Introduction

The demand for rare earth elements (REEs) is increasing due to their vital role in clean energy solutions like hybrid and electric vehicles, and wind turbines [1]. Primary industrial sources of REEs include bastnäsite, monazite, loparite, xenotime, and ion-absorption clay minerals [2, 3]. Recently, attention has turned to extracting REEs from alternative raw materials, including silicate minerals, posing unique processing challenges [3, 4]. Allanite, an epidote-group silicate mineral serving as a host for REEs, is widespread in the Earth’s crust, showing economic potential in regions like Australia, Asia, Turkey, Canada, the United States, and Europe [5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20].

Allanite has the simplified structural formula A2M3Si3O11(O,F)(OH), where A is primarily occupied by Ca and REEs, and M is occupied by Al and Fe [21, 22]. Typically, allanite’s total REE content ranges from 14 to 33 wt%, with light REEs (La to Eu) constituting over 90% [23]. Thorium and U are common A-site constituents, with Th ranging from 2–3 wt% and U in the < 0.1 wt% range [23]. Over geological timescales, allanite undergoes metamictization due to self-radiation damage from incorporated U and Th, resulting in partial or complete transformation into defect-rich crystalline and radiation-amorphized nanoregions [24,25,26,27,28]. Metamictization is a crucial factor in allanite’s processability, as the disordered state increases its susceptibility to chemical alteration and solubility in acids [23, 27, 29, 30].

While allanite hasn’t been economically utilized for REE production, the proposed dissolution and processing schemes mirror those of current REE minerals [3, 4, 10, 31,32,33,34]. Metallurgical tests on allanite dissolution [10, 16, 35,36,37,38] suggest efficient decomposition and high REE recoveries (up to 95%) through sulfuric acid baking at elevated temperatures (200 to 650 °C), followed by extended water leaching (e.g., 24 h) at near-boiling temperature [34, 35, 37]. However, these processes pose significant drawbacks, including high energy and equipment costs and the generation of greenhouse gases and acidifying substances [39, 40]. Although detailed emission values for the initial flowsheets in the processing of allanite are not provided, notable examples of current mines, such as China’s Bayan Obo deposit and Australia’s Mt. Weld deposits—accounting for over half of global REE production—reveal that the acid baking and water leaching stages in their flowsheets contribute to 10–20% of greenhouse gas emissions and 30–40% of acidifying substance emissions in the overall REE supply chain [41, 42]. This paper addresses these challenges by presenting a case study of laboratory-scale test work exploring an alternative low-temperature approach for allanite dissolution in sulfuric acid.

Materials and Methods

Samples

In this study, the allanite concentrate was prepared from samples taken from the allanite-bearing mineralized zone in Otanmäki, central Finland [20]. The concentration, carried out at Oulu Mining School, University of Oulu, involved crushing and screening the material to − 1 mm, followed by combination of hand-picking and magnetic separation, isolating paramagnetic allanite. The resulting concentrate was homogenized, divided for testing, serving as the primary feed (“coarse feed”). To explore particle size impact, specific subsamples were finely ground to create a finer grain size (“fine feed”).

Mineralogical Analyses

Quantitative mineralogical analysis of the feed was made by semi-automated scanning of polished grain mounts at the Centre of Material Analysis, University of Oulu, using a Zeiss ULTRA plus field emission scanning electron microscope (FE-SEM) equipped with an Oxford instruments X-MAX 80 EDX (Energy dispersive X-ray spectroscopy) detector and Oxford Instruments’s INCAMineral automated mineralogy (AM) software [43]. Quantitative X-ray diffraction analysis (QXRD) was done at the University of Oulu by Rigaku SmartLab 9 kW X-ray diffraction system and PDXL2 software for Rietveld quantification analysis.

Chemical Analyses

Elemental analysis (REEs, Al, Ca, Si, Fe, Th) of solid samples and solutions were performed primarily with inductively coupled plasma optical emission spectrometry (ICP-OES) at the University of Oulu by Agilent 5110 VDV ICP-OES instrument and Perkin Elmer AvioTM 500 ICP-OES instrument at the University of Jyväskylä. Prior to analysis, the solid samples were decomposed using lithium borate fusion and dissolution in concentrated nitric acid. Chemical analyses of solutions containing trace levels of REEs, Th and U were carried out with inductively coupled plasma mass spectrometry at the University of Oulu by Agilent 8900 Triple Quad ICP mass spectrometer (ICP-MS). The calculation for metal recovery in the leach solution was determined using the following formula: wt% of the element dissolved or leached/total weight of the element in the feed * 100.

For comparative purposes, elemental analysis of a representative sample of the allanite concentrate and a certified reference material (REE-1) were determined at Bureau Veritas Minerals Ltd (Canada) by ICP-OES/-MS (protocols LF300 and LF100, analytical package LF200), which included fusion of sample powder with lithiumborate and dissolution in nitric acid and total C and S by LECO combustion analysis (method TC000). The results for the quality control samples analyzed by ICP-OES at the University of Oulu indicated mostly good correspondence for all the REEs (for more detail refer to Electronic Supplementary Material, ESM Table S-I). However, slight deviations were observed in the case of praseodymium (Pr). Consequently, the Pr quantification data was excluded from the interpretation of the results.

Leaching Test Procedures

Various parameters were investigated in the experimental setup, including leaching time (ranging from 4 to 0.5 h), temperature variations (including an ice bath, 22 °C, and 50 °C), variations in H2SO4 concentration (3 M, 2 M, 1 M, and 0.4 M), and alterations in feed particle size (both coarse and fine). Initial parametric testing involved 1 g of the feed combined with 20 mL of sulfuric acid. Subsequently, the feed batch size was scaled up to 10 g (with 200 mL H2SO4) and 100 g (with 2000 mL H2SO4), maintaining a constant slurry density of 2% solids. The individual tests followed the same basic procedures which included the addition of H2SO4 solution to a carefully weighted dry concentrate in an Erlenmeyer flask. The mixture was continuously stirred to improve mass transport conditions in the leaching vessel. The parametric optimization work was carried out with at least three parallel experiments in series, to observe the repeatability of the procedures.

Results

Characterization of the Allanite Concentrate

The bulk chemical analysis of the allanite concentrate indicates a total rare earth oxide (TREO) content of 8.1 wt%, with strong enrichment in light rare earth oxides (LREOs) over heavy rare earth oxides (HREOs), accompanied by elevated Th and U contents (Table 1). Allanite-(Ce) is the dominant mineral (~ 60 wt%; Table 2) and the main REE-bearing mineral in the concentrate sample. The average TREO content within allanite-(Ce) is measured at 15.9 wt%, with dominance in LREEs over HREEs and notable enrichment of Ce [20]. Thorium content in allanite-(Ce) typically ranges from 0 to 0.9 wt% (average 0.2 wt%), while U content is consistently below 0.1 wt% [20]. Particle size analysis reveals median (D50) particle sizes of 327 µm and 108 µm for the “coarse” and “fine” feeds, respectively (for further details, refer to ESM Figs. S-1 and S-2).

Table 1 Chemical composition of the feed (allanite concentrate)
Full size table
Table 2 Minerals found in the allanite concentrate together with their structural formulae, densities, and abundances (wt% values) determined by FE-SEM AM and XRD Rietveld analyses
Full size table

X-ray diffraction patterns of the feed exhibit distinct diffraction lines attributed to gangue minerals, with no discernible patterns from allanite-(Ce) (for additional information, refer to ESM Figs. S-3 and S-4), implying its metamict and amorphous nature. To estimate the degree of its metamictization, the quantification of amorphous content in the feed was performed using Rietveld/XRD whole-pattern fitting analysis, employing the internal crystalline standard method, and introducing zincite (ZnO, 100% crystalline) as the reference standard. The results indicate that the feed has high amorphous content (~ 64 wt%; Table 2) that is attributed to the metamict allanite-(Ce).

Leaching Tests

The results and observations of the sulfuric acid leaching tests are illustrated in Figs. 1 and 2 and Table 3. In the following, the discussion is focused on the behavior of REEs and major gangue elements Fe, Al, Th, Ca, and Si in the tested procedures.

Fig. 1
figure 1

The effects of experimental parameters on dissolution recovery of LREEs, HREEs, and major gangue (Ca, Si, Al, Fe, and Th) to the leachate using sulfuric acid leaching. Effects of a H2SO4 concentration, b dissolution temperature, and c dissolution time on REE recovery, and d recovery of main impurity elements in optimal conditions. Error bars indicate standard error

Full size image
Fig. 2
figure 2

Backscattered electron images of leach residues after leaching tests using 100 g coarse feed, 2000 mL of 2 M H2SO4, 3-h leaching time at 22 °C. a Corroded allanite particle alongside uncorroded gangue particles, covered by fine-grained gypsum. b Abundant, fine-grained gypsum formed in the dissolution stage. Adr andradite, Aln allanite-(Ce), Gps gypsum, Kfs K-feldspar, Qtz quartz, and Zrc zircon

Full size image
Table 3 Results from sulfuric acid leaching experiments utilizing 100 g coarse feed, 2000 mL of 2 M H2SO4, 3-h leaching time at 22 °C
Full size table

Effects of Acid Concentration and Leaching Temperature

First tests were carried out with different acid concentrations using 2 h dissolution time at room temperature (22 °C) and sufficient dissolution in terms of REE recovery was obtained with a 2 M H2SO4 solution, and this molarity was used in the rest of the experiments (Fig. 1a). Leaching at room temperature proved to be better in comparison to ice bath and 50 °C conditions (Fig. 1b). The low recoveries obtained at 50 °C could be caused by REE sulfates formed at higher temperatures [31]. In ice bath conditions, the recoveries were low and could be explained by low reaction rates, although REE sulfates should be most soluble at low temperatures [44].

Effect of Leaching Time

In terms of dissolution time, the optimal average leach recovery of approximately 90% for LREEs and approximately 70% for HREEs was achieved in small-scale tests employing 1 g and 10 g feed sizes with a 3-h leaching time (see Fig. 1c). Longer dissolution times appeared to result in a decrease in solubilities of REEs (Fig. 1c), possibly due to reprecipitation of REE as double sulfates [31]. When the amount of feed was increased to 100 g, the REE recoveries obtained with 3-h leaching time decreased slightly (approximately 80% for LREEs and 60% for HREEs, Table 3) with on average 66% of mass of the solids being dissolved. A leachate rich in LREEs is produced (Table 3), and its REE grades and recoveries closely mirror those associated with allanite-(Ce), suggesting a near-selective dissolution of this mineral. A LREE-rich leachate is generated with its REE grades and recoveries strongly reflecting those REE carried by allanite-(Ce) (Table 3), thus likely indicating its near selective dissolution. The major gangue (Ca, Si, Al, Fe, and Th) that mostly originate from allanite-(Ce) are extracted to a minor to moderate extent (Fig. 1d and Table 3).

Effect of Particle Size

The impact of feed particle size on sulfuric acid leaching was investigated with a 4-h leaching time. Intriguingly, a marginally higher recovery of REEs was observed with the coarse feed compared to the fine feed, showing an improvement of 11% for LREE and 8% for HREE. This outcome contrasts with the conventional expectation that particle size reduction would enhance dissolution rates and thereby metal recovery due to increased surface areas available for reaction. However, the potential factors contributing to this unusual behavior were not explored in further detail. It is conceivable that some passivation of the outer surfaces of allanite grains and/or REE reprecipitation effects may occur more frequently in the fine feed due to higher solubility rates [45].

Additional Observations

Based on SEM-BSE imaging of leach residues from tests carried out in optimal conditions at the largest tested feed sample size (100 g coarse feed and 3-h leaching time with 2000 mL of 2 M H2SO4 at 22 °C), only sporadic, strongly corroded particles of allanite-(Ce) are found (Fig. 2a). More refractory REE bearing minerals, such as zircon, show no obvious signs of dissolution (e.g., Fig. 2b) and are likely to account for some of the losses in REE, notably HREEs [20]. Rietveld analysis and SEM-BSE imaging of the solid residue from sulfuric acid leaching indicated that the residue contains abundant gypsum (~ 20 wt%; see Fig. 2a and b and ESM Fig. S-4), which is observed to form during the dissolution stage. The formation of gypsum also prompts the question whether there is a drift in the solution pH versus time, but this was not recorded. It is possible that minor to trace amounts of REEs are incorporated to gypsum structure, although the REE contents of gypsum in the leach residues remained below detection limits in SEM–EDS analyses.

Discussion

This study demonstrates the effective and selective dissolution of allanite from a relatively coarse concentrate sample (particle size distribution D50 = 327 µm) using a mild dilute (2 M) H2SO4 solution at 22 °C. The observed low-temperature dissolution behavior is primarily attributed to the heightened chemical reactivity of metamict allanite-(Ce) within the examined concentrate, allowing for milder chemical conditions for dissolution compared to many other REE-bearing minerals [3, 31]. The achieved overall recoveries, around 80% for light REEs and 60% for heavy REEs, are consistent with reported results from previous tests on other allanite-bearing materials, including acid bake–water leach processes. For instance, a study [46] reported LREE recoveries of 77–80% and HREE recoveries of 55–78% by treating allanite-bearing material with sulfuric acid at 200 °C for 2 h, followed by 24 h of water leaching at 90 °C.

In our study, impurity elements from allanite (e.g., Th, Fe, Al, Si, and Ca) show higher concentrations in leachates, particularly for Fe, Al, and Si, compared to other studies using acid bake–water leach tests on allanite-bearing materials [10, 35, 46]. This increased gangue element concentration in the REE-rich leach solution could complicate downstream processing for REE recovery, especially with the interference of Fe and Al with common extractants [47,48,49], requiring more effort in their removal. The elevated thorium (Th) content (> 2000 mg/L) in the leach solution also poses a challenge, as Th disrupts the separation process. Typically, REE refineries lack the capability or willingness to recover Th, leading to potential health and environmental risks [40, 50]. This highlights the need for hydrometallurgical procedures to eliminate Th from REE-rich sulfuric acid solutions [51, 52] in flowsheets utilizing allanite-based feedstock before advancing to REE purification, whether for safe disposal or potential recovery as a by-product [53, 54].

Despite the challenges caused by the higher dissolution of gangue elements, the results of this study are promising, successfully demonstrating the potential of low-temperature sulfuric acid leaching of allanite. However, the scalability of such a process remains speculative, and additional parametric research is central. One significant parameter to be investigated for upscaling is the acid/solid ratio in the leaching procedure, as it governs the total volume of acidic solutions involved and is likely to impact the complexity of leachate compositions.

Conclusions

  1. (1)

    This paper discusses the potential for economically exploiting the rare earth elements (REEs) in allanite, a common REE-bearing silicate mineral. Previous methods involving sulfuric acid baking and prolonged water leach at high temperatures have shown efficient decomposition of allanite but are economically and environmentally costly. This paper introduces an alternative approach, examining the dissolution behavior of allanite at low temperatures using sulfuric acid.

  2. (2)

    Parametric optimization reveals selective dissolution of allanite from a coarse concentrate sample at 22 °C, with overall REE recoveries of approximately 80% for light REEs and 60% for heavy REEs, being at a similar level to those previously reported in acid baking of allanite. However, impurity elements from allanite, such as Th, Fe, Al, Si, and Ca, show higher concentrations in leachates compared to other studies, potentially complicating downstream processing. Particularly, the elevated Th content in the leach solution poses a challenge for separation processes, emphasizing the need for procedures to eliminate Th before REE purification.

  3. (3)

    The success of the direct, low-temperature leaching is attributed to the high chemical reactivity of metamict allanite-(Ce) in the concentrate, allowing milder dissolution conditions compared to other REE-bearing minerals. However, scalability remains uncertain, and further research, particularly on parameters like acid/solid ratio, is essential for upscaling the process and understanding its potential challenges and implications.