Fertilizer International 534 Sep-Oct 2026
11 September 2026
Integrating REE recovery into phosphate processing flowsheets
ADVANCED P&K PRODUCTION TECHNOLOGY
Integrating REE recovery into phosphate processing flowsheets
Phosphate producers have an opportunity to pursue rare earth element (REE) recovery from phosphogypsum (PG). Treating PG as a strategic resource, by unlocking its critical mineral potential, is a way of converting a liability into a cleaner, saleable product. Lyndsay Tran, Khanh Pham and Gabriela Olvera of Hatch look at the options for REE recovery from PG and their relative merits.
Introduction
Rare earth elements (REEs) are a group of 16 chemically similar metallic elements that play a key role in permanent magnets, catalysts and electronics. They include 15 lanthanides, plus yttrium, and commonly occur together in natural mineral deposits. In 2018, REEs were officially designated as critical minerals by the US government.
Current global mining and refining of REEs is highly concentrated, being limited to a small number of operations and jurisdictions. To counter this, many governments have accelerated efforts to secure their own domestic supply chains due to rising REE demand from electrification and national defence. This has involved expanding the search for REEs outside of conventional ore deposits to encompass secondary and unconventional sources – including phosphate rock and phosphate industry by-products.
The phosphate ore mined for use in phosphate fertilizer production contains varying amounts of REEs locked within apatite. This natural rock feedstock is digested with sulphuric acid during wet process phosphoric acid production. During this process, calcium sulphate (CaSO4) is formed and precipitates in large volumes as a phosphogypsum (PG) by-product.
Between 70-85% of the original rare earth content of the apatite partitions into the PG. As a consequence, this phosphate industry by-product has received increased research attention over the past decade due to the significant opportunity to recover these REEs commercially.
Not scarce – just difficult to isolate
Rare earth elements are chemical similar. Indeed, the near-identical configuration of their outer electron shells causes them to behave as a single chemical entity. The group is commonly subdivided into light REEs (LREEs, La-Sm) and heavy REEs (HREEs, Eu-Lu and Y). HREEs are more valued, as they are generally less abundant than LREEs, and offer a higher market premium.
Conventionally, rare earth elements are sourced from deposits such as bastnasite, monazite and ion-adsorption clays. Despite their name, rare earths are not scarce in an absolute sense. Cerium, for example, has an abundance comparable to copper in the Earth’s crust. Instead, the ‘rare’ in rare earths refers to the fact that they rarely occur in high concentrations and, historically, have been difficult to isolate and purify into individual elements.
RARE EARTH MARKETS AND END USES
Demand for rare earths is being driven by a few major applications. The most important application is permanent magnets, with these constituting roughly 95% of total rare earth consumption by value in 20251.
NdPr (or NdFeB) permanent magnets are made from an alloy of neodymium-praseodymium, iron and boron, with dysprosium and terbium often alloyed-in to impart performance at high temperatures. They are strongest type of permanent magnets available commercially.
Catalysts are another significant market for rare earths. Major types include lanthanum and cerium-based fluid catalytic cracking catalysts and automotive emission-control catalysts. Other significant uses include alloys for rechargeable batteries and optical glass for smartphones, displays and other applications. For permanent magnets, demand is concentrated in three major sectors:
• In electric vehicles, NdPr magnets are preferred for traction motors due to their high energy density and resistance to demagnetisation at high temperatures.
• In wind turbines, permanent-magnet direct-drive generators are favoured over the traditional gearbox-coupled design, as eliminating the gearbox removes the most failure-prone component in a drivetrain6. This is a significant consideration, especially for offshore installations where access to maintenance is limited and downtime is costly.
• In defence applications, NdPr magnets are used in guidance, steering and actuation systems for munitions, aircraft and naval vessels. Demand from this sector is heavily reinforced by government stockpiling efforts. The US Defense Logistics Agency, for instance, is building a strategic stockpile of critical minerals to develop defence systems including several rare earths7.
It is generally the last category, the defence sector, that has changed the need to secure a rare earth supply chain from a commercial planning question into a matter of national security and policy. A critical mineral supply chain that is majority-controlled by a single country – as is the case for REEs currently – poses geopolitical risks for other countries, particularly in national defence, and threatens their ability to build up domestic reserves.
In 2024, China and Myanmar collectively accounted for 83% of global HREEs mining and 76% of global LREEs mining, with China also holding additional stakes in overseas mines1. In downstream processing, China also represented 91% of global refined output and 94% of permanent magnet production1.
The combined demand from the EV, wind and defence sectors has turned rare earth supply into a matter of national policy in multiple jurisdictions around the world, driving a deliberate push to decentralise the supply chain. Several governments, particularly the United States, have responded with direct financial and policy support aimed at establishing independent domestic value chains. Three companies, MP Materials, USA Rare Earth and Arafura Rare Earths, are among those receiving financial incentives and policy support.
Phosphogypsum: a strategic source of rare earths
Rare earth elements are known to occur in phosphate rock due to their isomorphous substitution for calcium in the apatite lattice. Minor amounts are also present as discrete mineral inclusions within the ore, such as monazite, xenotime, allanite and rare earth carbonates12.
Published literature reports typical REE concentrations in apatite in the range of 0.01-0.1% by weight3. While modest compared to conventional REE resources, the total concentration can be significant considering the global throughput and consumption of phosphate rock by the phosphate industry. Several studies have also reported a higher proportion of HREEs in phosphate deposits, compared to conventional rare earth ores, a favourable characteristic given their higher market value.
Tracking where REEs report and partition as they travel through a conventional wet-process phosphoric acid plant helps identify where the recovery opportunity lies (Figure 1). Studies on REE distribution have found that less than 10% of the apatite-derived REE content reports to the tailings stream during beneficiation13. Following acidulation and filtration, approximately 70-85% of the REEs partition to PG, while an estimated 15-30% reports to clarifier sludge11. The phosphoric acid product itself retains a comparatively small fraction, typically less than 1,000 ppm14. Consequently, given this distribution, PG has been the primary focus of REE recovery research within the wet-process flowsheet.
PG has, however, long been considered a regulatory liability for phosphate producers. PG can contain residual phosphoric acid, high moisture content and naturally occurring radionuclides, such as uranium, thorium and radium. In the United States, PG is classified under EPA guidelines as Technologically Enhanced Naturally Occurring Radioactive Material (TENORM).
Wet-process phosphoric acid production generates roughly 4-5 tonnes of PG per tonne of phosphoric acid produced. Estimates from the International Fertilizer Association in 2025 put annual PG output at 245 million tonnes2.
While PG can be treated via dewatering, impurity removal and other processing steps, the treatment costs are often significant compared to the value of the clean gypsum. Consequently, long-term stacking remains the most cost-effective approach for many phosphate producers.
Rare earth recovery thus provides an opportunity to create a value-added REE co-product while reducing the land requirements for gypsum stacking (Figure 2). Given the 245 million tonnes of PG generated annually2, the total recoverable REE inventory could be significant. This would allow for meaningful valorisation of a material that has historically been considered a waste stream.
Methods of rare earth recovery from PG
Recovery from PG has been studied extensively at laboratory scale, and a number of pilot projects have been completed5,11,15. However, no process or technology has been successfully commercialised to date. That is partly because rare earth co-recovery has historically been less economically attractive than it is today.
The main technical constraints across all recovery pathways and options are mineralogy and chemistry. PG is often resistant to acid attack because it is contained within the fraction of the phosphate ore that remains undissolved during the initial acidulation.
Physical beneficiation is also limited in its ability to recover rare earths: PG particles are often very fine, which can hinder further liberation through grinding or flotation. Any process aggressive enough to liberate rare earths also tends to co-extract radionuclides and other impurities, requiring radioactivity management within the flowsheet. This can have direct implications for capital and operating expenses (CAPEX and OPEX), as well as health, safety and environmental (HSE) management.
Most published studies on REE recovery focus on PG taken directly from the filters. While recovery from stacked PG has also been studied, this is technically more difficult, as the mineral structure of PG can change due to recrystallisation over time15.
Many hydrometallurgical processes have been studied over the past decades. Methods include direct acid leaching, resin-in-pulp (RIP) extraction, bioleaching and solvometallurgy. In parallel, various pretreatment strategies, such as carbonate conversion and hemihydrate/ anhydrite conversion, have been developed to enhance REE recovery efficiency.
The following section focuses on the two most common REE recovery methods for PG – direct acid leaching and RIP – and discusses carbonation pretreatment as a means of improving extraction performance.
Direct acid leaching
Most published work on acid leaching of PG focuses on the use of sulphuric acid, due to both its low cost and availability for most phosphate producers. Prior to leaching, physical pretreatment of the PG feed, including mechanical activation or ultrasonic treatment, is often adopted to increase access to rare earths ahead of the acid contact.
In terms of morphology, a large percentage of the REEs present in PG occurs within the gypsum crystal lattice as an isomorphous substitution for calcium11. This stable, structurally-bound REE is resistant to leaching under high-sulphate, acidic conditions. However, a small percentage of the REEs present also occurs as an amorphous phase adsorbed onto particle surfaces. This is comparatively easy to leach from PG – and therefore the most readily recoverable fraction of the total REE present.
Reported leaching efficiencies for REEs from PG vary widely across the literature, ranging from around 20-90% (Table 1). Removal efficiency is highly sensitive to liquid-to-solid (L/S) ratio, acid concentrations, temperature and residence time.
Higher REE recoveries are generally achieved through a combination of high L/S ratio and extended residence time, as this allows more dissolution of gypsum. However, these requirements increase reagent consumption and equipment sizing, both of which contribute to higher OPEX and CAPEX, respectively.
To overcome the challenges associated with sulphuric acid, there has been ongoing research on alternative acid leaching (Table 2) with hydrochloric acid (HCl), nitric acid (HNO3) and even phosphoric acid (H3PO4).
Sulphuric acid is, however, typically the least expensive of the three acid leaching options, on a cost-per-mole-of-H+ basis. It is also commonly produced on site at phosphoric acid production sites. Its marginal cost, therefore, is usually closer to the internal production cost than to the merchant price.
In comparison, neither hydrochloric nor nitric acid are readily available at phosphate production sites. Their use would therefore require external purchase and the delivery of corrosive acids via pipes or trucks, along with the necessary storage, handling and containment infrastructure.
Phosphoric acid leaching of PG has been investigated less extensively, although the available literature suggests that a high L/S ratio is required due to the chemical stability of the REEs present. The on-site availability of phosphoric acid is advantageous, as is the opportunity for plant integration by recycling.
The large volume of phosphoric acid required for leaching is, however, likely to incur extra capital costs, due to the additional equipment capacity and the larger recycle loop needed to maintain the plant’s phosphoric acid production. Furthermore, the integration of recycled acid would introduce extra (and potentially unwelcome) complexity at an existing wet phosphoric acid plant.
Resin-in-pulp extraction
Resin-in-pulp (RIP) is another commonly investigated technology for rare earth recovery from PG. Both strong cation exchange resins and chelating resins have been tested for REE adsorption using this approach.
In the RIP process, resin is contacted with PG slurry and sulphuric acid in a continuously stirred tank reactor (CSTR). Gypsum has low solubility in acidic slurry and, as it dissolves and recrystallises, a small fraction of REEs continuously enters solution and becomes available for capture.
The resins used in this process are selected to have a high affinity for dissolved REEs with functional groups that preferentially bind to REE ions relative to other dissolved species.
After capture, the REE loaded resin is separated from the barren PG by screening based on particle size. The barren PG solid is filtered while the filtrate and wash solution are recycled to the loading step.
The loaded resin then undergoes elution, typically with sodium chloride (NaCl) or similar alternatives. During elution, sodium ions exchange with REE ions and desorb them into solution. The resin is then regenerated using an acid wash, typically with sulphuric acid, and finally recycled to the loading step.
Studies have shown that resin-in-pulp can achieve REE recoveries from PG of up to 75%4. The equipment required for this process is scalable and not uncommon in other hydrometallurgical industries, such as gold and uranium recovery.
Compared to direct acid leaching, the RIP process has a much lower acid consumption, with potential opportunities for acid recycling. However, carrying out ion exchange at scale generally requires consistent feed conditions to control the cycle. Consequently, pretreatment may be required in a commercial flowsheet – adding to the capital costs of the project.
Carbonation pretreatment
This method involves carbonation with salts such as soda ash (Na2CO3) or ammonium bicarbonate (NH4HCO3) to pretreat the PG ahead of leaching. The carbonation reaction converts calcium sulphate into calcium carbonate. This is advantageous as the REEs can be more easily leached when present in carbonate.
In this method, PG slurry is fed to a stirred tank reactor and contacted with a carbonate salt. The resulting REE-bearing carbonate slurry is filtered, producing a solid cake and a filtrate rich in sodium sulphate or ammonium sulphate. This filtrate stream has potential as a value-added product, especially for the fertilizer market.
The filter cake containing rare earth carbonate is then leached with acid. Leaching efficiency is improved, compared to direct leaching of untreated PG under comparable L/S ratio, temperature and reaction time conditions. One study that converted calcium sulphate to calcium carbonate, through carbonation with sodium carbonate, managed to leach up to 87% of the REEs present using citric acid5.
Despite this, carbonation pretreatment is more reagent-intensive than direct leaching. This is due to the stoichiometric consumption of carbonate reagents, with one mole of reagent being required per mole of PG treated. By contrast, the consumption of sulphuric acid in direct leaching is primarily controlled by REE dissolution and associated gangue reactions.
Also, instead of producing a potentially saleable gypsum by-product, carbonation pretreatment converts PG into calcium carbonate and sodium sulphate, both of which may have limited commercial outlets – depending on purity, logistics and local market demand.
Technical and economic barriers
Increasingly, many critical minerals are being sourced from non-conventional, lower-grade resources in order to meet projected demands, with lithium recovery from clays and brines and cobalt recovery from mine tailings being two examples. Rare earth recovery from PG therefore reflects this broader industry trend. Yet, historically, several techno-economic barriers have limited the development of phosphogypsum as a REE resource.
From a technical standpoint, REEs are difficult to isolate from their PG host due to the underlying chemistry and mineralogy. This constraint stems from gypsum’s low solubility in sulphate-rich acid media and the tendency of rare earths to substitute within the gypsum crystal lattice.
From an economic standpoint, processing these by-products also carries costs that are not always apparent. Carbonation pretreatment, for example, recovers REEs at high efficiency but converts PG into a different by-product that carries its own handling and disposal costs.
Additionally, for these types of midstream recovery processes, existing plant facilities must be modified to accommodate the new process routes. This poses a risk to the recovery and quality of the primary product, especially when extra levels of complexity are added, such as recycle streams.
The time is now
Historically, the lowest-cost option available to producers for handling and storing PG has been gypsum stacks. This default assumption is no longer true, however, with new regulatory and economic drivers encouraging the greater valorisation of PG.
The push to diversify rare earth supply has led to major economic incentives being specifically aimed at lowering the market entry barrier for new participants. Since 2020, government-led funding for rare earth projects has reached the multibillion-dollar range7. This includes investments, loans, promised funds, trade deals, etc.
In 2025, the US Department of Defense took a $400 million equity position in MP Materials, paired with a ten-year offtake agreement guaranteeing a $110/kg price floor for NdPr products7. USA Rare Earth also secured up to $1.6 billion in US government funding, alongside $1.5 billion in private sector investment in early 2026, bringing the total available capital for their integrated mine-to-magnet strategy to roughly $3.1 billion8.
In Australia, Arafura Rare Earths has advanced its Nolans project with a $1.6 billion funding package backed by export credit agencies in the United States, Canada, Germany and South Korea9. Nolans is particularly relevant to this discussion because its ore body contains significant phosphate minerals alongside its rare earth content, enabling the co-production of merchant-grade phosphoric acid (MGA) to offset mining costs.
Guaranteed price-floors and offtake agreements have helped de-risk rare earth recovery projects and protect producer margins from oversupply, price crashes and broader market volatility.
Government R&D funding aimed at establishing rare earth processing outside of China is also specifically relevant to PG. In December 2025, the US Department of Energy’s Office of Critical Minerals and Energy Innovation issued a notice of funding offering up to $134 million of support for demonstration-scale facilities to recover rare earths from unconventional feed streams, including feedstocks from acid mine drainage, mine waste or other deleterious materials10. These types of programmes present significant funding opportunities for research and development into REE recovery from PG.
Competitive advantages for phosphate producers
Global demand for phosphoric acid is increasing from two directions, both of which will ultimately contribute to increasing PG generation: firstly, conventional fertilizer applications and, secondly, lithium iron phosphate (LFP) battery production. This extra demand will ratchet up the pressure on phosphate producers as they face increasing scrutiny over their land use and closure liabilities for gypsum stacks.
More positively, tighter environmental regulation of phosphogypsum may push phosphate producers to pursue rare earth and radionuclide recovery as a way to convert a liability into a cleaner, saleable product. What is potentially beneficial for these projects is the higher proportions of HREEs in phosphate ores, a factor which materially increases the potential revenue from the rare earth co-product.
A large driver of cost in many mining and metallurgy projects is often the cost of mining the resource in the first place. The main advantage with pursuing rare earth extraction from PG, therefore, is that the primary mining and beneficiation are already covered by the existing phosphate production process. This means the incremental capital and operating cost associated with rare earth recovery may be significantly lower than for a greenfield project.
Lastly, and perhaps most importantly, phosphate producers can bring a competitive edge to this market. While the current high influx of investment is enabling junior miners and start-ups to develop rare earth projects, incumbent phosphate producers bring industrial maturity, access to skilled labour, project execution experience and established supply chain networks. For new participants, in contrast, establishing these capabilities can require additional cost and time.
Phosphate producers are therefore well placed to tap into the REE opportunity, and leverage their existing assets, while addressing the growing PG by-product problem within their own operations.
Hatch is here to help
Hatch is uniquely positioned to help new and existing phosphate producers navigate the technical and commercial transition into rare earth co-production. In 2018, Hatch was engaged by Arafura Resources as the lead engineering study manager for the Nolans project and completed the definitive feasibility study (DFS) in 2019. Recently, Hatch was awarded the engineering, procurement and construction management (EPCM) contract for Nolans in Arafura’s final investment decision (FID) published in May 2026.
Hatch brought a similar depth of experience when entering a strategic partnership with USA Rare Earth in 2023 – by supporting the construction of two rare earth separation and magnet production facilities in Texas and Oklahoma. Hatch’s extensive experience across both the phosphate and rare earth sectors enables us to offer clients complete engineering solutions, from prefeasibility through procurement to construction.
References
