium.

Processes to recycle individual phosphors are based on physicochemical separation or chemical attack through hydrometallurgical routes. Red phosphor Y2O3:Eu3+ has the highest intrinsic value as it contain large concentrations of yttrium and europium.

However, a major issue in recycling rare earths from lamp phosphors is the removal of mercury, which is highly toxic.

In most countries, rare earths are not recycled from EoL lamp phosphors, which are landfilled or stored in containers.

A two-liquid flotation process, which was developed by the Waseda University and the University of Tokyo, Japan, in 2008 reported successful reclamation of Y2O3:Eu3+, with purity and recovery levels higher than 90%.

In 2010, Rhodia developed a flow sheet to recover rare earths from a mixture of halophosphate and rare earth phosphors, using HNO3, NaOH and molten sodium carbonate (NaCO3). Rhodia planned to build two facilities at La Rochelle and Saint-Fons, France, to recover rare earths from EoL fluorescent lamps.

Siemens’ former subsidiary, OSRAM, developed a process to recover all rare earths from used phosphors using several selective leaching steps after mechanical separation of the coarse parts of the EoL fluorescent lamps in 2012.

Despite some successful achievements, recycling rare earths from downstream waste products of the high-tech industry - also known as urban mining - is not a consolidated process and several projects have been developed to establish its economic feasibility and industrial scalability.

Complementary sources

Marcus Reuter, director of technology management at Finnish engineering group Outotec Oyj, told IM earlier this year that in order to boost low recycling rates the rare earths industry needs to shift from a material-centred to a product-centred approach.

The difficulty of recycling rare earths in e-waste is due to the fact that the “designed mineralogy” found in manufactured products is often more complicated than the simple primary mineralogy of geological deposits, Reuter said.

Improving recycling rates will allow the rare earths industry to take advantage of recycling not only against potential supply shortages, but more as a complementary strategy to reduce supply/demand imbalance and boost prices of REOs and end-products, guaranteeing more sustainability.

However, owing to the low concentrations of rare earths in high-tech products and the complexity of their extraction from EoL products, recycling will not be able to meet global rare earths demand in a market where global consumption of a resource grows at a rate of more than 1% per annum.

Meanwhile, the presently depressed rare earths market is not putting pressure on end-users to seek alternative sources for these minerals, and new projects outside China are likely to enter into production only after 2016.

However, the large cost of these projects in terms of processing, waste management and shipping rare earths might be an increasing burden for downstream industries.

Moreover, rare earth mining projects that extract critical dysprosium and yttrium from xenotime or monazite contain radioactive thorium and uranium as by-products, that need to be handled and disposed of.

The decay process of the two radioactive elements thorium and uranium can also generate other isotopes such as radium, radon, actinium and protactinium, which could leak into rare earth carbonates in the processing stages.

This adds an extra level of process complication in the production of rare earths from primary sources, increasing costs and probability of mineral losses.

Reclamation of rare earths from e-waste represents a solution to these problems, as there are no radioactive by-products in EoL gadgets.

The most positive benefit of rare earths reclamation in the downstream industry is the possibility to solve what is known as the ‘balance-problem’ - the need to maintain an equal demand and supply of individual rare earth elements at any time, due to the co-presence of rare earths in geological ores.

This is clearly not possible, as different elements have different applications in different markets. For instance, while demand for neodymium and dysprosium in permanent magnets is projected to steadily increase in the near future, the use of yttrium and europium in red phosphors is expected to decrease as the market is already developing rare earth-free substitutes.

The problem is further exacerbated by the fact that rare earth elements occur in different ratios in mineral deposits, with cerium and lanthanum being far more abundant than neodymium and dysprosium. This implies that mining rare earths will almost always produce large amounts of lanthanum, and cerium, which need to be sold to prevent oversupply.

Recycling critical rare earths such as neodymium and dysprosium from end-products will prevent additional mining of light REOs, however, limiting their oversupply in the long term.

“The balance problem explains why even for countries with large primary rare earth resources, such as China, recycling of rare earths is becoming an issue,” Binnemans said.

“A drastic improvement in EoL recycling rates for rare earths is a strategic necessity, even more so in countries possessing no or few rare earth deposits,” he added.

According to Binnemans, this is possible only by developing fully integrated recycling routes to increase recycling rates.

These are determined by collection rate of old scrap and the recovery rate - the recycling process efficiency rate.

Binnemans forecast global recycling rates to vary from 32-56% by 2020, on the basis of global collection and recovery rates estimates in three main end-products such as permanent magnets, NiMH batteries and lamp phosphors.

In order to increase global collection rates Binnemans urges legal enforcement, collection schemes and international co-operation within urban mining industries.

In addition, more efficient and environmentally-friendly flow sheets, including dismantling, sorting, pre-processing and extraction processes can increase recovery rates.